[PATCH v10 0/5][BZ 10871] Month names in alternative grammatical case

[PATCH v10 0/5][BZ 10871] Month names in alternative grammatical case

[Rafal Luzynski]

This is the 10th version of my patches, this time with a hope that
it may be the final version, hence no more "RFC" tag in the subject.
But, of course, the review is still needed.

Changes since the previous version (v9):

* The first patch of the series has been considered complete and
  loosely related with the rest therefore has been accepted and
  pushed to stable. [1]
* Commit messages corrected.
* ChangeLog changes removed (temporarily, actually moved to other
  patches which are not published here).
* Fixed bug in strptime(), thanks Zack Weinberg for finding and
  fixing it.
* Order of matching the month names in strptime() is consequent:
  basic month name, abbreviated, alternative, abbreviated alternative.
* Some tests for strptime() added.
* Everything rebased to the current (oh, sorry, yesterday's) master.

This series publishes the patches 1, 3, and 5. These patches are skipped:

0002 Regenerate locfile-kw.h + ChangeLog to the previous commit (bug 10871).
0004 Again regenerate locfile-kw.h + ChangeLog to the previous commit (bug
10871).

They contain the changes to the automatically generated files and
should be eventually merged into the patches 1 and 3, respectively,
before pushing to the git repo but should not be reviewed here.
For those who want to examine, they are available in bugzilla [2]
and my github repo. [3] Also note that in order to actually see
any difference you also need the locale data which are not here
but also in bugzilla and github.  Unfortunately, my COPR repo [4]
is not updated this time, this means that the packages in the main
Fedora repositories are more recent and will overwrite my packages.

Again, note that the backward compatibility patches which may be
still around most probably will be rejected.

Regards,

Rafal


[1] https://sourceware.org/git/?p=glibc.git;a=commitdiff;h=f301e53
[2] https://sourceware.org/bugzilla/show_bug.cgi?id=10871
[3] https://github.com/rluzynski/glibc
[4] https://copr.fedorainfracloud.org/coprs/rluzynski/genitive/

[PING][PATCH v10 0/5][BZ 10871] Month names in alternative grammatical case

[Rafal Luzynski]

16.11.2017 12:43 Rafal Luzynski <digitalfreak@lingonborough.com> wrote:
>
>
> This is the 10th version of my patches, this time with a hope that
> it may be the final version, hence no more "RFC" tag in the subject.
> But, of course, the review is still needed.
>
> Changes since the previous version (v9):
>
> * The first patch of the series has been considered complete and
> loosely related with the rest therefore has been accepted and
> pushed to stable. [1]
> * Commit messages corrected.
> * ChangeLog changes removed (temporarily, actually moved to other
> patches which are not published here).
> * Fixed bug in strptime(), thanks Zack Weinberg for finding and
> fixing it.
> * Order of matching the month names in strptime() is consequent:
> basic month name, abbreviated, alternative, abbreviated alternative.
> * Some tests for strptime() added.
> * Everything rebased to the current (oh, sorry, yesterday's) master.
>
> This series publishes the patches 1, 3, and 5. These patches are skipped:
>
> 0002 Regenerate locfile-kw.h + ChangeLog to the previous commit (bug 10871).
> 0004 Again regenerate locfile-kw.h + ChangeLog to the previous commit (bug
> 10871).
>
> They contain the changes to the automatically generated files and
> should be eventually merged into the patches 1 and 3, respectively,
> before pushing to the git repo but should not be reviewed here.
> For those who want to examine, they are available in bugzilla [2]
> and my github repo. [3] Also note that in order to actually see
> any difference you also need the locale data which are not here
> but also in bugzilla and github. Unfortunately, my COPR repo [4]
> is not updated this time, this means that the packages in the main
> Fedora repositories are more recent and will overwrite my packages.
>
> Again, note that the backward compatibility patches which may be
> still around most probably will be rejected.
>
> Regards,
>
> Rafal
>
>
> [1] https://sourceware.org/git/?p=glibc.git;a=commitdiff;h=f301e53
> [2] https://sourceware.org/bugzilla/show_bug.cgi?id=10871
> [3] https://github.com/rluzynski/glibc
> [4] https://copr.fedorainfracloud.org/coprs/rluzynski/genitive/

PING - any feedback about it?  Currently my biggest concern is the
documentation, changelog and commit messages.  I hope that the
code itself is correct (but a review is still welcome).

