Skip to content

Meeting minutes 2012 09 25

robander edited this page Sep 25, 2012 · 5 revisions

Attendance: Robert, Eliot, Jarno, Kris, George, Radu

Item 1: DITA-OT 1.7 Milestone 3

Posted yesterday. Mostly contains fixes, low level items that are not visible to end users. Also pulled in all of Kris's updates.

Item 2: DITA-OT 1.6 latest fix pack status

Released 1.6.3 last week, contains several fixes. Currently have a branch open to allow fixes to be checked in for 1.6.4, but no planned release date (will only release if something significant comes up or a large number of minor items are collected).

Item 3: Plans for DITA-OT 1.7 Milestone 4

Estimated release date: October 29

Note to those hoping for any significant new features: should begin contribution process; DITA-OT unlikely to be able to review / accept any major new features that begin the process in the final two milestones.

Flagging items from Robert - continue work that began earlier this iteration, check in some updates in this milestone. Has some impact on developer docs for new output types.

Jarno:

  • working on removing dependency to ditaext property (causes all DITA topics to change to .xml or .dita in the temp directory), make better use of the format attribute to recognize DITA topics.
  • Background work: cleaning up way file lists are generated. While doing that, likely create a new version of the "dita.list" file that does not have the comma separated lists, use an XML file that removes any blocks on what can go in a file name.
  • Depending on time/interest, may work on github tracker items

Call for assistance from Jarno: in DITA there is a duplicates attribute in linklist/linkpool, to determine whether duplicate links should be created. Shouldn't be difficult, just means creating if/else statement in rel-links.xsl file. Looks like the specification declares this is "mostly unsupported" - Robert says language like that is generally a sign that something did not work in 1.0, and no tool developer has tried to update it, otherwise the language would have been clarified.

Item 4: Other topics

Jarno: changed names of branches in Git. Old names were based on IBM China branch names in SourceForge. New system is simpler and more in line with other projects; uses "gitflow" branching model, which is supported in many tools. Now the latest code is in the branch "develop", which should also make it easier to continue working with the branch as each release finishes. Bug fixes will be developed in branches that have a prefix of "hotfix/" but those are deleted after merged back to master. This means we will be relying more on tags, removing old branches names that no longer have meaning.

Other note: our XSL for a while has had the ability to reference other XSL files using plug-in IDs. Jarno making use of this to move troff/ code into a plug-in, makes it easier to drop that entire plug-in, as well as to more easily locate files (all troff code encapsulated in one spot). Moved TROFF and RTF because they are much less used; core items like preprocess and XHTML are staying put for now. Will likely move eclipsecontent as well - Robert expects that is not used much.

Eliot: Ran in to item recently: output of mappull process, expected topicref's that use a keyref would get the effective href derived from the keyref onto the reference. Challenge - in XSLT, can resolve a keyref, but haven't checked if resolved map has keys from submaps in way that precedence is preserved.

DITA for Publishers: trying to get latest version out, with new HTML5 transform, updates to PDF/HTML processing, more complete map driven framework. Doubt will be solid enough to contribute to 1.7. Don't want to dump a bunch of half tested code into the toolkit. Also have to update for 1.6 and 1.7, so more testing there.

Radu encountered issue with issue tracker from SourceForge - looking at the older bug - UI of sourceforge has changed. XML code that was listed in the bug is now missing, only there if you view the source of the comment. Jarno points out - we have trackers in github for every SF item. Only reason to keep SF visible to public is that many items have attachments, which could not be copied over to github.

Item 5: Next contributor meeting

Meeting to kick off Milestone 5 planned for October 30

Clone this wiki locally