Hacker News new | comments | show | ask | jobs | submit login
Firefox marketshare revisited (andreasgal.com)
390 points by ronjouch 12 months ago | hide | past | web | favorite | 513 comments



One additional cause of new Chrome installs taking over from Firefox: bundleware. Chrome is foisted upon users as install-by-default bundleware when users install or update lots of different apps, especially free antivirus apps on Windows. Just clicking "Continue" when your free antivirus on Windows updates will cause Chrome to be installed and set as the default browser. Here's an image of Avast tricking you into installing Chrome: http://imgur.com/hNZLbmL

I've had to fix this for three family members previously as they were using a free antivirus and couldn't figure out why their browser looked different and didn't have an ad-blocker now.


I'm actually kinda glad that Windows 10 forces the user to be very explicit when changing the default browser. Yeah, they use it to push Edge, of course. Better than programs changing it without the user knowing.


Exactly. Though I wouldn't be surprised if Chrome finds a way around it. Chrome adds itself pinned to the taskbar as it installs even though Microsoft guidelines explicitly forbid it and it is purposely a little difficult to do.


Just to clarify, did you mean that pinning something to the taskbar is a little difficult, or did I misunderstand? It is very simple to pin something to the taskbar, requiring only one right click and one left click. That's why I thought maybe I misunderstood your question. I was just curious. And I completely agree it's stupid the Chrome doesn't even respect Microsoft developer guidelines. It's normal behavior for Chrome or any other Shmoogle crapware lol!


If they found a way around it, Microsoft would patch it out, certainly. But I doubt it.


Chrome was the first app I saw that found the technical way around to force taskbar pinning. Microsoft, sadly, didn't patch that out.


I think they must have recently fixed this, in Win 10 Pro at least. I installed Win 10 Pro yesterday, then Chrome and it did not automatically pin itself. Actually, if I remember correctly it even had a message at one point during the install describing how to pin it.


It is fairly easy to disable this functionality. I have seen at least one app do this already.

https://www.google.com/amp/www.winhelponline.com/blog/edge-h...


Ironic that this comment highlights another annoying Google thing. Looks like a link to Google.com but it's an amp page.


Why is it ironic that a company that does one thing that annoys you also does another thing that annoys you?


As scummy as those installs are, it highlights Google's killer advantage: money. It generates so much cash from advertising that it can pay any company for more distribution (even Firefox!), which in turn generates more cash... Unstoppable until anti-trust kicks in.


In addition to Chrome's bundling deals which override your default browser settings, major Windows updates now appear to reset your default browser to Edge every time.


Does this actually happen? I recently installed the big Windows 10 "Creators update" and my default browser remained exactly the same (Chrome), or is this something that depends on what "tier" of Windows 10 you have?

I remember this unfortunately happening on Windows 7/8 upgrades to 10, but I've yet to encounter it on my personal machine in a Win 10 point release, at least as far as I can recall.


Oddly enough, Windows 10 Pro seems to leave my default browser alone most of the time (I do seem to recall it messing with it once, but I may be wrong), but I've had it attempt to switch my default media viewers away from VLC and Irfanview multiple times. Quite annoying.


That is because they still haven't fully adopted "new" file format association APIs (introduced with Windows 7).


Likely related to the feeling that Windows 10 resets file associations is that Windows 10 changed the way this data is stored and restores the default value if the registry keys are manipulated directly.[1] As a side-note, I don't think any of the major updates have reset my browser choice, but I have gotten an irritating pop-up suggesting that I should try Edge.

[1] https://blogs.msdn.microsoft.com/oldnewthing/20170517-00/?p=...


I know lots of settings get reset when you get new Insider builds. It annoyed me constantly. Perhaps OP is on those?


Yep. This is super-scummy behavior. No surprise Google and Microsoft engage in it!


But that's not fair, Google's already paid for that install! ;)


Ahhhh I was wondering why my desktop suddenly started opening links in Edge!


Isn't that in direct opposition to their EU antitrust settlement? How does MS get away with this?


I'm actually wondering more how Google is getting away with those screenshots from the submission. That's exactly the same behavior that it was recently fined for wrt product search.


One reason might be that EU doesn't have many browser vendors who would put in a official complaint.


Why wouldn't MS complain? If MS can be good for anything, helping keep Google in check ranks #1 on my list.


They reached a deal last year to stop complaining to regulators about each other:

https://www.recode.net/2016/4/22/11586336/microsoft-google-a...


This should be proof of illegal collusion and grounds for a gigantic fine for each of them. It's just like the no-poaching agreement they got in trouble with a few years ago.


That would require the US to have a functioning justice department.


IANAL, but I think that would count as "compelling speech" ("You MUST complain to regulators about each other!") which is explicitly forbidden in the States (not sure about Europe, but I think it's the same).


We're not talking about individuals here, we're talking about corporations. Collusion between companies for the sake of negatively affecting the market (such as establishing a cartel) is illegal in any jurisdiction with proper, functioning laws. As I mentioned before, this is similar to the no-poaching agreement which they got into trouble (not nearly enough though) before.


Lobbying and corruption ? At the very least they have friends at the right places.


That was my first thought from those screenshots as well. They are the dominant web portal for the world with more views than Facebook. Their position gives them a crazy advantage in the browser wars.

Firefox pulling ahead of IE back in the early 2000s was a sign of open source disruption taking on a monopoly:

https://xkcd.com/198/

Chrome is the total opposite; well at least initially. They pumped millions into Mozilla and then took the parts they liked from Gecko and Webkit and created Chrome. Originally closed source, even though we have the open source Chromium today, most people used the official Google branded and integrated version.

I feel like with Microsoft and IE and the anti-trust cases from back in the day, we're see a return to what the author termed web browser 'monoculture.' The author does make a point that at least Chrome isn't stuck in the past.

Chrome, FireFox and Edge all seem to be doing the rolling release thing today, which is vital to us not getting stuck in IE6 land again. (Not sure if Safari does this yet). I've recently started using Vivaldi, but I do miss my dedication to Firefox and all the years of usage, plugins and bug reports.


They probably only do it outside of EU.


Good grief.


If I remember, wasn't the version installed by Avast a custom chromium fork until relatively recently? Then there was some security issue, and google basically stop the AV's to knock it off? I think Comodo also did something like this too, though theirs was more of a mess, because Comodo.

EDIT: Found it: https://news.ycombinator.com/item?id=11057532


Sometimes. The screenshot I showed was when Avast specifically offered me Chrome. Not the Comodo Chrome knockoff. Avast offers up different things at different times depending on who is paying what. Chrome has been bundleware with a ton of software.


I recall even the standalone flash installer coming with Chrome bundled (or at least pushing for it) at one point.


Indeed, I'll never forget how my first exposure to Chrome came as a result of updating Flash and finding that it had not only installed Chrome, but set it as my default browser. That was the day that I began weaning myself off of Google services (still don't know if I'll ever make it off of Gmail, though...).


The irony is that Chrome bundles (Pepper) Flash, so a Firefox user would download the Flash installer, which downloads the NPAPI Flash plugin, Chrome, and Chrome's Pepper Flash plugin. The NPAPI Flash plugin would be downloaded but never used after the Flash installer changed the user's default browser to Chrome.


Flash is now significantly broken on Firefox under Ubuntu anyway (Bugzilla #1374559) and it's been classified as a wontfix (P5), so continuing the irony, downloading Chrome is probably closer to what the user wanted to do.


> closer to what the user wanted to do

This happened years and years ago, and it wasn't long after that I decided that I didn't want Flash installed on my computer ever again. :P Sad that Shumway didn't manage to pan out, though.


I got off Gmail back around 2013. Unfortunately it's made my e-mail even more unreliable than back when I uses Google's shitty e-mail service:

http://penguindreams.org/blog/how-google-and-microsoft-made-...


The hardest part about switching from Gmail is simply the time it takes to migrate all your accounts over to the new address.

Some providers offer a switching service though.

My three email providers don't (Kolab, Posteo, Autistici) but the few hours I spent changing my email addresses everywhere was worth it.

Gmail really doesn't offer anything you'll miss once you make the switch.


Yahoo mail is unlimited, has a sane folder structure, and in spite of the fear-mongering, is less likely to be attacked than GMail nowadays.


AVG also started doing this recently. I get a lot calls from family members and friends who aren't tech-savvy and can't differentiate between browsers. They think their "bookmarks" disappeared but the true reason is Chrome was bundled with something (e.g. AVG) and replaced Firefox as default browser. I called AVG on twitter but their answer was just typical PR BS: "We're sorry for your inconvenience. Please remember that you can opt-out from this". Yeah, they have clean hands because they put very small checkbox in very small font in installer.



While I don't agree with the technique, I would certainly not call it "tricking". Especially in your example, where it very clearly (more than most) describes what will happen if the boxes remain checked. People who blindly click "next", "ok", or "continue" buttons are tricking themselves, not being tricked. Especially when dealing with "free" software.


If the title of the page was "Would you also like to install Chrome and make it your default browser?" with a Yes and No button, then it wouldn't be a trick. Instead, the title of the page is "Complete your Avast program update" and the button is "Continue". Then, in very small letters in the lower left, it says what it's really going to do. This design is called a dark pattern and it is quite literally designed to trick you while remaining technically legal and "upfront" about what it is doing.

