No, I tell my friends and family to use Firefox. There is more than one decent browser out there.
I don't like the idea of everything using the Chromium engine; monoculture _is_ bad and we shouldn't put our eggs in the one basket. That said, IE and Edge have historically been terrible both compatibility and performance-wise. It's not just websites that aren't tested in Edge - my experience with Edge is that all aspects of performance (initial load time, ui responsiveness, web site responsiveness) are terribly slow.
I applaud the death of IE and Edge. One thing I'm worried about though is that Chromium may take over the market and we end up with websites working only in Chromium based browsers and other important browsers like Firefox and Safari get left in the dirt.
It's not really better though - having only two real browsers is not good for open standards or the open web.
This will be especially true when Firefox inevitably shutters it's in-house efforts in a couple of years due to the amount of sites that won't work with it, and starts using Chromium too.
Staring into my crystal ball tells me Firefox will become "janky" in the eyes of users on account of how many sites don't load on it like they do Chromium (because developers will only test on the most popular browser, because THIS IS WHAT HAPPENS LITERALLY EVERY TIME).
> This will be especially true when Firefox inevitably shutters it's in-house efforts in a couple of years due to the amount of sites that won't work with it, and starts using Chromium too.
It seems very unlikely:
- Mozilla has been in a far worse uderdog position before during the IE6 era, with way more incompatible sites and less funding.
- Mozilla is not for profit. Fighting for the open Web is one of their goal. It always has been. They are not perfect, but their track record is damn good compared to almost any player of their size and impact in the tech world.
- Mozilla strongly invested in their own tech, including rewritting the browser rendering engine and taking huge risks such as create a bloody hole new language, rust, in the process. To my knowledge, the "oxydation" project has been a success so far, and rust is proving everyday that it's a positive force in the world as well.
- Firefox is the only decent mobile browser. I can't navigate the web without the ublock extension. I just can't.
- Mozilla keeps innovating. Their last brillant idea, the tab container, is worth switching on it's own.
- Mozilla has the hardcore geeks on their side. Even during the V30 to v50 transition period where Firefox was, at the time, clearly an inferior product, we kept using it to support it for the sheer ideal of it. We hoped it would come back from it, and it happened: Firefox is now a fast, lean and fantastic browser again.
- Privacy concerns are (FINALLY !) being taken in consideration from the crowd. And Chrome is terrible at this, so moving to a chromium core, while technically not related at all because you can set it up the way you like, would carry the stigma.
All in all, I'm incredibly optimistic about Mozilla et Firefox's future despite the market share taking a serious hit.
> Mozilla has the hardcore geeks on their side. Even during the V30 to v50 transition period where Firefox was, at the time, clearly an inferior product, we kept using it to support it for the sheer ideal of it. We hoped it would come back from it, and it happened: Firefox is now a fast, lean and fantastic browser again.
I use Firefox. I always have. My experience is totally different to yours. Before FF57 it was a single process and it ran nicely on a 4GB machine with a 2009 Intel Atom processor. Afterwards it became much hungrier for memory and processor. I had to buy a new computer. (I tried Chrome of course but it is hungrier.)
I said "during the V30 to v50 transition period where Firefox was, at the time, clearly an inferior product", compared to the competition, not anterior version of itself.
The later is a more complicated matter, as it was 10 years ago, with different expectations, hardware, user base and web.
And I pointed out that prior to v57 FF was a clearly superior product to the competition for me. For me FF v52 was also superior to v57. Sadly v52 is now out of support and the result is that I had to get a new computer.
For me the order of utility is FF pre v57 > FF post v57 > Chromium based browsers.
The multiprocess architecture makes it run faster on newer machines, but slower on older ones. Have you tried disabling it (limiting to one process) though?
>Firefox is now a fast, lean and fantastic browser again.
This is not my experience. On my MacBook Pro it is slow and tends to cause system lockups on a regular basis. Everytime there is a new announcement about how the new Firefox has improved performance and stability I give it a try and each time am disappointed.
