OKAY SO OUR FIRST PRESENTER IS MIKE EDEN WITH CAME BRIDGE UNIVERSITY. >> GOOD MORNING. >> WE WROTE THIS PAPER [INDISCERNIBLE] SOME OF THE MORE FINER TECHNICAL DETAILS, GOT HIS ADDRESS AT THE END. SO I WILL SHARE WITH YOU SOME OF THE EXPERIENCES WE HAD IMPLEMENTING BITS INTO OUR BIT PRODUCTION WORK FLOWS AT CAMBRIDGE. GO OVER A LITTLE BIT OF HISTORY AND HOW WE GOT TO THAT POINT OF NEEDING CHANGE. A LITTLE BIT ABOUT HOW THE CHANGE HAPPENED AND SOME OF THE CHALLENGES WE FACED DURING IMPLEMENTATION AND THE BENEFITS OF MOVING TO BITS. SO WE'VE GOT A FAIRLY BIG HISTORY OF [INDISCERNIBLE] FROM SGML HEADERS FOR ONLINE JOURNAL DELIVERY OF JOURNAL CONTENT THROUGH OUR OWN TYPE SETTING SYSTEM WHICH WAS UPGRADED TO HAVE SEMANTIC MARK UP IN THEIR 3-4 TEXT ETD. AND AROUND ABOUT 2000, IT WAS THE AMBITIOUS GOAL OF CREATING ETD THAT WOULD SERVE BOTH BOOKS AND JOURNALS AND THE CAMML, WAS BORN AND ALONG THAT WE HAD A MORE SPOKE VERSION CALLED CAVEAT AND ALL WAS GOOD. AND WE HAD THIS, IT WAS GREAT. HOWEVER, CAMML, PROVED TO BE QUITE DIFFICULT. AS AN EXAMPLE THIS IS A VERY SIMPLE PARAGRAPH WITH [INDISCERNIBLE] TEXT AND THAT WAS--IT BECAME QUITE DIFFICULT FOR OUR VENDORS AT THE TIME WHO WERE ENCOURAGED TO STOP IMPLEMENTING XML FIRST WORK FLOWS AND WE JUST, WE STRUGGLED FOR ABOUT 2 YEARS WITH CAM ML, DURING WHICH TIME, IT WAS DECIDED THAT ACTUALLY THIS WASN'T THE BEST, THE BEST SCENARIO, SO WE SPLIT UP 2 DTDs, WE HAD A SEPARATE 1 FOR JOURNALS, CJML AND 1 FOR BOOKS, CBML AND THESE LOOSELY FOLLOWED THE NLM DTDs AT THE TIME. IT WAS MUCH EASIER FOR TYPE SETTERS TO START--[INDISCERNIBLE]--SO THESE ARE A LOT MORE FAMILIAR TO PEOPLE TO TALK WITH THEM, LONG USED THERE, MORE FAMILIAR WITH THE NLM. SO AS I SAID, IT WAS MUCH EASIER FOR SUPPLIERS TO INTEGRATE INTO THEIR SYSTEMS AND FROM THAT POINT OF VIEW, IT WAS STILL HOURS, IT WASN'T IN THE [INDISCERNIBLE] DTD, IT WAS CBML AND WE ALWAYS FELT THAT HAVING A PROPRIETARY DTD WAS AN ADVANTAGE. IT GAVE US FREEDOM. IT GAVE US FREEDOM TO DO WHATEVER WE WANTED. IF WE GOT A NEW MODEL OR NEW TYPE OF CONTENT COMING IN, WE CREATE NOW 1S, NEW MODELS WITHIN THE DTD, WE OWNED IT, IT WAS OURS. HOWEVER THAT FREEDOM SORT OF BECAME ABUSED SHALL WE SAY. AND IT LEAD TO FALL DOWN, SLIP DOWN ADDING NEW BITS AND BEEN VERY BESPOKE TO EACH PRODUCT. WE HAD A BROAD TEMPLATE THAT WAS USED AT THE TIME BUT THIS WAS RELATIVELY FLAT. IN AS MUCH AS ANY OF THE STARTERS WERE THERE, THEY WERE UNIQUE, THEY CONTAIN STRUCTURAL INFORMATION, SO IF WE CONSIDER A LIST, WE HAD A BULLET LIST WITHIN EXTRACT, BULLET LIST WITHIN OTHER ENVIRONMENTS, SO IT QUICKLY GOT THIS EVER INCREASING LIST OF STYLES WITHIN THE WORK TEMPLATE. IT BECAME SPRAWLING. AND FREQUENT AD HOC AMENDMENTS DIDN'T REALLY JUST MAKE THINGS EASIER, 1 OF THE PROBLEMS WE FOUND WAS THAT BECAUSE OF THIS EARLY PROCESS WAS OPEN TO A LITTLE BIT OF INTERPRETATION AND IT WAS DIFFICULT TO IMPLEMENT FOR THE TYPE SETTERS, THIS WAS BEFORE WE WERE DOING XML PROOF, XML FIRST, SO THE FIRST TIME WE SAW THE XML WAS AT FINAL DELIVERY. AND WE'RE PICKING UP AREAS WITHIN THE DTD, OR PROBLEMS WITH THE DTD IN MARK UP AT FINAL STAGE, EACH SORT OF LATE IN THE BIT WORK FLOW, TO BE PICKING UP THOSE SORTS OF THINGS, SO THEY WOULD HAVE TO BE RUSHED QUITE QUICK THE NOT ONLY FROM THE DTD, BUT VALIDATION SCRIPT, AND INCIDENTALLY WE SEE CBML, MOVED AWAY AND HAD STARTED USING SCHEMEATRON. ALONG WITH THE PRESSURES TO PRODUCE WORK FLOWS AND SIMULTANEOUS PUBLICATION OF PRINT AND ONLINE DIGITAL CONTENT, THESE CHANGES WERE MADE INCREASINGLY LATE IN THE DAY AND RUSHED AND SOME TRENDS MADE TO VALIDATION MAY HAVE HAD ADVERSE EFFECTS ON OTHER CONTENT COMING THROUGH. SO IT WAS VERY, VERY CLEAR AND OBVIOUSLY ALL OF THIS WAS A REGIONAL [INDISCERNIBLE] AND IT WASN'T THE RIGHT TYPE OF PLACE TO APPLY THAT RESOURCE. SO A REVIEW WAS A WAY FORWARD. AT THAT POINT THE XML THING--WERE AWARE OF MODELS THERE. WE HAD A COLLEAGUE WHO HAD BEEN FEEDING INTO THE BOOKS DISCUSSIONS FOR A BIT. BUT [INDISCERNIBLE] SAID NO, WE NEED TO BE OPEN AND ASK ALL THE RIGHT QUESTIONS. SO, THEY HAD BEEN OUT FOR CONSULTANCY AND WE HAD iTECHS COME IN AND BUILD AND LOOK AT OUR XML AND RECOMMEND AN UNBIASED WAY OF LOOKING FORWARD. SO BITS WAS RECOMMENDED RATHER THAN REWORKING THE PROPRIETARY DTD WE GOT OR ANY OTHER THING. ONE OF THE--POINTS WAS CONSIDERED--WAS XHTML, HOWEVER WE DISCOUNTED THIS ON THE GROUNDS THAT ALL OF OUR AUTOMATED WORK FLOWS WE GOT, WE FELT THAT GETTING THAT IN THE SPECIFICATION VALIDATION WOULD HAVE BEEN AND FOR THE BREDTH OF CONTENT WOULD HAVE BEEN FAR TOO MUCH WORK THAN GETTING A STRUCTURED XML. SO IT WAS DURING THIS TIME, DURING THE REVIEW, IT WASN'T JUST THE XML STANDARD DTD CHANGE, THAT WE NEEDED TO TAKE HOLISTIC APPROACH AND LOOK AT THE WHOLE ECOSYSTEM. HTML OUTPUTS, OBVIOUSLY WE'RE ENCOURAGED TO ASK ALL THE QUESTIONS FOR EVERY POINT LOOKING AT ALL THE NOOKS AND CRANNIES TO SEE WHAT IT WAS WE NEEDED TO DO. EVEN DOWN TO WHO WAS GOING TO DO THIS WORK, DO WE OUTSOURCE, DO WE DO IT INTERNALLY, DO WE GET THE EFFECTS ON PRODUCTS. HOW WERE WE GOING TO MANAGE THAT MIGRATION AND SUPPORT THE EXISTING WORK FLOWS. A BOOK IN PRODUCTION COULD BE QUITE A LONG TIME. IF YOU HAVE IT IN 1 FORMAT, YOU NEED TO SUPPORT THAT GOING THROUGH AND INDEED WE WERE STILL SEEING SOME CBML BOOKS COMING THROUGH LAST YEAR. SO INITIAL DECISIONS MADE AND CONCEPTS, [INDISCERNIBLE] WE HAD 3 MAJOR CONCEPTS THAT CAME OUT, THE FIRST WAS NOT TO SUBSET. AND WE'VE HAD--I MEAN, HEARD YESTERDAY, A LOT ABOUT SUBSETTING, TO SUBSET OR OR DON'T SUBSET, WE DECIDED NOT TOO AND HAVE THE FULL SUITE OF BITS AT OUR DISPOSAL. WE ALSO DECIDED WE HAD TO HAVE THE COPY EDITORIAL [INDISCERNIBLE] 100% TO THE CUP-BITS MODEL AND THIS WOULD BE VALIDATED THROUGH A RICH SET OF RULES FOR SCHEMEATRON. SO WE HAD TO LOOK AT THIS. NOW A COUPLE OF OTHER POINTS MAJOR REASONS FOR TAKING BITS IS THAT WE HAD ALREADY MOVED TO JATS, JOURNALS WORK FLOWS SO THE THOUGHT OF BEING ABLE TO AT SOME POINT OR OTHER ALIGN ON BOTH OF THESE, THIS PROJECT DIDN'T COVER THAT. BUT BEING ABLE TO DO THAT AT SOME POINT IN THE FUTURE WAS CONSIDERED A WORTH WHILE GOAL. IT WAS ALSO THOUGHT THAT BITS BEING A STANDARD ACROSS PUBLISHERS WOULD BENEFIT FROM OTHER PUBLISHER INPUT. WE COULD IF YOU LIKE, RIDE ON THE COAT TAILS OF BEING PART OF A BIGGER PICTURE, RATHER THAN AS WE HAD DONE UP UNTIL THAT POINT BEEN VERY BLANKET IN THIS IS OUR STANDARD AND WE DO WHAT WE WANT AND OPENING UP AND SHARING AND BENEFITING FROM A WIDER--WIDER GROUP. MORE OF THE NLM SUITE OF DTDs IS MORE ALONG THE LINES OF STM AND THE BOOKS HAD QUITE A LARGE PERCENTAGE OF HNS CONTENT. WE DID FEEL THIS EACH STILL SERVED THAT CONTENT, TOO. ONE OF THE EXCEPTIONINGS WHERE WE HAD THE IMPLEMENTATION, WE HAD SOME HIGH DESIGN TEXTBOOKS AND THEY [INDISCERNIBLE] PARKED UP FOR A MOMENT AND THEY STILL STOOD IN THE FORMAT THEY WERE ALREADY IN. SO--APOPTOTIC SO WE HAD A COUPLE OF CHALLENGES ON OUR PLAN AND 1 WAS DOCUMENTATION. WE NEEDED TO INSURE THAT WE HAD DOCUMENTATION THAT WAS FIT FOR DIFFERENT AUDIENCES, WE WERE GOING TO HAVE EDITORS, COPY EDITORS HAVING TO REFER TO THIS INFORMATION. WE HAD TYPE SETTERS AND WE HAD THE DEVELOPMENT TEAM AS WELL. SO RICH DOCUMENTATION FOR DIFFERENT AUDIENCES AND DOCUMENTATION WRITING, AS I'M SURE MOST PEOPLE WOULD AGREE. NEVER A TOP PRIORITY. IT'S ALWAYS QUITE A DIFFICULT THING TO DO. AND THERE WERE OTHER TASKS TO DO. NOT NECESSARILY COMPLEX TASKS, BUT LABORIOUS NONETHELESS. SO BECAUSE WE WEREN'T SUBSETTING, OR EVEN IF WE HAD, WE HAD TO CREATE MODELS OR UNDERSTANDING OF WHAT OUR USE OF THE DTD. SO ALLOW FOR ALL THE HYPHEN TYPE ATTRIBUTES AND VALUES DOTTED ACROSS THE DTD AND HOW THEY WERE GOING TO BE FED BACK INTO THE COPY EDITORIAL PROCESS AND THROUGHOUT THE HTML AND HOW THEY WERE VALIDATED INTO THE [INDISCERNIBLE]. SO THAT PROCESS WENT ROUND. SO WE HAD STANDARD TEMPLATE WHICH HELPED DOCUMENTATION WHICH HELPED INFORM SUPPLIES AND WE WENT THROUGH A CYCLE OF EVENTS OF DEVELOPMENT. SO THERE WERE ALSO CONCEPTUAL HURDLES. NEEDED HELP WITH THE EDITORIAL STAFF IN GAINING NEW UNDERSTANDING OF CONTENT AND HOW--WHAT WAS GOING TO BE REQUIRED OF THE AUTHORS AND COPY EDITORS WITHIN THIS NEW MODEL. AND HOPEFULLY MAKE THIS ANALOGY OR BREAK DOWN WHERE IT WAS CLEAR THAT TECHNICAL STUFF COULD BE CONSIDERED AS LUMPERS IN THEIR CONCEPTS. COMBINING THINGS INTO A SIMILAR OBJECT, ABILITY LISTED, ABILITY TO LIST NO MATTER WHERE IT IS, HOWEVER THAT EDITORIAL WHERE MUCH WERE CONSIDERED SPLITTERS WHERE THEY WOULD SEE ABILITY LISTED, 1 ENVIRONMENT TO BE INHERENTLY DIFFERENT INVOLVED AND HENS THIS IS THE TEMPLATE BEFORE HAND. SO IT WAS A BIT OF A JUMP BUT IT WAS USEFUL TO GO THROUGH BECAUSE IT GAVE AN UNDERSTANDING ACROSS THE BUSINESS FOR THE DIFFERENT TEAMS. SO THE CONCEPT OF START AND END CONTENT MARKERS IN THE COPY EDIT CODES WHOSE SOLE JOB IT WAS TO MARK AN EXTENT OF AN ENVIRONMENT WITHIN THE DOCUMENT WHERE WITHIN THERE GENERIC STYLES, PARAGRAPHS, LISTS COULD BE VALIDATING AND CONTROLLED. SO FOR EXAMPLE HERE, GENERIC TEXT BOX WILL BE MARKED. WE HAVE 2 MARKERS, A BEGINNING AND AN END AND EVERYTHING ELSE IN BETWEEN. THE THAT WOULD THEN FLOW THROUGH INTO THE BITS CONVERTED TO BOX TEXT AND THAT BIT OF INFORMATION WILL FLOW THROUGH TO THE HTML. THAT MEANT WE HAVE LINEAGE CONTENT ALL THE WAY THROUGH FOR THE WORD FILE ALL THE WAY THROUGH TO DIGITAL PRODUCTS. THERE WAS A DESIRE AT THE PRESS TO AGAIN OWN THINGS. WE ALL WANTED TO ALSO OWN AND CREATE AND OWN THE WORD TO BITS TRANSFORM. HOWEVER, FOR THIS PROJECT AND FOR EXPEDIENCY BECAUSE WE WERE GIVEN QUITE TIGHT LEAD LINES, THAT WAS PUT ON THE BACK BURNER AND WE FELT THAT THAT ACTUALLY--THE VENDORS WHO DO THIS GENERALLY ANYWAY, WOULD HAVE TO VALIDATE AGAINST OUR--OUR CURATOR, I THINK AS SASHA SAID YESTERDAY, TRUST AND VALIDATE. SO YES, WE TOOK THE CHANCE THAT WE WOULDN'T NECESSARILY DO THAT WITH THIS IMPLEMENTATION. BUT WE WOULD CONSIDER IT IN THE FUTURE. NOW ON TO THE FUTURE, THIS WAS GOING TO BE A MUCH GREATER SET OF RULES NOT SUBSETTED IN DTD. SO 1 OF THE ASPECTS OF THIS IS WE REFER TO THE SCHEMATRON USING HTTP CORE--SO THAT MEANT THAT WE COULD REFER, THE CONTENT WAS ALWAYS VALID AGAINST ITS VERSION OF THE SCHEMATRON, SOMETHING WE DIDN'T HAVE IN CBML AND WITH ALL THE CHANGES WE HAD IN CBML LATER ON, YOU COULD GO AND GET SOME XML FROM LAST YEAR, AND IT WOULDN'T VALIDATE BECAUSE IT WAS ALWAYS GOING AGAINST THE LATEST BECAUSE THAT WAS QUITE AN IMPORTANT PIECE FOR US. BUT CHALLENGES CHALLENGES WERE IN ORDER WHEN YOU CONSIDER A SIMPLE MODEL LIKE THIS FOR THE TITLE, IT'S EASY TO MAKE THAT RELATIONSHIP WITH THE PARAGRAPH. SO WHEN YOU CONSIDER SOMETHING LIKE THAT TRY TO VALIDATE ORDERING AND OPTIONS WITHIN THAT BECOMES A LOT MORE COMPLEX TASK. SO WE HAD THE USE OF, THIS COULD BE A LARGE SET OF ELEMENTS THAT WERE THERE, JUST MILDLY PRESENT AND THEN WE CONSIDER THE SCHEMATRON SORT OF LIKE A BLACK LIST OF WHAT WE DON'T WANT THERE AND IT'S DIFFICULT TO DISTINGUISH THE DIFFERENCE BETWEEN THOSE TYPES OF AREAS AND VALIDATION AND THE USE OF ROLE THAT I THINK WAS PICKED UP YESTERDAY AS WELL WAS IN INCREDIBLY USEFUL IN THIS PROCESS TO DIFFERENTIATE THE DIFFERENT CATEGORIES OF AREA. SO AS YOU CAN SEE HERE--OOH, I HAVE MY SLIDES OUT OF ORDER--SO WE HAVE HERE EXAMPLES OF SOME SIMPLE ROLE WHERE WE'RE DEFINING DIFFERENT TYPES OF CHAPTER PART CAN'T WITHIN EXIST A BOOK TYPA CHAPTER SO WE'RE STRUCTURING AND VALIDATING A MODEL THE PARTS THAT CAN'T EXIST WITHIN THE CHAPTERS BUT THE OPPOSITE WOULD BE TRUE FOR THE CHAPTERS AND THE PARTS. WE'VE GOT A MORE COMPLEX TYPE OF ROLE WHERE THIS IS REFERRING TO STATIC EXTERNAL FILE THROUGH THE VARIABLE REQUIRED ATTRIBUTES WHERE WE'RE DEFINING ALL THE ELEMENTS THAT HAVE ATTRIBUTES THAT ARE REQUIRED, THAT ARE REQUIRED WITHIN THE DTD, SO THIS MIGHT HAVE BEEN SOMETHING WE WOULD HAVE USED IF WE ALTERED THE DTD, WHERE THIS ATTRIBUTE HAS TO DEFINITELY BEEN THERE, WHERE WE WOULD CHANGE THE DTD TO BE A REQUIRED ATTRIBUTE. WE DON'T DO THAT. WE CAN ENFORCE THAT HERE USING SCHEMATRON. AND THEN HERE IS THAT PLACE OF MODELING BOOK TYPE TYPE OF ROLE WHERE WE HAVE A--FOR THE CONTEXT PARAGRAPH OF JUDGES WHICH IS A JOURNAL WHICH IS A LEGAL PUBLICATION, DESCENDANTS IN THERE MUST HAVE AT LEAST 1 NAMED CONTENT OF A CONTENT TYPE JUDGE. SO THERE MUST BE A JUDGE CORRECTLY CAPTURED WITHIN ANY PARAGRAPH OF JUDGES WITHIN AN INR BOOK. AND THAT WOULD HAVE BEEN SOMETHING THAT INCIDENTALLY IN CBML WE WOULDED JUST CREATE AN ELEMENT AND HAD THAT THERE, IT WOULD HAVE BEEN A LOT MORE SIMPLELE BUT THIS IS THE TRADE OFF. AND MOVING TO AN EXTERNAL STANDARD. SO THE USE OF ROLE. SO WE LOOK AT THE USE OF STRUCTURES WHETHER SOMETHING IS REQUIRED OR DISALLOWED AND THAT WOULD BE CONSIDERED FATAL AND THOSE SPECIFIC STRUCTURES WITHIN REQUIRED AND MUST FOLLOW A CERTAIN CONFIGURURATION AND WE WOULD ERROR THAT. FOR EXAMPLE, WHEN ATTRIBUTE X HAS ELEMENT 1, HAS VALUE ZED, THEN ELEMENT 1 MUST HAVE NO CHILDREN. SO IT PROVIDES THAT DISTINCTION. AND THEN WE HAVE THE 1 FOR--TOP GRAPHICAL TO HAVE 1 WHERE WE MAY HAVE A LOOK AT CONTENT. AND THEN THE INFORMATION ABOUT THE BOOK WE CAN PROVIDE INFO FOR EXAMPLE. THIS BOOK HAS 12 CHAPTERS USUALLY AT THE ROOT LEVEL. AND THOSE WE'RE EXPLAINING THAT ALL THE COP TENT MUST VALIDATE OR GO INTO THE DTD AND FOR THE SCHEMATRON. SO, WHERE ARE WE NOW? WELL, WE'RE FINALLY OUT OF THE PROJECT AND INTO BUSINESS AS USUAL. IT'S BEEN A BIT OF A SLUG, WE'VE STILL GOT SOMETIME TO GO. WE EXPECT THINGS TO BED DOWN OVER THE NEXT YEAR, PROBABLY ANOTHER YEAR, BEDDING IN, REFINEMENT AND HOPEFULLY HAVE SEEN MOST PRODUCTS. WHAT NEXT. WE'RE GOING TO REVISIT THE WORD TO BITS TRANSFORMS. AND WE WILL LOOK AT THOSE HEAVY DESIGN BOOKS AND SEE IF WE CAN BRING THOSE BACK IN TO BITS. SO THE CHALLENGES THAT WE FACED, SO PROJECT WORK AND THE ACTUAL DOING THE PROJECT WHILST MAINTAINING THE CURRENT STANDARD, DEFINING THE NEW 1, ALL THE WORK WE NEEDED TO DO, RESOURCES THIS PARTICULAR ISSUE, WE WERE GIVEN A FAIRLY AGGRESSIVE SCHEDULE OF ABOUT 5 MONTHS AND IT TOOK ABOUT 9 MONTHS FROM THE GREEN LIGHT TO FIRST BEING PUT THROUGH BITS MODEL. IT WAS A MUCH NEEDED REVIEW OF CONTENT CAPTURE. AND ENDED UP WITH A LOOK AT WHOLE WORK FLOW. WHOLISTIC, GOOD TO DO THAT AND MAKING SURE WE HAD A GOOD SYSTEM TO GO FORWARD WITH. WELL, WE HAVE A STANDARD THAT'S I THINK GOING TO BECOME IF HE'S NOT ALREADY THE DE FACTO STANDARD FOR BOOKS AND DATA CAPTURE FOR ACADEMIC WORK. IT WAS A STRONG CASE FOR IT, WE HAVE JOURNALS NOW. JATS IS THE STANDARD NOW AND IT'S USEFUL AS I SAID BEFORE, THE IDEA OF IT BEING A SHARED STANDARD SOMETHING THAT HAS INPUT FROM THESE OTHER SOURCES. AND THE UNEXPECTED BENEFITS WERE WHAT WE GOT FROM THAT SHARE AND COMMON UNDERSTANDING BETWEEN DIFFERENT FRACTIONS OF THE PRESS. PEOPLE WHO HAVE NORMALLY BEEN INVOLVED IN CONSIDERING MODELING AND MARK UP HAVE NOW GOT A GREATER UNDERSTANDING OF THEM. WE HAVE TO START UP CREATE IN THOSE PARTS AND EDITORIAL IS VERY OFTEN DISTANT FROM MARK UP. WE'VE ALSO GOT A STANDARD THAT'S GOING TO BE EASIER FOR DISTRIBUTIONS. WE HAVE SOURCES LIKE SCHOLARS PORTAL NOW, IMPLEMENTING BITS AS THEIR STANDARD. THAT MEANS IT WILL BE EASIER FOR US TO DELIVER CONTENT TO THEM AND THE REUSE OF THE CONTENT BECOMES A LOT EASIER. THIS IS A STANDARD VOCABULARY SO--OOH. SO WE FEEL OUR CONTENT IS WELL SERVED BY BITS AND IT'S BEEN A DIFFICULT BUT WELL-WORTH WHILE PROCESS TO GO THROUGH. AND WE DO FEEL OUR CONTENT IS WELL SERVED BY THIS STANDARD. I'D LIKE TO LEAVE JUST 1 LAST THOUGHT AND THIS IS SLIGHTLY OFF PAYS BUT THERE'S BEEN A LOT OF TALK ABOUT [INDISCERNIBLE] SCHEMATRON. AND I KNOW TALKING THROUGH THIS WITH VARIOUS TEAMS AND VARIOUS OTHER NONTECHNICAL STAFF, THE WORD SCHEMATRON BRINGS UP AND CONJURES UP AUTOBOTS AND DECEPTY CONS AND OTHER STRANGE TRANSFORMING CARS INTO PEOPLE'S HEAD AND BEFORE THIS PRESENTATION MY SON ASKED ME WHAT SCHEMATRON IS. AND I ASKED HIM WHAT DO YOU THINK SCHEMATRON IS, AND HE CAME UP WITH THIS. WHICH IS PRETTY HID AND I SAID WHAT IS IT? HE SAID IT'S A HELMET THAT WOULD HELP YOU COME UP WITH SCHEMES AND IT WOULD MOSTLY BE USED BY VILLAINS. [LAUGHTER] , I DON'T THINK THERE'S MANY VILL ANNS IN THE AUDIENCE BUT FOR US, IT'S VITALLY IMPORTANT AND IT GIVES US A MEANS FOR RESTRICTING MODELS AND STRUCTURES FOR ELEMENTS AND ATTRIBUTES EFFECTIVELY DOING THE SAME JOB AS SUBSETTING. AND WE FELT LIKE IT'S SOMETHING WE CAN'T DO WITHOUT AND IT GIVES US THE ABILITY TO FLEX NOW AND USE OTHER CHUNKS OF THE FULL SUITE OF BITS, A LOT EASIER THAN IF WE HAD BEEN MAKING DTD CHANGES EVERY 6 MONTHS SO IT'S INCREDIBLY IMPORTANT TO US. AND I'M SURE THERE'S A LOT OF PEOPLE IN THE PUBLISHING INDUSTRY WHO PREFER THE HELMET. SO ON THAT NOTE, I'LL THANK YOU AND ANY QUESTIONS. >> DEBBIE LAPPAR, MODEL CITIZEN LIABLE TECHNOLOGIES, MY FIRST QUESTION WAS A THANK YOU. CAMBRIDGE WAS THE MOST ACTIVE PARTICIPANT IN THE BITS GROUP AND THEY KEPT TRYING THINGS AND BRINGING US STUFF AND ANYTHING WHY DOESN'T IT DO THIS AND WHY DOESN'T DO THIS AND WHY DOESN'T IT DO THIS AND WE WOULD GO THAT'S EYE GOOD IDEA. THANK YOU YOU WERE INCREDIBLE LOAMACYY USEFUL AND HELPFUL AS COMMITTEE MEMBERS AND MY OTHER THING IS I'M SORRY ALSO A COMMENT. I WAS VERY IMPRESSED BY YOUR USE OF CONTENT TYPE. THAT'S EXACTLY WHAT THE ATTRIBUTE IS SUPPOSED TO DO. IT'S FOR INFORMATION CLASSES SO YOU THE PUBLISHER CAN CAPTURE WHAT YOU NEED TO AND DO INFORMATION CHECKS THROUGH SOMETHING LIKE SCHEMATRON. THAT'S WHAT IT'S DECIPED FOR AND NOBODY USES IT. THANK YOU. >> YES. >> HI, CAROLINE WEBER FROM ARIES, SYSTEM, CAN YOU TALK ABOUT RUNNING THROUGH A CONTENT DOC. >> WE CAPTAIN REALLY TALK ABOUT THAT WE LEFT IT--CAN'T REALLY TALK ABOUT THAT. WE LEFT THAT FOR OUR VENDORS. THEY'VE BEEN DOING THAT WITH OUR JOURNALS AND VARIOUS OTHER WORK FLOWS BUT THEY WILL TAKE THE WORD AND TRANSFORM IT. WE WILL WANT TO DESIRE WE STILL HAVE A DESIRE OF OWNING THINGS AND AT SOME POINT IN THE FUTURE ONCE THIS SETTLES DOWN, WE WILL WANT TO CONSIDER OWNING THAT TRANSFORM AS WELL. >> HI. >> THANKS, THAT WAS VERY INTERESTING. I WILL HAVE TO THINK ABOUT A NUMBER OF THINGS YOU SAID. ONE OF THE THINGS I THINK IS INTERESTING WITH THE STRATEGY OF PUTTING SCHEMATRON SO CENTRAL IS THE QUESTION OF MANAGING A RULE SET. AND HOW IS IT THAT YOU SPECIFY WHAT YOU'RE GOING TO DO, HOW IS IT THAT YOU TEST THEM, PUBLISH THEM, ALL THAT. WHAT CAN YOU TELL US. >> OKAY, AS I MENTIONED TOM IS AN EXCELLENT ARCHITECT AND I THINK HE WOULD WOULD HAVE BEEN MUCH BETTER PLACED. IT ISN'T EASY. AND WE THOUGHT THAT WASN'T EASY AND ACTUALLY CHANGING A DTD ON AN ONGOING BASIS WAS MUCH MORE DISRUPTIVE TO THE PROCESS AND WHILE THE SCHEMATRON CAN BECOME MORE COMPLEX, I KNOW WHERE YOU'RE HEADING WITH THAT, IT IS A HEADACHE. BUT HAVING TO CONSIDER THAT WHOLE WORK FLOW AND PLEASE CONSIDER THERE'S A LOT MORE COMMUNICATION WITH THIS EDITORIAL AND THIS CONTENT AND SO THE REAL EXAMPLE MADE AND ASSUMED THAT WE HAVE THIS, SO IT CAN'T BE SOMETHING ELSE. THAT THERE HAS TO BE--IF YOU TALK TO EDITORIAL, THAT IS LIKE THAT, OH BUT IT CAN ALSO EXIST HERE OR IT DOES THIS, SO THERE'S A LOT MORE COMMUNICATION GOING ON AND--IT'S A LITTLE BIT MORE LIKE THE INTEND TO DO THAT. YES, YOU'RE RIGHT. >> WELL, IN MY EXPERIENCE, THERE ARE 2 THINGS TO KEEP IN MIND WHEN YOU START ON THIS. NUMBER 1 IS GETTING SOMETHING SIMPLE THAT'S NOT ACTUALLY--IT DOESN'T INFLICT, DAMAGE, RIGHT? I MEAN, NUMBER 1 IS THAT YOU DON'T WANT YOUR SCHEMATRON TO LIE TO YOU, RIGHT? AND THAT'S ACTUALLY A CONCERN. DEVELOPERS ARE PERFECTLY WILLING TO DEVELOP POLICY IF YOU LET THEM OR IF THEY DON'T KNOW THEY'RE DOING THAT. AND THAT'S HOW YOU END UP IN THAT SITUATION, RIGHT. SO YOU NEED SOMETHING SIMPLE BUT YOU NEED TO EXERT THAT LEVEL OF CONTROL FROM THE VERY GET GO THAT YOU'RE GOING TO NEED WHEN YOU SCALE UP IN COMPLEXITY, WHICH MEANS THOSE COMMUNICATION CHANNELS YOU JUST ALLUDED TO ARE REALLY THE CRITICAL POINT, RIGHT? IT'S ACTUALLY OPENING UP THOSE--THE FEEDBACK LOOPS WHEREBY YOUR EDITORIAL OR DEVELOPERS AND SO FORTH CAN COMMUNICATE ABOUT WHAT THESE NEEDS ACTUALLY ARE THAT YOU AVOID THAT SORT OF BLIND SIDEDNESS. SO, YOU KNOW IT'S EASY TO ENGINEER EARLY ON AND THAT ALSO IS TRUE OF THESE COMMUNICATION CHANNELS AND THE SORT OF SUPPLEMENT AT INFRASTRUCTURE WE BUILT FOR EXAMPLE, INLINE DOCUMENTATION FOR THE SCHEMATRON THAT YOU HAVE, YOU'VE TAKEN THE TROUBLE OF EMBEDDING THE SCHEMATRON PRESUMABLY THAT WILL BE USEFUL, I MEAN THAT'S WHY IT'S THERE, AND YET, IT'S ANOTHER THING THAT NEEDS TO BE MANAGED AND EXPOSED. RIGHT? SO YOU DON'T WANT THEM TO MISREPRESENT THINGS EITHER AND AS YOU KNOW, DEVELOPER IS PERFECTLY ABLE TO GOING FROM AND SWITCHING UP THE RULE SO AND SWITCHING UP THE DOCUMENTATION SO THEY'RE OUT OF SYNCH. SO THE NEW FUNCTIONALITY THAT YOU INTRODUCE, YOU ARE INTRODUCING SOMETHING THAT NEEDS TO BE MANAGED. >> YES, IT DOES HAVE THAT REASON THAT YOU SAID, WITH THE USE OF SUBVERSION AND TESTING THROUGH XMLT, AND OUR OUTPUTS [INDISCERNIBLE] WE CAN TRY TO MANAGE THAT BECAUSE IT'S NOT LIKE--IT'S A WORTH WHILE GOAL. >> SO HAVE YOU ENDED UP CLUSTERING YOUR SCHEMATRON RULES BY DISCIPLINE OR BOOK TYPE OR THINGS THAT ARE TRUE FOR ALL-- >> NO, I DON'T THINK SO. WE HAVE A SET OF--AS YOU ALLUDED TO, DIFFERENT SETS, WE HAVE THE SORT OF COMMON BITS AND PIECES. I WILL HAVE TO LEAVE--I KNOW TOM WILL BE QUITE HAPPY TO ANSWER QUESTIONS ON THINGS SO --I WILL HAVE TO LEAVE THAT BUT YES, THEY ARE COMPILED AT THE POINT THAT WE PROCESS THOSE. >> HOW DO YOU FACILITATE GETTING THE EDITORS TO TELL YOU WHAT THEY WANT TO THE SCHEMATRON TO CHECK. YOU SAID IT INCREASED THE COMMUNICATION WHICH IS GOOD AND THE AMOUNT OF DISCUSSION BUT HOW DO YOU SOLICIT THE REQUIREMENTS FROM THE EDITORS. >> AT THE POINT THAT THERE'S A NEW STRUCTURE OR NEED, THAT CAN STILL COME UP EITHER--EITHER NOW HOPEFULLY A LITTLE BIT EARLY ON, SO COPY EDITORS, WE HAVE QUITE A COMPLEX, THE TEMPLATE WE HAVE IS QUITE COMPLEX BUT IT'S EXTENSIBLE BECAUSE IF THE STRUCTURAL INCREASE WE HAVE THERE, HOPE LOOFULY WE'RE NOT PICKING UP STUFF AT THAT POINT, THISSEN AT THAT POINT WE HAVE EARLY COMMUNICATION, WE HAVE THIS NEW THING SO THEY'LL SIT WITH THE DEVELOPER AND THE XML GUYS AND GO THROUGH IT LIKE THAT. AS IN WHEN. >> [INDISCERNIBLE] >> IT WAS INTERESTING YOU MENTIONED THAT USING JATS FOR JOURNALS AND BITS FOR BOOKS, HAVE YOU REALIZED BENEFITS FROM USING JATS FOR JOURNALS AND BITS FOR BOOKS? >> NOT COMPLETELY YET. NO. , BECAUSE OF THE WAY--WE'RE SORT OF LEAP FROGGING, THE BUSINESS WAS 2 SEPARATE ENTITIES REALLY WORKING ON AND THEN NOW A SINGLE ACADEMIC GROUP AND 1 OF THINGS WE WOULD LIKE TO CONSIDER AFTER THIS IS TO ALIGN THEM PROPERLY AND CONSIDER--JATS IS SUBSETTED FOR EXAMPLE. WE WOULD LIKE TO PROBABLY CONSIDER THE SAME MODEL THAT WE'VE DONE WITH BITS NOW IN TRYING TO ALIGN THEM SO THEY'RE NOT LEAP FROGGING THAT WE HAVE THIS SINGLE SORT OF STANDARD. [APPLAUSE ] >> SO IF YOU VILLAINS HAVE ANYMORE QUESTIONS FOR MIKE, YOU CAN GET HIM AT THE BREAK. HEY, WE'RE MISSING SOMEBODY. >> OKAY, BOTH OUR AUTHORS OF THE NEXT PRESENTATION ARE HERE TODAY. BRUCE ROSENPLUM, WHO I THINK YOU ALL KNOW SEEMS TO BE INVOLVED IN EVERYTHING. HE'S WITH THE FTS GROUP WHICH WE HEARD ABOUT YESTERDAY. HE'S ON THE BITS GROUP WHICH WE HEARD ABOUT THIS MORNING, AND HE'S ALSO 1 OF THE ORIGINAL AUTHORS OF THE NLM BTBs AND IS MEMBER OF GOOD STANDING OF THE JATS STANDING COMMITTEE. GOOD STANDING. CAITLINGEBHARDHAS BEEN WORKING FOR 2 YEARS WITH BRUCE, AT INERA, SHE'S WRITTEN A WONDERFULLA PAPER. I HOPE YOU ALL GO AND READ IT SO PLEASE, WELCOME. >> THANK YOU VERY MUCH. I HAVE A FEW ADMISSIONS TO MAKE BEFORE I START GIVING THIS TALK. THE FIRST IS I'VE NEVER GIVEN A TALK BEFORE THAT'S NOT MY OWN WORK. THIS IS CAITLIN'S WORK, SHE DID A TREMENDOUS JOB AND I'M HAPPY TO PRESENT THE PAPER. THE SECOND IS THAT I'M NOT A MATHMETICIAN, ON THAT SCORE, I AM TOTALLY UNQUALIFIED TO TALK ABOUT MATH, THE LAST MATH COURSE I HAD WAS CALCULUS IN HIGH SCHOOL, ANY ERRORS IN MATH, PLEASE FORGIVE ME, I WAS AN ENGLISH MAJOR. WHAT WE'RE TALKING ABOUT TODAY IS ALL THE OF THE DIFFERENT WAYS THAT AUTHORS CAN PUT MATH INTO WORD AND ALL OF THE WAYS YOU HAVE TO GET IT OUT IN A JATS WORK FLOW? WHY ARE WE DOING THIS? WELL FROM A BIG PICTURE PERSPECTIVE BECAUSE MATH SHOULD IDEALLY BE IN A FORM THAT'S DISCOVERABLE, ACCESSIBLE AND SUSTAINABLE IN TERMS OF PUBLICATION ENVIRONMENT. AT A MORE MICROLEVEL, THE FACT THAT YOU HAVE TO SUPPORT MULTIPLE MODELS IN MATH FOR THIS, WE TALKED ABOUT THIS YESTERDAY. YOU CAN HAVE PLAIN ASY TEXT, YOU CAN HAVE GRAPHICS FONT CHANGES, YOU CAN HAVE MATH ML, SO 4 DIFFERENT WAYS OF HAVING MATH. IN ADDITION THERE ARE MULTIPLE WAYS YOU CAN HAVE MATH EMBEDDED IN MICROSOFT WORD. AND JUST A QUICK SHOW OF HANDS. HOW MANY OF YOU THINK THERE ARE 3 WAYS OF INSERTS MATH INTO MICROSOFT WORD? HOW MANY OF YOU THINK 4? HOW MANY OF YOU THINK 5? FEWER HANDS EACH TIME? HOW MANY OF YOU THINK 6? NO HANDS WENT UP--1? OKAY. THERE ARE 6 WAYS OF INSERTS MATH INTO WORD. THAT'S WHY THIS PAPER EXISTS IS BECAUSE WE WERE FINDING OURSELVES SPENDING MORE AND MORE TIME TALKING WITH THE PUBLISHERS WE WORK WITH ABOUT WHAT ARE ALL THE DIFFERENT WAYS CAN YOU ENCOUNTER MATH IN A WORD FILE AND THEN HOW DO YOU WRANGLE THAT INTO JATS. A SIDE NOTE, KAVA IF YOU ARE LISTENING THIS NOTE IS FOR YOU SINCE YOU'RE NOT HERE THIS YEAR. THIS PAPER IS NOT ABOUT TECH. TEX AND LATEX ARE GREAT REQUIREMENTS ESPECIALLY IN PHYSICS, THIS PAPER IS NOT ABOUT TEX, THIS PAPER IS ABOUT PEOPLE WHO PUT MATHEMATICS IN MICROSOFT WORD AS MUCH AS THEY SEEM MATH TO THE MATHEMATICIANS. BUT TO LOOK AT THIS WE NEED TO GO BACK AND LOOK AT A HISTORICAL LOOK AT THIS. THE THIS IS SAY MANUSCRIPT PROVIDED BY ALBERT EINSTEIN TO A PUBLISHER, WHAT WOULD AUTHORS DO IF THEY HAD A TYPE WRITER, THEY WOULD ACTUALLY TYPE OUT THEIR TEXT, OTHERWISE THEY WOULD WRITE IT LONG HAND AND THE EQUATIONS WOULD BE WRITTEN IN LONG HAND. THIS IS HOW MATHEMATICS WOULD BE COMMUNICATED. TYPE SETTING OF ALL OF THAT WAS ENORMOUSLY PAIN TAKING PROCESS WITH A HOD METTLE AND TRIED TO MAKE IT LESS PAINFUL, BUT NO MATTER HOW YOU SLICED IT, IT WAS PAINFUL. PHOTOCOMPOSITION WHICH WAS INVENT INDEED THE 40S AND COMMERCIALIZED IN THE 1950S SORT OF HAD THIS TASTE GREAT LESS FILLING. OVERALL TYPE SETTING BECAME LESS PAINFUL IN THE PHOTOCOMPOSITION ERA THAN THE HOT METAL ERA, BUT NOT FOR MATH WHICH IS THE LESS FILLING PART. AND IN FACT, FAMOUSLY DONALD KNUTH'S GOT THE PAGE BACK IN THE 1970S LOOKED AT IT AND SAID THE TYPE SETTING IS HORRIBLE, THEY HAVE DESTROYED MY OFF AND WENT OFF AND INVENTED TEX AS A REABS TO TYPE SETTING AND AND HE EXTENDED THAT IN 1985. FUNDAMENT WHAT TEX, X AND L ATEX, THEY'RE TYPE SETTING PROGRAMMING LANGUAGES, POWERFUL, AND COMPLEX. THEY HAPPEN TO BE THE 50 CHOICE FOR AUTHORING FOR MOST MATHEMATICIANS AND THE AMERICAN MATHEMATICAL SOCIETY WILL NOT AUTHOR A PAPER UNLESS IT'S IN