I think the argument is that those alternatives already existed before. Twitter was not being prioritized, it was essentially mirroring the content already available in RSS, mastodon, etc. So effectively, there’s now one less place where the news will be visible.

However, I do agree with the move, but only because Debian being a FOSS initiative should stay away from proprietary platforms and promote FOSS, even if it means effectively “shutting off” a portion of users who don’t wanna leave the twitter bubble.

Were they using Twitter to provide exclusive updates not available anywhere else?

My impression from the post is that they are publishing the exact same updates in multiple locations, including mastodon at https://framapiaf.org/@debian …so just because they were publishing in that one extra site to make it accessible to a particular subset of people does not mean all other people were being shut off from receiving updates.

However, I do agree with the move, but only because Debian being a FOSS initiative should stay away from proprietary platforms and promote FOSS.

I don’t know enough detail about ATproto, but I wonder if it’s technically possible to block access to posts without also blocking federation. From what I’ve heard the functionality is more modular than Activitypub (content indexing being a separate service from content hosting) so I wouldn’t be surprised if it wasn’t possible.

I think there are situations that fsync does not cover very efficiently, to the point that it can cause timing issues that lead to some bugs / incompatibilities. The timing issues might be rare, but that doesn’t mean the overall efficiency is the same. It would be interesting to see benchmarks of fsync vs ntsync.

I’ll probably release on March, since there’s a release every 2 months and 6.13 was on Jan 20th.

But those are small fries, not “the provider of games”

They have less to loose, then. That’s just as dangerous, if not more.

I’m a small fry too, would you run a binary I send you without any form of sandboxing?

we don’t run games as root

No, we typically run them with the same user that stores all our useful private data and that we typically type our passwords with.

Also, why are you OK with that level of sandboxing? don’t you want more “control”? You say containers are bad, but using user roles to protect parts of the system is ok? why are you not running all as root if you want “control”?

we are speaking about Wine, so what they see is limited to WINEPREFIX

Not really, by default you have access to other drives (Z:\ being /, the fs root), wine is not a perfect sandbox, it’s not designed for that… and if you actually did want it to become one (which ultimately would also lead to a need for memory separation to fight memory-leak attacks) then it would not be that different from what’s being pursued. You’d be essentially building the container in a custom version of wine shipped by Valve on Steam, it does not make any difference in terms of “control”.

Currently, in order for Android app to appear in the official Store, developer has to allow Google to repackage their app and sign it with Google key. So while we can inspect what is there in the code of the app in git, we don’t really know what lands on our phones if installed via Google Play

You can still open an APK and decompile it… it being signed with a specific key is no different than the digital signatures some attach to their emails, it’s a way to prove authenticity, not a way to encrypt the message… you can open the email without having to even care about the signature.

We have no control over what they put in those containers

Most games on Steam are proprietary software you don’t control to begin with. It seems reasonable to keep them encapsulated in containers (+1 if you run Steam on flatpak or so) rather than granting them the capacity to run amok in the entire system, which we would have even less control over.

It seems contradictory to want to remove barriers that are preventing the software from taking more control, and at the same time complaining about how they are having too much control.

Personally, I would be happy even if it didn’t translate it but were able to give some half decent transcription of, at least, English voice into English text. I prefer having subtitles, even when I speak the language, because it helps in noisy environments and/or when the characters mumble / have weird accents.

However, even that would likely be difficult with a lightweight model. Even big companies like Google often struggle with their autogenerated subtitles. When there’s some very context-specific terminology, or uncommon names, it fumbles. And adding translation to an already incorrect transcript multiplies the nonsense, even if the translation were technically correct.