Yes, it is less scummy than burying it in an EULA and giving the user basically no notice, but it is still really scummy.


Does anyone know why they do this? Do they earn some extra money per each Chrome install or something like that?


Google is paying the app makers for the bundleware per install.


When other companies are doing this, we're usually pretty quick to call it by a different name: spyware, malware.

Bundleware makes it sound like it's just some innocent "bloat", while in reality this effectively hijacks all web traffic of innocent users and sends it to Google.

How is this not criminal?


As a web developer, I'm actually quite fine with this. Considering your family members probably fall under the "potential IE users" category.


All of them were using an up-to-date copy of Firefox previously with ad and tracker blocking configured with exceptions for their financial sites and with Flash configured to automatically update via Windows service as well.


It is frightening that as a web developer you lack the comprehension of why a single dominant browser is quite bad for the web. Have considered a career change, maybe growing vegetables is something you'd be good at.


I think the parent is saying that's it preferable to IE being that one single browser.

And honestly, your reply was childish and and a bit douche-bag-ish.

There's no need for this here - (I assume) we're all adults, and don't need to resort to playground tactics.

(If you actually are under 16, then I take the above and I'm sorry - welcome to HN!)


I think the "Why?" section's conclusions are off the mark. It basically blames Google's advertising of Chrome for Firefox's decline, and even goes so far as to say "Firefox’s decline is not an engineering problem."

While I don't doubt that Google's advertising of Chrome has drawn away some Firefox users, I also don't think that we can ignore or deny the many controversial changes to Firefox that have likely had an impact, too.

Just off of the top of my head I can think of things like:

* Frequent breakage of extensions when first switching to the more rapid release schedule.

* Frequent and disruptive UI changes that didn't bring users much benefit, such as Australis.

* Removing the ability to easily disable JavaScript.

* Taking many years to get multiprocess support working. (Not that I'm suggesting they should have rushed it, of course.)

* The inclusion of Pocket and Hello.

* Sponsored tiles.

* Users who report experiencing poor performance and high memory usage.

* Disruption caused by requiring signed extensions.

* The removal of support for OSes or OS releases that are moderately older, but still do have active users.

I'm sure there are others that I'm forgetting.

Even if they seem minor, those are the kinds of things that can cause users to switch away from Firefox, or not even start using it in the first place. Losing a small number of users for a variety of minor reasons can add up very quickly, as well. Furthermore, those issues don't really have anything to do with Google or Chrome.


From my experience, the reasons why people switched to Chrome have been because it renders pages much smoother and everything generally looks better. These were the original reasons that they moved over to Firefox from IE as well. I personally helped a number of relatives and friends make these switches.

Late last year, after many years on Chrome, I gave Firefox another serious look and I have switched back. Firefox has improved tremendously and I would prefer to give my support to Mozilla from a philosophical standpoint (the Chrome team does a lot of good work with regards to pushing forward the features of the web and its security but at the end of the day, Chrome is still a strategic piece of Google's business machine and not a philanthropic effort)

While I have my reasons for using Firefox, I don't see a compelling reason for most users already happy with Chrome to switch back. The average web user that I know doesn't really understand where web browsers come from and isn't very interested in learning about it. They just care whether the browser runs better or worse for the tasks that they do. (Except many still hate IE and will not even try Edge because the logo looks similar enough - that's a branding issue that Microsoft has)

What irritates me now are more and more sites that only work with Chrome (where they literally throw up a page that blocks access and says go download Chrome). These are sites that are not Google properties so I'm not blaming Google for this bad behavior, but again, I would like to support the diverse browser landscape that has existed to this point. I guess my main complaint to Google is to please stop popping up dialogs about Chrome across all of your properties. The browser I'm using works perfectly fine thank you, and you should be supporting the open web with your products anyway.


For me the compelling features of Firefox over chrome are

* better search/address bar behaviour (particularly in finding relevant bookmarks. Chrome wants to turn everything into a Google search)

* Integration with Firefox on android (which I need because it supports ad-blocking extensions)


Definitely agree with the address bar, I'll add a few:

* Being able to disable unnecessary features and phoning home using about:config is great

* Extensions look and feel more native (this will probably change because Mozilla has decided that cloning Chrome is the way to go)

* Extensions are more capable, still no decent side tabs in Chrome

* They're not an ad agency, so they don't ban extensions they don't like or nag you when you install something unapproved

* Font rendering manages to not look terrible

I'll admit though, Chrome still kills Firefox on UI speed and in many security technology ways. My biggest worry is that Mozilla will fail to achieve Chrome UI speed while ditching the things that make Firefox unique today.


It's quite hard to find any relevant setting in about:config.

Extensions used to be more capable, this is about to end. Decent side tabs in chrome is called vivaldi, (actually a decent chrome is vivaldi).

Mozilla effectively bans extensions they don't like since the made signed extensions mandatory.

I disagrre on the speed and performance difference, with 150+ tabs opened at all times firefox works while chrome struggle to deal with 50 tabs. All this on a core i5 16GB RAM SSD laptop. I guess YMMV here.


Vivaldi seems like more of the same, sends statistics information with no way to disable it. On Firefox, all of that is in about:config and easily copied around as user.js.


> Mozilla effectively bans extensions they don't like since the made signed extensions mandatory.

Not true. They sign very liberally and you can even host signed extensions for your own users exclusively without listing them on addons.mozilla.org at all.


That doesn't help, when you really needed an unsigned extension, or at least signed with your own key (and enroll your own key to firefox installation).

For example, FreeIPA used to have an extension, that configured Firefox to your own domain (enrolled an root signing certificate, configured trusted domains for GSSAPI, etc. - all the dangerous things). But because the extension was customized for your own domain, obviously, it could not be signed.

So, it was killed instead. Nowadays, you get a list of steps, you have to do by hand. On every desktop.


Getting an extension reviewed can take months. They periodically publish how many have been in the queue for over 10 days, but they otherwise dont say how long the tail is. It's long.


I started using Vivaldi after struggling with slowdowns in Firefox. Even with the new multi-processor window support enabled, my browser still frequently slowed to a crawl.

I miss some of the extensions of Firefox and Vivaldi does have some interesting bugs, however development on Vivaldi seems rapid. Recently they finally combined the web page inspector into the browser (it use to open a separate window).

I've loved Firefox for years and would honestly rather use it, but the performance problems turned me away.


I switched away from firefox once they announced they were moving to the Chrome extension model (and thereby sank their "better extensions" selling point.)


they expanded the API so it's way less limited than chrome extensions. I don't understand how that's not known by folks who read HN at this point.


Sure. But in practice everyone is just going to make extensions that are compatible with "all" the browsers, and there are developers who have simply decided not to rewrite their firefox extensions.

http://fasezero.com/lastnotice.html


doubtful, otherwise folks woudldn't have spent so much time on stuff like sqlitebrowser and other bits.


It is less limited than chrome extensions API, but still way more limited than the old API. Still no Tree Style Tabs or DownThemAll!.


There's a Tree Style Tabs-like WebExtension. It has a few quirks, though. It runs in the sidebar. It doesn't hide the top tabs (though you can do this with Stylish). It has its own menu for tabs which does not include some options (like send tab to device).

There was a very good non-nested side-tabs extension in Firefox Test Pilot, but it has expired.


Tab Center (the extension from Test Pilot that expired) is not an WebExtension. It is a classic extension that will stop working in FF56.

Yes, I'm using it ;)


Try Pale Moon, it's great.

They are keeping the current Extension model.


However long that will last. They are already considering rebasing on a newer Firefox codebase for their UXP plans.


Quoted below is from Pale Moon roadmap [0]

Long-term plans

Our long-term plans for Pale Moon involve (potentially) moving our browser to the UXP (Unified XUL Platform) that is currently being worked on alongside the browser. This will at its earliest be somewhere in 2018.

Plug-ins

Pale Moon supports NPAPI plug-ins. Unlike Firefox, we will not be deprecating or removing support for these kinds of plug-ins. This means that you will be able to continue using your media, authentication, and other plug-ins in Pale Moon like Flash, Silverlight, bank-authenticators or networking plug-ins for specific purposes.

[0]: https://www.palemoon.org/roadmap.shtml


the address bar behaviour on chrome is so obnoxious! i can never find stuff that is on my history -- with firefox, couple of keywords and i can find almost anything.


Is this also the case with vivaldi ?

I miss opera where it would also search in page content from cache, not only URL or title.


Sadly it seems that Vivaldi, for now at least, only search the url and title.


the address bar behavior is actually inferior to what opera offered back when it was a browser.


>These aren't Google properties so I'm not blaming them for this bad behavior,

No, they should very much be blamed for it! Proper web design should follow cross-platform standards and implementations. They are part of the problem if they force users to choose one or the other.


I think he means he's not blaming Google for the behavior of other website owners.


Yep, sorry poor grammar there. I've edited so hopefully it's clearer now.


Proper web design should follow cross-platform standards and implementations.

To be blunt: says who?