TEX. IT'S A FIRST CHOICE FOR PHYSICIST AND IN A FEW OTHER SISPLINGS, SOME PEOPLE WOULD USE A PARTICULARLY ECONOMICS AND COMPUTER SCIENCE. BUT IT'S NOT THE TOOL OF CHOICE OF THE REST OF ACADEMIA. NOW IF YOU HAVE A CASUAL CONVERSATION WITH A PHYSICIST, THEY WILL EXPRESS THAT THE MEDICAL EXAMINERS DON'T DO EVERYTHING IN THIS LATEX. OUTSIDE OF THE PEOPLE USING TEX AND LATEX, WHICH IS ALL THE REST OF THE WORLD AND PROBABLY ABOUT 90-KNIFE% OF THE PAPERS THAT ARE AUTHORED WHAT WAS HAVE IS THIS TRANSFORMATION FROM THE MID80S WHERE PEOPLE WERE LARGELY USING TYPE WRITERS TO AROUND 2000 WHEN PEOPLE WOULD PRETTY MUCH SHIFT TO USING COMPUTERS AND THIS WAS DRIVEN BY A COMBINATION FACTORS, LOWER PC COST AND INCREASING PC POWER, A HUGE FACTOR WAS EASY TO LEARN AND EASY TO USE GRAPHICAL USER INTERFACES, AND THE RISE OF EASY TO USE COMMERCIAL WORD PROCESSING APPLICATIONS MOST NOTABLY WORD, PERFECT, MICROSOFT WORD AND THE OTHER FACTOR IN THIS WAS THERE BECAME BOTH FORMAL AND INFORMAL IT SUPPORT FOR ORGANIZATIONS WHERE SCHOLARS WERE WORKING. I GAVE A TALK A COUPLE WEEKS AGO AND I MENTIONED A ORGANIZATION NOW DEFUNCT COWLAD THE BOSTON COMPUTER SOCIETY WHICH SPRANG UP TO HELP PEOPLE WHO WERE PRETTY GEEKY AT THAT POINT FIGURE OUT HOW TO USE THEIR COMPUTERS THAT ORGANIZATION DISSOLVED BY THE 1990S BECAUSE IT WASN'T NEEDED ANYMORE. COMPUTERS BECAME SIMPLE ENOUGH THAT PEER MORTALS COULD USE THEM. ULTIMATELY BY 2000 WHAT WE HAD WAS THE DOMINANCE OF MICROSOFT WORD AND THE MAJORITY OF PAPERS BEING SUBMITTED TO PUBLISHERS THAT WERE NOT IN LATEX, OR TEX, AND THE MAJORITY OF PAPERS WERE COME NOTHING MICROSOFT WORD FORMAT AND THIS WAS DRIVEN BY FACTORS LIKE WORD HAD SUPPORT FOR THINGS LIKE ACADEMIC AUTHORS THEY NEEDED, SPECIAL CHARACTERS, TABLES ACCIDENT IMAGES, THEY HAD SUPPORT WHICH IS THE REASON FOR THIS PAPER. IT ALSO HAD COOL CITATION AND REFERENCE MANAGEMENT AND HERE'S THE BOTTOM LINE, IT CAME TO DOMINANCE BECAUSE MICROSOFT FIGURED OUT HOW TO BUNDLE A BUNCH OF APPLICATIONS INCLUDING WORD AND EXCEL INTO A LOW PRICE PACKAGE AND THAT NEEDED A COMPELLING REASON TO PURCHASE IT, THERE WERE THOSE THAT SAY THERE WERE THE ONLY REASON TO PURCHASE IT BECAUSE THEY DON'T LIKE WORD. ONE MORE BIT OF HISTORY BEFORE WE DIVE BACK INTO WORD. THAT'S SHORT HISTORY ON MATH MARK UP IN TERMS OF ACTUALLY PUBLISHING CONTENT. OF COURSE WE HAD TECH AND LATEX, AND OF COURSE THERE WAS SGML, AND THE IN THE DAYS OF SGML, THE AMERICAN ASSOCIATION OF PUBLISHERS ACTUALLY GREATED A MATH MARK UP MODEL WHICH SGML AND THAT WAS ADOPTED WITH VARIATION, I SEE ACKNOWLEDGMENT FROM THE AUTHORSHIP OF THE MODEL IN THE FRONT ROW, I THINK. THAT WAS ADOPTED WITH VARIATIONS INTO ISO12083. ALL OF THIS FOCUSED ON THE PRESENTATION OF MATH, NOT THE STRUCTURE OF MATH. SO INLET MID90S A WORKING GROUP WAS FORMED AND ULTIMATELY MATH ML CAME OUT OF THIS AS W3 C RECOMMENDATION AND IT DESIGNED TO SOLVE THE PRESENTATION PROBLEM AND GO BEYOND THAT AND CELEBRATE RE-USE OF MATH AND XFL PLATFORMS. SO IT HAS 2 MODELS, PRESENTATION XML WHICH IS CONCERNED WITH THE FORMATTING OF MATH AND CONTENT MATH ML WHICH IS MORE ABOUT THE MEANING AND SEMANTICS OF THE MATH. LITTLE SURPRISED THAT MATH ML HAS NOT BEEN AS SUCCESSFUL AS--HOWEVER MATH ML BECAUSE IT WAS A W3 C RECOMMENDATION AND HAD MORE POWER THAN THE OLD AAP MATH FRAGMENT WAS QUICKLY ADOPTED AND BY 2001 BLACKWELL HAD ADOPTED IT INTO THEIR DTD VERSION 4. AND LSEVERE MADE THE AOPTION AND ON THE HEELS OF THIS IN 2003 WHEN THE NLM DTD1 CAME OUT, MATH ML WAS PART OF THAT. AND IN FACT IN THE EARLY DISCUSSIONS WE HAD IN IF 2002 ABOUT HOW TO CREATE THE NLM DTD, 1 OF THE MAIN REASONS IT WAS MODULAR WAS BECAUSE IT HAD TO ACCOMMODATE MULTIPLE MATH MODELS, MATH ML, TEXT, AND BECAUSE WE FIGURED THAT THERE WAS SUCH WIDE USE OF THE OLD AAP AND 1208 MODEL. WE THOUGHT WE WERE GOING TO HAVE TO INCORPORATE INTO THE NLM DTD, THIS WAS MAKING IT MODULAR, AND IT TURNED OUT THINGS MOVED QUICKLY AND WE TO TO HAD BRING THAT MODEL IN. SO MATH ML ULTIMATELY BECAME IN THE FIRST DECADE OF THIS CENTURY, THE PREFERRED XML, EMPHASIS ON XML HERE REPRESENTATION FOR MATH IN TAG CONTENT. AND THERE ARE BENEFITS TO MATH ML. IT DOES MAKE MATH MORE DISCOERABLE, DOES MAKE IT MORE ACCESS CLINICAL COMMUNITY AND YOUR ARCHIVE IS MORE SUSTAINABLE THAN HAVING YOUR MATH PRESENTED AS GRAPHICS. NOW WE GET TO THE HEART OF THE PAPER. WHEN I SAID THERE WERE 6 WAYS OF GETTING MATH INTO WORD BEFORE, I MEANT IT. YOU CAN ENTER MATH AS TEXT, GRAPHICAL IMAGES, WORD FORMULA FIELDS, HOW MANY KNOW ABOUT WORD FORMULA FIELDS? ONLY A FEW HANDS. YEAH, WE'LL GET TO THAT. DESIGN SCIENCE MATH TYPE. MICROSOFT EQUATION EDITOR SPEED AUTOAND WHAT MICROSOFT CALLS THE EQUATION BUILDER. AND I WILL TALK ABOUT ALL OF THESE IN MORE DETAIL. SO THE FIRST VERY SIMPLY IS, MATH AS PLAIN TEXT, THIS HAS BEEN AVAILABLE SINCE WORD WAS FIRST INTRODUCED BECAUSE IT'S WHAT YOU CAN TYPE AT THE KEYBOARD, WHAT YOU CAN ADD WITH SPECIAL CHARACTERS AND FONT BASED CHANGES LIKE SUPER SCRIPT, SUBSCRIPT AND ITALICS. IT WORKS WELL FOR MATHEMATICAL FORMULAS AND IT CAN BE SHOWN ON A SINGLE LINE. HOWEVER IT WASN'T ALL THAT SIMPLE BECAUSE IN THE DARK AGES IT COULD BE CHALLENGING. THERE WAS AICISM SYMBOL FONT THAT WAS AVAILABLE BUT THAT WAS LIMITED TO 200 CHARACTERS AND MATHEMATICIANS NEEDED FAR MORE OF THAT SO WE HAD A PROLIFERATION OF MATHEMATICAL FONT. SO WHEN WE RECEIVED ADOBE PI FONT SET IN THIS 6 POETIC PARTS AND THERE WERE MATHEMATICCA FONTS AND GOD KNOWS HOW MANY OTHERS WERE CREATED WHEN YOU COULD ONLY HAVE 230 OR SO CHARACTERS IN A COMPLOAT FONT AND THERE WAS A COMPLETE LACK OF STANDARDIZATION ON THIS, SO WE HAD AUTHORS WHO WOULD MAKE THEIR OWN FONT. AND THEN OF COURSE, THEY WOULD SUBMIT THE PAPER AND THE PAPER WOULDN'T RENDER CORRECTLY BECAUSE THEY HADN'T SUBMITTED FONT WITH IT. SPECIAL SYMBOLS ARE A MUCH EASIER THING TO DO WITH TODAY BECAUSE OF UNICODE. THERE IS EXTENSIVE SUPPORT FOR SPECIAL CHARACTERS BOTH WITHIN THE NEWER MICROSOFT FONTS THAT SHIFT WITH 2007 AND LATER AND ALSO THE STICKS INITIATIVE AND IT'S MUCH EASIER TO ADD THEM BOTH THROUGH WORDS INSERT SYMBOL MENU AND 1 OF MY FAVORITE WORD TRICKS IF YOU TYPE THE HEXADECIMAL CODE FOR AUNE CODE VALUE, HIGHLIGHT THAT CODE AND PRESS ALT-X, AND SO IF I TYPE 03 D2 AND HIGHLIGHT THOSE CHARACTERS AND PRESS ALT-X, I WILL GET A GREEK BETA AND IF I HIGHLIGHT BETA AND PRESS ALTASD-X, I WILL GET THEUNE CODE VALUE AGAIN. SO IT'S A USEFUL TRICK WITH MICROSOFT WORD. CONVERTING THESE TEXT EQUATIONS TO XML, IN THEORY YOU CAN JUST MAKE THEM PLAIN TEXT. YOU CAN ALSO IF THEY'RE IN LINE, YOU CAN WRAP THOSE IN-LINE FORMULA, YOU CAN ALSO HAVE DISPLAY FORMULA THAT ARE ENTERED AS PLAIN TEXT AND YOU CAN SEE THIS COME NOTHING FROM AUTHORS AND SO BOTH OF THESE FRAGMENTS WOULD BE COMPLETELY VALID, I'M NOT SURE IF THE EQUATION WAS AT ALL, IT I WAS AN ENGLISH MAJOR AND I HAVEN'T HAD A MATH COURSE IN 40 YEARS, AND THEN, YOU COULD VERY SIMPLY, IF IT'S A DOCK X FILE, YOU CAN OPEN THAT UP, TAKE THE DOCUMENT.X FRAGMENTS, XML YOUR WAY OVER TO JATS WITH OBVIOUSLY A MUCH MORE COMPLICATED TRANSFORM BUT THEN CAN YOU GRAB THIS MATH AND WATCH OUT IF NEEDED. WITHOUT AUGHT WITH CODING BECAUSE WITH FONTS THERE ARE LOTS OF AUTHORS WHO ARE USING OLD FONTS FOR MATHEMATICS AND YOU HAVE TO WORRY ABOUT WHAT FONT, WHAT CODING AND DO A MAPPING FOR A SPECIAL CHARACTER SET. NOW THIS LEADS TO AN INTERESTING STRUCTURE. WHEN IS MATH REQUIRED? SOME PEOPLE WOULD BE HAPPY TO HAVE NLESS THAN 5 IN A PARAGRAPH OTHER THAN AN ITALIC AND OTHERS WOULD SAY NO IT'S AN EQUATION AND IT NEEDS TO BE MARKED WITH THE INLINE FORMULA. SO THEN HAVE YOU A QUESTION OF WHAT CONSITUTES AN EQUATION. IN LESSON 5, OKAY HA ABOUT A RANDOM ATALLIC IN THE MIDDLE OF TEXT, IS THAT AN EQUATION AND THERE ARE ALL KINDS OF PHILOSOPHICAL WARS THAT RAGE OVER THIS. I KNOW 1 PUBLISHER, I'LL LEAVE THEIR NAME OUT, THAT INSISTERS THAT ABSOLUTELY ANYTHING THAT EVEN REMOTELY SMELLS OF MATH, MUST BE MARKED UP AS A FORMULA, IN FACT, THAT PUBLISHER IN THEIR PREJATS PROPRIETARY DTD DID NOT HAVE ELEMENTS FOR SUPER CRYPT AND SUBSCRIPT BECAUSE IF YOU NEEDED IT HAD TO BE AN EQUATION. INTERESTING APPROACH TO SOLVING THE PROBLEM. IF THESE EQUATIONS DO NEED TO BE MATH ML OR TEX IN WHAT YOU CREATE OUT OF MICROSOFT WORD, THERE ARE 2 WAYS TO APPROACH IT. FIRST IS YOU CAN HAVE SOMEONE GO THROUGH AND REKEY ALL OF THESE LITTLE FRAGMENT INTO'S A PROPER MATH EDITOR AND WE WILL GET TO WHAT THOSE ARE IN A MINUTE. OR WE CAN TRY TO WRITE A SCRIPT THAT WILL PICK THESE OFF AUTOMATICALLY. CAN YOU DO THAT WITH 80% ACCURACY BUT ULTIMATE GOAL MAYLY YOU WILL HAVE HUMAN REVIEW BECAUSE FIGURING OUT WHAT THE BOUNDS OF THE EQUATION ARE WILL BE A BIT CHALLENGING, IT'S OFTEN CONTEXTURAL. I'VE ALWAYS PERSONALLY AND THIS IS OPINION AS TONY WOULD SAY, I'VE QUESTIONED THE WISDOM OF NEEDING TO MAKE ALL THESE THINGS EQUATIONS, SOME PEOPLE INSIST ON IT FOR THESE PRESENTED RESULTS. THE OTHER THING YOU'LL RUN INTO WHEN HAVE YOU KEYBOARD AND TEXT IS THAT AUTHORS CAN BE CREATIVE SO TO SPEAK AS TO HOW THAT I ENTER THINGS. SO I'VE GOT A LITTLE TABLE HERE WITH A COUPLE EXAMPLES OF SOMETHING THAT'S WRONG AND SOMETHING THAT'S RIGHT. NOW THEY LOOK VERY SIMILAR. BUT PEOPLE USED VARIOUS TRICKS WITH FONT FACES AND MAYBE FONTICIZES AND SO TO TRY TO SET THESE UP, SO IF YOU TRY TO READ THESE WITH A SCREEN READER YOU'LL GET VERY INTERESTING THINGS. SO FOR EXAMPLE, IN THE WRONG COLUMN, THE LEFT SIDE, THAT IS AN UNDERLINED PLUS SIGN, NOT A PLUS/MINUS CHARACTER. SO THAT WILL NOT SOUND LOGICAL TO SOME BE WHO'S BLIND AND THIS IS BEING RED BY A SCREEN READER. OR THE NEXT PIECE DOWN AND THIS IS A REAL TASTE, THE AUTHOR OF THAT LEFT HAND NEGATIVE 0.5 KNEW THEY NEEDED A MINUS SIGN. BECAUSE THEY KNEW THAT A MINUS SIGN WAS LONGER THAN A HYPHEN BUT THEY COULDN'T FIGURE OUT HOW TO INSERT A MEANUS SIGN. SO WHAT DID THEY DO? THEY INSERTED AN UNDERSCORE CHARACTER AND MADE IT SUPERSCRIPT AND THEN TO GET EXTRA SPACE, THEY THROUGH A THIN SPACE BEFORE THE 0. YOU HAVE TO REMEMBER THAT AUTHORS OF SCHOLARLY PAPERS ARE SOME OF THE SMARTEST PEOPLE IN THE WORLD. WE'RE TALKING ABOUT NOBEL PRIZE WINNERS, BUT THAT DOESN'T NECESSARILY MEAN THEY'RE SMART ABOUT HOW THEY USE MICROSOFT WORD. SO YOU WILL SEE INCREDIBLY CREATIVE USE OF MICROSOFT WORD. SO BE FOREWARNED, YOU SHOULD PLAN TO DO SOME EDITORIAL CLEAN UP IF YOU'RE GRABBING KEYBOARDED MATH. SO MODEL NUMBER 2, MATH AS GRAPHICS THIS IS SIMPLE. THERE ARE PEOPLE WHO WILL FIT THERE AND TAKE WINDOWS PHOTOSHOP AND DRAW, QUAGSS BECAUSE THEY DON'T KNOW THERE'S EQUATION EDITOR IN WORD. THEY FORGOT TO CHECK OUR FORMAL OR INFORM AT IT DEPARTMENT TO LEARN HOW TO INSERTA AN EQUATION. WHEN YOU GET THESE THERE'S NO AUTOMATED WAY TO INSERT THEM INTO MATH ML, YOU WILL HAVE TO CAREFULLY REKEY THEM, IF YOU'RE A PUBLISHER THAT HAS A VERY SMALL VOLUME OF MATH AS MOST MEDICAL JOURNALS ARE, YOU COULD PROBABLY LEAVE EVERYTHING AS GRAPHICS IN YOUR WORK FLOW AS TOMMY POINTED OUT YESTERDAY. MATH ML IS NOT THE RIGHT SOLUTION OR TECH FOR A PUBLISHER THAT GETS TO DISPLAY EQUATIONS A YEAR: THE 1 THING TO KEEP IN MIND WITH GRAPHICS THOUGH IS THAT TYPICALLY THESE ARE GOING TO BE BIT MAP GRAPHICS, SO THEY'RE NOT SCALABLE, THEY WON'T CONSISTENTLY ALIGN WITH THE TEXT. THEY'RE NOT MACHINE READABLE YOU MIGHT WANT TO CONSIDER ALT-NOTE FOOT TEXT, WE DISCOVERED THAT THE INLINE GRAPHICS IN JATS DOES NOT HAVE A LONG DESK ELEMENT WITHIN THE REQUEST HAS BEEN MADE FOR THAT IN 1.2. THESE ARE NOT ACCESSIBLE, YOU COULD USE ALT-X BECAUSE IT'S AVAILABLE IN THERE WITH INSIDELINE GRAPHIC. METHOD NUMBER 3, OBSCURE EQUATION AND FORMULA FEELS. THESE HAVE BEEN AVAILABLE SINCE WORD 1.0. THIS IS THE OLDEST FORM IN WORD, THEY'RE ALL ABOUT HIDDEN. IT TOOK ME 6 CLICKS IN A CURRENT VERSION OF WORD TO GET TO THE POINT THEY COULD INSERT AN EQUATION, SO RATHER THAN CLICKING THE IPT GREATER CERT EQUATION, IT TOOK 6 CLICKS TO GET TO THIS SO IF YOU WANT TO KNOW HO DO IT, READ THE PAPER, WE HAVE IT IN THE FOOT NOTE. HOW DO YOU DETECT THESE THINGS? WELL THERE IS A FEATURE IN WORD AND THIS IS ANOTHER COOL FEATURE YOU CAN GO TO IN THE OPTIONS DIALOGUE SO THAT EVERY FIELD WILL BE SHADED WITH GRAY. SO THE FIRST THING YOU WANT TO THE DO IS TURN ON FIELD SHADING AND THEN TO DISTINGUISH THESE, YOU HAVE TO RIGHT CLICK ON IT AND YOU WILL SLEEP APNEA AND OBESITYY EDIT FIELD ON THE DROP DOWN MENU AND THAT'S THE ONLY WAY YOU CAN FIGURE OUT THESE ARE THERE, SHORT OF WRITING WORD MACROS. THERE ARE HOWEVER LIMITED XML CONVERSIONS, OPTIONS FOR THESE. IF THERE IS A SINGLE LINE AT THE KEYBOARD YOU COULD REKEY IT AS TEXT. YOU COULD TAKE YOUR DOC OR DOC X FILE AND SAVE IT TO HTML AND YOU HAVE A LOW SCALE GRAPHIC IF YOU WANT A GRAPHIC, YOU CAN USE THAT OR YOU CAN USE THIS TOON CONVERT TO MATH TYPE AND THEY BECOME MORE MANIPULATABLE. WE DON'T OFTEN SEE THESE, MICROSOFT BURIED THAT FORTUNATELY MOST PEOPLE CAN'T FIGURE OUT HOW TO FIND IT. BUT EVERY ONCE IN A WHILE YOU'LL SEE A PAPER THAT HAS USED THE EQUATION OR FORMULA FIELD. THIS LEADS US TO METHOD NUMBER 4 WHICH IS DESIGN MATH TYPE. MATH TYPE HAS BEEN AROUND SINCE 1987 AND THIS WAS DESEENED BY OR CREATED BY DESIGN SITES AS WHAT YOU SEE, WHAT YOU GET EQUATION EDITOR, TO SERVE AS AN ADD IN, NOT JUST TO MICROSOFT OFFICE BUT ADOBE INDESIGN WORD PRESS AND ANY APPLICATION THAT COULD WORK WITH OLAY OBJECTS FROM A TECHNICAL GEEK SPEAK PERSPECTIVE AND 1 THING ABOUT MATH TIME IS IT PROVIDES A LOT OF DIFFERENT WAYS TO GET IT OUT OF WORK, INCLUDING GOING TO MATH ML, EPS OR GIF AND YOU CAN WRITE CUSTOM TRANSFORMS FOR THIS. AND IN FACT, THERE IS A FULL MATH TYPE SDK, IT IS AVAILABLE FOR PREE. THE URL IS IN THE SLIDES AND THE PAPER. THIS GIVES IPT GREATER STRUKSS ON HOW YOU CAN WRITE YOUR OWN CUSTOM TRANSLATORS TO GO FROM MATH SITES INTERNAL FORMAT TO WHATEVER INTERNAL TARGET YOU WANT. IF YOU WANT A TRANSFORM TO WRITE AAM MATH, IF YOU THOUGHT IT WOULD BE USEFUL YOU COULD DO THAT WITH DOCUMENTATION. THERE'S ALSO API FOR ACCESSING MATH TYPE DIRECTLY ITSELF, SO CAN YOU INTEGRATE MATH TYPE IN TO HAVE A HUGE AMOUNT OF AUTOMATION IN A WORK FLOW WORKING WITH MATH. THIS LEADS ME TO SORT OF THE COUSIN OF MATH TYPE WHICH IS THE MICROSOFT EQUATION EDITOR, METHOD NUMBER 5. IN 1991 MICROSOFT REALIZED THAT THE FORMULA FIELDS WEREN'T REALLY WORKING WELL FOR SERIOUS MATHEMATICIANS AND I DON'T KNOW IF THEY LOOK TIME OR THE WILL BUT THEY LICENSEDDA I SUBSET OF MATH TYPES WHICH BEGAME KNOWN AS MICROSOFT EQUATION EDIT AND ABOUT BUILT THAT INTO MICROSOFT WORD AND IT WAS WELL RECEIVE FEDERAL YOU LOOK BACK ON REVIEWS AND PC MAGAZINES AT THE TIME THEY SAID HEY, THIS IS GREAT, IT'S MUCH, MUCH, EASIER TO USE THAN FIELDS BUT IT IS A SUBSET OF TYPES, IT LACKS TEMPLATES, SYMBOLS AND EXPORT AND CONVERSION FEATURES. HOWEVER IT DID,A LOW AUTHORS TO WORK WITH A MUCH RICHER SET THAN FORMULA FIELDS. HOWEVER, BECAUSE IT'S THE SUBSET OF MATH TYPE AND IT DOESN'T HAVE ANY OF THOSE EXTRA FEATURES FOR CONVERSION AND EXPORT, THERE'S NO NATIVE WAY TO TAKE THESE EQUATIONS AND GET THEM INTO AN XML WORK FLOW. SO YOUR OPTIONS OR THE SORT LIST IF YOU'RE SETTING ASIDE MATH FOR A SECOND WOULD BE TO SAFE THE FILE AS HTML AND AGAIN GET NONSCALABLE GRAPHICS OR UPGRADE TO MATH TYPE. AND THIS IS NOT A REQUIREMENT FOR THE AUTHORS YET TO UPGRADE. THE AUTHORS CAN WORK WITH THE MICROSOFT EQUATION EDIT BUT THE EDITORIAL AND PRODUCTION TEAMS WOULD NEED TO HAVE MATH TYPE ON THEIR SYSTEMS AND THIS IS QUITE COMMON. THE PUBLISHERS WILL EQUIP EDITORS AND PRODUCTION PEOPLE WITH THE FULL FEATURED MATH TYPE. SO GOING FROM MATH TYPE TO XML, MATH TYPE CAN GO BOTH WAYS. IT CAN TAKE MATH ML AND LATEX IN THE MORE RECENT VERSION, 6.5 AND LATER AND CONVERT THEM INTO MATH TYPE OBJECTS AND IT CAN TAKE ANY MATH TYPE OR MICROSOFT EQUATION OBJECT AND CONVERT THOSE TO MATH ML OR LATEX, FROM MICROSOFT WORD. AND BECAUSE OF THIS RICH SET OF CONVERSION FEATURES, MATH TYPE HAS BECOME THE TOOL OF CHOICE FOR PUBLISHERS TO WRANGLE ALL OF THIS DIFFERENT FROM WORD OUT TO A JATS WORK FLOW. SO WHEN YOU'RE GOING--WHEN YOU'RE WORKING WITH THE EICATION EDITOR MAGS AND MATH TYPE, FIRST OF ALL IF YOU HAVE MATH TYPE ON YOUR SYSTEM AND YOU LOAD A DOCUMENT THAT HAS EQUATIONS THAT WERE DONE WITH EQUATION EDIT MATH SITE CAN DETECT THOSE AND AUTOMATICALLY CONVERT THEM INTO MATH TYPE OBJECTS. ONE OF THE THINGS TO REALIZE IS THAT THOSE MATH TYPE OBJECTS ARE NOW MATH TYPE OBJECTS THEY ARE NOT EQUATION EDITOR OBJECTS. SO THEY'RE NOT BACKWARDS COMPATIBLE. AND SO WHILE MATH TYPE CAN WORK WITH MATH TYPE AND EQUATION EDITOR OBJECTS, EQUATION EDITOR CAN CANNOT WORK WITH THE MATH TYPE OBJECTS. IF YOU HAVE A SITUATION WHERE YOU CONVERTED THESE OBJECTS TO MATH TYPE FORMAT AND YOU NEED TO GIVE THE PEOPLE BACK TO AN AUTHOR WHO DOESN'T HAVE MATH TYPE. THEY CAN THEN ACTUALLY DOWNLOAD FOR FREE, THE 30 DAY TRIAL OF MATH TYPE, WHICH WILL ALLOW THEM FOR 30 DAYS TO HAVE ACCESS TO OLOGY OF THE MATH TYPE FEATURES AND THEN AFTER 30 DAYS WILL ACTUALLY RUN IN A LIGHT MODE SO CAN YOU EDIT THOSE EQUATIONS USING THE SUBSET OF FEATURES THAT WOULD BE AN EQUATION EDITOR. SO YOUR AUTHORS WOULDN'T BE STUCK IF YOU GIVE THEM BACK A PAPER THAT HAS MATH TYPE EQUATIONS AND THEY DON'T WANT TO PAY FOR MATH TYPE. THIS IS THE CONVERT EQUATIONS DIALOGUE FOR MATH TYPE. AS CAN YOU SEE IT HAS A BUNCH OF SOURCE FORMATS ON THE LEFT, MATH TYPE AND EQUATION EDITORS AND EQUATIONS, MICROSOFT EQUATION FIELDS, WORD 2000 LATER OR AML EQUATIONS AND THESE CAN BE CONVERTED EITHER TO MATH TYPE OBJECTS OR FORMATS SUCH AS CAN YOU GO TO LATEX OR MATH ML WITH THIS DIALOGUE. THE CONVERT EQUATION FEAT SURACTUALLY USEFUL TO RUN EVEN ON MATH TYPE EQUATIONS BECAUSE SOMETIMES WE FIND THAT EQUATIONS CREATED WITH OLDER VERSIONS OF MATH TYPE IF YOU RUN THEM THROUGH THIS WITH THE NEWER VERSION OF MATH TYPE, THE OBJECT IS CLEANED UP AND THAT HELPS GET A CLEANER TRANSFORMATION FROM MATH TYPE TO MATH ML OR LATEX OR WHATEVER FORMAT YOU'RE CHOOSING TO CONVERT TOO. AN IMPORTANT NOTE: WHENEVER YOU RUN ANY KIND OF TRANSFORM WEATHER IT'S WORD TO XML OR WHETHER IT'S EQUATION BUILDER MATH TO MATH TYPE ANY KIND OF TRANSFORM, YOU SHOULD REPROOF THE EQUATIONS JUST? CASE AN ERROR CAME UP, IT DOESN'T ALWAYS HAPPEN BUT IT DOES IF YOU RUN A CONVERSION ON IT. SO 1 OF THE NEAT FEATURES IN MATH TYPE IS YOU CAN ACTUALLY SAVE THE EQUATIONS NOT JUST TO MATH ML AND LATEX, BUT YOU CAN ALSO CONVERT THEM TO GRAPHICS AND FINALLY TO A FORMAT THAT'S OTHER THAN BIT MAP. YOU CAN GO AND ALSO GO TO EPS. AND THIS IS REALLY USEFUL FOR WORK FLOWS, WHERE YOUR RENDERING ENVIRONMENT MAY NOT HAVE SUPPORT FOR MATH ML. FOR EXAMPLE, IN DESIGN SUPPORT FOR MATH ML, THERE ARE THIRD PARTY RENDERS THAT CAN ADD IT BUT IT CAN'T RANDOMIZED TRIALER MATH ML SO A NUMBER OF PUBLISHERS WE WORK WITH ACTUALLY EXPORT WHEN THEY CONVERT WORD FILE TO JATS, THEY WILL EXPORT THEIR EQUATIONS IN 2 FORGEATS AND ETERNAL ELEMENTS AND HAVE THAT FORMAT EPS AND ALSO XML OR EITHER MATH ML OR TEX AND ALTERNATIVES IS GREAT FOR THAT SCENARIO, SO THE DESIGN THEY WILL RENDER FOR EPS AND EVERYTHING ELSE THEY WILL RENDER FOR MATH ML OR TEX. A COUPLE THINGS TO WATCH OUT FOR WITH MATH TYPE. AUTHORS CAN CREATE EQUATIONS TYPICALLY BY MISUSING SUPER SCRIPT AND SUBSCRIPT WITHIN MATH TYPE THAT DO NOT CONVERT TO MATH ML CORRECTLY. SO CAN YOU CATCH THIS BY PROOFING IF YOU USE THE MATH TYPE API, THERE ARE ACTUALLY WAYS TO TRAP THAT ERROR. AND GIVE WARNINGS ON IT, BECAUSE IT'S NOT SOMETHING AUTOMATICALLY SHOWS UP SO YOU HAVE TO BE CAREFUL OF THAT THE OTHER THING IS IF YOU ARE CONVERTING EQUATIONS TO EPS FORMAT, AND EPS DOES NOT NECESSARILY HAVE UNICODE SUPPORT IN THE WAY THAT MATH TYPE SET IT UP AND SO, SOMETIMES CERTAIN SYMBOLS WILL NOT COME THROUGH CORRECTLY IN PARTICULAR. I KNOW THAT SOME OF THE PUBLISHERS WE ARE WORK WITH HAVE PROBLEMS WITH THE EURO SYMBOL AND WITH ACACCENTS ON WITH GREEK OR LATIN LETTERS. SO BE CAREFUL WITH THE EXPORT. DEEP BREATH AND A SIP OF WATER. THIS PAPER MIGHT NOT HAVE BEEN NECESSARY IF WE ONLY HAD 5 WAYS OF INSERTING AN EQUATION. BUT MICROSOFT MUDDIED THE WATERS ENORMOUSLY IN 2007 WHEN THEY BROUGHT OUT OFFICE 2007 AND THEY ADDED WHAT THEY CALLED THE EQUATION BUILDER AND YOU CAN TELL IT'S AN EQUATION BUILDER EQUATION BECAUSE WHEN YOU CLICK ON IT IN WORD, YOU GET THE FUNKY HANDLEOT LEFT AND THE DOWN ARROW ON THE RIGHT WITH THE SCROLL ZONE. WHY EQUATION BUILDER? THAT'S WHAT A LOT OF US ASKED OURSELVES WHEN MICROSOFT ANNOUNCED THIS FEATURE, WE HAVE MATH TYPE: IT WORKS WELL, WHY DO WE NEED YET ANOTHER EQUATION EDITOR. WELL, FIRST OF ALL, UNDER THE HOOD, IT USES SOMETHING KNOWN AS OPEN MATH MARK UP LANGUAGE. SO SOMETIMES YOU WILL HEAR PEOPLE REFER TO THESE EQUATIONS AS NOT EQUATION BUILDER BUT OMML, SO I WILL USE THOSE INTERCHANGEABLY AS I TALK BUT MICROSOFT ACTUALLY DID HAVE GOOD REASONS FOR IT. FIRST OF ALL TYPOGRAPHY, MATHEMATICIANS REALLY CARE ABOUT TYPOGRAPHY. REMEMBER CONAILED KNUTH, HE CARED ABOUT TIOGRAPHY AND ADMITTEDLY, - NAYSER IN MICROSOFT WORD THAN MATH TYPE AND PART OF THAT IS BECAUSE OF THE CAM BRIA MATH FONT. THEY DID A NICE JOB BUT A ALSO WANT A TIGHTER INTERIGRATION OF MATHEMATICS WITH WORD, THEY WANTED THEM TO AUTHOR FOOT NOTES INTO AN EQUATION, HOW DO YOU REPRESENT THAT IN JATS XML. SHRKS I HAVEN'T SEEN AN AUTHOR DO IT. THEY WANTED PEOPLE TO BE ABLE TO PUT COMMENTS IN AN EQUATION, HOPEFULLY THEY WILL PUT IT AWAY BEFORE X, ML. THEY WANTED FINDING A REPLACE TO WORK THROUGHOUT THE DOCK YOU WANTS INCLUDING INEQUATION, SO THESE WERE ALL GOOD REASONS, HOWEVER. MICROSOFT FORGOT TO TALK TO THE SPM COMMUNITY BEFORE THEY RELEASE THE FEATURE, AND BEFORE THEY WENT TO THE DOC X FILES FORMAT SO THE COMBINATION OF THIS NEW EQUATION EDITOR AND ALSO THE RADICAL USER INTERFACE OVERALL, THEY WENT TO THE IBT--INTEGRATEON INTERPHASE, THE SCM COMMUNITY JUST SORT OF SAID, AHH... NO. SO THERE WERE PUBLISH COMMENTS THAT WERE MADE AT THE TIME AND THE SCM COMMUNITY SAID UH-UH, NOT YET, NOT SO FAST. AND SO WHAT HAPPENED AMONG OTHER THINGS IS A LOT OF PUBLISHERS WEREN'T WILLING TO UPGRADE YET FROM 2002, 2003, TO WORD 2007 BUT MICROSOFT MADE A CONVERTER SO YOU COULD LOAD DOC X FILES INTO IT EXCEPT THE MATH BECAME FUZZY IMAGES IN THOSE FILES AND YOU COULDN'T GET MATH XML AT THEM. SO THE WAY A LOT OF PUBLISHES DEALT WITH THIS IS FOR YEARS NO DOC X, THEY SAID WE'RE NOT GOING TO ACCEPT DOC X FILES. THINGS HAVE CHANGED A LITTLE BIT IN THE INTERVENING 9 YEARS. SINCE THAT WAS RELEASED, PUBLISHERS UPDATED THEIR SYSTEM SO THAT EVERYONE HAS A DOC X COMPATIBLE VERSION OF MICROSOFT DOC X OFFICE. BUT PUBLISHES RECOGNIZE THEY COULDN'T CONTROL THE AUTHORS AND THEY WERE GOING TO SUBMIT MATTHEWSING WHATEVER TOOL SEEMED TO BE MOST HANDY OR MOST CO CONVENIENT. ALTHOUGH ADMITLY RAY YEAR AGO I DID HEAR A STORY OF A PUBLISHER WHO SHALL REMAIN NAMELESS, AND THEY REJECTED THE PAPER BECAUSE THEY USE TED MICROSOFT EQUATIONS FOR HIS WORD BUILDER AND THEY SAID OUR PROGRAM DOESN'T ACCOMMODATE THESE AND COULD YOU CONVERT THEM. THE AUTHOR WAS NOT HAPPY ABOUT IT AND I DON'T KNOW IF HE TOOK THE PAPER ELSEWHERE. HOWEVER, THAT STATED PUBLISHER IS STILL OFTEN HAVE PREFERENCES. THIS WAS FROM FROM 2 WEEKS AGO. WE THINK IT'S CURRENT KNOW. WE RECOMMEND USING MATH TYPE FOR DISPLAY AND INLINE EQUATIONS AS IT WILL PROVIDE THE MOST RELIABLE OUTCOME. IF THIS IS NOT POSSIBLE, EQUATION EDITOR IS ACCEPTABLE. NOW EQUATION EDITOR IN THIS CONTEXT COULD MEAN EITHER MICROSOFT EQUATION EDIT OR EQUATION BUILDER, I'M NOT SURE. I HAVE NOT QUERIED FOR CLASSIFICATION. I DON'T KNOW. BUT THE POINT IS THERE ARE PUBLISHERS WHO DO PREFER GETTING PATH TYPE FOR EQUATION. ONE OF THE OTHER REASONS WHY THIS BECAME SUCH A STICKING POINT WHEN MICROSOFT BROUGHT OUT THIS NEW FEATURE IS BECAUSE THEY DID HAVE AN OMML TO MATH CONVERSION FEATURE BUT IT WAS INCREDIBLY WELL HIDDEN, POORLY DOCUMENTED IN THIS FACT FOR THE FIRST 6 MONTHS AFTER OFFICE 2007 WAS RELEASED IF YOU TRIED TO GOOGLING YOU GOT 0 HITS. IT WAS ALSO BUGY. THE ORIGINAL RELEASE OF OFFICE 2007, A LOT OF EQUATIONS, THE MATH ML WAS NOT VALID FOR MATH NL BTD AND BY SERVICE BACK 2 OF WORD 2007 IT WAS A LOT BETTER BUT THAT TRANSFORM WAS NOT PRODUCTION READY UNTIL WORD 2010 CAME ALONG. THE TRANSFORM ITSELF FOR THOSE WHO WANT TO PEEK UNDER THE HOOD, YOU WILL FIBBED THESE 2 XSLs AND THESE ARE USED BY MICROSOFT TO CONVERT FROM OMML, TO MATH ML, AND ALSO FROM MATH ML TO OMML, AND YOU HAVE MATHEMATICCA, CAN IT PUTS ON THAT THE BOARD, CAN YOU PASTE IT INTO MICROSOFT WORD AND GET EQUATION BUILDER MATH OBJECTS. HOWEVER IF YOU WANT TO COPY AND PASTE MATH ML OUT OF WORD, YOU NIGHED TO FIND THE OBSCURE SETTING IN THE EQUATIONS DIALOGUE TO COPY IT ON TO THE CLIPBOARD, NOT OMML FORMAT. THE XSLTs CAN BE USED OUTSIDE OF WORD SO IF YOU CREATE A CONSTRUCT THAT DECONSTRUCTS WITHOUT OPENING UP WORD, YOU CAN USE THESE XSLs, YOU CANNIES THIS IF IT'S COMMERCIAL, CAN YOU USE IT AT NO CHARGE SO YOU CAN INDGREAT THAT INTO A CONVERSION PROCESS. SO 1 OF THE QUESTIONS THAT WAS ASKED IS WE HAVE MATH ML. WHY DID MICROSOFT ACTUALLY NEED YET ANOTHER MARK UP LANGUAGE FOR MATH WITHIN MICROSOFT WORD. WHY DID THEY NEED OMML, WHY DIDN'T THEY SHOVE MATH ML INTO MICROSOFT WORD. WELL A COUPLE OF REASONS. TO ILLUSTRATE THIS WE HAVE A SIMPLE EQUATION AND WE HAVE THE OMML MARK UP AND THE MATH ML MARK UP FOR IT. OMML WAS ALIGNED WITH MICROSOFT WORD'S OVERALL INTERNAL MARK UP, MORE THAN MATH ML WAS. AND IN PARTICULAR OMML WAS USED TO CREATE MATH ZONES AND THE PROBLEM WITH MATH ML IS IT COULD ONLY BE USED FOR MATH AND OMML, CAN YOU STICK ANYTHING YOU CAN STICK IN A WORD PARAGRAPH SO YOU COULD HAVE THINGS LIKE FOOT NOTES OR COMMENTS SIT NOTHING THE MIDDLE OF AN OMML EQUATION. ALSO WITH OMML, THE ARGUMENTS ARE EXPLICIT TAGS, WHEREAS WITH MATH ML, THEY'RE DETERMINED BY POSITION. SO MICROSOFT HAD GOOD REASONS IF TAR THIS. IF YOU WANT TO LEARN MORE ABOUT THE BACKGROUND ON IN, MURRAY SERGEANT WHO DEVELOPED OMML WROTE A FAIRLY LONG TECH NOTE THAT'S ON THE UNICODE SITE EITHER 25 OR 28-FOOT NOTES AND IT'S IN OUR PAPER AS A REFERENCE. SO CAN YOU LEARN MORE ABOUT THAT. CONVERTING HOWEVER, IN AND OUT OF OMML, HAD TRICKS, FIRST EQUATION BUILDER IN SITE, CAN'T CONVERT BETWEEN THE 2. MATH IS THE INTERMEDEIARY SO IF YOU REMEMBER THAT CONVERT EXPECTATIONS QUAGSS DIALOGUE FROM A FEW MINUTES AGO, IF YOU CONVERT FROM OMML, TO MATH SITE, WHAT'S HAPPEN SUGGEST MATH TYPE TAKES THE OMML, AND CALLS MICROSOFTAND CONVERT TO MATH TYPE AND MATH TYPE CONVERTS THAT MATH ML INTO MATH TYPE. AGAIN, IT'S ESSENTIAL TO REPROOF WHEN YOU CONVERT BECAUSE--WHEN YOU'RE CONVERTING BECAUSE HERE YOU'VE DONE 2 CONVERSIONS, FORMAT TO MATH ML AND MATH ML TO FORMAT B. YOU CAN GO EQUATION BUILDER TO MATH TYPE AND THAT'S QUITE SIMPLE JUST BY USING THE MATH TYPE CONVERT EQUATION FEATURE; GOING MATH TYPE TO EQUATION BUILDER IS HARDER BECAUSE MICROSOFT HASN'T DEVELOPED ANY AUTOMATION AROUND