My opinion is that none of those actually matters. The "one" single reason Chrome won market share was the same playbook how Google Search Engine Started.
Speed.
I watched hundreds of users, given the choice, or doing so themselves installing / using Chrome by their own selection. It was faster. The user experience was far better.
You could advertise the heck out of your product, but if it wasn't any good it wouldn't matter. For nearly 10 years Google Chrome was undoubtedly the fastest browser on the planet. It wasn't until Quantum landed on Firefox before you could at least have a few metric of performance wins going to Mozilla, and that is also partly because Chrome has slowed down a lot in recent years.
I don't know who you are, but having watched “hundreds of people” installing a web browser sounds a bit dubious…
Personally, I have discussed with a few dozens of websites users and most of them would answer the question “what web browser do you use” with “Google” no matter if they were using Chrome, Firefox, or IE (there are few Mac users in my country).
Firefox users had it installed on their computer by some computer-savy relatives, and most of Chrome users didn't know were it came from (most believing it was already there beforehands) which is a good marker of people getting it bundled with other pieces of software.
There's definitely jobs where this is possible... IT support at a medium sized company, various user research, etc.
Gmail was bait to lure millions into the Googleverse, where they usually don't leave.
Personally I'm kind of happy to see Microsoft Edge gaining some marketshare at Chrome's expense. (even if it's mostly just Chrome)
for the longest time chrome advertising on google.com were actually malware : (
If someone claims to have done some work with tech, chances are they're telling the truth
Really the parent commenter is right: Chrome is dominant only because of its huge marketing budget.
P.S. Chrome is also buggy as hell and crashes pages all the time, which Firefox doesn't do.
Are you comparing current Chrome to current Firefox? Relatively recently, compared to the overall trend we're seeing here, Firefox got much faster than it had been: https://blog.mozilla.org/blog/2017/11/14/introducing-firefox...
The discussion above, however, is about how Chrome came to grow so much while Firefox shrank so much, and for that the question is how did Firefox and Chrome compare historically. There was ~10y from Chrome's release until Quantum came out.
(Disclosure: I work for Google, speaking only for myself)
Even opening a select box is slow, if you have several 1000 tabs.
Funny enough, opening a new window (and keeping all the other tabs open) makes it fast again, if you are in that window.
Seems like Firefox has some slow code paths, like notifying every tab about going fullscreen?
Is it possible to have 1000 tabs open in Chrome? When I had like 50, it ate my whole 8GB last time I tried it.
In Firefox I frequently have 1000 tabs with no issues. (almost none, recently I noticed that there is some memory leak and the ram slowly disappears).
This is the reason why I went back to Firefox after using Chrome for like 2 weeks, it is great for 1-10 tabs.
Chrome kills them regularly on it's own, which means they simply crash with out of memory.
For me, Chrome generally handles killed renderers better, e.g. it shows a popup about crashed extensions and just reloads them. Firefox requires me to quickly disable and enable adblock and I lose the build in screenshot extension permanently (except disabling and renabling it with about:config). But for chrome a renderer often just freezes and only a day later there is a popup about it which offers to kill it.
Firefox seems to handle the tabs better overall, but at some point session resume fails with javascript errors in the shell. Chrome seems to handle session restore a bit better.
Also FF was shitty at stability, it didn't have multi process architecture for the longest time and I've had a page crash my entire browser (especially in the flash days) regularly.
And then chrome started pushing excellent dev tools (even in FF nightly I get dev tools hanging on complex sites because of how slow they are)
Chrome was the better browser back then and it won the power users and developers, then it took over the casuals as it was advertised on google and it was recommended as the best by everyone - it took FF too long to catch up and now a lot of sites don't even test compatibility - I've been trying to switch back to FF before this fiasco for a month and it wasn't a smooth experience compatibility wise.
WHo cares? A Ferrari and a Lamborghini in 2020 provide about the same speed experience. What matters is that for the past 10 years Mozilla had the speed of an old Volvo versus the Ferrari that was Chrome/Chromium in comparison. The fact that Mozilla merely caught up in the end is rather a proof that they lost technical leadership a long time ago.
Speed's always been similar for me, Firefox has generally been more reliable (partly due to being able to block so much crap, probably). Chrome used to break with alarming regularity and need a total reinstall. Sometimes there was a site which wasn't compatible with Firefox, and if I needed it I'd open it in a Chrome incognito window.
