Everything feels very snappy compared to 0.19.3

Looks like a Chenbro rm-42000

It’s a damn shame it’s going not free open source, I Just switched my lab over to nomad and consul last year and it has been incredibly smooth sailing.

:wq

RiP

Mailcow includes rspamd which learns spam using bayes analysis. Just move messages to the junk folder and it learns, after a month or so of training I get very very little actual spam, and no false positives.

Is robotics 19.2 was awesome to see pages of text load at a time vs scrolling in at 300baud. Telex was wild.

476M ./postgres 1.1G ./pictrs

After 3 weeks

Woah, wefwef is awesome! Using it now!

How does it compare to invidious? Worth switching?

The discussion is more interesting than the video

What a fantastic but subtle way to differentiate instances

This is true, It is much safer to rely on providers at the protocol layer vs the application layer.

I had no problems with mlem

Is the blocker here that each instance is a single postgres/Lemmy process? I imagine a clustered inplementstion of the Lemmy backend could be used to shard individual communities to dedicated containers when they reach a given size, proxies through a community away load balancer? More to manage but would let instances scale up/down as needed. There are costs associated with this, but those of us who run instances do it because we like playing this game.

This is quite the conundrum that the fediverse requires corporations like fastly/cachefly/cloudflare to stay afloat/responsive.

Primarily rely on zfs for file system replication.

I have primary/redundant nas on site, then a single node offsite connected via vpn.

On my list of things to tinker with is zettarepl.

I will do you one better, I have my instance behind 3.

WAN -> haproxy -> traefik ingress w/ letsencrypt -> Lemmy nginx -> Lemmy-ui

I can probably remove the lemmy nginx but it only uses ~10mb of ram and didn’t want my changes getting in the way when sorting federation issues(which work fine!).

It’s so good!

And at the very least, there’s a record of the discussions and thought processes behind why this was or wasn’t chosen.

What if each instance had a message broker distribute updates in a pub/sub topics oriented fashion? Does the activitypub spec specify that instance X must http post updates to instance Y or is there room for implementations to get creative?