Hey,

I don’t there’s much more you can do other than to get in touch with that instance’s admin. I suggest sending them a direct message, they should be able to help by modifying stuff directly in the database.

The 2FA implementation is unfortunately pretty buggy still.

I have a Lenovo ThinkPad T430 that is by now almost 10 years old, it runs perfectly on Linux and is a fantastic choice even today. It’s built like a tank and that Intel i5 powering it is immortal. DDR3 RAM is dirt cheap now and it takes up to 16Gb, you can swap its HDD to a SATA SSD (if not done already) and batteries for it are still cheap and plentiful.

If you’re looking for something affordable for software dev, I can’t think of a better choice for $200-$300

https://www.lenovo.com/lt/lt/laptops/thinkpad/t-series/t430/

They really built this one right, they don’t make them like this anymore.

This is a Lemmy-ui bug, not related to CDN/cache.

I had submitted a bug report for it some time ago but it’s not that urgent in the great scheme of things so they didn’t get around to it yet: https://github.com/LemmyNet/lemmy-ui/issues/1865

(the report speaks about this happening after you log in, but it’s the same thing).

Awesome project and the tool looks great - nice work!

There was an existing tool which was similar: https://federation-checker.vercel.app/

Yours is definitely nicer.

Hey, thanks for your feedback.

Good point on the Admin account potentially not being e-mail verified, I’ll add that bit in there.

Lemmy stores a record of users in two places:

• local_user is where local accounts are stored. Each entry has an id and a person_id.
• person gets an entry for every user@instance that your Lemmy instance learns about, including users which were created on your instance. local_user entries are mapped to person entries using the person_id.
• local_user doesn’t actually keep track of user@instance - it’s all stored under person and referenced via the person_id.

This means that to free up the user names, you would also need to delete the respective entry from person. There are a few ways you can achieve this:

• You could grab a list of all person_id numbers where email_verified=‘f’, then use this list to delete from person.
• You could just delete the specific username you want to free up from person (where name=‘username’).

Edit: This also explains your confusion around the id of the Admin username - id would’ve been ‘1’ and that’s the one that matters. person_id was ‘2’ because your instance likely learnt of another user before the Admin somehow. I just checked mine and person_id is also 2, so it seems like the standard thing it does.

I didn’t worry about this too much since these were all randomly generated junk usernames that nobody would ever miss and I didn’t think it was worth the extra hassle to try and delete them from person too, since it would be a bit cumbersome especially if there’s a lot of them.

Anyway, to free up those usernames just delete them from person using one of the options above.

I hear you on the need to delete junk data, my hope is the devs will eventually include some decent tools for that. It’s a good suggestion, I may put something together when I have a bit of time.

A solution you can try meanwhile is to use an admin account to ‘purge’ the stuff you want to delete - purging does remove everything from the server. Annoyingly you don’t get the option to purge your own posts as the admin, but you can use a second admin account to work around this.

That project is a great find, thanks! A real time saver, I should have these marked up shortly.

I’ll do some more thinking on how we represent Cloudflare instances in general, I think for the time being I may just include a short note for them as people will wonder why they’re marked anyway.

Edit: This is now done, all Cloudflare instances marked and an explanation added.

Hey, thanks for your feedback. I like your idea of labeling Cloudflared services, reporting is indeed a bit tricky for those especially if they use “Always online” to serve cached copies while the instance is down. I have some ideas on how to combat that, but labeling them also makes sense.

I can add tags against services - I have done this for ani.social as a proof of concept, I think it works but I welcome feedback. Sorting through the entire list is a bit daunting and will take me a while, but I’ll get there.

Manually adjusting availability is a can of worms that I don’t want to open, I’d rather we try to find other ways to level the playing field.

Yes it was changed some time ago: https://github.com/LemmyNet/joinlemmy-site/pull/170

I think you’re one “active user” short of being back on there:

I believe you need 5 to be listed.

I’ve updated the post to include clickable URLs. Lesson learnt that I shouldn’t rely on clients to treat domain names as links.

Thanks for your feedback. I like your idea, I’ll have a look into what I can do to make it happen.

Absolutely!

Uptime Kuma Github: https://github.com/louislam/uptime-kuma

Note that I am not the creator of Uptime Kuma, https://github.com/louislam gets all the credit there.

It’s just a play on words: Le(mmy) + Stat(us) or Le Stat(us)

I think it’s short and easy to remember, the fact that it’s also the name of a famous vampire is just a bonus.

Https HEAD requests - I find it yields accurate results for Lemmy and it’s extremely light on bandwidth.

https://lestat.org

Hey thanks for your feedback. I’ve made them all URLs now.

Search and sort are tricky as it’s not something Uptime Kuma status pages can do, I’ll look into what can be done there.

Of all the ones to miss!

Thanks for pointing it out, I’ve added it now.

Have you thought of self-hosting mailcow? https://github.com/mailcow/mailcow-dockerized

It’s very easy to set up and fairly straightforward to maintain, if you have a static IP and it’s not impossible to get a PTR record then I highly recommend it. Yes you’re self hosting your own mail server but mailcow vastly simplifies this.

Alternatively plonking it on the right VPS can also work.

Hey, that is weird behaviour, I haven’t encountered it in my testing with Chrome. There are a few things you can try to figure out what’s causing it:

1. Try clearing your browser cache (cookies, offline site data, etc): https://support.google.com/accounts/answer/32050?hl=en&co=GENIE.Platform%3DDesktop - sometimes, especially after Lemmy updates, things can be a bit wonky until you do this.
2. If the problem persists, try using another instance to see if the same thing is happening there too. You’re welcome to try lemdit.com if you wish.

This will help you narrow down the problem.

I suspect there may be two separate things going on:

1. The comment behaviour you mentioned may be related to browser cache and recent Lemmy updates introducing infinite scrolling.
2. The generic page error is likely related to the instance you are using. lemmy.world has suffered through a few DDoS attacks recently so it’s possible that you are getting those errors when the instance is down due to those attacks. If it happens again, then perhaps take a screenshot and post it here.

Let me know how you go.

That’s fine too, the short of it is it’s entirely up to you.