THIS. AS FAR AS THEY'RE CONCERNED IT DOESN'T EXIST BECAUSE THEY NOW HAVE EQUATION BUILDER SO WHAT YOU CAN DO IS GO MATH TYPE TO MATH ML TO EQUATION BUILDER AND CAN YOU USE MATH TYPE TO AUTOMATE GOING TO MATH ML BUT YOU HAVE TO WRITE A MACROTO GO THROUGH THE WORD DOCUMENT TO PICK THEM UP AND PUT THEM IN THE GOBBING YOU WANT AND THIS HAS BEEN DOCUMENTED IN MURRAY'S INSTITUTE. THE WHAT YOU IF WANT TO GO TO TEX, I KNOW A PUBLISHER WHO USES TEX FOR IT RATHER THAN MATH ML. WELL GETTING TEX OUT OF MATH M, ML IS A NEW FEATURE, WHO WILL NEED IT? WELL THEY WERE WRONG. SO CAN YOU GO FROM MATH TYPE AND CONVERT TO OMORKS ML. AND THEN FROM MATH--I'M SORRY, YOU CONVERT THE OMML TO MATH TYPE AND THEN EXPORT TO TEX WHICH IS ACTUALLY A VERY COMMON WORK FLOW, OR YOU COULD ESSENTIALLY EXPORT OUT THE OMML OR MATH ML AND THEN USE THIRD PARTY SCRIPTS WITH SOMETHING LIKE 10 DOC TO CONVERT TO THE MATH NL OR OMML TO TECH. SO THERE ARE A COUPLE OF WAYS WE CAN SLICE THAT SO MOST PUBLISHERS GO TO MATH ROUTE EVEN THOUGH THEY HAVE TAKE THE TIME TO REPROOF THE EQUATIONS. WHAT ABOUT OMML, TO SCALABLE GRAPHICS. WELL, HEY FROM MICROSOFT'S PRESPECTIVE, YOU CAN MAKE A PDF FROM A WORD FILE SO WHY WOULD YOU MAKE SCALABLE GRAPHICS OUT OF THE EQUATION BUILDER MATH. WELL THE ANSWER IS NOT EVERYONE TYPES THE PDF IN MICROSOFT WORK. SO 1 OF THE GLARING OMISSIONS FOR MICROSOFT WORD IS THAT YOU CAN'T CREATE SCALABLE GRAPHICS OF THOSE OMML EQUATIONS. AND THIS IS ANOTHER REASON WHY A LOT OF PUBLISHERS WE KNOW WILL CONVERT THE OMML TO MATH TYPE TO HARMONIZE IT IN THEIR WORK FLOW BEFORE THEY GO AND DO EVERYTHING ELSE. SO 1 WAY YOU CAN GET THE SCALABLE GRAPHICS OUT IS WE CAN GO OMML TO EPS, THE OTHER THING CAN YOU DO IS YOU CAN GO OMML, TO MATH ML AND USE THIRD PARTY TOOLS TO GO TO SCG AND THERE ARE A NUMBER OF WAYS CAN YOU DO THAT IPT GREATER CLUING MATH JACKS. A FOOT NOTE BY THE WAY, IF YOU ARE TRYING TO FIGURE OUT WHAT AN EQUATION IS, WHEN YOU CLICK ON IT IN A WORD FILE, ON THE LEFT, IF YOU GET THE 4 HANDLES LIKE THAT, IT'S A MATH TYPE EQUATION, IN THE CENTER IF YOU GET THE FUNKY HANDLE ON THE LEFT AND THE CONTROL BAR IT'S EQUATION BUILDER AND ON THE LEFT IF IT'S FUZZY THAT MEANS SOMEBODY SAVED THE FILED FROM DOC X TO DOC FORMAT SO THE EQUATION WAS DOWN CONVERTED TO A GRAPHIC BUT NO FEAR ON THIS, MICROSOFT DID THIS 1 REALLY, REALLY WELL. YOU WON'T ARE HE OFTEN SAY THIS ABOUT MICROSOFT, IF YOU RESAVE THAT FILE TO DOC X FORMAT, THAT WILL DO WHAT IT'S CALLED REHYDRATION, IT WILL TURN MAGERALLY BACK INTO A BUILDER EQUATION. SO IF YOU SEE THAT FUZZY MATH, DON'T PANIC, DON'T REKEY IT, SAVE THE DOCUMENT IN DO DOCK X FORMAT. WHAT ARE THE WORK FLOW IMPLICATION SAYS OF ALL OF THIS. WELL A FEW THINGS TO THINK ABOUT FROM THIS FROM THE BIG PICTURE PERSPECTIVE, THE FIRST IS IN YOUR OWN WORK FLOW, DECIDE WHAT IS IN-LINE MATH AND WHETHER OR NOT YOU'RE GOING TO TAG IT. IS IT CRITICAL TO MAKE SURE ALL THE MATH IS TEX ORACY TEXT OR HAVE AN INLINE MATH TAG. THE OTHER THING IS, DO YOU WANT TO HAVE ALL OF YOUR EQUATIONS IN TAGGED IN WHAT I'LL CALL THE SAME XM, MRKS L OR DIALECT. IF SO YOU WILL NORMALIZE ALL YOUR MATH TYPE ORI QUAGS BUILDER. NOW WHAT DO I MEAN BY THIS? SAME EQUATION, EQUATION BUILDER, SML AND MATH TYPE MATH ML, IT IS DIFFERENT. I HONESTLY--AGAIN, I WAS AN ENGLISH MAJOR HAVEN'T HAD A MATH COURSE IN 40 YEARS, I DON'T KNOW WHICH 1 OF THOSE MAY BE BETTER OR IF THERE'S EVEN A DIFFERENCE BUT IF YOU CARE OR WANT CONSISTENCY BE WARE THAT THE MATH ML IS NOT THE SAME FROM THESE 2. AND SO THAT MAY BE A REASON YOU WANT TO THINK ABOUT NORMALIZING YOUR EQUATIONS UP FRONT IN YOUR WORKFULLY TO 1 FORMAT OR THE OTHER. A QUICK CLOSING NOTE BEYOND PRODUCTION. WHY DO WE SEE MATH SO OFTEN AS IMAGES WHEN WE LOOK AT SCHOLARLY ARTICLES PARTICULARLY ON THE WEB? HOW MANY OF HAVE YOU TRIED TO LOOK AT MATH ML IN A BROWSER? HOW MANY OF HAVE YOU BEEN SATISFIED WITH THE EXPERIENCE ACROSS A LARGE RANGE OF BROSSERS? I DON'T--GENE, CAN YOU PUT YOUR HAND DOWN? [LAUGHTER] MOST BROWSERS DON'T SUPPORT IT, SOME DID AND HAD IT GO AWAY, CHROME FOLKS SAID MATH IS NOT NOTHING WE'RE GOING TO DO AND INTERNET EXPLORER DIDN'T EVEN TRY, THEY JUST DIDN'T CARE. IF YOU WANT TO KNOW WHAT YOUR BROWSER CAN DO SEE CANIUSE.COM. BREAKING NEWS HOW MANY HAVE SEEN THESE BLOG POSTS IN THE LAST WEEK AND HALF. THESE ARE THE MATH EQUIVALENT OF GOING VIRAL. SOMEONE TOLD ME THEY HEARD THAT THE FIRST POST BY PETER A WEEK AND HALF AGO HAD 20,000 HITS. DO THAT MEAN PEOPLE CARE ABOUT MATH ML? THAT'S--THANK YOU! FOR THOSE WHO DIDN'T HEAR IT ON THE VIDEO A LOT OF PEOPLE JUST SAID YES. THE THIS IS A FASCINATING SET OF BLOGS AND I'M CERTAINLY NOT GOING TO GET INTO THE MIDDLE OF THE PETER AND PAUL DISCUSSION EXCEPT THAT PETER SAID MATH ML IS A FAILED WEB STANDARD, NOTICE THE EMPHASIS ON THE WORD WEB BEFORE YOU READ ALL OF THESE AND I DON'T THINK WE HEARD THE LAST OF THESE YET. BUT AS A CLOSING NOTE BE FAMILIES WITH MATH JAX, THAT'S THE MOST COMMON WAY THAT PEOPLE ARE TRYING TO RENDER MATH IN A WEB ENVIRONMENT FROM EITHER MATH ML OR TEX. AND THIS ALLOWS YOU TO PUT OUT MATH ML THAT IS SCALABLE, SEARCHABLE, ACCESSIBLE. IT CAN BE INTERNATIONAL CLASSIFICATION GRITTED WITH VARIETY OF ACCESSIBILITY TOOLS INCLUDING BENETECH'S MATH ML CLOUD. THE EPUB IS A SAD STORY BECAUSE EPUB 3 SUPPORTS MATH ML, CONTENT MATH NL CAN BE EMBEDDED WITHIN THIS IF YOU'RE DOING A BOOK, BUT THE DEVICES ARE NOT UP TO THIS. THE AND THE A LOT OF THE EPUB READERS DO NOT SUPPORT MATH ML SO YOU HAVE TO RESORT TO USING GRAPHICS AND I THINK MIKE DEAN'S WITH CFA IS HERE, HE GAVE A GOOD TALK ABOUT THIS A FEW YEARS AGO, PLEASE READ HIS PAPER. YEP. I KNOW THAT I'M KEEPING YOU FROM COFFEE. THIS IS MY LAST SLIDE. REALLY JUST GENERAL THOUGHTS INCORPORATING MATH INTO XML REQUEST BE A CHALLENGING PART OF ANY WORK FLOW AND IN PARTICULAR MATH, IN WORD TO XML, THE SIMPLE ANSWER IS IT'S COMPLICATED WHICH IS WHY WE RIGHT THIS PAPER AND YOU CAN BE SUCCESSFUL BY MAKING SURE YOU UNDERSTAND ALL OF THE IDIO SIRCHG ROSETTAYS, AND THERE'S A HUGE AMOUNT OF DETAIL IN THE PAPER. SO PLEASE READ THE PAPER IF YOU WANT ALL THE OTHER UGLY URPD BELLY OF THIS. BUT MOST IMPORTANTLY UNDERSTAND YOUR GOALS AND MAKE SURE YOU HAVE A SET OF POLICIES, MOVE FORWARD FROM ALL OF THAT. THANK YOU VERY MUCH. >> ONE OR 2 BURNING QUESTIONS. >> I WILL GO WITH GENE BECAUSE GENE KNOWS A LOT ABOUT MATH. >> NOT REALLY A QUESTION, GENE REPRESENTING MYSELF, YOU CAN AM TO EPUB STANDARDOT WORKING GROUP. WE ARE ACTUALLY DEPRECATING THE MANNER IN WHICH YOU CAN SWITCH FROM MATH IMAGES TO MATH ML AND DO FALL BACKS. AND NOBODY'S TALKED ABOUT HOW THAT WILL GO FORWARD. >> THANK YOU FOR RAISING IT, IT'S IMPORTANT POINT. I THINK THEY'RE STEAKING A STEP BACKWARDS. >> I THERE ARE A COUPLE POINTS I CAN MAKE ABOUT SWITCH IS THAT IT'S HIT AND MISS WHERE IT'S INCORPORATED. >> IT'S INCORPORATED IN [INDISCERNIBLE]. THE IN GOOGLE PLAY, DON'T EVEN ASK ME ABOUT KINDLE AND I DON'T WANT TO TALK ABOUT ILUX. BUT MORE AND MORE PEOPLE ARE STARTING TO GET MATH JAX INTO THE EPUBS OR REARS, I THINK THE MOST RECEIPT 1 TO COME OUT WAS HELEKIN. SO YOU CAN GO AND LOOK BUT THE THING IS THAT THE STUFF THAT SUPPORTS IT IS USUALLY PURPOSE BASED SO SAFARI LOOKS ON LINE, THE NEW SAFARI ONLINE SUPPORTS MATH JAX BUT NOT THE OLD 1. , OKAY. THANK YOU. >> THANK YOU. >> [INDISCERNIBLE] WHAT. >> OKAY, DEBBIE. >> FORGIVE THE QUESTION: CAN YOU DISCUSS WORD SCIENCE OOML, AND MATH 2 VERSES MATH ML3? >> MATH ML SUPPORT COMING FROM MICROSOFT IN THE TRANSFORT IS MATH ML2, NEITHER OF THEM HAVE BEEN UPGRADED TO MATH ML 3 AT THIS POINT. >> ONE LAST QUICK, FROM WENDLE AND THEN WE'LL BREAK FOR COFFEE. >> YOU MENTIONED SVG ONLY IN THE CONTEXT OF GETTING IT OUT OF MML, DO YOU KNOW ANY WAY TO GET SVG WITHOUT MATH ML. >> NOT OFF THE TOP OF MY HEAD IT WOULD BE AN INTERESTING QUESTION TO GOOGLE AROUND AND SEE IF THERE'S SOMETHING? CAITLIN DO YOU KNOW? YEAH, OKAY, NEITHER OF US KNOW OFF THE TOP OF OUR HEAD. OKAY, THANK YOU VERY MUCH. ENJOY YOUR COFFEE BREAK BE BACK IN 25 MINUTES. [APPLAUSE ] >> OKAY, WE'RE NOW ENTERING INTO A TOOL PHASE. FOR JATS-CON. THE WE HAVE GARRETT AND WIND WITHLE FROM THE NEXT SET, SO GARRETINAL LOCATION YOU ARE FROM [INDISCERNIBLE]. >> IT'S LIKE THE TYPE SETTING SYSTEM AND THERE IS MANY PRONOUNCEMENT OPPORTUNITIES FOR SOME CALL IT TEX AND SOME CALL IT TEX AND E-STAND FIST ARE [INDISCERNIBLE]. LET THAT'S WHERE ISM FROM. >> WELL WE'RE LOOKING IF ORDER TO YOUR PAPER. >> OKAY, THANK YOU. [ APPLAUSE ] SO MAYBE A COUPLE MORE WORDS WE HAD MARTIN TOGETHER WITH FRANCESCA PRESENTING ON SOMETHING BASED ON OPEN SOURCE FRAMEWORK TRANSSPECT AND AND THIS TALK HAS SOMETHING TO DO WITH TRANSPECK AND CONCEPT OF HTHL REPORT BUT ALSO IS USING SOMETHING ELSE THAT WENDELL, IT'S OXYGEN FRAMEWORK FOR JATS. AND WHAT I DID IS ADOPT A GENERATED IRBIC JATS KIT FRAMEWORK FOR SPECIFIC JOURNAL QUALITY TRACKING REQUIREMENTS AND ALSO FOR SUBMISSION PACKAGE BUILDING FOR THE [INDISCERNIBLE] PLATFORM. AND WHAT I--THIS IS PART 1 SO ADOPTING AN OXYGEN FRAMEWORK FOR PUBLISHER SPECIFIC NEEDS. ADDING SOME 150 SCHEMATRON RULES, PARTLY FOR SATISFYING [INDISCERNIBLE] PACKAGES REQUIREMENTS PARTLY SATISFYING INTERNAL CONVENTIONS SUCH AS TRACKING ISSN, ET CETERA AGAINST JOURNAL NAMES. AND THE SECOND PART WILL BE ABOUT TAKING WHAT'S IN THAT FRAMEWORK WHILE--SO TAKING ALL THE GOODNESS OFF THE XFL ASPECT INCLUDING [INDISCERNIBLE] EXPECTATION AND IT OF COURSE, AND RUNNING THAT'S ON A DIFFERENT PLATFORM, NOT WITHIN OXYGEN BUT ON A WEB PLATFORM. DOESN'T NECESSARILY HAVE TO BE ON THE WEB BUT ON A DIFFERENT PLATFORM WHICH ARE TRANSSPECT, WHICH ARE XPROC PIPELINES AND THE IDEA IS TO ESCAPE A BIT OF THE VENDOR LOOK THAT IS STILL THERE WHEN YOU ARE PROVIDING SUCH A QUALITY TRACKING TOOL ONLY FOR OXYGEN SO WHAT [INDISCERNIBLE] REQUIRES FROM THAT TYPE SET IS THAT THEY RUN THE SCHEMATRON TECHS AND USE THE BUILDING TYPES THAT DIDN'T USE OXYGEN AND WE--IN THE FIRST PLACE WE TRY TO GIVE THEM THE SCHEMATRON IN FACT THAT'S ALREADY OPEN SOURCE. I WILL GIVE YOU THE LINKS LATER SO YOU CAN ADOPT ALL THIS FRAMEWORK FOR YOURSELF: BUT THE TYPE SETTERS HAD SOME DIFFICULTIES IN MAKING THE SCHEMATRON RUN AT THEM, SO WE THOUGHT IT MIGHT BE A GOOD IDEA TO PROVIDE THIS AS A WEB SERVICE, THEY JUST UPLOAD THE CONTENT AND GET VALIDATION RESULTS AND ENHANCE PACKAGE FOR [INDISCERNIBLE]: SO THE CUSTOMER THAT WE DID THAT FOR WAS HOGREFE, WE HAD A PRESENTATION 2 YEARS AGO THAT WORK FLOW THAT WAS IMELEMENTED FOR HOGREFE, AND MAYBE RECENTLY, 2 OR 3 YEARS AGO, METAPRESS WAS DELIVERING JOURNAL CONAT THE PRESENT TIME THROUGH METAPRESS AND OF COURSE THEY THEN HAD TO SWITCH. AND IN THE COURSE OF THESE CHANGES, THEY THOUGHT, OKAY, WE ARE STICKING WITH A REALLY ANTIQUE APA DTD WHICH IS KIND OF DERIVED FROM AND ALSO QUITE ANTIQUE NLM DTD AND THEY THOUGHT IT MIGHT BE GOOD IDEA TO MOVE TO JATS NOW: IN FACT, THEY HAD TO BECAUSE ATYPON REQUIRES JATS OR THEIR OWN DERRIFFATIVE. MENT AND IN THE COURSE OF THESE CHANGES THEY WANTED TO INVEST IN THEIR DOCUMENTATION, RECENT DOCUMENTATION AND MORE AUTOMATED QUALITY TRACKING PROCESSES. SO THIS IS WHAT THE TYPE SETTERS DID BEFORE. THEY CREATED THE ARTICLE XML, I DON'T KNOW EXACTLY HOW THEY DO IT. THEY CERTAINLY TYPE SET THINGS WITH THIS DESIGN OR 3 B 2 OR WHATEVER. AND I DON'T KNOW WHICH KIND OF XML WORK FLOW THAT IS. IN THE END, THEY HAVE XML. AND THEY MUST BUNDLE IT TO PACKAGES. AND PREVIOUSLY, THERE WERE ALSO LOTS OF THESE CONSTRAINTS THAT WE NOW HAVE TRANSFERRED TO TO SCHEMATRON ROOTS AND THEY'VE BEEN WRITTEN IN WORD FILES ESSENTIALLY AND THERE WAS A BIT OF EXCESS IT FOR QUALITY MATRIX BUT NOT THAT MUCH SO IT WAS A RATHER MANUEL PROCESS THAT WAS REQUIRED FOR SIGNIFICANT RESOURCES ON BEHALF OF THE PUBLISHER AND THEY WANTED TO AUTOMATE THIS. LET THE NEW WORK FLOW LOOKS LIKE THIS: THE TYPE SETTER CREATE NOW JUST XML AND NOT THE APA CONTENT ITEM XML. AND THEY THEY CAN NOW CREATE THE DELIVERY PACKAGES AUTOMATICALLY, SO THEY ARE LOTS OF CONVENTIONS THAT ARE THAT ATP YON REQUIRES STRUCTURE THIS WILL ALL BE CHECKED AND SOMETIMES ALSO AUTOCORRECTED BY THE PACKAGES SCRIPTS. THIS IS NOW A SINGLE HTML PAGE AND LENGTH BUT IT'S--I WILL SHOW IT IN A MINUTE. THE DERIVED FRAMEWORK THAT WE CREATED CONTAINS ALSO DOCUMENT TEMPLATES FOR ARTICLES FOR THE ACCOMPANYING ISSUE, XML THAT DESCRIBES THE TABLE OF CONTENTS AND WHAT IS IN THE PACKAGE AND SUBMISSION MONEY WHICH IS A TINY XML FILE THAT NEEDS TO BE ON THE ROOT DIRECTORY OF THE PACKAGE THAT IS DELIVERED TO ATYPON, QUALITY CHECKS ARE PERFORMED BY SCHEMATRON RULES AND THE SCHEMATRON RULES YOU CAN USE PROS THERE, YOU CAN LINK TO ADDITIONAL DOCUMENTATION IF IT THERE IS ANY. THESE RULES ALSO KIND OF DOCUMENTATION, SO, THEY COULD ACTUALLY REDUCE THE AMOUNT OF WRITTEN DOCUMENTATION, RELYING ON ISSUES BEING REPORTED AT THE PLACE WHERE THEY OCCURRED AND AS I SAID, EVERYTHING WAS DELIVERED AS AN OXYGEN FRAMEWORK AND THE HTML PAGE LOOKS IN THE BEGINNING, LOOKS LIKE WHAT WENDELLWROTE FOR AN OLDER VERSION OF HIS FRAMEWORK THAT I SHAMELESSLY [INDISCERNIBLE] ON GITHUB. I DON'T THINK THAT WENDELLWILL MIND THAT SO BASICALLY IT DESCRIBES HOW THEY INSTALL THAT PACKAGE AND THEN ONCE IT IS INSTALLED, WHAT DOCUMENT TEMPLATE IT PROVIDES THE ARTICLE ISSUE AND MONEY FIRST AS I SAID. IF THEY CREATE SUCH A TEMPLATE ARTICLE, IT LOOKS LIKE THAT AND THEN IMMEDIATELY SCREAMATRON WILL COME IN AND REPORT ERRORS BECAUSE IT'S FULL OF PLACE HOLDERS, THE TEMPLATES THEMSELVES ARE NOT VALID IN THAT SENSE, THEY NEED TO BE FILLED IN. AND THIS IS ALSO A THING THAT I WILL DEMONSTRATE LATER. THIS IS ALSO AN ISSUE OF XML THAT IS LACKING THE TABLE OF CONTENTS. SO ON THE ATYPONPLATFORM, THEY ARE ABLE TO GENERATE A TABLE OF CONTENTS FROM THE UPLOADED CONTENTS AND THE ARTICLE SEQUENCE. BUT YOU IF YOU WANT TO HAVE A KIND OF FANCIER TABLE OF CONTENTS WITH CATEGORY OF HIRE ARCHIC STRUCTURE THEN YOU HAVE TO CREATE IT YOURSELF AND IF YOU DO CREATE IT YOURSELF, YOU HAVE TO HAVE SOME CONVENTIONS FOR THE HIERARCHICALLY STRUCTURED TABLE OF CONTENTS AND THIS IS THE THING THAT THE PIPELINE ALSO DOES PRIOR TO PACKAGING THE WHOLE THING AS A ZIP FILE IN THE BRIDGE. AND THEN THERE ARE INSTRUCTIONS ON HOW TO NAME THE DIRECTORIES AND HOW TO NAME THE FILES WITH THEM AND AND WHAT THE OVERALL DIRECTORY LAY OUT IS AND IN THE END, THEY WILL APPLY THIS TRANSFORMATION SCENARIO, AND MOST OF YOU WHO ARE USING OXYGEN ARE PROBABLY FAMILIAR WITH THE XIT SKEANARIATIONS, THIS IS AN XIT SCENARIO THAT DOES HOTS MORE THAN EXXIT CAN DO. IT DEMONSTRATES THE WHOLE PROCESS OF ASSEMBLING AND CHECKING AND FIXING THE PACKAGE. SO, AND I'M GOING TO LIVE DEMO THIS NOW, SO WISH ME LUCK. [LAUGHTER] FIRST I HAVE TO IDENTIFY--SO FIRST WE HAVE A STRUCTURE TO WHAT WE HAVE THERE IS YOU HAVE ARTICLE DIRECTORIES FOR THAT ISSUE DELIVERY. YOU CAN LOOK INTO THESE ARTICLES AND YOU INSTANTLY GET SCHEMATRON MESSAGES AND THESE ARE ACTUAL ARTICLE DELIVERIES THAT WHOEVER RECEIVED FROM THERE, FROM THE TYPE SETTING AND MAYBE ALSO DELIVERED TO ATYPONIN THAT FORM, SO THIS SCHEMATRON RULE IS ENFORCED SINCE A COUPLE OF WEEKS AND THE PACKAGE IS FROM DECEMBER, SO MAYBE IT JUST WASN'T CHECKED THEN AND THEY SUBMITTED IT LIKE THAT AND THIS IS OF COURSE A PROCESS OF CONTINUOUSLY IDENTIFYING NOW SOURCES OF ERRORS OR NEW SOURCES OF RENDERING ERRORS ON [INDISCERNIBLE] PART AND TRACING IT BACK TO ERRORS IN THE SOURCE THAT ARE TO CONTINUOUSLY IMPROVE THE CHECKING RULES. AND WHAT THE TYPE SETTERS DID HERE IS APPARENTLY HAVE AN MT--YEAH, HAVE AN MTXREF IN ANOTHER XREF THAT HAS CONTENT AND THIS SEEMS TO BE SICKLICAL, I DON'T KNOW. IT HAPPENS IN MULTIPLE ARTICLES HERE. THEY IDENTIFIED THAT I THINK IT WAS A HINT FROM ATYPONSMAZE THAT MOST ELEMENTS IN THEIR USE OF XREF AREN'T EMPTY AND WE COULD EASILY REPORT ALL MT REPORTS FROM GRAPHIC AND NONGRAPHIC MHTD INSED OF [INDISCERNIBLE] AND WE ARE DOING THIS IN PRINCIPLE AND IN JATS WE CAN ALSO USE MT AND THE REND--RENDERING IS THAT THEY WILL BE GENERATED WITH CONTENT. BUT THE REQUIREMENT HERE IS THAT THERE IS NO GENERATED CONTENT AND AS I'M LOOKING AT THIS, PROBABLY NOT AN ARTICLE BY JEFF BECK BECKETAL, AND WE DON'T KNOW THAT BECAUSE THERE'S NO ALL ATTACHED TO THAT. [LAUGHTER] , YEAH, MAYBE AT A LATER STAGE WE CAN HAVE A SCHEMATRON RULE THAT WOULDED--WOULD REQUIRE THAT AT LEAST ALL OF US THAT ARE NOT EQUIPPED WITH AN ORCHID AS A SIDE NOTE. SO THIS IS WHAT IS REPORTED HERE AND HERE'S AN ARTICLE THAT I DELIBERATELY INVALIDATED. SO WOO HAVE SOME SCHEMATRON RULES HERE SO THIS USED TO BE A 10 SO WE BOTH CHECK ALSO FOR MONTHS. THEY WANT TO HAVE THAT IN DECIMAL FORMAT AND THEY WANT TO HAVE ALWAYS A STRING DATE TOGETHER WITH THE DATE AND--AND THEY CHECK AGAINST A LIST OF PERMISSIBLE PUBLICATION TYPES AND REFERENCES SO WE CAN FIX THAT. SO THESE ARE JUST EXAMPLES FOR THESE CHECKS AND THEN I DTD VALIDATE THAT SO WE HAVE THIS STANDARD DTD VALIDATION HERE. AND THEN WE GET BACK TO THESE DTD ERRORS LATER BECAUSE THEY CONSITUTE A PROBLEM WHEN YOU CREATE PURE HTML REPORTS BUT WE'LL LOOK AT THAT DATA. SO THIS IS THE DOCUMENTATION THAT IS THERE. CERTAINLY IT CAN BE--CAN BE ENHANCED. SO IF THE--THIS IS THE THING THAT DOROTHY MENTIONED YESTERDAY. SO WHEN THEY HAVE A LIST OF JOURNALS MAINTAINED AS AN HTML FILE. SO THIS WILL BE ACTUALLY--THIS WILL BE--THIS HTML FILE WITH ALL THEIR JOURNALS, AND THERE ARE JOURNAL IDs, THIS WILL BE SLOPED IN BY THE SCHEMATRON TRACK AND THEN IT WILL MAKE SOME USING XXITs AND ALL KINDS OF ADVANCED STUFF WHICH SORRY THE SIDE NOTE WAS THE CAUSE THAT THE EXTERNAL TYPE SETTERS COULDN'T MAKE THE SCHEMATRON RUN. SO THIS--THERE'S VALIDATION OF THEIR ISSN AGAINST THAT LIST. SO IF I CHANGE THAT HERE, THEN A NEW MESSAGE WILL BE CREATED THAT TELLS ME I NEED TO CHANGE THAT BACK. IT DERIVES THAT BACK FROM THE ABBREVIATION JOURNAL AND WHEN I CHANGE THAT IT WILL TELL ME THAT JOURNAL WITH THAT ABBREVIATION DOESN'T EXIST. ALSO THE FULLY EXPANDED NAME WILL BE--SO, AND THIS WAS A FREQUENT CAUSE OF TROUBLE AND THEY HAD TO CHECK EVERYTHING MANUELLY. SO THESE ARE ABOUT JUST 150 OF THE SCHEMATRON RULES THAT APLOY HERE. ALSO CHECKING THESE IN CONTRAST GROUPS OR MIXED CITATIONS THE IN BETWEEN CONTENT OR XIT ELEMENTS THEY ARE IN BETWEEN CITATIONS OR IN CITATIONS HAVE YOU A LIST--IF YOU HAVE A LIST OF 6 OR MORE AUTHORS, THEN THERE SHOULD BE AN ELIPSIS SIGN AFTER THEM, AFTER THE FIFTH OR SO, AND WE CAN MAKE THESE KIND OF DETAILS, CHECKING RULES ALSO USING INFORMATION SUCH AS THE ARTICLE LANGUAGE SO THAT WE CAN HAVE DIFFERENT RULES FOR GERMAN, FRENCH, ENGLISH ARTICLES OR IF THERE'S INTERNATIONAL PSYCHOLOGY PUBLISHER, WITH OFFICES IN MANY COUNTRIES AND THEY ARE IN SWEDEN AND SWITZERLAND AND 3 LANGUAGES ALONE AND IN THE U.S. ALSO. SO THERE'S A COUPLE OF ALSO LANGUAGE SPECIFIC CONVENTIONS THAT WILL BE ENFORCED HERE. YEAH. THAT'S WHAT I DID IN THIS THE NEXT DEMO AND THE NEXT DEMO WILL BE USING THE XPROC FROM OXYGEN AND IN ORDER TO DO THAT, I WILL OPEN THE MANIFEST AND THEN SELECT A TRANSFORMATION SCENARIO, ISSUE SCENARIO AND IT WILL RUN. SO X PROC WILL ASSEMBLE THINGS AND AGAIN, SCHEMATRON VALIDATE THINGS BECAUSE YOU CAN RELY ON THE TYPE SETTERS TO HAVE CHECKED EVERYTHING SO IT WILL RECHECK ALL THE CONTENT AND THEN IT WILL COAT A SIMPLE HTML LIST THAT CORRELATES ALL THE WARNINGS AND ERRORS OF THESE ARTICLES GIVING ALSO THE EXPATH LOCATION SO THEY COULD GO BACK TO THAT ARTICLE AND FIX IT. GO IMMEDIATELY TO THAT LOCATION. BUT IN ADDITION IT WILL ALSO CHECK OTHER THINGS REGARDING THE FULL PACKAGE, NAMING CONVENTIONS AND IT WILL GIVE INFORMATIVE INFORMATION WHICH KIND OF ADDITIONAL PAGES ARE THERE WITH ADVERTSIZEMENTS AND THINGS LIKE THAT. -- ADVERTISEMENTS AND THINGS LIKE THAT. SO HERE'S ANOTHER THING. I MENTION THAD I INVALIDATED 1 OF THESE XML FILES AND I WON'T RECEIVE ANY SCHEMATRON CHECKS FOR THAT BECAUSE IT JUST FAILS TO BE PATHS BECAUSE IN THE EXTRA PIPELINE I STIPULATED DTD VALIDATE TRUE. SO IF THE--IF THE ARTICLE IS DTD INVALID THEN IT WON'T BE READ AT ALL. THE IDEA IS THAT IT'S EASY FOR USERS OF OXYGEN TO GO BACK TO THE ARTICLE AND FIX THESE ISSUES. THEY WILL BE REPORTED IN PLACE. BUT WE'LL SEE LATER THAT THIS IS A BIT PROBLEMATIC FOR USERS THAT DON'T HAVE OXYGEN. THE. YEAH, I ALREADY MENTIONED THE FANC [INDISCERNIBLE] IN THE FILE SYSTEM THAT'S BEEN CREATED HERE, IT'S A LINK SO OXYGEN OPENS THIS HTML SLIDE SO THIS IS THAT. AND IT CONTAINS A LINK TO THAT ZIP FILE AND IF I--IF I OPEN THAT LET ME FIRST GO BACK TOTE ISSUE FILE. IT'S AUTHORED SO THIS ENDS WITH ISSUE-META, THIS WOULD BE THE TABLE OF CONTENTS AND THE TYPE SETTERS DON'T HAVE TO MANUELLY ADD A TABLE OF CONTENTS WITH THEY HAD TO DO BEFORE BECAUSE IT CAN BE AUTOGENERATED AND NOW I'M HERE IN THE ZIP FILE. IT HAS BASICALLY THE SAME DIRECTORY STRUCTURE SPORTSMANSHIP IF I LOOK AT THE IMPROVED ISSUE XML, THEN WE SEE IT CONTAINS A TABLE OF CONTENTS AND IT'S A GROUP 1 THAT ATYPON'S PLATFORM WON'T CREATE BY ITSELF BUT IF YOU WANT TO HAVE A MORE FANCY TABLE OF CONTENTS YOU HAVE TO CREATE IT YOURSELF OR HAVE YOUR XPROC CREATE IT AND THIS IS WHAT HAPPENED HERE. AND ANOTHER SO TO SAY AUTOCORRECTION IS--AND FOR THE TIME BEING I WILL CLOSE THE GROUP FILE AND RETARP TO THAT ARTICLE. THERE'S ANOTHER REQUIREMENT FOR ATYPONPACKAGES THAT OTHER GRAPHIC OR OTHER MEDIA REFERENCED IN THESE ARTICLES RESIDE IN DIRECTORIES OF THEIR OWN. SO, THE WAY THE TYPE SET ARE CREATED HERE IS NOT LEGAL FOR PACKAGE DELIVERY BECAUSE THE ARTICLES ARE ON THE SAME LEVEL: IT SHOULD BE MORE LIKE IN THIS ARTICLE THAT THERE'S A GRAPHIC FOLDER BUT OF COURSE IT'S CONVENIENT FOR AUTHORS BECAUSE THE LINKS TO THE GRAPHIC FILES THEY DON'T HAVE A GRAPHIC SUBDIRECTLY. THEY JUST LINK TO GRAPHICS WHETHER WE HAVE THEM, YEAH, HERE THIS IS A GRAPHIC, THIS IS AN ALTERNATIVE TABLE RENDERING SO IT'S INCLUDED AS A GRAPHIC WITHOUT A PATH BUT IN THE FINAL DELIVERY IT HAS TO BE IN THE GRAPHIC FOLDER AND IF THE TYPE SETTERS CREATED LIKE THAT, THEN IT WOULD BE A BROKEN IMAGE AND FOR CONVENIENCE OF THE TYPE SETTERS, WE SAID OKAY, WE WILL AUTOCORRECT THIS WHEN CREATING THE PACKAGE AND JUST TO LET YOU KNOW THAT IT HAS BEEN AUTOCORRECTED. THIS IS THE 50 ARTICLE. IT NOW AS A GRAPHIC SUBJECTRY WITH THESE ALTERNATIVE TABLE RENDERINGS. SO THIS IS NOT THE MOST COMPLEX, OR XPROC APPLICATION ON EARTH BUT STILL CAN BRING SOME CONVENIENCE AND QUALITY ASSURANCE WITH IT. OKAY, SO THIS IS THIS AND OKAY WHAT I DIDN'T MENTION WAS, UNDO, CLOSE TAB, THERE'S ANOTHER CONVENTION FOR NAMING THE FINAL ZIP FILE. SO IF WE DIDN'T HAVE A ZIP FILE INSHALLLY, SO IT WAS JUST NAMED AFTER CONVENTIONS THE EXTRA PIPELINE WILL GO LOOK INTO THE ISSUE XML AND MAYBE ALSO IN THE SUBMISSION MONEY FEST AND THEN GET A CURRENT TIME STAMP. SO THIS IS FROM 7 MINUTES AGO. SO IT--JUST ADDS THE TIME STAMP TO THE NAME. YEAH. SO SOME OF THE TYPE SETTERS DIDN'T WANT TO USE OXYGEN. THEY WANT TO STICK WITH THEIR OWN SOLUTIONS, BUT STILL FOGREFEWONDERED IF THEY COULD BE APPLIED TO THE SLIDES THAT THEY DELIVERED. SO WE THOUGHT ABOUT CREATING A WEB SERVICE THAT WILL ACCEPT THE RAW PACKAGES SO WHAT THEY EDITED ON THE FILE SYSTEM SUBMITTED AS A ZIP FILE. AND THEN RUN THE SAME CHECKS ON THAT. AND FINALLY RETURN THE REPORT AND ENHANCED PACKAGE. AND OPTIMALLY, THIS WEB SERVICE SHOULD PROVIDE WATCHED FOLDERS AND ALSO IN HTTP, API INTEGRATION AND CONTENT MANAGEMENT. SO, YEAH, I MEAN THE EASIEST APPROACH WOULD BE THE HTML REPORT WHAT IS LACKING IS BASICALLY THE--I HAVE ANOTHER LOOK AT THAT. BASICALLY THE INDEPTH VALIDATIONS OF THAT FILE, SO THEY WOULD UPLOAD IT, RECEIVE THIS ERROR MESSAGE AND THEN WOULD HAVE TO FIX THAT AND REUPLOAD IT AGAIN AND THEN THEY WOULD WOULD DISCOVER ALL KINDS OF NEW ERRORS IN THE FILE THAT INITIALLY WASN'T PASSED. SO WE SAID IT WOULD BE DESIRABLE TO ALSO BE--YOU KNOW BE ABLE TO DISMRI THESE VALIDATION MESSAGES HERE BUT THIS IS KIND OF DIFFICULT WITH DTD VALIDATION AND XPROC, SO YOU CAN'T GET A A LIST OF ERRORS AND MAYBE DISPLAY IT. ALSO, IT WAS DIFFICULT WITH OTHER KINDS OF VALIDATIONS WITH RELAXING. YOU WOULD WOULD ONLY GET AN ERO THAT YOU COULD CATCH WITH SOME XPROC CATCH STATEMENTS AND AT THAT STAGE WE PATCHED [INDISCERNIBLE] THE RELAXING VALID DATA TO NOT ONLY CREATE, WELL ERROR MESSAGES AS ERROR MESSAGES BUT TO CREATE MESSAGES THAT CONTAIN THE X-PATH OF AN ERROR AND THE CORRESPONDING MESSAGE AND WE ARE USING THIS PATCHING FOR OUR HTML REPORT IN ORDER TO INSERT THE VALIDATION ERA AT THE LOCATION WHERE IT HAPPENED WITH AN HTML RENDERING OF THE CONTENT. SO WE THOUGHT WE WOULD TRY THE REPORTS BECAUSE WE THOUGHT IT WOULD BE BETTER SCHEMA TRANSPORTS AND IT LOOKS BETTER THAN THESE HTHM REPORTS AND I WILL SHOW YOU HOW THAT'S--HOW THAT WORKS. SO I WILL FIRST CREATE A ZIP FILE OF MY RAW CONTENT. SO THIS IS NOT THE GENERATED ZIP FILE. I WILL GO TO THE--TO THE LEVEL WHERE THE MONEY FEST RE--MANIFEST RESIDES. AND UPLOAD IT. AND WILE THIS PIPELINE WORKS IT TAKES A BIT LONGER BECAUSE IT HAS TO GENERATE HTML RENDERING OF EACH ARTICLE FIRST, I CAN TELL YOU A BIT MORE ABOUT THESE HTML REPORTS BECAUSE FOR THIS APPLICATION, IT'S A BIT OF AN OVERKILL TO USE THESE. THE PRIMARY USE FOR THESE HTML REPORT SYSTEM TO COLLECT AREAS OF A PIPELINE OF MULTIPLE CONVERSION STEPS AS I PRESENTED LAST YEAR IN NOVEMBER IN AMSTERDAM. SO THIS IS AN EXAMPLE GOING FROM IDML TO A DIVERSE IMMEDIATE PERFORMANCE AND AFTER THAT YOU HAVE SCHEMATRON TRACKS FOR EXAMPLE, IN THE END DESIGN, YOU CAN CHECK FOR UNANCHORED FRAMES IN DESIGN AND ALL KINDS OF OTHER THINGS AT CERTAIN INTERMEDIATE STEPS ALSO MISSING METADATA IF IT YOU HAVE METADATA FROM SOME SOURCE AND ALSO IDPS, PUB CHECK RULES OR SCHEMATRON TOOLS, THAT CAN YOU RUN ON YOUR EPUB AND SHOULD THE PRIMARY PURPOSE FOR THESE HTML REPORTS TO COLLECT MULTIPLE ERRORS FROM MULTIPLE CONVERSION STEPS AND TO PRESENT THEM IN A SINGLE HTML RENDERING AND AN EXAMPLE OF THIS WOULD LOOK LIKE THAT. THIS IS NOT--NOT--THE ISSUE SEEMS TO BE THAT THE SCREEN HAS BEEN KIND OF ON THE RIGHT HAND SIDE. SO I CAN EASILY SCROLL HERE. BUT YOU SEE, KIND OF GERMAN BUT YOU YOU SEE THESE MESSAGES IT'S CALLED PSEUDO, ELLIPSE SIS. SO HERE THEY USE THE 3 DOTS INSTEAD OF THE HEX 2026 ELIPSIS SIGN AND THEN THEY WILL GET THESE ERRORS AND A COUPLE TYPOGRAPHY ERRORS. AND ERRORS ON ILLEGALLY USED STYLE NAMES SO THIS IS A LEX AN VALIDATION ERROR. XML THAT'S GENERATED HERE IS T. I. AND EVERYTHING ELSE WILL BE RENDERED INTO A SINGLE HTML FILE. YEAH, SO THIS IS JUST--THE WHOLE CONCEPT OF THE HTML REPORTS IS A BIT LARGER THAN WHAT WE USE HERE. BUT GOING BACK TO THAT CHECK, IT HAS FINISHED AND WE CAN--SO WE HAVE HERE THE SAME OUTPUT. IT'S AGAIN, A TIME STAMP. IT'S--IT'S CENTRAL EUROPEAN TIME FOR TIME STAMP. SO THERE ARE SERVERS IN GERMANY. AND WE HAVE HERE THE REPORT THAT WILL GIVE YOU AN HTML RENDERING OF THE WHOLE CONTENT OF THE METADATA HERE AND WAS KIND OF A BIT OF A PITY THAT THE RIGHT HAND COLUMN IS CUT OFF. BUT YOU SEE