DocBook Technical Committee Meeting Minutes: 17 Sep 2002 ======================================================= | The DocBook Technical Committee met on Tuesday, 17 September 2002 | at 01:00p EDT (10:00a PDT, 17:00GMT, 18:00BST, 19:00CEST, 02:00JST+) | for 90 minutes. | | Agenda | | 1. Roll call Present: Paul Grosso Dennis Evans (:15-) Dick Hamilton Nancy (Paisner) Harrison (:10-) Richard Lander Sabine Ocker Michael Smith (-:30?) Tim Teebken Norman Walsh Regrets: Bob Stayton | 2. Accepting the minutes[1] of the previous meeting Accepted. | 3. Review of the agenda None. | 4. Review of open action items {5 min} | | a. Bob to write a more detailed proposal for instance-based generated xref text Completed[2] ACTION: Norm to put on the agenda for the next meeting | b. Bob to investigage RFE #558443 include storage info in metadata Completed[3] ACTION: Norm to post a solution that uses existing metadata | c. Norm to construct some sort of IPR template Completed[4]. ACTION: Norm to put on the agenda for the next meeting | d. Dick to experiment with SiberSafe for help authoring background Continued. | e. Norm to follow-up on RFE 473365: Allow optional in funcprototype Completed[5]. | f. Norm to follow-up on RFE 562343: element Continued. | g. Norm to post combined table markup for CALS+HTML Continued. | h. Nancy to summarize what DITA and TEI have to say about general principles | for adding new markup NP: TEI does it like we do: define things in PEs and classes, recommend that people modify things by redefining existing things or suppressing elements. Like DocBook, they say if you do anything outside this we can't promise that tools will work. Darwin is different, they expect people to create domains and topics. Domains are the hierarchical models. Topics are their basic element category. So you can modify and customize either by adding a domain specialization and putting the constituents together differently or you can extend a topic definition. They restrict you to using the existing building blocks. You can extend a topic, but it has to be stuff that's already there. They don't let you create new elements that don't already have content that's there. There's more flexibility in creating new domains. They have an API domain, but you could create a specific object-oriented API domain. That would be fine because a method would use existing topics. TEI and Darwin differ in that they've already modularized the gross structural level so that instead of having a single basic DTD they have a set of them. For example, TEI has base and a few different hierarchies and they expect people to come up with some more. This does make it easier for people to think about modification, becauase you're starting with something that isn't so large. | i. Norm to put principles for new markup the agenda for next month. Completed. | j. Richard to provide some concrete examples of what might need to be added | for HTML Help Continued. | k. Sabine to provide some concrete examples for HTML Help Continued. | l. Norm to check with submitter on RFE #571998 Completed[6]. | m. Paul to investigate what the content model for a bidi element needs to be | and why simply using a nested phrase is insufficient (or is it?). Completed[7]. ACTION: Norm to put on the agenda for a decision next month | n. Norm to investigate namespace convention in MathML Module (RFE #582705) Continued. | o. Norm to post solution for RFE #582822 for discussion Completed. | p. Norm to publish SVG Module Completed[8]. | q. Norm to republish simplified ToC proposal Continued. | r. Norm to propose solution to submitter of RFE #473365 Completed[9]. | 5. Next meetings | | Our next telcon is 15 Oct 2002. Need to update the reservation beyond October. ACTION: Sabine follow-up in email when Sun's hosting is resolved. | Is anyone going to be at XML 2002? Apparently just Norm. | 6. What about "this version" markup Obsoleted by Bob's action item "b." | 7. Simplified 1.0 We're now in a one-month countdown. If no bug reports are received in a month, we'll make it 1.0 officially. | 8. Priciples for new markup NW: In addition to new markup principles, I've been working on a tool that may let us generate the various schema versions from a single master "meta" source. More updates as I have them. RL: Microsoft is only interested in W3C XML Schema implementation. We took a look at the existing implementation and decided to reimplement it using our own methodology. RL: There's a lot of interest in making this information public, but we haven't pursued that yet. I'll probably have news before the next meeting. NW: I'm not surprised that a different methodology was chosen. ACTION: Norm to put schema discussion on the agenda for next month. RL will attempt to get the Microsoft Schema published before the next meeting. NP: I'm impressed with TEI and Darwin. It would be really nice to divide up hierarchies so that at we have, for example, and Article module that includes the stuff articles need and then other modules wouldn't include that stuff. DE: I would like to have different content models in different contexts. NP: If we provided guidelines for getting into DocBook for the five or ten most common things that people need to do with it or the five or ten most common critical tasks, I think that would be very helpful. TT: This sounds a lot like the pagetype discussions that we were having before. NP: What I'm thinking of is supplying the outer shells. So instead of just DocBook, we'd have the article, api, command line, and html help DTDs. NW: I've never seen this work very well in practice. I'd love to see some examples. DE: I'd like to have RefEntry be a subset all the way down to the individual elements. NW: I'd