That’s it. Sorry you couldn’t make it.

On Sep 25, 2015, at 9:01 , Majewski Stefan wrote:

Hi all,

thanks for the extensive protocol to the meeting. Unfortunately, the meeting before has not been ended early enough for me to join. Regarding the f2f, I have put the following on my agenda

Sun 25th Council all day Mon 26th Council all day Tue 27th smaller groups of people available working on issues, I'm in. Wed 27th morning, Council work

Have I got this somewhat right?

cheers, Stefan

Am 21.09.2015 um 14:05 schrieb Majewski Stefan:

...

Hi,

thanks for the poll, depending on the details of public transport and the meeting before, I might be able to join via airport wi-fi. I'll try.

cheers, Stefan

Am 21.09.2015 um 13:11 schrieb Hugh Cayless:

...

There’s been no further participation in the Doodle Poll for our September telecon, so I think we should have it this Thursday, September 24th, at 07:00 PDT / 10:00 EDT / 15:00 BST / 16:00 CEST. This way, we have a chance to do last minute things before our scheduled freeze on October 1.

I’ve started a draft agenda at https://docs.google.com/document/d/1oz3MpPLyslDUudj33Tn_q9YH8j0hFei0sd8ChTcD... https://docs.google.com/document/d/1oz3MpPLyslDUudj33Tn_q9YH8j0hFei0sd8ChTcD...

All the best, Hugh

-- Mag. Stefan Majewski Projektmanager Abteilung Forschung und Entwicklung Österreichische Nationalbibliothek

Josefsplatz 1, 1015 Wien Tel.: (+43 1) 534 10-434 E-Mail: stefan.majewski@onb.ac.at Skype: stefan.majewski.onb.ac.at

-- tei-council mailing list tei-council@lists.tei-c.org http://lists.lists.tei-c.org/mailman/listinfo/tei-council

PLEASE NOTE: postings to this list are publicly archived

So did I. On Thursdays I am always in a different city at a different job (donning my hat as Community College Cataloguer). Not sure if I can connect from there (no camera, no headset, etc.) but will try. pfs On Mon, Sep 21, 2015, at 09:02, Syd Bauman wrote:

Ack! Sorry, missed the post about the poll entirely. I've put mine up. I'll have to miss a TAPAS meeting, but should be able to make Thu 09-24 10:00 unless the train is late (10% chance).

...

There’s been no further participation in the Doodle Poll for our September telecon, so I think we should have it this Thursday, September 24th, at 07:00 PDT / 10:00 EDT / 15:00 BST / 16:00 CEST. This way, we have a chance to do last minute things before our scheduled freeze on October 1.

I’ve started a draft agenda at https://docs.google.com/document/d/1oz3MpPLyslDUudj33Tn_q9YH8j0hFei0sd8ChTcD... -- tei-council mailing list tei-council@lists.tei-c.org http://lists.lists.tei-c.org/mailman/listinfo/tei-council

PLEASE NOTE: postings to this list are publicly archived -- Paul Schaffner Digital Library Production Service PFSchaffner@umich.edu | http://www.umich.edu/~pfs/

On 24/09/15 03:02, Syd Bauman wrote:

What am I supposed to do to work on something other than xenoData? Is the expectation that I'll generate a new branch to work on, say, the tweaks needed to discussion of language tags?

That's my understanding too. My belief is that in the Glorious New Age of Gittery we are expected to have multiple copies of the repo, one for each "branch" we want to work on. This is not quite as bonkers as it sounds, since each "branch" (if I understand correctly) is not a copy but a bunch of pointers. Or something.

Well, the 'rebase' method failed after some 28 CONFLICT error messages. (And it forced me to $ git checkout -- Exemplars/tei_lite_fr.nvdl first.) I then tried the the 'merge' method, and while it seems much more reasonable, I still have three conflicts to deal with. And while you're right, Hugh, the conflicts are marked in the file in the same way as they are with Subversion, it's still a bit more confusing. With Subversion I'm also used to having complete files with each revision and my local working copy. But more importantly, here in git-land, I'm not 100% sure which version represented in the file is which. I think "master" indicates what's in the head version in our main development branch ('cause that branch is called "master"), and "HEAD" indicates what's in the head version of whatever branch I've checked out, in this case sydb-xenodata. Right? If someone could verify that, it would give me a little more confidence to go on. Failing that, if someone could tell me whether the schema our xml-model PIs point to is supposed to be http://tei.it.ox.ac.uk/jenkins/... or http://tei.oucs.ox.ac.uk/jenkins/... and whether it should be .../artifact/release/xml/tei/odd/p5.nvdl or .../artifact/P5/release/xml/tei/odd/p5.nvdl that would do. (Yeah, I could probably figure that out myself by just trying those 4 URLs.) Oh, and while I'm at it, does P5/Makefile still have VCS := svn and if so, why not "git", now?

