@cnx I've been surprised myself to find out that in 2024 it's apparently very hard to have a domain registered with a registrar without having any form of hosting attached (and pay for it).

@tshirtman There are usually scheduled tootctl scripts that admins run to clean up stuff older than a certain threshold. But the cache folder itself can't simply be wiped without basically everything on the instance breaking (including emojis and avatars), and it's often very painful to fix it.

It seems like one of those bad design decisions where the cache folder has been used as a dropbox for almost everything, including things that the software is supposed to store long-term.

So even with the cronjobs scheduled to run every day and very little local traffic, an instance connected to a few big relays ends up having many GBs of media cache that it can't touch.

Ok, git.platypush.tech has successfully been migrated from Gitea to Forgejo.

It didn't take long - it is indeed a drop-in replacement, but the systemd configuration packaged with Arch required a bit of tweaking to point to the previous Gitea paths and some permissions changes, *only then* it's an actual drop-in replacement.

Luckily I've always used `git` as a username rather than `gitea`, so that part, after a few changes in the configuration, won't require any more changes.

Tip: never, ever use the default username provided by your code forge (gitlab, gitea, forgejo...). Code forges come and go, and one after the other they are doomed to enshittification. Code repositories, configurations, scripts and documentation are there to stay instead. So never use a username that is tied to your code forge, or you'll have a hard time changing it later.

@santiago I'm now in the process of migrating to #Forgejo, which is supposed to be a fully FOSS fork and drop-in replacement for Gitea - let's see how long this lasts before pushing all of its users to their cloud offering too...

@bob well, given my hatred for Github after their latest decisions, I'd be more than happy to take over a federated "webgit" project if there's really enough momentum around it.

@BitAcadie let's do the math together.

Pulling some publicly available data (some of that referenced in the articles you mentioned), #Bitcoin:

- Processes 230,000 transactions per day
- Consumes about 110 TWh per year

That accounts to an average of 1.3 MWh per transaction (1.1e14/2.3e5/365).

So the estimate from the MIT article is actually quite accurate.

The average US household consumes 11 MWh per year.

Therefore, in order to process a single Bitcoin transaction, in average, we need more energy than the average US household consumes in a month. Numbers don't lie.