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]
On 05/12/15 11:49, Lou Burnard wrote:
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]
To clarify, this is only happening where the datatype of an attribute is being supplied, and looks like another bug. The (clickable) name of the dataSpec concerned is also no longer visible, though it is present in the github.io version. I attach 3 screenshots, showing a bit of the spec for specDesc/@key as displayed by the current web GL, the github.io version, and my latest build to show what I mean
Glad to say that nearly all of my yesterday winges are no longer an issue. After a morning wasted barking resolutely up the wrong tree, I checked out a fresh copy of the lb42-pureodd-2 branch and spent the rest of the weekend bringing it up to speed. If you check out commit 5e9864a you should now have something which will run the Makefile successfully to completion. If you have also installed the pureodd branch of the Stylesheets, you should see sensible HTML web pages. Even more amazingly, if you run "make pdf" you will get a not at all bad version of the Guidelines in PDF. https://github.com/TEIC/TEI/commit/5e9864a6b15b6541f81dba1367c570a31d24aea0 But I still dont like the class information being repeated inside the Pure ODD section, and I still don't think element names should be clickable inside examples. But I now declare this branch ready for merging -- as soon as at least one other person has managed to run the makefile to completion! On 05/12/15 11:49, Lou Burnard wrote:
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]
Just in case it's useful, I've created a new Jinks job to build the lb42-pureodd-2 branch. However, the main build has been broken for a few days by either Lou or Syd: http://teijenkins.hcmc.uvic.ca/job/TEIP5-Test/lastFailedBuild/changes So that needs to be fixed before we can expect the rest to build OK. Cheers, Martin On 15-12-06 01:35 PM, Lou Burnard wrote:
Glad to say that nearly all of my yesterday winges are no longer an issue. After a morning wasted barking resolutely up the wrong tree, I checked out a fresh copy of the lb42-pureodd-2 branch and spent the rest of the weekend bringing it up to speed.
If you check out commit 5e9864a you should now have something which will run the Makefile successfully to completion. If you have also installed the pureodd branch of the Stylesheets, you should see sensible HTML web pages. Even more amazingly, if you run "make pdf" you will get a not at all bad version of the Guidelines in PDF. https://github.com/TEIC/TEI/commit/5e9864a6b15b6541f81dba1367c570a31d24aea0
But I still dont like the class information being repeated inside the Pure ODD section, and I still don't think element names should be clickable inside examples.
But I now declare this branch ready for merging -- as soon as at least one other person has managed to run the makefile to completion!
On 05/12/15 11:49, Lou Burnard wrote:
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]
On 06/12/15 23:16, Martin Holmes wrote:
Just in case it's useful, I've created a new Jinks job to build the lb42-pureodd-2 branch.
Not sure that I understand the implications of that, but hoorah anyway.
However, the main build has been broken for a few days by either Lou or Syd:
http://teijenkins.hcmc.uvic.ca/job/TEIP5-Test/lastFailedBuild/changes
So that needs to be fixed before we can expect the rest to build OK.
Yes. The detest expected output has changed as a result of Syd's new schematron rule, so that needs to be updated. I have fixed this in the lb42-pureodd-2 branch.
Hi Lou, On 15-12-07 03:43 AM, Lou Burnard wrote:
On 06/12/15 23:16, Martin Holmes wrote:
Just in case it's useful, I've created a new Jinks job to build the lb42-pureodd-2 branch.
Not sure that I understand the implications of that, but hoorah anyway.
The idea is that this should predict how the main branch would behave if you merged the PureOdd branch into it. However, if the two branches get out of sync in other ways, then that won't hold so well. Perhaps you should apply your fix for Syd's Schematron problem to the main branch too. Cheers, Martin
However, the main build has been broken for a few days by either Lou or Syd:
http://teijenkins.hcmc.uvic.ca/job/TEIP5-Test/lastFailedBuild/changes
So that needs to be fixed before we can expect the rest to build OK.
Yes. The detest expected output has changed as a result of Syd's new schematron rule, so that needs to be updated. I have fixed this in the lb42-pureodd-2 branch.
Have fixed the detest.log in expected results in the dev branch so that the test should not fail. Mr Jenkins doesnt seem to tell me when failures occur any more, so I can't tell whether this worked or not though. On 07/12/15 13:41, Martin Holmes wrote:
... However, if the two branches get out of sync in other ways, then that won't hold so well. Perhaps you should apply your fix for Syd's Schematron problem to the main branch too.