You have two choices if you want to work on something in another branch. You can, as Lou has, make a second copy of the repo and have it use the master branch. This makes decent sense for what Lou is doing I think, given his need to have master to hand as a reference while he works on Pure. Most people though, if they want to work on a different branch, just do $ git checkout master #or sydb-xenodata, or whatever you want in your repo, and then you have a working copy of whatever branch you want. Your other branch is still there, and perfectly safe. If you have uncommitted changes, then git will ask you nicely to please do something about them (probably commit them) before you switch branches. If you’re not sure which branch you’re on, $ git status will tell you. You can create a new branch anytime you want: $ git checkout -b mynewbranch If you want to get your branch up-to-date with master, there are two options. Your choice will depend on what you’ve been doing. If you have only a small number of changes in your branch, unlikely to conflict with anything, then you can run $ git rebase master #after you’ve checked out sydb-xenodata which will essentially pretend you branched off the current HEAD of master rather than wherever you originally did, and replay your commits on top of that. If you have a situation where you’ve changed files and master has changed the same files, you should probably do a merge: $ git merge master This may result in your having to do a merge commit, which git will set up for you. If there’s a conflict, you’ll need to resolve it and then commit your changes. Conflicts are marked just as they are in svn.

HEAD is your stuff, yes (or at least the state of the branch you were in when you ran the merge). Yeah, we should go ahead and change the default VCS in master.

On Sep 29, 2015, at 22:21 , Syd Bauman wrote:

Well, the 'rebase' method failed after some 28 CONFLICT error messages. (And it forced me to $ git checkout -- Exemplars/tei_lite_fr.nvdl first.)

I then tried the the 'merge' method, and while it seems much more reasonable, I still have three conflicts to deal with. And while you're right, Hugh, the conflicts are marked in the file in the same way as they are with Subversion, it's still a bit more confusing. With Subversion I'm also used to having complete files with each revision and my local working copy.

But more importantly, here in git-land, I'm not 100% sure which version represented in the file is which. I think "master" indicates what's in the head version in our main development branch ('cause that branch is called "master"), and "HEAD" indicates what's in the head version of whatever branch I've checked out, in this case sydb-xenodata. Right?

If someone could verify that, it would give me a little more confidence to go on. Failing that, if someone could tell me whether the schema our xml-model PIs point to is supposed to be

http://tei.it.ox.ac.uk/jenkins/... or http://tei.oucs.ox.ac.uk/jenkins/...

and whether it should be

.../artifact/release/xml/tei/odd/p5.nvdl or .../artifact/P5/release/xml/tei/odd/p5.nvdl

that would do.

(Yeah, I could probably figure that out myself by just trying those 4 URLs.)

Oh, and while I'm at it, does P5/Makefile still have

VCS := svn

and if so, why not "git", now?

...

You have two choices if you want to work on something in another branch. You can, as Lou has, make a second copy of the repo and have it use the master branch. This makes decent sense for what Lou is doing I think, given his need to have master to hand as a reference while he works on Pure. Most people though, if they want to work on a different branch, just do $ git checkout master #or sydb-xenodata, or whatever you want in your repo, and then you have a working copy of whatever branch you want. Your other branch is still there, and perfectly safe. If you have uncommitted changes, then git will ask you nicely to please do something about them (probably commit them) before you switch branches. If you’re not sure which branch you’re on, $ git status will tell you. You can create a new branch anytime you want: $ git checkout -b mynewbranch If you want to get your branch up-to-date with master, there are two options. Your choice will depend on what you’ve been doing. If you have only a small number of changes in your branch, unlikely to conflict with anything, then you can run $ git rebase master #after you’ve checked out sydb-xenodata which will essentially pretend you branched off the current HEAD of master rather than wherever you originally did, and replay your commits on top of that. If you have a situation where you’ve changed files and master has changed the same files, you should probably do a merge: $ git merge master This may result in your having to do a merge commit, which git will set up for you. If there’s a conflict, you’ll need to resolve it and then commit your changes. Conflicts are marked just as they are in svn. -- tei-council mailing list tei-council@lists.tei-c.org http://lists.lists.tei-c.org/mailman/listinfo/tei-council

