|
Edit: taking it to D&D.
|
#
?
Apr 4, 2014 08:07
|
|
|
|
| # ? May 16, 2024 23:23 |
|
|
I have to use Windows at work. Why doesn't Firefox have a 64 bit version for Windows?
|
|
#
?
Apr 5, 2014 20:06
|
|
|
Boris Galerkin posted:I have to use Windows at work. Why doesn't Firefox have a 64 bit version for Windows? You can run 32-bit applications on 64-bit machines and the benefits to running a 64-bit copy of Firefox are marginal so the team halted development on it. If you really think you need a 64-bit build of Firefox there are several 3rd party ones, last I checked Waterfox and Pale Moon were the two big ones.
|
|
#
?
Apr 5, 2014 20:08
|
|
|
Nightly has a 64 bit build
|
|
#
?
Apr 5, 2014 20:22
|
|
|
64 bit builds of the same software generally gobble up substantially more RAM as well, so you should prefer 32 bit on any architecture. Unless there's a specific use case like "this database software won't support instances larger than 4GB unless you have a 64 bit build and I need databases bigger than 4GB" you shouldn't sweat it much.
|
|
#
?
Apr 5, 2014 20:24
|
|
|
Pryor on Fire posted:64 bit builds of the same software generally gobble up substantially more RAM as well, so you should prefer 32 bit on any architecture. Unless there's a specific use case like "this database software won't support instances larger than 4GB unless you have a 64 bit build and I need databases bigger than 4GB" you shouldn't sweat it much. No they don't. Also there's nothing wrong with caching stuff in RAM in the first place. You get no performance boost from leaving your RAM unused.
|
|
#
?
Apr 5, 2014 20:51
|
|
|
Pryor on Fire posted:64 bit builds of the same software generally gobble up substantially more RAM as well, so you should prefer 32 bit on any architecture. Unless there's a specific use case like "this database software won't support instances larger than 4GB unless you have a 64 bit build and I need databases bigger than 4GB" you shouldn't sweat it much. The only major thing that gets bigger in 64bit apps are pointers, and they don't usually make up a significant percentage of an application's memory. The text and images on a web page will take roughly the same using either architecture. It's not hard to hit the 32bit wall with a browser. Keeping multiple windows open each with multiple tabs, visiting sites that are image and content heavy can get you there. There's really not reason to still be running 32bit apps on a 64bit machine.
|
|
#
?
Apr 5, 2014 21:13
|
|
|
Ok? The two dozen packages I've had to roll out to VMs and had to look at have all used 10-40% more RAM when going from a 32 to 64 bit build, since we're paying for hundreds of VM instances we need to know exactly how little RAM we can get away with, and stick with 32 bit binaries as a result. Not sure what you mean by "caching in RAM" in this context, but yeah if you have ample unused memory it doesn't matter either way.
|
|
#
?
Apr 5, 2014 21:13
|
|
|
OhPlz posted:The only major thing that gets bigger in 64bit apps are pointers, and they don't usually make up a significant percentage of an application's memory. The text and images on a web page will take roughly the same using either architecture. Not at all, it depends on the compiler and data structures used in the application. It's not that simple.
|
|
#
?
Apr 5, 2014 21:15
|
|
|
Pryor on Fire posted:Ok? The two dozen packages I've had to roll out to VMs and had to look at have all used 10-40% more RAM when going from a 32 to 64 bit build, since we're paying for hundreds of VM instances we need to know exactly how little RAM we can get away with, and stick with 32 bit binaries as a result. Not sure what you mean by "caching in RAM" in this context, but yeah if you have ample unused memory it doesn't matter either way. Most 64 bit applications take advantage of the fact that they're no longer stuck in a memory window from 1985. This usually includes caching much more stuff in RAM then they would in 32 bit builds, because it's way harder to actually run out of RAM.
|
|
#
?
Apr 5, 2014 21:17
|
|
|
Pryor on Fire posted:Not at all, it depends on the compiler and data structures used in the application. It's not that simple. There's a lot involved, but it's a fair generalization for desktop apps. A web browser is going to be eating memory for pictures and other content. Other than managing the memory heap, there's not a lot going on there in terms of data structures. Lists of links, html nodes, css rules, etc will get into that, but would be drowned out by a handful of high res pictures, byte-for-byte. I've done 32bit to 64bit conversions at several companies. I can't think of any where there were significant differences in the memory footprint from one to the other. In the early days of Windows doing 64bit perhaps, but the compilers and various libraries have matured since then. Something like a database would be a different case. They're you're generally talking about massive amounts of small pieces of data. The "data structure" overhead there could be quite significant compared the actual data being managed.
|
|
#
?
Apr 5, 2014 21:34
|
|
|
Well in my experience 64 bit versions have always used like I said 10-40% more RAM. I was able to save my company what will work out to almost a million dollars a quarter by utilizing 32 bit binary versions of various unix packages/libraries instead of 64 bit, and downgrading the weight of the Amazon VM instances as a result. A quick glance at both the 64 bit and 32 bit versions of firefox with no plugins, just opening this single page in one tab reveals 159MB of RAM usage vs 118MB of RAM for the exact same version, which works out to a 35% increase in RAM usage. This is a bit higher than I expected really, wonder what the difference is with a ton of plugins and 20 tabs open and music playing.
|
|
#
?