A lot of people and most organisations aren't making websites as a charitable exercise. They're doing it with a goal in mind, such as bringing in money directly or indirectly, or raising awareness of a cause they care about.

Whatever that end goal is, they need to use the web to communicate effectively with their visitors. If those visitors are mostly using one particular browser and they can achieve better progress towards their end goal by optimising for that browser, that is what a lot of them are going to do.

I don't think this is necessarily healthy for the long term future of the World Wide Web, but I also don't think it's reasonable to blame people with a job to do for choosing the most effective tools available to do that job.


> people with a job to do

No such thing. A job, like an order, is not something that actually exists outside of the actions of people. Both the people giving orders and the ones following them remain responsible. They can pretend to leave the court room by dozing off but they remain in it for those who haven't, and what you seem to see as "putting blame" is simply pointing out what is already present and cannot be removed.


This is a very short sighted decision, apart from technical demo, designing a website to only work with a specific browser is actually costly, doesn't last and significantly limits your reach.


It's 2017, the era of evergreen browsers and living standards. As much as you or I may wish otherwise, anything built for the web today may not last and may have significantly limited reach within a matter of months. Every major browser developer has shown a willingness to cut off established functionality, even things that have worked for many years, if its suits their purposes. (Either that or they've abandoned older, non-evergreen browsers more or less completely.) The only standards that matter in web development today are the de facto standards of what the browsers actually do, just like the bad old days of IE vs. Netscape.

As I said before, this might well be bad for the long term future of the World Wide Web as a resource for society, but in this business you have to play the hand you're dealt, and the browser developers hold all the important cards.


Somehow it seems you live in a quite different world than I do or maybe I didn't understand what you meant here.

Those still work in every browsers I've tried so the major browsers developer have to cut more established functionality: https://justinjackson.ca/words.html http://motherfuckingwebsite.com/


Firefox is the inferior product.

I use firefox because I want a browser to exist that isn't hellbent on knowing exactly who I am in order to maximize profits.


Are you aware that mozilla actually ran on money from google out of using firefox to tell them exactly who you are in order to get a share of the maximized profits ?


I downvoted you because what you said is extremely misleading. Mozilla had a contract with Google to use them as the default search engine. They didn't "tell them exactly who you are..."


Are you trying to tell me that people at mozilla are not aware that google can tell who you are with an impressive degree of precision based on your usage of google search engine ?

Mozilla is not telling google obviously because they don't know who you are. But it's a technicality as they enable and empoyer google to do it by themselves, in exchange for millions of dollars. For a few years over 95% of mozilla revenue in millions of dollars came from google for exposing their user privacy while most users were not aware of this and mozilla boasting being a white knight for privacy.

Then when they finally decided to do something about it, the chose to help yahoo artificially inflate their usage stats to improve the value in the upcoming sale, but only in the US where they had lost the most marketshare, in Europe where firefox was still relevant and where there are alternatives that actually respect user privacy mozilla chose to keep google as the default search engine.


While technically true, using Google as your default search engine goes a long way to that goal.


Chrome uses Google as default search too though.


I wasn't aware that Mozilla shares every single web page you visit with Google. Do you have a source?


AFAIK they don't. Firefox phones home to mozilla and according to this comment has about 20 default settings phoning home to google[1].

But my point was simply about directing users' searchs to google search engine, which is enough to expose yourself to google.

[1]: https://news.ycombinator.com/item?id=14809837


Of course they don't; it's not true.


>What irritates me now are more and more sites that only work with Chrome (where they literally throw up a page that blocks access and says go download Chrome).

Any example of this?. Because I've never visited such a site.


Inbox during the Beta.

WhatsApp during the Beta.

Signal's desktop client.

Google Earth.

Google Translate works, but has 3 helpful dialogs telling you to install Chrome.

Google Search on Mobile Firefox is heavily restricted in functionality, and tells to install Chrome.

Under some situations, Google will replace the first 4 search results and replace them with a "your browser is outdated, install Chrome" while also adding a top bar saying the same, and a dialog at the top right.


Chrome font rendering on Windows is still broken. Fonts that look find everywhere else are different in Chrome and usually look worse. Why?!


I don't know the exact reason why, but it probably has to do with the fact that Edge and Firefox both use Direct2D for rendering and DirectWrite for text, while Chrome uses Skia (which can call into DirectWrite or GDI for text rendering I believe, but may not do so in the exact same way).


Skia is now the default on firefox/windows when direct2d isn't available.

http://techdows.com/2016/10/firefox-52-skia-enabled-by-defau...

What does that actually means in practice? I'm guessing the only time you actually get skia is when running in a pretty old hypervisor environment or on a server/BMC without hardware acceleration.

Which IMHO, having your application _LOOK_ different depending on hardware acceleration capabilities is sorta stupid.


> Which IMHO, having your application _LOOK_ different depending on hardware acceleration capabilities is sorta stupid.

Which is one of the reasons I transitioned from liking Web back into loving native development (there are plenty of other reasons though).

Doing discussions with customers about pixel differences across browsers, specially if it is the same browser on different OSes, is anything but fun.

That and the request for features and behaviors only possible in native UIs.


Should file a bug, though.


direct2d ? is that a thing on linux ?


No. Firefox uses skia on both Linux and macOS.


Linux is arguably worse. For whatever reason, Chrome absolutely refuses to display some unicode fonts (that are installed and render fine in Firefox). CJK and other Asian fonts seem to be worst in this regard.


> more sites that only work with Chrome (where they literally throw up a page that blocks access and says go download Chrome).

Any specific examples of this? A URL, or a couple of them?

When making such an assertion, it would be nice to minimally provide a way for others to see for themselves.


Only time I've encountered something like this was OK GO's All Is Not Lost music video released in 2011, it came with a warning that you could skip but then the performance were so bad you could not enjoy it.

it was available at http://www.allisnotlo.st/ but now features an error message about google dropping python 2.5


For me, the deal breaker for Firefox was Youtube full-screen playback performance. On my old Core 2 Duo E6600 machine, Youtube stuttered playing 1080p videos while it didn't on Chrome.



I'd be surprised if any of those, up to the last three, was a big deal for more than 1% or so of users. And I am inclined to think the original article probably has it right about simple saturation marketing as the cause of most user switching.

But performance. Firefox very often outperforms Chrome in microbenchmarks and computationally-intensive code in my tests, but in the real world an awful lot of sites really are much more responsive in Chrome.

For me as a user, most recently an update to the FastMail web UI a couple of weeks ago made it lamentably slow in Firefox -- just mousing over the folder tree caused CPU spikes and lag in updating -- and in the end I switched to opening FastMail in a separate Chrome instance while continuing to use Firefox for everything else. I've just switched it back to Firefox as I type this, to see whether anything has improved.

The web app I'm working on as a developer just now also has problems updating as smoothly in Firefox as in Chrome, and I'm not at all sure whether we'll be able to do anything about it.

I can't think of an example at the moment of a site that feels faster in Firefox.

I believe I have come to think of Firefox as a web browser, and Chrome as a platform for web apps. Things written to be web apps are almost always more responsive in Chrome, even though many of their components (number-crunching work) really do run measurably quicker in Firefox.


I use Firefox (Linux) and Fastmail, and definitely don't have any problems: the web UI is just as fast as always and have no CPU issues. Maybe it's some addon or something?


Can definitely be due to addons. That innocuous Emoji keyboard addon [1] for example, caused horrible page load performance and unresponsiveness (due apparently to parsing the entire DOM in order to replace/insert emojis); filed a bug at [2]. Hope the move to WebExtensions lets Mozilla provide less opportunities for extension developers to shoot users in the foot.

[1] https://addons.mozilla.org/en-US/firefox/addon/emoji-keyboar...

[2] https://bugzilla.mozilla.org/show_bug.cgi?id=1339822


>hope the move to WebExtensions lets Mozilla provide less opportunities for extension developers to shoot users in the foot.

If there are firefox users left to shoot.


I think you're right about this one. Sorry about that. It seems to have started interacting badly with Ghostery -- which I had been using for years, but which did have an update recently as well. I tried a different ad-blocker and it's been a lot quicker since.

I think I jumped to conclusions because the change roughly coincided with a visible update to FastMail (the one that added the little progress spinners to the folder tree).


I'll add, for the record, that simply trusting the site in Ghostery appears to be enough to speed it up again.

(My earlier post lost a couple of HN points after I posted that reply, even though it was many hours after the original discussion. I wonder whether people had upvoted it because of potentially useful material about Firefox and FastMail and then unvoted when I admitted that particular bit was bogus. It would be rather encouraging if that were the case)

Later edit: I spoke too soon. After editing an email extensively, the whole UI slows down again.


Yeah me too - and even the last three(perf/mem, extension breakage, and OS support) are partially questionable - FF has essentially always been better behaved in low-memory scenarios that chrome; and desktop OS support is (again, to this day) better that chrome's. Even XP still has FF 52 ESR support up to june 2018, and v52 isn't that old yet.


It seems to me like Firefox has a lot of tiny delays everywhere, and they're just long enough to be distracting. It's like going back to a hard disk after having used an SSD.

