Support Forum discussions

Does Firefox have a fundamentally structural and architectural problem?

  1. Below is a very quick "grab" of some posts relating to the issues that have dominated the Firefox Support Forum in the last week. When you look at the overall content a possibly startling factor jumps out of the web pages. Consider this: Included below are very few of the very many examples of more of the same which all relate to four typical errors - browser crashes, page loading issues, dropped internet connections and badly hanging/freezing screens. They all point to the one thing. Memory depletion.

    Does Firefox have a fundamentally structural and architectural problem which hogs memory? If it has been a long-standing issue, which one poster claims that it has, (first post at the top of the list - credibility unknown), then it can become costly to change once implemented. Maybe this is what we are staring at, which is horribly embarrassing for Mozilla. Not something that it would want published if it be the case. Worse, there is nothing we can do in the forums to help the clients re memory hungry issues. I hope this is not the case, but looking at the consistency of error, all being the same upon same, upon same, upon same it really does look like an across the board memory problem.

    I hope I am wrong. Just checking the titles alone to these posts below is cause for concern. What do others think?

    https://support.mozilla.org/en-US/questions/1072363 https://support.mozilla.org/en-US/questions/1073459 https://support.mozilla.org/en-US/questions/1072517 https://support.mozilla.org/en-US/questions/1073448 https://support.mozilla.org/en-US/questions/1073441 https://support.mozilla.org/en-US/questions/1070481 https://support.mozilla.org/en-US/questions/1073394 https://support.mozilla.org/en-US/questions/1073386 https://support.mozilla.org/en-US/questions/1072930 https://support.mozilla.org/en-US/questions/1071950

    https://support.mozilla.org/en-US/questions/1070481 https://support.mozilla.org/en-US/questions/1071916 https://support.mozilla.org/en-US/questions/1071865 https://support.mozilla.org/en-US/questions/1073390 (May be a security software problem) https://support.mozilla.org/en-US/questions/1073396 https://support.mozilla.org/en-US/questions/1072214 https://support.mozilla.org/en-US/questions/1073329 https://support.mozilla.org/en-US/questions/1071802


    I will close thIs thread as it is rather long and drifting over related subjects. I will add some comments to serve as a quick indication of its content

    Below is a very quick "grab" of some posts relating to the issues that have dominated the Firefox Support Forum in the last week. When you look at the overall content a '''''possibly''''' startling factor jumps out of the web pages. Consider this: Included below are very few of the very many examples of more of the same which all relate to four typical errors - browser crashes, page loading issues, dropped internet connections and badly hanging/freezing screens. They all point to the one thing. Memory depletion. Does Firefox have a fundamentally structural and architectural problem which hogs memory? If it has been a long-standing issue, which one poster claims that it has, ''(first post at the top of the list - credibility unknown)'', then it can become costly to change once implemented. Maybe this is what we are staring at, which is horribly embarrassing for Mozilla. Not something that it would want published if it be the case. Worse, there is nothing we can do in the forums to help the clients re memory hungry issues. I hope this is not the case, but looking at the consistency of error, all being the same upon same, upon same, upon same it really does look like an across the board memory problem. I hope I am wrong. Just checking the titles alone to these posts below is cause for concern. What do others think? https://support.mozilla.org/en-US/questions/1072363 https://support.mozilla.org/en-US/questions/1073459 https://support.mozilla.org/en-US/questions/1072517 https://support.mozilla.org/en-US/questions/1073448 https://support.mozilla.org/en-US/questions/1073441 https://support.mozilla.org/en-US/questions/1070481 https://support.mozilla.org/en-US/questions/1073394 https://support.mozilla.org/en-US/questions/1073386 https://support.mozilla.org/en-US/questions/1072930 https://support.mozilla.org/en-US/questions/1071950 https://support.mozilla.org/en-US/questions/1070481 https://support.mozilla.org/en-US/questions/1071916 https://support.mozilla.org/en-US/questions/1071865 https://support.mozilla.org/en-US/questions/1073390 (May be a security software problem) https://support.mozilla.org/en-US/questions/1073396 https://support.mozilla.org/en-US/questions/1072214 https://support.mozilla.org/en-US/questions/1073329 https://support.mozilla.org/en-US/questions/1071802 -------- I will close thIs thread as it is rather long and drifting over related subjects. I will add some comments to serve as a quick indication of its content * [https://support.mozilla.org/en-US/forums/support-forum-contributors/711430?last=66241&page=1#post-66103 #post-66103 ] Mark's comprehensive answer, including mention of improvements and work in progress * [https://support.mozilla.org/en-US/forums/support-forum-contributors/711430?last=66241&page=1#post-66134 #post-66134] Tyler's comment on Firefox being generally stable, and listing likely causes of issues. * [https://support.mozilla.org/en-US/forums/support-forum-contributors/711430?last=66241&page=1#post-66166 #post-66166 ] Tyler's comments regarding HelpDesk and on addon signing * [https://support.mozilla.org/en-US/forums/support-forum-contributors/711430?last=66241&page=2#post-66185 #post-66185] Madalina's comments on the escalation process * [https://support.mozilla.org/en-US/forums/support-forum-contributors/711430?last=66241&page=2#post-66181 #post-66181] Dingeroo's start of comments on own issues with malware and crashes ** Follow in https://support.mozilla.org/en-US/questions/1074326

    Modified by John99 on

  2. Thanks for helping out, Dingeroo! And double thanks for going above and beyond and reporting a trend in support cases. We really appreciate it when contributors take a critical eye and show interest in improving things beyond just answering questions.

