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

Stylesheets release starter question
by Syd Bauman 28 Sep '18

28 Sep '18

Who am I supposed to ask for the TEI-C GPG key passphrase? Whoever it is, could you send it (off list) to me or Raff? Thanks! (Unless you already have it, Raff.)

1 0

when to meet? Doodle poll invite
by Syd Bauman 27 Sep '18

27 Sep '18

Per Martina, here is our Doodle poll for scheduling future TEI Council monthly teleconference calls. This poll is only about which day of week and what time, not about 1st, 2nd, 3rd, 4th, last, or whatever week of month. All times are UTC, and thus shift an hour for DST. 12:00Z, earliest time ------- -------- ---- 07:00 EST, 08:00 EDT 12:00 GMT, 13:00 BST 13:00 CET, 14:00 CEST 21:00Z, latest time ------- ------ ---- 16:00 EST, 17:00 EDT 21:00 GMT, 22:00 BST 22:00 CET, 23:00 CEST I can't speak for others, but here in USA [1] we will "fall back" from DST to standard time on Sun 04 Nov; except that Florida is trying to exempt itself from ever using standard time -- they want to use DST all year. So Sarah, you should keep abreast of that. A good list of time changes is available at https://www.timeanddate.com/time/dst/2018.html Notes ----- [1] Except for AZ and HI, which do not observe DST at all. But we don't have any Council members in either of those states. :-)

4 4

Stylesheets release freeze -- better late than never
by Syd Bauman 27 Sep '18

27 Sep '18

We have just created the release branch for Stylesheets (which we should have done last week). So feel free to push to 'dev' branch, but nothing should go to 'release-7.46.0' branch unless approved by me or Raff. Could Martin & Peter create appropriate jobs in Jenkins, please? Thanks.

2 1

Meeting in 20 minutes
by Scholger, Martina (martina.scholger＠uni-graz.at) 27 Sep '18

27 Sep '18

Hi all, we start in 20 minutes. https://docs.google.com/document/d/1uGtDk4AoOvHXTV6EzdSeyo-D5ItsAz7PBnKtHIg… --Martina

2 1

Re: [Tei-council] meeting?
by Scholger, Martina (martina.scholger＠uni-graz.at) 27 Sep '18

27 Sep '18

Hi Syd, I actually sent out the agenda and hangouts link on Tuesday. Hope everyone got it. Just in case: Our conference call is today, starting at 12:00Z-13:30Z (08:00-9:30 EDT, 13:00-14:30 BST, 14:00-15:30 CEST) This is the agenda: https://docs.google.com/document/d/1uGtDk4AoOvHXTV6EzdSeyo-D5ItsAz7PBnKtHIg… Hangouts: https://hangouts.google.com/hangouts/_/brown.edu/teicouncil Best wishes, Martina -----Ursprüngliche Nachricht----- Von: Tei-council <tei-council-bounces(a)lists.tei-c.org> Im Auftrag von Syd Bauman Gesendet: Donnerstag, 27. September 2018 00:08 An: TEI Council <tei-council(a)lists.tei-c.org> Betreff: [Tei-council] meeting? RELAX NG namespace in Schematron Welcome aboard, Vanessa! Martina -- are we in fact meeting tomorrow, and if so a) at what time?[1,2] b) can you post the URLs for both the Hangout and the agenda & minutes sooner rather than later? All -- I'm working on the "normal" ODD->Schematron pathway (odds/extract-isosch.xsl, i.e. the process that creates a standalone Schematron schema, not the one that embeds Schematron into RELAX NG) in response to Stylesheets #250. The code in that program always generates an <sch:ns prefix="rng"> element to declare the RELAX NG namespace in the output Schematron. I presume when it was put in it was needed because some Schematron rule or other used "rng:" in its @context or @test attribute. But the string "rng:" does not occur in *any* attribute in all of P5 now. (And the only elements in the RELAX NG namespace are inside <egXML> or in data.XXX, soon to be deprecated.) So I am planning to nuke the automatic generation of <sch:ns prefix="rng"> in this routine. (It will still get generated if there are definitions that specify it on their @ns, or an <sch:ns> element that defines it, or it is a namespace in force for one of the definitions.) If anyone knows of a reason it should stay, please speak up soon (or revert my change). Notes ----- [1] 12:00Z = 08:00 EDT, 13:00 BST, 14:00 CEST 13:00Z = 09:00 EDT, 14:00 BST, 15:00 CEST Sorry, I don't know your time zone, Vanessa. [2] IIRC we agreed to 13:00Z, but then again, I thought we agreed not to have a teleconference the last Thu of Oct, too. _______________________________________________ Tei-council mailing list Tei-council(a)lists.tei-c.org http://lists.lists.tei-c.org/mailman/listinfo/tei-council

