AAUP Web Seminar: XML Workflow Case Study
1 like814 views
The University of Michigan Press has been transitioning to digital workflows since the 1990s, including collaborations and the creation of digital culture books. Key strategies include adopting XML-first workflows and developing new digital scholarship products. The emphasis is on establishing effective digital properties to ensure long-term sustainability and relevance in publishing.
![XML at University of Michigan Press A case study 2008-2010 April 2010 Karen Hill Assistant Director and Digital Manager [email_address]](https://image.slidesharecdn.com/aaupxml2010-100429145125-phpapp02/85/AAUP-Web-Seminar-XML-Workflow-Case-Study-3-320.jpg)
![Questions? Karen Hill [email_address]](https://image.slidesharecdn.com/aaupxml2010-100429145125-phpapp02/85/AAUP-Web-Seminar-XML-Workflow-Case-Study-26-320.jpg)
AAUP Web Seminar: XML Workflow Case Study
- 1. “ The first concern is not the backlist. The first concern is to have creation properties in place today that mean that you won’t be worried about 2010 books in 2013. So the first thing to do is go to a digital workflow.” --Mike Shatzkin quoted in Book Business March/April 2010
- 3. XML at University of Michigan Press A case study 2008-2010 April 2010 Karen Hill Assistant Director and Digital Manager [email_address]
- 4. University of Michigan Press Some digital history… 1990s: Press experiments with electronic publishing, including CD ROMs and the web-based Journal of Electronic Publishing February 2006: UM Press and UM Library’s Scholarly Publishing Office agree to collaborate on digitalculturebooks, a joint imprint publishing works both online and in print and experimenting with new business models October 2006: DCB titles launch with online text and print options August 2008: UM Press hires Digital Manager, committing resources to develop digital versions of its backlist and to build a more digital-first front list October 2008: Press launches investigation into XML as a way to support digital migration of content April 2009: Press products enter into digital distribution and dissemination paths via third-party distribution July 2009: UM Press officially joins with the library to focus energy on developing products and services that support digital as well as print scholarship. October 2009: HathiTrust digital library creates free-viewing portal for UM Press scholarly content October 2009: The Press becomes a part of MPublishing, a University of Michigan Library organization dedicated to disseminating scholarly works in all formats and to bring scholarly publishing communities at the library and across the university together in close collaboration and partnership. Library appoints Maria Bonn as Associate University Librarian for Publishing to oversee and guide MPublishing operations November 2009: digitalculturebooks re-launches with a new web site that begins the first stage in developing web communities for Press and other scholarship November 2009: Press web site allows for multi-format purchasing (audio, Kindle, PDF, etc.) March 2010: Press launches site to allow online commentary (public or private) of key Press works
- 5. 2008: Facing (and embracing) the Digital Future
- 7. Back to the Future (2010) Publishers Lunch Survey March 2010 http://www.aptaracorp.com/index.php?/eBook-survey1.html 46% of publishers responding do not produce ebook formats of their products
- 8. Back to the Future (2010) Perseus Distribution Survey February 2010 Polled 70 publishers on ebook/digital format
- 9. Back to the Future (2010) Google Editions Survey March 2010 Quick poll to determine file formatting options from its customer base
- 10. Now Back to 2008: What were we thinking? UM Press defined the following needs to as part of its strategy to become a digital scholarship publisher: Build cross-searchable collections and collaborations Create reflowable text for e-readers Link text to multimedia elements Present text in various stages of production (e.g., online review) Chunk text for reuse (iPhone apps) Own its content’s future
- 11. October 2008: Where do we start? Everyone, and no one, was an expert. Did we want to follow others, or figure out our own way? Hired an independent xml expert to help us start Pros: Minimizes the follow-the-colleague solution that may not be best for you Minimizes the potential for internal bias Allows more open discussion and critique of options Cons: Expense Getting expert up to speed on priorities, strategies, products Being brutally honest about ourselves
- 12. 2008: Assumptions Authors cannot do xml tagging CMS systems will not be purchased or built UMP remains independent from vendors No cut-over migration. Titles and products move to xml workflow over time
- 13. 2008: Problem Statement Given the strategy, assess current formats for suitability and scalability; Define any upstream workflow processes affected Determine if current digital states (PDFs and Google Search files) continue to be practical for growing digital revenue or reach