HERE FOR THE ARTICLE THAT I VALIDATED, THE RELAXING MESSAGE AND YOU ALSO SEE HERE THE SCHEMATRON MESSAGESS FOR EXAMPLE, [INDISCERNIBLE] NOW REQUIRES 4.2 MANIFEST AND SUPPLY IS STILL 4.1. SO YOU SEE THAT ALL THESE ERRORS THAT HAVE BEEN PRESENT IN OXYGEN ALSO CATEGORIZED AS ERRORS AND WARNINGS ARE PRESENT HERE IN THAT REPORT. AND THAT'S ESSENTIALLY WHAT WE DID. SO TAKE ALL THE GOODNESS OF THE XML STACK AND JUST PORT IT TO ANOTHER RUN TIME AND I THINK THIS IS REALLY THE GOOD THING THAT THAT YOU HAVE THIS. IT'S ALSO IN GETHUB PROJECT, THERE'S ALSO NODULE, NOT MUCH CODE FOR THAT WEB CHECKING APPLICATION. IT'S MOSTLY REUSE OF OTHER MODULES AND 1 OF THESE MODULES IS THE FRAMEWORK THAT RUNS IN OXYGEN AND THAT I FORM FROM [INDISCERNIBLE]. SO I WANT TO THANK YOU VERY MUCH. [ APPLAUSE ] >> JUST A QUICK QUESTION, FOR YOUR WHOLE GRAPHIC RULES FOR [INDISCERNIBLE] RATHER THAN ATYPON? >> I THINK I GAVE ESTIMATE THAT THESE 155 RULES IN TOTAL TO BE 60% FOR ATYPON[INDISCERNIBLE] NAMING CONVENTIONS, ET CETERA AND 40% FOR HOGREFE'S CONTENT RULES AND IT WOULD CERTAINLY BE A GOOD IDEA TO SEPARATE THAT. SO IF ANOTHER CUSTOMER THAT USES YOUR PLATFORM WILL DEPLOY THEIR OWN SCHEMATRON RULES SO THEY COULD JUST CUT OFF HOGREFE'S RULES AND REPLACE THEM AS THEIR OWN. [ APPLAUSE ] >> IT IS MY PLEASURE TO INTRODUCE WENDELLPIEZ, AND DESPITE HIS YOUNG AGE, HE'S BEEN AROUND FOR A WHILE AND HE'S WORK WIDE ME AND HE GRADUATED AND HAS HIS OWN CONSULTING SERVICE. DO I TELL THEM ABOUT THAT OR DO YOU? >> NO. >> PIEZ, CONSULTING CHECK HIM OUT. ANYWAY, HANDS TOGETHER FOR WENDELL, PLEASE. [ APPLAUSE ] >> THANK YOU DEBBIE, THANK YOU GARRETT FOR THE WATER, THANK YOU GARRETT FOR THE TALK THIS ACTUALLY IS NOT BACKWARDS. OH I'M GOING AFTER GARRETT, GARRETT--OH, OF COURSE, YOU KNOW IN MANY WAYS I FELT ANXIOUS ABOUT THAT BECAUSE GARRETT HAS BUILT ON MY WORK AND DONE SOMETHING THAT I COULD NOT DO SO IN THAT SENSE I'M SORT OF LIKE THIS WILL BE COMPLETE DEFLATION? AND YET AT THE SAME TIME I THINK IT'S ACTUALLY THE RIGHT WAY TO PRESENT THIS BECAUSE WHAT YOU JUST SAW WAS THE DEVELOPMENT OF AN EXTREMELY SOPHISTICATED SET OF TOOLS TO DO A REAL DEAL WITH A REAL WORLD SET OF PROBLEMS THAT'S QUITE COMPLEX AND NEEDS THE ATTENTION OF PROFESSIONALS WHO REALLY KNOW WHAT THEY'RE DOING BUT AT THE SAME TIME, GARRETT'S FIRST WAY OF TELLING THAT HE DIDN'T DO IT ALONE BECAUSE HE FORKED MY EARLIER PROJECT THAT I'M TRYING TO UPDATE NOW AND GET STARTED AND I THINK THAT'S FANTASTIC. THAT'S EXACTLY THE MODEL THAT I'M HOPING TO PROMOTE BY DOING THIS WORK AND PUTTING IT UP ON GITHUB, THE IDEA IS THAT YOU SHOULD FORK THIS. IT'S NOT COMPLETE. IN FACT IT'S INCOMPLETE BY DESIGN. THE INTENTION IS TO DO THE 80% EASY PART AND LEAVE THE 20% HARD PART FOR THE PEOPLE WHO ARE BETTER ABLE TO DEAL WITH IT BECAUSE INFORMATION IS FRACTAL. RIGHT? WE HAVE EDGES THAT ARE FUZZY AND TO GET--YOU ZOOM IN AND THEN YOU SEE MORE DETAIL AND YOU ZOOM IN AND SEE MORE DETAIL AND THOSE EDGES ARE WELL ALL THE ACTION HAPPENS AND THAT'S WHAT MAKE THIS IS WORK INTERESTING AND DIFFICULT. AND IN ORDER TO BUILD SUCCESSFUL SYSTEMS WE HAVE TO BE AWARE OF THAT, BUT AT THE SAME TIME IT'S REALLY, REALLY HELPFUL IF SOMEBODY HAS ALREADY LAID A FOUNDATION FOR US AND AS DEBBIE JUST REMINDED YOU, I WAS VERY FORTUNATE AND PRIVATE SECTOR LEMMINGED TO WORK WITH [INDISCERNIBLE] AT THE FIRST FEW YEARS AND AT THE BEGINNING OF JATS DEVELOPMENT SO I'M HOPING TO TAKE ADVANTAGE OF SOME OF THE INSIGHT I DID THERE AND DO SOMETHING I CAN DO WHICH THE EASY PART. AND THEREBY GIVE YOU FOUNDATION ON WHICH YOU CAN BUILD THE KIND OF THING THAT GARRETT WAS JUST SHOWING. SO TO GIVE YOU JUST A BIT OF BACKGROUND OF THIS, THE ORIGINAL OXYGEN FRAMEWORK FOR JATS IS A PROJECT I PUT UP IN 2012. IT WAS VERY KIND OF QUICK AND DIRTY ASSEMBLAGE OF THINGS INCLUDING THE PREVIEW XXLTs THAT ARE AVAILABLE FROM NLM, SOME CSSTHAT I HAD DEVELOPED AT MULBERRY WHICH THEY SAID, OH, GO FOR IT, TO HANDLE JATS DATA SHARINGA AND SO FORTH AND SO ON, WHAT YOU'RE LOOKING AT HERE IS OF COURSE THE NEW FRAMEWORK AND THE FACT THAT IT LOOKS EXACTLY LIKE WHAT GARRETT JUST SHOWED YOU IS ACTUALLY AN INDICATION OF THE ARCHITECTURE I'M DESCRIBING BECAUSE MANY OF THESE COMPONENTS GO BACK QUITE A LONG WAY, RIGHT? AND IF IT'S SOMEWHAT UGLY AND NOT ENTIRELY POLISHED THAT GOES ALONG WITH THE THEME HERE BECAUSE POLISHING IT ISN'T REALLY MY JOB. I DON'T KNOW HOW TO POLISH THINGS THAT WELL, CERTAINLY NOT TO YOUR SET OF STANDARDS SO THAT'S NOT SOMETHING I'M GOING TO ATTEMPT. BUT WHAT I CAN DO IS PUT THINGS TOGETHER AND KIND OF EXPOSE AND PUT TOGETHER SOMETHING DHS IS PERHAPS GOOD ENOUGH TO IMPROVE, AND THAT'S REALLY THE GOAL HERE. MAKE SOMETHING GOOD ENOUGH TO IMPROVE, THE FACT THAT GARRETT HAS DONE THAT IS XREAMLY REASSURING TO ME BECAUSE I KNOW IT CAN BE DONE AT LEAST ONCE. THEN OF COURSE THE NEXT STEP WILL BE UP TO YOU AS WELL AS UP TO ME. SO THE OLDER VERSION, BASICALLY HAD THE TRANSFORMATIONS THAT WERE PROVIDED BY NLM AND IT HAD CSS, AND SHOULD JUST A COUPLE OF OTHER THINGS BUT OXYGEN HAS COME A LONG WAY IN 4 YEARS AND ADDED MANY, MANY POWERFUL FEATURES AND SO PARTLY IN THE--WITH THE IDEA OF DEMONSTRATING THOSE BECAUSE MANY OF THESE AS YOU'LL SEE ARE THINGS THAT ARE NOT REALLY VERY GENERIC. IT'S WHEN YOU GET INTO THE VERY, VERY SPECIFIC STUFF THAT YOU GET INTERESTING STUFF. SO I'M IN SOME WAYS THE FUTURES YOU'LL SEE HERE ARE MORE DEMONSTRATIONS OR PROOF OF CONCEPT TO KIND OF ALERT YOU TO THE POSSIBILITY OF DOING THINGS SOMEWHAT DIFFERENTLY, YOURSELF IN YOUR OPERATIONS IN A WAY THAT'S APPROPRIATE TO HOW YOU NEED TO PROCESS YOUR DATA. SO THE BASIC IS THAT YOU HAVE OF COURSE THIS STRUCTURED FORMAT THE XML EDITORS, THE KIND OF TOOL THAT MOST OF US HAVE BEEN FAMILIAR WITH FOR SOMETIME, OXYGEN HAS BEEN DOING GOOD WORK IN THIS AREA IN THE LAST FEW YEARS. SO THAT'S ABLE TO POLE IN THE CSSSTYLE SHEET IN THE STYLING, AND THE RECENT VERSIONS OF OXYGEN, CAN YOU MIX AND MATCH STYLE SHEETS SO FOR EXAMPLE, YOU CAN ADD A STYLE SHEET WHICH WILL DROP SOME ENHANCEMENTS IN, SO FOR EXAMPLE, YOU WILL SEE HERE THAT NOW I HAVE A PULL DOWN THAT LETS ME PICK MY LIST TYPES STRAIGHT FROM THE MOUSE AND NOT HAVE TO GO INTO THE ATTRIBUTE IN ORDER TO DO THAT, AND SAME THING WITH THE XREF HERE, IT KNOWS IT'S CROSS REFERENCED WITH ID REFS AND THEREFORE IT'S ABLE TO HANDLE THAT FOR ME AND THESE ARE ALL FEATURES OF OXYGEN THAT SIMPLY PROVIDE A BIT OF SUPPORT. SO WHAT I WILL DO HERE IS BASICALLY A QUICK RUN THROUGH OF THE FEATURE SET THAT I INSTALLED HERE AND THEN,--YEAH, SURE. YOU WANT TO READ THE TEXT, THIS IS ALL ON THE WEB: SO WE HAVE SOME NICE TAGS--OH I'M SORRY. SO WE CAN WORK WITH OUR TAGS, THINGS ARE DRAG AND DROPPABLE, THIS IS ALL NICE STUFF THAT OXYGEN HAS BASICALLY PROVIDED FOR US. ONE OF THE THINGS THAT I BUILT INTO THE NEW FRAMEWORK IS BETTER DOCUMENTATION OUT OF THE TAG LIBRARY SO FOR EXAMPLE, IF YOU SCROLL THE LIST OF ELEMENTS THAT ARE OFFERED THERE, OF COURSE AS OUR OWN VALID ELEMENTS AND YOU GET THE LITTLE POP UPS AND THESE ARE PULLED OUT OF THE TAG LIBRARIES, I HAD A LITTLE BIT OF HELP FROM MULL BURY TO RANKLE THIS DATA SO I DIDN'T HAVE TO SCRAPE OFF THE WEB WHICH WAS REALLY HELPFUL AND AS YOU CAN SEE HERE, YOU CAN DOCUMENTATION FROM THE TAG LIBRARY AND THEN A LINK. SO YOU CAN JUMP STRAIGHT TO THE TAG LIBRARY FOR THAT TAG. MENT LET'S SEE, WHAT ELSE HAVE I GOT HERE? WELL, I WILL HAVE TO COME BACK TO SENSIBILITY. PREVIEWS AND PROOFS. IT WILL GENERATE USING XLT WILL GENERATE HTML. IT WILL GENERATE PDF, BUT THE PDF SUPPORT RELYS ON A COMMERCIAL F. O. FORMATTERS AND THAT'S AN IMPEDIMENT FOR THE TARGET AUDIENCE HERE. BECAUSE, I MEAN OF COURSE, THERE'S MORE THAN 1 TARGET AUDIENCE BUT THE THINKING HERE IS MORE THAN 1 IMPORTANT THAT ARE JATS USERS AND THEY'RE POTENTIAL USERS AND THEY'RE LOOKING INTO IT ANOTHER TARGET IS OBVIOUSLY YOU GUYS, SO YOU'RE NOT GOING TO BE HAPPY WITH THE PDF SUPPORT IN THIS THING. BECAUSE IT'S TOO RUDIMENTARY AND YOU WILL NEED YOUR PDFs TO BE MUCH MORE CAREFULLY DESIGNED TO YOUR OWN REQUIREMENTS AND NEEDS WHETHER IT'S PROOFING OR PUBLISHING OR WHAT HAVE YOU. SO, HAVING JUDGED THAT YOU ARE NOT REALLY VERY WELL SERVED BY THE CURRENT PDF STYLE SHEETS, THE OTHER AUDIENCE WOULD BE CRITICAL IN THAT THEY WOULD NEED AN F. O. EDITOR AND WE'RE DOWN PLAYING THE PDF PRODUCTION HERE. NOW HE CAN COME BACK TO THIS ISSUE BECAUSE I AM INTERESTED IN THE QUESTION ABOUT GENERATING PDFs AND HOW USEFUL THEY ARE FOR DIFFERENT USE CASES SO FEEDBACK ON THAT WHOLE ISSUE IS VERY WELCOME. BUT CERTAINLY WE HAVE XLT. LET'S SEE, WHERE'S MY TRANSFORM BUTTON GO. SO CAN YOU SIMPLY GENERATE A PAGE. >> IT'S VERY SIMILAR TO THE OLD PAGE BUT IT'S GOT--IT'S GOT A FEW IMPROVEMENTS AND/OR DID I MENTION CAN YOU CUSTOMIZE THIS? YOU DON'T HAVE TO OPEN A NEW BROWSER EVERY DARN TIME YOU RUN THAT THING BECAUSE WE HAVE ANOTHER SCENARIO THAT LETS YOU REFRESH IT. WE'VE GOT THE PRETTY STYLING WHICH I SHOWED YOU A BIT ABOUT, OF COURSE, YOU CAN GO BACK TO THE TEXT MODE IF YOU LIKE THAT AND MUCH OF THE FUNCTIONALITY IS STILL AVAILABLE SO HERE FOR EXAMPLE, CAN YOU SEE A TERM IS VALID. I CAN INSERT A TERM AND THE DOCUMENTATION SUPPORT IS THERE. WE ALSO HAVE LET'S SEE IF I CAN FIND A--WE SHOW THE ATTRIBUTES VIEW. IF YOU ROLL YOUR MOUSE OVER AND LET IT SIT THERE FOR A BIT, YOU WILL GET DOCUMENTATION IN OUR ATTRIBUTES. WE DO HAVE SUPPORT FOR CONVERTING OASIS TABLES INTO HTML, SO OUT OF THE BOX THESE COME WITH DTD VERSIONS THAT ARE BOTH VERSIONS OF TABLES. THAT'S NOT BECAUSE WE NEED IT IN THIS FRAMEWORK AS OASIS TABLES BUT WE HAVE CODE AVAILABLE TO DO IT FROM EARLIER PROJECTS. WE CAN BUNDLE THAT IN. WE HAVE THE NEW FEATURES THAT COME WITH THE OXYGEN INTERFACE, YOU SAY A LITTLE BIT OF THAT. AND THEN YOU HAVE SOME REALLY NICE SCHEMATRON. I THINK WHAT I WILL DO HERE IS JUMP OVER INTO A BITS EXAMPLE BECAUSE YOU CAN SEE THIS HAS A BIT MORE FEATURES IN IT THAN IN THE ARTICLE. HERE SEEN A SECOND AGO. SO FOR EXAMPLE, HERE YOU'VE GOT A LITTLE WINDOW THAT ALLOWS YOU TO ISIGN THE ID TO A SECTION OR I'M SORRY TO A BOOK PART. THERE'S EXCELLENT SCHEMATRON SUPPORT AS GARRETT WAS JUST SHOWING YOU IN OXYGEN. SO IF I GO INTO MY VALIDATION SCENARIO ISSUES I AM ABLE TO CONFIGURE IT TO VALIDATE TO ANY OR ALL OF THE BUNDLED VALIDATION CONCERN HERE SO WE HAVE THE DETECTED DTD AND IN ADDITION WE HAVE WITH A BITS DOCUMENT WE HAVE 3 DIFFERENT SCHEMATRONS THAT CAN BE USED. SO FOR EXAMPLE, IF I'M JUST--WELL, LET'S JUST GAR FOR THE USE. , IT'S ABLE TO DETECT A BITS IN SID AND SOME OF YOU HAVE SEEN THE SUPPORT FOR SCHEMATRON 86 FIXES AND THAT'S LIPO OVER THERE, SO LET'S SEE IF I CAN--IT ALLOWS YOU TO MAKE YOUR CORRECTION RIGHT HERE. SO THE FRAMEWORK IS WIRED UP WITH LOGIC TO GENERATE ID FAR THAT SECTION THERE FOR YOU SO YOU CAN PICK EITHER OF 2, CAN YOU HAVE A NUMBERED SECTION, NUMBERS REALLY THROUGH THE DOCUMENT OR CAN YOU NAME IT AFTER ITS TITLE AND OXYGEN WILL ADD THAT. YOU CAN SEE THAT ID HAS APPEARED OVER HERE NOW IN THE ID WINDOW AND YOU KNOW YOU GET THE IDEA. OXYGEN ALSO HAS A NICE NEW FEATURE WHERE YOU CAN REFACTOR AT THE DOCUMENT LEVEL AND IT'S DONE WITH XSLT, IT'S EASY TO SET UP, YOU WRITE YOUR XSLT AND YOU WRITE A DECLARAATIVE XML DOCUMENT THAT TELLS THIS WHAT THIS IS AND WHERE IT GOES AND IT DROP ITS INTO YOUR FRAMEWORK AND THAT'S AVAILABLE HERE. SO FOR EXAMPLE, YOU CAN SEE THESE ARE THE 1S THAT OXYGEN COMES WITH AND THIS IS A USEFUL AND POWERFUL SET OF FEATURES FOR PEOPLE WHO DON'T WANT TO MEASURE TRANSFORMATIONS TO MAKE GLOBAL CHANGES TO THEIR DOCUMENT SO YOU SHOULD CHECK IT OUT ON OXYGEN IF YOU HAVEN'T DONE THAT YET. YOU CAN NAME ATTRIBUTES AND ELEMENTS AND TWEAK VALUES AND SO FORTH STRAIGHT FROM THESE SETS OF SCREENS. SO THE FRAMEWORK I HAVE 2 TRANSFORMATIONS THAT DO NOTHING BUT PROVIDE A SMALL LEVEL OF CONVERSION MPLET THAT'S REALLY INTERESTING. THE SMALLER CONVERSION IS PRIMARILY ASSIGNING A CURRENT FORMAL PUBLIC IDENTIFIER WHAT IS MEAN IS THE JATS 1.1, FPI IN THE CASE OF THIS TRANSFORMATION OR IN THE CASE OF THE SECOND 1, THE BITS 2.0 FPI WILL BE ASSIGNED TO YOUR DOCUMENT SO THAT THE FRAMEWORK WILL TURN ON. UNLIKE THE OTHER VERSION, THIS FRAMEWORK DOES NOT TURN ON WHEN IT SEES ANYTHING THAT RESEMBLES NLM, WHICH WAS A GOOD IDEA BACK THEN BUT PROBABLY NOT NOT A GOOD IDEA NOW, SO THIS FRAMEWORK WILL WAKE UP IN OXYGEN, IF YOU OPEN A DOCUMENT THAT HAS A CURRENT 1.1 DTD NAMED. NOW THAT'S A TOPIC WHICH NEEDS SOME GOING INTO BECAUSE OF COURSE OBVIOUSLY MOST OF US DON'T YET HAVE THAT SO YOU HAVE THE OPTION WILL OF UPGRADING AND SEE IF IT'S VALID WHEN IT ISN'T, YOU GO AND MAKE YOUR CORRECTIONS, ALTERNATIVELY. YOU COULD HAVE A STYLE SHEET THAT MADE THOSE CORRECTIONS FOR YOU AND THAT STYLE SHEET COULD GO RIGHT HERE. IN OTHER WORDS, THIS--THIS LITTLE UTILITY I HAVE HERE TO ADD AN FPI, ALSO COULD BE EXTENDED EDED SO THAT YOU CAN MIGRATE YOUR OLDER JATS INTO THE NEW VARIANT OF JATS OR OF COURSE, YOU MAY BE IN A SITUATION WHERE THAT'S NOT A VIABLE OPTION, YOU WANT TO CONTINUE SUPPORTING YOUR OLD DTDs AND YOUR TOOLS THAT'S NOT DIFFICULT TO DO EITHER BECAUSE A FORM WORK CAN BE CLONED: IT CAN BE EXTENDED. CAN YOU CREATE YOUR OWN EXTENSION WHICH INHERENTS ALL THE LOGIC FROM THE PARENT FRAMEWORK AND WHICH YOU SET IT UP TO SAY, USE MY CATALOG TO POINT TO MY DTDs AND RECOGNIZE THESE DOCUMENTS USING FORMAL PUBLIC IDENTIFIERS OR SOME OTHER SET OF CRITERIA THAT OXYGEN CAN USE TO CREATE THAT BINDING. SO THAT YOU CAN USE THIS WITH BASICALLY ANY DTDTHAT'S JATS LIKE. YOU COULD TRY DTI AND IT WOULDN'T BREAK. SO THEN FINE'S AS A DEMONSTRATION HERE I ALSO HAVE A FACTORY TRANSFORMATION CALLED NORMALIZED IDs AND THIS IS SIMPLY A DEMONSTRATION OF HOW THIS OUGHT TO WORK HERE MPLET I WILL RUN THIS 1 SO CAN YOU SEE IT. I WILL RUN ON THE CURRENT FILE. AS YOU CAN SEE, YOU CAN RUN IT ON BATCHES, YOU CAN PROCESS ENTIRE FOLDERS FULL OF XML THIS WAY, AND THEN I CAN PREVIEW IT AND--IT WANTS ME TO SAVE THINGS, THAT'S OKAY. AND THEN IN MY PREVIEW AND YOU GUYS MIGHT HAVE TROUBLE SEEING THIS AT THIS SIZE BUT WHAT YOU CAN SEE IS THAT ALL OF THE IDs HAVE BEEN RELABELED. IS THAT VISIBLE? LIKE YOU CAN SEE HERE THAT, YOU KNOW THIS BOX XID HAS BEEN RELABELED THIS. NOW OF COURSE THAT RAISES AN INTERESTING QUESTION WHICH IS WHAT KIND OF [INDISCERNIBLE] ARE YOU GOING TO USE FOR ID GENERATION, AND THERE IS LOGIC IN THE TRANSFER THAT I WROTE IT'S ROBUST AND USEFUL. IT'S A COMPROMISE BETWEEN A SEMANTIC IDENTIFIER THAT LETS YOU NAVIGATE THINGS AND SOMETHING THAT WON'T BE TOO COMPLICATED. BUT, YOU KNOW EVERYBODY'S GOING TO HAVE THEIR OWN FLAVOR OF THIS. SO AGAIN IT'S SOMETHING YOU WOULDN'T WANT TO REPLACE OR CUSTOMIZE. I CAN SEE HERE THERE'S AN ISSUE WITH MY ID THAT YOU CAN SEE MY ID IS STARTING WITH AN UNDERSCORER. I BELIEVE THAT'S ACTUALLY LEGAL, RIGHT? DEBBIE IS UNDERSCORE A NAME STORE CHARACTER? I'M SORRY. >> [INDISCERNIBLE] >> IT DIDN'T CAN'T BE AN UNDERSCORE, SO THAT 1 WILL NOT VALIDATE. NOW I CAN TELL YOU WHAT THE SOURCE OF TAKEN--THEY BUG IS, IS THAT THIS DOCUMENT DOESN'T HAVE A AN ID ON IT, SO WE'LL JUST ADD 1. THIS IS A PREVIEW--SO YOU HAVE BASICALLY AS YOU CAN SEE THERE, MANY POSSIBILITIES, AND ALSO MANY QUESTIONS RAISED BECAUSE THIS IS GOOD ENOUGH TO SHOW, BUT NOT GOOD ENOUGH TO USE. I HAVE TO IMPROVE IT YET IT'S ALL THERE SO CAN YOU--YOU CAN ACTUALLY GO TO IT AND USE IT. SO IT'S ALL THERE. LET'S SEE, WHAT ELSE HAVE WE GOT TO SHOW YOU? OH SMART PASTE IS INTERESTING. SO THIS IS GOING TO BE LIKE TOTAL FLYING BY WIRE HERE. WE'RE GOING TO--SORT OF GO TO A RANDOM DOCUMENT IN--YOU KNOW IN WEB SPACE AND WORK. THIS IS FUNCTIONALITY THAT IS PROVIDED BY OXYGEN DEVELOPERS AND HE SAID OH I DON'T KNOW JATS AND YET HE DID A GOOD JOB WITH IT. IT ISN'T PERFECT BUT IT'S ACTUALLY SURPRISINGLY GOOD. SO YOU PASTE THAT OUT OF THAT WEBPAGE. YOU GO BACK IN THE OXYGEN. I'LL DROP IT INTO DOCUMENTS HERE. IT LOOKS LIKE IT'S GOING TO DO IT AS PLAIN TEXT. YEAH, OH ACTUALLY IT'S NOTY, SOPHISTICATED, YOU CAN SEE THAT YOU'RE GETTING A PASTE. IT'S NOT PERFECT BUT IT'S DOING A JATS CONVERSION AND ONCE AGAIN THIS IS SOMETHING THAT'S--IT CAN BE--WE CAN COME BACK TO THIS IF THIS IS A FEATURE--I'M NOT ENTIRELY SURE THIS IS SHOULD GO WE NEED BECAUSE IT'S NOT CREATE INDEED AN XML EDITOR BUT IT IS, CERTAINLY LET ME KNOW AND WE CAN GIVE MORE TAPINGS TO THIS AREA. ONCE AGAIN IT'S A SET OF QUESTIONS ABOUT WELL, HOW DO YOU--HOW DO YOU DEFINE THE SCOPE OF THE PROBLEM FIRST AND YET AGAIN, THIS IS XLT UNDER THE HOOD SO IF WE CAN DEFINE OUR REQUIREMENTS WE CAN WRITE THE CODE TO DO WHAT NEEDS TO BE DONE SO THAT'S A FUNCTIONALITY FOR YOU. I'VE BEEN SHOWING YOU NEW FEATURES IN OXYGEN, AND THEN FINALLY ON THE BIT SIDE, PARTLY BECAUSE I WANTED TO FIGURE OUT HOW THIS WORKED, I FIGURED OUT XPROC IS A REALLY REALLY DEFINED SET OF TOOLS SO IF YOU HAVE NOT LOOKED AT XPROC, AS A PIPELINE MIGHT BE INTERESTED IN DOING SO BECAUSE YOU CAN ADDRESS MANY, MANY HIRE LEVEL PROCESSING REQUIREMENTS BY WRITING PIPELINES THAT COMBINE PROCESSES TOGETHER. AND SO HERE, SIMILAR TO WHAT GARRETT WAS SHOWING YOU, I HAVE A VERY GENERIC XPROC PIPELINE THAT PRODUCES EPUB, RATHER THAN SHOW YOU EPUB, I CAN SHOW YOU EPUB, IT DOES WORK AND IT PRODUCES A EPUB AND VALID WORK AND THEN I HAVE THE PROBLEM OF WHAT KIND SHOW SHOW YOU IN THIS WINDOWS AND IT'S TERRIBLE AND IT WILL LOOK UGLY AND WE KNOW THIS BUT YET ONCE AGAIN IT'S OPEN SOURCE, SO YOU KNOW THE 80% IS IN PLACE AND CAN YOU DEFINE WHAT YOUR EPUBS NEED TO DO AND THEN GO IN AND AMEND THE CODE IN ORDER TO MAKE IT HAPPEN IN YOUR EXTENSION OF THIS FRAMEWORK. WHAT I WILL SHOW YOU IS THAT I'VE ALSO GOT A SCENARIO WHICH RUNS AN XPROC WHICH WILL PRODUCE WHAT I'M CALLING AN EPUB MOCK UP, AND IT'S THE EPUB THAT'S UNPACK INDEED YOUR FILE SYSTEM SO CAN YOU SEE EXACTLY WHAT THE ARCHITECTURE OF THAT UPUB IS WITHOUT HAVING TO OPEN IT UP IN YOUR UPUB EDITOR OR UNZIP IT OR SOMETHING IF YOU'RE DEVELOPING EPUB, YOU WILL BE DOING A LOT OF. SO YOU KNOW FOR SOME OF US THIS MAY BE SHORTENING THE DEVELOPMENT LOOP A LITTLE BIT AND IN ADDITION IS SUGGESTS THAT WE HAVE A POSSIBLE FUNCTIONALITY HERE THAT THIS COULD BE AMENDED SO THAT RATHER PRODUCING AN EPUB AND PRODUCING A STATIC WEB SITE BECAUSE HAVE YOU A WHOLE SET OF WEBPAGES THAT ARE INTERLINKED. SO, THAT'S BASICALLY WHAT MY EPUB IS SUPPOSED TO LOOK LIKE, ONLY YOU'RE LOOKING AT IT IN A BROWSER AND BASICALLY WHAT THE EPUB XPROC DOES IS IT TAKES THIS AND BUNDLES IT UP AND ADDS THES INFRASTRUCTURE BITS AND DOES THE PACKAGES AND SO FORTH. SO THAT'S WHAT THE UPUB PREVIEW LOOKS LIKE. SO AS I SAID THE HARDEST PROBLEM HERE IS IN A SENSE DEFINING A SET OF FUNCTIONALITIES THAT WILL BE GENERIC AND GOOD ENOUGH TO IMPROVE WITHOUT GOING TOO FAR IN THE DIRECTION OF THE SPECIFIC FUNCTIONALITYS THAT THE PARTICULAR APPLICATION WILL NEED AND YOU KNOW, IT'S ACTUALLY AN OPEN QUESTION WHETHER THAT IN BETWEEN SPACE EVEN EXISTS AND AS I WAS GOING INTO THIS PROJECT, I KIND OF LIKE--I HAD MY DOUBTS ABOUT THAT. THE MAIN REASON BEING THAT BECAUSE HAVING RELEASED THE OLDER FRAMEWORK IN 2012, I REALLY HEARD NOTHING BACK FROM ANYBODY. I DIDN'T KNOW IF ANYBODY WAS USING IT. IT WAS CLEAR--IT NOT REALLY CLEAR THAT IT WAS GOOD TO IMPROVE AT THAT POINT. HOWEVER GARRETT'S WORK HAS REASSURED ME THAT THAT MAY NOT BE ALTOGETHER THE CASE. SO THE LITTLE LIGHT BULB OR HOPE AT PANDORA'S BOX IS PIPING UP AND SAYING ALL MAY NOT BE LOST HERE. SO THAT'S MY YIELD TO YOU, IS PLEASE GIVE ME FEEDBACK AND GIVE IT TO ME, SO WELL, FOR ME IT'S NOT YET GOOD ENOUGH TO IMPROVE SO--OR, IF--IF FOR EXAMPLE YOU HAVE A MAJOR COMMITMENT FOR ANOTHER TOOL SET AND OXYGEN IS NOT THE ORIZEON, MAYBE THAT'S NOT SOMETHING I NEED TO HEAR BUT IF BUT IF THERE'S SOMETHING THAT IMPAIRS ON THE DEVELOPMENT OF THAL AND WHICH YOU KNOW MIGHT PROMOTE AND IMPROVE IT, THAT'S SOMETHING I'M REALLY, REALLY EAGER TO HEAR ABOUT. THE OBVIOUS LEAP THERE'S NO POSSIBLE WAY OF GUARANTEEING THAT ANY FUNCTIONALITY THAT YOU CAN AM UP WITH WILL MAKE IT INTO THIS. AND YOU KNOW, THERE'S ALSO CLEARLY CERTAIN AMOUNT OF DISCUSSION, IT'S HARD FOR 1 OTHER THAN THINKING QUIETLY TO FIGURE OUT WHAT IS GOING TO BE GENERICALLY USEFUL FOR EVERYBODY I THINK THE WAY WE COME TO THAT IS MORE BY COMMUNICATING AND SO THAT'S WHAT I'M--INVITING YOU TO DO. AS I SAID THIS IS ON GIT HUB, SO IT'S EASIER TO FORK, EASIER TO DOWNLOAD AND DOWNLOAD IT AND RAN SACK IT, YOU DON'T HAVE TO DO ANYTHING PUBLIC IF YOU WANT, YOU IT PULL IT DOWN AND UNZIP IT AND DO WHAT YOU WANT WITH THE PIECES BECAUSE THIS IS AN OPEN SOURCE PROJECT AND THAT'S ALL IN SCOPE. BUT ON THE OTHER HAND, YOU KNOW IT'S ON GIT HUB WHERE WE HAVE WIKI AND ISSUES LISTED AND SO FORTH SO YOU CAN PIPE UP. AND OBVIOUSLY IF YOU DON'T WANT TO DO THAT IN PUBLIC, CAN YOU SEND ME E-MAIL IN PRIVATE, THAT WORKS TOO. SO AT THAT POINT, I THINK I'M GOING TO LEAVE OFF. I'M SURE THERE ARE THINGS I'VE FOCUS ON THE GOTTEN WHICH I WILL REMEMBER LATER AND I CAN'T HELP THAT NOW WE HAVE QUESTIONS FOR FEEDBACK AND I WOULD BE HAPPY TO HEAR FROM YOU. [ APPLAUSE ] >> [INDISCERNIBLE] >> I HAVE. >> [INDISCERNIBLE]. >> SO READING THEM IS AN EPUB READER. CORRECT? >> CORRECT. [INDISCERNIBLE] EPUB READER FROM [INDISCERNIBLE]-- >> DULY IT DOES BETTER THAN A COUPLE OF THE OTHERS I LOOKED AT. >> AND IT'S JUST DOING THINGS RIGHT NOW THAT ARE SOMEWHAT HARD LATELY, SO WE WANT TO SHOW IT, WE CAN BUT-- >> RIGHT, WELL UNREGISTERED FORT PLATE--UNFORTUNATELY, I FOUND THIS IN A BAG BECAUSE I WANTED A DEMO OFF A WINDOWS MACHINE SO IT'S NOT ON THIS THING BUT I HAVE LOOKEDDA IT AND IT LOOKS PRETTY GOOD AND NOW OF COURSE WHAT THAT MEAN SYSTEM THAT ALL THE BLEMISHS OF THE AWKWARD CSSARE EXPOSED BECAUSE THE HEY, THE GUY WHO LOVES CSSBARELY KNOWS WHAT HE'S DOING, BUT THERE ARE CSSEXPERTS IN THE AUDIENCE WHO COULD SPIFF IT UP AND IF THEY'RE NICE AND GENERIC PULL THEM UP INTO THE FRAMEWORK SO THAT'S SOMETHING WE COULD DEFINITELY DO. THANK YOU. REEDIUM--REFRESH YOUR R-E-A-D-I-U-M. YES? >> WONDERING HOW MANY PUBLISHERS ACCEPT THE PUBLISHMENTS OF JATS ARTICLES? >> WELL NOT VERY MANY. I'M GLAD YOU ASKED THAT BECAUSE IT GOES TO WHAT MAKES THIS GREAT ENOUGH TO IMPROVE. THERE ARE A FEW THINGS I WANT TO SAY THAT BEAR ON THAT 1 NUMBER IS OXYGEN IS INTEGRATED VERY SUCCESSFULLY INTO ALL KINDS OF WORK FLOWS AS A SECOND STEP AND RIGHT? SO IN OTHER WORDS OXYGEN IS CAPABLE OF CALLING A CONVERSION ROUTINE AND, YOU KNOW PEOPLE ARE--ARE SIMPLY USING A SERVICE THAT'S PROVIDED BY ANOTHER PIECE OF SOFTWARE TO DO THEIR INITIAL CONVERSION PRODUCING JATS STANDINGS AND A SINGLE OPERATION HAS A COUPLE OF WEEKS OF DOING THAT. BUT I THINK TO THE MEAN POINT. THIS IS NOT NECESSARILY AN AUTHORING ENVIRONMENT, RIGHT? IN OTHER WORDS THERE ARE A LOT OF WAYS YOUD USE THIS, YOU COULD USE IT AS A PLATFORM FOR AN AUTHORING ENVIRONMENT THAT'S WHAT I WOULD WOULD DO BECAUSE I'M MOSTLY RIGHT XML, THESE DAYS, IT COULD A QA PLATFORM. WILL RIGHT? SO THAT YOU COULD USE IT AND FOCUS ON THIS, THE SCHEMATRON AND AUTOCORRECTION FEATURES AND SO FORTH AND SUPPORT YOUR--SUPPORT YOUR QA PROCESSES THROUGH SOMETHING LIKE THIS. YOU COULD USE IT AS A STARTING POINT FOR PRODUCTION, SO IF YOU HAVE PEOPLE WHO ARE PRODUCING EPUBS AND KNOW HOW TO GET INTO THE GUTS OF EPUBS AND COMFORTABLE WITH XLT, A HUGE AMOUNT OF THE WORK IS ALREADY DONE FOR YOU AND GETTING OUT EPUBS WHICH ARE THE DESIGN OTHER THAN ISSUES WHICH IS WHAT PEOPLE ARE MOSTLY HAPPY AND NOT NECESSARILY SO HAPPY ABOUT THE PROBLEM TO GET UNCOMPRESSED PLAIN TEXT FILE INTO A ZIP WHICH IS SOMETHING YOU HAVE TO DO WHEN YOU BUILD EPUB, WELL GUESS WHAT I'VE DONE THAT. THAT'S HERE. SO ALL THOSE HARD ASSPECS OF PRODUCTION ARE STREAMLINED BY BEING INTEGRATED INTO PIPELINES THAT USE STANDARD TECHNOLOGIES LIKE XMLT AND XMLSO AND SO FORTH AND IF, THIS IS GOOD ENOUGH TO IMPROVE, THAT MAKES IT GOOD ENOUGH FOR THE PURPOSE OF THE PROJECT. SO, YOU KNOW I WOULD LIKE EVERYBODY TO THINK WIDELY ABOUT WAYS IN WHICH THIS MIGHT BE USED, AUTHORS ENVIRONMENT, QA ENVIRONMENT, PRODUCTION ENVIRONMENT, AND BASIS FOR INTERCHANGE. ONE OF THE LARGER THEMES THAT EMERGED FOR ME IN THIS CONFERENCE THAT'S INTERESTING IS THAT WE DON'T NECESSARILY EVEN KNOW WHAT WE MEAN BY DATA INTERCHANGE WHEN WE TALK ABOUT IT AND THERE'S INTERFERON-GAMMA T-CELL CHANGE ACROSS ORGANIZATIONAL BOUNDARIES BUT THERE'S ALSO CHANGE WITHIN THE ORGANIZATIONS AND 1 OF THE THINGS, 1 OF THE GREAT THINGS ABOUT XML THAT YOU KNOW, I REMEMBER, YOU KNOW WORKING WITH A VERY, VERY FIRST SET OF XMLTOOLS, BACK WITH TOM SCHEDEBBIE AND YOU KNOW, UP AND ROCKVILLE AND IMMEDIATELY WE FIGURED--WE FOUND THAT WE WERE ACTUALLY USING MULTIPLE TOOLS ON XML AND THAT WASN'T JUST BECAUSE WE WANTED TO TEST THE XML AND MAKE SURE IT WAS IN FACT CONFORMING, THAT WAS 1 REASON TO QUALITY CHECK OUR OWN WORK BUT ANOTHER REASON TO USE LOTS OF DIFFERENT TOOLS WAS BECAUSE THEN WE GOT THE ADVANTAGES OF LOTS OF DIFFERENT TOOLS AND YOUR XML CAN MOVE VERY, VERY NICELY FROM 1 TOOL TO ANOTHER AND I THINK FOR ORGANIZATIONS THAT CAN,A FORD TO WORK THIS WAY, I MEAN IT'S NOT NECESSARILY AN OBVIOUS THING FOR EVERYBODY BUT HAVING A QA PLATFORM THAT WORKS IN A CERTAIN WAY FOR 1 GROUP OF PEOPLE AND THEN HAVE ANOTHER QA PLATFORM THAT WORKS IN ANOTHER WAY FOR ANOTHER GROUP OF PEOPLE AND COMBINE THEM YOU GET MUCH BETTER QUALITY RIGHT? BECAUSE YOU'RE HAVING THE 2 PLATFORMS WORK OFF EACH OTHER NOW I ACKNOWLEDGE THE SAME PUBLISHER CAN'T DO THAT BUT AT THE SAME TEE WOO HAVE TOOLS THAT CAN BE USE WIDE OTHER TOOLS AND SO FORTH THAT MEANS THAT ONCE AGAIN YOU HAVE SOME SORT OF--SOME SORT OF USE CASE FOR SOMETHING LIKE THIS. >> PAUL JOHNSTON. WHAT'S IT WRITTEN IN? HOW DO YOU INDGREAT--IT LOOKS COOL. I LIKE THE JATS-- >> SORRY JATS IDE, INTEGRATED DEVELOPMENT ENVIRONMENT. >> YES. >> WHAT'S IT WRITTEN IN? HOW DO YOU EXTEND IT? >> OKAY. IN OXYGEN EVERYTHING IS UNDER [INDISCERNIBLE]. MENT AS YOU CAN SEE THIS IS A TREE OF THINGS IS SO LARGE THEY HAVE A SEARCH BAR AT THE TOP BUT BASICALLY THE WAY THIS WORKS IS, THAT OXYGEN COMES SET UP WITH FRAMEWORKS FOR ALL KINDS OF DIFFERENT DOCTRINE TYPES AND HERE CAN YOU SEE THAT I'VE GOT THESE 3 DIFFERENT JATS KIT FRAMEWORKS TURNED ON AND THERE'S ALSO A LOWER LEVEL 1. THIS GUY WHICH IS A LOW PRIORITY FRAMEWORK AND IT WILL TURN ON, IT WILL NEVER TURN ON BECAUSE IT'S NOT ACTUALLY BOUND. THIS IS A PARENT OF THE 3 SPECIFIC FRAMEWORKS I MADE. ONE FOR AUTHORING, 1 FOR PUBLISHING AND 1 FOR BOOK. I CAN HEAR THE QUESTION LURKING? WHAT ABOUT GREEN? WELL, ASK THAT QUESTION--[INDISCERNIBLE]. SO THE WAY I WOULD DO DO IT BUT I WOULD GO TO THE GENERIC 1 AND AS YOU SEE, YOU HAVE DUPLICATE OR EXTEND. DUPLICATE CREATES A COPY OF IT WHICH YOU CAN HACK. BUT EXTEND ALLOWS YOU TO CREATE A NEW 1. AS CAN YOU SEE THE DEFAULT IS NAMED WITH THE EXTENSION BECAUSE YOU ADDED ALL THIS AND THEN ALL THE WORK IS DONE IN THESE TABS, RIGHT? SO BASICALLY YOU'VE GOT A SCHEMA TAB WHICH I NEVER SET UP