On my phone the Firefox browser is almost unusably slow and slow enough that it can't replace Chrome. Mozilla has released other mobile browsers which perform faster but they lack the features I need in my browser.
If one doesn't want to rely on anecdata, here [1] is a benchmark. Based on my experience and this benchmark, I don't how such a negative opinion can be supported.
I don’t experience any lockup’s. I switched to Firefox as my main browser on Linux and macOS over a year ago full time. Google Meet finally works properly with it, so I don’t even have to use chrome for that last thing.
I don’t know what you’re doing, but I’ve never had a lockup in that entire time (and I run Beta), as far as I can tell, it is the fastest of all my options.
The very last thing that I notice any difference on is energy usage, for which Safari is still king. I don’t use Firefox on my iPhone, as there’s not much point (I wish Apple would change their restriction here).
My anecdotal experience shows a) less of a battery hog than Chrome for like 40-50 tabs b) now works with all my sites (exception being Google drive share box has blinky UI) c) it doesn't keep begging me to "sign in"
For this it's become my main work browser. At home I mainly use safari and it works great for 15-ish tabs.
I seem to recall that there was a bug on macOS which made it eat too much CPU on non-native screen resolutions (UI scaling) for some reason, which likely explains drastic differences in how people perceive Firefox there.
> - Privacy concerns are (FINALLY !) being taken in consideration from the crowd
I don't know how true this is when considering all the calling-home (split over multiple settings in about:config making it difficult to disable) in modern firefox versions.
> Mozilla has the hardcore geeks on their side.
A lot of geeks are moving away from it to forks or to other browsers like qutebrowser, which actually makes senses considering that mozilla keeps trying to sabotage the poweruser demographic.
> I don't know how true this is when considering all the calling-home (split over multiple settings in about:config making it difficult to disable) in modern firefox versions.
I agree but between a broken leg and cancer, you choose the broken leg.
> A lot of geeks are moving away from it to forks or to other browsers like qutebrowser, which actually makes senses considering that mozilla keeps trying to sabotage the poweruser demographic.
This has always been the case. We try alternatives, that's what we do. It's sane, and no matter Mozilla's behavior, we would do it.
While I'm with you on the variety side, lest not forget that Chromium is at least open source, permitting other vendors to do a move like this instead of being left in the cold (and in this way probably committing back to the project). The flip side here is that building a fully featured browser engine from scratch that is "just" good enough to render 80% of pages correctly is a herculean, almost impossible task in 2018. Browsers went from graphic tools to browse the Internet to a full blown Operating System with sandboxed arbitrary code execution, intricate cross-origin rules, staged caching, full vector graphics animation systems and whatnot.
Although I would have loved even more choice (or MS championing the superb Firefox rendering engine instead), I think we're already well off with two fully featured and completely open source browser engines.
> building a fully featured browser engine from
> scratch that is "just" good enough to render 80%
> of pages correctly is a herculean, almost impossible
> task in 2018.
Absolutely!
> I think we're already well off with two fully
> featured and completely open source browser engines.
Well, that's the problem. The death of Edge brings us one step closer to a browser monoculture.
Which is apparently something developers want, since apparently none of them were working in the industry the last time we had a monoculture.
> This will be especially true when Firefox inevitably shutters it's in-house efforts in a couple of years due to the amount of sites that won't work with it, and starts using Chromium too.
Watching Mozilla align Firefox's extension API with Chromium's, I'm a little surprised that they haven't already made the move. If browser evolution continues along the current path, I predict Firefox will switch to Chromium within 3 years.
> Staring into my crystal ball tells me Firefox will become "janky" in the eyes of users on account of how many sites don't load on it like they do Chromium (because developers will only test on the most popular browser, because THIS IS WHAT HAPPENS LITERALLY EVERY TIME).
It's already happening. Until a year or two ago, the only time I ever opened Chrome was when a page using some sort of experimental API ran slowly or not at all on Firefox. Now, the number of mainstream websites I see with serious glitches in Firefox is increasing by the month.
