[pkg-gnupg-maint] Behavioral change for pinentry-qt 1.2.0-1

[Gard Spreemann]

On June 21, 2022 11:26:13 PM GMT+02:00, Daniel Kahn Gillmor <dkg at fifthhorseman.net> wrote:
>Thanks for looking into this, Gard!
>
>On Tue 2022-06-21 17:00:04 +0200, Gard Spreemann wrote:
>> I finally got around to this. Good news: The bisection showed me that
>> the problem is fixed upstream. I've filed #1013318 against pinentry-qt,
>> with an attached patch based on upstream's commits, and also an MR on
>> Salsa [1].
>>
>> [1] https://salsa.debian.org/debian/pinentry/-/merge_requests/1
>
>I took a look at the MR.  It looks not entirely unreasonable, but i'm
>afraid i'm not well-versed enough in QT to know exactly what all these
>changes do.
>
>For example, i see changes related to how the icons are identified that
>look like they come from upstream's
>7d5c123f802abce11c711d57e8796d58d6ff1a16 ("qt: Embed icons to avoid
>theme searches"), but they aren't accompanied by the .svg files
>themselves.  is changing the icon really related to the window sizing
>issue you've identified under sway?  won't changing these icon
>references without shipping the with pinentry cause cause a problem?
>
>Is there a way to provide a more targeted fix, at least until upstream
>releases 1.2.1?
>
>           --dkg

Hi,

Yeah, I must admit that I did a sloppy job: The last commit is the one that fixes the problem, based on bisection. Then I went back as many commits before it as was needed to get a patch that applied cleanly (and code that built and behaved).

I understand your aprehension. I'll ping upstream and ask if they can consider releasing a 1.2.1 soon. They might not be aware of the problem, as the fixing commit's intent seems to be different.

 -- Gard
-- 
Sent from my Android device with K-9 Mail. Please excuse my brevity.