BECAUSE I RELY ON CATALOGS AND SCHEMA RESOLUTIONS DO THAT BUT CAN YOU PUT YOUR CATALOG IN HERE. THE AUTHOR TAB, HAVE YOU A PLACE WHERE YOU CAN SET UP ALL THE GSSAND SO FORTH SO ALL THE BINDINGS, ALL THE COMPONENTS ARE SET UP AND RIGHT? AND THEN THE COMPONENTS THEMSELVES ARE IMPLEMENTED IN EITHER CSSOR XLT OR SCHEMATRON WITH OR WITHOUT QUICK FIXES OR CAN BE EXTENDED WITH THE JAVA OR AVIs SO CAN YOU DO MORE MANIPULATIONS BY PUTTING THEM IN UP IN JAMILA OFA AND ALL THAT IS DONE THROUGH HERE. IT WILL TELL THE TEMPLATE WHERE TO GO TO LOOK FOR TEMPLATES, THESE ARE THE TRANSFORMATION SCENARIOS AND THIS IN TENSION YOU CAN ADD THEM OR REMOVE THEM AND THAT MEANS THAT YOU KNOW, YOU DON'T--BECAUSE IT'S AN EXTENSION YOU HAVEN'T LOST THE RELATIONSHIP WITH THE PARENT SO THAT YOU KNOW, FUTURE CHANGES CAN BE INCORPORATED. BUT ONCE AGAIN YOU CAN FORK TOTAL IF YOU WANT TO. VALIDATION HERE IS THAT YOU'RE SETTING A VALIDATION SCENARIOS, WHERE YOUR SCHEMATRON IS PUTTING AND SO FORTH. SOPHISTICATEDY IT'S A BIT OF FUSSING TO KIND OF FIND WHERE THE COMPONENTS ARE AND THEN YOU FINALLY LOCATE THE THING. THE FRAMEWORK IS BUNDLED AS A SET OF FOLDERS AND YOU HAVE TO JUST LOCATE THE AREA WHERE YOU WANT TO EDIT OR PUT A NEW CS S SOME PLACE AND PUT THE BINDING IN HERE AND EVERYTHING JUST WORKS. FELT I HAVE FOUND THIS TO BE VERY WORKABLE. YOU KNOW WHAT I MEAN BY THAT IS WHEN I HAVE PROBLEMS THEY'RE EASY TO FIX. AND BECAUSE FIRST WHEN YOU'RE--DOING THIS KIND OF PROJECT YOU'RE RUNNING INTO MANY SORTS OF PROBLEMS YOU KNOW EXACTLY WHAT KIND OF--THERE'S A DESIGN ASPECT OF THIS, OF HOW DO I SET THIS UP, WHAT THE DEFAULT SHOULD BE, YOU KNOW HOW ARE THINGS GOING TO BE PRESENTED AND SO FORTH. DOES THAT ANSWER THE QUESTION? IT'S NOT HARD. NOBODY WANTS TO TALK ABOUT GREEN? WELL I CAN WIRE IT UP WITH GREEN. I DID NOT WIRE IT UP WITH GREEN BECAUSE MAINLY BECAUSE I WANT TO IMPLY THAT GREEN WOULD BE SUPPORTED. NOW THAT DOESN'T MEAN GREEN ISN'T GOING TO WORK. WHAT IT DOES MEAN IS THAT THE GREEN PARTS OF GREP ARE GOING TO SORT OF FALL THROUGH: THEY'RE GOING TO BE LOST BUT THEY'RE GOING TO HAVE TO DO SOMETHING TO ACTUALLY GET THEM TO DO WHAT YOU WANT, RIGHT? SO GREEN IS KIND OF THE TEST CASE FOR YOU KNOW PUSHING THIS A LITTLE BIT, RIGHT? AND TO SUPPORT GREEN WOULD BE SIMPLY A MATTER OF MAKE THANKSGIVING EXTENSION AND PUTTING IN A CATALOG THAT POINTED TO YOUR GREEN DTDs AND THAT'S HOW HARD IT WOULD BE AND THEN ALL THESE STYLE SHEETS AND EVERYTHING WOULD WORK IN YOUR GREEN. >> [INDISCERNIBLE]. >> LUNCH TIME. >> [ APPLAUSE ] [ APPLAUSE ] [ APPLAUSE ] >> OKAY, SO BEFORE WE BREAK FOR LUNCH, DO WE HAVE A REPORT FROM THE SOCIAL DIRECTOR? NO? NOT YET. OKAY. WE'LL HEAR FROM THE SOCIAL DIRECTOR MAYBE AFTER LUNCH ABOUT OUR EVENING EVENT. THE IF YOU ARE SIGNED UP FOR AN OPEN SESSION AND HAVE YOU NOT BROUGHT YOUR SLIDES, IF YOU HAVE THEM, SLIDES ARE NOT REQUIRED, BRING THEM UP, IF YOU WANT THEM BEFORE LUNCH OR AFTER LUNCH? >> PREFERABLILY BEFORE SO WE DON'T HAVE TO STRESS ABOUT IT. ALSO IF YOU'RE USING YOUR MACHINE CHECK OUT THAT WE HAVE THE RIGHT HOOK UPS PLEASE BEFORE YOU GO TO LUNCH. THANK YOU. >> OKAY, SEE YOU AT 1:30. >> OH WAIT, WE HAVE 10 PEOPLE SIGNED UP SO THOSE WHO ARE SPEAKING, YOU WILL GET ABOUT 10 MINUTES, I REALIZE 10 TIMES 10 IS A HUNDRED BUT NOT 90 BUT WE'LL MAKE IT WORK. >> OKAY. >> DO YOU NEED MORE, CAN YOU TAKE MORE? >> IF YOU'RE REALLY WOULD LIKE TO SPEAK AND YOU HAVE THAT BURNING DESIRE JUST, YOU KNOW LET US KNOW, WE'LL WORK IT OUT. >> WELCOME BACK TO AN EXCITING PART OF THE PROGRAM. FIRST UP IS STEPHEN, WHO ASK TALKING TO US ABOUT HOW USING JATS TO HELP ENGLISH AS A SECOND LANGUAGE AUTHORS. >> OKAY. I'M THE INTEGRATION MANAGER WITH ADVANCED EDITING, WE PROVIDE SERVICES FOR PUBLISHERS AND FOR AUTHORS TO HELP AUTHORS WHO HAVE ENGLISH AS A SECOND LANGUAGE THROUGH THE PATH TO PUBLICATION. IT WON'T BE A SURPRISE TO KNOW AUTHORS HAVE PROBLEMS NAVIGATING THEIR WAY THROUGH THIS, AND LOTS OF DIFFERENT STAGES THAT THEY WILL ENCOUNTER THOSE PROBLEMS. I WON'T GO THROUGH EVERY ONE. SUFFICE TO SAY WE HAVE PAID SERVICES THAT WILL HELP AUTHORS THROUGH EACH OF THESE DIFFERENT AREAS. OF ALL OF THESE KIND OF DIFFERENT WAYS IN WHICH THE AUTHORS MAY HAVE PROBLEMS, OBVIOUSLY THEY HAVE LOTS OF QUESTIONS ALONG THE WAY. IT'S OUR JOB TO KNOW WHAT THESE QUESTIONS ARE PROBLEMS THEY ARE HAVING, ALL THAT KIND OF THING. DURING ONE OF THE SURVEYS THAT WE CARRIED OUT WITH OUR AUTHORS THEY TOLD US THE MOST DIFFICULT THING THEY ENCOUNTERED DURING THE PREPARATION OF MANUSCRIPTS WAS CHOOSING WHICH JOURNAL TO SUBMIT TO SO WE CREATED A FREE SERVICE ON OUR WEBSITE, AND THAT IS THE EDANZ JOURNAL SELECTOR, ENABLING THE AUTHORS TO SEARCH AND FOR JOURNAL, BY JOURNAL NAME, PUBLISHER NAME, FIELD OF RESEARCH, FILTER THOSE SEARCHES SO HE CAN THAT FIND OUT WHICH HAVE OPEN ACCESS, PUBLICATION FREQUENCY, ALL THAT KIND OF THING THAT HOPEFULLY WILL HELP THEM MAKE A DECISION AS TO WHICH JOURNAL THEY ARE GOING TO SUBMIT TO. WE ARE BASED IN ASIA, I'M IN BEIJING, WE HAVE A CHINA OFFICE AND OFFICE IN KOREA. AND THAT IS WHERE THE MAJORITY OF OUR PAID CUSTOMERS ORIGINATE FROM. TO GO ALONG WITH THOSE DIFFERENT LOCATIONS OBVIOUSLY WE HAVE LOCALIZED SITES AND THE JOURNAL SELECTOR IS AVAILABLE IN LOCAL LANGUAGES ON EACH LOCALIZED SITE. SO YOU CONSIDER -- YOU CAN SEE THAT'S WHAT IT LOOKS LIKE, YOU CONDUCT YOUR DIFFERENT SEARCHES THROUGH THIS INTERFACE. AS I MENTIONED YOU CAN DO THAT THROUGH THE JOURNAL, THROUGH THE PUBLISHER, THROUGH THE FIELD OF RESEARCH, BUT JATS HAS TO COME INTO IT SOMEWHERE ELSE, OR I WOULDN'T BE ALLOWED TO STAND HERE. THE WAY WE DO THAT, WE ALLOW THE AUTHORS TO BE ABLE TO SEARCH FOR JOURNALS USING THEIR ABSTRACT. SO THEY CAN SUBMIT THEIR ABSTRACT INTO THE INTERFACE AND OUR ALGORITHMS WILL ANALYZE THE ABSTRACT AND THEY WILL GO AND LOOK AT THE ABSTRACTS IN OUR DATABASE, ABSTRACTS THAT COME IN IN JATS, AND THEY WILL COME BACK WITH RESULTS TO FIND WHICH JOURNALS ARE PUBLISHING RESEARCH, WHICH MOST CLOSELY MATCHES THAT OF THE AUTHOR'S ABSTRACT. AND THIS IS THE KIND OF RESULT THAT YOU WOULD GET FROM THAT. SO YOU CAN SEE WE'VE GOT A COUPLE OF JOURNALS LISTED THERE, HOPEFULLY YOU CAN SEE THAT ANYWAY. AND THEY AUTOMATICALLY ARE ORDERED ACCORDING TO THE ORDER OF BEST MATCH. YOU CAN SEE ON THE LEFT-HAND SIDE THERE WE'VE GOT RIGHT-HAND SIDE, LEFT-HAND SIDE, OH, THERE. [LAUGHTER] THERE IS THE AREA THAT YOU CAN DO ALL OF YOUR FILTERING, FILTER ACCORDING TO IMPACT FACTOR, SCI, OPEN ACCESS PUBLICATION FREQUENCY, ALL THIS KIND OF THING. AND SO THE AUTHORS CAN THEN SEE WHICH JOURNALS ARE PUBLISHING RESEARCH MOST RELEVANT TO ABSTRACT THEY SUBMITTED AND THEY CAN THEN GO INTO THOSE PARTICULAR JOURNALS, REALLY, REALLY TINY PICTURE HERE, SO WHAT I'VE DONE IS BLOWN EACH OF THESE DIFFERENT AREAS UP. AT THE TOP OF THAT PAGE WE HAVE THE VARIOUS JOURNAL INFORMATION. THE SUMMARY OF THE JOURNAL, IMPACT FACTOR, WHO THE PUBLISHER IS, BLAH, BLAH, BLAH, BLAH, OPEN ACCESS OPTIONS. YOU CAN ALSO SEE WE KIND OF MAKE IT AS EASY AS WE CAN FOR THE AUTHORS TO BE ABLE TO GET INTO THESE JOURNALS BY HAVING SUBMISSION URLs. WE HAVE THE LINKS TO THE INSCRIPTIONS FOR AUTHORS AND LINKS TO ACTUAL SITES OF THE JOURNAL AS WELL. WHAT WE'RE ALSO ABLE TO DO IS TAKE THOSE RESULTS OF THE MATCHING THAT WE FOUND THROUGH THE ABSTRACTS THAT WE'VE GOT WITHIN THE JATS FILES AND GIVE THE AUTHORS THE EXAMPLES OF THE ARTICLES THAT MOST CLOSELY MATCH THEIRS THAT HAVE BEEN PUBLISHED WITHIN THESE DIFFERENT JOURNALS. SO IN A WAY WHAT WE'RE DOING HERE AS WELL IS KIND OF DRIVING TRAFFIC BACK INTO THE JOURNALS FOR THE AUTHOR WHO MAY HAVE ALREADY WRITTEN THE MANUSCRIPT, BUT THERE MAY BE SOME RESEARCH THEY WEREN'T FAMILIAR WITH, WEREN'T AWARE OF, WE'RE ABLE TO HIGHLIGHT THAT AS WELL AS HAVING DETAILS OF MANUSCRIPTS, THE ARTICLE THAT MOST CLOSELY MATCH, WE'RE ABLE TO PROVIDE OTHER LINKS. JUST DRIVING TRAFFIC BACK THROUGH INTO THE JOURNAL WEBSITE. THAT IS KIND OF THAT. I CAN STAND HERE FOR A FEW MINUTES NOW AND TAKE UP MY FULL TEN MINUTES. [LAUGHTER] I DIDN'T REALIZE HOW QUICKLY I WAS GOING TO GO THROUGH THAT. SOME STATISTICS, WE HAVE OVER 11.5 MILLION ABSTRACTS IN OUR DATABASE. WE HAVE AGREEMENTS WITH PubMed CENTRAL, VARIOUS PUBLISHERS WHO ARE DEPOSITING WITH US DIRECTLY, AND HIGHWIRE ALLOWED US TO USE THEIR CURRENT SERVICE AS LONG AS WE HAVE PUBLISHER CONSENT TO BE ABLE TO DO THAT. AND WE HAVE 28-POINT 28500, MAJORITY FROM PMC. IF YOU'RE A PUBLISHER WHO IS NOT NECESSARILY LISTED IN PMC, AND WOULD LIKE TO TALK ABOUT MAYBE SUBMITTING YOUR ABSTRACT INTO THE JOURNAL SELECTOR, THEN COME AND TALK TO ME. WE HAVE OVER 57,000 USERS EACH MONTH. THE HIGHEST USAGE COMES FROM -- LET ME THINK. BRAZIL, CHINA, INDIA, THE UNITED STATES. I'M NOT GOING TO CRACK ANY JOKES ABOUT ENGLISH BEING SECOND LANGUAGE IN THOSE COUNTRIES BECAUSE I'M IN THE UNITED STATES AT THE MOMENT AND I WOULD QUITE LIKE TO GET BACK OUT AGAIN. [LAUGHTER] SO VERY GOOD. THAT IS THE EDANZ JOURNAL SELECTOR. MY E-MAIL ADDRESS, SHOULD ANYBODY WANT TO CONTACT ME ABOUT COOPERATION, AROUND HOW WE MIGHT BE ABLE TO INGEST CONTENT, THE COMPANY TWITTER ADDRESS, MY TWITTER ADDRESS. THANK YOU VERY MUCH. [APPLAUSE] >> QUESTIONS? OKAY. >> COOL, THANK YOU VERY MUCH. >> ALL RIGHT. >> NEXT UP TALKING TO US AND I HOPE I DON'T BUTCHER HIS NAME OR TOPIC, IS MATHEUS KEEPERRING, I GOT A THUMBS UP, ABOUT A MANUSCRIPT AUTHORING TOOL THAT'S USER FRIENDLY AND JATS COMPATIBLE >> I'M (INDISCERNIBLE) AND I'VE LEFT THERE AND GONE FULL TIME WITH A DIFFERENT COMPANY CALLED MANUSCRIPTING. PAPERS WAS A REFERENCE MANAGER. WHAT I'VE BEEN TRYING TO DO FOR THE PAST YEAR IS WITH MANUSCRIPTS IS REPEAT THE USER USE WE MANAGED TO GET WITH PAPERS, EASY TO ADOPT POWERFUL ENVIRONMENT FOR SOMEBODY TO DO EVERYTHING ABOUT PAPER WRITING. AND WHY AM I TALKING ABOUT IT HERE? NOT ONLY IS IT A SCIENTIFIC WORK PROCESSER, IT'S ALSO A STRUCTURED APPROACH TO THAT PROGRAM. SO WHAT YOU'RE LOOKING AT HERE ARE AUTHOR METADATA TOOL, CLOSELY MODELED AFTER JATS, WE ALREADY EXPORT THE MATTER INTO JATS FOR EDITORIAL MANAGER INGEST WHICH I THINK WE'LL HEAR NEXT. SO YOU CAN ENTER AUTHOR DATA, AFFILIATIONS, GRANTS, ALL OF THAT STUFF AND WE FOR INSTANCE FORMAT THAT IN A SPECIFIC WAY IN THE EDITOR AND WE'RE ABLE TO TRANSFORM THAT INTO OTHER FORMATS. AND WHAT I'M GOING TO FOCUS ON IS MORE KIND OF THE ANGLE OF STRUCTURED WRITING FROM THE POINT OF VIEW OF HOW DOES IT ACTUALLY HELP AN AUTHOR BECOME MORE PRODUCTIVE AS OPPOSED TO HOW DOES THAT ACTUALLY ADD MORE WORK TO THE AUTHOR BECAUSE I THINK IT'S REALLY THE FORMER CASE, THAT WE CAN ACTUALLY HELP THE AUTHOR BECOME FAR, FAR MORE PRODUCTIVE IF THE MATTER THAT YOU'RE WRITING IS TO BEGIN WITH STRUCTURED. HERE IS ONE EXAMPLE. WE HAVE A DOCUMENT OF WELL OVER A THOUSAND DOCUMENT TEMPLATES IN A TOOL AND WORK WITH PUBLISHERS SUCH AS ELIFE FOR GIVING THE TOOL FOR FREE FROM THE PUBLISHER POINT OF THROUGH AND ALSO FREE FROM THE AUTHOR POINT OF VIEW FOR SUBMITTING TO CERTAIN TITLES. FOR INSTANCE I'VE ACTIVATED THE FREE MODE FOR DOING ELIFE RESEARCH ADVANCE, RESEARCH ARTICLE SUBMISSIONS, AND WHAT IS A TEMPLATE FROM OUR POINT OF VIEW? BASICALLY WHAT WE DO WITH THIS TOOL IS WE GIVE YOU A PLACE HOLDER FOR THE SECTION HIGH AERODYNAMIC, ENCODE STYLE INFORMATION IN TEMPLATES. WHAT ARE THE REQUIRED FIGURE FILE FORMATS, WHAT ARE THE -- SHOULD THE FIGURES BE EXPORTED AS A SEPARATE FILE AT THE END OR IN LINE OR AT THE END OF THE SAME FILE? LOTS OF DIFFERENT STYLE RULES THAT HAVE TO DO EITHER WITH WHAT THE DOCUMENT LOOKS LIKE OR KIND OF REQUIREMENTS ON WHAT SHOULD BE INCLUDED. HERE IS ANOTHER EXAMPLE. WE'RE ABLE TO TELL THAT IN THIS DOCUMENT TYPE THE ABSTRACT SHOULD BE AT LEAST 300 WORDS LONG, IN THIS CASE IT ISN'T. WE INDICATE TO THE USER, SO IT DOESN'T BECOME A PROBLEM AT SUBMISSION STAGE BUT SOMETHING THEY NATURALLY WORK TOWARDS FROM THE BEGINNING OF THE WRITING PROJECT. HERE IS ANOTHER EXAMPLE, THE SEPARATION OF DOCUMENT TREE, SECTIONS, THE ACTUAL BODY AS A SEPARATION IF THAT HELPS YOU. IT'S SUBTLE. I'LL POINT OUT WE'RE GOING TO REORDER CONTENT I CAN THE SAME THING WITH SECTIONS HIERARCHY. IT TENDS TO BE HELPFUL BASED ON CUSTOMER FEEDBACK WHERE YOU START FROM A ROUGH IDEA OF WHAT YOU'RE GOING TO WRITE AND KEEP ITERATING AS YOU GO, THAT WAS MY APPROACH TO WRITING, ROUGH IDEAS. OUTLINER TOOL AND THEN START WRITING INTO SPECIFIC PARTS OF THE BODY. ANOTHER THING ABOUT THE OUTLINE IN THE TOOL IS THIS FOCUS MODE, LIKE A TINY ICON, IF I CLICK THAT. ONLY THAT SPECIFIC SECTION IS NOW AVAILABLE FOR WRITING. SUPPOSE THIS WAS NOT A FINISHED PRODUCT, I COULD FOCUS ON THE THING I WORRY ABOUT TODAY AND NOT THE REST BEING OVERTLY UNFINISHED OR JUST FOCUS MY MIND ON A SPECIFIC DOCUMENT SECTION. ANOTHER THING ABOUT A STRUCTURE IS FIGURES. NON-DESTRUCTIVELY WE KEPT VECTOR FORMATTED IMAGES, INTO A SINGLE IMAGE, BOTH HAVE A TITLE COMPONENT AND A CAPTION THAT GOES WITH THE ENTIRE PANEL. ALL THIS IS RETAINED IN A NEAT CLEAN FORM AND CAN BE EXPORTED INTO VARIOUS FORMS. ANOTHER EXAMPLE IS TABLES. SO TABLES ARE DICTATED BY TABLE STYLES IN THIS DOCUMENT SO FOR INSTANCE I'M NOT ABLE TO DO -- LET'S BOLD THE HEADER AND BOLD THE FOOTER AND TABLE, THAT'S IMPLICIT BASEDDED ON WHERE IT IS IN THE TABLE. I CAN HIDE AND SHOW HEADERS AND FOOTERS, THERE'S A LOT OF CONTROL GIVEN OVER HOW THEY LOOK BUT THEY LOOK CONSISTENT. EQUATIONS WORK THE SAME WAY. THERE'S A COUPLE OPTIONS GIVEN FOR THEM. WE LET YOU DO EQUATION EDITING IN LINE AND PARAGRAPH IN THE TOOL BUT FIRST YOU CAN CHOOSE WHETHER THEY ARE NUMBERED, WHETHER THEY HAVE A CAPTION THAT'S PRESENTED OR NOT, THAT CAN BE TRANSFORMED INTO VARIOUS OUTPUT FORMS WE SUPPORT. SYMBOL EDITING AS WELL, NOTHING TO SAY ABOUT THAT. CODE LISTINGS, WE HAVE ONE OF THE ENTITY TYPES, CODE, NOT SO MUCH AIMED AT JOURNAL ARTICLE. I WANTED IT FIRST FOR TECHNICAL DOCUMENTATION, BUT THAT'S ANYWAY ONE OF THE SUPPORTED TYPES. WE SUPPORT EXPORTING OUTPUT FORMATS, IMPORTING AS WELL. IMPORT AND EXPORT WORK. WE IMPORT AND EXPORT TEXT, MARK DOWN, IF YOU COPY AND PASTE OUT OF THE APP, THE CLIPBOARD, AND PASTE INTO A TEXT EDITOR BY DEFAULT THAT'S MARKED DOWN. WE HAVE SORT OF A LOT, IT'S PART OF PROJECT, INTERPLAY WITH OTHER SYSTEMS AND OTHER FILE FORMATS, AND THAT IS GREATLY MADE EASIER BY THE FACT THAT THE DATA MODEL IS GRANULAR AND STRUCTURED. THERE'S A COUPLE OF DIFFERENT INTEGRATION POINTS FROM A PUBLISHER POINT OF VIEW THAT WE HAVEN'T EXPLORED TOO MUCH. WE HAVE A COUPLE OF BRANDED COMPONENTS. ONE IS PRESSROOM, THE TRANSFORMATION TOOL, THE IN FACT JATS TRANSFORMATIONS AS WELL, MANUEL IS AN AUTOMATION TOOL THAT USES DESCRIPTING HOODS, THERE'S A JAVA SCRIPT API IN THE SYSTEM AS WELL, AND FOR INSTANCE THE WAY HOW WE DO CITATION TOOL INTEGRATION WORKS THROUGH THAT SCRIPTING API SO THERE'S A LOT OF WAYS TO CREATE, AUTOMATE AND SUMMARIZE IN DIFFERENT WAYS. THE ANOTHER TOOL WE BUILT BASED ON THE DATA MODELEL IS A TEMPLATE BUILDER TOOL. YOU'RE LOOKING AT OUR INTERNAL VIEW ON WHAT IS A DOCUMENT TEMPLATE. SECTION HIERARCHY, A SECTION CALLED ABSTRACT WHICH IS OFF THE CATEGORY ABSTRACT, A WORD COUNT OF 350 INCLUSIVE OF REFERENCE. THERE'S LIKE SEVEN OTHER TABS OF STUFF THERE, OVER 90 RULES WE CAN APPLY AS VALIDATION RULES. AGAIN, THE COST FOR THIS FROM THE PUBLISHER POINT OF VIEW AND THE END USER POINT OF VIEW TO GET STARTED IS NOTHING. FROM THE PUBLISHER POINT OF VIEW WE'RE NOT MONETIZING THAT AT ALL ABOUT BUSINESSES MODELEL IS END USERS ARE HAPPY TO PAY FOR THE FULL LICENSE FOR USING UNLIMITED DOCUMENTS, NOT JUST ONES WE PREPARE THE TEMPLATES FOR. IF YOU WANT TO HEAR MORE JUST GET IN TOUCH. THAT'S IT. [APPLAUSE] >> QUESTIONS, COMMENTS? >> THANKS. >> OKAY. NEXT UP WE HAVE CAROLINE WEBBER WHO IS TALKING TO US ABOUT SIMILAR TOPIC, USING JATS TO INGEST MANUSCRIPTS >> OH, IT IS OPEN, OKAY. I THOUGHT THIS WOULD BE THE HARDEST PART. I'M CAROLINE WEBBER FROM ARIES SYSTEM, LONGTIME LISTENER, FIRST-TIME PRESENTER. I THOUGHT I WOULD GET THAT JOKE IN BEFORE SOMEBODY ELSE THOUGHT OF IT. I'M GOING TO TALK ABOUT A NEW PRODUCT THAT WE HAVE, NEW FEATURE OF INGESTING MANUSCRIPTS FROM SYSTEMS SUCH AS THE ONE THAT MADDIE JUST SPOKE ABOUT. ARIES BRIEFLY, I DID STEAL THIS FROM SALES SO I HAVE TO GO OVER A FEW THINGS ABOUT THE COMPANY. WE WERE FOUNDED IN 1986, PRIVATELY HELD, MORE THAN 6700 PUBLICATIONS USE OUR SYSTEM. WE PROCESS ALMOST TWO MILLION MANUSCRIPTS ANNUALLY. AND WE HAVE THREE MAJOR PRODUCTS, EDITORIAL MANAGER WHICH IS THE PEER REVIEW AND MANUSCRIPT SUBMISSION SYSTEM, PRODUCTION MANAGER IS WEB-BASED PRODUCTION TRACKING, AND COMMERCE MANAGER FOR E-COMMERCE. AND WE'RE USED BY A LARGE NUMBER OF PUBLISHERS, WILEY, SPRINGER NATURE, LDBW, QUITE A FEW OF THE BIG ONES, AND EL SEVEER USES A BRANDED. NOW HOW WE USE THE JATS, IN THREE WAYS. INGEST IS WHAT I'M FOCUSING ON. I'LL TALK ABOUT TWO OTHER WAYS WE USE IT. IMPORT FROM -- THAT'S AUTOMATIC MANUSCRIPT IMPORT IF YOU'RE NOT COMING FROM A SUBMISSION PARTNER, IF YOU'RE COMING FROM ANOTHER JOURNAL. AND WE ALSO TRANSMIT JATS OUT. SO INGEST IS A WAY TO GET SUBMISSIONS INTO EDITORIAL JOURNALS FROM SUBMISSION PARTNERS, AND WHO ARE SUBMISSION PARTNERS? WELL, MADDIE JUST SPOKE, AND PAUL JOHNSON IS ANOTHER SUBMISSION PARTNER SPEAKING IT BELIEVE AT 3:30. LANGUAGE EDITING SERVICES, AUTHORING TOOLS. HERE IS A LIST OF THE ONES WE'VE SIGNED UP SO FAR. AND EDITTAGE IS ALSO HERE, A LOT OF PEOPLE WE SIGN UP ARE ACTUALLY IN THIS ROOM. AND HOW IT WORKS IS INSTEAD OF THE AUTHORS SUBMITTING DIRECTLY TO THE JOURNAL ON EDITORIAL MANAGER, THEY WILL SUBMIT THROUGH THE SUBMISSION PARTNER AND THE SUBMISSION PARTNER THEN CREATES THE FILES AND THE JATS-XML. ONCE IT GETS INTO EDITORIAL MANAGER THROUGH THE INGEST, IT APPEARS IN A FOLDER ON THE EDITOR'S MAIN MENU CALLED TRANSFERRED SUBMISSIONS, AND AT THAT POINT IT WILL SAY WHO IT'S BEEN TRANSFERRED IN FROM AND THE EDITOR CAN DECIDE WHETHER TO SEND BACK TO THE AUTHOR OR MOVE ON IN THE WORK FLOW, SENDING BACK TO THE AUTHOR IS NOT A PUNISHMENT IT'S JUST TO GET THE AUTHOR TO FILL IN A FEW MORE DETAILS THAT THE JOURNAL MIGHT NEED. SO KEY INGREDIENTS TO THE PRODUCT ARE THE JATS. WE ALSO USE ORCID AND FUNDING INFORMATION, FundRef I.D.s. IT'S AN OLD-FASHIONED WAY BUT IT DOES WORK, THROUGH FTP. SO THE SUBMISSION PARTNERS FTP IN TWO FILES, A TRIGGER FILE WITH THE ARIES JOURNAL CODE AND THEIR SUBMISSION PARTNER LICENSE CODE AND THEN A ZIP FILE THAT CONTAINS THE JATS-XML METADATA FILE AND ANY FILES THAT NEED TO BE INGESTED. WE FIND THE FTP SERVER, FIND THE FILES AND START THE INGEST PROCESS. HERE IS THE MEAT OF WHAT WE READ FROM JATS. I JUST PULLED UP -- WE READ A SUBSET -- WHEN I SAY SUBSET I DON'T MEAN FORMAL SUBSET, WE READ A SMALL NUMBER OF TAGS. WE DO READ THE SUBMISSION PARTNER LICENSE CODE WHICH I JUST KIND OF JAMMED INTO THE JOURNAL I.D. TYPE EQUALS PUBLISHER BECAUSE I DIDN'T KNOW OF A GOOD PLACE TO PUT IT. THE SUBMISSION PARTNER NAME, AND THEN THE SUBMISSION PARTNER'S DOCUMENT IDENTIFIER TO PREVENT DUPLICATE INGEST. WE'RE GOING TO READ INFORMATION ABOUT THE CORRESPONDING AUTHOR AND CO-AUTHORS, AND WE CAN READ IN THE RINGGOLD I.D. FOR AUTHOR INSTITUTION IF PROVIDED. WE LOOK FOR THE CORRESPONDING AUTHOR IN THAT EDITORIAL MANAGER, IF THEY ARE NOT ALREADY THERE THEN WE DO PROXY REGISTRATION WHICH JUST MEANS REGISTRATION RECORD IS MARKED AS NOT SELF REGISTERED AND WHEN THE USER COMES IN THEY ARE ASKED TO VERIFY THE INFORMATION WE READ IN SOME MANUSCRIPT INFORMATION, ABSTRACT KEY WORDS, DOI IF IT HAS ONE ALREADY WHICH IT PROBABLY WON'T, ARTICLE TYPE IF THAT'S KNOWN, CLASSIFICATION, SECTION CATEGORY AND THEN THE POPULAR CUSTOM META ELEMENT IS USED TO READ IN CUSTOM STUFF THAT IS CLASSIFIED AS ADDITIONAL MANUSCRIPT DETAILS IN EDITORIAL MANAGER, AND THESE ARE CUSTOM FIELDS THE JOURNAL CAN SET UP TO CAPTURE ADDITIONAL INFORMATION ABOUT THE MANUSCRIPT, IN ADDITION THE FUNDING INFO, AND AS I MENTIONED THEN IT APPEARS IN THE TRANSFERRED SUBMISSIONS FOLDER ON THE EDITOR MAIN MENU AND WE IMPORT THE FILES. AND THEN AT THAT POINT TYPICALLY THE EDITOR WOULD SEND IT TO THE AUTHOR SO THAT JOURNAL SPECIFIC QUESTIONS CAN BE ANSWERED. AND THEN THE CORRESPONDING AUTHOR IS REGISTERED, THEY WOULD GET THAT INFORMATION, THEY CAN LOG IN AND THEN THE SUBMISSION THEN GOES INTO THE NORMAL WORK FLOW PROCESS FOR WHATEVER THAT JOURNAL IS. WE ALSO USE A VERY SIMILAR PROJECT, AAAS USES THIS, IF YOU HAVE A JOURNAL THAT'S NOT ON THE EM, AND I HAVE AN OPEN SOURCE MEGA JOURNAL ON EM, THE SISTER JOURNALS CAN TRANSFER IN TO THE MEGA JOURNAL USING A SIMILAR PROCESS. WE USE JATS AND FTP. WE DON'T NEED THE LICENSE CODE BECAUSE IT'S NOT A MANUSCRIPT LICENSE PROVIDER, IT'S ACTUALLY ANOTHER JOURNAL. THEN WE USE JATS TO TRANSFER OUT. WE EXPORT OUT JATS 1.0, 1.1, ALSO NLM 3.0, AND YOU CAN SET UP AN FTP OF THAT ANYWHERE, EITHER TO ANOTHER PRODUCTION SYSTEM OR IT CAN BE SENT FROM WITHIN OUR PRODUCTION SYSTEM TO ANOTHER VENDOR USING OUR PRODUCTION TASK TRANSMITTAL FEATURE. AND THAT'S IT. ANY QUESTIONS? [APPLAUSE] >> QUESTIONS? >> YES. >> JEFF FROM NLM. YOU SAID YOU USE THE PROVIDER DOCUMENT I.D. TO KEEP YOU FROM CRASHING DUPLICATE SUBMISSIONS. >> TO KEEP FROM IMPORTING THE SAME MANUSCRIPT TWICE. >> YEAH. EVER HAVE TO WORRY ABOUT GETTING THE SAME MANUSCRIPT FROM MORE THAN ONE PARTNER? >> NOT AS OF YET BUT WE DO HAVE SOME DUPLICATE SUBMISSION CHECKING BUILT INTO THE SYSTEM. THAT'S ANOTHER FEATURE. BOTH ON THE AUTHOR NAME AND WE ALSO HAVE A DUPLICATE SUBMISSION CHECK THAT GOES AGAINST THE DATABASE AND LOOKS FOR SIMILAR TITLES, ET CETERA. SO THAT HASN'T BEEN A PROBLEM SO FAR. >> RIGHT. THANK YOU. >> ANY OTHER QUESTIONS? WELL, THANK YOU. [APPLAUSE] >> YOU'RE DOING AN AMAZING JOB KEEPING TO TEN MINUTES. FANTASTIC. NEXT UP IS (INDISCERNIBLE) TALKING ABOUT JATS-CON ASIA 2015. >> OKAY. I MENTIONED LAST YEAR WE HAD A JATS-CON ASIA MEETING LAST YEAR ON OCTOBER 19, THE IT WAS HOSTED BY XSPA AND THE JAPAN SCIENCE AND TECHNOLOGY AGENCY, JST. THANKS TO JST, WE HAD A VERY GOOD FACILITY, AND TRANSLATION, SO JAPANESE AUDIENCE COMFORTABLE IN OUR LISTENING TO INVITED PRESENTERS. SO THIS IS A DR. HIDEAKI TAKEDA WHO TALKED ABOUT OPEN SCIENCE AND IDENTIFIERS, SUCH AS DOI, ORCID, AND OTHER IDENTIFIERS WHICH WILL HELP, YOU KNOW, MAKE SCIENTIFIC INFORMATION, YOU KNOW, FLOW VERY FLEXIBLE AND ALSO BRUCE TALKS ABOUT THE JATS AND ROLE IN SCHOLARLY PUBLISHING, AND MR. NAKANISHI TALKS ABOUT THE JAPANESE LANGUAGE PROCESSING ON JATS-XML, AND ALSO AUTOMATIC TYPESETTING USING XSLT. THIS WAS A SIMILAR TALK WHICH WAS PRESENTED LAST YEAR HERE. ALSO MR. KOSALA TALKED ABOUT HIS IMPLEMENTATION OF THE JATS FOR THE ICUSHI DATABASE, ABSTRACTING DATABASE, FOR JAPANESE MEDICAL ARTICLES, WHICH IS SORT OF SIMILAR TO PubMed BUT IN JAPANESE LANGUAGE. ALSO CHANDI TALKED ABOUT CHALLENGES IN IMPLEMENTING A MULTI-LINGUAL PUBLISHING W ORKFLOW, BASIS OF HIS TALK YESTERDAY. AND CHOON SHIL LEE IN THIS AUDITORIUM TALK ABOUT JATS FOR KOREAN MEDICAL JOURNAL DATABASES. AND FINALLY THE LADY FROM JST TALKED ABOUT HOW JATS IS USED IN JOURNAL PLATFORM CALLED J-STAGE, AND FOR JAPANESE -- FOR JOURNALS PUBLISHED IN JAPAN. THAT'S ABOUT IT. WE HOPE WE HAVE ANOTHER MEETING SOMETIME SOON BUT NOT THIS YEAR BUT IF WE HAVE THE PLAN WE'LL LET YOU KNOW. THANKS. [APPLAUSE] >> I THINK EVERYONE SHOULD ACKNOWLEDGE THAT MR. TOKIZANE DID A WONDERFUL JOB ORGANIZING THE MEETING AND WE LOOK FORWARD TO HAVING ANOTHER ONE IN THE FUTURE. THANK YOU FOUR THAT. >> YOU'RE WELCOME. [APPLAUSE] >> OKAY. NEXT UP, TWICE I'VE TRIED TO GET SOMETHING OUT OF HIM AS FAR AS A TOPIC. ERIC LANDERS WILL BE SPEAKING ABOUT SOMETHING HE HAS YET TO REVEAL >> GOOD AFTERNOON. I DON'T MEAN TO BE SECRETIVE BUT WITH ALL THE INFORMATION COMING AT ME I KEEP CHANGING EXACTLY WHAT I WISH TO ACTUALLY SAY. SO I'M WITH CAS. YOU CAN TAKE A LOOK AT CAS.ORG IF WE'RE INTERESTED, THE WORLD'S PREMIER SUPPLIER OF CHEMICAL INFORMATION. I'M AT AN INTERESTING PART OF THE WORKFLOW, BASICALLY I'M RECEIVING PUBLISHERS' OUTPUT SO I'M CONSUMING PRETTY MUCH YOUR END PRODUCT AND USING IT WITHIN MY WORKFLOWS. I WILL ADMIT TO BEING A BUILD OF A LURKER WHEN IT COMES TO THE JATS. I WAS HERE FOUR OR FIVE YEARS AGO WHEN THE JATS WAS BRAND NEW, AND I WANTED TO MENTION WHAT WE'VE DONE SO FAR WITH IT ON ONE PROJECT. WE'VE BEEN ABLE TO USE THE JATS AND THE BITS TO SUPPORT NEW INPUT THOUGHTS FOR CONTENT THAT WE USED TO HAVE ENORMOUS DIFFICULTY PROCESSING. WE NOW HAVE JUST NORMAL DIFFICULTIES PROCESSING. THESE SYSTEMS ARE ACTUALLY RATHER ROBUST. WE HAVE AT THIS POINT IN THE LAST THREE YEARS PROCESSED ABOUT 650,000 ARTICLES AND BOOK PARTS USING SYSTEMS. SO IF THERE'S INTEREST, I MAY ACTUALLY TRY TO WRITE UP WHAT WE'RE ACTUALLY DOING. ONE OF THE THINGS THAT MY COLLEAGUE POINTED OUT IS THAT WE DO A JATS TO JASON CONVERSION FOR INTERNAL SYSTEMS. SO I MAY WRITE THAT UP AND SEE IF I CAN GET IT PAST THE COMMITTEE. THE ONE THING I'D LIKE TO ASK YOU IS, OH, A NUMBER OF DISCUSSIONS HERE SO FAR HAVE BEEN ABSOLUTELY MARVELOUS. SEVERAL OF THOSE HAVE REALLY PIQUED MY INTEREST FOR WHAT I'M ABOUT TO ASK. SO SOME OF YOU ARE WARNED THAT I'LL TRY TO CONTACT YOU. BUT I'M ABOUT TO EMBARK ON A LARGER PROJECT THAT WE'RE PLANNING TO USE THE JATS AND THE BITS AS THE PREFERRED METHOD OF REPRESENTING NON-PATENT PUBLISHED PUBLICLY AVAILABLE CONTENT WITHIN THE CAS WORLD, AND I AM INTERESTED IN DISCUSSING WITH ANYONE WHO HAS HAD EXPERIENCE IN LARGE SCALE CONVERSIONS OF THIS TYPE OF THE SORTS OF PROBLEMS WE RAN INTO. SOME DISCUSSIONS HAVE BEEN ALONG THOSE LINES, AND THOSE HAVE BEEN VERY NICE. BUT IF THERE'S ANYONE ELSE THAT HAS ANY INFORMATION OR INSIGHTS THEY WISH TO SHARE, I WOULD DEFINITELY BE INTERESTED. I'M ALSO INTERESTED IN HOW PEOPLE USE THE JATS OR THE BITS TO REPRESENT SOME OF THE MORE INTERESTING TYPES OF CONTENT SUCH AS DISSERTATIONS, TECHNICAL REPORTS. SO I WOULD APPRECIATE ANY CONTACT. THANK YOU VERY MUCH. [APPLAUSE] >> OKAY. NEXT UP IS SOMETHING WE'VE HEARD ABOUT SEVERAL TIMES DURING THE CONFERENCE, IT'S THE JATS FOR R. AND MARY WILL BE TALKING TO US ABOUT THAT >> OKAY. I JUST RECEIVED ELABORATE INSTRUCTIONS ABOUT THE MICROPHONE. HI, EVERYBODY. I'M MARY FROM CANADIAN SCIENCE PUBLISHING, TALKING TO YOU ABOUT THE GOINGS ON JATS4R. YOU MAY HAVE HEARD JATS4R'S TALK LAST YEAR AT JATS-CON, A LONGER PRESENTATION, JUST IN CASE YOU'RE SUPER EXCITED I'LL GO OVER IT. JATS 4 USE, A GROUP OF PUBLISHERS, ARCHIVES, VENDORS AND OTHERS IN SCHOLARLY PUBLICATION INTERESTED IN FINDING COMMON TAGGING PRACTICES THAT WOULD OPTIMIZE THE REUSABILITY OF JATS IN MACHINES. WHY DO WE CARE ABOUT REUSABILITY? WELL, REUSABILITY AFFECTS EVERYBODY. THE LONGER ANSWER, XML DOESN'T JUST GO INTO HTML AND A PUBLISHER'S WEBSITE AND LOOKS PRETTY. IT DOES DO THAT BUT IT GOES A LOT OF OTHER PLACES. IT GOES TO DIGITAL CATALOGS, IT GOES TO AGGREGATOR SITES, ARCHIVES, DATABASES, TO GOOGLE SCHOLAR, THERE'S A LOT OF PLACES IT GOES I PROBABLY