On the other hand, Firefox Focus on mobile seems to run pretty fast and comes with enough ad blocking to make the web bearable.

If it had either tabs or a way to open multiple processes then I would probably ditch the other mobile browsers.


I love Firefox Focus, but it's not suitable for your main mobile browser. What it is is a substitute for Chrome Custom Tabs.

Full Firefox on Android has a slow startup time if it's been pushed out of memory, which it usually is, because it's big. Firefox Focus is a way of quickly opening links from other apps without paying the cost of opening Firefox, while keeping an acceptable level of ad-blocking and privacy.

You should have both installed, set Focus as your default browser, but liberally use the "Open in Firefox" menu on Focus when you have real browsing to do.


I switched to Chrome in ~2011 or 2012 after being a Firefox user and evangelist since... was the first one Phoenix or Firebird? Anyway, that one.

Reasons:

- Firefox got bogged down with just a few tabs open, and caused beachballs (OSX/macOS) systemwide. Chrome was snappier and didn't harm my system's overall responsiveness with several times as many tabs open. This was the main reason.

- Dev tools. Liked Chrome's better.

- Profile handling was, at the time anyway, better.

- IIRC Firefox didn't do per-tab crashing at the time, while Chrome did, which aided overall stability.

Advertising had nothing to do with it. Chrome was just way, way better, especially its (apparent, which is mostly what matters) resource footprint.

Now I'm mostly on Safari, even though it's the worst mainstream browser, just because I gain 1-2hrs of battery life using it over Chrome or Firefox.


> Dev tools. Liked Chrome's better.

They definitely are. To this day I'm baffled I'm not sent to the debugger when clicking a line reference in the console.

But I still use Firefox as main browser. Since the pages I visit and the pages I develop are always in different places it's pretty easy to have one browser for development only.


For a while, Firefox was crashing Linux as well, probably from some accelerated graphics bug. (It used to happen sometimes when ArsTechnica would play an ad, for example.)


Whichever version of Ubuntu introduced pulseaudio crashed any time Flash was used on a website, often in ads. May have been that. I mean, Pulseaudio crashed xwindow (or just stopped working) all the time in that version, but Flash was the easiest way to kill it.


I switched to Chrome around that time too, for many of the same reasons (mainly performance and stability).

I've since switched back because Chrome is a terrible memory hog and I can have tons of tabs open on Firefox with no impact on performance (as long as I don't actually load them), and I don't have problems with crashing the way I used to.


Chrome recently introduced some changes to background tabs (to a bit of grumbling from sites that wanted to use background resources/service workers):

https://developers.google.com/web/updates/2017/03/background...

And also - they're aggressively throttling background tabs:

http://blog.strml.net/2017/01/chrome-56-now-aggressively-thr... https://news.ycombinator.com/item?id=13471543

I'm on Chrome Canary - and there's been noticeable improvements in memory/responsive wise for a few months now.

I have around 320+ tabs open, spread over 2 Chrome profiles (around 160 per profile).


Yahoo as the default search engine didn't help, either.

Firefox is about to shoot itself in the foot again. Soon, all old add-ons will stop working, as Firefox tries to get add-on developers to change to their new WebExtensions API. (Which is almost, but not quite, compatible with Google's add-on format.) Many developers are not bothering, and will drop Firefox.


Yep, I'll be limping along on the "old" version of Firefox so that I can use Tree Style Tabs, which can't be made compatible with the new extension framework. And my startup may or may not update our own Firefox extension this year. We're waiting to see how things shake out — how many people update to the latest version of Firefox, or whether folks are tied to the old version because they have so many legacy extensions.


https://addons.mozilla.org/en-US/firefox/addon/tree-tabs/ use the new extension api AFAIK. Seems you have to go through some hoops to hide the regular tabbar though.


Unfortunately it draws its own context menu which looks very ugly, doesn't behave native and interferes with dom.event.contextmenu.enabled=false.


Well if you decide to stick with an older version of Firefox, you might just as well stick with PaleMoon instead. At least it's still getting some new features and security fixes.


Would love more details!



I suggest that you switch over to development for Pale Moon. There are lots of positives behind it, and all that I've heard is that they have no plan at the moment to switch over to WebExtensions.


I second switching to Pale Moon.

Someone has forked Tree Style Tabs so that it works with latest Pale Moon.


The testpilot program from mozilla has a beta test for sidebar tabs. This is still not tree, but can be a decent alternative if you need tree style tabs to reclaim screen real-estate.


The sidebar tabs testpilot extension is also not a WebExtension, so it will stop working at the same time as TreeStyleTabs.

Tab Center Redux <https://addons.mozilla.org/en-US/firefox/addon/tab-center-re... is a fork of the testpilot extension that is a WebExtension, so it works fine in Firefox Nightly. It's naturally more limited than the original extension, though: it can't hide the top tab-bar, it has to draw its own context-menu, it can't be shown at the same time as any other sidebar, etc.

It's still nicer than the top tab-bar for those of us with a bunch of tabs open, though.


Tab Center (the extension for sidebar tabs from testpilot) is already down, with no replacement. For existing users, it will stop working with FF56.


Yup, I fear for the future of tree style tabs. So much so, I am developing an alternative browser that embeds Chromium but has a tree-style tab interface from Qt [0] (disclaimer, nowhere near ready yet).

0 - https://github.com/cretz/doogie


Looks interesting!


It was the only choice.

Set your User-Agent to Firefox or IE Edge and Windows OS. You'll soon see "install chrome" pop ups/banners/warnings that take up a portion of the screen all over Google properties.

At that point why not just self-uninstall?

Out of non-Google search engines, Yahoo makes most sense even if they got no money from the deal. Maybe DuckDuckGo but unfortunately it's still not as good.

Regarding extensions, it's better this way because nobody is bothering with current API. Most new extensions are chrome-only.


yahoo as the default search engine is the same as google as the default search engine. It is a matter of money and selling to the highest bidder, though the move away from google was probably motivated by chrome marketshare.

This is a US only move, in the EU it's still google and mozilla got a truckload of criticism for this.


They do need to make money in some way, ever since Google didn't renew their contract with Mozilla (which iirc was 98% of their revenue back in the day) after Chrome started taking off.


Mozilla actually moved to Yahoo not because Google refused to renew their contract, but because Yahoo was flush with cash and desperate to gain any foothold at all in search engine share (which they did, briefly; a friend of mine at Google had the job of getting Firefox users to switch their default search back to Google). Yahoo just offered better terms than Google at the time.


It's mozilla who didn't renew with google in the US, because yahoo needed to appear better for negotiating their price tag and made a better offer.


Honestly, as a firefox die-hard who finally gave up, all of those issues were dwarfed by the performance one. The only one I even bothered to config away was the search-engine change.

I stopped using firefox because of performance. Nothing more, nothing less.


It was the extension signing that caused me to move on. I had written several small, but useful, extensions for my own personal use. I knew they were harmless, yet Firefox made it difficult for me to actually use them.

If I'm remembering this right, I think there was initially an about:config option for disabling the signature checks. But that was eventually removed from the stable releases. The workarounds were to waste my time getting the extensions signed, or to use some special unbranded build, or to use the Nightly or Developer Edition releases. None of those were acceptable to me. Then I learned about the planned WebExtensions changes, and knew it was time to move on.

I'm aware of the security-related reasons that were used to justify such changes. But for me they ended up taking away the main benefits that Firefox offered, namely being easy to extend, and giving me the freedom to use the browser as I see fit.


Firefox developer edition allows using unsigned webextensions by toggling xpinstall.signatures.required to false in about:config


waterfox allows unsigned extensions.

It seems to me disabling the xpinstall.signatures.required setting has no effect anymore, at least the last time I tried using it, it had no effect.


Firefox developer edition is the former Aurora (development) channel. It is not the stable release channel.


You could use the unbranded builds. They build them specifically for your use case. https://wiki.mozilla.org/Add-ons/Extension_Signing#Unbranded...


FWIW, developer edition is now the Beta channel.


It's not labeled as stable, that does not mean it is unstable.


IMO it is really shortsighted to treat your best users and evangelists so poorly. This is the absolute creme de la creme of your user base- they should be pandered to, not locked out.


Mozilla seems to have gotten infected by the same paternalism that is riding the likes of Gnome into the ground.


Something I could never quite understand. I'm using Firefox but occasionally run Chrome for a few minutes. I do have a couple of extensions installed in Firefox. To me, Chrome might be faster, or maybe not. But honestly, I couldn't care less: even if Firefox takes a second, or two or even three more to show me a page sometimes, so what? I mean, two seconds? I guess I can wait that long, even if I look at tens of websites each day (which I'm not even sure I do).

If somebody gave me a Ferrari for free with the caveat that there's a guy sitting on the passengers seat who keeps track of where I'm going at all times, I guess I'd still keep driving my current car (hint: it's not a Ferrari).

And before the downvote reflex sets in in some of you: I'm not saying that you should be like me. If you like Chrome, great, good for you! It's just that the speed difference to me personally has never been a good enough reason to switch. YMMV.


