#3285 Request for exception: lxqt-wallet 4.0.0 update
Closed: Accepted by zbyszek. Opened by farchord.

Hello, I've built lxqt-wallet v4.0 in a sidetag. I honestly didn't think that submitting an exception for this was needed, so I was simply planning on letting the packages of the consumers get built in the same side tag.

But I was told (https://bugzilla.redhat.com/show_bug.cgi?id=2321970) that an exception request needed to be submitted to fesco.

One of the package consumers is already rebuilt, would it be ok to be allowed to submit this?


Metadata Update from @ngompa:
- Issue tagged with: updates policy exception

Metadata Update from @ngompa:
- Issue set to the milestone: Fedora Linux 41

Btw the two reverse dependancies for this are zulucrypt and sirikali. The latter has already been rebuilt in the sidetags. I was waiting on the last one before pushing anything, I apologize I didn't know an exception was needed for this.

I feel like I'm missing some context here - the why side of why you think you need to push a breaking change to a stable branch isn't really answered here or in the linked bug.

It is part of the LXQt 2.0 stack, but was released very late relative to the rest of the stack.

That's an answer to a "what" question, but still not to "why" :)

To be 100% honest, I figured that if I can get the reverse dependancies built with the new version, there was no problem.

I was wrong... To be honest, once I saw Ian's answer in the tracker, I thought I was getting trolled, hence the slightly condescending answer afterwards...

I knew about ABI breakages, that's why I went very much out of my way to make sure the reverse dependancies were being rebuilt on it. But now, thinking in hindsight, "What if the packager(s) didn't respond?"

But all that is besides the point, once I saw the policy...

The main reason why I wanted this in ASAP was because this is a conversion from qt5 to qt6.

Right now, lxqt is a smorgenborg of qt5 and qt6 libraries because, as @ngompa stated, they are.... releasing stuff out of sequence which is very annoying.

So I was really trying to get this in.

Admittedly, looking at the sources, it doesn't seem like there was much in the way of breakage API-wise, only that it's now built with Qt 6. And its reverse dependencies can be built with Qt6 as well (and not crash) once this is updated to Qt 6.

Yeah when I started the whole thing, I figured there was a possibility that building part of the lxqt stack out of order might cause issue so I did a dnf repoquery --whatrequires on the package.

I actually wasn't expecting two unrelated packages. I didn't look at the source code deeply (I don't have the expertise for that) but did notice that they can now be built in both qt5 and qt6. That was the reason why I started on this route.

Oh and in case someone asks....

The version of the package is 4.0. Don't ask me why, but the lxqt stack's versions are a bit all over the place. For LXQt 2.0, some are 1.0, some are 2.0, some are 4.0. I don't think there's any 3.0 from memory.

LXQt Wallet was a 3.x version before.

I don't see a problem with this update. +1

Weak +1 too. Primarily because the work's already been done, and reverting the changes would be more work than pushing them through.

After a week:
APPROVED (+4, 0, 0)

Metadata Update from @zbyszek:
- Issue tagged with: pending announcement

Announced.

Metadata Update from @zbyszek:
- Issue close_status updated to: Accepted
- Issue status updated to: Closed (was: Open)

Metadata Update from @zbyszek:
- Issue untagged with: pending announcement

Metadata