Here are my notes from the San Jose listening session.
Thanks to Dawn Stevens for facilitating the session in-person!
Survey:
DITA 1.1 = 0
DITA 1.2 = 8
DITA 1.3 = 0
no DITA = 3
Specializations = 3
Constraints = 4
Filtering = 7
Branch filtering = 0
Conref/keyref libraries = 7
Keys = 6
Troubleshooting = 0; 1 specialized in DITA 1.2
Release management = 0
MathML = 1
Translation = 6
Linking videos = 3
Learning and Training = 0
Machinery task = 0
Hazard statement = 1 planning to
Highlight element constraints = 5
Separate DITA vocabulary release schedule = 2
DITA 4 Publishers = 2
Dynamic Publishing = 2
Trish Billard
Hewlett Packard Enterprise
100 writers
transitioning from DocBook to DITA
thousands of pubs, and thousands of pages
translation
using 1.2 + troubleshooting specialization
SDL Knowledge Mgmt
Arbortext
plan to move to more dynamic in the near future
Topic based writing is working well
Want: a tag that can go anywhere that can be used for variables (ph is not allowed everywhere – uicontrol for example)
Recommendations for migration tips and tricks
Want proven best practices (what works well, what is no longer used) - reference properties (for example)
Joe Russo
Hitachi Data Systems
50 writers in US (convert as you go), 100 in Japan (determine all the combinations first, then convert everything)
converting dozens of docs (100-1000 pages)
using SDL, Xmetal
DITA 1.2
Intuitive Surgical (surgical robots)
PTC Windchill+Arbortext
DITA 1.2
converting some docs to DITA
conversion is most painful and expensive
want to know about user communities that use PTC
need better examples in the spec. would like better, searchable version of spec
Mark Whisler
Intel Programmable Solutions (formerly Altera)
40 writers, 1 full time architect
sought out an expert to create toolkit/fo etc
Ixiasoft + oXygen 18
producing HTML, PDF, webhelp
conversion was in Robohelp/Frame – used Stilo Migrate for conversion
cleanup with XSL and brute force, but that was excruciating (and stupid), mostly because of how the source content was written, rather than DITA itself.
Needed an internal spec before they got too far…
Make the spec more accessible. Have to drill down to find what need. Need better examples in the spec.
Migrating their web presence at the same time as DITA migration, so decisions were made that had impact.
Would like more docs on how to do specialization.
Yas Ettesam
Wind River
32 writers (info dev + customer experience)
DITA 1.2
Ixiasoft + Oxygen
Specialization
no translation
Working well: have an info architect, tools team
Getting engineers or from China to write content. Having to have the writers create and edit the content instead
No use for 1.3 for now. Want to know how far out for the specs
Toolkit works well and like to know what is changing
Updating transforms and toolkit is what was painful
Wind River makes real time operating systems. Many elements that they had to add to software and programming domains. Want more for Java (class, method, function, macros, libraries)
Constraints worked pretty well and they used it a lot.
Training content is produced in DITA (but not using the Learning specialization)
Fought pressure to adopt a tool until information model was complete!
Steve Liang
VMWare
More from tools side
100 writers, 1000 pubs
13 langs moving to 20 langs
DITA 1.2
SDL + Xmetal
DITA-OT 1.5, 2.x
Reuse has worked well
SDL tools are difficult
Conversions expensive
Grace Murao
FICO
adopted DITA 1.2 almost 10 years ago
SDL + Xmetal
upgrades are painful
SuiteSolutions for SuiteHelp
PDF, ePub (dita4publishers)
created massive DITA writers guide, which was very helpful
Framemaker files were their baseline for look and feel and structure, so haven’t deviated a lot
Used a fair amount of constraints, but also based on what was available in Frame
Output constantly has issues (some due to writers not structuring properly)
7 different languages – DITA has been helpful there
have to ask yourself "what do they really need" before they migrate or convert. Try to clean up writing ahead of time.
Getting new writers up to speed quickly is a challenge
How to reuse documentation in the new world of social media?
Do use some embedded media, but need to make it available on different channels
Learning XSL and XSL:FO is really a challenge. Would be nice if output configuration was much easier
Jane Chow
| KLA-Tencore |
Not using DITA yet
Evaluating different vendors (SDL, Ixiasoft, Quark)
Trying to determine what tools would work better
Using Documentum now
33 writers (13 in US, some in India and Israel)
Hardware docs are mostly in Word now
have a lot of reuse potential
Vidya Kameswaran
| KLA-Tencore |
hardware documentation
evaluating DITA and SmartContent (Quark)
most writers are not DITA trained, so there is pushback on whether they want to adopt it. Many are not comfortable with tagging.