- 14. 2008: Analysis Phase Interview downstream recipients of current digital content (Scholarly Publishing Office) Interview internal staff Assess current university press environment Assess current technology in use: DocBook TEI-lite NLM ePub (or OEB) DITA ACLS DTD (HEB)
- 15. 2008: UMP Recommendations Four options or flavors: 1. DocBook for xml-first 2. Continue with PDFs 3. Split production with DocBook xml and PDFs to manage transition 4. Adopt Adobe InDesign to xml format for ePub (or Epub, or EPub, or EPUB…)
- 16. 2009: UMP Decision Adopt two main workflows: XML first workflow XML Post-Production workflow DocBook for xml-first titles – digitalculturebook series; some “enhanced” books; a few others Use Adobe’s InDesign—to—XML process to create ePub or DocBook xml as appropriate And of course, continue with PDFs as is currently acceptable by industry, and Adobe advancements may be of interest
- 17. 2009: Changing the Workflow to Accommodate Digital Where should we begin? XML-first workflow impacts the handoff phases of work as manuscripts pass from copyediting to Production xml editing tools DocBook designs Conversion testing Altering print-focused steps XML Post-Production is low-impact; affecting output at the end of the print production tasks – creating PDFs for the printer and ePub files for digital Output ePub and DocBook Review
- 19. 2009-2010: Projects to Help Define the New Workflows XML-First: About 10 at most in current pipeline. Pilot: The Accidental Teacher online review and commentable blog test for vendor xml conversion XML Post-Production: digitalculturebooks.org – DocBook Kindle, etc. - ePub Benetech – daisy xml
- 21. 2010 and Beyond Copyediting Tests: First-pass with DocBook tags? Highly segmented data (i.e., chunks): Enhanced ebooks and databases iPad and other mobile apps XHTML5? http://thinkubator.ccsp.sfu.ca/wikis/xmlProduction/XMLProductionStartWithTheWeb New ePub version http://www.openebook.org/idpf_groups/IDPF-EPUB-WG-Charter-4-6-2010.html
- 23. “ The first concern is not the backlist. The first concern is to have creation properties in place today that mean that you won’t be worried about 2010 books in 2013. So the first thing to do is go to a digital workflow.” --Mike Shatzkin quoted in Book Business March/April 2010
- 24. Experience So Far Define your own strategy that fits who you are Experiments build experience Talk with your colleagues – learn new things or just simply validate Don’t let your content and its formats be “owned” by others
- 25. A Few Tips from UMP Colleagues Remove display triggers from source materials Simplify a list of press-specific tags Work with vendors on scripts to eliminate conversion problems Think about your artwork, footnotes, etc. in digital form Standardize what you can (e.g., Copyright page) Get xml savvy and use the free software to get familiar with it Prioritize what you want to accomplish
- 26. Questions? Karen Hill [email_address]
Editor's Notes
- #16: that DocBook is the best choice for UMP (and for most publishers!) -- but for not all the same reasons. For me, the main reason to support DocBook is that the standard schema is complete yet inflexible, so there are lots of simple tools available for it that integrate into other publishing workflows. TEI tries to meet scholars' needs for texts, making it so flexible that you can't write generic tools for it.
- #17: XML first: Need to managing costs and staff time during this roll-out of new processes.
- #18: Remove as much of the display triggers as possible in your files. Simplify your own tags so conversion is easier. Post Production results: Don’t stress on imperfections of conversions, or be willing to pay. Work with vendors on scripts to eliminate, ways to simplify (like Copyright page)
- #25: All books do not need to be xml upstream, at least today. Post-production may be fine. Try this stuff out– don’t let the technical challenges stop your own knowledge from growing. We’ve done the process management internally with redirected staff and some workstudies/temps. Costs are mostly around conversion from old formats to new ones. Can be significant– hence the tendency of some to allow vendors to do for free with strings or limited control over output.
- #26: All books do not need to be xml upstream, at least today. Post-production may be fine. Try this stuff out– don’t let the technical challenges stop your own knowledge from growing. We’ve done the process management internally with redirected staff and some workstudies/temps. Costs are mostly around conversion from old formats to new ones. Can be significant– hence the tendency of some to allow vendors to do for free with strings or limited control over output.
- #27: All books do not need to be xml upstream, at least today. Post-production may be fine. Try this stuff out– don’t let the technical challenges stop your own knowledge from growing. We’ve done the process management internally with redirected staff and some workstudies/temps. Costs are mostly around conversion from old formats to new ones. Can be significant– hence the tendency of some to allow vendors to do for free with strings or limited control over output.