1 0

meeting? RELAX NG namespace in Schematron
by Syd Bauman 27 Sep '18

27 Sep '18

Welcome aboard, Vanessa! Martina -- are we in fact meeting tomorrow, and if so a) at what time?[1,2] b) can you post the URLs for both the Hangout and the agenda & minutes sooner rather than later? All -- I'm working on the "normal" ODD->Schematron pathway (odds/extract-isosch.xsl, i.e. the process that creates a standalone Schematron schema, not the one that embeds Schematron into RELAX NG) in response to Stylesheets #250. The code in that program always generates an <sch:ns prefix="rng"> element to declare the RELAX NG namespace in the output Schematron. I presume when it was put in it was needed because some Schematron rule or other used "rng:" in its @context or @test attribute. But the string "rng:" does not occur in *any* attribute in all of P5 now. (And the only elements in the RELAX NG namespace are inside <egXML> or in data.XXX, soon to be deprecated.) So I am planning to nuke the automatic generation of <sch:ns prefix="rng"> in this routine. (It will still get generated if there are definitions that specify it on their @ns, or an <sch:ns> element that defines it, or it is a namespace in force for one of the definitions.) If anyone knows of a reason it should stay, please speak up soon (or revert my change). Notes ----- [1] 12:00Z = 08:00 EDT, 13:00 BST, 14:00 CEST 13:00Z = 09:00 EDT, 14:00 BST, 15:00 CEST Sorry, I don't know your time zone, Vanessa. [2] IIRC we agreed to 13:00Z, but then again, I thought we agreed not to have a teleconference the last Thu of Oct, too.

2 1

Next Meeting: September 27
by Scholger, Martina (martina.scholger＠uni-graz.at) 26 Sep '18

26 Sep '18

Hi all, our next conference call is on Thursday September 27 (08:00 EDT, 13:00 BST, 14:00 CEST). Here is the agenda - please add your topics: https://docs.google.com/document/d/1uGtDk4AoOvHXTV6EzdSeyo-D5ItsAz7PBnKtHIg… I'm happy that Vanessa will join us for the meeting. For the call we usually use this Google Hangouts connection: https://hangouts.google.com/hangouts/_/brown.edu/teicouncil. Vanessa, you should already have access to the minutes. Please let me know if not. Best wishes, Martina

3 5

Congrats and welcome
by Scholger, Martina (martina.scholger＠uni-graz.at) 24 Sep '18

24 Sep '18

Dear all, congratulations to everyone who got (re-)elected to the TEI Technical Council for the next two years! I'm delighted that Hugh, Raff, Sarah and Syd will be joining us for another run of office and I want to welcome our new member Vanessa Bigot-Juloux to the Council. Looking forward to a productive year! Best wishes, Martina Martina Scholger Zentrum für Informationsmodellierung - Austrian Centre for Digital Humanities Elisabethstraße 59/III 8010 Graz Austria martina.scholger(a)uni-graz.at<mailto:martina.scholger@uni-graz.at> https://informationsmodellierung.uni-graz.at<https://informationsmodellierung.uni-graz.at/> | https://gams.uni.graz.at<https://gams.uni.graz.at/> Chair of the TEI Technical Council | http://tei-c.org Institut für Dokumentologie und Editorik e.V. | https://i-d-e.de<https://i-d-e.de/>

1 0

tei website links
by Lou 18 Sep '18

18 Sep '18

Of the six links on the page http://www.tei-c.org/support/Learn/ all but two are currently broken. Is no link checking in place or is this a temporary issue?