Also I'd like to notify that we have discussed this with the Polish
translator and the Polish locale [5] [6] will be updated with
alternative month names as soon as these patches are pushed.
For other locales at least I'd like to try to contact translators.

Regards,

Rafal


[5] https://sourceware.org/bugzilla/show_bug.cgi?id=10871#c119
[6] https://github.com/rluzynski/glibc/blob/master/localedata/locales/pl_PL

[PING^2][PATCH v10 0/5][BZ 10871] Month names in alternative grammatical case

[Rafal Luzynski]

30.11.2017 12:37 Rafal Luzynski <digitalfreak@lingonborough.com> wrote:
>
>
> 16.11.2017 12:43 Rafal Luzynski <digitalfreak@lingonborough.com> wrote:
> >
> >
> > This is the 10th version of my patches, this time with a hope that
> > it may be the final version, hence no more "RFC" tag in the subject.
> > But, of course, the review is still needed.
> >
> > Changes since the previous version (v9):
> >
> > * The first patch of the series has been considered complete and
> > loosely related with the rest therefore has been accepted and
> > pushed to stable. [1]
> > * Commit messages corrected.
> > * ChangeLog changes removed (temporarily, actually moved to other
> > patches which are not published here).
> > * Fixed bug in strptime(), thanks Zack Weinberg for finding and
> > fixing it.
> > * Order of matching the month names in strptime() is consequent:
> > basic month name, abbreviated, alternative, abbreviated alternative.
> > * Some tests for strptime() added.
> > * Everything rebased to the current (oh, sorry, yesterday's) master.
> >
> > This series publishes the patches 1, 3, and 5. These patches are skipped:
> >
> > 0002 Regenerate locfile-kw.h + ChangeLog to the previous commit (bug 10871).
> > 0004 Again regenerate locfile-kw.h + ChangeLog to the previous commit (bug
> > 10871).
> >
> > They contain the changes to the automatically generated files and
> > should be eventually merged into the patches 1 and 3, respectively,
> > before pushing to the git repo but should not be reviewed here.
> > For those who want to examine, they are available in bugzilla [2]
> > and my github repo. [3] Also note that in order to actually see
> > any difference you also need the locale data which are not here
> > but also in bugzilla and github. Unfortunately, my COPR repo [4]
> > is not updated this time, this means that the packages in the main
> > Fedora repositories are more recent and will overwrite my packages.
> >
> > Again, note that the backward compatibility patches which may be
> > still around most probably will be rejected.
> >
> > Regards,
> >
> > Rafal
> >
> >
> > [1] https://sourceware.org/git/?p=glibc.git;a=commitdiff;h=f301e53
> > [2] https://sourceware.org/bugzilla/show_bug.cgi?id=10871
> > [3] https://github.com/rluzynski/glibc
> > [4] https://copr.fedorainfracloud.org/coprs/rluzynski/genitive/
>
> PING - any feedback about it? Currently my biggest concern is the
> documentation, changelog and commit messages. I hope that the
> code itself is correct (but a review is still welcome).
>
> Also I'd like to notify that we have discussed this with the Polish
> translator and the Polish locale [5] [6] will be updated with
> alternative month names as soon as these patches are pushed.
> For other locales at least I'd like to try to contact translators.
>
> Regards,
>
> Rafal
>
>
> [5] https://sourceware.org/bugzilla/show_bug.cgi?id=10871#c119
> [6] https://github.com/rluzynski/glibc/blob/master/localedata/locales/pl_PL

PING^2 - actually I have nothing more to say about it. Shortly,
my biggest concern now is the documentation. Also I'm trying to
keep my github fork [3] up to date rebasing against the recent
changes in locale data.