Are you saying that you only look at 10s of websites or 10s of pages? As a (frontend or full-stack) web developer you end up looking at many 100s of page views a day, and those extra seconds matter. Both in maintaining a semblance of flow, and in real cumulative time. I switched to Chrome the very first time I tried it, very soon after it was released, and never looked back because it made me happier and more productive.


That's exactly why I wrote "YMMV" and put a big focus on stressing that this is how things work for me. I'm not a web developer and so my usage pattern is certainly much different from yours (if you are one).

I'm glad to hear that you found a web browser that makes you happy. So have I.


The problem is that when you get more and more tabs going, Firefox's single-threadedness becomes more and more painful. When one misbehaving tab locks up (or crashes) the whole browser, that's bad.


This has been fixed. Firefox is now fully multi-process on all release channels.


Unless you have an add on that isn't compatible, like the one that ubuntu for some reason bundles with the browser out of the box.


It looks like the development of that extension stopped in 2014 but it's still bundled with Firefox in Ubuntu https://packages.ubuntu.com/search?keywords=xul-ext-ubufox

That could explain why it doesn't support multiprocessing. I disabled it because it doesn't do much. This is a list of its functionality from https://ubuntu-mate.community/t/what-are-the-ubuntu-firefox-...

* Enable searching for missing plugins from Ubuntu software catalog

* Add the following options to the Help menu

Get help on-line

Help translating Firefox

Ubuntu Release Notes

* Set homepage to Ubuntu Start Page

* Display a restart notification after upgrading Firefox

* Add ask.com to the search engines. You can uninstall this if you prefer to use a pristine Firefox install.


I hate it when people who know how to disable them don't do it and complain about it. It is really Ubuntu's fault for bundling an addon that is not compatible with multiprocesses.


I have lots of tabs open all the time, it's not really an issue for me most of the time. But it's true that every now and then it happens that something goes wrong and CPU consumption in Firefox stays way up. (I blame plugins/extensions though.) In such a case, I don't mind killing the process and restarting Firefox to remedy that.


It is similar to Chromes memory usage, add one tab after another and you will hit memory wall pretty fast (100 tabs?), with firefox 1000 tabs is not a problem.

That's why I switched back from Chrome after using it for a month few years ago.

Now I would switch because Chrome is the new IE, some developers don't test on Firefox, they say "just use Chrome", no WAY.


Firefox is now a multiprocess browser.


Firefox is multiprocess now.

I find it uses probably 50% of the memory that Chrome does in my typical use-cases as well (4-5 windows open with around 10-20 tabs in each).


Why run Chrome at all then? What's wrong with Opera, Iridium or even Chromium?

I think majority of people just use Chrome because they either don't know better or because when it came out, it was legitimately a cooler more innovative browser than Firefox at the time.

Only very few people need a few extra codecs that Chrome provides or maybe the bundled Flash plugin.


I just tried on two popular websites, espn.com and cnn.com, and Firefox was slower to first paint on both, noticeably. No extensions installed on Firefox but a few on Chrome. It's slower.


Maybe I wasn't quite clear with what I've been trying to say: even when Firefox is slower, it does not matter to me.

If it took a minute to render, it would matter. If it's a matter of a few seconds, it doesn't.


That's fine, but I wonder if that carries over to most people. There have been plenty of studies that suggest perceivable slowness has a large effect on user engagement. Amazon famously did a study on their website that showed 100ms of latency cost them 1% of sales.


I wouldn't be surprised if there's a large percentage of people to whom rendering speed matters. Of course, you'd have to have some frame of reference: if you've never experienced Chrome's faster rendering times, you might not think of Firefox as being particularly slow.

However, I also know that folks like my parents who do not deeply care about IT and performance in general don't really care too much. They do not spend their day in front of the screen like some of us do, but rather look something up once or twice a day. In the greater scheme of things, the difference in rendering times across different browsers doesn't make a measurable difference in their lives.


IIRC there is no significant rendering times differences, there is a perceived difference though. Too bad I don't have the URL of the test but it was featured on HN a little while ago.


Firefox was noticeably faster for me on first paint when I just tried both of those. Caching effects? CDN latency? Some part of the network? Nondeterministic browser behaviors? who knows.

This gut-check test isn't particularly useful.


The point of the OP was

> If somebody gave me a Ferrari for free with the caveat that there's a guy sitting on the passengers seat who keeps track of where I'm going at all times

The price you pay is much higher than you think.

Think twice, you have a choice.


I did both in incognito fwiw, it prevent the caching effect. It's not a thorough study, and I'm not concluding that it is, but i'm not doing a study when picking a browser. I just use it and stuff is slow, that's all I have time to try.


Such anecdotal evidence is invalid. Try the scientific method instead.


> there's a guy sitting on the passengers seat who keeps track of where I'm going at all times

You seem to be implying that Chrome tracks every site that you go to and shares it with Google. Care to cite a source on that?


It is called an "End User License Agreement", most users check the box without reading it. To give Goog some credit, they do tend to state their intent plainly(not legal jargon) in the first couple paragraphs. Of course, sometimes they have to edit/whitewash out the "creepy" aspects.

Edit: link

https://www.pcmag.com/article2/0,2817,2329414,00.asp


Perhaps this (under 'Browser History')?

https://support.google.com/websearch/answer/54068

Of course, this can be turned off, but it is very likely to be turned on if a user has set up an Android device (lots of prompts to do so).


This was the big part of why I dumped firefox. Others included pending depreciation of addon apis, screwy behavior with dark gtk themes and unreadable text, issues where video would straight on skip or freeze max headroom style, yahoo partnership, addon signing requirement, failure to add synced reading lists to desktop browser in favor of pushing pocket.

In short screw you too mozilla.


Same here. I read a bit about multi process having arrived, so I'm planning to give ff another try soon. But at least all the electrolysis pre release versions I tried ended up disappointing. It's a real shame, I'd love to switch away from chrome.


have you tried vivaldi ?


It's crowded, there are way too many things I don't need. I prefer getting the flavor of feature I need by choosing an extension.


Firefox nowadays feels every bit as fast as Chrome, if not faster.


Not for me and definitely not on Linux, see https://bugzilla.mozilla.org/show_bug.cgi?id=1280523 for example.


Just run it in Wine (currently may not work).


Most people will just use Chrome instead.


Haven't tried it on Linux, just Windows.


The problem Firefox had for a long time was the terrible Flash plugin combined with a lack of per-tab processes. That meant that a single flash element on one page would slow down the entire browser. I removed the flash plugin completely a couple of years ago and havent looked back.


As a fellow Firefox user (writing this on Firefox, btw), I've heard people complain about performance issues all the time with Firefox vs. Chrome... and I've never understood the problem. Firefox with ~20 addons starts up like a charm for me, loads every page I need in moments, and has no problem doing anything I need it to do. Chrome, on the other hand, gums up my machine to the point that just opening a Chromium-based browser usually ends with my machine frozen like an iceberg, me pulling the battery, booting into safe-mode, and uninstalling the product completely.


Same here. I like the idea of Firefox. Every few months I reinstall it and try it for a day or two, and it's always just painfully slow compared to Chrome. In most other respects I actually prefer Firefox, but the other differences are just much less important to me.

Ohh, and the Chrome dev tools are just better. So that helps. But if Firefox weren't noticeably slow I'd use it without hesitation.


If you're willing to spend a bit of time on this, I'd really appreciate a list of at least some of the things that feel slow compared to Chrome when you try to use a current Firefox nightly. It's a lot easier to fix things we know about than things we don't know about. ;)


It must be something subjective or system-dependant then because I switched from Chrome to Firefox for that same reason.


It is totally subjective. Except for the cases where it's accumulated cruft and dubious extensions (adblock plus for example).


Performance is really good with the new multi threaded implementation.


You have forgotten a few, such as dropping alsa and making pulseaudio a hard dependency blaming alsa when actually it was their own implementation that was falling behind and they did not want to fix it.

But the issue here is that mozilla do not listen to user feedback and just push whatever they feel like pushing with an attitude and some hostility towards unhappy users. or as pointed out in the mozillazine forums: they're " making far-reaching and very short-sighted decisions in a vacuum. "

It seems inside mozilla they're convinced that firefox is great and answers users' needs, while users feels that firefox is not that good and getting worse. There's quite a gap between mozilla marketing and the reality, which shows that firefox fails to deliver on its promises.


I'm a Firefox user and quit Chrome for the same reason I stopped using Windows after Windows 7: to keep a bare minimum of privacy when using the internet. Here are a few additions to that list: * Firefox Extensions now must be signed, and there is no workaround, not even one for users willing to delve deeply into settings x,y,z. The rationale for making it difficult to add unsigned extensions is sound, but there is none for making it impossible. * Firefox claims to be privacy-friendly, but they use an identifier when calling home that, along with other data, uniquely identifies the user. Firefox claims that this communication is encrypted, but this is not enough for state actors, and likely not for others. * There are about 20 default settings in Firefox that call home to Google behind the scenes. * All of these settings are incredibly confusing and there is no standard documentation on how any of it works, so even for a technically inclined person it is near-impossible to simply use the browser without spewing your data all over the place- and that's _before_ actually calling up e.g. the NY Times and its 50 trackers, data-miners, beacons, ad networks, social networks, etc. * The shameful treatment of Brendan Eich (creator of JavaScript and co-founder(!)) still rankles.