CAN'T EVEN IMAGINE. ALL OF THESE PLACES HANDLED XML MACHINE SYSTEMS THAT HAVE TO BE PROGRAMMED BY HUMANS ULTIMATELY BUT THEY ARE SYSTEMS AND THOSE SYSTEMS NEED TO BE ABLE TO RELIABLY FIND WHATEVER PIECES OF INFORMATION IN THE XML THAT THEY NEED TO DO WHAT THEY ARE SUPPOSED TO DO WITH THAT ARTICLE XML. SO IN OTHER WORDS, XML GETS REUSED LOTS OF PLACES FOR LOTS OF REASONS. BUT JATS IS A VERY FLEXIBLE STANDARD, AND WE'VE HEARD LOTS OF TALK OVER THE LAST DAY AND A HALF OR SO ABOUT THE MANY WAYS THAT THINGS CAN BE HANDLED IN JATS. IN JATS THERE ARE MANY ROADS THAT LEAD TO ROME. YOU CAN TAG A GIVEN OBJECT IN MANY WAYS AND STILL HAVE IT BE VALID JATS-XML. HERE IS A CONCRETE EXAMPLE OF A WAY, MORE THAN ONE WAY, TO TAG A LICENSE, FOR EXAMPLE, LET'S SAY FOR AN OPEN ACCESS ARTICLE. AND THERE ARE OTHER WAYS OF DOING THIS WHICH WOULD ALSO BE VALID BUT HERE IS TWO. AND THE POINT HERE IS THAT A SYSTEM THAT NEEDS TO BE ABLE TO FIND THE URL FOR THE LICENSE THAT APPLIES TO THIS CONTENT SO IT KNOWS IT CAN ACCESS THAT CONTENT AND REUSE THE ENTIRETY OF IT OR MAYBE IT'S JUST THE FIGURE IN THAT ARTICLE OR WHATEVER, IT NEEDS TO BE ABLE TO FIND THAT. IT'S NOT USEFUL OR IT'S A LOT MORE CHALLENGING LET'S SAY, A LOT LESS EFFICIENT, IF THAT SYSTEM IS GOING TO ENCOUNTER MANY DIFFERENT WAYS OF PRESENTING THAT INFORMATION. SO IT'S A LOT BETTER IF THERE'S A SYSTEMATIC WAY, PREDICTABLE WAY OF FINDING THE INFORMATION. I ACTUALLY HAVE -- JUST BY THE BY, THIS IS OUR JATS4R'S BRAND SPANKING NEW WEBSITE, JUST BORN JUST THIS LAST FRIDAY, I'LL GET TO THE WHERE AND WHY IN A SEC. ON THIS SITE THERE'S A LINK ON THE ACTUAL RECOMMENDATION FOR LICENSING BECAUSE WE HAPPEN TO HAVE RECOMMENDATIONS FOR THAT. I JUST WANT TO MAKE SURE THAT I TRANSMIT THE IDEA THAT THIS IS NOT JUST FOR OPEN ACCESS CONTENT. REUSABILITY AFFECTS EVERYBODY WHO HANDLES CONTENT IN SCHOLARLY PUBLISHING BECAUSE ULTIMATELY, SYSTEMS THAT AFFECT ALL OF OUR CONTENT NEED TO BE ABLE TO PROCESS, EXCHANGE, STORE, RETRIEVE, REUSE, ALL OF THAT CONTENT AND THAT'S WHY COMMON TAGGING PRACTICES ARE IMPORTANT. SO WHAT HAVE WE BEEN DOING FOR THE LAST YEAR WHICH IS SUPPOSED TO BE THE POINT OF THIS UPDATE. WE'VE BEEN DEVELOPING RECOMMENDATIONS. SO WE HAVE A NUMBER OF RECOMMENDATIONS, I DON'T KNOW IF YOU CAN SEE THAT. SO FAR WE HAVE GENERAL RECOMMENDATIONS, I'LL PULL THAT UP, ON HOW TO STRUCTURE AN XML DOCUMENT SO IT'S A GOOD SENSIBLE WELL-FORMED XML DOCUMENT AND THIS PAGE IS NOT SO MUCH ABOUT JATS AS IT IS JUST ABOUT MAKING A REALLY GOOD SOLID XML DOCUMENT. BUT WE ALSO HAVE FOR THOSE OF YOU WHO LISTENED IN, PROBABLY EVERYBODY HERE DID ON DATA CITATIONS YESTERDAY, INTERESTED IN CITING THAT PROPERLY, WE HAVE WORKED AND FINALIZED A DOCUMENT ON DATA CITATION RECOMMENDATION, SO THIS WOULD BE OUR RECOMMENDATIONS FOR PRACTICES TO TAG DATA CITATIONS SO SYSTEMS CAN RELIABLY FIND WHAT THEY NEED TO FROM THAT CITATION. OKAY. SO I'LL JUST SHOW YOU, THERE ARE -- WE HAVE A PAGE THAT SHOWS OUR CURRENT AND NEXT UP RECOMMENDATIONS, SO CURRENTLY WE'RE WORKING ON VERSIONING FOR ARTICLES, SO BY VERSIONING I MEAN HOW TO INDICATE WHICH VERSION OF AN ARTICLE A SYSTEM IS LOOKING AT AT A GIVEN TIME. SO IF I HAPPEN TO FIND SOME ARTICLE XML AND GO, OKAY, IS THIS THE VERSION OF RECORD, THE JUST ACCEPTED, WHAT AM I LOOKING AT, SO IT'S A WAY TO INDICATE THAT IN THE XML, AND THEN WE THINK WE'VE DECIDED THAT WE'RE GOING TO WORK ON AUTHORS AND AFFILIATIONS, CAPTIONS AND TITLES LIKE FOR EXAMPLE FIGURES AND TABLES, THAT KIND OF THING. FUNDING, CUSTOM META, KEY WORDS AND SUBJECT. OUR GOAL IS TO PUBLISH FOUR RECOMMENDATIONS PER YEAR. SO WE'VE BEEN A BIT SLOW ON THAT IN THE LAST YEAR SIMPLY BECAUSE WE REALIZE THAT THERE'S A LOT OF STUFF THAT HAS TO BE PUT IN PLACE BEFORE YOU CAN KIND OF -- THE RECOMMENDATIONS IS ONLY SORT OF THIS SURFACE PART OF WHAT YOU SEE THE WORK FOR JATS4R, THERE'S A LOT OF WORK THAT GOES INTO ORGANIZING THE GROUP, AND THE OTHER THING WE REALIZED IS THAT AS XML BECOMES SOMETHING THAT IS MORE AND MORE STANDARDS, IT BECOMES MORE IMPORTANT WE BELIEVE FOR PUBLISHERS, ESPECIALLY TO BE A LOT MORE FAMILIAR WITH AND COMFORTABLE WITH THEIR OWN XML, BECAUSE AS WE'VE HEARD MANY TIMES, CONSISTENCY IS VERY, VERY IMPORTANT. XML IS ALL ABOUT CONTENT. IT'S NOT REALLY -- YES, IT'S TECHNICAL BUT AT ITS HEART IT'S ABOUT CONTENT AND THE PEOPLE WHO KNOW THAT CONTENT THE BEST ARE OFTEN SOME OF THE BEST RESOURCES FOR LOOKING AT XML, ONLY XML'S KIND OF SCARY FOR A LOT OF PEOPLE. SO WE HAVE MADE AN XML LEARNING CENTER, AND WHAT WE HOPE TO ACCOMPLISH WITH ANY IS WE HAVE TWO PAGES. WE HAVE FREQUENTLY ASKED QUESTIONS PAGE AND ANYBODY CAN ASK QUESTIONS BY CONTACTING US, BUT MORE IMPORTANTLY WE HAVE -- WE'RE GOING TO START AN XML TOPICS PAIN. THIS WILL BE -- SORRY, ARTICLE SECTION. THIS IS GOING TO BE A SECTION OF XML ARTICLES ON ALL SORTS OF TOPICS FROM VERY BASIC THINGS TO, YOU KNOW, WHY WE EVEN USE XML OR JATS-XML, MUCH MORE TECHNICAL ARTICLES ON MAYBE XML-RELATED TECHNOLOGIES SO A RESOURCE FOR ANYONE WHO WANTS TO LEARN AT THEIR OWN PACE AND MAYBE DOESN'TLY HAVE THE -- DOESN'T NECESSARILY HAVE THE TIME AT WORK OR DOESN'T WANT TO BOTHER THE TECHNICAL PEOPLE THERE. WE WANT XML TO BE ACCESSIBLE FOR ANYBODY WHO HAS TO TOUCH IT SO THEY CAN HELP MAKE DECISIONS AND MAYBE FREE UP SOME OF THE MORE TECHNICAL RESOURCES IN THEIR ORGANIZATION TO WORK ON OTHER THINGS. SO THAT'S WHAT WE HAVE BEEN DOING, AND OF COURSE I MENTIONED THAT WE HAVE THIS WEBSITE AND THE PURPOSE OF THE SITE IS TO GIVE YOU A ONE-STOP-SHOP FOR ALL INFORMATION ABOUT JATS4R AND WHAT IT'S DOING, ALL OF THE RECOMMENDATIONS AND A PLACE TO GO AND LEARN ABOUT XML IF YOU HAVE NOTHING ELSE TO DO. SO I SORT OF LIED BECAUSE I SAID THIS WAS AN UPDATE ON JATS4R BUT THAT'S ONLY PARTLY TRUE. IT'S ALSO A SHAMELESS APPEAL FOR PARTICIPATION BECAUSE THE THING ABOUT COMMON TAGGING PRACTICES IS THEY ARE ONLY COMMON IF LIKE LOTS OF PEOPLE DO THEM AT THE SAME TIME. SO THAT'S WHY WE HAVE A PART OF OUR WEBSITE DEVOTED TO TELLING YOU ALL ABOUT HOW YOU CAN HELP US. THERE ARE A LOT OF PEOPLE HERE WHO ARE PART OF -- I SHOULD GO BACK TO THE HOME PAGE BECAUSE RATHER THAN - WE HAVE A NUMBER OF ORGANIZATIONS, MANY OF WHICH WHOSE REPRESENTATIVES HERE ARE HERE TODAY, HERE IS JUST A FEW OF THE SORT OF MORE ACTIVE ORGANIZATIONS THAT ARE PARTICIPATING IN JATS4R AND MAKING RECOMMENDATIONS. SO I SHOULD HAVE MENTIONED AT THE BEGINNING AND WAS REMISS BUT MELISSA HARRISON FROM E-LIFE WHO YOU HEARD FROM AND MIKE EDEN FROM CAMBRIDGE UNIVERSITY PRESS IS HERE, FROM ATYPON AND MR. CHRISTOPHER MALONEY FROM NCBI, TO NAME A FEW PEOPLE HERE SUPPORTIVE OR ACTIVELY WORKING IN JATS4R. SO HOW TO PARTICIPATE? YOU THOUGHT I FORGOT ABOUT THAT, HMM? ONE THING WE REALLY NEED IS SAMPLES, BECAUSE WE DON'T WANT TO JUST DREAM UP RECOMMENDATIONS FOR THINGS OUT OF OUR OWN HEADS, HOW WE SHOULD DO THEM. WE'D LIKE TO LOOK AT HOW MOST PUBLISHERS ARE TAGGING THINGS IN THE WILD AND TRY TO USE ANY EXISTING CLEARLY COMMON TRENDS AND HOW PEOPLE ARE TAGGING THINGS TO SORT OF BASE OUR DECISIONS WHEN WE START TALKING ABOUT RECOMMENDATIONS. SO TO DO THAT WE NEED SAMPLES AND JUST TO MAKE IT EASY YOU CAN SEND SAMPLES I BELIEVE TO OUR -- THERE'S A WAY TO UPLOAD ON THE GITHUB REPOSITORY, TO MAKE IT EASY FOR THIS PARTICULAR DAY YOU CAN E-MAIL ANY SAMPLES OR TELL US YOU'D LIKE TO SEND SAMPLES BY E-MAILING JATS 4R@GMAIL.COM. IF YOU'RE A PERSON WHO IS KEEN ABOUT AUTHORS AND AFFILIATIONS AND YOU WANT TO WEIGH IN, WE WILL ANNOUNCE AHEAD OF TIME WHAT THE CALL IS GOING TO BE TALKING ABOUT SO THAT YOU CAN CHOOSE TO JOIN OR NOT JOIN. BUT IF YOU'RE TOO SHY TO JOIN THE CALL OR YOU DON'T REALLY -- IT DOESN'T FIT WITH YOUR TIME SLOT THEN YOU CAN GO SUBSCRIBE TO OUR GOOGLE E-MAIL LIST, THESE ARE THE LINKS BY THE WAY, AND YOU CAN ACTUALLY -- YOU'LL BE ABLE TO FIND OUT WHERE THE LINKS ARE TO THE PUBLIC GOOGLE DOCS WHERE WE DEVELOP THE RECOMMENDATIONS. AND FROM THERE YOU CAN HELP OUT BY COMMENTING. AND REVIEWING ON YOUR OWN TIME. YOU DON'T HAVE TO DO IT WHEN WE'RE TALKING ABOUT IT. CAN YOU GO THERE ON YOUR OWN TIME AND TAKE YOUR TIME AND LOOK THROUGH WHAT'S THERE AND SAY, WELL, THIS MAKES SENSE, THIS DOESN'T BECAUSE WE DO THIS, OR WHATEVER. OR EDIT, IF YOU SEE SOMETHING THAT'S NOT -- FACTUALLY INCORRECT, ALL EDITS ARE WELCOME. YOU CAN CONTRIBUTE WORK ON THE VALIDATOR TOOL. THE VALIDATOR TOOL IS OUR TOOL THAT IS BASED ON SCHEMATRON RULES, WHICH IT'S THERE FOR THE USER TO CHECK XML FOR -- I HATE THE WORD COMPLIANCE, IT SOUNDS VERY SERIOUS, BUT JUST TO SEE HOW YOUR XML STANDS UP AGAINST THE RECOMMENDATIONS THAT JATS4R HAS MADE BECAUSE WHEN WE MAKE A RECOMMENDATION WE ALSO MAKE A VALIDATOR RULE TO GO WITH IT. IT MIGHT NOT BE A FAIL, JUST INFORMATION, HEY, FYI, YOUR XML IS THIS AND THIS IS WHAT WE RECOMMEND KIND OF THING OR IT MIGHT BE THAT IT'S ACTUALLY A FAILURE OR JUST A WARNING. SO THERE'S ALL OF -- WAY MORE INFORMATION ABOUT ALL THIS STUFF ON THE WEBSITE SO I'M NOT GOING TO GO INTO IT RIGHT NOW BECAUSE I PROBABLY ONLY HAVE LIKE ONE MINUTE LEFT, AND SO THIS IS -- SORRY, THE LAST THING YOU CAN DO TO HELP IS TO FOLLOW US ON TWITTER BECAUSE WE REALLY NEED TO HAVE -- TO SPREAD THE WORD ABOUT THIS. THERE'S A LOT OF PEOPLE WHO ARE ALREADY PARTICIPATING AND SUPPORTING BUT WE FEEL THIS IS REALLY IMPORTANT BECAUSE COMMON TAGGING PRACTICES THAT MAKE SENSE FOR MACHINE SYSTEMS MAKE ACCEPTS FOR ALL OF US. AND THAT'S IT. [APPLAUSE] >> I WANT TO RUN BACK HERE BECAUSE I HAVE A QUESTION. >> HI, LAURA RANDALL FROM NLM. MY QUESTION IS HOW DOES THE GROUP DECIDE ON WHAT RECOMMENDATIONS TO WORK ON? >> WELL, THAT'S A GOOD QUESTION BECAUSE WE JUST -- WE HAD A MEETING THE DAY BEFORE THE CONFERENCE HERE AT NLM AND THAT'S ONE OF THE THINGS WE TALKED ABOUT. WE DECIDED THAT ON A RUBRIC, ACTUALLY, WE HAVE FOUR CRITERIA THAT WE USE. ONE IS HOW UBIQUITOUS IS THE OBJECT WE'RE CONSIDERING. BY OBJECT FOR EXAMPLE I MEAN PERMISSION, CAPTIONS ON FIGURES AND TABLES OR AUTHORS AND AFFILIATIONS, THINGS LIKE THAT, OBJECTS. SO HOW COMMON IS THAT OBJECT AMONG PEOPLE'S XML? OBVIOUSLY AUTHORS AND AFFILIATIONS OR SOMETHING LIKE CITATIONS, EVERYBODY HAS THOSE. SO THAT'S A VERY -- THAT GIVES IT SOME POINTS THERE BECAUSE IT'S VERY COMMON AND AFFECTS EVERYBODY. SO ANOTHER CRITERION WOULD BE HOW MUCH VARIATION IS THERE IN JATS FOR TAGGING THAT OBJECT, SO MAYBE IT'S SOMETHING THAT JUST HAPPENS IN JATS OR JUST HAPPENS TO BE ONLY ONE OR TWO WAYS TO DO IT, SO IT'S NOT THAT BAD, BUT SAY FOR AUTHORS AND AFFILIATIONS WHERE YOU HAVE SIX DIFFERENT ROADS LEADING TO THE SAME PLACE, AND THAT'S SUCH A KEY PIECE OF METADATA IN AN ARTICLE, THAT'S A PROBLEM WHEN YOU HAVE SYSTEMS THAT ARE TRYING TO RELIABLY GET INFORMATION OUT OF THE TAGGING THAT'S AROUND AUTHORS AND AFFILIATIONS. SO SOMETHING THAT HAS NATURALLY A LOT OF VARIATION POSSIBILITY IN JATS WOULD ALSO A HIGH -- A GOOD CANDIDATE FOR RECOMMENDATIONS. AND THEN I THINK WE SAID THAT IF THERE WAS A LOT OF -- A PARTICULAR REALLY STRONG PAIN POINT AROUND A GIVEN OBJECT, MAYBE IT'S NOT THAT COMMON AND MAYBE THERE ISN'T THAT MUCH VARIATION BUT IT'S REALLY, REALLY IMPORTANT AND I CAN'T THINK OF A GOOD EXAMPLE OFF THE TOP OF MY HEAD, BUT PEOPLE ARE TALKING ABOUT IT AND IT'S SOMETHING THAT WOULD RELIEVE A LOT OF PAIN FOR MAYBE A SMALLER GROUP OF PEOPLE, THAT WOULD WEIGH IN. AND THEN JUST TRYING TO THINK OF WHAT THE FOURTH THING IS. I THINK WHETHER IT'S SORT OF A NEW OBJECT. SO IF IT'S SOMETHING -- A CONCEPT THAT'S RELATIVELY NEW IN JATS AND AGAIN I CAN'T THINK OF AN EXAMPLE RIGHT NOW, BUT SOMETHING THAT HASN'T REALLY HAD A CHANCE TO AGE AND MELLOW A LITTLE BIT, JUST TO SEE HOW PEOPLE ARE GOING TO TAG IT, THAT'S SOMETHING WE MIGHT NOT BE AS LIKELY TO MAKE A RECOMMENDATION ON BECAUSE THERE HASN'T BEEN AN ESTABLISHED SORT OF FEELING ABOUT HOW THAT SHOULD BE I GUESS COMMONLY. DOES THAT MAKE SENSE? ANYTHING ELSE? >> HI, (INDISCERNIBLE) FROM JOHN BENJAMIN. I'M A BIT CURIOUS WHAT THE RELATIONSHIP BETWEEN YOUR RECOMMENDATIONS AND THOSE OF THE TAG SETS DOCUMENTATION, WILL THEY BE SYNCED? SO FAR I'VE BEEN SENDING MY INDIAN FRIENDS TO THE TAG SET DOCUMENTATION WHICH IS FULL OF GUIDELINES, BEST PRACTICES, AND EXAMPLES. AND NOW YOU ALSO HAVE GUIDELINES OR RECOMMENDATIONS THAT WOULD BE A STANDARD AT SOME POINT. >> WELL, WE TRY TO BE INFORMED BY WHAT THE TAG LIBRARY SAYS AS MUCH AS POSSIBLE. THAT'S HOW IT'S BEEN SO FAR. SO FOR EXAMPLE A WHILE AGO WE WERE DISCUSSING, SAY, DATE TYPE ATTRIBUTES, WHICH HAVEN'T TURNED INTO ANY KIND OF RECOMMENDATION YET BUT WE WERE TALKING ABOUT THEM. AND WE DECIDED THAT SINCE MANY OF THE DATE TYPE ATTRIBUTES LIKE RECEIVED, ACCEPTED, I CAN'T REMEMBER THE ENTIRE LIST, WHICH ARE SUGGESTED BY THE TAG LIBRARY HAVE BEEN USED FOR SO LONG AND SO COMMONLY THAT IT WOULD BE RIDICULOUS FOR US TO STRAY FROM THAT AND JUST MAKE UP OUR OWN JUST FOR THE SAKE OF BEING DIFFERENT. SO I GUESS THE ANSWER IS THAT WE REALLY TRY TO SORT OF LOOK AT WHAT THE TAG LIBRARY SAYS AND IF IT'S SOMETHING THAT SUGGESTS A COMMON TAGGING PRACTICE WELL I GUESS THAT WOULD BE A SERIOUS FACTOR IN WHAT WE DECIDE TO DO. >> THE TAG LIBRARY THAT WE PUBLISHED ALONG WITH THE SCHEMAS DESCRIBE WHAT YOU CAN DO. WE'RE VERY CAREFUL NOT TO TRY TO SAY WHAT YOU SHOULD DO IN THERE. THE TAG LIBRARIES ARE NOT BEST PRACTICES FOR TAGGING. THERE IS SOME BEST PRACTICE INFORMATION IN THE ESSAYS THAT ARE INCLUDED BUT THE JATS4R GROUP TRIES TO MAKE REALLY A VERY NARROW SET OF RECOMMENDATIONS SO THAT YOU CAN CHOOSE TO FOLLOW THEM OR NOT. BUT IF YOU LOOK AT THEM YOU WILL GET GUIDANCE AND YOU DON'T NECESSARILY GET GUIDANCE FROM THE GENERAL TAG LIBRARIES. >> THANK YOU. [APPLAUSE] >> NEXT UP IS A VOICE WE JUST HEARD FROM, THAT WOULD BE JEFF BECK TALKING ABOUT A JATS COMPATIBILITY MODEL. >> WE HAVE JATS 4R CARDS, IF YOU WANT ONE I'LL GIVE IT TO YOU. I LIKE TO GIVE THEM OUT BECAUSE I FORGET. WE HEARD YESTERDAY ABOUT THE FTS, THE STANDARD TAG SUITE. WE'VE ALSO HEARD FOR THE PAST TWO DAYS ABOUT BITS. SOMEONE EVEN MENTIONED TAX PUB THE OTHER DAY. THESE ARE ALL EXTENSIONS OF JATS. RIGHT NOW JATS IS A NISO STANDARD. THE STS, THE STANDARDS TAG SUITE, IS IN PRODUCTION, I'M NOT SURE I'VE GOT THE WORDS RIGHT. IF IT'S WRONG I HOPE MANY WILL CORRECT ME. IT'S IF PRODUCTION AT NISO. AND THE BTS, BOOK TAG SUITE, IS A PROJECT WE HAVE HERE AT NCBI. I THINK BRUCE MENTIONED THAT THE GROUPS ARE ALL RATHER -- INCESTUOUS IS A BAD WORD. [LAUGHTER] BUT IT'S PROBABLY ACCURATE. AND WHAT WE REALLY WORRY ABOUT IS THAT AS MORE AND MORE EXTENSIONS ARE MADE, THAT THE EXTENSIONS WILL DO THINGS OR MAKE ELEMENTS OR MAKE CHANGES TO EXISTING ELEMENTS THAT MAKE THEM NO LONGER COMPATIBLE WITH JATS. SO A GROUP OF US GOT TOGETHER, IT WAS DEBBY AND TOMMIE FROM MULBERRY, LAURA AND LAN DALE AND MYSELF, AND WE STARTED THINKING ABOUT HOW WE COULD SET UP SOME GUIDELINES FOR IF YOU WANT TO MAKE AN EXTENSION OF JATS, THAT IS COMPATIBLE WITH JATS, THESE ARE THE THINGS YOU NEED TO CHECK. NOW, ANYONE CAN TAKE THE JATS AND MAKE AN EXTENSION AND THERE ARE NO RULES ABOUT WHAT YOU DO AND WHAT YOU'RE GOING TO DO WITH IT. BUT IF YOU'RE LOOKING TO MAKE ESSENTIALLY ANOTHER MEMBER OF THE FAMILY AND I HOPE THAT BITS AND STS WILL ALL TRY TO STAY WITHIN THE JATS FAMILY, THERE'S SOME THINGS YOU CAN DO AS YOU'RE MAKING EXTENSIONS THAT ARE REALLY GOING TO SCREW THINGS UP. SO WE TRIED TO LOOK AT SOME ITEMS AND WRITE DOWN SOME RULES ABOUT WHAT NOT TO DO IF YOU WANT TO STAY COMPLIANT WITH OR COMPATIBLE WITH JATS. AND I HAVE A COUPLE OF EXAMPLES. WE HAVE A NUMBER OF THINGS THAT NEED TO BE CHECKED. BUT SOME EXAMPLES SO THAT IT'S NOT SO OBSCURE, ONE IS YOUR ELEMENT AND YOUR ATTRIBUTE NAMES, ONE OF OUR COMPATIBILITY CHECKS IS IF YOU HAVE -- IF YOU'RE GOING TO USE AN ELEMENT FROM JATS IT SHOULD BE USED FOR THE SAME THING IN WHATEVER YOUR EXTENSION IS. A GOOD EXAMPLE OF THIS IS WE HAVE A STATEMENT ELEMENT IN JATS AND IT'S DESCRIBED AS SOMETHING LIKE STATEMENT IS FOR THEOREMS AND LEMMAS, NOT LEMMINGS, SO STATEMENTS OF THEORY, NOW IF SOMEONE WAS COMING TO MAKE A JATS EXTENSION AND THEY DECIDED, WELL, THEY DIDN'T NEED TO TALK ABOUT THEOREMS, BUT THEY DID NEED TO KEEP TRACK OF STATEMENTS THAT PEOPLE MADE IN COURT UNDER OATH, THEN THAT APPLICATION OF THE ELEMENT NAME, THAT USE OF THE ELEMENT NAME STATEMENT SHOULD NOT BE ALLOWED BECAUSE IT HAS COMPLETELY DIFFERENT MEANING THAN STATEMENT IN THE GREATER SET, IN THE FAMILY OF ELEMENT NAMES. SO YOU COULD JUST MAKE UP ANOTHER ELEMENT NAME LIKE WITNESS STATEMENT OR WHATEVER YOU LIKE, BUT WE DON'T WANT ELEMENT NAMES CRASHING FOR THINGS THAT ARE DIFFERENT MEANINGS. SIMILARLY, IF YOU'RE GOING TO MAKE AN ELEMENT FOR THEOREMS AND LEMMAS AND THINGS LIKE THAT YOU PROBABLY SHOULD USE STATEMENTS BECAUSE IT'S ALREADY BUILT AND HOPEFULLY TOOLS THAT EXIST, ALREADY KNOW HOW TO USE IT, YOU CAN RIDE THAT JATS WAVE WITH YOUR EXTENSION. WE ALSO WANT TO AVOID HAVING ELEMENTS AND ATTRIBUTES WITH THE SAME NAME. WE HAVE ONE OR TWO IN JATS NOW, AND THIS IS SO NEW THAT I'M NOT EMBARRASSED BY THAT. BUT MAYBE NEXT YEAR I'LL BE EMBARRASSED BY THOSE. ANOTHER THING THAT WE LOOK FOR AND THIS IS ONE THAT YOU DON'T REALLY REALIZE MIGHT BE A PROBLEM IS WHITE SPACE IN ELEMENTS. WE WENT FOR THE LONGEST TIME TRYING TO -- WITH A QUALITY -- A TEST ON WHAT THE ELEMENT CONTENT WAS, WHETHER IT WAS ELEMENT ONLY OR ALLOWED TEXT, AND IT TOOK US A WHILE, A NUMBER OF REVISIONS OF OUR DOCUMENT TILL WE REALIZED WHAT WE WERE REALLY TALKING ABOUT THERE IS HOW THE WHITE SPACE IS EXPECTED TO BE HANDLED WITHIN THOSE ELEMENTS. IF YOUR ELEMENT HAS ELEMENT ONLY CONTENT, THEN ANY WHITE SPACE WITHIN IT IS -- IF YOUR ELEMENT IS DEFINED AS ELEMENT ONLY CONTENT, THEN ANY WHITE SPACE WITHIN IT IS, YOU KNOW, WHO CARES, IT CAN COME, IT CAN GO, IT CAN BE HUGE, IT CANNOT BE THERE. IF YOU HAVE TEXT ALLOWED WITHIN YOUR ELEMENT, WHETHER IT'S TEXT ONLY OR MIXED CONTENT, OR EMPTY, BECAUSE WE CONSIDER EMPTY TO BE A TEXT, A TEXT TYPE FIELD, THEN YOU'RE ESSENTIALLY FOLLOWING THE NORMALIZED SPACE RULES THAT WHITE SPACE IS PRESERVED BUT THE WHITE SPACE CHARACTERS AREN'T NECESSARILY PRESERVED. SO EVERYTHING WILL COLLAPSE, ALL WHITE SPACE WILL COLLAPSE INTO A SINGLE WHITE SPACE, SINGLE-SPACE THE CHARACTER, BUT NONE OF THEM ARE GOING TO BE INSERTED AND NONE OF THEM ARE GOING TO BE REMOVED. THE THIRD CASE IS FOR ELEMENTS WHERE WE NEED TO PRESERVE SPACE, LIKE FREE FORMAT AND CODE, AND THOSE ARE ELEMENTS WITH ALL THE WHITE SPACE CHARACTERS MUST PRESERVED. SO IF YOU HAVE AN ELEMENT, IF WE HAVE AN ELEMENT IN JATS THAT IS A MIXED CONTENT MODEL, YOU REALLY SHOULDN'T DEFINE IT, PRE-DEFINE IT IN YOUR EXTENSION, AS AN ELEMENT ONLY MODEL BECAUSE ANY SYSTEM THAT'S TRYING TO DEAL WITH IT IS GOING TO SCREW THE WHITE SPACE UP SOMEWHERE. SO WE'VE WRITTEN SEVERAL DRAFTS OF THIS AND FLOATED THEM AROUND TO A COUPLE OF PEOPLE WHO MAY OR MAY NOT BE IN THIS ROOM. WE'VE GOTTEN SOME GOOD FEEDBACK. WE'RE GOING TO WORK ON ANOTHER REVISION AND HOPEFULLY HAVE SOMETHING FOR THE COMMUNITY, I THINK WE SAID SHORTLY, SO YOU CAN READ INTO THAT WHATEVER YOU WILL. SO I'M NOT SURE HOW WE'RE GOING TO ANNOUNCE IT, BUT IT SHOULD BE AVAILABLE TO THE PUBLIC. SO ANY QUESTIONS? YES? >> A COMMENT. THIS IS EVAN. I WAS ONE OF THE ONES ASKED TO REVIEW THIS DOCUMENT. THIS IS A MAGNIFICENT CONTRIBUTION TO THE COMMUNITY. WE SHOULD ALL BE VERY GRATEFUL TO THE FOUR OF THEM, THE WORK THEY HAVE DONE ON THIS, AND WHAT THIS WILL MEAN FOR THE FUTURE OF THE JATS COMMUNITY. SO THANK YOU >> OH, THANK YOU, EVAN. IT WAS A MIND-BENDING EXPERIENCE. [LAUGHTER] THANK YOU. [APPLAUSE] >> MIND BENDING AND SLIGHTLY MIND NUMBING. I NICKNAMED THE PROJECT NUMB FOR THE NISO UBER META MODEL. IT'S CHANGED NAMES AND HAS MORE MEANINGS. NEXT UP SPEAKING OF BENDING THINGS, I ASKED CHARLES O'CONNOR WHO IS OUR NEXT SPEAKER WHAT HE WANTED TO TALK ABOUT, AND HE SAID HE'S THROWING A GRENADE. I THOUGHT, NO. AND THEN HE SAID, BENDING JATS. SO CHARLES. >> OKAY. I'M CHARLES O'CONNOR. I WORK FOR FAREDDEN JOURNAL SERVICES, FORMERLY KNOWN AS DARTMOUTH JOURNAL SERVICES. WE RECENTLY CHANGED OUR NAME. WE ARE IN THE ECOSYSTEM OF THIS COMMUNITY. WE'RE A VENDOR, LOWEST ON THE TOTEM POLE. WE'RE A FILE SUPPLIER, SUPPLYING FILES TO A RANGE OF HOSTS. PROBABLY EVERY HOST OUT THERE. WE FOUND THAT WE GET SOME ODD REQUESTS FROM HOSTS AND I THINK MY PURPOSE HERE IS TO RAISE SOME AWARENESS ABOUT SOME OF THE THINGS THAT WE'RE ASKED TO DO AND HOPEFULLY, YOU KNOW, GET SOME COMMUNITY SUPPORT TO TIGHTEN THINGS UP AND I THINK JATS 4 REUSE IS A GREAT START, BY THE WAY. ANYWAY, SOME OF THE THINGS WE'RE ASKED TO DO, AND I THINK MY COLLEAGUE ISLA REEBER, I THANK HER FOR GATHERING THESE FOR ME. THESE ARE A SMALL CHAMBER OF HORRORS. FOR EXAMPLE, FOR CORRESPOND CAN GO AUTHOR DATA WE HAD A CUSTOMER WHO DIDN'T WANT IT TO APPEAR WITH THE ABSTRACT, THE ONLINE HOST SAID MAYBE DON'T GIVE IT TO US. AND WE THOUGHT WASN'T A GOOD IDEA. THEN THEY SUGGESTED THAT WE PUT THE CORRESPONDING AUTHOR DATA AS A BACK MATTER FOOTNOTE. SIMILARLY, AUTHOR AFFILIATIONS, CUSTOMER WANTED THEM TO APPEAR AT THE BOTTOM OF THE DOCUMENT AND WE WERE TOLD, HEY, WHY NOT PUT THEM AS BACK MATTER FOOTNOTES AGAIN. BY THE WAY, I SHOULD POINT OUT THESE ARE EXAMPLES DRAWN FROM ALL SORTS OF ONLINE HOSTS, NO ESPECIALLY BAD GUYS OR ESPECIALLY GOOD GUYS IN THIS SPACE. LONG SEQUENCES OF CHARACTERS, WE'VE BEEN ASKED SOMETIMES TO -- BECAUSE BUNK SEQUENCES CAN BE A BIT OF A PAIN, TO THROW ZERO WITH PACES IN THE XML. HEY, THAT WOULD BE GREATS IN THE HTML IF YOU NEEDS TO BREAK THE LINE BUT IT REALLY DOESN'T BELONG IN THE XML. VERY COMMONLY, YOU KNOW, WE'RE ASKED TO MAKE AN ENTIRE TITLE BOLD WHICH DOESN'T MAKE SENSE. SOMETIMES WE'RE ASKED TO ADD LINE BREAKS WHERE THEY DON'T BELONG AND EVEN I THINK THE JATS RECOMMENDATION IS TO NOT USE BREAK EXCEPT IN A TABLE HEAD OR OTHER PLACES. WE HAD A CONTINUOUS PUBLISHING WORKFLOW WHERE WE HAPPENED TO HAVE THE FOLIO DATA, EVEN THOUGH IT WAS CONTINUOUS PUBLISHING ON AN ARTICLE BASIS, SO SINCE IT WAS SHOWING UP IN THE ETOC, AND THAT WASN'T, YOU KNOW, DESIRED AT THE ARTICLE STAGE BUT WAS DESIRED AT THE ISSUE STAGE, WELL, WHY NOT JUST NOT GIVE IT TO US. SO, YOU KNOW, THIS HAS AN IMPACT ON US IN A LOT OF WAYS. THE IMPACT, THE CONSEQUENCES FOR FILE SUPPLIERS IS THAT IT'S DIFFICULT TO MAINTAIN DOCUMENTATION, IF YOU HAVE, YOU KNOW, HOST X WANTS NOT JUST THIS WAY BUT THIS WAY, THIS WAY, THIS WAY AND THIS WAY, IT'S HARD TO DOCUMENT THAT, KEEP THAT UP TO DATE. AND THAT OF COURSE CAN LEAD TO ERROR BECAUSE, YOU KNOW, IT'S, YOU KNOW, WE DON'T HAVE ONE STANDARD TO HIT. AND EVEN MORE SO CONTRIBUTING TO ERROR IS THAT WE CAN'T REALLY BUILD MODULAR TOOLS FOR QA'ING. WE SUPPLY -- WE HAVE A LOT OF PUBLISHERS THAT WE SUPPLY FILES FOR, A LOT OF ONLINE HOSTS WE SUPPLY FILES TO, AND SO IF THERE'S SUCH VARIATION IN THE TARGET XML AND IT'S BY PUBLISHER, BY PUBLISHER, BY PUBLISHER, THEN IT'S REALLY DIFFICULT FOR US TO BUILD TOOLS FOR QA. LAST YEAR I PRESENTED ON ONLINE WEB EDITING, JATS XML, WHATEVER SYSTEM, AND I CAN TELL YOU THAT FROM THE PERSPECTIVE OF SOMEBODY WHO BUILDS TOOLS, IT'S VERY DIFFICULT TO BUILD TOOLS THAT DO SOMETHING THAT DOESN'T MAKE SENSE. IT'S EASIER TO MAKE -- [LAUGHTER] IT'S EASIER TO MAKE A TOOL THAT DOES SOMETHING THAT MAKES SENSE. I DON'T KNOW IF THAT MAKES SENSE. AND I'M SURE, YOU KNOW, THE RESPONSE FROM THE PEOPLE HERE WHO ARE NOT FILE SUPPLIERS OR VENDORS MAY BE, WELL, THAT'S YOUR PROBLEM, AND WHO REALLY CARES, BUT THERE ARE CONSEQUENCES FOR PUBLISHERS AS WELL. INFORMATION CAN BE LOST OR MISIDENTIFIED, IF IT'S, YOU KNOW, IF SOMETHING'S IN A FOOTNOTE INSTEAD OF IN A CORRESPONDENCE LINE THAT'S NOT WHERE A SYSTEM WOULD BE LOOKING FOR IT, CAN HAVE XML CONTAINING UNNECESSARY FORMATTING, IT MAKES THE XML LESS PORTABLE. YOU MAY NOT WANT TO FOREVER BE WITH ONLINE HOST X, YOU MAY WANT TO MOVE TO Y AT SOME POINT. IT'S ALL OF YOUR -- AND I JUST WANT TO MAKE ONE POINT. IT'S NOT THAT AN ONLINE HOST X HAS THESE TAGGING REQUIREMENTS AND Y HAS THOSE. THAT'S NOT SO BAD. IT'S THE VARIATION WITHIN THEM. SO IF YOU ARE PORTING FROM ONE ONLINE HOST TO ANOTHER, YOU KNOW, IT WOULD BE NICE IF THAT WERE A PREDICTABLE AND ROBUST WAY TO DO THAT. ALSO, THERE CAN BE PROBLEMS WITH DOWNSTREAM DEPOSITS. WE'VE HAD TIMES WHERE WE, YOU KNOW, WE DO WHAT WE'RE TOLD, WE TRY TO BE GOOD ABOUT IT. IT HITS PMC AND THEY GO, WHAT'S THAT? [LAUGHTER] AND THAT'S -- SO THAT'S, YOU KNOW, WE HAVE PROBLEMS WITH DOWNSTREAM DEPOSITS SOMETIMES, OBVIOUSLY WE BOW TO TMC. IT MAKES THE XML LESS SEARCHABLE AND MACHINE SEARCHABLE AND ALL OF THOSE POINTS MADE IN THE JATS FOR REUSE, IF THE MACHINE IS LOOKING FOR CORRESPONDENCE LINE IN A CORRESPONDENCE ELEMENT, THAT'S WHERE IT'S LOOKING. IT'S NOT GOING TO GO MINING THE FOOTNOTES TO FIND THEM. SO I HAVE A LITTLE BIT OF A PLEA IS TO PUBLISH THOSE TAGGING GUIDELINES. LET US KNOW WHAT YOU WANT, CODIFY THEM, AND POINT NUMBER TWO, STICK TO THEM. I HAVE A CUSTOMER WHO WANTED TO DO INLINE SUPPLEMENTS. THERE WERE TAGGING GUIDELINES ON HOW TO DO INLINE SUPPLEMENTS. THIS CUSTOMER SAID, OKAY, SO THIS IS WHAT YOU WANT, THIS IS HOW YOU'RE DOING IT FOR PUBLISHER Y OVER HERE. I SEE THEY HAVE ONLINE SUPPLEMENTS. AND THE RESPONSE WAS, NO, THAT'S NOT HOW THAT OTHER PUBLISHER IS DOING IT. THEY ARE DOING IT A COMPLETELY DIFFERENT WAY. AND SO STICKING TO THEM IS KEY. YOU KNOW, XSLT IS A GREAT TOOL. IT CAN REARRANGE THINGS. IT CAN FORMAT THINGS. AND THAT'S REALLY THE WAY TO KEEP DISPLAY OVER HERE AND CONTENT OVER HERE SO WE KNOW THAT'S NOT REALLY, YOU KNOW, THERE'S OVERLAP THERE. AND BUILDING TOOLS LIKE THE PMC STYLE CHECKER AND JATS 4 REUSE. WE HAVE A LAUNCH FEATURE FROM EXPORT THAT YOU CAN USE TO THEN HAVE THAT XML GO THROUGH THE PMC STYLE CHECKER, WHICH IS GREAT BECAUSE IT'S A GREAT TOOL. IT FINDS ALL SORTS OF THINGS THAT YOU DON'T SEE IN A WORD FILE LIKE SOMEBODY HAS LIKE A MATH TYPE OBJECT AND THEN AN EMPTY MATH TYPE OBJECT RIGHT NEXT TO IT, THE PMC STYLE CHECKER SAYS YOU HAVE TWO MATH ELEMENTS WITHIN ONE DISPLAY EQUATION, AND, YOU KNOW, IT'S INVISIBLE IN THE WORD FILE. SO TOOLS LIKE THAT ARE FANTASTIC AND WOULD MAKE FOR BETTER TAGGING FOR ALL OF US. THAT'S MY PLEA. [APPLAUSE] >> WENDELL. >> WENDELL PIEZ, OLD TIME RELIGION NEVER HURTS. I SUMMARIZE WHAT YOU TOLD US, FORMATTING CONTENT, YOU KNOW, WE KNOW THEY GO LIKE THIS, BUT NO, NO, NO, KEEP LIKE THIS THIS, RIGHT? IN OTHER WORDS, I THINK ALL THE VIALS US DESCRIBED ARE VIOLATIONS TO THE RULE OF HOW TO ARCHITECT LAYERS OF SYSTEMS TO OPERATE THEM SUSTAINABLY. AND WHEN YOU VIOLATE A RULE THAT, YOU KNOW, THAT MESSES WITH THE ARCHITECTURE, THINGS BEGIN TO FALL OFF. AND WE SEE THAT. AND WE LEARN IT FROM EXPERIENCE. AND A LOT OF THE FRUSTRATION WE EXPERIENCE COMES FROM PEOPLE HAVING LEARNED THE OLD LESSON. KNOWING THAT, CAN YOU TELL YOUR CUSTOMERS THERE ARE PEOPLE WHO HELP WITH THIS. I KNOW PEOPLE THAT WOULD LOVE TO HELP THEM WITH