- Tei-council - lists.tei-c.org

Pure ODD progress
by Lou Burnard 09 Dec '15

09 Dec '15

I am now at the stage of being able to generate web pages in the lb42-pureodd-2 branch. To get there, I have had to comment out of the makefile the two "EXTENDED" test files testp4compat.xsd and testi18n.xsd (the latter was already commented out in fact). Can anyone remember/reconstruct what these tests were intended to do? I am also now using Hugh's latest Stylesheets, so have some further comments to make on them: Firstly, what appears to be a bug: in the pureODD section, I see nothing at all inside the <content> element. e.g. in emph. This looks OK at http://teic.github.io/TEI/ref-emph.html but not in the version I have locally generated, so either the former is not up to date, or I have messed something up. Secondly, I still don't want to see class declarations inside that section Thirdly, I don't want to see element names clickable inside examples. This is not what we currently do; I find it distracting; and it's strictly speaking wrong, since example contents are not in the TEI namespace! :-) [this also doesn't happen in the github.io version]

4 13

anyone for @exclude ?
by Lou Burnard 08 Dec '15

08 Dec '15

Wandering the anyXML rabbit warren, I came upon a new side passage which may be worth considering. This is to define in pure ODD a new attribute @exclude on the <content element, the effect of which is to specify one or more elements which should be excluded from said content model. We could implement this in DTD land (if we still did DTDs) with an exclusion exception; in Relaxng of course it has to be a schematron rule. It would enable us to do two things we currently do rather obscurely (1) prevent <teix:egXML> from containing <teix:egXML> (2) prevent <s> from containing <s>. And there may well be others. Unless anyone can plausibly show that this is bonkers, I will create a ticket.

1 0

Contributors
by Hugh Cayless 07 Dec '15

07 Dec '15

Dear All, Thanks for the productive meeting earlier today. I’d said I wanted to broach the topic of keeping Contributors engaged over email, partly because it’s been a little contentious in the past and partly because I think it’s something that deserves more deliberation than we can give it in a single meeting. I’ll start by setting out how (as I understand it) things work now and then paint a picture of things as I’d like them to be. I mean the latter as something for you all to react to, so don’t hesitate to do so. I’d particularly like reactions from those of you who are new to Council. This is really going to be an ongoing conversation, I believe, and not something we can just fix. Martin has volunteered to be a champion for Contributors and I’m hoping that he and any other outgoing members who want to remain active participants will be vocal about what works and what doesn’t work for them. Council is a highly effective group that does a lot of good work, so we need to find a balance that doesn’t mess with our effectiveness and cohesiveness. The status quo: Engagement with TEI work happens via the TEI and Stylesheets repos on GitHub. Each of these has issue trackers, and there are two teams that have access to make changes to TEI, the Technical Council and "TEI Contributors". Incidentally, if the new folks will send me their github ids, I’ll add you to the Council team. The Tech Council team is for current Council members and the Contributors team is for a) former Council members who wish to remain active and b) anyone else we decide should have the ability to push directly to the TEI repo. This being GitHub, we can also take pull requests from people on neither team. Discussion of TEI stuff happens on this mailing list, on the issue trackers, in our monthly telecons, and in our Face to Face meeting. Only current Council members are on the mailing list, and people are removed as they cycle off Council. Only Council members attend the telecons and the F2F meetings, though occasionally "outside experts" are invited. Council tends to operate on consensus, though from time to time contentious issues are put to a vote. My ideal world: Contributors would be able to be as involved as they want to be in ongoing TEI development. That might include staying on or joining the mailing list, attending telecons, and even F2F meetings (though the TEI wouldn’t be able to pay for them unless they were formally invited of course). The main difference between Council Members and Contributors would be that the former have a vote when things come to a vote and would in general have more responsibility for development and maintenance. I feel that we don’t do enough to foster an active community of people who are TEI experts, and I think part of that is Council being a bit mysterious and exclusionary. It’s hard to get onto Council, especially if you’re not already part of a community of TEI users (so that you have name recognition in the elections) and people who leave have tended to just drift away (often leaving unfinished business). As I mentioned above, we’ve had a version of this discussion before, and I think mine was a minority view then, but I wanted to raise it again now to get the reactions of our incoming members and the perspectives of our outgoing members. What should we do? As you might have guessed, I have strong opinions, so I’m going to shut up for a while in the hope that a productive discussion develops without me driving it :-).

