DITA is an OASIS "standard" that has been steadily supplanting other
"standards" like Docbook to become a dev-ops go-to for information
architecture: http://docs.oasis-open.org/dita/v1.2/os/spec/DITA1.2-spec.html

Having said that, it comes down to a business-versus-bureaucracy thing.
 One of the rough ideas bandied at Apachecon was to develop "a strategy
to encourage more business users," and typically that means minimizing
the bureaucracy of the project (as in: don't scare away business
managers who are assessing OFBiz on behalf of their organization).

DITA's advantage over things like wiki-only or HTML-only (or trying to
maintain wiki markup plus HTML markup plus whatever else) is its ability
to store content as single-sourced files and then transform those files
from their native XML format into appropriate end user formats (e.g.
XHTML or PDF/FOP or SCORM or even some wiki markups depending on the
plugged-in extension to the transformation engine).  For information on
the most popular transformation engine (which takes raw DITA files and
transforms them from their native XML format to a format that end users
can read), try this: http://www.dita-ot.org/1.8/.

Most OFBiz documentation contributors, understandably, have been
comfortable gaining professional experience using things like Docbook
(which is by no means obsolete but rather is suffering from the recent
migration toward DITA), so any embrace of DITA could necessitate some
kind of bureaucratic dictate that "all contributors will now write
documentation based on the DITA standard," which is not likely to go
over well with either the writer-contributors or the business managers
assessing OFBiz while hoping to avoid overly-bureaucratic open source
projects.

Probably the best approach, at least for the next few years, is to
encourage doc commits of DITA-based source files while continuing to
accept with thanks everything writers are willing to contribute.  Some
of the Open Toolkit engine transformations (which at the end of the day
are ant-target-thingies) might be used to backward-transform someone's
Docbook contribution into DITA format.  A goal worthy of debate on the
developer ml is to consider a day "three years from now" when:

- most tech writers are contributing DITA files
- transformations exist to turn other contributed docs into DITA files
- transformations exist to turn single-sourced DITA files into
appropriate end user documents (wiki pages, etc.)

The biggest advantage of single-sourcing documentation is that when
OFBiz framework specs change you need only edit one file to enable
necessary updates to information destined for multiple end user doc
targets, instead of wondering if there's another wiki page (or PDF or
whatever) out there somewhere that got overlooked.

Beyond this overview I'm too tech-tetched to help much with putting the
transformation engine through a set of tests or anything like that.
Those who know Ant should have few problems experimenting here & there
(resources permitting of course).

I will add some final links to peruse:

The Derby project appears to use DITA as the basis of its documentation:
https://db.apache.org/derby/manuals/index.html

The FOP project appears to be debating whether or not to tighten up the
integration/coupling of DITA and FOP:
https://mail-archives.apache.org/mod_mbox/xmlgraphics-fop-dev/201403.mbox/%[email protected]%3E

The OpenOffice project appears to have been considering DITA as a format
for its documentation-related source files (although Confluence markup
as an engine output target doesn't appear to be implemented inside
Apache or elsewhere):
https://cwiki.apache.org/confluence/display/OOOUSERS/User+Documentation+Plan




On 14-11-26 07:38 AM, Pierre Smits wrote:
> Todd, all,
> 
> Don't you just love those acronyms.... What is meant by that (DITA)?
> Remember, the audience is diverse.
> 
> DITA as opposed to PITA?
> 
> Al jokes apart. Documentation is important with respect to adoption. From
> all angles: business, development, deployment/implementation, etc.
> 
> Regards,
> 
> Pierre Smits
> 
> *ORRTIZ.COM <http://www.orrtiz.com>*
> Services & Solutions for Cloud-
> Based Manufacturing, Professional
> Services and Retail & Trade
> http://www.orrtiz.com
> 
> On Wed, Nov 26, 2014 at 4:11 PM, Todd Thorner <[email protected]>
> wrote:
> 
>> Quick opinions on the rough ideas...
>>
>> 1. Turn the "kernel" into something based around containers (e.g.
>> docker).  This approach could also apply to many dev-ops considerations
>> that are internal to the project.
>> 2. Make DITA the basis of all documentation.
>>
>>
>>
>> On 14-11-26 04:13 AM, Sharan-F wrote:
>>> Hi Everyone
>>>
>>> Please see below for the link to the notes from the meeting we held at
>>> Apacheon in Budapest last week.
>>>
>>>
>> https://cwiki.apache.org/confluence/display/OFBIZ/Apachecon+Workshop%3A+19th+November+2014
>>> <
>> https://cwiki.apache.org/confluence/display/OFBIZ/Apachecon+Workshop%3A+19th+November+2014
>>>
>>>
>>> A key point to remember is that no decisions were made and that these are
>>> just the notes from our discussions.
>>>
>>> As you will see, a lot of ideas/proposals came up that need community
>>> feedback, discussion and opinion.
>>>
>>> Please feel free to provide any feedback or comments using this mailing
>> list
>>> thread.
>>>
>>> Thanks
>>> Sharan
>>>
>>>
>>>
>>> --
>>> View this message in context:
>> http://ofbiz.135035.n4.nabble.com/Notes-from-Apachecon-EU-Budapest-Meeting-tp4658991.html
>>> Sent from the OFBiz - User mailing list archive at Nabble.com.
>>>
>>
> 

Reply via email to