Regards,

Rafal

[PING^3][PATCH v10 0/5][BZ 10871] Month names in alternative grammatical case

[Rafal Luzynski]

6.12.2017 23:32 Rafal Luzynski <digitalfreak@lingonborough.com> wrote:
>
> 30.11.2017 12:37 Rafal Luzynski <digitalfreak@lingonborough.com> wrote:
> >
> > 16.11.2017 12:43 Rafal Luzynski <digitalfreak@lingonborough.com> wrote:
> > >
> > > This is the 10th version of my patches, this time with a hope that
> > > it may be the final version, hence no more "RFC" tag in the subject.
> > > But, of course, the review is still needed.
> > >
> > > Changes since the previous version (v9):
> > >
> > > * The first patch of the series has been considered complete and
> > > loosely related with the rest therefore has been accepted and
> > > pushed to stable. [1]
> > > * Commit messages corrected.
> > > * ChangeLog changes removed (temporarily, actually moved to other
> > > patches which are not published here).
> > > * Fixed bug in strptime(), thanks Zack Weinberg for finding and
> > > fixing it.
> > > * Order of matching the month names in strptime() is consequent:
> > > basic month name, abbreviated, alternative, abbreviated alternative.
> > > * Some tests for strptime() added.
> > > * Everything rebased to the current (oh, sorry, yesterday's) master.
> > >
> > > This series publishes the patches 1, 3, and 5. These patches are skipped:
> > >
> > > 0002 Regenerate locfile-kw.h + ChangeLog to the previous commit (bug
> > > 10871).
> > > 0004 Again regenerate locfile-kw.h + ChangeLog to the previous commit (bug
> > > 10871).
> > >
> > > They contain the changes to the automatically generated files and
> > > should be eventually merged into the patches 1 and 3, respectively,
> > > before pushing to the git repo but should not be reviewed here.
> > > For those who want to examine, they are available in bugzilla [2]
> > > and my github repo. [3] Also note that in order to actually see
> > > any difference you also need the locale data which are not here
> > > but also in bugzilla and github. Unfortunately, my COPR repo [4]
> > > is not updated this time, this means that the packages in the main
> > > Fedora repositories are more recent and will overwrite my packages.
> > >
> > > Again, note that the backward compatibility patches which may be
> > > still around most probably will be rejected.
> > >
> > > Regards,
> > >
> > > Rafal
> > >
> > >
> > > [1] https://sourceware.org/git/?p=glibc.git;a=commitdiff;h=f301e53
> > > [2] https://sourceware.org/bugzilla/show_bug.cgi?id=10871
> > > [3] https://github.com/rluzynski/glibc
> > > [4] https://copr.fedorainfracloud.org/coprs/rluzynski/genitive/
> >
> > PING - any feedback about it? Currently my biggest concern is the
> > documentation, changelog and commit messages. I hope that the
> > code itself is correct (but a review is still welcome).
> >
> > Also I'd like to notify that we have discussed this with the Polish
> > translator and the Polish locale [5] [6] will be updated with
> > alternative month names as soon as these patches are pushed.
> > For other locales at least I'd like to try to contact translators.
> >
> > Regards,
> >
> > Rafal
> >
> >
> > [5] https://sourceware.org/bugzilla/show_bug.cgi?id=10871#c119
> > [6] https://github.com/rluzynski/glibc/blob/master/localedata/locales/pl_PL
>
> PING^2 - actually I have nothing more to say about it. Shortly,
> my biggest concern now is the documentation. Also I'm trying to
> keep my github fork [3] up to date rebasing against the recent
> changes in locale data.
>
> Regards,
>
> Rafal

PING^3. It's been a month since I've stopped working on this and
about 1.5 month since I have received any feedback. Can we please
finish this reasonably long before the end of this development cycle?