10 17

PDF and Pure ODD...
by Hugh Cayless 03 Dec '15

03 Dec '15

Tried this out, and decided to wait until we’ve got a version that’s current with the HEAD of dev. Right now it won’t build because of expired deprecated stuff that’s been fixed in the dev branch. Hugh

1 0

Telco this morning, right?
by Martin Holmes 03 Dec '15

03 Dec '15

According to my calendar, we have a telco in about 45 minutes. Is that right? Cheers, Martin

7 11

Agenda and Hangout
by Hugh Cayless 03 Dec '15

03 Dec '15

Hi all, Sorry for the delay. Moving offices is a headache. Agenda: https://docs.google.com/document/d/1f27Q0VLFN26oDNKgJ8hIXg97MF_9LeYavb1Wd9h… Hangout: https://talkgadget.google.com/hangouts/_/x3ri7iimwlh5ogr3t265zhe6n4a See you in 15 minutes!

1 0

LearnGit at Codecademy
by Raffaele Viglianti 02 Dec '15

02 Dec '15

Codecademy has just released a free interactive course for learning Git. Some council members may find it useful! It cleared some things up for me as well. (Also interesting how it has you write literature rather than code) https://www.codecademy.com/learn/learn-git Best, Raff

1 0

Telecon and updates
by Hugh Cayless 30 Nov '15

30 Nov '15

Dear All, I’ve made some progress on the Pure ODD XSLT. Please browse around http://teic.github.io/TEI/index.html <http://teic.github.io/TEI/index.html> and be on the lookout for display bugs or things you’d like done differently. Although we had enough for a quorum in the poll for meeting times, we’d either be excluding new members, which I’d rather not do, or current members who really need to be there, so I’ve added some more dates at http://doodle.com/poll/wyuc3qiv9rkx4db6. Please update your availability when you get the chance. Whether we have a year-end release or not, I think we’re getting close to the time when we should go ahead and merge the Pure ODD branch. Thoughts? All the best, Hugh

6 7

Bugfix for Roma available for testing
by Martin Holmes 29 Nov '15

29 Nov '15

Hi all, A new Roma is here: <http://www.tei-c.org/Roma-4.18/> with my minor bugfix for the title element included, along with other things (I know not exactly what) that had been added since the last release found its way onto tei-c in 2013. If you have any experience with Roma at all, please give it a bit of a go to make sure nothing is broken. Cheers, Martin

3 4

Adding processing model to github
by James Cummings 26 Nov '15

26 Nov '15

Hiya, To the 'processing-model' branch, based on the 'lb42-pureodd' branch, I've added the specs of elements needed for the processing model, modified that for elementSpec, and added the draft prose to the TD chapter. Hurrah. (Feel free to edit and improve the prose.) I don't promise I've not broken anything and haven't yet added the cross-references to the guidelines text in the TD chapter. (But that is straightforward and I'll do so soon.) What I *haven't* added so far are the tests that Sebastian had written for the processing model because a) believe it or not I've never actually added tests to the TEI build system and so now ask for help in doing so and b) I'm not sure the tests Sebastian was doing are specifically the tests we want. In this case what we want are tests that test that our schemas are allowing/disallowing the various model* and related elements in the right place and used only in ways that they are allowed. Sebastian's tests actually test the parsing of these processing models and the output they generate. Is that what we want? Remember, what we're doing here is adding the ability for documenting processing models, not adding TEI Simple, or any particular implementation of doing things with that processing model documentation. For example, in the TEI-Simple repository there are two types of test in the build.xml ant file: The 'tests' target: https://github.com/TEIC/TEI-Simple/blob/master/build.xml#L197 validates the TEI Simple ODD against the processing-model schema, and validates various instances of TEI Simple against the TEI Simple schema. That doesn't actually test the processing model elements are used properly. The 'pm' target: https://github.com/TEIC/TEI-Simple/blob/master/build.xml#L259 first creates some TEI simple XSL from the processing model documentation in the TEI Simple ODD, then creates a variant of this from teipm-ms.odd (which was a test variant requested by Council demonstrating using it for Manuscripts), and then runs various test files through both those XSL in order to produce HTML. Again, this seems to be testing the processing of the documentation of processing models, rather than the elements. (e.g. we don't seem to test 'list' or 'person' in this way?). Could someone with more experience with the TEI test system either help me design some new tests or help me migrate some of these into the processing-model branch? Many thanks, -James -- Dr James Cummings, James.Cummings(a)it.ox.ac.uk Academic IT Services, University of Oxford

1 0