4 ms·
At this point, I know this link almost by heart. This is not useful. It doesn't mention Firefox at all. It doesn't contradict at all what I'm saying. It doesn't
by jraph 2y ago
At this point, I know this link almost by heart. This is not useful. It doesn't mention Firefox at all. It doesn't contradict at all what I'm saying. It doesn't hint at you being right neither.
But okay, let's dig in anyway. PureOS, a FSF-approved distro listed by your link, has a FSF mirror and this mirror contains Firefox [1]. Mistakes (to be proven) aside, this makes for quite a strong endorsement from the FSF itself, actually.
Now I expect some more substantive argument or I won't answer anymore. I'm a friend, I'm heavily biased toward FLOSS, it's not like I'm trying to downplay your activism around free software, but we need strong, convincing and correct and kind, gentle and respectful arguments if we want this to work. It's not even like I'm trying to defend Firefox at all cost, I don't like some recent moves from Mozilla including this one.
[1] https://mirror.fsf.org/pureos/pool/main/f/firefox-esr/ https://mirror.fsf.org/pureos/pool/main/f/firefox-esr/
- fsflover 2y agoThanks for sharing your thoughts. You are right in principle that FLOSS apps can and do have telemetry/surveillance included. In practice, it's much, much less frequent than for any proprietary apps. If we go even further and make the FLOSS definition as strict as it gets (i.e., stick to FSF), then such problem almost disappears. Does any other FSF-approved distro include Firefox? I guess Firefox should not belong to PureOS. If you read the criteria for the free distros, you will see that anything that downloads or even mentions nonfree software should be excluded. Even Debian was excluded for this reason, including one with only "main" repository enabled. Firefox mentions DRM and downloads it in just one click. AFAIK PureOS didn't include Firefox in the beginning, but the users demanded it. See also: https://www.fsf.org/news/fsf-condemns-partnership-between-mozilla-and-adobe-to-support-digital-restrictions-management https://www.fsf.org/news/fsf-condemns-partnership-between-mo... From the above link: > Use a version of Firefox without the EME code Oh, I found that Firefox in PureOS doesn't include the DRM part: https://forums.puri.sm/t/not-possible-to-enable-drm-content-streaming-in-firefox/19339 https://forums.puri.sm/t/not-possible-to-enable-drm-content-....
- jraph 2y ago> You are right in principle that FLOSS apps can and do have telemetry/surveillance included Note that I was just replying to your "Firefox is non-free" statement, not to this. I do agree that the tracking story is way better when using free software in general (but that indeed, it's not a guarantee, especially when using official installers from sloppy software editors). ---- Code that is under a free software license that downloads (and runs) non-free software is still free software. This can be seen as an anti-feature, F-droid would call this one "Non-Free Addon" [1], but this is a separate matter. For a distro to be approved by the FSF, it needs to only include free software, and also match additional, more restrictive criteria such as, indeed, not promoting non-free software. But this doesn't qualify some specific software, it qualifies a distribution. In addition to DRM, Firefox also has other issues, including these: - it allows non-free addons on its extension repository, and even recommends (promotes) some of them (also Non-Free Addons) - it includes features that depends on non-free services, like pocket (Non-Free Network Services [2]) I don't like them, but that doesn't make Firefox proprietary. Worst case, it isn't suitable as is for a FSF-approved distro. Which, again, goes beyond the quality of being free software. The fsf.org blog post you link to (which I'm also familiar with) shares concerns (which I mostly share, indeed), but doesn't state that it makes Firefox non-free software. There's one point that I would debate from this link: > We agree with Cory Doctorow that there is no meaningful distinction between 'installing DRM' and 'installing code that installs DRM.' The distinction is meaningful, because embedding the DRM code in Firefox itself would make Firefox non free. However I agree that it makes Firefox promote horrible software without even warning the user, which is not great at all. tl;dr: The concepts of FSF-approved distro and free software should not be mixed up. They are both useful but separate concepts. I could even see someone being full pro free software but thinking that FSF-approved distro goes too far. [1] https://f-droid.org/docs/Anti-Features/#NonFreeAdd https://f-droid.org/docs/Anti-Features/#NonFreeAdd [2] https://f-droid.org/docs/Anti-Features/#NonFreeNet https://f-droid.org/docs/Anti-Features/#NonFreeNet
- fsflover 2y ago> But this doesn't qualify some specific software, it qualifies a distribution. I never thought about it. Why wouldn't it be the same for any software? A distribution is nothing else but a (large) set of software. Do you think one could find a software that downloads something nonfree in the FSF Free Software Directory? I do not expect that. Nowadays, in our connected world, it doesn't matter if the software nonfree or it downloads a nonfree piece. It's effectively the same thing.