I agree, this is a wild reactionary shift to the issues they've seen with kbin. Unless the community "consensus" includes people actually reviewing and testing this is just going to put the repo admins in a tough situation when they have to merge in some broken commit the community voted on.
this post was submitted on 09 Nov 2023
9 points (100.0% liked)
Fediverse
19 readers
2 users here now
This magazine is dedicated to discussions on the federated social networking ecosystem, which includes decentralized and open-source social media platforms. Whether you are a user, developer, or simply interested in the concept of decentralized social media, this is the place for you. Here you can share your knowledge, ask questions, and engage in discussions on topics such as the benefits and challenges of decentralized social media, new and existing federated platforms, and more. From the latest developments and trends to ethical considerations and the future of federated social media, this category covers a wide range of topics related to the Fediverse.
founded 2 years ago
It looks like they're still working out what they want their process to be:
https://github.com/MbinOrg/mbin/pull/34
Seems like your concern is addressed there:
Pull Requests require at least one (1) other maintainer approval before the PR gets merged (built-in peer review process).
The mbin fork happened when kbin development was looking a lot less active. In any case, it's not necessarily bad to have a diversity of approaches. Due to their differing organizational structures, mbin will likely tend to have more features and more rapid development, but also potentially more bugs, while kbin remains more stable.
I cant follow the convo to tell if this is the actual state of things or just something thst was being discussed but:
16 Maintainers MAY merge incorrect patches from other Contributors with the goals of (a) ending fruitless discussions, (b) capturing toxic patches in the historical record, (c) engaging with the Contributor on improving their patch quality.
What an idea.
From the PR comments:
Maintainers MAY merge incorrect patches from other Contributors with the goals of (a) ending fruitless discussions, (b) capturing toxic patches in the historical record, (c) engaging with the Contributor on improving their patch quality.
I asked around and asked in the C4 specification matrix room.
And the reason is actually simple. If you merge bad code, have a record of proof in git (pull requests aren't forever it's only a github/gitlab thing).So the idea is if you merge bad code you have proof in the git record that there is a bad actor. You can always revert the commit again or fix it. And the record can act as a proof in case the community want to get rid of bad actors.
I think it’s an interesting experiment. If it survives natural human tendency to mess with things, then it could be a good process discovery that is further refined and implemented elsewhere.
I don’t have any reason to believe it’ll be a success, but that shouldn’t stop people from at least trying.
If it's any consolation, you likely won't have to worry about using it, as its liable to be unusable.
Sounds like it'll be a disaster
It seems to automatically pull all changes from kbin anyway so I don't know about this consensus approach.
Hmm, that seems like not such a good look from Ernest. According to google translate:
I know, honestly it was on purpose. I noticed that forks sync changes immediately with /kbin. I wanted to check how they deal with this much-announced community-based qualitative code review. Answer: they can't cope. Quite an obvious bug was accepted in PR and domerged into the main branch :P It now works properly on the rifle ;)
Hopefully everyone can play nice and work together productively.
seems like you are saying ernest put thru an intentionally malicious PR to see what would happen? And what happened was exactly what is described? I mean, ya, thats what people will do.
It wasn't entirely intentional, it was actually my mistake. But I held off on pushing the hotfix for a while. It was a development branch, so these kinds of bugs were permissible - in this case, it just changed the order of related posts, nothing serious. It was quite easy to spot and fix. Slow and cautious acceptance of pull requests, something I spent a lot of time on, was the main accusation from the creators of forks. Hastily accepting them was a problem for me. I personally considered a consensus similar to that, but now I see it doesn't make sense. Someone needs to take responsibility. Personally, I believe that forks are the best thing that could have happened to the project.
It is good to really see your true nature now. I'm also think the fork is the best thing that could have happened for the community. It's a pity that you never started a conversation, but instead you still try to do mean things like this.
Oh c'mon, don't be mad. It's just a wrong sorting of posts, it's in an edge case, and seriously it wasn't intentional. I just wanted to check how such management looks in practice, how many merge accepts are needed, etc. I didn't mean to do anything wrong that could cause harm. I even push the same code to my instance to facilitate your tests ;)
But you're right - that's just my nature. I approach PR with very limited trust, whether they're mine or from others.
I know your approach on PRs. Hence the main reason of the fork. The community does believe in their people and the good in mankind. Only 1 approval is required from another maintainer for now. We are using C4 way of working.
I assure you that I didn't intentionally push incorrect code into the repository. These were my first lines of code in a really long time. I simply got involved in other things that I wanted to finish first, and I noticed the edge case in the meantime, but it wasn't a priority. I saw that you were syncing and I was hoping to benefit a bit from it once you fixed it. I didn't expect the review to happen so quickly. By the way, I was genuinely curious about how this project management method works because, you know, I've always avoided such an approach. Merloy, you know how much I owe you, and I appreciate what you've done for the project, as well as the other Mbin contributors. Our overall visions haven't always been the same, and I think it's great that kbin has been forked. You see for yourself how my work looks until the release - there are many things I'll be refining over time. That's why I've put a hold on all other PRs, and now I want to focus on this.
“True nature” in this case appears to be slow and cautious. Shocking stuff!
Ow.. it was 100% intentional. You said it yourself: "Wiem, szczerze mówiąc było to celowe. Zauważyłem, że forki synchronizują od razu zmiany z /kbin.". https://karab.in/m/karabin/p/340377/Usterka-z-crosspostami-nie-zawsze-sa-przyporzadkowane-odpowiedniemu-watkowi-matce-at-ernest#post-comment-510980
@melroy I don't think you can really be upset about anyone putting through bad code. According to the philosophy as I understand it, bad code (intentionally so or otherwise) is a useful contribution and you are basically soliciting it. You supposedly have some way other than code review to ensure nothing harmful gets through and it has to do with the reputation of the contributor. Since you already knew @ernest and clearly have a bad opinion of him, how did it happen?
I did not and could not review the PRs themselves. So I am just going on the information as presented here. Sounds like @ernest put through some code (either into kbin or mbin not clear on that) which he knew was not 100% highest quality but which error was not critical or devastating. And that it could easily be found and fixed. Partially he did this to learn more about this governance model. A model which has apparently been developed in direct opposition to his own. Is it approximately accurate?
If so, sounds a bit mischievous at the worst.
I really can't recommend Tyranny of Structurelessness highly enough.
Well I don't have a bad opinion about him (those are your assumptions), we just didn't agree on how a community project would/can work.
If however he did introduce intentionally a bug in kbin, just because of Mbin that's downright childish. The Mbin community does try to test all the incoming PRs (not just kbin sync PRs) on various instances apart from unit-tests, etc. We just do not want to depend on a single maintainer, hence a different way of working in the project.
He saying Mbin can't handle the kbin changes that is just not true (Odpowiedź: nie radzą sobie), at least we try to keep in sync (eg. for API comparability for upcoming mobile clients). But I'll leave it this, I'm not going to waste any more energy. I hope you understand.
Thanks for your recommendation.
Yep, in Polish we call it skrót myślowy.
load more comments
(17 replies)
Thanks for your feedback.
We do have code reviews in GitHub and discussions on Matrix. We updated the README that reflect our latest way of working. As stated in the comment section we are also working on it in PR: https://github.com/MbinOrg/mbin/pull/34. Feel free to comment on that.
As a person who hangs around in repos but isn't a developer that sounds totally insane.
Why do you hang around there then? So you can write articles like this in an attempt to stir the shit? What is there to gain from that, for the fediverse?
I'd also like to know how you hang out there then, as I can't seem to find a person called Density hanging around? I might not be looking in the right place, of course