https://github.com/dessalines/jerboa/issues/891
If you downvote any post in jerboa 0.0.35 you actually upvote it. Seems like anything gets merged? If this goes on like this, there could be a lot of trouble
Edit: Thanks to quick reaction seems like a simple an quick fix. I hope that things will be fine in the future
It’s rough, there are two main developers working on Jerboa, one other person reviewing code, 5 or 6 other people helping squash bugs on the regular. And a few dozen people that even bother to submit PRs every now and then.
The hundreds to thousands of people using the app that aren’t helping to develop it: y’all are the bug testers. It’s good to bring up issues like this one, but it’s the coders that have to go in and fix the issues, hope that they don’t break others in the process (a missing question mark broke all thumbnail displayings at one point mid-development). Lots of issues coming in and not enough hands to both fix them and check that it’s still all working fine. Even the unit tests are coded in by people.
In the next couple days I will try to release something that works well enough for my standards on my fork. But it takes some patience and a little less whining, there’s a reason why we’re in alpha.
I can totally understand the lack of people working on the project. As skeptic I am, I must say that you have total right to not believe me, but I’m an open source developer too with an own project, with literally no time at all since years. But what I’m trying to say: if the devs have no time to review PR-s, and features, please just don’t merge them at all. Please tell me that I’m wrong, but this seems like PR-s getting merged without proper reviews. This is far more worse than not merging any PR-s
I haven’t looked through the Jerboa codebase very closely yet (I downloaded it, but that’s about as far as I have got so far), but it sounds like they either need more code coverage for their unit tests or they need to improve the quality of the unit tests. Good unit tests should catch a lot of bugs introduced by PRs. There is no QA team to do regression testing for each release, and the main developers can’t be expected to run manual testing for every PR. The developer of the PR should be doing that, but clearly they are not.
Without a link to the PR you refer to this is a bit of a vague ticket. Did it work in a previous version? And if so, which version?
Also I think your tone is not constructive and you have to keep in mind that this is all open source software being quickly developed to deal with the huge influx of new users. It’s good to create github issues for these things, I also agree that it’s a silly bug if true but remember the human on the other side, they owe us nothing :-)
You are right, I should have tested it on an older version of jerboa. I hope that the developers or other users can prove me wrong
git blame shows this bug was introduced in v0.0.35 when they moved from websockets to http.
A developer of an open source application being attacked my an entitled user. A story as old as time, yet very sad to see over and over again.
Dear user, you are getting something for free. Open source even. If you don’t like how it works, fork it and develop your own, or do your part in helping out debug and investigate. Or just stop using it.
This is not a big corporation with dozens or hundreds of devs working on an app that you pay for with ads or premium subscriptions. This is a free app, developed by volunteers. Please be nice. You can complain, but be civil about it.
I’m sorry if this seems like an attack, the whole point I’m trying to make is that jerboa might not be as ready as we think for being a lemmy client. Open source projects are extremely hard, an mostly never rewarding, and I totally understand that. I hope that those who have the skill and time will make it better, but right now it can contain flaws that make basic features unusable, and who knows what actually runs in the background
I’m trying to make is that jerboa might not be as ready as we think for being a lemmy client.
I’m one of those random guys who try to improve things submitting a PR now and then fucking up things in the progress and trying to fix them again.
What I want to say is: I think you might be right saying Jerboa might not be as ready as many think. It started as a side project of one of the two guys who develop Lemmy full time. Dessalines even writes in Jerboa’s README that people can submit issues but he will most likely not be able to work on them.
Now Lemmy exploded: A lot of people started contributing to Jerboa (including me) with vastly varying skill levels (me definitely on the lower end), there’s no clear route the app will go in the future, no or little standards we can use as orientation.
And actually I’m not surprised at all that Jerboa might not be ready because this app went from “Just a little side project with 150 stars on GitHub on June 3” to “Probably most popular Lemmy client with 787 stars on June 27”. This has been less than four weeks. I don’t think any project is able to adapt to this influx of requests, changes, attention in this short time.
Who knows what runs in the background?
In an open source project? Really?
Could not have said it better!
I am not sure what the general expectation was from the mass influx of people coming from Reddit, I can say it saddens me to see the daily barrage of complaints. Lately it is not even constructive criticism, it’s just all out bitching about this or that…
Lemmy is ALPHA software, and suffering from a HUGE influx of new users. There are and will be growing pains.
Jerboa is ALPHA software, and suffering from a HUGE influx of new users. There are and will be growing pains.
Either get involved with development, move on to something else, or deal with it. Posting a daily rant (most of which have already been stated, and addressed or will be) is helping absolutely no one.
You should look at the code in Lemmy lol. I get that people think open source means more eyes, but it doesn’t mean higher quality.
Depends. Not under these conditions, with just a few devs stressing to both add features and fix bugs, as users comes crashing in.
It’s obvious that the code is not going to be high quality right now. But hopefully once the storm is over.
There is a reason companies don’t try and stress devs with tight deadlines unless they want poor quality. It takes time to do things right, and now the devs are probably just patching stuff together to get it out the door.
And seems it’s kind of a false accusation because the bug wasn’t even introduced by an PR. (If you follow the recent discussion on github.)
It clearly says alpha version, which means it will be full of bugs and unstable. Alpha has a specific meaning in software development.
I understand that people who are not devs don’t know that though… :)
From what I understand, it’s a bit of a mad run right now to try and both add features and fix bugs, and there are not many devs. It’s only a matter of time before something stops working or produces weird results. Just be on with it. It’s alpha.
Does this happen in the latest version, v0.0.36-alpha? I’d also like to point out the full version name includes the phrase “alpha”. This is not a fully released product and will contain bugs.
You make it sound like there was a pull request that deliberately changed the functionality of the voting system when it’s most likely just a bug. The main developer of the app also has to maintain the back-end code for all of Lemmy itself as well as the web interface. There’s no need to be so accusatory on an alpha release of an app, a bit of respect goes a long way.