    The short answer is, yes Firefox has a handful of well known architectural problems which contribute to crashes, slowness, and high memory usage. In keeping with Mozilla's dedicatation to openness and transparency, they make no attempt to hide these problems.

    Now for a longer answer: When it was first created Firefox was a best-in-class piece of software, with a unique multi-platform architecture which blew away all other browsers at the time. Even 5 years is a long time for one piece of software to be incrementally updated. But Firefox is 12+ years old right now. Architectural choices made 10 or 12 years ago may have been great for the time, but now prove difficult to change. However, the problems are well known and we are working on fixing many of them by seriously and fundamentally changing certain parts of the architecture of Firefox.

    Here's an example of three of these problems which are just about to get better:

    1. The next version of Firefox, Firefox 40, will also be the official first release of a 64 bit version of Firefox for Windows. Not being a 64bit piece of software is a huge contributor to crashes, slowness, and memory miss-management on Windows (which the majority of FF users use). If you're on Windows, you can get the 64bit installer for Firefox Developer Edition here: https://www.mozilla.org/en-US/firefox/developer/all/

    2. Unlike newer browsers, Firefox runs as a single monolithic process. This makes memory management almost impossibly difficult, and means a crash in one component can easily crash the whole browser. A project to re-architect Firefox to use a different process for each tab, called Electrolysis (or e10s for short), has been underway for some time. E10s is almost ready for prime-time, and is now turned-on by default in Firefox Developer Edition (formerly Alpha/Aurora) and Nightly.

    I actually run Developer Edition and the normal current Firefox at the same time every day. I use Developer Edition for work, and the current/normal version of Firefox for my personal things. After having "normal" Firefox open for 3 hours, with 5 tabs, and having hardly used it, it is using almost 1738mb of memory. But Firefox Developer Edition, which has been open for 7 hours, with 30 tabs, and I've been using constantly, is currently using just 342mb of memory. If I were using the 64bit version of Dev Edition it would probably be doing even better. More about that here: https://wiki.mozilla.org/E10s

    3. Another big problem is malicious third-party add-ons. Firefox 40 also starts the first part of a 3 release (40,41,42) roll-out plan for add-on quality improvement which begins with warning users when add-ons have not been approved and properly "signed." More about that here: https://wiki.mozilla.org/Addons/Extension_Signing

    There are probably another 10 or so big contributors to the most common memory and crash related issues on the forums. Rest assured that we are well aware and working towards fixes.

    Again, thanks a ton for answering all those questions on the support forums and for making this post. Let me know if you'd be interested in getting involved in Mozilla Support semi-officially in a more specialized role. I'm looking for volunteers to take on special projects like bug tracking on the support forums, tracking security updates for plugins, and other things that I think you might find engaging. Contact details are in my profile.

    Cheers!

    Mark Schmidt Firefox Support Lead @ Mozilla

    Thanks for helping out, Dingeroo! And double thanks for going above and beyond and reporting a trend in support cases. We really appreciate it when contributors take a critical eye and show interest in improving things beyond just answering questions. The short answer is, yes Firefox has a handful of well known architectural problems which contribute to crashes, slowness, and high memory usage. In keeping with Mozilla's dedicatation to openness and transparency, they make no attempt to hide these problems. Now for a longer answer: When it was first created Firefox was a best-in-class piece of software, with a unique multi-platform architecture which blew away all other browsers at the time. Even 5 years is a long time for one piece of software to be incrementally updated. But Firefox is 12+ years old right now. Architectural choices made 10 or 12 years ago may have been great for the time, but now prove difficult to change. However, the problems are well known and we are working on fixing many of them by seriously and fundamentally changing certain parts of the architecture of Firefox. Here's an example of three of these problems which are just about to get better: 1. The next version of Firefox, Firefox 40, will also be the official first release of a 64 bit version of Firefox for Windows. Not being a 64bit piece of software is a huge contributor to crashes, slowness, and memory miss-management on Windows (which the majority of FF users use). If you're on Windows, you can get the 64bit installer for Firefox Developer Edition here: https://www.mozilla.org/en-US/firefox/developer/all/ 2. Unlike newer browsers, Firefox runs as a single monolithic process. This makes memory management almost impossibly difficult, and means a crash in one component can easily crash the whole browser. A project to re-architect Firefox to use a different process for each tab, called Electrolysis (or e10s for short), has been underway for some time. E10s is almost ready for prime-time, and is now turned-on by default in Firefox Developer Edition (formerly Alpha/Aurora) and Nightly. I actually run Developer Edition and the normal current Firefox at the same time every day. I use Developer Edition for work, and the current/normal version of Firefox for my personal things. After having "normal" Firefox open for 3 hours, with 5 tabs, and having hardly used it, it is using almost 1738mb of memory. But Firefox Developer Edition, which has been open for 7 hours, with 30 tabs, and I've been using constantly, is currently using just 342mb of memory. If I were using the 64bit version of Dev Edition it would probably be doing even better. More about that here: https://wiki.mozilla.org/E10s 3. Another big problem is malicious third-party add-ons. Firefox 40 also starts the first part of a 3 release (40,41,42) roll-out plan for add-on quality improvement which begins with warning users when add-ons have not been approved and properly "signed." More about that here: https://wiki.mozilla.org/Addons/Extension_Signing There are probably another 10 or so big contributors to the most common memory and crash related issues on the forums. Rest assured that we are well aware and working towards fixes. Again, thanks a ton for answering all those questions on the support forums and for making this post. Let me know if you'd be interested in getting involved in Mozilla Support semi-officially in a more specialized role. I'm looking for volunteers to take on special projects like bug tracking on the support forums, tracking security updates for plugins, and other things that I think you might find engaging. Contact details are in my profile. Cheers! Mark Schmidt Firefox Support Lead @ Mozilla