Regards,

Rafal

[PING^4][PATCH v10 0/5][BZ 10871] Month names in alternative grammatical case

[Rafal Luzynski]

12.12.2017 23:49 Rafal Luzynski <digitalfreak@lingonborough.com> wrote:
>
>
> 6.12.2017 23:32 Rafal Luzynski <digitalfreak@lingonborough.com> wrote:
> >
> > 30.11.2017 12:37 Rafal Luzynski <digitalfreak@lingonborough.com> wrote:
> > >
> > > 16.11.2017 12:43 Rafal Luzynski <digitalfreak@lingonborough.com> wrote:
> > > >
> > > > This is the 10th version of my patches, this time with a hope that
> > > > it may be the final version, hence no more "RFC" tag in the subject.
> > > > But, of course, the review is still needed.
> > > >
> > > > Changes since the previous version (v9):
> > > >
> > > > * The first patch of the series has been considered complete and
> > > > loosely related with the rest therefore has been accepted and
> > > > pushed to stable. [1]
> > > > * Commit messages corrected.
> > > > * ChangeLog changes removed (temporarily, actually moved to other
> > > > patches which are not published here).
> > > > * Fixed bug in strptime(), thanks Zack Weinberg for finding and
> > > > fixing it.
> > > > * Order of matching the month names in strptime() is consequent:
> > > > basic month name, abbreviated, alternative, abbreviated alternative.
> > > > * Some tests for strptime() added.
> > > > * Everything rebased to the current (oh, sorry, yesterday's) master.
> > > >
> > > > This series publishes the patches 1, 3, and 5. These patches are
> > > > skipped:
> > > >
> > > > 0002 Regenerate locfile-kw.h + ChangeLog to the previous commit (bug
> > > > 10871).
> > > > 0004 Again regenerate locfile-kw.h + ChangeLog to the previous commit
> > > > (bug
> > > > 10871).
> > > >
> > > > They contain the changes to the automatically generated files and
> > > > should be eventually merged into the patches 1 and 3, respectively,
> > > > before pushing to the git repo but should not be reviewed here.
> > > > For those who want to examine, they are available in bugzilla [2]
> > > > and my github repo. [3] Also note that in order to actually see
> > > > any difference you also need the locale data which are not here
> > > > but also in bugzilla and github. Unfortunately, my COPR repo [4]
> > > > is not updated this time, this means that the packages in the main
> > > > Fedora repositories are more recent and will overwrite my packages.
> > > >
> > > > Again, note that the backward compatibility patches which may be
> > > > still around most probably will be rejected.
> > > >
> > > > Regards,
> > > >
> > > > Rafal
> > > >
> > > >
> > > > [1] https://sourceware.org/git/?p=glibc.git;a=commitdiff;h=f301e53
> > > > [2] https://sourceware.org/bugzilla/show_bug.cgi?id=10871
> > > > [3] https://github.com/rluzynski/glibc
> > > > [4] https://copr.fedorainfracloud.org/coprs/rluzynski/genitive/
> > >
> > > PING - any feedback about it? Currently my biggest concern is the
> > > documentation, changelog and commit messages. I hope that the
> > > code itself is correct (but a review is still welcome).
> > >
> > > Also I'd like to notify that we have discussed this with the Polish
> > > translator and the Polish locale [5] [6] will be updated with
> > > alternative month names as soon as these patches are pushed.
> > > For other locales at least I'd like to try to contact translators.
> > >
> > > Regards,
> > >
> > > Rafal
> > >
> > >
> > > [5] https://sourceware.org/bugzilla/show_bug.cgi?id=10871#c119
> > > [6]
> > > https://github.com/rluzynski/glibc/blob/master/localedata/locales/pl_PL
> >
> > PING^2 - actually I have nothing more to say about it. Shortly,
> > my biggest concern now is the documentation. Also I'm trying to
> > keep my github fork [3] up to date rebasing against the recent
> > changes in locale data.
> >
> > Regards,
> >
> > Rafal
>
> PING^3. It's been a month since I've stopped working on this and
> about 1.5 month since I have received any feedback. Can we please
> finish this reasonably long before the end of this development cycle?
>
> Regards,
>
> Rafal