like to see some examples of that too. RL: As part of my work, I've been exploring this issue. When I get around to publishing this work and some of the related documentation, I think it'll be some of the stuff we're looking for. One thing that came to mind was the work [Norm] has done on Website. NP: One possibility is for us to try purging them. Not actually purging them, but add another module that is the simplified inlines. Just 10 or 11 inlines and say when you do something simple, you can just pick that group. ACTION: NW to post more detail about the PE reorganization with some documentation. NP: I think many people would find simplicity more compelling than completeness. DH: Another thing we might think about is the fact that having subsets is important and we could work with vendors and tools suppliers to make subsets that work for tools and environments. You wouldn't need to change the DTD at all, just configure the tool. SO: I think it's the ramp-up that's intimidating. Access to all of DocBook is overwhelming. To make it easier to figure out what sorts of subsets of elements are required to achieve the functionality that you need. | 9. Discuss Help Authoring Continued pending action items. | 10. Review of Requests for Enhancement {40 min} | | To browse a specific RFE, enter the URL (on one line): | | http://sourceforge.net/tracker/index.php?func=detail& | group_id=21935&atid=384107&aid=XXXX | | where XXXX is the RFE number. | | 598994 arg vs option DE: There is room for confusion. Doing cmdsynopsis is hard for an author. I don't think this is something we could easily rewrite to make it easier. DE: Option and optional are only similar in spelling, they have completely different meanings. ACTION: Norm to write better documentation. Post for discussion. ACTION: DE to find out if he can post the internal style guide documentation on this issue. | 606071 allow glossary in place of glossdiv Explain that it would complicate processing. You really only want to go down one level and that's not possible if you make it recursive. Rejected. | 609061 Add new Step sibling element Alternative SO: Steps have the semantic of being sequential, so the thought was that branch would be less semantically confusing. NP: I assume the content module would be like itemizedlist. SO: No, the content model of branch is the same as step. RL: If you think about W3C XML Schema, you have sequences and choices and it seems like this is more-or-less what we're talking about. NW: I think that's what we get here: procedures contain a sequence of steps, altenatives allows one to choose among several branches. NP: Alternatives to first step? SO: Yes. PG: If procedures are sequences, and alternatives are choices, why call the thing inside branch? SO: I think it could be inferred from the content model, but the group that made this request thinks its ambiguous and confusing to writers. RL: That's like saying people get confused if they have multiple folders with the same name. PG: And we already have 'title' reused all over the place. NP: But these are the same thing, and using a different name confuses the issue. DE/SO: They aren't the same. NW: Maybe a couple of straw polls will help us settle the issues. Poll 1: The proposal as written attempts to prevent alternatives from occurring recursively (that is, a step with alternatives occurring as a descendent of some other alternative step). Are you in favor of allowing or excluding alternatives: Allow: 4, Exclude: 2, Abstain: 3 Poll 2: Are you in favor of using 'branch' or 'step' inside 'alternatives': Branch: 2, Step: 3, Abstain: 3 NW: Not very useful. We seem to have general approval of the idea; we need to work out the details. Take it to email. | The following RFEs have been moved to the bottom of this agenda in order | to make sure that all RFEs get fairly discussed. | | 514435 Allow reference within refentry Mike: can you summarize where this issue stands? | 558443 include storage info in metadata Closed. Use existing elements. | 564776 process/service/daemon/server markup ACTION: Norm to find record of previous discussion. Didn't we fall back on systemitem? Adjourned. | 565716 URL and URN markup | 565905 Add xrefstyle attribute | 570068 Online values for pubwork | 571996 Add 'namespace' inline element | 571998 Add 'default' inline element | 573419 Add bidirectional text overrides | 573812 Allow blockinfo on blockquote | 574880 Add annotation element | 582705 Namespace convention for MathML Module | 582822 VARARG and FUNCDEF together | | The following RFEs are awaiting action items | | 431411 RFE 70: add generic linking capability | 472229 Allow HTML tables | 522552 Add title attribute to element | 413389 Enhance METHODNAME and VARNAME | 436067 splitting tech.char.class | 473365 Allow optional in funcprototype | 482818 Simplify ToC content model | 562343 element | 565637 Associate non-inline image with link | | The following RFEs are identified as V6.0 or later | | 531851 Remove inline person name elements | 532088 Remove RevHistory from qandaentry | | [1] http://lists.oasis-open.org/archives/docbook-tc/200208/msg00002.html [2] http://lists.oasis-open.org/archives/docbook-tc/200209/msg00000.html [3] http://lists.oasis-open.org/archives/docbook-tc/200207/msg00010.html [4] http://lists.oasis-open.org/archives/docbook-tc/200209/msg00005.html [5] http://lists.oasis-open.org/archives/docbook/200209/msg00152.html [6] http://sourceforge.net/tracker/index.php?func=detail&aid=571998&group_id=21935&atid=384107 [7] http://lists.oasis-open.org/archives/docbook/200208/msg00172.html [8] http://lists.oasis-open.org/archives/docbook/200209/msg00145.html [9] http://lists.oasis-open.org/archives/docbook/200209/msg00152.html