Do you mind listing mainstream sites that have serious glitches on Firefox? It's my main browser, and I can't remember one single mainstream site that glitches on FF.
I just ran into a somewhat important website that didn't work right in Firefox. I tried creating an online bank account with Discover. The register account process only worked with chromium. Once I had an account, though, I could use it through Firefox.
It counts, sure, but frankly there’s too many competitive reasons for a non-search Google property to not work well in Chrome. An example of this being that Gmail doesn’t work particularly well in desktop Safari, nor in Edge. Chrome though, no problems...
9/10 times when I encounter a bug either when doing web dev or just browsing the web, it's because of Chrome/WebKit improperly implementing standards.
It's not Firefox's fault if people aren't testing their websites on both browsers, just like it wasn't Firefox or Chrome's fault whenever people only tested their webpages with Internet Explorer and its mess of exclusive APIs and deviations from standards.
You seems very certain that Firefox will switch to Chromium but the recent trend is going in the exact opposite direction :
- Firefox Focus is switching from Chromium Engine to Gecko Engine.
- Huge amount of work in WebRender, they're starting to test it. If it lives up its promise Chromium Engine might fall far behind in terms of performance.
So maybe you're right but from what we can tell right now, the current trend for Mozilla is to remove the last pieces of Chromium and bet everything on a new generation engine which is not Chromium.
So basically it is already really hard to support all web browsers out there, since there is no "standard". So what you do is you limit yourself to a few, and display "Best played in browser X", because browser Y has these strange issues, and might introduce others with a new update, and the same for browser Z and F.
So if you make rich web content such as games, the ecosystem will automatically push itself to a limited amount of web browsers.
Not to mention that fixing it for one browser may well break it for another. Usually it leads to compromising but even then still favouring one browser (usually Chrome) over another.
- Build some webapp and get it working correctly in Chrome
- Test on Firefox. It works 90% of the time. 9% of the time it's relatively easy to support. 1% of the time it isn't worth the hassle.
- Test on Edge. It works 80% of the time. 10% it's an easy enough fix. 10% you have to move heaven and Earth to fix and when you contact Microsoft, they know about the issue but won't do anything to help.
- Test on IE. It works 60% of the time. The other 40% of the time you start looking for a new job so you don't have to write the same code a second time for IE.
Edge guesses for each number on a page if it might be a phone number. The guess depends on your locale. If it is considered phoneable, edge removes all javascript and replaces it with a link to skype. Which is fun if skype is not installed, it just pops up a messagebox complaining you need to install something, and it's not going to tell you what exactly. The error is only half translated and lost any meaning in the process. Can't put a DOM breakpoint on it either, the bloody browser just ignores everything you do to that DOM node.
Fun, now a few rare numbers on one of our sites seem phone number enough for the dutch locale to break the site. Tell the user to switch to the french locale as workaround, or get any other browser. MS wants you to change the html to say: hey edge, this is not a phoneable number.
WTF, microsoft? Way to shoot yourself in the foot. What else did you hide in there that will bite me one day?
This. I raised an issue where Microsoft fucked up download prompting in IE9 (identified in beta released 2010) and broke ClickOnce launched via Javascript redirection. They haven't fixed it yet as of today (including in Edge!) and rep contacted me and said they deleted the case we had open.
We rewrote the app so it didn't use ClickOnce in the end.
I just heard about that and I'm still kind of puzzled. App Store money, yeah, but isn't Tumblr like 10% SJW kids / fandoms and 90% porn? Seems like they're shooting themselves in the foot.
The main reason developers prefer chrome over firefox is the devtools hands down. When firebug was amazing in the jquery era, we used firefox but now chrome's devtools is simple, fast and full featured to use. If firefox's devtools is as good as or better than chrome's maybe things can change!
I used to think this too, but After seeing some colleagues using FireFox dev tools and making the switch (back) to Firefox myself, I realised that there really isn’t that much between the two and in fact it’s mire a matter of what you’re used to.
Much like the debates about other developer tools in fact (text editors et al).