Sorry for the noise but I really wouldn't like to leave it until
the last day of this development cycle so...

PING^4

With these changes, [7] can I assume that the reviews so far are
sufficient and I should commit this week?

Regards,

Rafal


[7] https://sourceware.org/ml/libc-alpha/2017-12/msg00616.html

Re: [PING^4][PATCH v10 0/5][BZ 10871] Month names in alternative grammatical case

[Carlos O'Donell]

On 12/20/2017 04:30 PM, Rafal Luzynski wrote:
> PING^4
> 
> With these changes, [7] can I assume that the reviews so far are
> sufficient and I should commit this week?

Your patches are on the "desirable this release" list:
https://sourceware.org/glibc/wiki/Release/2.27#Desirable_this_release.3F

This means it should get one more review before commit for 2.27.

Who had reviewed these patches and provided their signed-off-by?

-- 
Cheers,
Carlos.

Re: [PING^4][PATCH v10 0/5][BZ 10871] Month names in alternative grammatical case

[Rafal Luzynski]

21.12.2017 02:48 Carlos O'Donell <carlos@redhat.com> wrote:
>
>
> On 12/20/2017 04:30 PM, Rafal Luzynski wrote:
> > PING^4
> >
> > With these changes, [7] can I assume that the reviews so far are
> > sufficient and I should commit this week?
>
> Your patches are on the "desirable this release" list:
> https://sourceware.org/glibc/wiki/Release/2.27#Desirable_this_release.3F
>
> This means it should get one more review before commit for 2.27.

I read this as a reassurance that the patches will be committed.
However, I'm getting worried because I know that Christmas and New
Year is a period when contribution rate drops.  Some potential
reviewers may be unavailable.  Also I may be unavailable for some
(but not all) days.  Just in case, if I become unresponsive when
it's time to commit somebody please commit this for me.  Also please
note that I'd also like to commit the locale data patch for pl_PL
applying the alternative month names.  For other languages I did not
yet talk to the translators.

> Who had reviewed these patches and provided their signed-off-by?

Zack Weinberg, [8] and I'm pretty sure you mean reviewed-by.
Although I'm not sure this tag still stands because the patches
have changed little since then.  Bugzilla [9] seems to be a good tool
to check what is the difference between the old and the new version
of the same patch.

Regards,

Rafal


[8] https://sourceware.org/ml/libc-alpha/2017-11/msg00320.html
[9] https://sourceware.org/bugzilla/show_bug.cgi?id=10871

[PING^5][PATCH v10 0/5][BZ 10871] Month names in alternative grammatical case

[Rafal Luzynski]

21.12.2017 02:48 Carlos O'Donell <carlos@redhat.com> wrote:
>
>
> On 12/20/2017 04:30 PM, Rafal Luzynski wrote:
> > PING^4
> >
> > With these changes, [7] can I assume that the reviews so far are
> > sufficient and I should commit this week?
>
> Your patches are on the "desirable this release" list:
> https://sourceware.org/glibc/wiki/Release/2.27#Desirable_this_release.3F
>
> This means it should get one more review before commit for 2.27.
> [...]

PING^5

Although my patches are on the "desirable this release" list it's
also likely they will be deferred to 2.28 if there are no more
reviews.

Sorry for my impatience but we are really short of time.

