federico3

signal is designed not to trust the server

Unfortunately this is not enough. A malicious Signal server can mount a timing correlation attack and infer the social graph of an user. Having a centralized server makes it more difficult to mitigate such risk.

free people from proprietary gardens, yet FOSS has actually been one of the biggest creators of such gardens

Forgive the nitpick, but FOSS is not creating walled gardens, companies are. (After all, software has no willpower... yet)

I'm surprised nobody pointed out that tyres are extremely polluting:
https://www.theguardian.com/environment/2022/jun/03/car-tyres-produce-more-particle-pollution-than-exhausts-tests-show

Element (and other crucial components of the Matrix ecosystem) received many rounds of investments including https://element.io/blog/element-raises-30m-as-matrix-explodes/ (These are investments, not donations.)

I would not be surprised if the usual bait & switch lock-in mechanism happens here as well.

Then yes there’s EEE danger. Hopefully the Mastodon developers will resist that.

Unfortunately developers can do very little to prevent that. EEE works by first attracting a large userbase into a service and later on prevent them from leaving. It's up to instances admins and users to defederate to prevent EEE.

This is debatable. The GPL allow redistribution of a given version of the software without additional restriction. If the user receives that copy knowing in advance that redistribution will lead to retaliatory actions this can be treated as an additional restriction.

Here's a summary: https://ploum.net/2023-06-23-how-to-kill-decentralised-networks.html

Software licenses cannot solve every problem and AGPL is still the best option.

There are many larger problems related to FOSS including freeloading, right of repair, surveillance, lock-in... and they require social solutions rather than new licenses.

That's besides the point. Of course it's always possible to create new communities on new instances, and import posts from various sources, but the original community would be still gone.

If an instance is shut down or becomes unusable for a long time there is no way to automatically migrate users to a new instance. Additionally, there is also no guarantee that all users will move to the same alternative instance. This can also cause unnecessary conflict around which alternative instance becomes the "legitimate" successor.

This is not correct. APT always verifies cryptographic signature unless you explicitly disable it. Yet it's very important to understand who is signing packages. What kind of review process did the software go through? What kind of vetting did the package maintainer themselves go through?

If software is signed only by the upstream developer and no 3rd party review is done by a distribution this means trusting a stranger's account on a software forge.

Update: the Debian infrastructure supports checking gpg signatures from upstream developers i.e. on the tarballs published on software forges.

What you are describing is just a local cache of !lemmy@lemmy.ml on your instance and it works only if it has been populated before the downtime of lemmy.ml. If lemmy.ml never comes back to life nobody can post to !lemmy@lemmy.ml proper. All the communities on in would be dead.

Context is important. It's possible that the software is distributed without any warning like that and that the termination of the support contract is done without citing the redistribution of previous versions as a reason. OTOH if the customers could prove that there's widespread knowledge of the retaliatory termination that could be equivalent to a (non-written) restriction that is indeed incompatible with the GPL