Yeah, it’ll redirect. If for some reason that messes with federation for the Sharkey instance since a lot of lemmy instances will keep trying to make requests to the site even when the instance is down, I’ll end up putting up an HTML page saying that Burggit has shutdown and link to dill.burggit.moe.

Honestly? Good riddance. If your opinion on which software we use is what makes you not want to use our instance and the fact that MAPs are allowed on our instance, which is just me feeling like I owe RQD2 a favor for having our backs when we got widely defederated on here makes you not want to join up, then fine. Don’t use it.

I spent weeks going over which fork of Misskey we’d like to use since they’re all similar but have their own set of issues. Those being, some aren’t maintained (Firefish), others (Iceshrimp) are moving completely to C#, or some are in the beginning stages of their fork (Catodon.) I had my reservations about choosing Sharkey because of the transfem stuff, but the thing is, that:

• They’re a soft fork of Misskey meaning that if the project dies, we have a way of going back to Misskey if the need arises
• They have the bubbles feature which is where a list of instances that the admin approves of are in its own separate timeline. A feature I especially desire because the complete fediverse timeline can get pretty noisy, IMO.
• They’re being actively maintained. Their Discord has a lot of activity. And their git repo has signs of life.
• It has 1:1 compatibility with the Mastodon API. Which means Android apps like Husky, Tusky, Fedilab all will completely work with the software.

So me being the pragmatic person that I am chose Sharkey for the reasons outlined above. Again, this was me taking a couple of weeks of my free time trying out the various forks of Misskey to see which ones that I like and will make sure we aren’t screwed somewhere down the line if the project ends up dying or if the maintainers have a falling out which kills the project.

Now for the MAP thing. I have been outspoken about 100% being opposed to MAPs on our Matrix and RQD2’s Matrix. They haven’t banned me because of this and they have chosen to remain federated with us despite us being more of a free expression-y instance that a lot of the peeps on there would find distasteful, I feel that I owe RQD2 a thanks for having our backs when we have been widely defederated, and that’s why I went with the compromise rule that I came up with.

This post came across as you pissing and moaning about how we run things despite the fact that we are self-funded and don’t ask for donations. We do this in our free time, so this honestly came across as a slap in the face as well as exceedingly ungrateful.

Going back on my, “This is the last post I’m making on Burggit.” Already apparently, lol. This was something that I thought of covering, but ultimately decided not to, since I thought the post was long enough as it is. Thank you very much for bringing this up.

A lot of our issues were due to the growing pains of our host who, don’t get me wrong, promptly responded and fixed issues as they popped up. The big reason we stuck with him is because he went to bat for us when someone did their typical crybully shit and tried to go above us and get us censored by our ISP despite the infringing content being completely legal in the country that the host is in.

I highly recommend kyun.host for this very reason because they will promptly tell any complainer to fuck off so long as the host they’re complaining about isn’t doing anything illegal. It really is censorship resistant.

I’d get pinged all the time on our Matrix that the server is down or is slow and other stuff which caused the components of the server, namely pictrs to crash and exhibit undefined behavior, since our image storage host was accessed via an NFS share to a spinning rust server with more ample storage mounted through a Wireguard tunnel. With each server going down all the time, it’d cause Lemmy to seriously break and require a manual fix to get it going again. Over time, I’ve gotten kinda burnt out from maintaining the site and slowly just mostly stopped using it altogether aside from a cursory glance every now and again to make sure it was still working, due to my increasingly waning lack of interest in perusing the site.

If it wasn’t for the funding issues, this instance probably would’ve remained on the same path to being shutdown just at a later date, most likely. Just because it’s too much for one person to maintain given all these situations along with the software still being of a beta quality piling onto our already existing issues. And frankly, I don’t have fun using it anymore and honestly completely dread working on it.