These days I've switched to Vivaldi for the work stuff (vertical tabs, "hibernate background tabs", tiling!), though I still debug JS in Chrome.
To win over non tech people, there should be at least one "killer feature" out of the box.
In case of Google Chrome it was speed.
In case of Google search, it was quality of results.
In case of whatsapp, it was availability on all kinds of devices.
Etc..
Or enforced by organization etc.. that's different thing.
For the first time or because they were already familiar with it? Once you've established yourself as the "default", it can be ridiculously difficult to convince people to switch to something else. Google has executed this strategy perfectly and they know that there's nothing Mozilla can do or anybody else will do. And that despite the fact that Google clearly abused its dominant position in another industry to establish dominance in this field. I'm sick of Google abusing its position and I'm sick of people defending Google because it's a "free market". No, that's not how it works.
Not very believable as FF actually gained a much bigger market share than default IE over time.
IE was a terrible browser. Everyone knew it and Firefox was better in every imaginable way especially for "web masters". The Firebug system was a godsend.
Firefox unseated IE because of how much better it was. Microsoft could not get its head straight in this particular case. Always a non-conforming, slow browser with far fewer features.
That's what it takes to unseat the default - you have to be multiples better before people take notice in serious ways.
And, Chrome did that beautifully by leaning into the one thing the masses cared about. Speed.
Those ads were so good - extremely viral. Did they show anything objective or real? Maybe... maybe not. It's what caught people's attention and then when they tested it themselves, they believed it.
Personally I tested it and it was, by a very noticeable amount - I was hooked. Took me a bit of time to get over the loss of Firebug, but mostly sentiment because Chrome's debugger was quite good.
If Firefox was the semi-truck, Chrome was the Ferrari. Both got me to A => B. With firefox I could carry anything, plugins, extensions, whatever I needed. But it was a lumbering giant. Chrome came with none of that at launch but it went fast - and people talked about it.
Those were the days.
The fact that Firefox was faster than IE and any technical expert you asked for recommendation would prefer it due to better standards compliance/extensions/having the only devtools were really the icing on the cake.
Google also has the kind of trust amongst end users, especially non-technical end users, that Mozilla couldn't hope to have since Google has their hands in so many pies. And as a technical friend/family member asked to make a recommendation, if anything the incentive is to recommend Chrome to reduce tech support calls when HangoutMeetDuoChat/YouTube/Photos/their preferred google product is acting weird in Firefox again.
It's hard at this point to see what any browser could do to build that kind of end user feature lead again this far into the game.
Nitpick, but Firefox never had the majority of the market or a bigger piece than IE. Firefox was on the way to reaching that but they peaked around 33% share before Chrome came along and ate everyone's lunch.
So, the statement does not mean anything anymore then. It should rather be: "it's difficult to unseat a technically superior product with an inferior one" which is patently obvious.
No, they did not. Electrolysis was around before WebExtensions. Sure, Electrolysis wasn't backwards compatible and a lot of existing extensions had to rewrite part of their code to be compatible, and abandoned extensions were broken forever - or would have been once mozilla removed compat shims (unless somebody picked up the slack, which people quite often did). This was still a lot better than breaking every extension and forcing a new extremely restricted WebExtensions API that was a clone of the Chrome APIs but came with their own bugs and a lot of "TODO: implement actual API".
Quantum is just a PR term. The impact of Quantum-related changes to extensions was tiny if present at all (if the extension already survived Electrolysis).
The only reason that old non-webext extensions do not work anymore in Firefox at all is that they helpfully removed the actually code that would run the extension (but not really last I checked, it's still around to support mozilla-owned sideloaded/builtin extensions such as pocket).
comm-central (Thunderbird, Seamonkey) even implemented their own implementation of what was removed to get (restartless) old-style extensions working again, falsifying the claim that forcing a WebExtensions-only change was a requirement to get e10s/Quantum going.
Underneath it all, to this day most of Firefox is still javascript + XUL/HTML on the frontend with bridges (XPCOM, etc) to expose the C components.
PS: e10s shipped in v53, "old" extensions were only removed in Quantum, which was v57 (but not really removed, as I noted, just disabled for everybody except mozilla's own stuff).
I ran (and sometimes still run) a personal fork of 60esr (aka >57 ;) where I added back "old" extension support by git-reverting a few hundreds of lines of code that they removed.. That thing was Quantum too, all the Quantum perf/speed improvements were still present, and yet magically still supported all the old extensions I cared about.