Regards,

Rafal

Re: [PING^5][PATCH v10 0/5][BZ 10871] Month names in alternative grammatical case

[Carlos O'Donell]

On 01/08/2018 02:08 AM, Rafal Luzynski wrote:
> 21.12.2017 02:48 Carlos O'Donell <carlos@redhat.com> wrote:
>>
>>
>> On 12/20/2017 04:30 PM, Rafal Luzynski wrote:
>>> PING^4
>>>
>>> With these changes, [7] can I assume that the reviews so far are
>>> sufficient and I should commit this week?
>>
>> Your patches are on the "desirable this release" list:
>> https://sourceware.org/glibc/wiki/Release/2.27#Desirable_this_release.3F
>>
>> This means it should get one more review before commit for 2.27.
>> [...]
> 
> PING^5
> 
> Although my patches are on the "desirable this release" list it's
> also likely they will be deferred to 2.28 if there are no more
> reviews.
> 
> Sorry for my impatience but we are really short of time.

Where are the final set of patches?

Could you please post the patch set to the list so I can review those?

-- 
Cheers,
Carlos.

Re: [PING^5][PATCH v10 0/5][BZ 10871] Month names in alternative grammatical case

[Rafal Luzynski]

8.01.2018 19:20 Carlos O'Donell <carlos@redhat.com> wrote:
> [...]
> Where are the final set of patches?

1. Most recent and up-to-date version in my github repo:
https://github.com/rluzynski/glibc

2. In Bugzilla as attachments, most probably nothing has changed
since I uploaded them:
https://sourceware.org/bugzilla/show_bug.cgi?id=10871

3. This list:
0/5: https://sourceware.org/ml/libc-alpha/2017-11/msg00569.html
1/5: https://sourceware.org/ml/libc-alpha/2017-11/msg00570.html
3/5: https://sourceware.org/ml/libc-alpha/2017-11/msg00571.html
5/5: https://sourceware.org/ml/libc-alpha/2017-11/msg00572.html
Patches 2/5 and 4/5 contain the changes in the automatically generated
files and in ChangeLog, they should not be posted.

4. Patchwork:
1/5: https://patchwork.sourceware.org/patch/24286/
3/5: https://patchwork.sourceware.org/patch/24287/
5/5: https://patchwork.sourceware.org/patch/24288/

Note that you should review only first 5 patches (github and Bugzilla
contain even 16.)  Those which provide the backward compatibility
most probably will be rejected.  Then there are locale data changes.
They are imported from CLDR.  There will be no change visible until
the locale data are changed.  BTW, as soon as the patches are accepted
and pushed to master I am going to push an update of the pl_PL locale.
This has been already consulted with the local community.  For other
locales I'd like to contact their local communities first although
most of them were interested in this change.  For small languages
(with low number of native speakers where it is unlikely to find
active contributors online) I will push the changes as well but not
immediately.

> Could you please post the patch set to the list so I can review those?

Sure, I will review and update the changes ASAP and repost them but
I think it's likely that there will be no more changes so you can review
them already.

Here is one place where I think I should update the ChangeLog and
commit message:

https://sourceware.org/ml/libc-alpha/2017-12/msg00616.html

Regards,

Rafal

Re: [PING^5][PATCH v10 0/5][BZ 10871] Month names in alternative grammatical case

[Carlos O'Donell]

On 01/08/2018 11:51 AM, Rafal Luzynski wrote:
>> Could you please post the patch set to the list so I can review those?
> 
> Sure, I will review and update the changes ASAP and repost them but
> I think it's likely that there will be no more changes so you can review
> them already.

My workflow is made easier if you are able to post patches to the mailing
list in a single thread, that way I can reply inline, and ACK the patches
one-by-one.

Please see how Adhemerval structured his C11 threads patch set post:
https://sourceware.org/ml/libc-alpha/2017-09/msg00871.html

If you can post it like that it helps me a lot. Thank you.

-- 
Cheers,
Carlos.