Firefox's problem is not lack of features. It's performance.
To be sure, Firefox isn't the worst by many metrics, but it's no longer the stand-out that it was circa 2005-2008. That's why Firefox users left it for Chrome.
And here we are in 2016, I feel that Firefox has a legitimate chance of fighting back if they get their act together. Chrome has become heavy on resource usage while Safari can be sluggish at times. Microsoft has done a great job with Edge but it's bound to Windows 10. This is the opportunity that Firefox should seize.
Yet, they seem to be permanently confused about what they want to do. Chucking Firefox OS was a brave and laudable choice (I say this as a former user, not just an owner, of Firefox OS). But then they work on this? Seriously? Just keep your heads down and concentrate your resources on making Firefox faster and more efficient. Everything else will come with it.
But there are no plans to put Servo in Firefox. Servo is a new embedable rendering engine that can be put in other apps. It's not aiming to replace Gecko in FF.
Browser vendor makes a browser engine, and you believe they don't plan to put it in the browser?
If you do a risky project that might fail or not pay off, and you don't want to get bad PR, you can do only two things: Keep it entirely secret, or develop it in the open but make no promises whatsoever. For an open source company, the choice is pretty easy.
I always thought it was a public secret that of course they're going to put it in the browser, but they don't dare promising that before they know it's going to work out. In fact, I wouldn't be surprised if nextgen Firefox is just going to be Servo + browser.html.
And considering onging projects (webextensions, e10s) it should be fairly obvious that even if they started today it would still be a multi-year project just to get to a point where pain-tolerant users might be willing to use it.
Nothing has changed. The singular goal of Servo is to make a great browser engine. There are lots of things that can be done with a great browser engine. Let's talk about what we can do with a great browser engine once it exists.
The reason why you can't find an answer as to whether Servo will replace Gecko is, in fact, that there is no answer to that right now. Any sort of wild speculation as to whether Servo is going to replace Gecko is extremely premature at this point.
I was just talking this week to one of the guys on the original chrome team. He told me in the beginning they were obsessed with speed -- there was a benchmark in place for time to load from cold boot that wasn't allowed to go up (80ms). So if you wanted to do a new feature you had to jump through hoops to make sure it didn't slow anything down.
Eventually chrome became mainstream, Google started advertising it on their homepage, and features started becoming a priority over speed. At some point the benchmark test case for load time broke and no one fixed it, and things started to slide.
I'm hoping servo provides the necessary investment and juice to restore Firefox to a good browser again (after many, many years).
TBH, I used to hate Firefox's performance, because even loading Facebook froze the entire browser for several msecs (enough to be perceptible), but recently I turned e10s[1] on and the browser feels almost as responsive as Chrome now.
The documentation seems to be outdated, there seems to be an A/B testing right now, and it's recommended to turn off some extensions[2], but this is how I forced e10s to be turned on in `about:config` in the beta:
Thanks for testing e10s! There's a simpler a better way to test it though, which is just setting
`browser.tabs.remote.autostart` to true.
These prefs that you mentioned won't stick (they will be reset after startup) and are just used for some internal state.
I'll make sure the info on that wiki page is up to date for people who want to test it.
Note: Multiprocess is not available in the Release channel yet. Please use our Developer/Test channels if you want to use it.
Also, if you have add-ons, multiprocess is by default blocked from running on Beta (but allowed on Nightly/Aurora). If you're on Beta and want to use it together with add-ons, look at that wiki page to see how to work around that. Or just use Aurora :)
What makes you think Edge is less resource hungry than Chrome? Last time I checked the resources used by Edge it was similar to Chrome and even worse in some cases. Any browser that implements a process-per-tab architecture is going to consume significantly more resources than one that doesn't.
Visiting The Verge, Hacker News and Ars Technica yields the following usage:
You'd think that the fact it sometimes crashes when trying to save a file being downloaded is also a bit of an issue. But here we are, two years in with this problem, every crash diligently reported to Mozilla, but all we get back is Hello, Pocket, Sync, Promoted Tiles and what have you. Priorities are certainly ain't set right somewhere in the development pipeline.
Firefox isn't losing market share because of minor performance differences.
We live in a world now where the browser that comes with your computer/phone/tablet is usually good enough for most people. Firefox isn't even a consideration anymore, and I say that as someone that uses Firefox as their main browser.
Firefox's popularity grew when IE was seen as bad even by non-technical users. There's not much selling points for the average person to replace their default browser anymore.
I've found it pretty handy, especially when doing remote code reviews or preliminary interviews.
It only shares the currently visible tab in a single window, and while enabled, that window gets a big banner and overlay video window that make it very difficult to accidentally share the wrong thing. Plus, WebRTC is encrypted between peers, which would thwart passive eavesdroppers.
Firefox bundles Pocket and this, and these are irremovable, without maintaining one's own builds. Actually, Chromium's default install is better at respecting my freedom than Firefox. This is incompatible with their mottos:
> We make it Firefox. You make it your own. -- The features you love. The privacy you trust. Our most customizable Firefox for Android yet. https://www.mozilla.org/en-US/firefox/android/
IIRC it was not easily disabled when it was first introduced, the control was added after community reaction. However, I just included as another example. Further, I don't know the situation now, but IIRC you can't remove (i.e. delete the code, not disable) zeitgeist and online searches, keeping Unity, those are hard dependencies, even though non-essential.
but the code does not execute unless you interact with it, as can be easily verified by inspecting source. move them out of your tab bar and never click them if it's a big deal to you.
Is this actually a problem that's really crying out for a solution? I mean, I can't say I do collaborative remote planning of trips or shopping online very often at all, but on the rare occasions that I have, sending an email or an instant message of some sort with links in it really does mostly seem to do the trick ("Do you prefer <url> or <url>?" - "I like the first one" - "Great, I'll buy that, thanks").
I feel like that's typical of most new apps/websites coming out. Trying to solve non-problems. That's why we never jumped onto using Slack at work. And now I'm feeling a bit validated about that decision, seeing as just a year later everyone's flocking away from it.
As I state there, Firefox Hello's user experience (before this change) was significantly superior to both Skype and Google Hangouts. But they seem intent on shipping a barely needed feature instead of providing excellent and usable video chat (where no other platform comes close, IMHO).
When I last used it a few weeks ago, the "feature" was a disaster -- every time you switch tabs the other person you're chatting with would see exactly the page you see. This should be an opt-in feature, not a possibly opt-out one! (they claim that in future versions there will be a way to disable tab-sharing... though I don't see it currently)
Imagine your Tesla Model S has an issue where it won't charge. You call your buddy who also has a Model S, and she's having the same issue with hers. So, you call Tesla and tell them, "I can't use my Model S because the battery won't charge!". They're like "no worries, send us the diagnostic information and we'll get it fixed in no time".
In a sigh of relief, you quickly head to your garage and use 1 of the the remaining 3 percent of your battery charge to send diagnostic data to Tesla...
12 weeks later, you're frustrated because you've been driving a sub-standard rental car, and you call Tesla again. "Why haven't you fixed this, yet?!". They reply, "We have! We just finished the update you've been waiting for. Go have a look!".
You run to the garage and plug the car in to the wall. You get in and look at the screen to see if it's charging. Your Model S's computer welcomes you with a message:
Drive with your friends now with Tesla Hello!
You can now see where your friends are while you drive!
Daily firefox user checking in. Every mozilla feature gets me upset because firefox was the poster boy for the underdog. I got behind them completely during the browser wars and have written firefox extensions as well.
But look at them today. Firefox OS. Ads on the initial browser page. Pocket. They were extremely slow to catch up with multi-process tabs. And firefox just freezes up arbitrarily periodically. So very sad.
Performance issues aside, Firefox has not been big on my privacy issues either. Running sync server is so incredibly complicated and it doesn't need to be that way. All it's doing is stash bookmark/password data and the amount of configuration required to run this on your own is mind numbing :/ Just see the wiki pages and they all say the pages need work.
Does anyone else remember Excite.com's chat application from ~20 years ago, which has the ability for people who were all on the same web page could chat amongst each other. I'm pretty sure you could navigate across pages and bring people with you, too. Very early precursor to this. I thought it was a lot of fun back then, but I was 11 years old.
Yes - I definitely remember shared browsing some time back in the mid/late 90's but I couldn't remember the app that did it. Excite doesn't ring a bell though, maybe there were more?
I don't understand what happened to Firefox or how they so comprehensively lost their direction.
The whole rationale of the browser was to be something small, tight and easily extensible and to get away from the unreliability of Netscape.
Now we learn they're deprecating full themes, changing plugins to be more limited (but again becoming more like Chrome), and filling the browser with pointless at best features. I can't recall ever wanting to share browsing. Screens when coding or helping sure, but it's a solved problem, so keep it and the kitchen sink out of the browser.
It's probably why its becoming so slow and flaky. The two key problems it was intended to fix.
> I see both version of this statement all the time: "Chrome is so slow. I switched to firefox![1]" "Firefox is so slow. I switched to chrome!"
I think it's simply that anecdotes are extremely unreliable when it comes to something as large and complex as a browser that's so sensitive to the specific content you browse. Firefox is slow from time to time for some people; Chrome is slow from time to time for some people.
Well I'm sure profiles are a part of it. It got bad enough that I completely removed all profile and plugin data, and clean reinstalled. Firefox remains the worst performing browser I have on this system by very large measure.
Five amazon tabs are enough to basically grind it to a halt. Chrome, IE, Opera and Palemoon (a fork of an older firefox) don't do that.
So I've reached a point where I can't do much more to make it usable without a debugger. I've no doubt others are fine, or find Chrome the slow one. If everyone had the issues I currently have it'd be the single highest priority bugfix in the project.
I'd maybe migrate to palemoon but I don't trust it on a security front.
Lots of negativity in these comments surrounding Firefox and Mozilla's direction with it. I've been using Rust, a Mozilla product, quite a bit recently and have even contributed to the project. I am throughly impressed with Rust, the language, the ecosystem, and most importantly the way it's maintained by Mozilla. It's an incredible gift by Mozilla to the community at large, and for that they've earned a lot of respect from me.
I agree that Firefox's performance is quite bad these days, but Mozilla is already working on that (with Servo). Pouring more devs into it is not likely to make progress on that front faster (too many cooks). They could have taken a more gradual approach to the problem, by slowly evolving the existing codebase, chipping away at performance. But they've been trying that for a long time. Servo is, in my opinion, the better plan. And with it also comes a massive security hardening. It's a long road, but one worth waiting for.
So, Mozilla has already dedicated the necessary resources to handle performance. What are they to do with the rest of their company's resources, and most importantly, what can they do to grow the company? Firefox Hello and other projects like it are part of that effort. We may not agree that Firefox Hello is a good idea. Personally I don't have a use for it either. But bless them for trying and experimenting.
And let's not forget that Firefox is a "free" product. I can't imagine Mozilla is swimming in money. They don't have the resources that Google or Microsoft have, and yet they still have a solid chance at leap-frogging their competition with Servo.
Personally, I continue to use Firefox as my daily driver, because it appears to respect my privacy more than Chrome, despite its performance issues.
So, Mozilla has already dedicated the necessary resources to handle performance. What are they to do with the rest of their company's resources, and most importantly, what can they do to grow the company?
Mozilla recently jettisoned Thunderbird, its email client, mostly because of its impact on Firefox development (https://groups.google.com/forum/#!msg/mozilla.governance/kAy...). Wouldn't a great native email client be a better way to grow the company than features like this? And shouldn't features be added to make the product better, not just because there are people sitting around with nothing else to do?
> I agree that Firefox's performance is quite bad these days, but Mozilla is already working on that (with Servo).
I've checked out servo and compiled and started it in release mode - it required 600MB RAM and 700% CPU to load google(!) with the same speed as my little firefox with tons of extensions - forgive my ignorance but I've expected more from the hype. I've tested some other websites but no luck - Servo is a heavy and slow beast.
> I am throughly impressed with Rust, the language, the ecosystem, and most importantly the way it's maintained by Mozilla.
Rust programs consume so much resources that it could just use a GC instead of the complex manual memory management. The syntax and the standard library is as 'modern' as in cpp - I'm not as impressed as I was a fan of Mozilla.
> Personally, I continue to use Firefox as my daily driver, because it appears to respect my privacy more than Chrome, despite its performance issues.
Search the internet for optimizations, my firefox feels like as fast as chrome - sometime even faster. It just needs some tweaks. Even if ff would be much slower I wouldn't leave it because of firefox's wonderful extension platform and its nice complete theme suites.
> Rust programs consume so much resources that it could
> just use a GC
I'm not sure where this is coming from. Everyone that I've spoken to about using Rust in place of managed languages has noted incredible reductions in memory usage. Indeed, this was one of the primary reasons for Dropbox adopting Rust for their core storage engine. Quoting from https://news.ycombinator.com/item?id=11283688 :
"The reasons for using rust were many, but memory was one of them. Primarily, for this particular project, the heap size is the issue. One of the games in this project is optimizing how little memory and compute you can use to manage 1GB (or 1PB) of data. We utilize lots of tricks like perfect hash tables, extensive bit-packing, etc. Lots of odd, custom, inline and cache-friendly data structures. We also keeps lots of things on the stack when we can to take pressure off the VM system. We do some lockfree object pooling stuff for big byte vectors, which are common allocations in a block storage system. It's much easier to do these particular kinds of optimizations using C++ or Rust."
And the above paragraph is using Go as a baseline, which itself already tends to use less memory than Java.
> I've checked out servo and compiled and started it in release mode - it required 600MB RAM and 700% CPU to load google(!) with the same speed as my little firefox with tons of extensions - forgive my ignorance but I've expected more from the hype.
Thanks for the bug report, but I can't reproduce. According to the built-in memory profiler, Servo uses 39 MB of explicit allocations to load google.com, which is well in line with other browsers even without a whole lot of optimization for memory use.
A large amount of CPU usage during initial pageload, on the other hand, is a feature, not a bug. Using more resources lets us get the page loaded faster, which not only shows you the page faster but also typically reduces power consumption since we can get the CPU into an idle state.
> Rust programs consume so much resources that it could just use a GC instead of the complex manual memory management.
Manual memory management isn't about some broad notion of "resource consumption": it's about having low-level control of resources, which can not only reduce memory usage but also eliminate pauses. I think it's very unlikely that the JavaScript garbage collector could be made thread-safe without regressing performance (which is what you presumably mean by "just use a GC").
> The syntax and the standard library is as 'modern' as in cpp - I'm not as impressed as I was a fan of Mozilla.
Just to name the first features that come to mind: Rust has a module system and hygienic macros, which are syntactic features not in C++. Rust comes with a standard library that includes networking, which isn't in C++.
> Search the internet for optimizations, my firefox feels like as fast as chrome - sometime even faster. It just needs some tweaks.
Those "tweaks" are in fact likely to be slowing down your browser.
> Thanks for the bug report, but I can't reproduce. According to the built-in memory profiler, Servo uses 39 MB of explicit allocations to load google.com, which is well in line with other browsers even without a whole lot of optimization for memory use.
If you say so, I'll check out a new version and try it again. I've tried it a few weeks ago on linux.
> A large amount of CPU usage during initial pageload, on the other hand, is a feature, not a bug. Using more resources lets us get the page loaded faster, which not only shows you the page faster but also typically reduces power consumption since we can get the CPU into an idle state.
Large CPU usage slows down other apps - that's one reason why I use ff instead of chrome. 8GB RAM seems too few nowadays...
> Manual memory management isn't about some broad notion of "resource consumption": it's about having low-level control of resources, which can not only reduce memory usage but also eliminate pauses. I think it's very unlikely that the JavaScript garbage collector could be made thread-safe without regressing performance (which is what you presumably mean by "just use a GC").
I was pointing to the promise that it shouldn't leak while it seems to be doing it seriously.
> Just to name the first features that come to mind: Rust has a module system and hygienic macros, which are syntactic features not in C++. Rust comes with a standard library that includes networking, which isn't in C++.
I don't really like Rust's module system but +1 for the hygienic macros.
> Those "tweaks" are in fact likely to be slowing down your browser.
Have you heard about experimental features like pipelining and prefetch? Or content restrictors like uBlock and disconnect? Or session+history limits? When I reset my ff settings(due to ff profile problems) the vanilla ff takes at least 2x as much time to load the same page. If these "tweaks" wouldn't present - I wouldn't be a happy ff user.
> Large CPU usage slows down other apps - that's one reason why I use ff instead of chrome.
The idea that apps should intentionally throttle their CPU usage to avoid "slowing down other apps" at the expense of performance is nonsensical. Apps should use all the resources in the machine to get the job done faster and go to sleep faster. The kernel, not userland apps, is responsible for multiplexing the system resources: that's literally its job, and only it is in a position to do so properly.
> I was pointing to the promise that it shouldn't leak while it seems to be doing it seriously.
Nothing you've cited has been evidence of a leak. Adding a global garbage collector would do nothing to reduce the (non-)"leaks" you observe.
> Have you heard about experimental features like pipelining and prefetch?
Enabling those via about:config tweaks and addons slows down your browser. Session and history limits do not affect your browser performance.
> The idea that apps should intentionally throttle their CPU usage to avoid "slowing down other apps" at the expense of performance is nonsensical. Apps should use all the resources in the machine to get the job done faster and go to sleep faster. The kernel, not userland apps, is responsible for multiplexing the system resources: that's literally its job, and only it is in a position to do so properly.
Did you knew that there are apps running - like IDEs, servers, games etc which requires a large amount of resources? If you run another resource hungry app it might affect your first app's runtime which would end up frozing both. Have you ever used resource hungry apps?
> Nothing you've cited has been evidence of a leak. Adding a global garbage collector would do nothing to reduce the (non-)"leaks" you observe.
600MB RAM for nothing is leak. Denial is futile.
> Enabling those via about:config tweaks and addons slows down your browser. Session and history limits do not affect your browser performance.
1. It seems like you don't have experience with browsers. Go and try them out instead of denying everything. Do you know what is prefetch at all? Or at least pipelining(cheat: https://en.wikipedia.org/wiki/HTTP_pipelining)?
2. Session and history limits DOES affect browser performance at long runtime. If you limit these, your browser will consume less RAM(+ it's easier to find in small history) and if another app wants to use its resources the os will put the memory on the swap then it will come back faster.
I really don't like this sort of comments, trying to make us feel guilty for criticising something. The stupidity of big browser producers makes the web worse for everyone, so it should not be allowed, even if done with good will.
About since 1992 there is a very good tool for sharing tabs: their URLs. Anything that can't be shared with a URL is probably too private to be shared anyways.
I'm a user, I don't care about Rust. Firefox may well be implemented in some object oriented brainfuck, but as long as it does not tuck into its interface irrelevant stuff that I can't remove, I won't touch Firefox. I actually recently considered a switch, and tried FF45, and what I saw: Pocket and this Hello, there, impossible to actually remove. I'd be fine with an addon that they officially developed/backed, and it'd be okay if they advertised it a bit, but this way, no. They certainly lost some people other than me with this inconsiderate manners of theirs.
I'm thankful to Mozilla to produce multi-platform software libre, but I also have some comments about their current priorities.
I think Firefox should try to gain users again with a good browser that it's fast and has great features expected in a browser. In this chart you can see how Chrome users increments and Firefox users decrement [1].
A feature to chat, call and share my screen it's not a feature that I expect from a browser. A button to store bookmarks in a third-party service it's not a feature that I expect in a browser.
A good printing functionality or to save a page as PDF it's something that I would expect from a browser (you can save a page as PDF in Firefox for Android but not in desktop).
Last year I made a little web project[2] that would take a group of images and generate a printable calendar. I'm fluent in CSS and HTML and it was pretty easy to make. To my surprise, Firefox has the same print dialog like from Firefox 1.0. I wasn't able to generate a decent pdf from their printing dialog, which was extremely basic and without a preview support.
Chrome and Safari were pretty good and I was able to get very good results. They both show you a preview of the page before sending it to print.
I am surprised to see that Firefox invests time and effort in things like VR[3] while lacking a good printing experience (I think a good printing experience would help lots of more people in concrete ways).
Finally, after seeing how they abandon projects like Persona, FirefoxOS or Thunderbird, I'm not very prone to invest time in their projects that maybe they will end up abandoning (I have the same feeling with Google).
Anyway, this is intended as feedback and hoping that they will start listening users about what they want from their browsers. Right now seems the opposite, Mozilla creative people trying to convince users that they want the things that they come up with.
77 comments
[ 3.5 ms ] story [ 142 ms ] threadTo be sure, Firefox isn't the worst by many metrics, but it's no longer the stand-out that it was circa 2005-2008. That's why Firefox users left it for Chrome.
And here we are in 2016, I feel that Firefox has a legitimate chance of fighting back if they get their act together. Chrome has become heavy on resource usage while Safari can be sluggish at times. Microsoft has done a great job with Edge but it's bound to Windows 10. This is the opportunity that Firefox should seize.
Yet, they seem to be permanently confused about what they want to do. Chucking Firefox OS was a brave and laudable choice (I say this as a former user, not just an owner, of Firefox OS). But then they work on this? Seriously? Just keep your heads down and concentrate your resources on making Firefox faster and more efficient. Everything else will come with it.
I'm pretty sure Firefox will be faster than any other browser once Servo is used inside Firefox instead of Gecko, check this out:
https://news.ycombinator.com/item?id=11175258
If you do a risky project that might fail or not pay off, and you don't want to get bad PR, you can do only two things: Keep it entirely secret, or develop it in the open but make no promises whatsoever. For an open source company, the choice is pretty easy.
I always thought it was a public secret that of course they're going to put it in the browser, but they don't dare promising that before they know it's going to work out. In fact, I wouldn't be surprised if nextgen Firefox is just going to be Servo + browser.html.
The reason why you can't find an answer as to whether Servo will replace Gecko is, in fact, that there is no answer to that right now. Any sort of wild speculation as to whether Servo is going to replace Gecko is extremely premature at this point.
Bugs to watch:
- https://bugzil.la/1135640 - (oxidation) Add rust compiler support
- https://bugzil.la/1243581 - (stylo) implement support for Servo's style system in Gecko
- https://bugzil.la/1151899 - Integrate the rust-url parser into necko
Such is detailed at Servo's roadmap. https://github.com/servo/servo/wiki/Roadmap
Specficially:
>> Internally committed 2016 goals
>> Ship one Rust component in Firefox Nightly, riding the trains
>> Experiment with the uplift of a major piece of Servo into Gecko
Eventually chrome became mainstream, Google started advertising it on their homepage, and features started becoming a priority over speed. At some point the benchmark test case for load time broke and no one fixed it, and things started to slide.
I'm hoping servo provides the necessary investment and juice to restore Firefox to a good browser again (after many, many years).
The documentation seems to be outdated, there seems to be an A/B testing right now, and it's recommended to turn off some extensions[2], but this is how I forced e10s to be turned on in `about:config` in the beta:
[1] https://wiki.mozilla.org/Electrolysis[2] https://www.arewee10syet.com/
These prefs that you mentioned won't stick (they will be reset after startup) and are just used for some internal state.
I'll make sure the info on that wiki page is up to date for people who want to test it.
Note: Multiprocess is not available in the Release channel yet. Please use our Developer/Test channels if you want to use it.
Also, if you have add-ons, multiprocess is by default blocked from running on Beta (but allowed on Nightly/Aurora). If you're on Beta and want to use it together with add-ons, look at that wiki page to see how to work around that. Or just use Aurora :)
Visiting The Verge, Hacker News and Ars Technica yields the following usage:
Something went terribly wrong along the way...
No, the real reason is Google's highly aggressive 'bundling' (i.e. crapware) strategy.
We live in a world now where the browser that comes with your computer/phone/tablet is usually good enough for most people. Firefox isn't even a consideration anymore, and I say that as someone that uses Firefox as their main browser.
Firefox's popularity grew when IE was seen as bad even by non-technical users. There's not much selling points for the average person to replace their default browser anymore.
It only shares the currently visible tab in a single window, and while enabled, that window gets a big banner and overlay video window that make it very difficult to accidentally share the wrong thing. Plus, WebRTC is encrypted between peers, which would thwart passive eavesdroppers.
> Committed to you, your privacy and an open Web. https://www.mozilla.org/en-US/firefox/desktop
> We make it Firefox. You make it your own. -- The features you love. The privacy you trust. Our most customizable Firefox for Android yet. https://www.mozilla.org/en-US/firefox/android/
This is like the case of online searches on Unity/Ubuntu. A non-essential, irremovable (non-destructively), annoying bit of software on my system.
As I state there, Firefox Hello's user experience (before this change) was significantly superior to both Skype and Google Hangouts. But they seem intent on shipping a barely needed feature instead of providing excellent and usable video chat (where no other platform comes close, IMHO).
When I last used it a few weeks ago, the "feature" was a disaster -- every time you switch tabs the other person you're chatting with would see exactly the page you see. This should be an opt-in feature, not a possibly opt-out one! (they claim that in future versions there will be a way to disable tab-sharing... though I don't see it currently)
</rant>
Hello only supports 1:1 chats, not group ones, so this may not work immediately for you.
(Edit: Link removed. It was nice meeting those of you who dropped by!)
Why? Keep that garbage out of the browser. Everything is too "social" and integrated already.
Just give me a "dumb" viewer - nothing more, nothing less.
In a sigh of relief, you quickly head to your garage and use 1 of the the remaining 3 percent of your battery charge to send diagnostic data to Tesla...
12 weeks later, you're frustrated because you've been driving a sub-standard rental car, and you call Tesla again. "Why haven't you fixed this, yet?!". They reply, "We have! We just finished the update you've been waiting for. Go have a look!".
You run to the garage and plug the car in to the wall. You get in and look at the screen to see if it's charging. Your Model S's computer welcomes you with a message:
Now, your battery is at 0.5%.That's Firefox.
[1] https://www.mozilla.org/en-US/firefox/os/ [2] https://login.persona.org/
But look at them today. Firefox OS. Ads on the initial browser page. Pocket. They were extremely slow to catch up with multi-process tabs. And firefox just freezes up arbitrarily periodically. So very sad.
Performance issues aside, Firefox has not been big on my privacy issues either. Running sync server is so incredibly complicated and it doesn't need to be that way. All it's doing is stash bookmark/password data and the amount of configuration required to run this on your own is mind numbing :/ Just see the wiki pages and they all say the pages need work.
I thought it was pretty useless back then...
The whole rationale of the browser was to be something small, tight and easily extensible and to get away from the unreliability of Netscape.
Now we learn they're deprecating full themes, changing plugins to be more limited (but again becoming more like Chrome), and filling the browser with pointless at best features. I can't recall ever wanting to share browsing. Screens when coding or helping sure, but it's a solved problem, so keep it and the kitchen sink out of the browser.
It's probably why its becoming so slow and flaky. The two key problems it was intended to fix.
I see both version of this statement all the time: "Chrome is so slow. I switched to firefox![1]" "Firefox is so slow. I switched to chrome!"
What's going on? Do browsers get full of history/cookies/profiles/plugins and slow down.
Is it just the act of switching that speeds things up?
[1] http://gizmodo.com/fuck-it-im-going-back-to-firefox-16854258...
I think it's simply that anecdotes are extremely unreliable when it comes to something as large and complex as a browser that's so sensitive to the specific content you browse. Firefox is slow from time to time for some people; Chrome is slow from time to time for some people.
Five amazon tabs are enough to basically grind it to a halt. Chrome, IE, Opera and Palemoon (a fork of an older firefox) don't do that.
So I've reached a point where I can't do much more to make it usable without a debugger. I've no doubt others are fine, or find Chrome the slow one. If everyone had the issues I currently have it'd be the single highest priority bugfix in the project.
I'd maybe migrate to palemoon but I don't trust it on a security front.