    Modified by Mark Schmidt on

  3. Mark Schmidt said

    If you're on 64-bit Windows, you can get the 64bit installer for Firefox Developer Edition here: https://www.mozilla.org/en-US/firefox/developer/all/

    Would be better to use Win64 Beta builds since that will be closest to Release. There is usually between six to twelve beta builds for a version. https://www.mozilla.org/firefox/beta/all/

    Aurora (aka developer Edition) and Nightly gets checkins almost everyday so it gets a update each of those days and sometimes can be unstable as a result.

    ''Mark Schmidt [[#post-66103|said]]'' <blockquote> If you're on 64-bit Windows, you can get the 64bit installer for Firefox Developer Edition here: https://www.mozilla.org/en-US/firefox/developer/all/ </blockquote> Would be better to use Win64 Beta builds since that will be closest to Release. There is usually between six to twelve beta builds for a version. https://www.mozilla.org/firefox/beta/all/ Aurora (aka developer Edition) and Nightly gets checkins almost everyday so it gets a update each of those days and sometimes can be unstable as a result.

    Modified by James on

  4. Dingeroo, If you want to follow some Historic (last few years) but recent improvements in memory handling, reporting and tools take a look at

    Glancing over there now I note

    Everyone here will be aware of

    Dingeroo, If you want to follow some Historic (last few years) but recent improvements in memory handling, reporting and tools take a look at * https://wiki.mozilla.org/Performance/MemShrink <br /> The project has been wound down now. * I found the developer's blogs a good read <br /> https://blog.mozilla.org/nnethercote/category/memshrink/ but now there are only occasional updates. Glancing over there now I note * Firefox 41 will use less memory when running AdBlock Plus <br />https://blog.mozilla.org/nnethercote/2015/07/01/firefox-41-will-use-less-memory-when-running-adblock-plus/ <br />Largely apparently due to a 14 year old bug being fixed. * The ADB issue probably did cause unexplained memory problems I did post a thread <br />''Memory Problems with Adblock Plus ( 19 Million users )'' <br /> [/forums/contributors/710319] Everyone here will be aware of * [[Firefox uses too much memory (RAM) - How to fix]]
  5. If you want to talk about the architecture of the code, I think it's best to speak to developers.

    There are forums listed at https://www.mozilla.org/en-US/about/forums/ Something like would probably belong in mozilla.dev.platform

    If you want to talk about the architecture of the code, I think it's best to speak to developers. There are forums listed at https://www.mozilla.org/en-US/about/forums/ Something like would probably belong in mozilla.dev.platform
  6. Thank you for your encouraging and generous words, Mark. I have sent you a private message and hope there can be a meeting of minds.

    Didn't realise there was a Dev's Edition of Firefox, but of course there would have to be - impressed by the volume of translations. The memory draw stats are surprising and explain why my data usage has climbed lately!

    Thanks for the "tour" through the links. Am concerned you may have the impression that I have some technical know how. I have almost zero, which is sure to disappoint. I guess I am what you might call merely a tad "street wise" around computing.

    A bit overwhelmed by all of the different "arms" of the Mozilla operation, from Sumo, to Wiki, to Support, to The Hub, to Contributor Forums and a swag of Mozilla this and thats. I guess I shall get my head around them all soon enough. Some reading waders should see me through.

    Let's hope the three phase Add-on /extension development moves along well. It may perhaps usher in a welcome reduction in site crashes and frozen screens.

    Thank you for your encouraging and generous words, Mark. I have sent you a private message and hope there can be a meeting of minds. Didn't realise there was a Dev's Edition of Firefox, but of course there would have to be - impressed by the volume of translations. The memory draw stats are surprising and explain why my data usage has climbed lately! Thanks for the "tour" through the links. Am concerned you may have the impression that I have some technical know how. I have almost zero, which is sure to disappoint. I guess I am what you might call merely a tad "street wise" around computing. A bit overwhelmed by all of the different "arms" of the Mozilla operation, from Sumo, to Wiki, to Support, to The Hub, to Contributor Forums and a swag of Mozilla this and thats. I guess I shall get my head around them all soon enough. Some reading waders should see me through. Let's hope the three phase Add-on /extension development moves along well. It may perhaps usher in a welcome reduction in site crashes and frozen screens.
  7. Thank you John and Chris also for your "link" tours. Interesting reading some of it, but I don't think I shall be rising to anything like "developer" status, thank you all the same Chris. I have no technical training, just street wise experience and that is not all that wonderful, either some times.

    Thank you John and Chris also for your "link" tours. Interesting reading some of it, but I don't think I shall be rising to anything like "developer" status, thank you all the same Chris. I have no technical training, just street wise experience and that is not all that wonderful, either some times.
  8. Those blogs I listed presume you know something about Firefox and its use, but try to explain a lot of the changes and improvements using minimal technical detail.

    There is lots to do whatever your previous experience. You've already demonstrated that you think about problems and don't just blindly repeat answers. ( N.B. Just posting canned responses & KB article links, helps people and may be an easy way of learning - about Firefox - and about our KB resources I would not wish to put off any new contributor choosing to to go down that route )

    Just needing to learn more about what we try to do ? Keep the waders on ! Look at and drill down through

    No need to be a developer to use Developer edition. The Developer Edition, was a re-branding of Aurora and now installs by default separately and with its own profile. That's handy as we had to do that manually before. I've had computer trouble one way or another for a while, (Spend too much time using them, not enough time fixing them.) but usually I use Aurora|Developer as the day to day browser with the Release only in reserve.

    You can expect to have compatibility issues with addons on Aurora, and need to recognise it is not fully stable but problems tend to be minor issues rather than it becoming unusable or crashing. You see new features about 12 weeks earlier than the Release. (Unless Mozilla is going to make a habit of surprising us with throwing things in direct to Release & Point Release). That gives us a chance to learn about test and document changes.


    Nightly is a different kettle of fish Great to see the latest changes. Useful to test bug fixes, but no pretence it is stable.

    Of course you may install as many versions of Firefox as you wish and with most modern computers easily run multiple versions and instances; simultaneously if desired !

    64 bit Firefox has been around for years on the Release if you are not a Windows user. (Or have another OS e.g. multibooting )

    Those blogs I listed presume you know something about Firefox and its use, but try to explain a lot of the changes and improvements using minimal technical detail. There is lots to do whatever your previous experience. You've already demonstrated that you think about problems and don't just blindly repeat answers. ( N.B. <sub> Just posting canned responses & KB article links, helps people and may be an easy way of learning - about Firefox - and about our KB resources I would not wish to put off any new contributor choosing to to go down that route</sub> ) <u>'''Just needing to learn more about what we try to do ? '''</u> Keep the waders on ! Look at and drill down through * [[Contributor News & Resources]] <u>'''No need to be a developer to use Developer edition.''' </u> The [https://www.mozilla.org/firefox/channel/#developer Developer Edition], was a re-branding of Aurora and now installs by default separately and with its own profile. That's handy as we had to do that manually before. I've had computer trouble one way or another for a while, (Spend too much time using them, not enough time fixing them.) but usually I use Aurora|Developer as the day to day browser with the Release only in reserve. You can expect to have compatibility issues with addons on Aurora, and need to recognise it is not fully stable but problems tend to be minor issues rather than it becoming unusable or crashing. You see new features about 12 weeks earlier than the Release. (Unless Mozilla is going to make a habit of surprising us with throwing things in direct to Release & Point Release). That gives us a chance to learn about test and document changes. ------- '''[https://www.mozilla.org/firefox/channel/#developer Nightly]''' is a different kettle of fish Great to see the latest changes. Useful to test bug fixes, but no pretence it is stable. Of course you may install as many versions of Firefox as you wish and with most modern computers easily run multiple versions and instances; simultaneously if desired ! '''64 bit Firefox''' has been around for years on the Release if you are not a Windows user. (Or have another OS e.g. multibooting )
  9. Dingeroo, don't worry about having technical training, people here come from many very different backgrounds and we learn by doing (you would actually be surprised how diverse our community is and how many members of the community have not had any formal computer training whatsoever)

    I know that it's a lot to take in with all the wikis and links and forums (we call it the Mozilla firehose) but everybody is here to help so take your time and feel free to ask around when you get stuck. And, most importantly, don't forget to have fun :)

    And thanks for starting this conversation, this is very interesting!

    Dingeroo, don't worry about having technical training, people here come from many very different backgrounds and we learn by doing (you would actually be surprised how diverse our community is and how many members of the community have not had any formal computer training whatsoever) I know that it's a lot to take in with all the wikis and links and forums (we call it the Mozilla firehose) but everybody is here to help so take your time and feel free to ask around when you get stuck. And, most importantly, don't forget to have fun :) And thanks for starting this conversation, this is very interesting!
  10. more options

    The words "Fundamental", "architectural" and "Structural" are very loaded and have very different meanings than I think Dingeroo is noticing in the forums.

    Firefox as a product by itself is very stable and mature. It runs very well on the majority of user's machines, and does so without major issues. Firefox is a product that is used by hundreds of millions of people around the world, and on a variety of systems with a variety of configurations for a variety of uses, so naturally, Firefox will fail on some of those use cases.

    Unlike Dingeroo's conclusion that these are memory issues however, the fact is that memory usage probably causes a very small percentage of the issues users report on SUMO. It is however, incorrectly called out by users as the cause because it's a very simple concept that has real world correlations. Plus, RAM is a number that is heavily called out in PC marketing and is large numbers equal faster, so if Firefox is using bigger numbers, that must mean something. Besides a subset of crashes (which will largely be fixed with e10s and x64 support), memory doesn't cause many common Firefox problems.

    Rather, issues tend to fall into the categories Mark called out above:

    • Crashes. Tend to be tied to graphics, malware, JS, memory, etc.
    • Video. This is largely caused by Flash being so horrible.
    • Malware. Self-Explanatory
    • Performance. This is caused by a large variety of issues, and isn't as simple as memory usage in most cases.

    The answer is no, there is not a fundamental issue in Firefox, obviously since the vast majority of Firefox users are VERY happy. There are however several general categories of issues that users tend to run into frequently, but many of them all have varying causes.

    BTW, a quick run down of the issues in the OP above, almost none of them are caused by memory issues.

    The words "Fundamental", "architectural" and "Structural" are very loaded and have very different meanings than I think Dingeroo is noticing in the forums. Firefox as a product by itself is very stable and mature. It runs very well on the majority of user's machines, and does so without major issues. Firefox is a product that is used by hundreds of millions of people around the world, and on a variety of systems with a variety of configurations for a variety of uses, so naturally, Firefox will fail on some of those use cases. Unlike Dingeroo's conclusion that these are memory issues however, the fact is that memory usage probably causes a very small percentage of the issues users report on SUMO. It is however, incorrectly called out by users as the cause because it's a very simple concept that has real world correlations. Plus, RAM is a number that is heavily called out in PC marketing and is large numbers equal faster, so if Firefox is using bigger numbers, that must mean something. Besides a subset of crashes (which will largely be fixed with e10s and x64 support), memory doesn't cause many common Firefox problems. Rather, issues tend to fall into the categories Mark called out above: *Crashes. Tend to be tied to graphics, malware, JS, memory, etc. * Video. This is largely caused by Flash being so horrible. * Malware. Self-Explanatory * Performance. This is caused by a large variety of issues, and isn't as simple as memory usage in most cases. The answer is no, there is not a fundamental issue in Firefox, obviously since the vast majority of Firefox users are VERY happy. There are however several general categories of issues that users tend to run into frequently, but many of them all have varying causes. BTW, a quick run down of the issues in the OP above, almost none of them are caused by memory issues.
  11. Dingeroo, your welcome for mentioning the Beta's option instead of the developer edition.

    Dingeroo, your welcome for mentioning the Beta's option instead of the developer edition.
  12. Woooowha-a-a ! How to respond to all quickly, succinctly and neatly! Where do I begin ...

    Probably best with the briefest first.

    Hi James, nice to see you responding here, partic. after the public caning I gave you for which I have been duly caned in return !

    Yup, am with you and likewise that is my question, would I need the Developer's Edition, being the non-techie that I am? Wouldn't the Beta be more up my ally, but I am open to the wisdom of others on this. (Chris, probably assumed I had better credentials). I have Win x32/x64.

    Madasan aka Madalina thank you for the encouragement (and the private caning - LOL!) Your words are an encouragement because I do love web community projects and getting involved, but had presumed technical training would be a pre-requisite. I would be chuffed to perhaps somehow be a minute part of making Firefox again the firestorm among browsers - it already has so many worthy attributes.

    John - From what I've read so far, much of the material you have linked is relatively non-techie, readable and interesting. You remind me of a past varsity Master, (a compliment), who used to shell reading assignments out in throngs, but please don't be deterred. The MemShrink project looks really interesting - keen to read up properly on the past and present MemShrink bugs.

    I went to the MDN page on Performance. Haven't really read a lot there and suspect much will be over my head. Example: I ran about:memory and then chose the tab measure memory report - all gobbledygook to me, to be perfectly candid. Nicholas Nethercote is obviously a Firefox Developer on the MemShrink team? I skim read his article on Firefox 41 in conjunction with your posts about the ADP add-on, its redoubtable popularity and memory hogging issues. I have to say the technical passages go zoom! straight over my head, but nevertheless I can grab the gist of progress and where things are headed - all worthwhile reading, thank you, John. That it took 14 years to land a bug fix, was truly humbling for me, recognising the obvious level of difficulty.

    Tyler - I bow to your greater knowledge regarding my use of adjectives. I admit, I was concerned that Firefox possibly had serious stability issues on account of memory leaks and you have put it in perspective by enumerating the global community of users, compared with reports in a mere Support Forum. Even if we must acknowledge there may be thousands who do not bother coming into the Support Forum and who switch back to I.E. (only because it is already installed on PC's), they are still a minutiae number, when compared with overall Firefox user numbers.

    The goal, of course, is to stop that leakage of custom, however small and to also steal market share from Google Chrome and I.E. The interesting thing is that in using Google Chrome, which - as I understand it and correct me if I am wrong - is the greatest memory hogger of all the browsers, I never experienced any browser crashing. Conversely, in the short time I have been using Firefox it has already crashed a couple of times. I find I cannot have too many tabs open, not as many as I could in Chrome, which is a real paradox, given that Chrome supposedly, according to tests I have read, hogs more memory.

    I am not questioning your knowledge, nor your definition of the issues at all, Tyler and think your response brings great perspective and accept your explanations in good faith. I do like to ensure that all viable and helpful information is aired. For example, I have never through "normal" use of my PC exceeded my monthly data limit. Since using Firefox I have had to "top up" on two occasions. This is important information.

    The move to x64 will certainly up the ante and Mark's comparisons of memory use are revealing, to say the least - i.e. after 3 hours as much as 1738mb of memory on the standard Firefox and after 7 hours as little as 342mb of memory on the Developer Edition. Great things to look forward to and exciting progress.

    I did wonder about any server issue and whether only high-end servers are commissioned, which I cannot imagine they would not be.

    Woooowha-a-a ! How to respond to all quickly, succinctly and neatly! Where do I begin ... Probably best with the briefest first. Hi '''James''', nice to see you responding here, partic. after the public caning I gave you for which I have been duly caned in return ! Yup, am with you and likewise that is my question, would I need the Developer's Edition, being the non-techie that I am? Wouldn't the Beta be more up my ally, but I am open to the wisdom of others on this. '' (Chris, probably assumed I had better credentials)''. I have Win x32/x64. '''Madasan''' aka Madalina thank you for the encouragement ''(and the private caning - LOL!)'' Your words '''''are''''' an encouragement because I do love web community projects and getting involved, but had presumed technical training would be a pre-requisite. I would be chuffed to perhaps somehow be a minute part of making Firefox again the firestorm among browsers - it already has so many worthy attributes. '''John''' - From what I've read so far, much of the material you have linked is relatively non-techie, readable and interesting. You remind me of a past varsity Master, ''(a compliment''), who used to shell reading assignments out in throngs, but please don't be deterred. The MemShrink project looks really interesting - keen to read up properly on the past and present MemShrink bugs. I went to the MDN page on '''Performance'''. Haven't really read a lot there and suspect much will be over my head. ''Example:'' I ran '''about:memory''' and then chose the tab '''measure memory report''' - all gobbledygook to me, to be perfectly candid. Nicholas Nethercote is obviously a Firefox Developer on the MemShrink team? I skim read his article on Firefox 41 in conjunction with your posts about the ADP add-on, its redoubtable popularity and memory hogging issues. I have to say the technical passages go ''zoom!'' straight over my head, but nevertheless I can grab the gist of progress and where things are headed - all worthwhile reading, thank you, John. That it took 14 years to land a bug fix, was truly humbling for me, recognising the obvious level of difficulty. '''Tyler''' - I bow to your greater knowledge regarding my use of adjectives. I admit, I was concerned that Firefox possibly had serious stability issues on account of memory leaks and you have put it in perspective by enumerating the global community of users, compared with reports in a mere Support Forum. Even if we must acknowledge there may be thousands who do not bother coming into the Support Forum and who switch back to I.E. ''(only because it is already installed on PC's)'', they are still a minutiae number, when compared with overall Firefox user numbers. The goal, of course, is to stop that leakage of custom, however small and to also steal market share from Google Chrome and I.E. The interesting thing is that in using Google Chrome, which - as I understand it and correct me if I am wrong - is the greatest memory hogger of all the browsers, I never experienced any browser crashing. Conversely, in the short time I have been using Firefox it has already crashed a couple of times. I find I cannot have too many tabs open, not as many as I could in Chrome, which is a real paradox, given that Chrome supposedly, according to tests I have read, hogs more memory. I am not questioning your knowledge, nor your definition of the issues at all, Tyler and think your response brings great perspective and accept your explanations in good faith. I do like to ensure that all viable and helpful information is aired. For example, I have never through "normal" use of my PC exceeded my monthly data limit. Since using Firefox I have had to "top up" on two occasions. This is important information. The move to x64 will certainly up the ante and Mark's comparisons of memory use are revealing, to say the least - i.e. after 3 hours as much as 1738mb of memory on the standard Firefox and after 7 hours as little as 342mb of memory on the Developer Edition. Great things to look forward to and exciting progress. I did wonder about any server issue and whether only high-end servers are commissioned, which I cannot imagine they would not be.
  13. about:memory - Handy if bug report needs filing A lot of it is gobledygook to me too. But if you have need to look at it again

    • Items have further info available on mouse-over as a tooltip
    • Did you notice the tutorial ?
    • It could be handy if a bug needs filing
    • If anyone claims Firefox is leaking memory rapidly it shows exactly what is using that memory. Not only will it produce a report it will produce another one showing the difference between two reports.

    So if someone says Firefox memory use goes up by 1GB in 30 minutes do a report and 30 minutes later do another and diff them. Just be aware they may contain privacy sensitive information. (May need to consider emailing link or file to named staff)

    Someone can of course look at their own reports and may find it is good clue as to what is eating up memory.


    Your Own Crashes

    ... in the short time I have been using Firefox it has already crashed a couple of times. I find I cannot have too many tabs open ...

    File a question in the support forum. Include the full troubleshooting formation as prompted. Include the recent submitted Crash IDs Post back here to let us know you have done that. Be prepared for suggestions that you test in Firefox's safemode with plugins disabled. And be aware that if the crashes are related to FlashPlayer use you may not get a satisfactory solution.

    What do you mean by a lot of tabs ? More than 10, 50, 100, 200, 1000 It may also depend on the site type: video streams or large interactive databases, will need more resources than a short page of text.

    Self Selection Bias & need for confirmation I can often be guilty of not remembering to confirm information. I know you try to empathise with users on the forum which is good. However if Firefox works for 98 out 100 or 999 out of a 1000 it is only those 1 in 50 or 1 in 1000 that look at the forum so if a common solution fails to help we need to consider less common causes, including users misunderstanding the situation or the steps required.

    '''about:memory - Handy if bug report needs filing''' A lot of it is gobledygook to me too. But if you have need to look at it again * Items have further info available on mouse-over as a tooltip *Did you notice the [https://developer.mozilla.org/en-US/docs/Mozilla/Performance/about%3Amemory tutorial] ? * It could be handy if a bug needs filing *If anyone claims Firefox is leaking memory rapidly it shows exactly what is using that memory. Not only will it produce a report it will produce another one showing the difference between two reports. So if someone says Firefox memory use goes up by 1GB in 30 minutes do a report and 30 minutes later do another and diff them. Just be aware they may contain privacy sensitive information. (May need to consider emailing link or file to named staff) Someone can of course look at their own reports and may find it is good clue as to what is eating up memory. ---------- '''Your Own Crashes''' <blockquote> ... in the short time I have been using Firefox it has already crashed a couple of times. I find I cannot have too many tabs open ... </blockquote> File a question in the support forum. Include the full troubleshooting formation as prompted. Include the recent submitted Crash IDs Post back here to let us know you have done that. Be prepared for suggestions that you test in Firefox's safemode with plugins disabled. And be aware that if the crashes are related to FlashPlayer use you may not get a satisfactory solution. What do you mean by a lot of tabs ? More than 10, 50, 100, 200, 1000 It may also depend on the site type: video streams or large interactive databases, will need more resources than a short page of text. '''Self Selection Bias & need for confirmation''' I can often be guilty of not remembering to confirm information. I know you try to empathise with users on the forum which is good. However if Firefox works for 98 out 100 or 999 out of a 1000 it is only those 1 in 50 or 1 in 1000 that look at the forum so if a common solution fails to help we need to consider less common causes, including users misunderstanding the situation or the steps required.
  14. more options

    Please note, your monthly data cap is in no way connected to Firefox's memory usage on your computer. Your data cap is used up by the websites you visit, the videos you watch, etc. The memory on your computer is used by programs on your computer to run. It has absolutely zero impact on your data cap. The fact that you've hit your data cap may be because of a change in browsing habits with Firefox, but that's about it.

    Please note, your monthly data cap is in no way connected to Firefox's memory usage on your computer. Your data cap is used up by the websites you visit, the videos you watch, etc. The memory on your computer is used by programs on your computer to run. It has absolutely zero impact on your data cap. The fact that you've hit your data cap may be because of a change in browsing habits with Firefox, but that's about it.
  15. LOL  ! ... well, there you go ... just illustrates how un-techie I am. I did question that after writing it, but was still not sure of the answer, so thanks for the heads up.

    I don't actually watch movies online - only a bit of YouTube and download music - has to be the culprit, where they are video clips of songs. I have a program/media player which copies anything on the screen or going through the sound card - legally and legitimately - from the web, its called Audials One. Is the only thing I do which is data heavy. Obviously have not kept track of that recently. I owe Firefox an apology - was blaming it !

    LOL ! ... well, there you go ... just illustrates how un-techie I am. I did question that after writing it, but was still not sure of the answer, so thanks for the heads up. I don't actually watch movies online - only a bit of YouTube and download music - has to be the culprit, where they are video clips of songs. I have a program/media player which copies anything on the screen or going through the sound card - legally and legitimately - from the web, its called Audials One. Is the only thing I do which is data heavy. Obviously have not kept track of that recently. I owe Firefox an apology - was blaming it !
  16. John, you have prompted me to go back to the memory reports and look at them properly, even if just the parent nodes to begin with - can appreciate the usefulness of differences in memory measurements over a time scale would be useful. Training I would need, if only to decipher what it is that many of the links are actually describing. Just identifying the add-ons looks a bit of a nightmare, without referring back to the tutorial.

    I would not wish to be the customer on the end of my analysis, but thank you for your confidence.

    The crashes in Firefox were at the outset of my experience with it and I did not associate them with Firefox - thought my machine needed a good clean out and took it to my PC mechanic and asked him to wipe as much as possible - with the exception of specific files to be copied to another external drive and imported after the clean.

    Only when I experienced the black/dark grey screens was I prompted to look at Firefox Support and saw that others were experiencing the same thing - months after the crashes. Since joining the community I have done an about:crash search, hoping to do exactly as you suggest, but all record has been wiped and I have had no crashes since the wipe.

    One could argue it was my PC's condition which caused the crashes. I tend to think it was how I was using the Firefox browser, with 30 or 40 tabs open at a time. I now keep that down to a max of 15. If playing a video, I close nearly everything to avoid a crash.

    John, you have prompted me to go back to the memory reports and look at them properly, even if just the parent nodes to begin with - can appreciate the usefulness of differences in memory measurements over a time scale would be useful. Training I would need, if only to decipher what it is that many of the links are actually describing. Just identifying the add-ons looks a bit of a nightmare, without referring back to the tutorial. I would not wish to be the customer on the end of my analysis, but thank you for your confidence. The crashes in Firefox were at the outset of my experience with it and I did not associate them with Firefox - thought my machine needed a good clean out and took it to my PC mechanic and asked him to wipe as much as possible - with the exception of specific files to be copied to another external drive and imported after the clean. Only when I experienced the black/dark grey screens was I prompted to look at Firefox Support and saw that others were experiencing the same thing - months after the crashes. Since joining the community I have done an about:crash search, hoping to do exactly as you suggest, but all record has been wiped and I have had no crashes since the wipe. One could argue it was my PC's condition which caused the crashes. I tend to think it was how I was using the Firefox browser, with 30 or 40 tabs open at a time. I now keep that down to a max of 15. If playing a video, I close nearly everything to avoid a crash.
  17. To get back to the thrust of my original post- in the wash-up my post basically concerns service to the customer and the quality of the product.

    At the time none of those posts listed had been resolved, bar one and a couple had not had any activity for nearly a week. What is the Firefox/Mozilla procedure for dealing with questions unanswered and/or unresolved after one week or even a after a few days? Is there a chain of action?

    To get back to the thrust of my original post- in the wash-up my post basically concerns service to the customer and the quality of the product. At the time none of those posts listed had been resolved, bar one and a couple had not had any activity for nearly a week. What is the Firefox/Mozilla procedure for dealing with questions unanswered and/or unresolved after one week or even a after a few days? Is there a chain of action?
  18. more options

    Looking through those threads, most of them do not have a last reply from the original poster, meaning we just have to wait until we hear back from them. The vast majority of sumo threads get a reply, and the original poster never replies back, meaning the thread never goes anywhere further. This is unfortunate, but there is only so much we can do to encourage people to come back to reply to their post.

    Looking through those threads, most of them do not have a last reply from the original poster, meaning we just have to wait until we hear back from them. The vast majority of sumo threads get a reply, and the original poster never replies back, meaning the thread never goes anywhere further. This is unfortunate, but there is only so much we can do to encourage people to come back to reply to their post.
  19. Yes, for those threads, I agree the ball is left in the original questioner's court. I understand that, but what is the Mozilla policy for preventing questions from slipping through the cracks, be it either no response from us or no appropriate closure to advise that their query is being referred to the technical division, (as opposed to being left un-pursued)? ... speaking only of issues which have been established and which the developers need to know about.

    Yes, for those threads, I agree the ball is left in the original questioner's court. I understand that, but what is the Mozilla policy for preventing questions from slipping through the cracks, be it either no response from us or no appropriate closure to advise that their query is being referred to the technical division, ''(as opposed to being left un-pursued)''? ... speaking only of issues which have been established and which the developers need to know about.
  20. more options

    We have an internal help-desk team which will follow up on questions that don't get answered after 3 days. As for bring issues to developers, my team (user advocacy) pro-actively looks out for issues that are commonly appearing among our users and brings them to engineering to be prioritized and resolved in future versions.

    For example, the upcoming add-on signing is being implemented partly because the UA team saw there was a significant number of users who are being infected by malicious add-ons, and we worked with the amo team and engineering to come up with a satisfactory solution.

    We have an internal help-desk team which will follow up on questions that don't get answered after 3 days. As for bring issues to developers, my team (user advocacy) pro-actively looks out for issues that are commonly appearing among our users and brings them to engineering to be prioritized and resolved in future versions. For example, the upcoming add-on signing is being implemented partly because the UA team saw there was a significant number of users who are being infected by malicious add-ons, and we worked with the amo team and engineering to come up with a satisfactory solution.
  1. 1
  2. 2