Is there anything wrong with PR review process?

[Dreamescaper]

Numerous times, I've seen the following situation. I encounter some MAUI issue. I find that it is already logged. Moreover - it even has a community PR fixing it! Except it is created a year ago, and has no recent activity whatsoever.
(The last ones which comes to mind is #24645, but again, it happened more than once)

I haven't created any PRs myself yet. But having this impression it makes me hesitant to even try, as it makes me think that my PRs would be ignored anyway.

It would be nice to have some visibility around this aspect.
Is it because of low PRs quality?
Lack of manpower to review them?
Or is my impression wrong, and it's not as bad?

[developer9969]

Its a pity , I think they dont have the manpower, I am in the same boat there are things fixed in PR that have been there for ages.. or things are moved from one sprint to the next .

The real answer on why, only the Maui team knows... My guess they are shortstaffed! very very frustrating...

[HuaFangYun]

I hope their review process can be faster, as this would resolve a significant portion of the issues quickly, leading to much higher user satisfaction.

[MartyIX]

Numerous times, I've seen the following situation. I encounter some MAUI issue. I find that it is already logged. Moreover - it even has a community PR fixing it! Except it is created a year ago, and has no recent activity whatsoever.

It depends very much on the change. For example, if you fix a bug for just one single platform, then it's quite easy to review your change. If you change code shared by all supported platforms, you can fix a bug on one platform and introduce another on another platform (saying that for the sake of argument).

Time and time again, it happens that a seemingly innocent change breaks something.

However, in short, things got better recently. There is this inflight process where PRs get reviewed and tested in a batch to test the PRs by CI more and to discover issues sooner.

Another aspect is that there are priorities. You might fix a bug but the MAUI team feels like another PR affects bigger set of customers, so I guess they consider it from this angle. This is something hard to put up with but it's simply like that so demonstrating that an issue affect many people helps.

I haven't created any PRs myself yet. But having this impression it makes me hesitant to even try, as it makes me think that my PRs would be ignored anyway.

Please do.

It takes effort and if you want your PR to be merged soon, then it's a good idea to make it easy to review it:

• Add a good test or tests.
• Comment on why you make such changes and what alternative approaches you considered.
• Personally, I feel like it's good make requested changes fast rather than slow because otherwise it gets tiring to review a PR.

[PureWeen]

@MartyIX 's comment is perfect

One additional note here @Dreamescaper is that reviews do tend to slow a bit as we are finalizing a GA release with fun new features. We are at a point where we are looking to address the current build up (for example we are down 60 PRs total count today compared to two weeks ago)

If you have specific PRs that are of high interest to you, please react and/or comment on them. We'd like to get through a lot of the easier to review PRs and the ones that are the biggest customer delighters over the next few months. We're heavily prioritizing community PRs right now, especially new ones.

We're also active in the MAUIVerse which is a good place to poke in case something has unfortunately slipped through the cracks.
https://discord.gg/UbqgEms5

[BernhardPollerspoeck]

2 Things stick stick out here for me:

Time and time again, it happens that a seemingly innocent change breaks something.

shouldnt this be prevented by tests and a robust architecture? This makes me flinch if i hear this, bedause either MAUI has bekome so big, that no developer can grasp the impact of a change, or its so fragile that changes are stable by luck.

reviews do tend to slow a bit as we are finalizing a GA release with fun new features

By the amount of bugs that are existent, the focus on features is a bit unsatisfying. we have a ton of vendors that make controls and we have all the base in the box. a framework full of features is not usable if it has so many bugs.

not sure if i even have to expect a answer here...

for the last 3 years it has gotten better bug wise, but still its so many, that ppl start questioning the choice to use maui.

[MartyIX]

Time and time again, it happens that a seemingly innocent change breaks something.

shouldnt this be prevented by tests and a robust architecture?

Lately, it stabilized and I believe it's in that state (🤞). So if there is a good test, then things do not break randomly. However, what I had in mind is that suppose that you want to add label.Align = justify feature, each platform implement it a bit differently, some older devices cannot do it, or it works by default but not in a special case (when WrapperView is involved), etc. So even though the feature appears trivial, the work can be non-trivial to make it work always. There is inherent complexity in trying to unify different platforms under one umbrella and you cannot remove it, you can just deal with it by good tests and testing.

This makes me flinch if i hear this, bedause either MAUI has bekome so big, that no developer can grasp the impact of a change, or its so fragile that changes are stable by luck.

There are things like:

• Android is not a single API, it's a set of SDKs. Similarly for iOS and Mac Catalyst. But it's hard to test on many devices for a single person. So I believe that people rely on the MAUI CI to check it for them. But then the feedback is limited because you cannot run 1 PR 100 times on CI as it would take too much time & resources.
• Contributors typically do not have all platforms (Android, iOS, Mac Catalyst, Windows), so they test on their primary platform typically.
• etc. etc.

for the last 3 years it has gotten better bug wise, but still its so many, that ppl start questioning the choice to use maui.

Everyone has to evaluate it himself or herself.

Personally, I'm happy to see people contributing fixes for their issue as everyone then benefits. There is IMHO even big enough community that if someone needs something fixed, then they can pay somebody to do the work. ... Or one can use alternative technologies, which may be more suitable for that particular job. Many options IMHO.