THEIR PROBLEMS INTERNALLY. THEY ARE NOT SAYING CAN YOU FIX SOMETHING THERE'S A REAL BETTER SOLUTION. >> EVERY SINGLE ONE IS A DISPLAY VERSUS CONTENT ISSUE, AND WE HAVE TO RESIST THE TEMPTATION TO GET DISPLAY OUT OF WHAT SHOULD BE CONTENT. THE MORE SUPPORT FROM THE COMMUNITY AND OTHERS, THE PUBLISHERS CAN ACCESS BECAUSE WE ALL HAVE AN INTEREST IN THE XML BEING CORRECT AND ROBUST. >> I SOMETIMES LIKE TO QUOTE AN OLD LINE, THE REASON WE OUR DISCIPLINE IS SO STRICT IS LIFE MAY BE EASY. >> INDEED. ALL RIGHT. >> NEXT IS CHUCK KOSHER, LOOKING AROUND, THERE HE IS. WHO APPARENTLY HAS A CORRECTION TO MAKE. >> THIS WILL BE QUICK. CrossRef DOES NOT REQUIRE ORCIDs TO BE VALIDATED WHEN YOU DEPOSIT THEM AND ENABLE CrossRef TO AUTOUPDATE PUSH THEM INTO THE AUTHOR'S PROFILE. SO I STEPPED OUT YESTERDAY WHEN THAT WAS MENTIONED, CAME BACK IN, MISSED THE WHOLE THING AND THEN HAVE ANSWERED THAT ABOUT FIVE TIMES SINCE YESTERDAY SO I'M MAKING A BLANKET STATEMENT RIGHT NOW. THANK YOU. [APPLAUSE] >> OKAY, NEXT UP IS PETER PLUMMER, DID I SAY IT RIGHT? HE HAS SOME ISSUES TO POINT OUT WITH THE JATS. >> IT'S MY FIRST TIME AT THIS CONFERENCE. ALSO AS A SPEAKER. I'M GOING TO TAKE FIRST THE POSSIBILITY TO HAVE A SIP. [LAUGHTER] JOHN BENJAMIN'S PUBLISHING, THAT'S ABOUT 70 JOURNALS PER YEAR, 150 BOOKS, AND WE'VE BEEN DOING PDF ONLY SO FAR, APART FROM METADATA EXPORTS AND JATS TO EVERYWHERE. SO I THOUGHT AT SOME POINT WE MAY NEED TO DO FULL TEXT, AND LAST YEAR ONE OF OUR JOURNALS TARGETS TRANSLATION JOURNAL (INDISCERNIBLE) TARGETS WHAT'S GOING TO GET COMMUNITY BASE TRANSLATIONS ASIDE, SO ALL THE ARTICLES COULD BE TRANSLATED. AND SHOULD BE HOSTED NEXT TO THE ORIGINAL. BUT THE ORIGINAL WOULD BE PDF FROM 25 YEARS OLD OR YOUNGER, AND THEN THE TRANSLATION WOULD BE ASML. THAT'S NOT NICE. I SAID LET'S TAKE THIS JOURNAL AS A PILOT TO SEE IF WE CAN DO FULL TEXT XML IN JATS AND IF WE CAN FIND THE RIGHT TAGS FOR OUR LITTLE PROBLEMS THAT WE MIGHT HAVE IN LINGUISTICS, ACTUALLY TARGETED NOT THAT DIFFICULT LINGUISTICALLY BECAUSE IT'S ABOUT TRANSLATION STUDIES, WHICH IS NOT SO THEORETICAL. BUT WE DID ENCOUNTER A COUPLE THINGS, AND ONE IS ABOUT PAGE NUMBERS. I HAVE THREE ISSUES, PAGE NUMBERS IS THE MAIN ISSUE. I KNOW THAT THE JATS COMMITTEE DOESN'T LIKE TO HAVE A PAGE NUMBER TAG BECAUSE IT'S ABOUT PRINT AND NOT ABOUT DIGITAL DOCUMENT. IT'S EVERYWHERE. AND PRINT IS FORMATIVE BUT THE FIRST AND LAST PAGE ARE PART OF THE PRINT DOCUMENT AND THEY ARE TAGGED IN THE METADATA. WE NEED A PAGE TAG IF WE WANT TO USE IT IN EPUB. WE NEED A PAGE TAG IF WE WANT TO MAKE A LINK FROM ANOTHER ARTICLE. SO SOMEBODY REFERS TO ANOTHER TARGET ARTICLE, THEN YOU HAVE A RELATED ARTICLE WITH A PAGE I.D., AND THEN WHERE DOES IT GO IF YOU CANNOT PUT A PAGE TAG THERE? SO FIRST WE HAD RELATED OBJECTS AS AN I.D. TO PUT IT IN RELATED OBJECTS. THAT WORKS BUT IT FEELS DIRTY. THEN WHEN WE DID A BOOK PILOT ON THE SIDE VENDOR CAME UP WITH RELATED ARTICLE WHICH IS SUPPOSED TO BE ABOUT ANOTHER ARTICLE. THEN WE FOUND IN THE DISCUSSIONS THERE'S SOMETHING CALLED TARGET. SOUND GOOD. BUT YOU CAN'T PUT IT EVERYWHERE, WHERE YOU WANT TO PUT IT. AND IT'S NOT MEANT TO BE FOR PAGE NUMBERS BECAUSE IT CAN BE A TARGET. BUT ALSO PLAINLY METADATA ITEM WITHIN THE BODY. BUT THEN THERE'S ANOTHER USE, SAY FOR INSTANCE THAT YOUR TEXT BOOK, WE'RE NOT TALKING BITS, AND YOU HAVE GO TO PAGE 200, SO EVERYBODY WHO MIGHT HAVE THE WEB-BASED VERSION THAT COULD BE BITS-BASED WOULD BE IN THE DARK, IF YOU DON'T HAVE PROPER PAGE I.D. SOMEWHERE. AND SO I HAVE THREE SOLUTIONS NOW. THEN MIKE HAD ANOTHER SOLUTION. FOUR SOLUTIONS. I THINK THE ABSENCE OF GOOD GUIDELINE AS TO HOW THE PAGE NUMBER SHOULD BE TAGGED CREATES THE VARIANTS BETWEEN VENDORS AND PUBLISHERS. SO THAT'S MY RANT BECAUSE RANTS ARE WELCOME. THEN I HAD TWO MORE INTERESTING THINGS. SO IF YOU MAKE A TRANSLATION TO FRENCH OR TO GERMAN OFF THE ENGLISH ARTICLE AND HAVE AUTOGENERATED TEXT FOR AUTHOR OR AUTHORS, EDITORS, OR TRANSTRANSLATORS, FRENCH FOR THE FEMALE VERSION AND [SPEAKING FRENCH] FOR THE GERMAN AUTHOR WOMAN, MEANS THAT WE NEED TO CAPTURE THAT INFORMATION SOMEWHERE. SO WE NOW HAVE [SPEAKING IN A FOREIGN LANGUAGE] IS FEMALE. IT DOESN'T REALLY SOUND RIGHT TO SAY THAT ON A NAME, SO I THINK IT'S GOOD. I WILL -- I MAY USE FORM IF NOBODY ELSE DOES AT SOME POINT TO SUGGEST SOMETHING BUT WE NEED GENDER AND I WOULD NOT GO INTO DISCUSSIONS ABOUT A THIRD GENDER OR ANYTHING BUT WE NEED A FEMALE GENDER BECAUSE LANGUAGES OTHER THAN ENGLISH NEED IT FOR LOCALIZATION. THEN MY THIRD ISSUE IS ABOUT LETTERS. OUR TYPE OF JOURNALS, WE OFTEN HAVE LETTER FROM THE EDITOR, LETTER TO THE EDITOR, AND I FOUND THAT THERE IS A SIGNATURE BLOCK FOR THE PERSON, WELL, THANK YOU, BUT THERE'S NO SALUTATION. SO I KNOW THAT THERE IS AN ADDRESS BUT I DON'T THINK ADDRESS WOULD BE APPROPRIATE FOR AN ADDRESS. SO THAT'S AN OPEN QUESTION. HOW WOULD YOU TAG "DEAR AUTHOR." OH, YES, I HAVE ANOTHER QUESTION, WHO IN THE HALL HAS HAD A NEED TO CAPTURE A PAGE NUMBER IN XML? SHOW OF HANDS. FOUR. OKAY. FIVE-ISH. DE GRUYTER, CAMBRIDGE, I DON'T KNOW YOUR PUBLISHER, WE'RE IN GOOD COMPANY. WELL, THAT'S BASICALLY IT. [APPLAUSE] >> THANK YOU. WHAT I WOULD SAY IN RESPONSE IS PLEASE USE THE COMMENT FORM AND MAKE SUGGESTIONS TO THE COMMITTEE SO WE MAY DISCUSS THESE THINGS AND PROVIDE EXAMPLES. >> THAT'S WHAT I LEARNED FROM THIS CONFERENCE. >> IF YOU TAKE NOTHING ELSE AWAY, TAKE THAT AWAY. OKAY. WE'VE DONE AN EXCELLENT TIME, EXCELLENT JOB WITH TIME. WE'RE NOW ON OUR 11th PRESENTER, AND THAT WOULD BE CHRIS MALONEY WHO IS TALKING TO ABOUT A DPD DIAGRAM. >> THANKS. IT'S NOT REALLY A TALK. I JUST WANTED TO SHOW YOU SOMETHING. A COUPLE YEARS AGO, YOU KNOW THE HIGH AERODYNAMIC DIAGRAMS, NEAR-FAR DIAGRAMS IN THE TAG LIBRARIES THAT ARE REALLY USEFUL SUGGESTING THAT MAYBE WE COULD -- SOMEBODY COULD MAKE THEM ANIMATED, IF YOU CLICK ON THEM THEY EXPAND SO YOU DON'T HAVE TO, YOU KNOW -- I DID THAT JUST FOR FUN. I WANTED TO SHOW YOU GUYS. THESE ARE ALTERNATE VERSIONS, TAG LIBRARIES. HOPEFULLY THIS WILL BE USEFUL TO SOME PEOPLE. IT SHOWS THE CONTENT MODEL OF THE FRONT ELEMENT AND THESE SYMBOLS ARE SEQUENCE, THESE SYMBOLS ARE FOR CHOICES. THAT'S A CHOICE OF ONE, QUESTION MARK MEANS OPTIONAL, YOU CAN CLICK HERE AND EXPAND IT. THE SO THERE'S THE ATTRIBUTES THAT GO IN THE ARTICLE, META, THAT'S THE CONTENT MODEL OF JOURNAL META, NESTED, SO FORTH. AND YOU CAN RIGHT CLICK, YOU CAN CLICK HERE, GO TO THAT ELEMENT DOCUMENTATION PAGE. IT'S HARD TO GET BACK TO THE EXACT SAME PLACE IN THE DIAGRAM THOUGH. OKAY. AND ONE MORE FEATURE, INTERESTED IN F PAGE, HOPEFULLY THAT WILL BE USEFUL. THAT'S ALL. THANKS. PROS APPLAUSE [APPLAUSE] [APPLAUSE] >> SINCE WE HAVE PEOPLE WATCHING ON THE VIDEO WHAT CHRIS SAID WAS JATSPAN.ORG, FOLLOW THE LINK, THE LINK IN THE HEADER TO THE TAG LIBRARY. HOPEFULLY THOSE OF YOU WHO ARE WATCHING REMOTELY GOT THAT. IF NOT, HIT US UP AND WE'LL SET YOU UP. OKAY. THAT IS IT. WE DID AN AMAZING JOB. I WANT TO THANK EVERYONE WHO HAD THE GUMPTION TO COME UP AND SAY THEIR PEACE AT THE OPEN SESSION. WE'LL MEET BACK HERE AT 3:30. >> TALKS OF JATS-CON 2016, IT'S MY PLEASURE TO INTRODUCE PAUL JOHNSTON FROM PUBREF.ORG, STARTED AS A VASCULAR SURGEON BUT DECIDED PROGRAMMING WOULD BE A LOT MORE FUN SO, YEAH, THAT'S AN INTERESTING CAREER SWITCH. HE HAS MOVED, COMPLETELY GIVEN UP SURGERY, ALTHOUGH THERE ARE SOME PEOPLE WHO CONSIDER SOFTWARE DEVELOPMENT A DIFFERENT KIND OF SURGERY, NOW AUTHORING TOOLS FOR SCHOLARS AND, PAUL, COME ON UP AND WE'D LOVE TO HEAR ABOUT THIS >> THANK YOU, BRUCE. THANK YOU TO THE MEMBERS. I'VE HAD A GREAT CONFERENCE, I'M LOOKING FORWARD TO APPLYING WHAT I'VE LEARNED, THINGS I WOULD BE TAKING AWAY FROM IT. IT'S GREAT TO BE HERE. I'VE DONE THOUSANDS OF PubMed SEARCHES AND IT'S BEEN SUCH A GREAT TOOL, IT'S AN HONOR FOR ME TO BE HERE SPEAKING WITH YOU. A QUICK SHOW OF HANDS, WHO KNOWS ABOUT MARKDOWN OR WRITES MARK DOWN? HALF THE PEOPLE. I WOULD ARGUE MARKDOWN IS KIND OF EVERYWHERE. DEPENDING ON YOUR PERSPECTIVE, THERE'S A TREMENDOUS AMOUNT OF CONTENT BEING GENERATED IN MODERN FORMS OF COMMUNICATION. I THINK A LOT OF THE PEOPLE THAT WOULD -- THAT HAVE TRADITIONAL OPINION PART OF SCHOLARLY PUBLISHING ARE USING FORMS OF COMMUNICATION. I'VE BEEN TO A FEW OF THESE CONFERENCES, WHERE ARE THE USERS GOING, WHERE ARE THE PEOPLE? THE PEOPLE ARE AT SITES LIKE THIS USING GITHUB AND SITES LIKE MATH OVERTHREW, STACK OVER OVERSNOW, STACK OR REDDIT. ALL THESE SITES ARE BASED ON MARKDOWN, THE SORT OF DOMINANT WRITING SYSTEMS, IT MARKDOWN COULD BE DOMINANT IN THE WORLD BUT PENETRANCE INTO SCHOLARLY PUBLISHING IS NIL AT THIS POINT. THE STORY OF MARKDOWN IS SIMILAR TO THE STORY OF JATS, AT LEAST THAT JATS IS ADDRESSING. JOHN GRUBER IN 2004, AN XML GEEK, A PROMINENT BLOGGER, YOU CAN PAY $10,000 A WEEK TO ADVERTISE ON HIS BLOG IN FIREBALL. IN FOUR HE WAS FRUSTRATED OF XML INTO MOVABLE TYPE. THIS IS STUPID, HARDER THAN IT HAS TO BE, I WANT TO WRITE AN E-MAIL IN AN E-MAIL-LIKE FORM AND USE THAT. HE WROTE MARKDOWN.PL, THAT BECAME A FOUNDATION FOR IMPLEMENTATIONS OF TODAY. THE STORY STARTS WITH AARON SCHWARTZ, BOY GENIUS IN THE STORY. AN AMAZING PERSON. YOU'LL READ SOMETIMES HE INVENTED RSS. HE DIDN'T, BUT AT AGE 14 HE WAS INVOLVED WITH DAVE WEINER AND RSS CORE GROUP AT THAT TIME. IF YOU PULL UP RFC3870, RDF SPEC, HE WAS THE AUTHOR WHEN I WAS 20 OR 22, HE WROTE A LANGUAGE CALLED ATX, PRECURSOR TO MARKDOWN, WORKING WITH JOHN GRUBER CLOSELY IN DEVELOPING THAT. HE WAS ONE OF THE CO-OWNERS OF REDDIT AND BECAME WEALTHY WHEN IT WAS SOLD TO CONDE NAST. HE WAS ARRESTED REGARDING ILLEGAL DOWNLOADING OF JSTOR ARTICLES. I DON'T CONDONE THAT BUT THE GOVERNMENT RESPONSE WAS AN INTERESTING TIME IN HISTORY WHEN WIKILEAKS WAS HAPPENING, JULIAN ASSANGE AND THE GOVERNMENT MADE A STRONG EXAMPLE OF HIM. HE COMMITTED SUICIDE TWO DAYS AFTER THE GOVERNMENT DECIDED NOT TO, YOU KNOW, GO TO -- NOT TO BASICALLY ACQUIT HIM BECAUSE NEITHER M.I.T. NOR JSTOR WERE INTERESTED IN PRESSING CHARGES. ANYWAY, SO HE WAS AN AMAZING PERSON AROUND I THINK IT'S POSSIBLE THAT HE COULD BE IN THIS AUDIENCE TODAY IF HE WAS STILL ALIVE. CONTINUING THE STORY OF MARKDOWN CONTINUES, THE GODFATHER NOW, THE PERSON HOLDING THE TORCH IS JOHN MacFARLANE, PROFESSOR AND CHAIR AT BERKELEY. HE HAS DEVELOPED PANDOC, A TOOL MOST PEOPLE ARE FAMILIAR WITH, AS WELL AS COMMONMARK, TAKING THE MARKDOWN PL SPEC AND SAYING LET'S FORMALIZE THAT, LET'S MAKE SURE IT'S AS CLEAN AS POSSIBLE, SO ANY PARSER THAT'S IMPLEMENTING MARKDOWN IS UNAMBIGUOUS. THAT'S A VERY ACTIVE COMMUNITY. IN TERMS OF GETTING MARKDOWN INTO SCHOLARLY PUBLISHING THERE HAVE BEEN A NUMBER OF EFFORTS, GENERALLY WON'T GO INTO DETAIL, YOU'RE FAMILIAR WITH NAMES, MARTIN FENNER, KARTIK RAM, BUT NOT SUSTAINED EFFORTS. THEY ALSO RELY ON USING PANDOC AS A CONVERSION TOOL. AS GOOD AS PANDOC IS, AS A GENERAL SORT OF DOCUMENT CONVERSION TOOL, SORT OF SUFFERS FROM ITS GENERALITY IN A LITTLE BIT THAT YOU CAN'T REALLY GO OUTSIDE OF WHAT PANDOC'S DOCUMENT MODEL IS EASILY, PROGRAMMING AN IMPENETRABLE PROGRAMMING LANGUAGE FOR MANY PEOPLE. SO MY THESIS IS TO TAKE THE IDEA OF PANDOC AT LEAST, PUT JATS AT THE CENTER OF THAT. AT LEAST THE SUPER FAMILIAR OF JATS/BITS, THEY ARE TRYING TO SOLVE STRUCTURE, PRESENTATION, TO BE ABLE TO TRANSFORM DOCUMENTS EASILY. JATS IS A STANDARDIZED DOCUMENT FORMAT, SO WE CAN KIND OF LEAN ON ALL THE PEOPLE WHO HAVE TAKEN IT THIS FAR AND LEARNED FROM THEIR EFFORTS. I THINK THERE'S A GREAT PROMISE IN USING JATS AS THE CENTRAL SORT OF COMMON DOCUMENT FORMAT FROM THE INCEPTION OF WHEN AUTHORS ARE STARTING TO WRITE ALL THE WAY THROUGH THE WORKFLOW WHEN IT'S GOING TO DIFFERENT FORMS IT'S CURRENTLY IN. SO THE DESIGN OF PIPEDOWN IS THAT IT SHOULD BE EASY TO LEARN, EASY TO WRITE AND MAKE IT AS SIMPLE AS POSSIBLE BUT NOT SIMPLER. SO THERE'S ACTUALLY MANY SORT OF LAYERS TO THIS ONION, AND THERE'S A DOCUMENT MODEL BASICALLY JATS, XML AND MARKDOWN ARE TEXT FORMATS WITHOUT FORMAL RECOGNITION OF A FILE SYSTEM, BOTH WORD AND PDF, THE TWO PROBABLY MOST DOMINANT FORMATS, ARE REALLY FILE SYSTEMS, NOT JUST DOCUMENTS. YOU CAN PUT ARBITRARY BINARY OBJECTS IN OR DOCUMENT OR PDF, THAT'S WHY THEY ARE SO USEFUL. PIPEDOWN MAKES THAT FORMAL, ANY TOOL THAT'S COMPATIBLE WITH THIS HAS TO PROVIDE INTERFACE. CONTAINERIZATION, DOCKER, ADDRESSES REPRODUCIBILITY ISSUES OCCURRING IN SCIENCE. I BELIEVE THE LIBRARY OF THE FUTURE IS LIKE A FORMALIZED DOCKER REGISTRY WHERE PEOPLE ARE CHECKING OUT OPERATING SYSTEMS, APPLYING RESEARCH AND BUILDING ON THAT. SO ANY FILE CAN RENDER THIS WELL, THERE'S ZIP FILE FORMAT AND ALSO A DOCKER CONTAINER. I WON'T GO INTO THOSE MUCH NOW. SO IT HAS A FORMALIZED WAY OF SAYING WHERE OBJECTS ARE WITHIN THIS FILE SYSTEM THAT IS CONTAINED WITHIN RELATIVE -- RELATIVE TO ROOT, CONTAINER OR DNS HOW MARKDOWN ALREADY WORKS. SO THE COMPILER, MOST MARKDOWN COMPILER PROCESSERS ARE NOT CONTEXT FREE, THEY PROCEED PARSING INTO A BLOCK STRUCTURE FIRST, PARSING INLINE STRUCTURE FROM THERE. THEN YOU GET THROUGH YOUR INITIAL AST OUT, TRANSFORMS TO GET THE FINAL JATS PRODUCTS. AND THE SYNTAX IS REALLY PRETTY MUCH THE SAME AS MARKDOWN, BUT KIND OF USING THIS LINK DEFINITION, A USEFUL OBJECT MAKING ASSOCIATION BETWEEN TWO THINGS, AND THEN APPLYING THAT SAME IDIOM IN DIFFERENT CONTEXT, IN A LIST CONTEXT YOU GET A DEFINITION LIST, WHICH IS QUITE USEFUL, OR IF YOU PUT IT IN A SECTION CONTEXT YOU CAN NOW APPLY DIFFERENT BLOCK STYLE TO SECTIONS. OR DO THINGS LIKE IMPORTING AND OTHER SORT OF RULE DEFINITIONS. BASICALLY SYNTAX IS THE SAME BUT YOU'RE GENERALIZING THAT PARTICULAR IDIOM. SO IF YOU DON'T KNOW MARKDOWN YOU'LL LEARN IT IN THE NEXT FIVE MINUTES. WE'LL DEMO THAT AND GO OVER THE BASIC FEATURES, SECTIONS, LISTS, THE RULES LIKE I TALKED ABOUT, EMPHASIS, EXTERNAL LINKS, GRAPHICS OR MEDIA, CITATIONS, ENTITIES, AND THEN WE'LL SEE HOW MUCH WE CAN GET TO BUT I THINK WE CAN COVER ALL THE INTERMEDIATE STUFF, MATHEMATICS, TABLE, PRE-FORMATTED CODE BLOCKS, DEFINITION LISTS, DEFINITION SECTIONS, FOOTNOTES, INTERNAL LINKS, HRFs BASICALLY, INLINE ANNOTATIONS AND GETTING OUT THERE, THERE'S NAMED ENTITIES, RULES, COORDINATES, PIPELINE, TRANSFORMATION OF TEXT, BINDING OTHER TYPES OF FILES TO, SAY, DEFINITION LISTS WHICH CAN SORT OF COMPREHEND INTO A LARGER TABLE FORMAT. INTERFACES AND TYPE ANNOTATIONS WE PROBABLY WON'T COVER. TO DO THAT I'M GOING TO COME OUT OF THE PRESENTATION AND THEN OPEN UP PIPELINE, APPLICATION WHICH IS LIKE A MASHUP OF GITHUB, WIZY WIG AND TEXT EDITOR THAT WORK TOGETHER. I WOULD START BY MAKING NEW PROJECTS, CREATE A PROJECT AND IMPORT FILES, START WORKING ON THEM HERE. IF I WANTED TO DO THAT. SO JUST DEMONSTRATE YOU CAN DO THAT. I'M GOING TO GO BACK TO SORT OF THIS ONE WHICH I'VE ALREADY SET UP. IN HERE I ALREADY HAVE SOME IMAGES THAT I'VE ALREADY UPLOADED. AND SOME TEST FILES. SO FIRST LET'S COVER SECTIONS. PROJECT, FILE. SO ON THE LEFT, TELL ME IF THE TEXT IS TOO SMALL, I CAN MAKE IT BIGGER AS I TYPE HERE, ON THE LEFT SIDE YOU GET THE UPDATES, COMPILES AND UPDATES IT, I CAN SWITCH OVER AND SHOW THE JATS UNDERNEATH. AS WE TYPE, WE'LL BE SORT OF DEMONSTRATING HOW THAT TRANSFORMS INTO JATS AND/OR THE RENDERED FORM. SO SECTIONS, WHEN YOU TYPE A SECTION, IT'S A POUND SIGN. AND YOU WOULD SAY TITLE AND THAT TRANSFORMS INTO SECTION TITLE THAT YOU'RE PROBABLY FAMILIAR WITH. THEN YOU KEEP TYPING FROM THERE. AND TYPICAL MARKDOWN PROCESSORS WILL -- IF YOU SAY ALL RIGHT, INSTEAD OF H-1, I WANT H-2 BLOCKS, SUBTITLE OR SUBSECTION, THEN IT GIVES A FLAT LIST, IT DOESN'T NEST THAT LIKE YOU WOULD EXPECT IN JATS. ONE OF THE FIRST CHANGES IS PUTTING YOU THAT INTO A NESTED STRUCTURE. SO IT BECOMES VERY EASY TO THEN KIND OF LAY OUT A DOCUMENT, YOU KNOW, METHODS, RESULTS AND SO FORTH. ALL RIGHT. THAT'S CLEAR? CONTINUE? ALL RIGHT. YOU'VE LEARNED SECTIONS. AWESOME. LISTS, LISTS ARE PRETTY EASY. TYPE A LIST. I'M GOING TO START A NEW FILE. I'LL START A NEW FILE. LISTS. BASICALLY YOU TYPE A LIST LIKE YOU WOULD IN AN E-MAIL. A, B,C, AND THAT BOOKS AN ORDERED LIST. OR IF YOU WANT A BULLETED LIST, SAME THING. I'M GOING TO INTRODUCE RULES. SOMETIMES THE WHITE SPACE WILL COLLAPSE, YOU WANT SOMETHING TO SLICE THE LIST IN HALF. SO A RULE ALLOWS YOU TO KIND OF DO THAT. BUT A RULE, IT'S A PARTITION MARK TRANSLATESSED INTO A HORIZONTAL RULE, ELIMINATED FROM DOCUMENTS. AND SO YOU CAN USE THESE AS COMMENTS. SO YOU HAVE A WAY TO PUT COMMENTS IN AND ALSO JUST BREAK UP CONTENT WHERE YOU WANT TO. OKAY. SO WE LEARNED LISTS, IF YOU WANT TO SAY, YOU KNOW, NEST A LIST, YOU SAY FIRST ITEM, WHAT YOU DO IS INDENT WITH FOUR SPACES, AND THEN TYPE THE NEXT LIST ITEM. AND MADE A TYPO BUT -- OH, I NEED TO PUT CONTENT ON THERE. AND THEN IT WILL CREATE A NESTED LIST FOR YOU. RENDERING IS WEIRD ON THAT. THERE IT GOES. SO IT'S REALLY EASY TO MAKE LISTS, NESTED LISTS AND SO FORTH IN MARKDOWN. AND THEN ALL THE FIVE SORT OF TYPES YOU WOULD EXPECT, ROMAN, ALPHA, ORDER, BULLET ARE THERE. MOVED FORWARD, RULES, WE COVERED THAT. THEY GET SPECIAL POWERS LATER. FOR EMPHASIS, SO IT'S EASY TO KIND OF EXPRESS -- IF YOU DIDN'T HAVE FORMATTING IN E-MAIL YOU MIGHT PUT SOMETHING LIKE -- ALL RIGHT. THIS IS IMPORTANT. AND THAT BECOMES ITALIC. TWO ASTERISKS, IT'S BOLD. ASK THEN THERE'S VARIOUS OTHER FORMS OF THAT. UNDERLINING CONTENT, TWO UNDERSCORES. I MAPPED ONE UNDERSCORE TO OVERLINE WHICH I DON'T KNOW HOW USEFUL THAT IS. HOW DO PEOPLE USE OVERLINE? THIS IS REALLY THE ONLY THING I COULD THINK OF, WRITING SCRIPTS OLD STYLE, ONE TAB TID WITH MEALS, AN OVERLINE. THAT'S SUPERFLUOUS. SUPER SCRIPT, SUBSCRIPTED FIRST BETWEEN TILDES. SO WATER, 10 TO THE 23RD. SO IT'S REALLY EASY TO MAKE ALL THOSE FORMATTING THINGS YOU WOULD EXPECT YOU HAVE STRIKE-THROUGH, PROGRAMMERS USE ALL THE TIME, LIKE A MONO SPACE HTML THAT MAPS TO COLD ELEMENT, A BLACK ELEMENT IN JATSZ JATS, AS FAR AS I CAN TELL, MAPPED IN MONO SPACE. THIS IS A VARIABLE NAME, MONO SPACED. THAT'S EMPHASIS. ANY QUESTIONS ON EMPHASIS? OKAY. KEEP MOVING FORWARD. NEXT SO LINKS. SO THERE'S FOUR WAYS, AT LEAST FOUR WAYS TO MAKE LINKS. THE FIRST ONE IS ... IT'S CALLED AN INLINE LINK. YOU SAY NIH AND YOU PUT THE LINK IN PARENTHESES. AND IT CREATES THE LINK FOR YOU OVER HERE. THIS IS THE EDITOR HERE, I CAN EDIT HERE BUT NOT FOCUSED ON THAT. THAT BECOMES AN EXTERNAL LINK THERE. OR YOU CAN TAKE THIS OUT, IT BECOMES A REFERENCE LINK. AND THEN YOU USE A LABEL, AND THAT LABEL IS USED TO ASSOCIATE THE HREF TO THE LABEL. SO THAT IS THE LINK DEFINITION. ALL RIGHT. YOU CAN ALSO JUST LEAVE OUT ONE OF THOSE, IT'S ALMOST LIKE A WIKIPEDIA-LIKE LINK, THAT STILL WORKS. OR YOU CAN DO AUTOLINKS, BETWEEN BRACKETS WHERE THE HRER EF TAKES ON THE LINK AND VALUE OF INTACT. THESE ARE LINKS. WE REUSE THIS IDIOM HERE, OVER IN DIFFERENT WAYS TO DO MORE INTERESTING THINGS. ONE WAY IS TO -- LET ME COVER MEDIA LINKS, THEY LOOK ALMOST THE SAME. IF YOU WANT TO PUT AN IMAGE IN, LET'S SWITCH BACK OVER. I'LL DO AN INLINE IMAGE HERE, I'M GOING TO ADDRESS SOMETHING THAT'S IN THE SAME PROJECT WHICH I ALREADY KNOW IS THE NAME FIGURE.PNG, ASTERISK IN FRONT OF IT, IT CREATES THAT AS A GRAPHIC, ALL TEXT IS HERE. THAT IS A GRAPHIC. SO ONE OF THE TRANSFORM STEPS, IF I FIND A GRAPHIC IN THE MIDDLE OF A PARAGRAPH, IT'S THE ONLY THING IN THE PARAGRAPH, LET'S MAKE THAT A GRAPHIC. IF I FIND IT IN TEXT, THEN I PROBABLY WANT IT INLINE GRAPHIC. AND THEN THAT ONE DOESN'T COMPRESS WELL BUT IF I WANT SAY AN ICON, SO YOU CAN DO BOTH INLINE AND DISPLAY GRAPHICS. SO THAT'S REALLY COOL. IT'S REALLY EASY TO DO. SO THEN ONE OF THE VARIATIONS, ALL RIGHT, IF THE HREF IS NOT ARBITRARY, IT'S TO ONE OF THE METADATA REPOSITORIES WE USE, THE TWO ONES IMELEMENTED ARE PubMed AND CrossRef, YOU WANT TO CITE THIS, YOU DON'T WANT TO BUILD A LINK. IF I SAY JONES ET AL., AND I DEFINE THAT HERE AND I'M GOING TO SAY THAT'S PubMed, GO BACK TO DO A QUICK SEARCH HERE. AND I'M GOING TO CREATE CITATION LIBRARY WHICH I'LL NAME PubMed, NOW I HAVE A SEARCH INTERFACE WITH PubMed. SO A FRIEND OF MINE OKAY. HERE IS AN ARTICLE, WHAT I WANT TO GET IS PubMed IDENTIFIER, I'M GOING TO RELOAD THAT. I NEED TO JUST LAY THE PubMed IDENTIFIER A LITTLE BIT MORE EASILY. AND FOR THE PURPOSES OF THIS DEMO I NEED AN EASIER WAY TO GET THAT I.D. SO IF I SEARCH FOR THAT AGAIN, GOING TO GET THE PubMed I.D. HERE, GO BACK TO -- WHERE WAS I, LISTS? NO, LINKS, THANK YOU. ALL RIGHT. TYPE THAT IN. OH, YEAH, YOU'RE RIGHT, THANK YOU. NOW IT WENT AND FETCHED THIS CITATION DATA. AND THEN WHAT THAT DOES IS IT CREATES THE BIBLIOGRAPHY REFERENCE REPRESENTS USE AND CREATED A REFERENCE WITH ELEMENT CITATION, AND THEN ONE IS THE MIX CITATION. MIX CITATION DOESN'T HAVE ANY SEMANTIC CONTENT, IT'S ONLY FOR HUMAN, YOU KNOW, LOOKUPMENT IF YOU WANT THE ACTUAL ELEMENT DATA, IT'S THERE TOO. HOPEFULLY THE BEST OF BOTH WORLDS INSTEAD OF ENVIRONMENT VARIABLE, CITATION STYLE, I'M GOING TO JUMP AHEAD, I CAN ACCESS ANY OF THE 7000 STYLES, I DON'T KNOW WHY WE NEED 7000 STYLES, BUT YOU CAN FORMAT YOUR REFERENCES, USING NATURE, SCIENCE, I LIKE THE IEE ONE, WHAT ARE SOME OTHER FAVORITE ONES? HARVARD? DOES THAT COME UP? THAT'S MAPPED TO SOMETHING ELSE. CHICAGO HAS A BUNCH. YOU HAVE TO REMEMBER IT'S CHICAGO, DATE, AUTHOR. CHICAGO, AUTHOR, DATE, A BUNCH OF CHICAGO STYLE. YEAH, OR I CAN PASTE IN LINKS THAT ARE DOIs OR PubMed I.D.s OR METADATA PROVIDER WHICH HAS A SCHEME AND SOMETHING AND THEN YOU CAN GET THIS CITATION DATA, YOU CAN DO THAT. CITATIONS BECOME REALLY EASY, IT BRINGS UP DO WE NEED CITATION MANAGERS ANYMORE? DO WE NEED TO HOARD PDFs? WE PROBABLY WILL FOR A PERIOD OF TIME BUT CrossRef IS OUR CITATION MANAGER, PubMed IS OUR CITATION MANAGER. ENTITIES, UNICODE, WHO ARE THE MOST PROMINENT USERS IN THE ENGLISH SPEAKING WORLD IN PROBABLY YOUR CHILDREN BECAUSE -- WHEN THEY USE SNAP CHAT AND SENDING HALF NAKED MESSAGES TO THEIR FRIENDS, OR USING THEIR, YOU KNOW, PHONE OR WHATEVER, THEY ARE USING SYNTAX LIKE THIS. OH MY GOD, AND THEN IT'S A COLON-COLON, AND THEN THEY ARE USING UNICODE PARAMETER ENTITIES BECAUSE ALL OF THE EMOJIS ARE UNICODE USERS, THE BIGGEST USER IS EMOJIS. WE TAKE ALL THE PARAMETER ENTITIES FROM JATS AND PUT THOSE IN HERE SO WE CAN SAY OH MY GOD, THIS SET, REAL NUMBERS. I'LL MAKE THIS BIGGER. IT'S KIND OF FUN. . YOU CAN EXPLORE THIS SET. ENCYCLOPEDIA IT MAKES IT EASY TO USE THE PARAMETER ENTITIES IN JATS AND THEY ARE AWESOME BUT JUST HARDER TO GET TO. SO THOSE ARE ENTITIES. ONE OF THE COOL THINGS YOU CAN DO WITH ENTITIES IS DEFINE THEM. IF YOU WANTED TO REUSE SOMETHING FROM A DOCUMENT, OKAY, THE MOUSE STRING, THEY ARE OFTEN LIKE THESE WEIRD THINGS THAT ARE LIKE MUTE, YOU KNOW, X4, THAT YOU WOULDN'T WANT TO TYPE OVER AND OVER AGAIN. I'M JUST GOING ASSIGN THAT THAT TO MY OWN ENTITY, C57, AND SAY C57 MOUSE STRING, REUSE THAT OVER AND OVER AGAIN THROUGHOUT THE DOCUMENT, EASY TO FIND VARIABLES, REUSE AND BUILD TEMPLATES OR PUT IN, YOU KNOW, THE RESULT VALUES THAT YOU'RE STILL WAITING FOR AND SUBSTITUTE IN ONE PLACE. SO THOSE ARE WHAT I'M CALLING NAMED ENTITIES. ALL RIGHT. SO INTERMEDIATE THINGS, MATHEMATICS, PRETTY EASY. WE'RE USING WHAT MOST MARKDOWN IMPLEMENTATIONS USE, LE-TEX, EINSTEIN'S EQUATION, ONE DOLLAR SIGN FOR INLINE MATH EQUALS MC SQUARED, YOU GET CITATION ALTERNATIVES WITH THE TACK M ATH ELEMENT, AND MML. DISPLAY MATH WORKS THE SAME WAY, BETWEEN TWO DOLLAR SIGNS, AND YOU GET DISPLAY FORMULA. MATHEMATICS IS EASY. I'M USING K-TECH AND/OR MATH JACKS. K-TECH RENDERS FASTEST SINCE IT'S A SYNCHRONOUS RENDERING PIPELINE BETTER FOR REALTIME DEMO, DOESN'T HAVE AS MUCH COVERAGE AS MATH JACK, IF YOU FAIL THAT IT WILL RETRY USING MATH JACK BUT IT'S NICER FOR EDITING ENVIRONMENT TO RENDER FASTER. SO TABLES, TABLES ARE STILL AN AREA WHERE YOU NEED -- YOU CAN DO BASIC TABLES AND BE EFFICIENT QUICKLY. TABLE SYNTAX IS YOU BASICALLY HAVE A HETERO, THEN KIND OF A CALL GROUP, AND THEN THE BODY. AND YOU START THAT OFF WITH LET'S SAY PATIENT CHARACTERISTICS. CALL GROUPS SEPARATED BY THREE DASHES, IT BECOMES HEADER ROW, AND THEN I CAN SAY 45, 55%, SOMETHING LIKE THAT. I SEE THIS NEW ENGLAND JOURNAL OF MEDICINE. SO USING THE XHTML TABLE MODEL, YOU GET A COLGROUP, TABLE HEADER, DID I DO THAT WRONG? I THINK THAT'S RIGHT. SO IT'S REALLY EASY TO MAKE TABLES. IN TERMS OF ALIGNMENT, COLON ON THE LEFT SIDE, LEFT ALIGNMENT. IF YOU WANT IT ON THE RIGHT SIDE, YOU DO RIGHT ALIGNMENT OR CENTER ALIGNMENT, WHICH I DON'T HAVE FULLY IMPLEMENTED BUT THAT'S HOW -- THESE ARE GITHUB FLAVORED MARKDOWN TABLES. PRE-FORMAT BLOCKS, CONSTANTLY PUTTING SYNTAX. IF I WANT TO TYPE JAVA CODE, BETWEEN THREE BACK TICKS OR THREE TILDES. IT'S LIKE, YOU KNOW, A RESERVED PRE-FORMATTED AREA WHERE I CAN TYPE ANYTHING I WANT HERE, AND IT PRESERVES THAT IN THE PRE-FORMAT BLOCKS. THE FIRST LINE THERE'S AN AREA CALLED THE INFO STRING, THE INFO STRING CONTAINS A LABEL WHICH MAPS TO A PROGRAM LANGUAGE, IT WILL GO AHEAD AND SYNTAX HIGHLIGHT. JAVA LANGUAGE WITH THAT, JAVA SCRIPT. OKAY. NOW INSTEAD YOU GET A CODE AND THEN SHOULD SYNTAX HIGHLIGHT THAT FOR YOU, THERE'S A BUNCH OF DIFFERENT SYNTAX HIGHLIGHTING SCHEMES AND COLOR SCHEMES FOR DIFFERENT LANGUAGES. OKAY. MOVING FORWARD, EVERYONE DOING OKAY? >> YES. >> DEFINITION LISTS ARE AWESOME. BUT THEY ARE TOTALLY UNDERUSED IN HTML. I MEAN, WHO HERE USES DD, DT, DL, ONE PERSON, A FEW PEOPLE. FOR WHATEVER REASON JUST THE DEFAULT RENDERING IN HTML JUST SUCKS AND NO ONE REALLY USES IT. BUT THEY ARE GREAT. AND IT'S EASY TO MAKE DEFINITION LISTS IN THIS PIPEDOWN LANGUAGE. DEFINITION LISTS. SO INSTEAD OF DOING A SIMPLE LIST LIKE A, B, C, YOU PUT A LABEL IN FRONT OF IT. SO FIRST NAME. LAST NAME. THEY BASICALLY RESEMBLE FORMS, RIGHT? DEGREE. YOU CAN IMAGINE TAKING A DEFINITION LIST AND ALSO REMAPPING THAT TO A CONTRIBUTOR. ALL RIGHT. YOU GET A DEFINITION LIST. I DON'T HAVE THE NESTING GOING ON FOR THAT YET BUT THAT SHOULDN'T BE TOO HARD TO ADD. BUT DEFINITION LISTS ARE REALLY QUITE USEFUL AGAIN BECAUSE IT'S AN ASSOCIATIVE ARRAY, AND THEN IT'S A MAP, YOU CAN THEN TRANSFORM THAT TO OTHER THINGS, YOU CAN IMAGINE HAVING LIKE A BIB TECH LIKE FORMAT PEOPLE CAN ENTER CITATIONS MANUALLY IF THEY WANT TO, IF THEY HAVE SOMETHING THAT ISN'T IN CrossRef, OR CONTRIBUTORS, AFFILIATION LISTS, FLEXIBLE KIND OF APPROACH TO TAKING A DEFINITION LIST AND THEN TRANSFORMING IT TO SOMETHING ELSE. DEFINITION LISTS ARE REALLY EASY TO DO. ALL RIGHT. DEFINITION SECTIONS, SO LET'S SAY I PUT IN A GRAPHIC HERE, SO ALL -- SO ALT, FIGURE. WHAT DID I DO WRONG? RELOADED THAT JUST IN CASE. I DON'T KNOW WHERE I WAS. LET'S TRY THAT AGAIN. SO I WANT A GRAPHIC. IT SHOULD BE FIGURED OUT PNG. ALL RIGHT, WELL, LET'S USE Q CUBE. I DON'T KNOW WHY IT'S NOT PULLING THAT UP BUT YOU SAW IT EARLIER SO LET'S JUST CONTINUE. IT WILL PROBABLY WORK IN A SECOND. WHAT'S THAT? (INAUDIBLE) OH, YEAH, I NEED THE DOT, THANK YOU. THAT SHOULD WORK. OKAY. WELL, SO