Docs are modular, but not down at a lower level.
How to find writers with DITA experience?
Louise Leck
Ring Central
3 writers (more offshore)
Use DITA as an interface for what is sent to a 3rd party
Use oXygen
How to help organizations with DITA adoption? business cases, implementation strategy/recommendations, output configuration needs to support graphic-intensive layout, embedded media recommendations
Want more DITA resources
Miles Gilster
Guidewire (insurance software)
17 writers (some distributed across Ireland, Poland, US)
DITA 1.2
Xdocs + oXygen 15 (for authors), 18 for architect
DITA-OT 1.8.5
PDF, webhelp
Small amount of content in DITA, most in unstructured Frame
Localization has worked VERY well. Saved 30% on traditional translation!
Expensive/difficult: in-house conversion.
Need more guidance on migrating content, pre-conversion prep
More wholistic examples
30-40% translation savings
one company reported 70% savings!
Exchanging content with training would save a lot, but its a tough sell.
Sanjay Tandon
Applied Materials
Word and unstructured Frame
5-year effort to transition to DITA
CMS was Sharepoint (very painful)
1200 page manuals, many graphics and links
Now using DITA 1.2 with SDL
8200 writers federated into different business units and different companies
Don’t have a single CMS and structure to their docs (each biz unit has different structures and needs)
Still all about the “Look and feel” of the output. Getting transforms to work right doesn’t work well. Only 1 guy can do the transforms.
How to get their corporate look and feel in the output
DITA is very overwhelming for first-time users
Tech pubs and training will use DITA, but when people see the structure, it is too hard. They want it easier in Powerpoint and Word (as example of ease of use)
Using Stilo Migrate and then 1-2 weeks to clean up the writing
3500 service engineers, but want to harness their expertise (especially for troubleshooting and other technical content)
Joe Padre
Tibco
30-40 writers in US, UK, India
PDF, webhelp
5-6 languages
SDL + Xmetal
Integration is way too reliant on CMS vendors
want more info on workflow
Stefan Millett
Sigma Tech / Ericsson
Migration in infancy, so looking for best practices
Leila
Brocade
60 writers
80,000 objects in repo
DITA 1.2
SDL Knowledge Center
Xmetal 9
Suitehelp and Oberon to help with stylesheets (but stylesheets are expensive)
not using constraints, but using a very limited set
75% reuse across 3 platforms! 5 platforms using DITA overall
conditions and variables are very helpful
Don’t see why DITA is so hard, since people write HTML all the time!
Stylesheets are very expensive.
Versioning in SDL works REALLY well
Fonto is being used for simple interface for contributors, but it is limited
Adobe Experience Manager is a challenge (metadata entry is manual)
More use case and examples would be much more helpful
unstructured Frame – brought it over all at once
Using Stilo for some migration
For more information, people go to:
Spec – appreciate the alphabetic listing of elements
oXygen, suite versions are easier to use
dita-users list
SDL user group
Ixiasoft user group
consultants
Don’t understand why you would WANT to use key scopes (from spec). Need more use cases for DITA features.
webinars
"Convince writers that tags are NOT for appearance, but for semantics. Think about how it can be used for search and new ways of using the content, not producing the same documents they used to.
You learn more about teaching DITA to others than being taught!”
What to do if you think you have a case for reuse? What is the structure /recommendation (put it in a common folder) above the current doc structure.
How to get around the lack of trust? Many people copy instead of reuse.
Educate on the paradigm shift away from the book model to the multi-channel knowledge base.
WYSIOOO – What You See Is Only One Option! Content can be used in many different ways!
Suggestions for Adoption TC:
conref-push power really isn’t communicated well.
Getting graphics to appear better (size vs resolution) in output
Getting output to look better (XSLT, Xpath)
Still very dependent on formatting. Promote the entire ecosystem (spec, style). How to overcome the limitations? Get the output more mature.
How to make content more searchable and findable?
Thanks and best regards,
--Scott
Scott Hudson
Content Strategist
Digital Aviation Learning & Development
Voting member:
OASIS DocBook TC, Publishers SC
OASIS DITA TC, Tech Comm SC, LW DITA SC, Learning Content SC
OASIS Augmented Reality in Information Products (ARIP) TC
Jeppesen |
Digital Aviation | Boeing
55 Inverness Drive East | Englewood,
CO 80112 | www.jeppesen.com
This document contains only administrative, uncontrolled data under U.S. International Traffic in Arms Regulations.