Tiff
Self Proclaimed Internet user and Administrator of Reddthat
- 119 Posts
- 404 Comments
That awkward moment when you are the person they are talking about when running beta in production!
beta.5 FIXES OUR pagination!
Wow… might I suggest the “newer” https://next.reddthat.com ? instead of old?
I know it might not be feature complete at the moment but it seems mlmym just… is not that great
I bet it’s something to do with the API / caching. Last month-ish we enabled NGINX caching to help with the “spikey” nature. It should only affect the main front end, but I hadn’t tested the frontends extensively.
I got the same issue. Even after modifying the sort settings it stayed being “hot”.
Bug:
subscribe & unsubscribe links are the wrong sizes on
communities?listingType=Local
“Bug”:
Your IP is visibleto
Suggestions:
- Padding on left side / center mode option
- Maybe its better on smaller screens, but on a 27" there is a tremendous amount of dead space.
- Comment threads are “clickable” like images?
- Instead of having to click the title, allow clicking the Comment Icon
- Instead of having to click the title, allow clicking the Comment Icon
Sent from next.reddthat.com :)
- Padding on left side / center mode option
Every bit will be used for copious amounts of coffee! 😜. The old server got really cranky, but this new one is raring to go! (Except for this db/pagination issue, but I think I found a good way to investigate).
Thankyou so much!
Thank you for articulating your thoughts. I always enjoy hearing the views of our users! I too will be looking at how this progresses. The more people we can migrate to the fediverse the better!
Though I can’t recommend getting a LW account 😛, purely because I want everyone on Reddthat. I think if you want to investigate Threads/etc, a mastodon type account would be the best course of action. As following single accounts on Lemmy is… non existent (I don’t count RSS feeds).
👀
👀Reverted 😓
As I was looking at mlmym’s docs, I thought I had forgot to turn dark mode on
Hmm… maybe you viewed your profile at the exact time when our database updated the ‘aggregates’? As I can’t seem to replicate that problem.
https://github.com/rystaf/mlmym/blob/ba4b330ca05df00a74817ed49ca71ea8848c4614/templates/activities.html#L13 it has been hardcoded to whatever user you are viewing. 'Cause technically (I guess?) the comment is by
username. I agree it does not make much sense. (Submitted an issue on mlmym for you)
Oh for threads.net? Once that happens I plan to re-evaluate how everyone feels.
If at the end of the day people feel strongly enough against it then we will keep it but I hope to open it back up again. As (I think) the defed only affects our communities, not content people post to other instances’ communities.
The way we “force federation” is to tell reddthat to “resolve” the comment/post via the API. It’s a roundabout way to force all of the content to still appear.
It’s interesting that it doesn’t trigger notifications but I’m not surprised it doesn’t and when you think about it makes some sort of sense.
The 5-7 day delay is when the actual comment/post is federated from LW as the “time behind” value is 6.3 days currently (and coming down).
I’m reaching out to one of the other admins who apparently wrote their own federation queue. That means we could return a 200 OK response instantly for everything and work on the queue on our own time.
There is still not that much movement on the dev side as it’s a complex problem and the devs don’t seem to want to develop something like a queue currently. Which I understand as it would be a completely new system and would need lots of testing…On-top of that, every “jump” in backlog results from some abusive kbin bugs. Kbin has some users who send a like and an unlike constantly. In the order of 50k requests in an hour. They like and unlike a post on LW. LW then has to federate ever like and unlike activity. Which ends up being 50k Activites that need to be sent per-server. So it makes federation lag for everyone.
So we’ve managed to catch-up a bit each day and then some kbin bug causes an issue and then we are now +1 extra day behind.
Hearing about the lack of notifications is disheartening. Is it so hard for a system to have some sort of quorum? :(
Hopefully we get something sorted in the next week or two… Otherwise I might block every vote coming in until we catch-up. (As votes account for over 90% of activities).
Tiff
Did 0.0.66 fix it as well for you?
The video uploaded fine, and plays and even has a nice thumbnail!
The problems might be your client. There even was differences between Lemmy-ui on mobile and lemmy-ui on web. There are a few clients that recognise when a link is a video and plays it/expands.
Bah! I forgot to let everyone know.
Since we updated to the beta version of Lemmy 3 days ago, it seems it’s broken.
I’ve already left Bug reports on Jerboa and Lemmy for the immediate issues. If we hadn’t found this out every Jerboa user would be in the same position.
Unfortunately the issue came about because I used the latest Lemmy build to get the metadata patch and test it out. Unfortunately there were database migrations I missed so we are just going to roll with it.
- Nutomic is adding back public_key in a patch, once that is back in I’ll rebuild our container and jerboa will be back working
- Eternity is a good alternative for an android app (which is what I mainly use).
I’m thinking it will be sorted by tomorrow? Hopefully. Otherwise I’ll make the code change myself and rebuild.
Sorry for the issue & headache. I only saw the issue in the last 24h-ish as I only have jerboa for the occasional testing.
Tiff
Swell, thankyou.