Cheers, Martin
However, the main build has been broken for a few days by either Lou or Syd:
http://teijenkins.hcmc.uvic.ca/job/TEIP5-Test/lastFailedBuild/changes
So that needs to be fixed before we can expect the rest to build OK. Yes. The detest expected output has changed as a result of Syd's new schematron rule, so that needs to be updated. I have fixed this in the lb42-pureodd-2 branch.
Hi Lou, Mr Jinks should write to the Council list "when a build fails, becomes unstable or returns to stable". It's building P5 now, so we'll see if it does. I've configured mine to build the PureODD branch after P5 if P5 succeeds. Cheers, Martin On 15-12-08 05:03 AM, Lou Burnard wrote:
Have fixed the detest.log in expected results in the dev branch so that the test should not fail.
Mr Jenkins doesnt seem to tell me when failures occur any more, so I can't tell whether this worked or not though.
On 07/12/15 13:41, Martin Holmes wrote:
... However, if the two branches get out of sync in other ways, then that won't hold so well. Perhaps you should apply your fix for Syd's Schematron problem to the main branch too.
Cheers, Martin
However, the main build has been broken for a few days by either Lou or Syd:
http://teijenkins.hcmc.uvic.ca/job/TEIP5-Test/lastFailedBuild/changes
So that needs to be fixed before we can expect the rest to build OK. Yes. The detest expected output has changed as a result of Syd's new schematron rule, so that needs to be updated. I have fixed this in the lb42-pureodd-2 branch.
Yes, that was my understanding, but I don't think he is. After I fixed the detest bug in the dev branch, I got a notification from his Oxonian incarnation, but the Victorian one remains silent sfaics. On 08/12/15 13:43, Martin Holmes wrote:
Hi Lou,
Mr Jinks should write to the Council list "when a build fails, becomes unstable or returns to stable". It's building P5 now, so we'll see if it does. I've configured mine to build the PureODD branch after P5 if P5 succeeds.
Cheers, Martin
On 15-12-08 05:03 AM, Lou Burnard wrote:
Have fixed the detest.log in expected results in the dev branch so that the test should not fail.
Mr Jenkins doesnt seem to tell me when failures occur any more, so I can't tell whether this worked or not though.
On 07/12/15 13:41, Martin Holmes wrote:
... However, if the two branches get out of sync in other ways, then that won't hold so well. Perhaps you should apply your fix for Syd's Schematron problem to the main branch too.
Cheers, Martin
However, the main build has been broken for a few days by either Lou or Syd:
http://teijenkins.hcmc.uvic.ca/job/TEIP5-Test/lastFailedBuild/changes
So that needs to be fixed before we can expect the rest to build OK. Yes. The detest expected output has changed as a result of Syd's new schematron rule, so that needs to be updated. I have fixed this in the lb42-pureodd-2 branch.
By the way, and FWIW, I asked Peter Boot for a copy of his ODD, and checked that it generated a valid RelaxNG schema using the lb42-pure-odd branch. Seems to have worked, once I added his namespace to the list of those excluded from macro.anyXML. On 08/12/15 14:03, Lou Burnard wrote:
Yes, that was my understanding, but I don't think he is. After I fixed the detest bug in the dev branch, I got a notification from his Oxonian incarnation, but the Victorian one remains silent sfaics.
On 08/12/15 13:43, Martin Holmes wrote:
Hi Lou,
Mr Jinks should write to the Council list "when a build fails, becomes unstable or returns to stable". It's building P5 now, so we'll see if it does. I've configured mine to build the PureODD branch after P5 if P5 succeeds.
Cheers, Martin
On 15-12-08 05:03 AM, Lou Burnard wrote:
Have fixed the detest.log in expected results in the dev branch so that the test should not fail.
Mr Jenkins doesnt seem to tell me when failures occur any more, so I can't tell whether this worked or not though.
On 07/12/15 13:41, Martin Holmes wrote:
... However, if the two branches get out of sync in other ways, then that won't hold so well. Perhaps you should apply your fix for Syd's Schematron problem to the main branch too.
Cheers, Martin
However, the main build has been broken for a few days by either Lou or Syd:
http://teijenkins.hcmc.uvic.ca/job/TEIP5-Test/lastFailedBuild/changes
So that needs to be fixed before we can expect the rest to build OK. Yes. The detest expected output has changed as a result of Syd's new schematron rule, so that needs to be updated. I have fixed this in the lb42-pureodd-2 branch.
Hi Lou,
It's definitely sending the email to the Council list; TEI Council shows
as one of the recipients when I get it (and James should see it too, so
he can confirm).
The Oxford box is sending from:
Oxford TEI Jenkins Server
Yes, that was my understanding, but I don't think he is. After I fixed the detest bug in the dev branch, I got a notification from his Oxonian incarnation, but the Victorian one remains silent sfaics.
