[staktrace]

It should only try to switch the user's default search engine if the user hasn't explicitly set a default search engine. You can verify this by install a Firefox 56 build from the ftp archive, explicitly changing the default search engine, and then letting it update to Firefox 57.

[Drdrdrq]

Not true - I had to explicitly remove Google (again) and set StartPage as default on all my installations. I think it happened with 57.0.1 upgrade though.

[staktrace]

Hm, if that's the case, then it's a bug. If you could file a bug at bugzilla.mozilla.org it would be much appreciated. Even though it already happened we can try and ensure it doesn't happen again.

[catwell]

Apparently it is a bug. It works for search engines shipped with Firefox (such as DuckSuckGo) but not for others (such as StartPage of Qwant). See https://twitter.com/gchampeau/status/938757076213518336

[feelin_googley]

It sounds like Mozilla just wants to keep users confined to a pre-approved list of search engines. Are these also business partners with Mozilla, like Google?

Here are some other HTTPS-enabled search engines to test. If I am not mistaken all have been mentioned on HN in the past.

    https://wiby.me 
    https://gigablast.com
    https://mojeek.com
    https://unbubble.eu
    https://ecosia.org
    https://searx.me
    https://symbolhound.com
    https://findx.com

[kazagistar]

It looks like a mistake, not malice.

[Nomentatus]

you mean that the omission of the above lesser known engines was a mistake and they will soon appear? I grant that you're responding to a "response" that should have been a new thread, instead, since it's really a new topic - but it's not alleging the switch was malice.

[bigbugbag]

I have a hard time believing this kind of happy coincidence benefiting their commercial partner could be a bug, it really looks like an intended feature.

[bigbugbag]

What do you mean much appreciated ? My experience with bugzilla is that it is a place where users are ignored and told they are not part of a large enough portion of the user base to be significant for using linux, or for using alsa or for needing an option to keep the previous interface.

I do not dare going close to mozilla's bugzilla, it has been a waste of my time almost every time and taking abuse and frustration is not exactly an enjoyable experience.

But hey this does not really matter anymore as after over 15 years of supporting mozilla I have had enough of its utter nonsense and have now left firefox with the firm intent to never return and have stopped supporting anything mozilla while actively hoping it will fail and disappear, the sooner, the better.

Way too much disdain for the people using firefox, too much discrepancy between the marketing and the actual thing. Alienating long time supporter one by one is probably not the smartest strategy but hey what do I know, mozilla is making hundred more millions with a fraction of the past user base so they're probably successful by their own metrics.

[roblabla]

I have no idea where you get all that. I've done a few reports on bugzilla, bugs in firefox on ArchLinux (so linux, and not a standardized setup). They've all been answered, and most have been fixed.

Sometimes, I've been told the setup I was using wasn't supported, but they'd accept a patch if I was willing to invest some effort in it. I did once, and the patch was indeed accepted.

Obviously, Mozilla cannot support every obscure platform, but then again neither does Chrome. I don't think your expectation match reality if you think that's the case.

[bigbugbag]

I get that from using bugzilla from time to time since around 2005 maybe. Last time and one of the worst was the ALSA debacle. IIRC first time was about integration in KDE.

To recap the ALSA one, mozilla pretended ALSA had shortcomings it actually does not have (some things about edge cases of netflix and 5.1 audio or something) and it turned out it was Mozilla implementation that was so bad and pretty much abandone that no one at mozilla wanted to work on it. It was taunted at people complaining that for ALSA to stay they should do it themselves or find someone to do it, until someone came forward and offered to do it and suddenly it was too late so it was not happening. The other reason invoked was that a very little number of people did not have libpulse installed which it turned out was false (because most distro disabled mozilla's spying) and irrelevant (libpulse is often pulled as a dependency on ALSA only system). So suddenly it was the fault of distro packagers and people caring about their privacy had it coming because they took the extra step of disabling telemetry (which is not supposed to be enable on release channel anyways).