<html:body xmlns="http://www.mozilla.org/keymaster/gatekeeper/there.is.only.xul">
<popupset id="mainPopupSet">
They are trying to replace XUL with HTML and made some progress towards that. For starters they removed most or all of XBL.
From the perspective of an add-on dev it wouldn't have made much of a difference, it used to be manipulating a XUL DOM, now would have been manipulating a mixed XUL/HTML DOM, in the future it would have been manipulating an HTML DOM, not much difference except for tag names and attributes names really, and the rest was js and css.
If one unzips this tiny XPI, one finds javascript code which can directly install a certificate into the master store.
So there is indeed a very powerful extension system available, but only to Mozilla itself who can sign such extensions.
Probably all the old stuff can be brought back easily.
Please put your comment in a blog! Back then I searched for these topics and found nothing
Not anymore. They removed/replaced a ton of stuff that only was used by extensions at that point. That might lower their maintenance burden, but for the most part it really was interfaces and code that hadn't been touched since the 1990, so it wasn't like it was a huge maintenance burden before.
That means it wouldn't be that easy to bring back the extension system where the old extensions would be able just run unmodified. But it would be easy enough to bring back newly created old-type extensions or modify existing old-style extensions to use whatever is present now instead of the old stuff that got removed.
But then again, old-style extensions had to check anyway if a new major release broke anything, and usually it did break some stuff for some extensions. It was the cost of doing business. Jetpack/WebExtensions could help with that; if you extension was simple enough to fit within their APIs.
In fx60 (esr) the removal of these "old" interfaces and other things hadn't progressed much yet, so I was able to just git-revert most of the removals that affected the extensions I care about and that was enough to run those extensions again without obvious bugs.
Sir, I would really appreciate if you could share a patch for it.
That's fucking infuriating. You should write up a blog post about how easily you were able to turn them back on in Quantum.
But here is the important bit regardless: https://gist.github.com/formerextensiondev/50f3ebf563048e75f...
And even back then I was too lazy to actually properly "fix" it as you can see.
v. Intentionally breaking other browsers subtly on Google properties.
i, ii, and v are directly from Microsoft’s 90ies/00s playbook and the regulators are sleeping.
Hard to say if incidents like that are intentional, passive aggressive, or accidental, but they happen pretty frequently.
Same with Youtube. Firebase console. GCP.
Which is why, for the life of me, I just can't see the logic of firing the Servo team.
It’s not like they binned servo and decided not to use it improve ff. It did improve ff.
They seem to have decided that’s done and finished now. And there’s no longer any need to work on it - whether that’s a wise move or not is obviously up for debate.
Torpedoing the technological future of their browser is bound to reduce that share, at which point Google will likely offer much less money.
They are basically shooting their one trick pony in a leg, deliberately. It’s.... not smart.
In the face of that, it would be reasonable to think that the cost is nowhere near worth the effort, compared to just adopting Chromium.
I think it'd be very easy for them to think that their goals for the modern web are not intrinsically tied to having their own, completely separate software.
That's because Firefox is slow. I'm saying this as someone who's been using it as the main browser for at least a decade.
My latest personal anecdote. Firefox trackpad scrolling is lagging on my brand new Precision 7750 with vanilla Ubuntu 20.04 installed. Chrome doesn't have the same issue, especially considering Google's unfriendliness towards Linux.
Spending time and money on solving these issues isn't going to generate new revenue. I find it hard to believe engineers wouldn't want to try fix performance issues. I find it easy to believe leadership at Mozilla would find it unimportant.
edit: clarification
Unfortunately, because Google is "In" just because it delivers mind-numbing ease via Youtube etc., this very vocal minority, claiming technical expertise scew impressions for the general population.
Similar to absolute views like :They do not care... I think they care, not about the particular problem, but the product and how it meets the requirements of a general audience swiftly and reliable. So the problem crossed their desk, they triaged it and closed it because it is an extreme use case which will not meet general requirements.
The same goes for all the multitudinous problems in that area.
From a Product perspective Chrome is the better product. I want to believe that Mozilla is capable of making FF the better product. Maybe it all boils down to Youtube and Facebook and in actuality it's impossible and they can't actually have a better browser. I don't really believe that. I do believe that Mozilla isn't on a trajectory to make that happen.
I wasn't referring to the tabs either, just as an example of people doing it wrong. No one stops them, but in the old days natural selection took care of it. These days people think, because they have a misconception, they have a right to an opinion which a misconception is not. This is known as "The terror of the masses". Desinformation and Fanboydoom(we against them) is based on that.