On 08/12/15 13:43, Martin Holmes wrote:
Hi Lou,
Mr Jinks should write to the Council list "when a build fails, becomes unstable or returns to stable". It's building P5 now, so we'll see if it does. I've configured mine to build the PureODD branch after P5 if P5 succeeds.
Cheers, Martin
On 15-12-08 05:03 AM, Lou Burnard wrote:
Have fixed the detest.log in expected results in the dev branch so that the test should not fail.
Mr Jenkins doesnt seem to tell me when failures occur any more, so I can't tell whether this worked or not though.
On 07/12/15 13:41, Martin Holmes wrote:
... However, if the two branches get out of sync in other ways, then that won't hold so well. Perhaps you should apply your fix for Syd's Schematron problem to the main branch too.
Cheers, Martin
However, the main build has been broken for a few days by either Lou or Syd:
http://teijenkins.hcmc.uvic.ca/job/TEIP5-Test/lastFailedBuild/changes
So that needs to be fixed before we can expect the rest to build OK. Yes. The detest expected output has changed as a result of Syd's new schematron rule, so that needs to be updated. I have fixed this in the lb42-pureodd-2 branch.
The tei-council list only allows posting by members and those specifically granted posting permission. I had previously set up ^.*@github.com as an allowed sender, but I've just added council@tei-c.org to the sender filter. So I hope the messages will start coming through now. Kevin On 12/8/15 10:28 AM, Martin Holmes wrote:
Hi Lou,
It's definitely sending the email to the Council list; TEI Council shows as one of the recipients when I get it (and James should see it too, so he can confirm).
The Oxford box is sending from:
Oxford TEI Jenkins Server
and mine is sending from:
Victoria TEI Jenkins
However, I don't actually see any of these messages from either server in the Council archives:
http://lists.tei-c.org/pipermail/tei-council/2015/date.html
so I suspect they're just being ignored; you may be seeing direct messages to yourself from the Oxford server, which is not for some reason happening on the Victoria server.
There was some discussion about whether we wanted these messages to go to the list or not, and I seem to remember the consensus was against it. How do we feel? Should we try to make this happen?
Cheers, Martin
On 15-12-08 06:03 AM, Lou Burnard wrote:
Yes, that was my understanding, but I don't think he is. After I fixed the detest bug in the dev branch, I got a notification from his Oxonian incarnation, but the Victorian one remains silent sfaics.
On 08/12/15 13:43, Martin Holmes wrote:
Hi Lou,
Mr Jinks should write to the Council list "when a build fails, becomes unstable or returns to stable". It's building P5 now, so we'll see if it does. I've configured mine to build the PureODD branch after P5 if P5 succeeds.
Cheers, Martin
On 15-12-08 05:03 AM, Lou Burnard wrote:
Have fixed the detest.log in expected results in the dev branch so that the test should not fail.
Mr Jenkins doesnt seem to tell me when failures occur any more, so I can't tell whether this worked or not though.
On 07/12/15 13:41, Martin Holmes wrote:
... However, if the two branches get out of sync in other ways, then that won't hold so well. Perhaps you should apply your fix for Syd's Schematron problem to the main branch too.
Cheers, Martin
> However, the main build has been broken for a few days by either Lou > or Syd: > > http://teijenkins.hcmc.uvic.ca/job/TEIP5-Test/lastFailedBuild/changes > > > > So that needs to be fixed before we can expect the rest to build OK. Yes. The detest expected output has changed as a result of Syd's new schematron rule, so that needs to be updated. I have fixed this in the lb42-pureodd-2 branch.
Thanks Kevin! People might get rapidly tired of it, of course. Cheers, Martin On 15-12-08 08:32 AM, Kevin Hawkins wrote:
The tei-council list only allows posting by members and those specifically granted posting permission. I had previously set up ^.*@github.com as an allowed sender, but I've just added council@tei-c.org to the sender filter. So I hope the messages will start coming through now.
Kevin
On 12/8/15 10:28 AM, Martin Holmes wrote:
Hi Lou,
It's definitely sending the email to the Council list; TEI Council shows as one of the recipients when I get it (and James should see it too, so he can confirm).
The Oxford box is sending from:
Oxford TEI Jenkins Server
and mine is sending from:
Victoria TEI Jenkins
However, I don't actually see any of these messages from either server in the Council archives:
http://lists.tei-c.org/pipermail/tei-council/2015/date.html
so I suspect they're just being ignored; you may be seeing direct messages to yourself from the Oxford server, which is not for some reason happening on the Victoria server.
There was some discussion about whether we wanted these messages to go to the list or not, and I seem to remember the consensus was against it. How do we feel? Should we try to make this happen?
