[svn.haxx.se] · SVN Dev · SVN Users · SVN Org · TSVN Dev · TSVN Users · Subclipse Dev · Subclipse Users · this month's index

Re: preparing for 1.14.1

From: Johan Corveleyn <jcorvel_at_gmail.com>
Date: Thu, 17 Dec 2020 11:20:01 +0100

On Sat, Oct 24, 2020 at 2:16 PM Yasuhito FUTATSUKI
<futatuki_at_yf.bsdclub.org> wrote:
>
> On 2020/10/24 1:06, Johan Corveleyn wrote:
> > On Thu, Oct 22, 2020 at 6:56 AM Nathan Hartman <hartman.nathan_at_gmail.com> wrote:
>
> >> - py38-windows-add-dll-directory--v2.diff posted by jun66j5 on 2020-05-11 in dev@ thread
> >> "[PATCH] fix unable to load *.pyd files with Python 3.8.x on Windows"
> >
> > (or a variation thereof ... I see that several patches were posted in
> > that thread, I'm not sure anymore which is which :-))
>
> There is at least 3 issues in this thread.

Sorry, for my late reply, and thanks Yasuhito for trying to keep these
things untangled, and keeping them on the radar!

> The issue that the subject originally addressed is resolved by the patch
> attached in the message, py38-windows-add-dll-directory--v2.diff.
>
> Date: Tue, 12 May 2020 02:08:37 +0900
> Message-ID: <db7f9979-18e6-b249-dd3f-7f56d2a22fad_at_gmail.com>
> https://svn.haxx.se/dev/archive-2020-05/0063.shtml
>
> +1 (again)

Okay, if I'm not mistaken this one was committed in r1883335, and has
been backported to 1.14.x now (in r1884483).

> Next, to make it possible to test swig-py with debug mode on Windows,
> Jun posted a patch py35-windows-debug-pyd.diff in the message
>
> Date: Fri, 15 May 2020 18:27:03 +0900
> Message-ID: <a41f7875-3ad9-269b-9e22-6229314d9af6_at_gmail.com>
> https://svn.haxx.se/dev/archive-2020-05/0086.shtml
>
> +1

Ah, indeed! This one has not been committed yet, I believe.

@Jun: could you please commit this py35-windows-debug.pyd.diff patch,
and nominate it for backport to 1.14.x too?

Now that I have my Windows dev environment more or less back
operational, I can easily reconfirm and vote+approve a backport
proposal of this change.

>
> And then, issue on handling of pools in swig-py, crash on test on
> debug mode, by negative ref count of pool object, reported on above
> message. Originally Jun posted the patch addressing it and then
> I posted a tweaked patch, global-py-pool-ref-count3-diff.txt, and
> waiting for reviews.
>
> Message-ID: <53ba0f87-9bea-1691-ffe0-479c9ec452b2_at_poem.co.jp>
> Date: Sat, 23 May 2020 21:15:56 +0900
> https://svn.haxx.se/dev/archive-2020-05/0120.shtml

Hm, this one is more difficult for me to comment on. I have not been
following that part of the discussion closely, and am not so well
versed in the deeper details of the swig bindings.

I think it would be best if this could be reviewed by Jun (or anyone
else who understands the swig python bindings sufficiently (more than
me at least)). Anyone wants to take a shot?

-- 
Johan
Received on 2020-12-17 11:20:18 CET

This is an archived mail posted to the Subversion Dev mailing list.