All that said, it's the least-worst in the browser world, for now. It seems pretty clear that they have some technically brilliant people as well.

Edit: seems that I need to learn how to write lists...


> seems that I need to learn how to write lists...

insert an empty line between each bullet point.

First without an empty line:

* bullet point without empty line between each other * bullet point without empty line between each other

then with an empty line inserted in between:

* bullet point with an empty line between each other

* bullet point with an empty line between each other


Yup, FF has just missed the mark on so many points that Chrome just blows past them. Any JS-heavy web applications (YouTube, HBO Go, Netflix) run like garbage in FF. Switch over to Chrome and they are buttery-smooth. Sucks to say this too as someone who's used FF since its beta.


For me Chrome and Firefox feel the same performance-wise. Are you sure you don't have some add-ons or other changes in your Firefox profile?


Try again from a brand new user account, chances are this issue will go away.


Performance is a big one for me. The Firefox UI feels slower, and every so often it hangs, which is frustrating to no end. When I used Chrome I took speed for granted - something I no longer can with Firefox.


Thank you! As soon as Pocket was "included" and the response to everyone complaining was a politely worded "fuck you" then I removed FF and haven't looked back.

It's understandable they have to find ways to make money, but those experiments alienated users. Once you've started down that path, there's no returning in a lot of users eyes, mine included.


So true. The Mozilla corp lost touch with their users. All early Netscape employees left, now it's official a non-profit but acts like big corp.

Firefox become unusable - shut down Firebug replaced by half-assed new DevTools, removed XUL based API, multi-process support that's still not working like Chrome1+/IE8+, still dog slow, can't handle more than a few tabs, Addons-website got useless as most addons aren't working anymore.

Sad, but Chrome is so much better, and Firefox is digging in a bigger rabbit hole with every new release. Would be great if we keep another open source competing browser around. Servo based browser could be a fresh start, but they need to focus now, in a year it can be too late.


People switch because Chrome is a superior product, just like Internet Explorer was a superior product to Netscape at one time. That's not a concern. The concern is using market position to engage in anti-competitive behavior, like bundling.


People begin to use more IE, becasue lazines and that IE was bundled with Windows. Netscape always was better that IE.


Netscape always was better that IE.

No, it wasn't. Were you there when NS4 came out?


Depends what sites you used.

Compared NN4 and IE4, NN4 was better. Compared to later IEs, IE were better.

Except for IE5 for Mac. That one was weird, it's CSS implementation was best from all the available browsers. But when you opened Slashdot (or another table heavy site), boy, you quickly switched back to Netscape.


Netscape 4 versus IE5 or 6 was pretty bad. They got stalled a few years while rewriting their browser. Firefox was better than IE almost from the very first alpha tests.


Netscape 4 was facing IE4 not IE5 or 6.

firefox was initially named phoenix and it was made as a workaround the performances issue of mozilla suite by removing everything unneeded to run a browser. Except mozilla suite fixed its performance issues way before phoenix was ready for prime time.


True about IE6, but IE5 was already out before Netscape 6 came out. Funny that the real 1.0 for Firefox came years later.


Netscape 6 was the "new one", with Gecko, XUL and stuff.

Firefox itself used to be an alternative - the main browser was Seamonkey, that included mail client, chat client and kitchen sink, just like the original Netscape did. Firefox started later as a lightweight alternative, just the browser.


This is so wrong. People used I.E. because it came bundled with their computer. Source: I was there.


You may have, but that's not why people who used and were fans of Netscape switched to IE voluntarily. While IE was adding CSS support and other consumer friendly features, Netscape was more interested in Communicator, the enterprise suite.


I don't remember fans of netscape switching to IE, though I do remember AOL buying netscape and making a mess of it with netscape 4 development stopped and netscape 5 abandoned.


AOL buying Netscape happened much later, after the war was over.


You're probably right, by the time AOL bought netscape it was a background news to me as I had moved from netscape for a while.


Ah the Mariner debacle.


Those are tech-y reasons that might have repelled tech-y users like us on Hacker News, but the bulk of the market are laymen who don't know the difference between a browser and a search engine on whom marketing is very efficient.


Yea, I've always been a huge fan of (and donor to) Mozilla, but in the last few years, I've been losing some respect for them. Getting embroiled in absurd political snafus, head-in-the-sand analysis of their situation which ignores product quality, etc.

I switched to Chrome a lot later than most of my friends, and actively try to switch to Firefox every once in a while, for literally the last 7 or 8 years. There has yet to be a single time where a couple days of usage didn't reveal the browser as far inferior, in ways that affect my day-to-day life materailly (multiprocess support being the biggest, most basic issue for a long time). I spend a LOT of my time in the browser, and I'm a very heavy user (usually about 100 Chrome tabs open total at any given time, with fairly high turnover). The performance and quality penalty I pay when using Firefox just isn't worth it.

I don't disagree with the article's claim that Google's advertising is having a big effect on a drop in FF usage; it's just bizarre for them to act like this is the only possible reason why people are switching.


> * Frequent breakage of extensions when first switching to the more rapid release schedule.

That's why I switched. It broke most of my extensions every time it upgraded. After the 3rd or 4th time it wasn't worth dealing with anymore. I switched to Chromium and I don't remember it breaking an extension.


I just skipped upgrading firefox, it's broken under debian anyway.


This is my list, all of which Chrome does better than Firefox (roughly related bug reports included):

* Restore the old settings. They copied Chrome's settings-as-a-tab with the UI just being HTML. But in Chrome I can at least search the settings. Why did Mozilla waste their time on copying the HTML-settings without also implemented the most useful feature? It was just a huge regression, because the UI is now non-native, many things aren't resizeable anymore and some other minor bugs where introduced, without any apparent benefit. https://bugzilla.mozilla.org/show_bug.cgi?id=1325286

* When you start Firefox two times in a row, the dialog "Firefox is already running, please close the running instance" or something like this pops up. Chrome doesn't have this problem, maybe just because its startup time is SO much better. https://bugzilla.mozilla.org/show_bug.cgi?id=489981

* On Linux: Integrate the tabs into the titlebar like Chrome does. https://bugzilla.mozilla.org/show_bug.cgi?id=513159

* Way too easy to quit the whole browser with Ctrl+Q (Chrome uses Ctrl+Shift+Q) https://bugzilla.mozilla.org/show_bug.cgi?id=52821

* Encrypt passwords with the keyring (like Chrome does) https://bugzilla.mozilla.org/show_bug.cgi?id=309807 (btw: that's the second most voted bug of the "Toolkit" product according to https://bugzilla.mozilla.org/page.cgi?id=productdashboard.ht... )

* No hardware acceleration on Linux (playing HD YouTube videos lags for me in Firefox out-of-the-box, perfectly fine in Chrome) https://bugzilla.mozilla.org/show_bug.cgi?id=1280523

* Speed and responsiveness of the UI in general are much better in Chrome. (no bug report link, sorry)

Notice how that there's no complain about the look of the main UI, still Mozilla decides to redo it yet again with project Photon ... Did I miss the bug report with lots of votes for that?

And regarding the bug reports (most of them reported years ago): There was a comment on Reddit a while ago where a GNOME (!) developer said something along the lines "We're not Mozilla, we're actually reading and answering our bug reports". That says something.


> When you start Firefox two times in a row, the dialog "Firefox is already running, please close the running instance" or something like this pops up.

This is actually due to shutdown (rather than startup) being too slow. Your profile is still in use from the instance of Firefox taking too long to shut down, so when you start a new instance it hits this error. This should be a little better with multiprocess, because web pages are run in a separate process, and we kill that process more quickly, so shutdown should be faster.


Also happens on startup for me and others. See for example this comment from the bug report:

I simply double-clicked the Firefox icon twice quicker than I ever normally would, and the Close Firefox error appears: "Firefox is already running, but is not responding. To open a new window, you must first close the existing Firefox process, or restart your system."


Indeed. This in particular when launching Firefox in response to opening urls from external sources.

Hit two of those quickly and hello dialog...


> When you start Firefox two times in a row, the dialog "Firefox is already running, please close the running instance" or something like this pops up.

On Windows, this only happens if the second instance starts with a specific command-line option (the name escapes me at the moment). Otherwise, the existing instance just opens a new window.


> Speed and responsiveness of the UI in general are much better in Chrome. (no bug report link, sorry)

> Notice how that there's no complain about the look of the main UI, still Mozilla decides to redo it yet again with project Photon ...

Are you seriously askong that?

Firefox is doing Photon (like they were doing Australis) because they're replacing their entire UI framework with a faster one.

Australis was the move from native GTK2 to XUL, Photon is a move from XUL to HTML5 for UI.


Firefox has never had a "native GTK2" UI. Australis was not a move from such a UI.

Photon is not a wholesale move away from XUL either.


While I do agree with a lot of the points:

> * Way too easy to quit the whole browser with Ctrl+Q (Chrome uses Ctrl+Shift+Q)

Saying that it's nitpicky to include this in your list would be huge understatement, it's straight out ridiculous. I have always found Firefox to be more responsive and less resource heavy than Chrome, so I don't know why you had problems with that.

