The entire app doesnt feel “smooth” everything is alright but scrolling is a nightmare. I’m wondering if everyone is experiencing this or just me? Using a Samsung a04.
EDIT: The website is perfectly fine no lags or stutters. I don’t think it’s an issue with too many people since opening posts and loading images are instant its the scrolling that seems too lag
EDIT 2: https://github.com/dessalines/jerboa/issues/445 Found this in git
Also having this issue. When I place my finger on the screen to scroll, it initially kinda “stick” a bit as if it wanted to act like a tap, not a swipe. But then it does start the scroll. And yes, the scrolling action is hard to describe, but certainly not smooth. No other app feels like this.
I’m on a OnePlus 6.
Same problem here, not an issue with other apps, just this one. Also sometimes I will be scrolling comments on a post and the page will randomly jump back up to the original post at the top.
On a Galaxy A53
Same here also. I have never noticed this in any other app.
Have this too. Jumps at like 5fps. Quite an old phone but it’s definitely the most pronounced on Jerboa
Have it too on a OnePlus 8T. It’s not that bad, it feels like framerate dips to maybe 40-60 Hz instead of 120Hz locked.
From previous Android dev experience, probably creating lots of views for posts in the list when scrolling fast.
Not guaranteed to be the answer, but it might just be the instance you are on being overwhelmed with users. My experience is great personally on a google pixel 3. If you were to try a less common instance, it should improve. Everything works great over here at sh.itjust.works.
Results may vary, but that seems to be a common sentiment. Lemmy was not designed for everyone to be on the same server like reddit is. Create an account somewhere not as popular and see if there is a difference.
Why would the instance affect the app’s scrolling? Not saying it wouldn’t, just not sure how it would.
I mean, it could be the app, but I believe it is potentially the instance because an instance with a lot of users is having to feed a whole lot of users at once. Which is much more resource intensive than the way the fediverse was designed to work by sharing with multiple instances to share the load of users constantly scrolling at the same time.
Imagine you were tasked with handing out a thousand hotdogs at a sporting event. You could do it yourself, but you are going to get overwhelmed very quickly and the line is only going to keep building up. Then imagine you have a hotdog delivery network where there are still 1000 hotdogs to be delivered, but you can share the load by having other vendors take 50 hotdogs each and giving them out at their station. The people get their hotdogs faster as they aren’t all waiting on just you and you aren’t overwhelmed being the hotdog overlord.
I personally believe that users need to start spreading out to new instances so we don’t overload the most popular instances. Lemmy was not designed to work like reddit where there is only one hotdog overlord. Lemmy is better built to be used at scale with much smaller vendors. Another neat feature of this concept is that the fediverse can’t be shutdown unless everyone stops selling hotdogs. If one goes down, everything still works fine, you just don’t have whatever that particular instance was giving away. Unlike reddit where if there is a problem, the whole site is down and nobody gets any hotdogs.
Okay.
But the UI scrolling shouldnt be affected by the server being hit by many requests imo. The only way I could see an impact is if the UI waits for something to load before allowing it to be scrolled in to view - but I’m no front end (app) dev. I did work on a content serving project as a full stack though, except that was a few years ago and browser based and only vanilla js and html on the front end. Still, scrolling issues were not impacted by content serving latency as it was easy to plug dummy tiles in until the server gave its content.
The hotdog analogy doesn’t make sense to me as that would be a Lemmy instance (all content serving server) latency issue, not a jerboah app UI component issue?
Spreading out over multiple instances is one option to reduce content serving latency, but it would also be interesting to see if a single instance can be distributed over multiple servers - I’m not sure of the lemmy capabilities at the moment though.
My guess is that that the UI is waiting for something from the server that hosts the user and therefore the content. That is more or less what I meant with my analogy. Keep in mind that I don’t know this for certain, I am just going off of things that I have heard about how the fediverse works. If a developer for the app or lemmy can weigh in for a more complete answer, that would be great. I could be completely wrong for all I know and am open to knowing more.
If single instances become large enough, finding a way to split those up may be necessary if you can’t just improve it by getting better hardware or optimizing lemmy to run better. I guess we see what becomes necessary.
I’m the only user in my instance and scrolling is definitelly laggy, and I feel it stops scrolling unnaturally sometimes, refresh rate also seems more like 40hz on my 120hz panel
I am curious to why that might be. It could be for several reasons not related to jerboa. Such as what hardware you are running on, network limitations or issues, whatever you are trying to reach is still trying to be collected by your instance. I don’t really know.
Which android version are you on? I am having the same problem on Andeoid 13 and seen many posts mentioning 13. I’m thinking could this be specific to Android 13 ?