I supposed those bug reports are still somewhere in bugzilla, if you wanna look for them. The KDE integration one is one of the longstanding open ones, and openSUSE provided a patch for better KDE integration at some point but it was refused for some reason. you can find it on AUR https://aur.archlinux.org/packages/firefox-kde-opensuse/

Even a simple one related to the ALSA debacle was ignored, see they forgot to mention dropping ALSA in the release notes somehow they overlooked this "detail" but they thought to include a warning message with a link to an URL that happen to be broken. A bug report was filed to report this along with the correct URL but the fix that should have taken about a minute to implement had not seen the light of day after two weeks and AFAIK did not happen at all.

I sincerely doubt KDE, debian, arch, ALSA are obscure platforms, and turns out chromium supports alsa (IINM firefox is the only linux browser with a hard dependency on pulseaudio).

To be honest with time I have learned not to expect better of Mozilla or more like to expect mozilla to not deliver on its marketing promises and to dismiss user feedback.

/edit

here's the recent iteration of the KDE integration bug, this one has been opened for 16 years: https://bugzilla.mozilla.org/show_bug.cgi?id=140751

here's the main one for dropping ALSA and making pulseaudio a hard dependency, it got locked and discussion continued on google groups somewhere: https://bugzilla.mozilla.org/show_bug.cgi?id=1247056 there are other with several duplicates such as: https://bugzilla.mozilla.org/show_bug.cgi?id=1345661

I can't find the main broken URL one with the duplicates, but it seems this one is about the same issue and had been marked as solved (Only took 3 months to fix a broken URL which was the only was to get details on why the browser suddenly stopped playing sound): https://bugzilla.mozilla.org/show_bug.cgi?id=1345439

To my knowledge there are many other examples in bugzilla in the last 10 years, maybe I've been unlucky and found myself in the worst of bugzilla making my experience not representative of the actual thing, but from what I gathered around on mozillazine, HN, distro forums and some other places my experience is not out of touch with what others have to say about theirs.

[jcranmer]

Judging from your comment, it sounds like the only bugs you filed are "YOU CHANGED THIS AND BROKE ME CHANGE THIS BACK NOW" kind of bugs, and probably for things that were consciously and intentionally done. That's not going to go anywhere because, quite frankly, they're not bugs, and most of the bug filers/commenters are unwilling to do anything constructive, like offer to maintain something.

My experience has been, rather, that Mozilla has had one of the most fantastically responsive bug filing systems. Actually getting bugs fixed in a timely manner is a different question, but that's independent of bug tracking systems.

[bigbugbag]

Well maybe you should not judge from comments then.

Reporting bugs and HN discussion is obviously not the same, at all. I know how to make bug reports and have not had bad experience like those on mozilla's bugzilla anywhere else, well maybe on gnome bugzilla and occasionally here and there because over 20 years of reporting bugs you are deemed to have the occasional bad experience.

I'm not saying your experience was bad or even close mine, I'm only talking for myself here. Maybe I have been impacted with some of the most controversial bugs and you've not, maybe I'm actually more in a niche segment of the user base than you are.

Anyways thanks for judging and misjudging my person based on assumption, it always nice to be targeted by passive aggressive online message from a random stranger.

The fun part in your comment is that constructive minded people do not expect or ask of bug reporters to offer to maintain something for they understand that reporting bugs is actually very constructive in itself, a helpful and much needed activity. Instead of antagonizing bug reporters try working with them, as "we're both in this working together to improve something for everyone". Assume good faith instead of malice and to understand that this an actual person reporting the bug and cut the patronizing, it will help. If you do not feel capable of doing that, just say nothing it will be more constructive.

[jcranmer]

The bug reports you linked elsewhere are not bug reports--note that I make a distinction between bug reports and feature implementation requests. For feature implementation requests, demanding that developers drop everything to implement features is absolutely not constructive (and yes, that does happen). Even for bug reports, shouting at increasingly shrill intervals is not constructive. I'll admit that I did it once when I was younger, and when I was finally pissed off enough to try to fix it myself... I found out that it was far more complex than I had thought.