But yes, you are right when you imply that Firefox seems to have prioritization problems, lots of them imo. However, it is understandable to me, making the UI looking prettier is for marketing, not usability. Most of these things you listed are not addressing a lot of users, on the other hand, having a flashier UI would address and (potentially) attract more users. But their management still needs to improve, and as a company, they should have better direction.


> Saying that it's nitpicky to include this in your list would be huge understatement, it's straight out ridiculous.

There's a bug report about it with lots of duplicates and 71 votes. I use Ctrl+W to close tabs, so losing work in other tabs is just one key away. For me, it isn't nitpicky.

> having a flashier UI would address and (potentially) attract more users.

You don't know that though. You think it will attract more users.

You'll need some kind of metric to know what people really want. "Gets often repeated in discussions" is one, "which bug reports get voted on" is another.


I agree although ctrl+shift+q is still suboptimal because ctrl+shift+tab is right next to it.


I don't get why closing the whole window need a shortcut at all.. Doesn't most OSes have a generic hotkey for that?


Edit: I see now that jhasse answered most of this already but I think mine adds some details so I'll leave it here.

I think alt-f4 closes the window while ctrl-q closes down Firefox entirely (in my KDE setup at least.)

On Windows ctrl-q doesn't work for me though.

As for why I sometimes use it it is because I can then do a restore session after restarting Firefox and get back everything.

(On Windows I have to find the now hidden menu and select Quit Firefox or something like that.)


It actually closes all Firefox windows. The generic hotkey for closing one window would be Alt+F4.

I definitely think that this should be handled by the OS though.


Nah, "Disable Ctrl+Q"[1] is one of the three extensions I require to browse the web, alongside uBlock Origin and NoScript. I have no idea what the Firefox devs are thinking with that shortcut.

[1] https://addons.mozilla.org/en-US/firefox/addon/disable-ctrl-...


Probably that ctrl+q is the historical cross-platform standard for quitting. Also you may have missed that not all keyboard layout are QWERTY.


QWERTZ also has Q next to W.


I should know, I'm using a QWERTZ keyboard. But those are not all keyboard layouts: QÜERTY, ÄWERTY, QZERTY, DVORAK, BÉPO, JCUKEN, WORKMAN, ŪGJRMV, MALTRON, etc.

Then there are the asian and middle east that do not even use the same alphabet. Though those may use the QWERTY as an underlying layout for compatibility.


At least on the Mac, I resolve this for browsers and other tools with the Keyboard pane of System Preferences (very powerful, if you've never dug through it). For that matter, I also tweak other things I don't want to easily hit using the keyboard, such as making it harder to Minimize.


You seriously think that removing the ability to disable javascript (well, not really, you still can - just not as easily) is in any way a factor? Which other browser makes this easier?

Pocket? Hello? Really?

Firefox memory usage has for years used less memory; basically since its inception. Apparently it's no different now: https://www.gizmodo.com.au/2017/03/whos-winning-the-browser-...

e10s? Come on, give me a break. I bet the vast majority of users have never heard of it, and of the others, most don't know what it's about to any useful degree, and of those that understand this feature, most probably wouldn't know the details of how the various multiprocess implementations actually compare. A vanishingly small proportion of the user base know of this feature, understand it enough, can compare this to other browsers, and then have a strong enough opinion to affect browser choice (and frankly, it's not obvious multiprocess is actually that great of an idea in the first place if you really do know what you're talking about - not one of the browsers actually separates every tab into a separate browser - for a reason!)

As to OS support - firefox still is the last browser to support XP, so I'm not sure what you're referring to. Version 52 was the last one; but that's on an extended support cycle until june 2018, which AFAICT is more than two years later than chrome's last v50. Microsoft hasn't "supported" XP with any reasonable browser... well, not ever (the highest IE version was 9!), and it hasn't supported the OS at all even with security patches for years (with certain notable exceptions).

As to disruptions caused by signed extensions - so that's why the appstore has failed and nobody is using windows anymore? I get it's annoying, but this is a pattern that's recurring all over the industry, and has for many years before FF made this step. If anything, I think it's more plausible FF is being punished because it was too slow to ban unsigned extensions! Because poor experiences based on bad or even malicious extensions do reflect on FF. And for that matter, signing isn't the real issue, it's add-on sandboxing/threading. Chrome got this "more" right, in that it's less likely for an novice extension author to accidentally bring chrome to a grinding halt. But precisely this feature is still causing lots of addon breakage because FF has not yet completely dumped the old, problematic add-on API, presumably because users really hate losing their cherished extensions (and for a reason). I've witnessed several addons that have chrome+FF equivalents where perf issues occured only in FF - which may have been the addon author's "fault" - but that's a really poor excuse.

Poor perf, and the expectation of poor perf sound like more reasonable guesses, but even there I'm not convinced this actually matters as much as you'd hope. Still, that's at least something. But then, the number of people you see working with unworkably slow setups for all kinds of reasons that apparently don't care enough to switch products suggests that even abominable perf isn't necessarily very impactful. Maybe this matters indirectly; in that power users that care influence others in their choices.


I think that many of the changes that Mozilla has made to Firefox that people in this thread are complaining about may not have directly driven away a large number of users, but, they indicate a serious problem with decision making within Mozilla.

It is clear that the only reason many changes were made, and features removed, was solely because Chrome did it. And Google has very different motivations and goals than Mozilla. Google wants to make money, and use Chrome as a pillar in their platform. So, by emulating Chrome so closely, not just does it indicate that the developers are making bad decisions, it also means that the browser will not be as good.

EXAMPLE: They proposed removing FTP support from Firefox, and the justification was just a link to an announcement that Chrome was doing it. [1] It makes sense for Chrome to do it from a business perspective, but it does not make sense for Firefox.

Or, better yet, I remember that there was talk of having Chrome switch back to using a native pdf renderer back from the javascript one. This sacrifices portability and arguably security for speed.

Or sometimes there are design decisions in Chrome that are outright hostile to the user, to help Google's partners, such as removing the "save as" option for html5 video. It is only a matter of time until Firefox makes it harder to download video, solely because Chrome is doing it. When Google does this, I at least understand that their sabotaging this functionality is part of their larger strategy. Mozilla doing it is just baffling.

I mean, the original Firebird went in the opposite direction as Internet Explorer 6. If Mozilla had the same culture back then, they would have put all of their resources into making an inferior clone of Internet Explorer.

Internet Explorer was a better user experience in a lot of ways, especially for the first few years. But people started moving to Firefox because it was worth it. The security, control, and flexibility was worth it. I specifically remember turning people onto Firefox because they were sick of ads, and there were special add-ons that they wanted.

If Mozilla wants Firefox to work, it needs to do what Chrome wont let you do. It needs to integrate aggressive ad-blocking. Let you have control over the content you view. I think that people would happily use Firefox if it empowered them.

1: https://bugzilla.mozilla.org/show_bug.cgi?id=1174462


I think you misundertood the justification of the "remove FTP" bug.

That bug was filed by the person who wrote most of the FTP implementation in Firefox and was one of the few people who maintained it. The reasons to consider removal were that it was a maintenance burden, extra security attack surface, and not really relevant to users nowadays. The posted link only spoke to that last point: that other widely used browsers were removing it and it wasn't being a problem for their users, apparently.

The first two reasons for removal were not clearly explained in the bug report initially, because they were obvious to both the bug filer and his intended audience: the networking module owners and peers.

See https://bugzilla.mozilla.org/show_bug.cgi?id=1174462#c23 for more details.

Would the summary being "consider removing FTP support" more accurately reflect intent? Probably. Did the actual engineers involved know what the bug was about? Yes.

> Or, better yet, I remember that there was talk of having Chrome switch back to using a native pdf renderer back from the javascript one.

I assume you mean Firefox? Speed was part of that discussion, but so were the various features pdf.js lacks but the proposed replacement supports.

> It is only a matter of time until Firefox makes it harder to download video, solely because Chrome is doing it.

If you assume that Mozilla is just aping Google. But if, on the other hand, you just have that as a preconceived notion and try to make all decisions fit that theory no matter what the actual reasons for them are... then you might just guess wrong on this.

I agree with you on the broad point that Firefox needs to do things Chrome can't or won't do. The worry with your "aggressive ad-blocking" suggestion (much as I would like that personally!) is that a likely outcome is a large enough number of sites blocking Firefox altogether that users stop being able to use it at all for normal browsing. If Firefox had a monopoly position in the market it _might_ be able to get away with that sort of move, but it doesn't have that position.


> I agree with you on the broad point that Firefox needs to do things Chrome can't or won't do.

Actually the point was that mozilla should stop removing features that are not in chrome first, if there are any left. The point is firefox should differentiate from chrome instead of being more and more similar because being similar to chrome just remove the appeal firefox had.

> If Firefox had a monopoly position in the market it _might_ be able to get away with that sort of move, but it doesn't have that position.

But when they had the better share of the market they were adamant not to do any adblocking and this may be related to the fact that 98% of their revenue came from advertising through google. Then again how would a website block firefox ? Using user-agent ?