Yes FF was slow some 5/7 years ago, but speed up quite a lot. See here https://arewefastyet.com/win10/overview?numDays=60 . You can, of course, choose the Benchmark proving your point, or look at the bigger, quite different, picture. And just because 200+ people write the same comment in spirit does not make it true (see benchmark-site) . Also, as I have pointed out elsewhere in this thread, it is not just about speed. I agree that a site should not be agonizingly slow to work with, but that is not the case with FF (see benchmark). As far as the reference to the trajectory is concerned, it is not s.th. that securely predicts outcome. You never know what the future holds. Just think of the situation of MS or Amazone some 10-15 years ago. Besides Youtube and Facebook have nothing to do with the browser; also note that Alphabet purposefully sabotaged their own website so that it performed bad or even breaks on a non-Chrome browser. https://www.zdnet.com/article/former-mozilla-exec-google-has...
I would also like to remind the reader and you that I never said that the fault is with the user, I merely stated that you cannot demand a product owner, ANY product owner, to cater to ALL and EVERY extreme edge-case.
But since this debate of ours, I fear so, because you tried to warp above said statement of mine, is in danger of devolving into a fan-boy flame thread, let me close this with a quote I read in a facility of a philosphical faculty, in hopes that it gets people thinking before they regurgitate other people's/trolls/campainers opinion/misconception unreflected: "People, eat feces, for a 1000 flies cannot lie."
The edges are where some of the most faithful users live. But you're right, it's becoming more and more clear, over various updates and certainly with the recent news, that Mozilla thinks they could do without the trouble and their PMs act accordingly.
Whether it's an issue that only one in a hundred users has to deal with, or an issue that every user has to deal with once in hundred of their actions in a browser session, an edge case is just an edge case.
They made a pretty huge deal about the Tab Groups (aka Panorama) feature of Firefox 4 in 2010; designed by Aza Raskin! https://web.archive.org/web/20101105212333/http://www.azaras...
But then it was removed in 2016 because few people really used it (or few people who had telemetry enabled, at least): https://support.mozilla.org/en-US/kb/tab-groups-removal
Relying on telemetry is heavily biased towards people who do not know what telemetry is.
This whole anti-telemetry outrage is so manufactured and boring.
(1) A significant portion of Firefox users think like me. In this case, Mozilla should not rely on telemetry because it excludes a sizeable portion of their userbase, and one that likely correlates with things like "believes in Mozilla's stated core philosophy" and "is passionate enough about browsers to recommend them to people".
(2) Not many Firefox users think like me. In this case, it's pointless for me to enable telemetry because they'll just ignore my needs anyway. "It is not worth supporting this feature, the data shows that 99% of people do not use it!" won't change if the number is actually 98.99%.
More importantly, it seems both of them have survived the move to Quantum. It was really hard for me to imagine going without Panorama.
https://addons.mozilla.org/fr/firefox/addon/tree-style-tab/
?
b.) With changing their extension API several times it was not clear the plugin would make it.
That was the big one that cost a lot of users, but it was a jusitifiable big leap for them to take.
But they have also been suffering papercut after papercut power user losses with various regressions and removal of functionality since then.
The actual users lost are minimal, but I suspect the overall perception of Mozilla's priorities, among those kinds of users, has suffered much more.
The reason Firefox has lost market share in my opinion is because:
i) Chrome is pushed by Google.com, which has a near monopoly on search
ii) Chrome was bundled and installed alongside other software
iii) Chrome is built into Android and Chromebook
iv) Chrome has a huge advertising budget
And all of these factors combine to make Chrome feel like the default portal to the web.
Also for a period of time Chrome was undeniably a better, more stable browser. It’s only relatively recently with Electrolysis and Firefox Quantum that Firefox has regained a technical parity or superiority. And those changes depended on specifying a proper UI and Extension API which is the reason why the browser is no longer so customisable.
But Firefox Quantum only exists because of Servo, and vital parts of Servo (parallel layout in particular) have not been integrated into Firefox. It’s bizarre to see Mozilla sabotaging its long term chance at technical leadership not only by scrapping Servo but firing the engineers who built it.
I had an untechnical family member say unprompted last week they want to move away from Chrome but couldn’t use Firefox Mobile because it’s too slow. At the same moment Mozilla leadership was sacking its staff working on parallelising the browser engine.