It's all the small things. Till recently there weren't even react devtools for Firefox. Almost any dev-oriented extension targets chrome first. Chrome has a great devtools protocol, and node conveniently happens to support it too, which is also why many node debugging tools will use chrome.
I use Firefox for casual browsing, but all work-y stuff I do in chrome. Sometimes I'll start debugging a site in Firefox and then some behavior is a bit "off" or missing and I just go back to Chrome.
And the only reason I'm using Firefox is because it's not Chrome. I wanted a different browser that looks a bit different to always be aware of the difference between work and personal.
What specifically do you find better in the Chrome devtools? I find Firefox devtools to be slightly nicer (but I don't really think either one has a massive edge over the other).
1.speed - FF tools always take time to open a script file. Chrome hickups only for huge files which is understandable.
2. Sourcemaps doesnt work most of the times in webpack bundles. Dont know why. (Maybe they tested only for chrome or something.)
3. This new FF devtools was developed after chrome took the dev mindshare and familiarity. I don't see FF devtools very compelling to be used all the time. No disrespect, but there is no incentive or visible improvement in productivity.
4. I'm a react dev now, the react dev tools plugin in FF is laggy. Not pleasant to use.
I can go on like this. But maybe for your use case, FF devtools is better.
For me the FF dev tools are not as good as Chrome's. For the basics I need both browsers offer the same functionality. However, on my machine, FF has performance problems with the dev tools open. Also, when I rearrange the width of the columns in the network tab, after some seconds and without knowing why, the columns return to their original width.
1. speed. On a million+ line code base chrome starts the app in about 5 secs. Firefox over a minute.
2. Blackboxing. Easy to setup and configure in chrome and you can preserve you settings.
3. Blackboxing combined with event breakpoints. Chrome lets you set event break points and combined with black boxing will drop you right into your code on an event. This is awesome.
Just tried a cold start, Firefox took less than 3 secs to start, granted it's on an SSD. What OS are you using? I've heard it's less optimized on a Mac than on Linux.
Not sure about that, because after all, most users use their smartphones and iOS has quite a big market share. So there's at least Chromium and Webkit. Web devs simply won't get away with testing only in Chrome that easily.
I'd like to see microsoft/mozilla/google collaborate on "THE" browser engine then all this compatibility bs goes away - they can all ship their own cruft and advertising but with a single rendering engine we can forget about multiple browsers forever
> It's not just websites that aren't tested in Edge - my experience with Edge is that all aspects of performance (initial load time, ui responsiveness, web site responsiveness) are terribly slow.
100% agree here. I'd be way more worried about monoculture if any other browser would be discontinued, but the MS browser history was and is just a shitshow.
<rant>
I can't even remember how many years MS and even some people in my vincinity were going on about how amazing the new IE (or later Edge) were and that MS would totally be changing their ways now. Usually, if you used their browser for more than five minutes, that sounded more like a bad joke.
To this day, the (properly updated) Edge on my dev machine does user interaction orders of magnitude worse than FF/Chrome. Tabs frequently stop functioning properly and even won't reload anymore, until you find out that some subprocess crashed and you'll have to close and re-open the site to try again. The adoption of new web standards happens at a crawl.
Yet, at the same time, their "Edge is totally a next-gen browser! Promise!" rhetoric leads to actual companies prohibiting users from getting and using an actually useful browser like FF.
Their dev story sucks, too. The best of their jokes was when someone wanted to convince me on how cool VS2015 was for HTML/Javascript development. Yes, I totally want a 16GB+ IDE that literally can take minutes to load from an SSD and frequently freezes to do nimble HTML editing m(
The funny thing is: They actually arguably fixed that one in the mean time (VS Code), but their browser politics remained. Perhaps this signifies the same shift there?
</rant>
TL;DR: I actually wouldn't mind that much, but MS has been repeatedly overselling and underdelivering for years and just slows down everyone else doing it.
TL;DR: I actually wouldn't mind that much, but MS has been repeatedly overselling and underdelivering for years and just slows down everyone else doing it.