PLEASE NOTE: postings to this list are publicly archived

I have no idea what the status of that file is. I wonder if it ended up tracked in git when it wasn't in svn? Cheers, Martin On 15-09-30 07:37 AM, Hugh Cayless wrote:

What is the deal with Exemplars/tei_lite_fr.nvdl? Make clean deletes it, but it’s in version control, and I have to mess with it every time I do something. Presumably it’s an exception to the rule that *.nvdl files are to be deleted? It doesn’t seem to be regenerated if you run make exemplars. If it’s important, I’ll modify the Makefile to ignore it, if it isn’t, then can we delete it?

...

On Sep 29, 2015, at 22:21 , Syd Bauman wrote:

(And it forced me to $ git checkout -- Exemplars/tei_lite_fr.nvdl first.)

I expect it did, but whenever you'd do an svn up it would silently put it back. Sent from my phone.

On Sep 30, 2015, at 11:43, Martin Holmes wrote:

Answering my own question: it _was_ tracked in svn, so this must be a relatively new problem. Perhaps make clean didn't delete it before?

Cheers, Martin

...

On 15-09-30 08:38 AM, Martin Holmes wrote: I have no idea what the status of that file is. I wonder if it ended up tracked in git when it wasn't in svn?

Cheers, Martin

...

On 15-09-30 07:37 AM, Hugh Cayless wrote: What is the deal with Exemplars/tei_lite_fr.nvdl? Make clean deletes it, but it’s in version control, and I have to mess with it every time I do something. Presumably it’s an exception to the rule that *.nvdl files are to be deleted? It doesn’t seem to be regenerated if you run make exemplars. If it’s important, I’ll modify the Makefile to ignore it, if it isn’t, then can we delete it?

...

On Sep 29, 2015, at 22:21 , Syd Bauman wrote:

(And it forced me to $ git checkout -- Exemplars/tei_lite_fr.nvdl first.) -- tei-council mailing list tei-council@lists.tei-c.org http://lists.lists.tei-c.org/mailman/listinfo/tei-council

PLEASE NOTE: postings to this list are publicly archived

I've temporarily made the Makefile route around it. Sent from my phone. On Sep 30, 2015, at 11:56, Syd Bauman wrote:

...

What is the deal with Exemplars/tei_lite_fr.nvdl?

The short answer is "I have no idea".

That said, if I had to guess, I would have guessed the reverse: that it is supposed to be deleted, but inadvertently found its way into version control.

The log isn't very illuminating.[1]

Notes ----- [1] All the commit messages I found that include this file. Note that I'm not at all sure I'm doing this right. | commit e7a66313d1e02cefac97a2f98fb8690e8f3cf339 | Author: Sebastian Rahtz | Date: Mon Mar 10 16:31:14 2014 +0000 | | extra NVDL file | | A P5/Exemplars/tei_lite_fr.nvdl | | | commit 0882c87b738a406a08a7261ef3d0be4ae0f9387e | Author: Sebastian Rahtz | Date: Sun Jan 12 15:33:55 2014 +0000 | | apply schematron constraints in p5odds.odd to Exemplar and Test .odd files | | M P5/Exemplars/tei_drama.odd | M P5/Exemplars/tei_enrich.odd | D P5/Exemplars/tei_lite_fr.nvdl | M P5/Exemplars/tei_lite_fr.odd | M P5/Exemplars/tei_tite.odd | M P5/Exemplars/tei_xinclude.odd | | | commit ee052e7d7a74d3b8d3bbae3405797356708f63f8 | Author: Sebastian Rahtz | Date: Tue Jul 9 08:43:04 2013 +0000 | | include graphics files with HTML of docs | | M P5/Exemplars/Makefile | A P5/Exemplars/tei_lite_fr.nvdl

...

Make clean deletes it, but it’s in version control, and I have to mess with it every time I do something. Presumably it’s an exception to the rule that *.nvdl files are to be deleted? It doesn’t seem to be regenerated if you run make exemplars. If it’s important, I’ll modify the Makefile to ignore it, if it isn’t, then can we delete it? -- tei-council mailing list tei-council@lists.tei-c.org http://lists.lists.tei-c.org/mailman/listinfo/tei-council

PLEASE NOTE: postings to this list are publicly archived