No Stupid Questions

Most aren't big enough to be worth the effort, either by size or userbase.

Torrents have little to offer when you aren’t pirating bulky media files. And don’t track frequent release versions, code repos, or integrate with distribution packaging ecosystems.

Is there any practical reason they would offer torrent downloads?

A lot of foss projects don’t make sense as a torrent download. For instance, most of my foss projects are libraries that people who write websites use to compile and bundle the JavaScript they ship for their website. It wouldn’t make a lot of sense to distribute this via torrent, since torrents don’t have a lot of the features you would want, like checking for new versions of the software. Because the majority of foss projects wouldn’t make sense to distribute via torrent, the developers on projects that would make sense to distribute via torrent are more likely to vend it via methods that there is more infrastructure for (like a package manager), or is easier to set up and maintain (like direct downloads). I think if you were interested in changing this part of foss culture, your best bet would be to create and publish a GitHub action that would package and upload a distributable to a torrent provider.

I think size is the biggest factor. OS ISOs are pretty big, so having a managed download is helpful. If the 100 GB triple-A games were open source I would certainly expect torrents. But the FOSS things I download directly are pretty small, and the vast majority are done through a package manager or docker compose. So there may be a Goldilocks zone in the middle where it'd be helpful, but in those cases I'd expect a small installer that downloads the bulk of it for you.

So not much benefit for the consumer, but what about the provider? Spreading the traffic would reduce load on the hosting server which is a positive. You'd still have to handle the bulk of the traffic until the seeders outpace the leechers, but on long enough timescales it'd be helpful. Except you can't really update a torrent, which means each release needs to start fresh. This still works for OS creators because updates tend to be far apart, there is a large user base, and there is still some market for older versions. For regular programs you may only get one of those three, at which point adding torrents may be more hassle than it's worth.

They do when it makes sense. Hosting small Giles is not a big deal, and small files would be a bad experience on torrent, since by the time you get peers to start the download you would have already finished it from a normal server, not to mention that you need to host the torrent file anyways. Also things that have lots of releases/patches are a bad experience, since people might be seeding an old version.

However, large files that only get sporadic releases, such as distros iso, can definitely benefit from it and you can usually find torrent links for them.

I don't think there's stigma. Most users use package managers to get their software. For large operating systems, a torrent does help make downloads faster and less expensive when many people begin to seed it (and many FOSS operating systems do offer torrents), but most projects won't benefit from that.

I personally would not want to go through the hassle of getting the magnet link, putting it in my torrent program, waiting for it to finish, verifying the signature (if there even is one) and the checksum, and only then manually extracting it so that I can use it.

I could be wrong but aren't torrents immutable. So only that particular release is valid. Should a new patch be released a new torrent will need to be disseminated.

Compared this to a http download where you can reuse the /latest/ folder for water is the most recent release.