Apr 5, 2014 22:33
|
|
|
And on my computer, opening Firefox and Waterfox to the same mixture of about 25 tabs with varying levels of content, and then leaving them to sit for 5 minutes, results in 995 MB of RAM used in Firefox and 999 MB of RAM used in Waterfox.
|
|
#
?
Apr 5, 2014 22:36
|
|
|
hifi posted:Nightly has a 64 bit build
|
|
#
?
Apr 5, 2014 22:37
|
|
|
The real reason Firefox isn't 64-bit is because they needed to make a new optimizing compiler for JavaScript that generated x64 code. The placeholder one they had was significantly slower than 32-bit. Firefox is pretty efficient with memory so there wasn't a lot of pressure to access more, and just jacking up caches isn't a sure win because memory management overhead can eat up more performance than you gain.
|
|
#
?
Apr 5, 2014 23:00
|
|
|
Pryor on Fire posted:A quick glance at both the 64 bit and 32 bit versions of firefox with no plugins, just opening this single page in one tab reveals 159MB of RAM usage vs 118MB of RAM for the exact same version, which works out to a 35% increase in RAM usage. IMO, that's too little memory usage for profiling. You want to ensure that the real data, in this case web content, is the majority of the memory utilization and not the executable itself and its dependencies.
|
|
#
?
Apr 5, 2014 23:01
|
|
|
If you really want 64-bit you can use Waterfox or Cyberfox. But Mozilla probably won't do it officially until Google does with Chrome.
|
|
#
?
Apr 6, 2014 02:16
|
|
|
Just use Linux Firefox!
|
|
#
?
Apr 6, 2014 17:45
|
|
|
Install Windows posted:Most 64 bit applications take advantage of the fact that they're no longer stuck in a memory window from 1985. This usually includes caching much more stuff in RAM then they would in 32 bit builds, because it's way harder to actually run out of RAM. It's true that an application can benefit from caching stuff, but it makes little sense in a web browser where the upcoming content is of infinite variance, unpredictable and the caching process itself would be intrusive (i.e. beginning to download and cache all of your bookmarks on startup or something). Caching like there's no tomorrow only really makes sense with applications where the data is mostly firm, its use predictable, and where there is a huge performance boost to be had by caching before opening the flood gates - the best example is a database.
|
|
#
?
Apr 7, 2014 00:47
|
|
|
Pilsner posted:It's true that an application can benefit from caching stuff, but it makes little sense in a web browser where the upcoming content is of infinite variance, unpredictable and the caching process itself would be intrusive (i.e. beginning to download and cache all of your bookmarks on startup or something). Caching like there's no tomorrow only really makes sense with applications where the data is mostly firm, its use predictable, and where there is a huge performance boost to be had by caching before opening the flood gates - the best example is a database. Web pages have plenty of static content. The same CSS, scripts, and images often appear many times in a Web site. Caching that stuff is going to reduce network usage and improve page load times.
|
|
#
?
Apr 7, 2014 03:07
|
|
|
dpbjinc posted:Web pages have plenty of static content. The same CSS, scripts, and images often appear many times in a Web site. Caching that stuff is going to reduce network usage and improve page load times. I'm not really informed enough to have an opinion on this, but are scripts / CSS large enough to really make a noticeable difference? Images I can see, but scripts and CSS are just a bit of text right?
|
|
#
?
Apr 7, 2014 06:40
|
|
|
Read posted:I'm not really informed enough to have an opinion on this, but are scripts / CSS large enough to really make a noticeable difference? Images I can see, but scripts and CSS are just a bit of text right? The firefox dev tools has a nice chart for that under the network tab if you click the number of requests in the bottom right corner. It's about half a second on the front page of this thread for me.
|
|
#
?
Apr 7, 2014 07:14
|
|
|
dpbjinc posted:Web pages have plenty of static content. The same CSS, scripts, and images often appear many times in a Web site. Caching that stuff is going to reduce network usage and improve page load times. hifi posted:The firefox dev tools has a nice chart for that under the network tab if you click the number of requests in the bottom right corner. It's about half a second on the front page of this thread for me. Pilsner fucked around with this message at 13:49 on Apr 7, 2014 |
|
#
?