I know what you mean, but that is mildly amusing and not entirely false.
I recently made the switch to Firefox from chrome. Webrender is a revolution in web technology since it primarily uses the GPU for rendering resulting in much faster and smoother websites. I'd urge everyone to give webrender a shot, it feels noticeably smoother than chrome.
Yes, webrender is fast, but keep in mind that if you're using webrender you will see bugs, even crashes, that you wouldn't otherwise. Use it only if you're willing to accept that in order to help Mozilla test and improve it.
Disclosure: I work for Mozilla but not on webrender.
It was a complete and totally different time for the web. I was waiting for folks to claim an open source project with outside contributors is the same as a proprietary company moving fast internally and shipping anything and evetything without following a standards process.
> proprietary company moving fast internally and shipping anything and evetything without following a standards process.
Well, that's the thing. What's to stop Google from doing that?
It won't be as overt as Microsoft ramming shitty versions of IE down people's throats for a decade, but please rest assured that Google's goal is profit and domination. I'm not saying they are worse than any other for-profit corporation, but they are a for-profit corporation.
> It was a complete and totally different time for the web
You lack a very basic sense of vision. You understand that the past is different from today, but seem unable to comprehend that the future might be different from today.
20 years ago: Microsoft was doing a pretty good job with the web! IE3/4/5 were consistently better than Netscape. Developers generally welcomed the IE monoculture because developing for both Netscape and IE was a real pain in the ass.
10 years ago: Microsoft had a stranglehold on the web and it suffered greatly, to the point where governments had to intervene (the EU mandating browser choice, etc)
Today: Google is doing a pretty good job with the web!
A modern browser is effectively on par with an Operating System complexity-wise. You need to implement a huge number of APIs (many of which are security-sensitive), you need to have a good UI, you need to be super fast while making all that play together nicely etc...
If you don't have a large team of developers and a few millions in the bank you basically don't have a chance.
As Google keeps pushing more and more features in the browser the bar keeps getting higher and higher. I'm actually surprised that Mozilla still manages to mostly keep up, but since it's mostly running on Google's money it's still not quite a relief.
> A modern browser is effectively on par with an Operating System complexity-wise.
I don't think this is even remotely true. Those same arguments (using apis, good security, good ui, speed, etc) could be made for any piece of non-trivial software. But there's no way a browser is "on par" in complexity with an operating system.
I maintain that. Actually I think a modern browser is probably more complex than many small kernels.
It's probably not the best metric but my checkout of the Linux kernel (a very advanced OS with support for many architectures and devices) is at about 14 million lines of C code (per sloccount), although if you remove device drivers and only count the "core" of the kernel you end up with about 3M lines. I don't have any browser source code available on my computer but a website[1] says that Firefox 20 (released in 2013) was around 4M lines of code and was rising quite fast.
Again, this is comparing apples to kernels but it shows that it's still within the same order of magnitude as far as code size is concerned. The maintenance burden alone on these large codebases is huge, you need teams of engineers just dealing with things like testing and regressions. The bar is extremely high for competition, you won't have two clever engineers write the next killer browser in their basement.
Consider the complexity of supporting all the dozens of CPU instruction sets and a few dozen more built-into-the-CPU hardware peripherals/extensions (e.g. SSE, hardware random number generators, etc). There's probably a thousand different combinations thereof across x86, ARM, MIPS, etc. There's only so many different kinds of hardware out there (and half as much quirky bullshit if you take out all the Sony laptops!)
Now consider how many websites that exist with all kinds broken code, use of dead APIs, old versions of, well, everything. Yet a modern browser can display those sites just fine 99% of the time. Users expect this!
Making a modern browser that works across all those nearly infinite combinations is considerably more difficult than making a kernel boot on a new board with a plethora of datasheets out in the wild to download at your leisure.
I won't really complain if there are just 2 render engines out there in the market, given that both are open source. It will make work hell lot easier for so many developers and the competition and development will stay healthy.