2 1

Update on Docker and Travis
by Peter Stadler 13 Sep '18

13 Sep '18

Dear all, as I reported during our Council meeting on Saturday(?) I wanted to clean up and reorganize our Docker images a little bit. Here’s a little update on my work so far: a) every DockerHub repo is now turned into an „automated build“. This results in a better GitHub integration since every commit to a GitHub repo will trigger a fresh built of the respective Docker image at DockerHub. Sadly, there is no way of converting an existing repo so I had to nuke the Jenkins and the teidev-docker repos and recreate them. Hence, all stars and pull statistics got lost :( b) because some repos do not get frequent commits, I added a cron job at my server to regularly trigger a rebuild of those images to keep them updated with their upstream images. Now that I’m writing this, I think that would make a nice Jenkins job. Will try and report back! c) I was originally thinking that I’d like to alter the teidev-docker image to inherit from our Jenkins image (so we need to maintain only one Dockerfile with the software dependencies) – but when I was working on this it occurred to me, that we actually could be doing everything with the Jenkins image alone (by overwriting the entrypoint and cmd directives) and don’t need a dedicated image for building the Guidelines etc. locally. If you want to give it a try, install Docker, navigate to your Stylesheets directory and enter the command: `docker run --rm -w /var/odd -it -v `pwd`:/var/odd --entrypoint "make" teic/jenkins:dev test ` This will run the tests on your machine, with your local files, with the software environment from the Jenkins Docker image. That’s cool, ain’t it? d) I was taking first steps with Travis for RomaJS and it’s great fun, because (again) it integrates with GitHub so seamlessly. What we gain is automated testing for every branch and every pull request, nice badges, and even Slack notifications ;) So, I added this Travis integration to the Stylesheets repo as well, using the Docker image and the above Docker command for testing. Another really cool thing IMHO, because (thanks to Docker) the commands run are using the very same software environment as our Jenkins does. If you want to check, have a look at the .travis.yml file in the Stylesheets repo which is driving the integration tests. [1] BTW, everybody who is a member of the GitHub TEIC organisation should be able to login and see/alter all the stuff at https://travis-ci.org/TEIC <https://travis-ci.org/TEIC>. Of course, these new opportunities bring up a few new ideas. Here’s the proposal bit: 1) since we do not need a dedicated Docker image for building locally, I propose to delete the repo at DockerHub. At GitHub, I propose to archive the repo. Instead, I will add a section to the Jenkins Readme on how to build locally with the Docker image. Any other reference on how to build the Guidelines should be updated as well and have at least a pointer to the Jenkins Readme and the Jenkins Dockerfile. That would serve as our main „infrastructure as code“ documentation on how to set up the build environment. 2) I propose to add Travis integration to the TEI repo as well and set it up to (only) run the Tests (i.e. `make clean validate test`). Since the Guidelines depend on the Stylesheets to build, Travis would need to get them from Jenkins lastSuccesfullBuild (as Jenkins is currently doing) 3) Once we have the Travis integration, I propose to turn our dev branches into "protected“ branches. That means, on the pull request pages on GitHub (and only there) the test results will be added. An example for how this looks like can be seen at [2]. The advantage is for the issuer of the pull request to directly see whether it’s working or not. If the tests are failing the pull request is marked as „blocked“, but as an administrator (and probably all Council members have this role?) you are still able to merge. 4) A mid term goal could be to use Travis for running the tests and have Jenkins only build the artefacts. We could even replace Jenkins completely but then we’d need another place to push our build artefacts to. Sorry for the lengthy mail. Best Peter [1] https://github.com/TEIC/Stylesheets/blob/dev/.travis.yml [2] https://github.com/eXist-db/exist/pull/2153

2 2