James McCoy wrote on Wed, 30 Oct 2019 02:03 +00:00:
> On Tue, Oct 29, 2019 at 11:21:04PM -, danie...@apache.org wrote:
> > +abort
>--^
> Left over debug code?
Yes, thanks. Fixed in r1869139.
On Tue, Oct 29, 2019 at 11:21:04PM -, danie...@apache.org wrote:
> Author: danielsh
> Date: Tue Oct 29 23:21:04 2019
> New Revision: 1869134
>
> URL: http://svn.apache.org/viewvc?rev=1869134&view=rev
> Log:
> * tools/dist/release.py
> (write_news): Validate the argument to the --news-release
Julian Foad wrote on Tue, 29 Oct 2019 20:54 +00:00:
> Daniel Shahaf wrote:
> > release_date = time.strptime(args.news_release_date, "-mm-dd") if
> > args.news_release_date else datetime.date.today()
> > … { 'date': release_date.strftime("mmdd"),
> > 'date_pres': release_date.strftime(
Daniel Shahaf wrote:
You don't do any input validation on the date in argv anywhere, so
Right. I'm not wanting to spend much time on niceties as I want to
change a lot of this to read data from somewhere else, but then it still
might want validating at this level.
[...]
release_date = tim
Nathan Hartman wrote:
+++ subversion/site/staging/doap.rdf Tue Oct 29 17:38:29 2019
+ Current 1.13 regular release
+ 2019-11-30
Should be 2019-10-30?
Good catch. Thanks. r1869128.
A good example of why I want to automate these updates.
- Julian
julianf...@apache.org wrote on Tue, 29 Oct 2019 17:22 +00:00:
> +++ subversion/trunk/tools/dist/release.py Tue Oct 29 17:22:40 2019
> @@ -1188,8 +1188,9 @@ def move_to_dist(args):
> def write_news(args):
> 'Write text for the Subversion website.'
> -data = { 'date' : datetime.date.today()
On Tue, Oct 29, 2019 at 1:38 PM wrote:
> +++ subversion/site/staging/doap.rdf Tue Oct 29 17:38:29 2019
> +Current 1.13 regular release
> +2019-11-30
> +1.13.0
>
Should be 2019-10-30?
On Tue, Oct 29, 2019 at 11:36 AM Daniel Shahaf
wrote:
> Nathan Hartman wrote on Tue, Oct 29, 2019 at 11:11:58 -0400:
> > [[[
> > def finish(self):
> ⋮
> > except smtplib.SMTPRecipientsRefused as detail:
> > sys.stderr.write("mailer.py: SMTP recipient(s) refused: %s: %s\n"
> >
Nathan Hartman wrote on Tue, Oct 29, 2019 at 11:11:58 -0400:
> [[[
> def finish(self):
⋮
> except smtplib.SMTPRecipientsRefused as detail:
> sys.stderr.write("mailer.py: SMTP recipient(s) refused: %s: %s\n"
>% (self.to_addrs, detail,))
> raise MessageSe
Thanks to everyone for reviewing and providing feedback!
I've incorporated it and I hope that the following is good enough to
be committed as a fix for SVN-1804...
Since last time, only SMTPOutput.finish changes. I think it will be
easier to just see the function in full rather than a diff...
Th
On 2019/10/29 19:37, Branko Čibej wrote:
If I understand your patch correctly, those symbolic links would be
created in the build directory, not the source directory? In any case,
if you commit that patch, you'll also have to fix the "clean-swig-py"
target to remove the symlinks.
Yes, the symbo
On Tue, Oct 29, 2019 at 11:40:35AM +0100, Stefan Sperling wrote:
> On Tue, Oct 29, 2019 at 11:37:17AM +0100, Branko Čibej wrote:
> > Our .tar.bz2 and .tar.gz packages contain source files generated by
> > Swig. A packager /can/ decide to regenerate them; I don't know why they
> > would want to, but
On 29.10.2019 04:30, Yasuhito FUTATSUKI wrote:
> On 2019-10-28 12:14, Nathan Hartman wrote:
>> Finally got back around to this...
>>
>> With these changes, I verified that:
>> * sending with SMTP works and mailer.py exits with 0 on success
>> * using a deliberately wrong configuration to trigger a
On Tue, Oct 29, 2019 at 11:37:17AM +0100, Branko Čibej wrote:
> Our .tar.bz2 and .tar.gz packages contain source files generated by
> Swig. A packager /can/ decide to regenerate them; I don't know why they
> would want to, but they can.
I had to do just that for quite some time in the OpenBSD pack
On 29.10.2019 01:47, Yasuhito FUTATSUKI wrote:
> On 2019/10/29 2:45, Branko Čibej wrote:
>> Because if that's the case, I would
>> prefer
>> making 3.0.10 our required minimum version. Remember that users who
>> build
>> from our tarballs do not need Swig on Unix, i
15 matches
Mail list logo