Open source doesn't reduce the issues with people writing to the implementation, not the spec - leading to unintentional interactions and trivia suddenly becoming impossible to change.
If edge dies, and FF shrinks further (both of which seem likely), FF will in effect be reduced to an alternative implementation of chrome; not a implementation of a web browser. Both Edge and FF already include chrome-quirk emulating features; you can expect those exceptions to become the norm.
Apple - for all it's wealth - isn't likely to bother bucking the trend here. The same forces pushing MS affect Apple too; and given their high-end only marketing and various political factors, they will be completely ignored in much of asia - and that's a lot of devs creating a lot of stuff that is likely to depend on chrome-only features eventually.
Safari is leaving itself in the dirt. It's the new IE with slow, quirky, and unreliable features. They made some advancements in privacy, but so did Firefox without giving up on actual rendering and standards.
Google owns Android, which is the Windows of the mobile world. They push a development monoculture based on their platform. They leverage their de-facto monopoly in some sectors to penetrate other sectors. They hoover up young developers and keep them in gilded cages that encompass as much of their lives as possible. The only difference is that their cash-cow is advertising rather than an office suite.
Google IS the new Microsoft. They are MS just before the Halloween Papers and the antitrust trial: rich, dominant, and mostly well-liked by the dev community at large.
Nah, wrong comparison. Microsoft mostly stayed in their corner (Software) while the others (Google, Amazon) are branching out to everything else. Things like Waymo will probably even take over the revenue of pure digital services (though it's tied into). Transportation is a big business.
Better analog would be Wayland-Yutani I guess. The corporations that are manifesting currently are nothing that the world has ever seen before in size and interconnections.
Not that I'm against it, it's just my observation. I'm an avid customer of both.
No, it doesn't even support basic features well, like SVG. The Mask implementation just wrong, you can crash Safari in iOS with a single SVG filter, only one color interpolation is implemented.
But worst of all, it seems like Apple just doesn't care. Bug reports don't seem to be read at all, while when reporting an issue for Chrome you usually get a reply within 24 hours.
I hear this a lot, but I find Safari to be an entirely usable browser day-to-day - the reduction in power usage is very noticeable vs Firefox or Chrome.
Same! I prefer Safari dev tools for inspecting the DOM, CSS prototyping, debugging, seeing request/response headers, local storage info. I use it every day, and find its interface much more intuitive and way less bloated. However Chrome devtools are better when it comes to performance profiling, and having extensions like React devtools make me wish that was a capability of Safari.
Meanwhile Chrome is a bloated resource hog that, despite its resource usage, still scrolls like early Android phones. It's embarrassingly bad compared to Safari's buttery smooth scrolling.
Is it? I find a lot of these performance views are very subjective depending on multitasking, hardware, gpu, extensions, etc. Chrome has always served me well, and the built-in sync and consistency on mac + windows is what I like best. I would rather compete on features and general standards compliance, and then let people choose their own experience on top.
When everything else runs silky smooth, using vanilla Chrome on vanilla macOS feels like a giant slug. For something so fundamental, I find it baffling Google is unable to do it properly.
You're talking about Google's own product. Chromium is open source and all that "bloatedness" can be taken out. There are already many companies out there that use the Chromium rendering engine have already taken the unneeded bits out and customised the browser with their own features.
I don’t think it’s really a reasonable comparison.
Safari does tend to release features a bit later than Firefox or Chrome, but it’s also an open-source, standards-compliant browser that includes almost all modern web tech that the others support. In daily use, I rarely find anything that isn’t supported, with the one exception of issues around WebRTC that have been fixed for some time.
Sorry safari isn’t great but it’s nothing like IE. It’s got an open rendering engine maintained by multiple corporations as well as contributions from others. WebKit the core of safari works on many different operating systems unlike IE that is basically a web view into activex controls for native windows only applications. Not at all comparable
I applaud the death of IE and Edge. One thing I'm worried about though is that Chromium may take over the market and we end up with websites working only in Chromium based browsers and other important browsers like Firefox and Safari get left in the dirt.