So, past few days, the instance I have my account in, seems to crash and burn every morning, when I would like to browse it. I wouldn’t otherwise mind waiting fir the few minutes it takes to the issue being fixed, but Jerbia decides it’s a great idea to PURGE EVERYTHING if there is any sort of hitch in the login.
Now, I do notice I get a some sort of javascript error every time this happens, and I suppose it’s a null reference error, but I also imagine the issue is not isolated to a situation where the instance itself is down, but can probably be replicated by a local network outage. And an easy solytion is to add a variable that tells the app that the logins used has been valit, with an option to change the password, rather than just lose everything. Zure, it might be cool if there was a tiny check to see if the instance exist, when loging in with a saved credentials, and notify about not being able to reach it, but the most important feature would be to have a “newly invalid login” to still exists within the app.
[EDIT]: To clarify: even if the logins on the app would still be there after the instance comes back up and the app has been restarted, the issue still presists, because there is no excuse for displaying the credentials as forgotten, and defaulting to anonymous, where you could simply display an error that implies the logins are still ther, but not reachable. Tint the username red, and add [login error] at the namespace. Or better yet, add a tiny ⚠️ icon in front of the handle, that will provide information aboit the error.
I am not looking how to solve this for myself, I am reporting a bug/oversight that will affect peoples retention rate with not just Jerboa, but with Lemmy as a whole. Now, I know Jerboa isn’t the only client out there, but I’d rather see it get better than try and find an alternative, but if the welcome for suggesting an improvement is “it’s not really a problem to me, so I don’t even care if it’s a good suggestion, you should just get over it too”, I’m rather tempted to jump ship if an alternative pops up.
While some people might consider this “just a quality of life thing, L2B, noob”, I’d say this is one of the things why open source centric apps rarely take off properly: the UX sucks because it’s dismissed as “unimportant, because I coded it, and I know how to use it and whst it does under the hood”.
Jerboa does not delete your account when it fails to reach the instance. Instead when it fails to retrieve your profile it falls back, to the defaults which is displaying the anon account.
I have a fix for this, and many other checks. But it still needs to be reviewed.
As the others said simply restarting the app makes another attempt.
btw your suggestion is already included in that PR
I’ve experienced the same problem… it’s happened to me multiple times.
Yeah, the most infuriating thing is that you can’t very rasily avoid it, because the best way to know an instance is down, is trying to check it.
Have you tried going into android app settings and killing jerboa? When it restarts, many transient login state issues get resolved when it logs in fresh.
To solve what exactly? The only reason I know the instance is out is because the instance has “out of order” statistics and a good issue tracker blog. By killing the app, I will just end up in the same situation, if the instance is still down. This is an issue that should not require user to “turn it off and on again”, but should be tought of by the developer while developing an app for a platform that is prone to these kind of outages.
The bottom line is, that not being able to reach an isntance should not nuke the local login information without a notice. I’m in the happy position that I don’t, at least yet, login to every instance separately, but I’d imagine if my way of keeping track of my instances, it’d be rather unhelpful for them to just be gone without a trace.
He’s right though, simply restarting the app is enough for me to be still be logged in once the instance is running again. It still shows you as logged in inside settings iirc.
Another option: Create a backup account on a different instance. If for example Lemmy.world is down, you can switch over to lemmy.ml. And you’ll notice that your login info for lemmy.world is actually still there. You can just switch back to your lemmy.world account to see if it’s back up yet.
I created a second account to work around this and it works, but that’s less than ideal.
I fixed the problem already, dessalines just needs to cut a release