I absolutely refuse to create an account on any of the popular instances, which would probably end up in a repeat of the Lemmy Admin Matrix room situation where I laughably got permanently banned on sight for “CSAM” even though I never posted any such thing, rofl. So I’ll leave this here, the hyperlink in that post references their Fediseer tool:

This is precisely another reason why I’m glad we’re shutting down, the threadiverse is so stifling that we have to be on our best behavior or else we get labeled through your dumb Fediseer tool. Sometimes wrongly, (IE: calling us bigots when we fully support the trans movement, as well as the overall LGBTQ movement as a whole, and any movement that’s focused on minorities.) labeling us as a bad instance. The fact that your tool doesn’t allow instances to opt out can create a completely slanderous situation if a clique of high profile instances decide that they don’t like you.

Burggit will live on in a frankly more fun microblogging fork of Misskey, called Sharkey in the wider fediverse. A big chunk of which, will not give two shits about the reputations passed down to instances by this tool wielded by a clique of popular Lemmy instances whose behavior would make a typical Reddit powermod blush, since it takes itself way less seriously: dill.burggit.moe

I really should write up a guide for it. I’ve just been procrastinating on it. The gist of it is that each timeline has the following:

Home: People you follow

Local: Everyone in your instance

Social: Home + Local

Bubbles/Reccommened: Servers the admin reccomend (like a network or webring)

Global: Every server the instance is federated with

I’ve chosen these instances for the bubble timeline:

varishangout.net

bae.st

pleroma.skyshanty.xyz

shota.house

kyou.social

breastmilk.club

lain.com

burggit.moe

uoh.cx

shortstacksran.ch

baraag.net

shrine.moe

dill.burggit.moe

akkoma.kikuri.moe

binky.fish

kitty.haus

cannibal.cafe

I personally lean more towards Misskey because I find it dead simple to maintain. Kbin/Mbin has a lot of moving parts to it.

I think I figured out the problem. Need to give it a while for federation to catch up.

Yeah, toss that condom by the wayside.

It depends on the policies of the home server. It’s technically allowed on Burggit’s space, but we’ve never had any requests for such a room on here. Most of what gets discussed is technical issues with the site, giving updates on matters, etc.

It’s similar in concept to the fediverse on here. You pick your home server which would be something like cutefunny.art, uddu.chat, matrix.org, etc. and you go to your destination room/space, in this case, you’d go to https://matrix.to/#/#burggit:burggit.moe

It’s less work for me. Since all I’m doing is a text replace at the reverse proxy level. I’m not having to change code for this since I’m just replacing what gets sent out. It’d be one thing if this breaks apps, but afaik apps just use the Lemmy API.

There’s been a DDoS hitting my host. It’s not directed at us since it’s hitting the /24 subnet. Not sure if you are on my home server, but I suggest getting an alt if you want to get on there, if this is the case. I can’t decouple the Matrix home server from the Burggit VPS because the domain is needed to point at the Burggit VPS.

Seems it’s resolved for now.

Running into heavy packet loss again for the moment…

Congrats!

This appears to be an app issue, then. I scrolled through my subscribed feed on the Boost app and even found a post from that community, and no crashes occurred there. Nothing I can do about it, I’m afraid. I guess report the issue to the maintainer and hope that you don’t get banned for the contents of the report (lol)?

It outright crashes. The logs don’t show the instance trying to send anything when it happens, so it’s purely a client-sided issue… Do try another community from an instance that’s fed’d that runs an older version of Lemmy to see if it happens with that one too (If you can find one.)

Which community was it? I’m curious to see what data the server tries to return, assuming I can reproduce this on my end.

E: I noticed that it’s running 0.19.0. Maybe try another server we’re fed’d with that runs an older version to see if this same behavior happens.

I tried setting subscribed to default like you had and am still not able to duplicate this. I forgot to point out that this is a fresh install, so maybe clearing data/reinstalling the app is in order?

I primarily use Burggit in a PWA on mobile since it seems to work pretty darn well by default. I’m on Android 13 too if that matters.

Not able to duplicate this on v23.11.29. You select the subscribed button and scroll down to achieve this crashing behavior?