> I think that many of the changes that Mozilla has made to Firefox that people in this thread are complaining about may not have directly driven away a large number of users, but, they indicate a serious problem with decision making within Mozilla.

Of course they have, I'm such an individual but I'm responsible for about one to two thousands installations of firefox over the last decade or so. Carefully replacing any sneaky install of bundled chrome each time I faced one, and so on.

It's been 2 or 3 years since I've gradually stopped doing this. I have no reason to keep helping a corp that keeps disappointing and not caring about users, selling an image through marketing (giving power and freedom to users) while doing the opposite in reality (removing power and freedom from users). I've barely made any new firefox installation and stopped replacing sneaky chrome/IE unless expressly asked.

Just with one individual they've lost I'd say about a few hundreds users over 3 years. Although a few hundred users is not even a blip on their radar, I'm not alone in this situation and numbers quickly add up to a significant amount. Though maybe not significant enough for mozilla's devs to care.

The point is, a limited number of individual users played a role (possibly significant) in firefox success, the same people now play a role in its demise. It seems mozilla does not understand this fact despite building a large church of evangelist fanboys.


Either Mozilla is infested by an me-too syndrome or they employ managers who do their current employer (Mozilla) a disservice (and their former employer a service). A big problem anyway.


I wouldn't care for a pdf viewer in the browser. What is the usecase for having this in-build?


Do you mean a native pdf viewer or any pdf viewer including the javascript one? It's very convinient to open pdf:s in the same tab and not having to download a file and open it in an external viewer.


I don't find it convenient at all, given that my native PDF viewer clearly outperforms the in-browser viewers, specially in large PDF files.

Change it to "open in external application" is one of my first changes after a fresh install.


But for small pdfs, the in-browser pdf viewer is typically faster. I use lots of PDFs and while I read most in adobe reader, I still appreciate the in-browser rendering: it's often good enough, and stuff like smooth scrolling and tabbed browsing actually works better, and the UI is less in-your-face. The windows PDF reader has (IMNSHO) ugly fonts, and no upside I can find whatsoever.


Turns out to display the document you have to download the pdf anyway. I had so many issues with this PDF in browser thing reported to me by users... And disabling it can be a hassle with firefox sometimes ignoring the setting or simply not offering the option. I don't know what purpose it is supposed to have but it only gave me more support requests.


Have it updated and sandboxed.


Useful for system without a package manager.


> Microsoft hasn't "supported" XP with any reasonable browser... well, not ever (the highest IE version was 9!)

Actually IE9 requires Vista, so XP is stuck with IE8.


You are, of course, completely right - I clearly misremembered that!


Removing the option the ability to disable javascript was a deal breaker for me. Luckily I could sort of restore this much needed feature through extensions.

The other browser that got it and offered a quick menu to disable/enable javascript was opera with the f12 or quick preferences menu, not sure if this option survived opera selling out and becoming a chrome skin.

Pocket was another thing not needed to disable in firefox, it was proprietary and had vulnerabilites when a better opensource alternative existed, mozilla got a lot of flak for this. Then they bought it for millions of $ to leverage the users to get into mobile.

Hello was again another unneeded thing to disable right away (yay more extensions!), it leaked the local IP and added vulnerability.It was such a popular feature that it got removed a few versions later.

Those 3 really infuriated me and a portion of the firefox users community.

I can't tell about the memory usage, I've bought a 16GB RAM laptop a few years ago to free myself from firefox memory issues, I do know that once in a while I have to quit firefox because it's consuming too much memory and cpu while idle.

by the time e10s arrived, there was no "vast majority of users" anymore for firefox. Having not heard of it means mozilla communication sucks. Anyways e10s is disabled in my firefox, I do not remember why.

I'm not sure what you mean by firefox being the last browser to support xp, 6 months ago I tried to install firefox on windows xp and installer denied to install telling me it does not support xp anymore. I think I installed pale moon instead and opera so maybe firefox is not the last one to support xp.

Signed extensions caused an uproar and some people left firefox over this because it's removing freedom from user and giving more power/control to firefox. I had to drop a couple extensions that refused to comply with mozilla demands or were not updated.

And there are other things like this, and when all this happens while bug report dating of years or decades are still waiting to be dealt with or closed as WONTFIX because you do not matter enough, well...


But isn't that exactly what extensions are for - to enable power users or simply niche use cases to flourish without bogging down the UI and codebase for everyone else?

I would have liked Hello (or really, any decent chat application that's not just some vehicle for the dreams of vendor lock-in by a large company) to succeed, but you're right, it flopped.

Firefox 52 was supported on XP service pack 2 (or later). Version 53 was released April 19, 2017. If you couldn't install FF on XP six months ago, you either were using a really really out of date XP (SP2 was released 2004!), or you misremembered, or you were using some beta or otherwise nonstandard FF.

I mean - I don't really disagree with you; heck I share some of the same frustrations. But what the heck is the alternative? Most of what you're mentioning is worse on every other browser.

What remains is perf and extensions; but I feel that those two issues are intertwined. FF extension APIs were, unfortunately, misdesigned. Not through malice or even incompetence, but because this turns out to be tricky. And so it was really easy to bog FF down with benign-looking extensions, and furthermore, it was difficult for FF to modernize with such low encapsulation. They should have bitten the bullet much earlier - instead, the problem kept festering.

When you say FF was a hog on a 16GB laptop, then it's almost certainly down to extensions - those very extensions you're frustrated they're forcibly modernizing. I don't think there's any good solution here.


IIRC extensions were the excuse for not adding features in firefox because it was supposed to be the most basic possible to avoid performance issues.

Thing is there is no need to bog down the UI for people who don't need certain advanced features (of which an adblocker is not), you can be smart about it. Once again look at opera did it. For example mouse gestures were disabled until you first tried to use them and at this point it would spawn a window asking you if you wanted to enable them.

I have no idea what the reasoning for adding hello into a web browser. Opera did it with an IRC client a couple decades ago and it stopped being relevant at some distant point in the past. There are enough chat applications around, no need to push an additional in the browser, especially when on the other hand you remove or do not add features on the pretense that you lack resources.

I do not misremember at all (well it was probably more than 6 months ago as this took place at the end of last year). I remember perfectly being sorry for cleaning up my archives of older version install files for firefox because the oldest I had was version 49 and this version had already dropped windows xp support for this hardware (and IIANM firefox dropped linux support with version 52 or 53). Faced with replacing fully functional CPU/ motherboard/RAM or dropping firefox, guess what was the chosen solution.

Even pale moon that's based on older firefox versions has dropped window xp support[1].

I wish there was alternatives, but AFAIK the only alternative is to rely on older releases. Next time I have such an xp machine between my hands I'll do more research and tests to hopefully find something that will fill the gap.

Windows xp sp2 has been released in 2004 but we still fill containers of such computers to ship to Africa through humanitarian programs. We have a container leaving in september of this year. Most technology oriented people from rich western world tend to overlook that the rest of the world has not the same privileges they have.

I agree that it is not difficult to have poorly coded extensions or misdeisgned API causing a drain on resources, but this a strong design choice by mozilla who consistently refused to add those features to their browser even if incredibly popular (heck they even remove them at times). So even if the extension is the reason for the performance issue, the responsibility still lies with mozilla.

I don't mind modernizing, what I mind is that said modernizing means removing a good 70% of what I had to add to firefox to turn it into a modern browser and make it useful to me. When they tout freedom and user power but remove control and power from my hands while ignoring user feedback or even knowingly messing users because well they're not a high percentage in their skewed metric.

I may badly expressed myself here, I didn't mean that firefox is a hog all the time, but that with my use case of couple hundred tabs open over weeks or months it happens once in a while that I have to close it to clear an unexplained CPU or RAM hogging. On the other hand it is common to have 35-75% CPU usage of one core for just being opened and idle which is something I don't understand. When this happened in opera a simple F12 / disable javascript was enough to stop this nonsense. My take is that it's not as much an issue with the browser than with those stupid js heavy websites, the shortcoming from mozilla is to have removed the ability to disable scripts.

It's probably extensions in most cases, but I have replicated the behaviour with a brand new user profile and a vanilla firefox without any extensions, it actually happens faster and more often without extensions (lack of ublock origin and noscript plays a significant role here I guess).

Anyways I'm disgruntled, frustrated and fed up with this mozilla bs over the years. I sincerely hope they disappear sooner than later or that people in charge are replaced by competent people and that a user centric browser emerges from this mess.

[1]: https://www.palemoon.org/PM_end_of_WinXP_support.shtml


The visual drops in data shows only one of the things you mentioned had a sizeable impact, and that was discussed in the article.

Almost everything you list are things that happen around a certain time and we don't see those events manifest in the graph.

The reason of course being that the vast majority of Firefox users are not expert and don't care or know about those advance features.

Maybe if you can link to a user satisfaction report...


I try firefox usually once a year to find out nothing changed about freezing, crashing, slow reactions and bad rendering on machine with 2GB RAM.

Meanwhile Chrome just works. It's even worse on Android, they don't have in FF even pull down to refresh and it's the slowest browser I tried.

More

Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | Legal | Apply to YC | Contact

Search: