Hi Jan!

On Tue, Feb 11, 2025 at 08:37:46PM +0100, Jan Christoph Uhde wrote:
> How about passing the fix to upstream? That way it would get reviewed by
> somebody else, as it is related to keepass I think it would be the correct
> choice. Not knowing the code, it would take me some time to verify the
> correctness of your code.

I usually do forward my fixes to upstream, but this time I did not do it for
two reasons:

1. In upstream dev branch this code no longer exists (replaced with different
   code in [1]), and upstream 5.15 branch is commercial and it is not possible
   to submit directly to it.

2. In Debian, we are regenerating qurltlds_p.h from our copy of publicsuffix
   (using this script [2]). The crash does not happen with unmodified Qt
   5.15.15 (or 5.15.16) code, so it can be considered Debian-specific.

   When/if upstream decides to update their qurltlds_p.h copy, they will have
   to do something similar. However Qt is getting end of life this year [3],
   so maybe this will never happen. Or maybe it already happened and then we
   will receive this fix after a year passes and it is published as LGPL.

If you have any suggestions what I can do, they are welcome.

[1]: https://code.qt.io/cgit/qt/qtbase.git/commit/?id=cba40055b1cb9f0b
[2]: 
https://salsa.debian.org/qt-kde-team/qt/qtbase/-/blob/master/debian/generateTLDs.sh
[3]: 
https://www.qt.io/blog/qt-5.15-extended-support-for-subscription-license-holders

--
Dmitry Shachnev

Attachment: signature.asc
Description: PGP signature

Reply via email to