IN A DEFINITION LIST, INSTEAD OF USING A TITLE HERE OR ANY TEXT, ANY OTHER TEXT, THERE'S A LOT OF ELEMENTS IN JATS WHICH HAVE THE FORM, THE BLOCK, OPTIONAL LABEL, OPTIONAL TITLE, AND SO YOU CAN REUSE THAT OVER AND OVER AGAIN TO SAY THIS IS MY LABEL SO IF I WANT FIGURE 1, WHATEVER IT IS, THEN INSTEAD OF TAKING THE SECTION YOU RECOGNIZE YOU WANT TO MAKE IT A FIGURE. SO TAKING A SECTION, RETRANSFORMING THAT TO A FIGURE BLOCK WHERE YOU HAVE THE LABEL, CAPTION PUT IN THERE, AND THEN THE GRAPHIC. AND THEN ANYTHING THAT'S ALSO SORT OF CONTAINED WITHIN THE SECTION, SO THIS IS GOING TO BE STUFF AFTER CAPTION, IF I PUT ANYTHING ELSE HERE IT'S KIND OF ALREADY SORT OF NESTED IN THAT BLOCK SO THIS BECOMES THE FIGURE CAPTION. AND SO IN HERE THEN I HAVE IT -- IT LOOKS E-LIFE-ISH, I HAVE A FIGURE LABEL, I CAN EDIT HERE. MY LABEL. FIGURE CAPTION. SO FORTH. SO IT'S REALLY EASY TO NOT ONLY MAKE FIGURES AND GRAPHICS BUT WRAP THAT IN APPROPRIATE MANNER THAT YOU -- THAT USERS WOULD EXPECT TO SEE IN A FINAL RENDERED OUTPUT AFTER THEY ARE PUBLISHED. YOU CAN DO THE SAME THING WITH THE TABLE. INSTEAD OF HAVING A BARE TABLE YOU CAN WRAP IT IN A TABLE WRAP. LET'S SEE. SO IF I PUT IN A SECTION ... I'LL MAKE A QUICK TABLE THERE, AND MY TABLE CAPTION, IT SAYS IT'S GOING TO BE TABLE 1, SO NOW INSTEAD OF GETTING A FIGURE I GET A TABLE WRAP. AND I GET THE TABLE, THE CAPTION, WITH THE TITLE. YOU HAVE TO REMAP THAT. SOMETIMES THERE'S SOME -- SOMETIMES THE TITLE'S IN THE CAPTION, SOMETIMES IT'S NOT. YOU HAVE TO BE CAREFUL FOR THAT, BUT EASY TO MAKE TABLE STRUCTURES, AND ONE SORT OF I LIKE THE APPEARANCE OF IS BASICALLY WRAPPING A CODE BLOCK, YOU CAN MAKE A BOX TEXT WITH SORT OF LIKE A FORMAL CODE LISTING, LIKE YOU SAW IN MANUSCRIPTS APP. SO TABLES, YOU CAN ALSO DO OTHER KINDS OF THINGS. YOU CAN PUT IN OTHER INTERESTING BLOCKS SO IF I WANT TO PUT STUFF INTO AN APPENDIX, NOW I'VE TAKEN IT, PUTS IT IN THE BACK OF THE ARTICLE, YOU CAN HAVE MULTIPLE APPENDICES. NOW I'M PUTTING IT IN THE FRONT, THE ABSTRACTION, AND SO YOU CAN KIND OF REUSE THAT FOR GLOSSARIES, DEFINITION LISTS IN GLOSSARIES, A LOT OF WAYS YOU CAN RECOMBINE STUFF TO MAKE THE SORT OF FAMILIAR JATS STRUCTURES. FOR FOOTNOTES, LET ME BACK UP HERE, GET MY TABLE. FOOTNOTES HAVE THE SYNTAX THAT LOOKS LIKE THIS WHERE IT'S SORT OF LIKE A BRACKET THING AND THERE'S A CAROT AND NUMBER. IF I WANT TO SAY FOR THE THIRD COLUMN THE DATE IS NOT AVAILABLE, I CAN DO THAT. IT SHOULD RERENDER THAT. WHICH ONE WAS I IN? WHICH FILE WAS I IN? SECTIONS? YEAH. OKAY. SO FOOTNOTES WHEN THEY ARE INITIALLY FOUND THEY WILL BUBBLE, AND THEY BUBBLE INTO THE NEAREST CONTAINING FOOTNOTE COMPATIBLE GROUP. SO FOR A TABLE THAT'S THE TABLE FOOTER. AND IF THEY BUBBLE ALL THE WAY TO THE ARTICLE, THEY GET PUT IN THE BACK. BUT SO YOU SORT OF HAVE TO WATCH THAT BUBBLING AND THEN PUT IT IN THE CORRECT FOOTNOTE CONTAINER. BUT IT MAKES IT REAL EASY TO MAKE FOOTNOTES. INTERN LINKS, HAVEN'T FULLY IMPLEMENTED. I HAVE A TABLE, TABLE 1, I SHOULD BE ABLE TO SAY SEE IN ANY OTHER PART OF THE DOCUMENT, SEE TITLE 1, YOU SHOULD BE ABLE TO MAP THAT AND MAKE THE APPROPRIATE REF. IT DOESN'T DO THAT YET BUT IT'S NOT VERY HARD TO DO. OKAY. WE'RE MAKING PROGRESS HERE. ANNOTATIONS, THIS IS A NEW ONE. HOW DO YOU PUT COMMENTS INTO A JATS DOCUMENT? YOU HAVE TO HAVE COMMENTS. THERE'S REALLY NO COMMENT BLOCK IN JATS. THERE IS A COMMENT ELEMENT BUT IT'S NOT USED -- DIDN'T SEEM TO BE INTENDED FOR THAT AT ALL. BUT WE CAN KIND OF DO USE THE LINK, INSTEAD OF THE LINK LOOKING LIKE HTDP, WHATEVER, I'M GOING TO GIVE YOU A COMMENT. MAKE IT LOOK LIKE A TWITTER OR SOMETHING LIKE THAT, THAT ASSOCIATES IT WITH A PERSON. SO THEN IT BECOMES A COMMENT, LIKES LIKE A COMMENT HERE. AND I'M USING NAME CONTENT, SPECIFIC USE COMMENT, THAT'S PROBABLY NOT RIGHT, MAYBE IT SHOULD CONTENT TYPE. WHAT'S THAT? >> PROGRAM. >> SPECIFIC USE? NICE. SO WITH THAT YOU CAN MARK COMMENTS, YOU CAN MARK UP ADDITIONS, INSERTION, THERE'S NO INS OR DEL IN JATS BUT THEY ARE KIND OF USEFUL. IF I WANT TO MARK INSERTED OR DELETED TEXT I COULD DO THAT PRETTY EASILY. YEAH, YOU NEED TO HAVE THOSE ABILITY FOR USERS TO ANNOTATE AS WELL. SO ADVANCED FEATURES, NAMED ENTITIES WITH THE C57, NAMED RULES BASICALLY -- HERE I'M IN SECTIONS BUT I WANT TO BASICALLY ASSEMBLE A DOCUMENT. GO AHEAD, BRUCE. >> FIVE MINUTES. >> OKAY. I WANT TO ASSEMBLE A DOCUMENT TOGETHER. I SHOULD BE ABLE TO SPLICE THAT IN, RIGHT? AND THEN NAME A DOCUMENT. CURRENTLY HAVE THAT MAPPED A LITTLE BIT DIFFERENT WAY WHERE I'M GOING TO IMPORT BASIC.P.MD, AND WITH SOME LUCK ... WELL, IT'S NOT WORKING RIGHT NOW BUT YOU CAN CREATE A MASTER FILE, IMPORT STUFF, JUST LIKE I HEARD ABOUT IN STS, JUST LIKE THAT, WHERE YOU CAN TAKE LARGER DOCUMENTS, COMPOSE THEM FROM SMALLER PIECES. SO I THINK I'LL SKIP COORDINATES, WHICH IS BASICALLY A GENERALIZATION OF FOOTNOTES. IMAGINE A FOOTNOTE BEING ONE DIMENSIONAL POINT, IF YOU HAVE A TWO-DIMENSIONAL POINT YOU WANT TO ACCESS SAY A ROW OR COLUMN OR ROW AND TABLE, OR SELECTION, OR MAYBE, YOU KNOW, A GRID ON A FIGURE, YOU WANT TO HIGHLIGHT SOMETHING IN A PIXELATED BIT MAP, YOU CAN DO THAT WITH COORDINATES. PIPELINE, BASIC TRANSFORMING THINGS. A FUNCTION WILL MAKE IT BOLD OR DO SUBSTITUTION OR CAPITALIZATION. YOU'VE GOT THAT. YOU CAN IMAGINE TAKING THE FILE AND THOSE LABELS MIGHT MAP FROM FIRST NAME LAST NAME E-MAIL, BIND THAT TO A DEFINITION LIST AND SPLAY IT OUT AS A FULLY POPULATED CONTRIBUTION LIST BECAUSE YOU SEE LIKE ARTICLES WITH A THOUSAND AUTHORS, YOU DON'T WANT TO TYPE THAT. WANT TO KEEP THAT IN EXCEL OR SOMETHING LIKE THAT. IT'S A WAY TO BE ABLE TO KIND OF DO THAT. SO I'VE COVERED A LOT. THANK YOU FOR YOUR PATIENCE. HOPEFULLY IT WAS CLEAR. I'M HAPPY TO ANSWER QUESTIONS. SO THANK YOU. [APPLAUSE] MARTIN? >> MARKDOWN, THERE'S SOME RESTRICTIONS FOR EXAMPLE I KNOW THAT YOU CANNOT DO CALL SPINS OR ROW SPINS IN TABLES, AND HOW DO YOU WORK WITH THOSE THINGS, JUST RESTRICTIONS. >> TABLES? >> DO YOU EXTEND MARKDOWN IN A WAY -- >> YEAH, RIGHT, YOU NEED TO EXTEND IT. ONE WAY YOU COULD DO THAT INSTEAD OF USING A VERTICAL PIPE SEPARATOR, YOU COULD ELIMINATE THAT AND INFER THAT, WELL, IN THE PREVIOUS COLUMN YOU HAD THREE CELLS, THIS ONE YOU ONLY HAVE TWO SO LET'S JUST MERGE THE TWO FOR YOU. OR YOU CAN MAKE THAT EXPLICIT SAY WITH A PLUS SIGN OR SOME OTHER LABELING. TECH IS ALSO -- YOU COULD SAY WELL, THERE'S DIFFERENT OPTIONS THERE. ANOTHER WAY TO ADDRESS THAT IS WITH THAT SORT OF COORDINATE REFERENCES WHICH I MENTIONED, SO YOU COULD MAKE A FOOTNOTE, WHICH WOULD ADDRESS THAT ROW, AND THEN ATTACH AND SAY CALL SPAN EQUALS TWO, ADDRESSING A PARTICULAR CELL AND APPLYING EXTRA ATTRIBUTE ONTO IT. I DIDN'T MENTION THERE ARE ACTUALLY THREE SPECIAL NAMED REFERENCES WHERE IF YOU DO SOMETHING LIKE THIS, I.D. IS A RESERVED WORD, YOU CAN ASSIGN AN I.D. TO AN ELEMENT, OR I CAN ALSO ASSIGN A CLASS TO IT, SO THIS IS AGAIN ONE WAY TO PUT EXTRA METADATA SO A CLASS, SO I KNOW THERE'S A CLASS CALLED UI SEGMENT HERE IN MY LIBRARY. OKAY. OR YOU CAN ALSO ACCESS STYLE PROPERTIES. SO A FUN ONE IS CSS TRANSFORM. I'M GOING TO RETATE THIS RE-- I'M GOING TO ROTATE THIS, 100 DEGREES, 180 DEGREES, YOU CAN PRETTY MUCH ACCESS THE PROPERTIES, WHITE LISTED STYLES TOO, BUT YOU CAN IMAGINE TAKING THAT AND THEN, YOU KNOW, SORT OF PINNING OR ASSOCIATING SPECIALIZED TYPES OF FOOTNOTES ON ELEMENTS TO DO DIFFERENT THINGS LIKE YOU'RE MENTIONING. YES? >> I HAVE A QUESTION AND COMMENT. THE QUESTION IS, IS THERE SOMETHING THAT'S SHAREABLE THAT WE COULD AS PUBLISHERS USE OR IS IT JUST FOR YOU? >> WELL, I'VE BEEN FOCUSING JUST ON ACTUALLY DEVELOPING IT AND GETTING IT TO WORK, AND IT'S STILL IN THAT PHASE, ALTHOUGH IT'S GETTING PRETTY GOOD SO RIGHT NOW IMPLEMENTATION IS RUNNING ON PUBREF.ORG. THIS IS INTEGRATED WITH ARIES AND E-JOURNAL PRESS, SO AT THE END OF THIS I CAN GO PUBLISH, I CAN SELECT 4000 DIFFERENT ARTICLES AND THEN I CAN ASSIGN A LABEL TYPE AND THEN IT WILL DO A CHECK TO MAKE SURE IT HAS ALL THE MINIMUM BASIC REQUIREMENTS, PACKAGE THAT UP AND IT WILL SEND IT TO ARIES. SO FOR PUBLISHERS THAT ARE ON ARIES OR E-JOURNAL PRESS OR THAT AREN'T, AND ARE INTERESTED IN EXPLORING NO SUBMISSION OPTIONS, I MEAN, THE WAY I THINK ABOUT AN ARTICLE SUBMISSION SHOULD BE IS IT'S A FORK. IF I'M GOING TO SUBMIT AN ARTICLE, I SHOULD HAVE LIKE A GITHUB-LIKE INTERFACE, REPO, WE MAKE TO MAKE GITHUB-ISH EASIER, FORK IT, CHANGE IT, MODIFY IT, WORK ON FORK AND SO FORTH. IF YOU'RE INTERESTING IN PURSUING THE NEW FANGLED WORK FLOW I DEFINITELY WANT TO TALK TO YOU. >> COMMENTS, IT'S USED WITH JATSDOWN BUT I SEE PIPEDOWN. >> I CHANGED THE NAME. >> I LOOKED IT UP. BASICALLY TELLING SOMEONE TO SHUT UP. PIPE DOWN. >> IT'S A LITTLE BIT FUNNY. PIPING, TRANSFORMATION, AND RATHER THAN HAVE IT BE ASSOCIATED WITH JUST JATS, I THINK IT DOES POTENTIALLY APPLY TO THE JATS SUPER FAMILY, AND, YOU KNOW, THERE WAS SOME COMMENTS, WHY ISN'T IT JATS? PEOPLE DIDN'T SEEM TO LIKE THE NAME. I CHANGED IT. >> THANK YOU VERY MUCH. >> THANK YOU. [APPLAUSE] >> THE LAST SPEAKER IS ABOUT THE LAST PERSON IN THE ROOM WHO NEEDS AN INTRODUCTION, BUT I WILL MAKE A COUPLE OF COMMENTS BY WAY OF INTRODUCING JEFF BECK. THE FIRST IS HE HAS BEEN RUNNING THE NLM DTV AND JATS PROJECT SINCE INCEPTION IN 2002. I REMEMBER A FATEFUL MEETING BACK IN APRIL OF 2002 WHEN IT GOT KICKED OFF. AND MORE RECENTLY HAS BEEN RUNNING THE BITS PROJECT SINCE THAT STARTED IN 2012. ON A MORE PERSONAL NOTE, I HAVE A MUCH KEENER APPRECIATION FOR HOW CHALLENGING IT IS TO BE A CHAIR OR CO-CHAIR OF THESE PROJECTS, NOW THAT I'M CO-CHAIRING STF, AND I HAVE ENORMOUS GRATITUDE FOR EVERYTHING YOU'VE GIVEN TO THESE PROJECTS THROUGH THE YEARS RUNNING THEM AND ONE FINAL NOTE BEFORE JEFF COMES UP IF I THINK THROUGH ALL OF THE TALKS THAT WE'VE HAD AT JATS-CON THROUGH THE YEARS, YOU'VE BEEN REALLY AT SOME AMAZING, AMAZING TALKS. REALLY TREMENDOUS TALKS. BUT I THINK MY FAVORITE OF ALL HAS NOTHING TO DO WITH ANY OF THE TECHNOLOGY BUT IT'S A TALK THAT JEFF GAVE CLOSING THE CONFERENCE FOUR YEARS AGO CALLED "BEWARE OF THE SPORK." HOPEFULLY I'LL HEAR SOMETHING AS ENTERTAINING TODAY >> I CAN'T GUARANTEE ANYTHING. >> WITHOUT FURTHER ADO, JEFF BECK. >> THANK YOU, BRUCE. THAT WAS VERY NICE. LET'S PUT THIS UP. THIS IS WHERE WE'LL BE FOR HAPPY HOUR AND DEPENDING HOW LONG I TALK, SO THIS IS AN EASY WALK FROM HERE, PROBABLY THE EASY WALK FROM A LOT OF YOUR HOTELS. I HOPE TO SEE EVERYONE THERE, EVEN IF YOU WANT TO DROP IN AND SAY HEY. I HAVE A LOT OF PEOPLE TO THANK BECAUSE THIS CONFERENCE TAKES A LOT OF PEOPLE TO RUN. OF COURSE, I WANT TO THANK THE CONFERENCE COMMITTEE AGAIN, LAURA AND BRUCE, AND DEBBIE AND TOMMIE WHO HAD TO LEAVE EARLY TODAY. I REALLY WANT TO THANK ALL OF OUR PRESENTERS AND AUTHORS, I WANT TO GIVE THEM ANOTHER HAND. [APPLAUSE] PLEASE CHECK OUT THE PROCEEDINGS IN THE NEXT COUPLE WEEKS, WE'LL GET FINAL VERSIONS OF PAPERS UP THERE, WE'LL GET SLIDES ASSOCIATED WITH THE ARTICLES, AND WE'LL GET THE VIDEOS SO YOU CAN WATCH THE PRESENTATIONS AGAIN. I THINK THAT'S A REALLY NICE RESOURCE. AND I HAVE A PROJECT FOR ONE OF MY PEOPLE IN THE BACK WHO IS GOING TO BE CONVERTING EVERYTHING OVER TO BITS 2.0 SO WE CAN USE SOME OF THE FANCY NEW FEATURES IN THE PROCEEDINGS FROM HERE ON OUT. FINALLY I WANT TO THANK THE STAFF OF THE ELECTRONIC LIBRARY SECTION OF THE INFORMATION ENGINEERING BRANCH OF THE NATIONAL CENTER FOR BIOTECHNOLOGY INFORMATION OF THE NATIONAL LIBRARY OF MEDICINE. THOSE ARE THE PEOPLE WHO HAVE BEEN MANNING THE DESK AND HANDLING YOUR REGISTRATION AND HANDING OUT PARKING TICKETS AND THE ONES YOU CAN GET OUT OF PAYING, NOT THE ONES YOU HAVE TO PAY. AND THEY HAVE BEEN HELPING YOU FIND THE RESTROOM AND THINGS LIKE THAT. I ESPECIALLY WANT TO THANK JANE DAVENPORT AND ALYSSA BRANNING WHO PUT IN EXTRA EFFORT THREE DAYS STRAIGHT GETTING OUR COFFEE AND SNACKS TOGETHER. I'D LIKE TO THANK ALL OF THEM. [APPLAUSE] AND WE SET ASIDE A COUPLE WATERS FOR THE LAST TALK. LAURA, IS THERE ANYONE I'VE FORGOTTEN TO THANK? >> OUR AV CREW HERE IN THE ROOM. >> AH, YES. >> AND THE VIDEOGRAPHER. >> EXCELLENT. WILL YOU THANK THEM FOR ME? >> GENTLEMEN, THANK YOU!. >> EXCELLENT, THANK YOU. SO I DON'T LIKE TO BE TOLD WHAT TO DO. OH, LET ME TAKE MY SLIDES. SOMEBODY SHOULD HAVE TOLD ME THAT. OKAY. I DON'T LIKE TO BE TOLD WHAT TO DO SO ANY OF YOU IN THIS ROOM WHO FEEL THE SAME, I'M GOING TO APOLOGIZE AHEAD OF TIME. I WANT TO START OFF WITH A LITTLE BIT OF A TRAGIC STORY THAT INVOLVES MY HOUSE. THIS IS NOT MY HOUSE BUT IT LOOKS NICE, FITS NICE ON A SLIDE. SEVERAL YEARS AGO, IT'S BEEN ABOUT TEN YEARS, WE EMBARKED ON A JOURNEY OF HOME REPAIR. I'VE ALWAYS DONE A LOT OF HOME REPAIRS MYSELF. BUT THIS WAS A VERY BIG PROJECT. AND IT WAS GOING TO INVOLVE MANY SYSTEMS AND WALL MOVING AND NEW HEATING AND AIR CONDITIONING AND PLUMBING AND BATHROOMS BEING REDONE, ROOMS MOVING AROUND. AND I KNEW THAT WE COULD -- THAT I COULD DO A LOT OF THAT MYSELF BUT I WOULDN'T HAVE BEEN FINISHED BY NOW AND I PROBABLY ALSO WOULDN'T STILL BE MARRIED. SO WE GOT IN TOUCH WITH A DESIGN AND BUILD FIRM IN THE AREA, AND THEY ARE MORE EXPENSIVE THAN IF YOU GO OUT AND HIRE SUBCONTRACTORS YOURSELF, BUT THEY COME IN AND THEY DO THE DESIGN, AND THEN WHEN YOU AGREE ON THE BLUEPRINTS, THEY WILL BRING A PROJECT MANAGER IN AND THEY WILL PLAN OUT ALL THE WORK AND THEY WILL MANAGE ALL OF THE SUBCONTRACTORS AND IT'S SUPPOSED TO BE SEAMLESS. WE MOVED INTO OUR GARAGE FOR ABOUT THREE MONTHS. UPSTAIRS. THERE WAS HEAT AND NO CAR. AND WE STAYED OUT THERE. AND WHILE ALL THIS WORK WAS BEING DONE. AND ONE OF THE RESULTS WAS WE SPENT A LOT OF MONEY. NOW, I THOUGHT IT WAS WORTH IT BECAUSE ALL THE MANAGEMENT WAS BEING TAKEN CARE OF, BUT THESE PROJECTS SOMETIMES TAKE ON A LIFE OF THEIR OWN, AND OUR PROJECT MANAGER TURNED OUT TO BE NOT THAT GREAT. SO OUR PROJECT TURNED OUT LIKE ... [LAUGHTER] NOT EXACTLY LIKE THIS. YOU MIGHT NOT RECOGNIZE ME BECAUSE YOU'VE NEVER SEEN ME IN SHORTS. [LAUGHTER] SO IT WASN'T THIS BAD BUT THERE WERE A LOT OF THINGS THAT WERE REALLY, REALLY WRONG. ONE OF THE BIGGEST PROBLEMS WE HAD WAS WITH THE TILE GUYS WHO JUST CAME AND WE HAD DIFFERENT TILES FOR DIFFERENT PLACES IN THE HOUSE, AND I CAREFULLY LABELED ALL OF THEM AND THEN WE CAME HOME FROM WORK ONE DAY AND WE FOUND THE BATHROOM TILE ON THE KITCHEN WALL AND ALL KIND OF THINGS. SO THAT ALL GOT TAKEN DOWN, AND TILED OVER, AND THEN THE FINAL THING WITH THE TILE GUY, AFTER HE FINISHED GROUTING THE SHOWER FLOOR, HE TOOK HIS GROUT WATER AND POURED IT DOWN THE SHOWER DRAIN WHICH OF COURSE CLOGGED -- TURNED INTO CONCRETE AND WHICH MEANT THAT THE FIRST FLOOR CEILING HAD TO BE OPENED UP AND THEN ALL OF THE PLUMBING FOR THE SHOWER HAD TO BE REDONE AND WE HAD SOME TROUBLE WITH THE HVAC AND THE ELECTRICIANS, THE PLUMBERS WERE GREAT, SO WE HAD A LOT OF THOSE PROBLEMS AND I WOUND UP FIGURING A LOT OF THIS OUT AND GETTING BACK TO THE PROJECT MANAGER WHO DISAPPEARED AFTER A FEW WEEKS AND IT WAS VERY, VERY FRUSTRATING. SO FINALLY WE GET TO THE END OF THE PROJECT, AND THEY TAKE THE SENIOR -- THEY SEND US THE SENIOR PROJECT MANAGER WHO WAS ALWAYS -- HAD ALWAYS BEEN ONE OF THEIR BEST CONSTRUCTION GUYS AND THEY SEND HIM OVER, WE'RE NEGOTIATING ABOUT HOW MUCH MONEY WE'RE GOING TO GET BACK OR HOW MUCH IS GOING TO BE LOPPED OFF THE FINAL BILL. AND WE GET INTO A DISCUSSION BECAUSE HE'S TELLING US HOW MUCH EXTRA THEY HAD TO SPEND BECAUSE THE SUBCONTRACTORS SCREWED UP. AND I THOUGHT, WAIT A MINUTE, THAT WAS YOUR PROBLEM. SO I WAS GETTING MORE AND MORE UPSET, AND I DON'T GET ANGRY A LOT. WELL, I USED TO GET ANGRY A LOT. [LAUGHTER] BUT I DON'T ALWAYS SHOW IT. AND I WAS SHOWING IT AT THIS POINT. AND I FINALLY TOLD HIM, THAT HE DIDN'T KNOW WHAT HIS JOB WAS BECAUSE HE THOUGHT HE WAS IN THE CONSTRUCTION BUSINESS. HE THOUGHT HE WAS GOING TO COME IN AND HE WAS A BUILDING GUY, AND IF YOU'RE A BUILDING GUY YOU COME IN AND YOU BUILD. WELL, NO, HE'S NOT A BUILDING GUY. I EXPLAINED TO HIM, HIS JOB WAS TO TAKE CARE OF ALL THOSE IDIOTS THAT HE INVITED INTO MY HOUSE. AND MAKE SURE THEY DIDN'T SCREW IT UP. AND IF THEY DID SCREW IT UP, IT WAS HIS RESPONSIBILITY TO FIX IT, NOT MY RESPONSIBILITY TO FIX IT. SO LET'S REVIEW. THIS, WHAT'S THIS? THIS. NOW KEEPING WITH ONE OF THE THEMES FROM EARLIER IN THE DAY, HERE IS AN XML. I USED TO SHOW CODES, SOMEBODY MENTIONED WORD PERFECT, I ALWAYS LIKED WORD PERFECT. ALSO THIS WAS A WHILE AGO, WE'VE GOTTEN THROUGH THAT, AND OUR HOUSE IS REALLY VERY NICE NOW. AND THE SHOWER STILL WORKS. SO WE'RE IN GREAT SHAPE. HAD ONE OF THOSE MILESTONE BIRTHDAYS RECENTLY, NOT LIKE LAST WEEK SO I DON'T WISH ME HAPPY BIRTHDAY AFTER WE'RE DONE. BUT IT WAS -- OH, WAIT A MINUTE. NO, THAT'S NOT THE GUY. [LAUGHTER] I HAD FORGOTTEN ABOUT THAT SLIDE. I DREW THAT BECAUSE THAT'S WHAT THEY DO IN MEDICAL JOURNALS, RIGHT? [LAUGHTER] I HAD ONE OF THOSE MILESTONE BIRTHDAYS, AND -- [LAUGHTER] THIS IS NOT THE ACTUAL CAKE. BUT YOU GET TO A POINT WHERE YOU HAVE A MILESTONE BIRTHDAY, YOU THINK BACK TO THE THINGS THAT HAVE CHANGED IN YOUR LIFE. AND I'VE BEEN MULLING THESE THINGS OVER, REALLY KIND OF WITH AN EYE ON NOT NECESSARILY TECHNOLOGY BUT I COME ACROSS SOMETHING, I START THINKING. AH, THAT SURE HAS CHANGED IN MY 25 YEARS OR SO ON THIS PLANET. AND I WAS LISTENING TO A PODCAST. NOW, THE PODCAST TOLD ME ABOUT THIS THING CALLED TASKRABBIT. EVERYBODY KNOWS ABOUT TASKRABBIT. TASKRABBIT, THIS IS THEIR SLOGAN. WE DO CHORES. YOU LIVE LIFE. IT'S REALLY NICE. YOU GO TO TASKRABBIT AND SIGN UP, EITHER AS A WORKER OR AS A NON-WORKER. SIGNING UP TO NOT DO WORK, RIGHT? THAT'S NOT THEIR TERM. SO IF YOU HAVE A CHORE THAT NEEDS TO BE DONE, I NEED MY GARAGE CLEANED OUT, YOU PUT IN A THING ON TASKRABBIT AND THEY TELL YOU THAT THEY ARE GOING TO -- YOU SAY WHAT THE JOB IS, HOW MUCH YOU'LL PAY, AND THEN THEY HAVE TASKERS WHO FILL OUT THEIR PROFILE AND THEN YOU GET HOOKED UP AND THEY COME OVER AND DO YOUR TASK FOR YOU. THIS SEEMS PRETTY AWESOME. I DON'T KNOW, IT'S REALLY KIND OF DIFFERENT THAN ANYTHING WE HAD, MEETING UP WITH PEOPLE TO DO YOUR TASK ON YOUR PHONE. I THOUGHT THAT WAS VERY INTERESTING. SO NOW I STARTED THINKING ABOUT PODCAST. AND WHEN I WAS A KID, I LOVED THE RADIO. WHATEVER HAPPENED TO RADIO? RADIO'S GONE DOWN THE TUBES. IT'S PROGRAMS BY CORPORATIONS THAT PLAY THE SAME SONGS OVER AND OVER AGAIN, AND ALL THE GOOD TALK STUFF IS GONE. NOW THERE'S ONLY, WELL, BAD TALK STUFF, AS FAR AS I'M CONCERNED. NONE OF THE FUN TALK STUFF. IT'S EITHER SPORTS OR POLITICS. SO I DISCOVERED PODCASTS AND PODCASTS ARE GREAT. YOU CAN GO LISTEN TO WHATEVER YOU WANT, WHENEVER YOU WANT. IT'S CONTENT THAT IS DISTRIBUTED, DECENTRALIZED, GET IT WHENEVER AND WHEREVER YOU WANT. ALL YOU HAVE TO DO IS FIGURE OUT WHAT YOU WANT TO LISTEN TO. FILTERING OUT IS ONE OF THE CHALLENGES, FINDING WHAT YOU WANT CAN BE A BIT OF A CHALLENGE. THAT GOT ME THINKING ABOUT TV. I MEAN, PEOPLE DON'T REALLY WATCH TV LIKE THEY DID WHEN I WAS A KID. THURSDAYS AT 8:00 WAS HAPPY DAYS. [LAUGHTER] HEEEY! EVERYBODY WANT HAPPY DAYS. YOU WENT TO SCHOOL AND TALKED ABOUT WHAT FONZIE DID THE NIGHT BEFORE. OH, HE'S GOING TO JUMP THE SHARK! IT WAS INCREDIBLE EVERYONE WATCHED THE SAME THINGS. WE HAD FOUR CHANNELS. NOW WE HAVE A JILLION CHANNELS, AND WE'VE GOT A JILLION CHANNELS THAT YOU CAN GET WHENEVER AND WHEREVER YOU WANT TO. I GET THEM ON MY PHONE IF I HAVEN'T LEFT IT BACK AT MY DESK. SO WE'VE GOT TV THAT'S BECOME DECENTRALIZED. WE DON'T HAVE TO GO TO FOUR CHANNELS AND LISTEN TO WHAT THEY TELL US IS ON THAT NIGHT. OH, AFTER HAPPY DAYS IT WAS LAVERNE AND SHIRLEY. ANOTHER WINNER. SO IT'S DECENTRALIZED BUT YOU GOT TO FIGURE OUT WHAT YOU WANT. I DON'T KNOW IF ANYBODY HAS ONE OF THESE APPLE TVs, BUT EVERYTHING'S ON THERE. AND THEN WHEN YOU CLICK ON ONE OF THOSE THINGS, EVERYTHING ON THERE. IT'S JUST ABSOLUTELY AMAZING. THEN I THOUGHT ABOUT CAR RENTALS. WELL, IT USED TO BE YOU GO TO HERTZ, YOU GET A CAR. THIS IS SOMETHING ELSE THAT'S BECOME DECENTRALIZED. YOU WANT A CAR? YOU GO TO THE INTERNET, PROVE THAT YOU'RE WHO YOU SAY YOU ARE, AND YOU GO DOWN THE STREET AND GET INTO A CAR. IT'S ABSOLUTELY AMAZING. THEY ARE ALSO DECENTRALIZED. AH, SO ONE OF MY PODCASTS I LISTENED TO, THESE GUYS ARE PRETTY ENTERTAINING, THEY ARE ECONOMISTS, THEY TALK ABOUT EVERYDAY THINGS. NOT JUST MONEY BECAUSE EVEN THOUGH I LOVE IT, I FIND IT KIND OF BORING. MONEY, THAT IS. SO THEY HAD THIS PODCAST ON HITCHHIKING, AND IT GOT ME TO THINK. WHEN I WAS A KID, HITCHHIKERS WERE EVERYWHERE. THEY WERE MOSTLY HIPPIES, NOT LIKE THE GOOD CLEAN HIPPIES WE HAVE TODAY. [LAUGHTER] THESE WERE THE REAL HIPPIES, THE REAL DEAL. AND SOMETIME ALONG THE WAY NOBODY HITCHHIKED ANYBODY. WHEN I WAS LOOKING UP THE LINK FOR THIS EPISODE, I SAW FOUR OR FIVE HAD ISSUES ON EPISODES ON HITCHHIKING. WE KNOW WHY PEOPLE DON'T HITCHHIKE ANYMORE, RIGHT? BECAUSE YOUR MOM ... SAYS YOU'RE GONNA GET MURDERED. [LAUGHTER] OR WORSE. BUT HAS ANYBODY DONE THIS? YOU GO TO YOUR PHONE. YOU SAY, HEY, I'M STANDING HERE BY MYSELF AND I NEED A RIDE. COME GET ME. [LAUGHTER] AND SOME GUY PULLS UP, YOU GET IN, HE DRIVES YOU WHERE YOU WANT AND YOU GET OUT. IT'S KIND OF LIKE HITCHHIKING. BUT THERE'S A LITTLE BIT OF A DIFFERENCE. IT'S REALLY THE SAME DIFFERENCE WE HAD WITH TASKRABBIT, I WOULD HOPE, AND THAT IS THERE'S SOME KIND OF TRUST THERE. I MEAN, IT'S NOT REAL TRUST. IT'S A FACT THAT ALL OF YOUR DETAILS AND ALL YOUR PAYMENT INFORMATION ARE IN THE SYSTEM, SO IF YOU MURDER YOUR UBER DRIVER, OR IF YOUR UBER DRIVER MURDERS YOU, THEY CAN TRACK YOU DOWN. BUT THAT'S KIND OF TRUST. SOMETHING TO BUILD ON. HOPEFULLY IT KEEPS A LOT OF PEOPLE FROM GETTING MURDERED. I HAVE TO TAKE A BREAK NOW BECAUSE I DON'T REMEMBER WHAT MY NEXT SLIDE IS. [LAUGHTER] OOOO! SO WE'VE HAD A LOT OF THINGS COME UP, A LOT OF TOPICS THAT HAVE COME UP, AND A BUNCH OF THEM HAVE BEEN KIND OF THEMES THROUGHOUT THE DISCUSSION, AND THERE'S LOTS OF THEM, OPEN ACCESS, AND OPEN ACCESS, ONLINE FIRST, AND OPEN ACCESS, PRE-PRINT, OPEN PEER REVIEW, DO IO, ORCID, OH, BITS, NISO, GREEN, BLUE, VERSIONS OF RECORD, ALL OF THIS. SO IT GOT ME THINKING ABOUT HOW THIS TECHNOLOGY THAT ALLOWS US TO DECENTRALIZE SO MANY BUSINESSES AND ALLOWS ACCESS IN MANY, MANY DIFFERENT WAYS, REALLY CHANGING HOW WE HANDLE JOURNALS AND JOURNAL ARTICLES. I'M ON TRICK PONY. I MAKE SOME ASSUMPTIONS, I'LL SHARE THEM AND WE CAN ARGUE ABOUT THEM LATER. BUT I THINK MOSTLY I'M SAFE BECAUSE I'M SURE I DIDN'T DREAM THEM UP. I HEARD THEM PROBABLY FROM SOME SMART PEOPLE. THE INTERNET WILL CONTINUE TO CHANGE THE WAIT INFORMATION IS DELIVERED. NOT REALLY GOING OUT ON A LEDGE THERE. I THINK IT'S GOING TO CONTINUE TO CHANGE REALLY EVERYTHING THAT WE DO. WE HAVEN'T SEEN THE LAST OF THE INTERNET, I'LL TELL YOU THAT. NEW AUDIENCES WILL WANT THINGS QUICKER. MY NEXT STATEMENT, I DID NOT WRITE DOWN. BUT I SAY IT ALL THE TIME IN THE OFFICE. SO DON'T QUOTE ME ON THIS. YOU PEOPLE ON TV, TURN AWAY. THE PEOPLE WHO WANT PBS FOR EVERYTHING ARE DYING OFF AND THEY ARE NOT GOING TO BE THE AUDIENCE OF THE FUTURE. THEY ARE GOING TO CONTINUE TO BE THE CONSUMERS FOR A WHILE BUT WE CAN'T BUILD OUR SYSTEMS ON AN AUDIENCE WHO IS LEAVING US. WE HAVE TO BUILD OUR SYSTEMS ON THE AUDIENCE THAT'S COMING. AND THEY WILL WANT THINGS QUICKER. THEY ARE VERY IMPATIENT, THESE MILLENNIALS. I HEARD YOU, AUDREY. [LAUGHTER] OTHERS WILL GET THEIR CONTENT OUT QUICKER, ONE WAY OR ANOTHER, BECAUSE THEY CAN. THEY HAVE THE TECHNOLOGY. I'M NOT MAKING ANOTHER 70s TV REFERENCE. THEY ARE GOING TO PUBLISH THEIR CONTENT AS SOON AS THEY CAN, WHEN THEY CAN, HOW THEY CAN, AND IT'S GOING TO HAPPEN, IT'S GOING TO MAKE PEOPLE UNCOMFORTABLE WHEN THEY START TO DO THIS MORE AND MORE, PROBABLY WHAT ALL THE PRE-PRINT TALK IS ABOUT THESE DAYS, AND I THINK WE CAN DECIDE THAT, NO, WE DON'T WANT TO DO PRE-PRINT BUT WE'RE NEVER NOT GOING TO DECIDE WE'RE NEVER GOING TO DO PRE-PRINTS BECAUSE THEY ARE GOING TO HAPPEN. SOME THOUGHTS ABOUT HOW THIS MAY OR MAY NOT WORK, IF PEOPLE GET TOO UPSAID -- UPSET I'LL RUN AWAY. FINALLY -- NO, DON'T LIKE AT THAT! [LAUGHTER] OKAY. SO SELECTION AND DISTRIBUTION, I THINK THESE ARE THE THINGS THAT WE REALLY NEED TO WORRY ABOUT WHEN WE -- I DON'T WANT TO USE CERTAIN WORDS -- GET JOURNAL ARTICLES OUT. SO LET'S TAKE A LOOK AT HOW THINGS WORK NOW IN A VERY GROSS SENSE. WE HAVE A JOURNAL. IT IS A JOURNAL. AND THERE ARE ARTICLES, SO PEOPLE ARE WRITING ARTICLES AND THEY THINK, OH, I'M GOING TO SUBMIT THIS TO THE GREATEST JOURNAL IN THE WORLD SO THEY MAKE THEIR SUBMISSIONS. SO ARTICLES GET CREATED, THEY GO TO JOURNALS, JOURNALS MAKE SELECTION. THEY SELECT WHAT THEY WANT IN THEIR JOURNAL. THIS IS KIND OF BASIC BUT STICK WITH ME. THEY MAKE A SELECTION. AND THEN THEY MAKE A DISTRIBUTION. AND I THINK THESE ARE THE TWO BIG VERBS, EVEN THOUGH THEY ARE NOUNS, THAT WE NEED TO WORRY ABOUT. NOW, I HAVE A FEELING THAT IN THIS MODEL, PUBLICATION IS AT THE DISTRIBUTION. AT THE DISTRIBUTION AREA BUT WE'VE GOT PEOPLE NOW, WE'VE HAD PEOPLE FOR TEN YEARS, MAYBE NOT TEN YEARS, WHO HAVE BEEN DOING THIS. THE ARTICLES ARE CREATED BY OUR OLD PURPLE AUTHORS, AND THEY SUBMIT THEM TO THE JOURNAL WHICH MAKES A SELECTION. BUT THIS JOURNAL HAS THE INTERNET. SO THE JOURNAL MAKES A DISTRIBUTION. THESE ARE OUR PREPRINTS FOR MANY OUR ONLINE FIRSTS. OR MAYBE OUR E-PUBLICATIONS, BUT AT THIS DISTRIBUTION, THIS PUBLICATION OF THE ARTICLE, IT MAY BE COLLECTED IN ISSUES, THEY MAY BE COLLECTED IN ISSUES LATER, BUT I THINK WE'RE GETTING THE POINT WHERE IF YOU'RE DOING THIS, ISSUES ARE UNNECESSARY. SO WHAT NOBODY TOLD THESE JOURNALS IS THAT THE AUTHORS HAVE THE INTERNET TOO. AND THEY ARE GOING TO WANT TO DISTRIBUTE THEIR ARTICLES AS SOON AS THEY CAN. IT RINGS TRUE, I HEARD IT THREE SLIDES AGO. THE INTERESTING THING IS IF I WAS THIS JOURNAL, I WOULD FEEL VERY CUT OUT OF THIS PROCESS. BUT WE'RE LIVING IN A WORLD OF INCREDIBLE INFORMATION. I REFUSE TO READ COMMENTS ON ARTICLES ON THE INTERNET BECAUSE 99% OF THEM ARE JUST RIDICULOUS, AND THERE'S SO MUCH OF IT. SO WE'RE GOING TO ACTUALLY HAVE A LOT OF PURPLE DOTS, MAKING ARTICLES, AND DISTRIBUTING THEM. I THINK THAT THE REAL -- THE REAL VALUE, THE REAL THING WE NEED, AND I HATE TO SAY, YOUR JOB AS A JOURNAL, IS GOING TO BE DOING THE SELECTION OF THE CONTENT, BLESSING IT, GIVING IT YOUR THUMBS UP, BECAUSE THERE'S GOING TO BE TONS OF STUFF OUT THERE, AND THERE'S TWITTER AND THERE'S THE FACEBOOK, AND PEOPLE NEED TO KNOW WHAT'S GOOD TO SPEND THEIR TIME ON BECAUSE TIME IS VERY IMPORTANT. HERE IT IS. WHEN EVERYONE HAS A VOICE, IT CAN GET PRETTY LOUD, AND IT'S OVERWHELMING. SO THOSE ARE MY THOUGHTS ON WHAT YOU ALL SHOULD BE DOING. I KNOW IT'S NOT VERY JAZZY, BUT IT'S BEEN BURNING INSIDE OF ME FOR A WHILE. AND I HAD TO GET IT OUT. SO THANK YOU FOR LISTENING PATIENTLY. THANK YOU ALL FOR COMING TO JATS-CON. AND I HOPE THAT WE WILL SEE YOU NEXT YEAR. [APPLAUSE]