In 99% of those cases you would want it to resolve .. components though.

It is about as common as using a database server for content though to use something like ES, Solr or similar software for search.

I don’t really see the point in forking a project like Mastodon unless you are already deeply involved with its development. It doesn’t do enough that you couldn’t rewrite it better (as in in a way you understand better and with lessons from the original taken into account) in the time it would take you to fully understand all the details of the existing code base.

You included the . at the end of the sentence in the URL so it doesn’t work properly.

Seems quite reasonable to remove illegal communities given how Lemmy works with the caching of all content locally. There is just a significant legal risk for the instance operators involved here.

!article_interop@lemmy.ml

The proposal fails to sufficiently motivate why existing protocols like OpenId Connect can’t be used given that trusting the user’s home instance seems necessary with this protocol too. The name also is confusingly close to WebAuthn.

Does Box<[T]> have the relevant trait implementations to be used in all the places where people misuse Vec?

It might be interesting to analyse which of the Hetzner IPs are in which data center.

I am not sure this will work as intended since part of the problem is that most of the lurkers who read the original and never interacted with it beyond that have already moved on to something entirely different by the time any mistake admitting, forgiveness,… happens.

I think you can’t read it too literally. Otherwise you also can’t discriminate against “people who wrote their comment later” and so you can never stop reading the comments for fear of discriminating between “early comment writers” and “late comment writers”.

a token that you’ve been authenticated against your home instance.

I assume you are talking about OpenId Connect (or OAuth 2.0 but that is basically what OpenId Connect is based on) here. The crucial bit that didn’t really work out with this is the part where users just specify their OpenId Connect provider at login time. All uses I have seen in at least 10 years have a fixed list of providers to choose from because of these trust issues.

Why would I trust a random Lemmy server admin with authentication for anything other than Lemmy?

Do you mean tribes/… on social media or do you mean social groups that develop from social media somehow? Either way I think you are over-estimating the importance of social media for people’s tribalism.

I don’t really see how this would be useful. Having purely archived data available in a software designed to show you new posts feels like a format/content mismatch.

Well, in case someone has the same idea, I just checked and the string kbinBot does not appear anywhere in the lemmy git repo.

I also must say the whole conspiracy nonsense in the comments over in that kbin link you posted really doesn’t look good for their user base. Most likely this is some sort of bug or compatibility problem.

Modern PHP is better than PHP5 but it still uses that brain dead execution model where every request starts the entire framework from scratch.

P2P based high-bandwidth uses seem really hard to implement on mobile. I can absolutely see why a mobile app wouldn’t be a priority for them since it is significantly harder to impossible depending on local mobile contracts.