Cheers, Martin
On 15-12-08 06:03 AM, Lou Burnard wrote:
Yes, that was my understanding, but I don't think he is. After I fixed the detest bug in the dev branch, I got a notification from his Oxonian incarnation, but the Victorian one remains silent sfaics.
On 08/12/15 13:43, Martin Holmes wrote:
Hi Lou,
Mr Jinks should write to the Council list "when a build fails, becomes unstable or returns to stable". It's building P5 now, so we'll see if it does. I've configured mine to build the PureODD branch after P5 if P5 succeeds.
Cheers, Martin
On 15-12-08 05:03 AM, Lou Burnard wrote:
Have fixed the detest.log in expected results in the dev branch so that the test should not fail.
Mr Jenkins doesnt seem to tell me when failures occur any more, so I can't tell whether this worked or not though.
On 07/12/15 13:41, Martin Holmes wrote:
... However, if the two branches get out of sync in other ways, then that won't hold so well. Perhaps you should apply your fix for Syd's Schematron problem to the main branch too.
Cheers, Martin
>> However, the main build has been broken for a few days by either >> Lou >> or Syd: >> >> http://teijenkins.hcmc.uvic.ca/job/TEIP5-Test/lastFailedBuild/changes >> >> >> >> >> So that needs to be fixed before we can expect the rest to build >> OK. > Yes. The detest expected output has changed as a result of Syd's new > schematron rule, so that needs to be updated. I have fixed this in > the > lb42-pureodd-2 branch. > >
They'll just get to see what it is like to be you or me. ;-) James -- Dr James Cummings, Academic IT, University of Oxford -----Original Message----- From: Martin Holmes [mholmes@uvic.ca] Received: Tuesday, 08 Dec 2015, 18:51 To: tei-council@lists.tei-c.org [tei-council@lists.tei-c.org] Subject: Re: [tei-council] Pure ODD progress Thanks Kevin! People might get rapidly tired of it, of course. Cheers, Martin On 15-12-08 08:32 AM, Kevin Hawkins wrote:
The tei-council list only allows posting by members and those specifically granted posting permission. I had previously set up ^.*@github.com as an allowed sender, but I've just added council@tei-c.org to the sender filter. So I hope the messages will start coming through now.
Kevin
On 12/8/15 10:28 AM, Martin Holmes wrote:
Hi Lou,
It's definitely sending the email to the Council list; TEI Council shows as one of the recipients when I get it (and James should see it too, so he can confirm).
The Oxford box is sending from:
Oxford TEI Jenkins Server
and mine is sending from:
Victoria TEI Jenkins
However, I don't actually see any of these messages from either server in the Council archives:
http://lists.tei-c.org/pipermail/tei-council/2015/date.html
so I suspect they're just being ignored; you may be seeing direct messages to yourself from the Oxford server, which is not for some reason happening on the Victoria server.
There was some discussion about whether we wanted these messages to go to the list or not, and I seem to remember the consensus was against it. How do we feel? Should we try to make this happen?
Cheers, Martin
On 15-12-08 06:03 AM, Lou Burnard wrote:
Yes, that was my understanding, but I don't think he is. After I fixed the detest bug in the dev branch, I got a notification from his Oxonian incarnation, but the Victorian one remains silent sfaics.
On 08/12/15 13:43, Martin Holmes wrote:
Hi Lou,
Mr Jinks should write to the Council list "when a build fails, becomes unstable or returns to stable". It's building P5 now, so we'll see if it does. I've configured mine to build the PureODD branch after P5 if P5 succeeds.
Cheers, Martin
On 15-12-08 05:03 AM, Lou Burnard wrote:
Have fixed the detest.log in expected results in the dev branch so that the test should not fail.
Mr Jenkins doesnt seem to tell me when failures occur any more, so I can't tell whether this worked or not though.
On 07/12/15 13:41, Martin Holmes wrote:
... However, if the two branches get out of sync in other ways, then that won't hold so well. Perhaps you should apply your fix for Syd's Schematron problem to the main branch too.
Cheers, Martin
>> However, the main build has been broken for a few days by either >> Lou >> or Syd: >> >> http://teijenkins.hcmc.uvic.ca/job/TEIP5-Test/lastFailedBuild/changes >> >> >> >> >> So that needs to be fixed before we can expect the rest to build >> OK. > Yes. The detest expected output has changed as a result of Syd's new > schematron rule, so that needs to be updated. I have fixed this in > the > lb42-pureodd-2 branch. > >
-- 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
participants (4)
-
James Cummings -
Kevin Hawkins -
Lou Burnard -
Martin Holmes