I agree that Firefox's performance is quite bad these days, but Mozilla is already working on that (with Servo). Pouring more devs into it is not likely to make progress on that front faster (too many cooks). They could have taken a more gradual approach to the problem, by slowly evolving the existing codebase, chipping away at performance. But they've been trying that for a long time. Servo is, in my opinion, the better plan. And with it also comes a massive security hardening. It's a long road, but one worth waiting for.
So, Mozilla has already dedicated the necessary resources to handle performance. What are they to do with the rest of their company's resources, and most importantly, what can they do to grow the company? Firefox Hello and other projects like it are part of that effort. We may not agree that Firefox Hello is a good idea. Personally I don't have a use for it either. But bless them for trying and experimenting.
And let's not forget that Firefox is a "free" product. I can't imagine Mozilla is swimming in money. They don't have the resources that Google or Microsoft have, and yet they still have a solid chance at leap-frogging their competition with Servo.
Personally, I continue to use Firefox as my daily driver, because it appears to respect my privacy more than Chrome, despite its performance issues.
Mozilla recently jettisoned Thunderbird, its email client, mostly because of its impact on Firefox development (https://groups.google.com/forum/#!msg/mozilla.governance/kAy...). Wouldn't a great native email client be a better way to grow the company than features like this? And shouldn't features be added to make the product better, not just because there are people sitting around with nothing else to do?
I've checked out servo and compiled and started it in release mode - it required 600MB RAM and 700% CPU to load google(!) with the same speed as my little firefox with tons of extensions - forgive my ignorance but I've expected more from the hype. I've tested some other websites but no luck - Servo is a heavy and slow beast.
> I am throughly impressed with Rust, the language, the ecosystem, and most importantly the way it's maintained by Mozilla.
Rust programs consume so much resources that it could just use a GC instead of the complex manual memory management. The syntax and the standard library is as 'modern' as in cpp - I'm not as impressed as I was a fan of Mozilla.
> Personally, I continue to use Firefox as my daily driver, because it appears to respect my privacy more than Chrome, despite its performance issues.
Search the internet for optimizations, my firefox feels like as fast as chrome - sometime even faster. It just needs some tweaks. Even if ff would be much slower I wouldn't leave it because of firefox's wonderful extension platform and its nice complete theme suites.
"The reasons for using rust were many, but memory was one of them. Primarily, for this particular project, the heap size is the issue. One of the games in this project is optimizing how little memory and compute you can use to manage 1GB (or 1PB) of data. We utilize lots of tricks like perfect hash tables, extensive bit-packing, etc. Lots of odd, custom, inline and cache-friendly data structures. We also keeps lots of things on the stack when we can to take pressure off the VM system. We do some lockfree object pooling stuff for big byte vectors, which are common allocations in a block storage system. It's much easier to do these particular kinds of optimizations using C++ or Rust."
And the above paragraph is using Go as a baseline, which itself already tends to use less memory than Java.
Thanks for the bug report, but I can't reproduce. According to the built-in memory profiler, Servo uses 39 MB of explicit allocations to load google.com, which is well in line with other browsers even without a whole lot of optimization for memory use.
A large amount of CPU usage during initial pageload, on the other hand, is a feature, not a bug. Using more resources lets us get the page loaded faster, which not only shows you the page faster but also typically reduces power consumption since we can get the CPU into an idle state.
> Rust programs consume so much resources that it could just use a GC instead of the complex manual memory management.
Manual memory management isn't about some broad notion of "resource consumption": it's about having low-level control of resources, which can not only reduce memory usage but also eliminate pauses. I think it's very unlikely that the JavaScript garbage collector could be made thread-safe without regressing performance (which is what you presumably mean by "just use a GC").
> The syntax and the standard library is as 'modern' as in cpp - I'm not as impressed as I was a fan of Mozilla.
Just to name the first features that come to mind: Rust has a module system and hygienic macros, which are syntactic features not in C++. Rust comes with a standard library that includes networking, which isn't in C++.
> Search the internet for optimizations, my firefox feels like as fast as chrome - sometime even faster. It just needs some tweaks.
Those "tweaks" are in fact likely to be slowing down your browser.
If you say so, I'll check out a new version and try it again. I've tried it a few weeks ago on linux.
> A large amount of CPU usage during initial pageload, on the other hand, is a feature, not a bug. Using more resources lets us get the page loaded faster, which not only shows you the page faster but also typically reduces power consumption since we can get the CPU into an idle state.
Large CPU usage slows down other apps - that's one reason why I use ff instead of chrome. 8GB RAM seems too few nowadays...
> Manual memory management isn't about some broad notion of "resource consumption": it's about having low-level control of resources, which can not only reduce memory usage but also eliminate pauses. I think it's very unlikely that the JavaScript garbage collector could be made thread-safe without regressing performance (which is what you presumably mean by "just use a GC").
I was pointing to the promise that it shouldn't leak while it seems to be doing it seriously.
> Just to name the first features that come to mind: Rust has a module system and hygienic macros, which are syntactic features not in C++. Rust comes with a standard library that includes networking, which isn't in C++.
I don't really like Rust's module system but +1 for the hygienic macros.
> Those "tweaks" are in fact likely to be slowing down your browser.
Have you heard about experimental features like pipelining and prefetch? Or content restrictors like uBlock and disconnect? Or session+history limits? When I reset my ff settings(due to ff profile problems) the vanilla ff takes at least 2x as much time to load the same page. If these "tweaks" wouldn't present - I wouldn't be a happy ff user.
The idea that apps should intentionally throttle their CPU usage to avoid "slowing down other apps" at the expense of performance is nonsensical. Apps should use all the resources in the machine to get the job done faster and go to sleep faster. The kernel, not userland apps, is responsible for multiplexing the system resources: that's literally its job, and only it is in a position to do so properly.
> I was pointing to the promise that it shouldn't leak while it seems to be doing it seriously.
Nothing you've cited has been evidence of a leak. Adding a global garbage collector would do nothing to reduce the (non-)"leaks" you observe.
> Have you heard about experimental features like pipelining and prefetch?
Enabling those via about:config tweaks and addons slows down your browser. Session and history limits do not affect your browser performance.
Did you knew that there are apps running - like IDEs, servers, games etc which requires a large amount of resources? If you run another resource hungry app it might affect your first app's runtime which would end up frozing both. Have you ever used resource hungry apps?
> Nothing you've cited has been evidence of a leak. Adding a global garbage collector would do nothing to reduce the (non-)"leaks" you observe.
600MB RAM for nothing is leak. Denial is futile.
> Enabling those via about:config tweaks and addons slows down your browser. Session and history limits do not affect your browser performance.
1. It seems like you don't have experience with browsers. Go and try them out instead of denying everything. Do you know what is prefetch at all? Or at least pipelining(cheat: https://en.wikipedia.org/wiki/HTTP_pipelining)?
2. Session and history limits DOES affect browser performance at long runtime. If you limit these, your browser will consume less RAM(+ it's easier to find in small history) and if another app wants to use its resources the os will put the memory on the swap then it will come back faster.
About since 1992 there is a very good tool for sharing tabs: their URLs. Anything that can't be shared with a URL is probably too private to be shared anyways.
I'm a user, I don't care about Rust. Firefox may well be implemented in some object oriented brainfuck, but as long as it does not tuck into its interface irrelevant stuff that I can't remove, I won't touch Firefox. I actually recently considered a switch, and tried FF45, and what I saw: Pocket and this Hello, there, impossible to actually remove. I'd be fine with an addon that they officially developed/backed, and it'd be okay if they advertised it a bit, but this way, no. They certainly lost some people other than me with this inconsiderate manners of theirs.
about:config
loop.enabled = false -> disable hello browser.pocket.enabled = false-> disable pocket
> They certainly lost some people other than me with this inconsiderate manners of theirs.
They're wasting time and resources into things the users don't really need.
why all the whining here?
I think Firefox should try to gain users again with a good browser that it's fast and has great features expected in a browser. In this chart you can see how Chrome users increments and Firefox users decrement [1].
A feature to chat, call and share my screen it's not a feature that I expect from a browser. A button to store bookmarks in a third-party service it's not a feature that I expect in a browser.
A good printing functionality or to save a page as PDF it's something that I would expect from a browser (you can save a page as PDF in Firefox for Android but not in desktop).
Last year I made a little web project[2] that would take a group of images and generate a printable calendar. I'm fluent in CSS and HTML and it was pretty easy to make. To my surprise, Firefox has the same print dialog like from Firefox 1.0. I wasn't able to generate a decent pdf from their printing dialog, which was extremely basic and without a preview support.
Chrome and Safari were pretty good and I was able to get very good results. They both show you a preview of the page before sending it to print.
I am surprised to see that Firefox invests time and effort in things like VR[3] while lacking a good printing experience (I think a good printing experience would help lots of more people in concrete ways).
Finally, after seeing how they abandon projects like Persona, FirefoxOS or Thunderbird, I'm not very prone to invest time in their projects that maybe they will end up abandoning (I have the same feeling with Google).
Anyway, this is intended as feedback and hoping that they will start listening users about what they want from their browsers. Right now seems the opposite, Mozilla creative people trying to convince users that they want the things that they come up with.
[1](https://en.wikipedia.org/wiki/Usage_share_of_web_browsers#/m...)
[2](https://github.com/sanbor/printable-calendar)
[3](http://mozvr.com/)