Apr 7, 2014 13:45
|
|
|
Pilsner posted:That's probably for processing the script. Downloading it will take almost no time. I'm afraid not. That's over half a second for a CSS file, mostly spent waiting for the server. No matter the size of the resource, every request introduces delays while waiting for the server to answer, and those delays add up pretty fast when every JS and CSS file is requesting more files. Pilsner posted:I don't think it makes sense to cache the CSS transformation... it would have to save a static HTML version of the page you're viewing and display that again later, which wouldn't be too smart for a forum at least. What are you on about? The CSS is static either way, it's not going to change every time you reload the page or load a new page on the same website. Unless you're talking about caching the resulting page once CSS and JS have been applied to it, which is something that some browsers do but only when using the back/forward buttons. Firefox has been doing it since 2006. Caching on the web isn't a new issue. Every browser has had cache both in memory and on disk since the netscape days. There are HTTP headers to specify how long a file should be cached, others to validate that a file in cache is still valid, lots of people have been thinking about it, designing cache management techniques and improving upon them for years. But I'm sure they'd appreciate your suggestion to cache “like the 20 latest visited websites, prioritizing the most visited or something.” If you really want no in-memory cache, you can go to about :config and disable browser.cache.memory.enable, or if you just want it to be smaller you can tune browser.cache.memory_limit (in kiB). The default value of 50MiB ought to be fine for most people though; if your instance of Firefox is eating up memory, it's certainly not because of its cache. If you're curious about memory usage, do check out about :memory, it's not always super accurate and it's not user-friendly but it still does an okay job of showing you where your memory is going.
|
|
#
?
Apr 8, 2014 15:10
|
|
|
So I have this problem with Firefox where Facebook will freeze and either crash to desktop or just force my entire laptop to restart. It only happens when that dropdown menu appears when you start to write a friend's name (or the name of a page). Its gotten to the point where I can't even use Facebook on my computer anymore. It doesn't happen on any other site and only happens with the dropdown menu. What could be causing the freezing?
|
|
#
?
Apr 11, 2014 01:10
|
|
|
JGBeagle posted:It doesn't happen on any other site and only happens with the dropdown menu. What could be causing the freezing?
|
|
#
?
Apr 11, 2014 01:16
|
|
|
Alereon posted:Odds are it's some add-on you have installed, disable anything Facebook related and try again. If that doesn't help, use the Reset Firefox feature mentioned in the OP, which makes a clean Firefox profile and reimports your data. I don't have any Facebook apps but I did update Java and Flash (why the gently caress do they not automatically update like everything else?  ) but I'm not really going to sit there on Facebook until it freezes again to see if that worked or not. Besides, I've followed what Firefox support has said for like the last ten versions that people have been complaining about. That and the inexcusable memory leak that still plagues me to this day.
|
|
#
?
Apr 11, 2014 02:05
|
|
|
JGBeagle posted:I don't have any Facebook apps but I did update Java and Flash (why the gently caress do they not automatically update like everything else?
|
|
#
?
Apr 11, 2014 03:00
|
|
|
I have a situation in which I frequently plug in and unplug audio devices (or conjure up some out of thin air with bluetooth) who in turn become the default audio device. I never had a problem with Firefox (and Flash, I guess) following along and immediately switching over to whatever output was preferred at each moment. I recently upgraded to 27 (from 23) and it seems to require a restart of the browser to detect that the list of audio devices has changed. Other programs still behave the way I'd expect. I guess what I'm asking is if someone with a usb audio interface or usb headset can confirm that this is now how Firefox 27 is expected to handle these things, or if it's some bizarre problem on my end.
|
|
#
?
Apr 12, 2014 12:38
|
|
|
Flipperwaldt posted:I guess what I'm asking is if someone with a usb audio interface or usb headset can confirm that this is now how Firefox 27 is expected to handle these things, or if it's some bizarre problem on my end. I have a similar issue with the 3.5mm front panel jack on my computer. File a bug?
|
|
#
?
Apr 12, 2014 18:19
|
|
|
Avenging Dentist posted:I have a similar issue with the 3.5mm front panel jack on my computer. File a bug? I dunno, I looked into the process of filing it as a new bug or adding some information to the one linked above and it's a little involved. I need an account for which I need to create a new email alias because they're going to make it public. That's already more of a headache than I need right now.
|
|
#
?
Apr 12, 2014 20:16
|
|
|
Alereon posted:It really sounds like you need to use Reset Firefox. The default browser doesn't really have noticeable memory leaks so if you're seeing that it's either an add-on you've installed or you have an issue with your profile, Reset Firefox will fix both. Check in Firefox > Add-ons to make sure you don't have any unexpected add-ons installed, such as something your antivirus program put there. It's also probably a good idea to run general malware scans. Tried it, still freezes. I should've mentioned that I had recently reinstalled Firefox in hopes of finally ridding myself of the memory leak and Facebook freezing and that didn't even work.
|
|
#
?
Apr 13, 2014 03:03
|
|
|
JGBeagle posted:Tried it, still freezes. I should've mentioned that I had recently reinstalled Firefox in hopes of finally ridding myself of the memory leak and Facebook freezing and that didn't even work.
|
|
#
?
Apr 13, 2014 03:44
|
|
|
Reinstalling Firefox is generally much less likely to fix a problem than resetting a profile. You can use the profile manger to create a new profile; if that profile works then you know the problem was somewhere in your profile, whether it's an add-on, plugin, or just general profile corruption.
|
|
#
?
Apr 13, 2014 03:52
|
|
|
I've only noticed that particular facebook crash bug since the site interface update, so it may just be a kink they need to work out.
|
|
#
?
Apr 13, 2014 22:07
|
|
|
Knowing if a clean profile helps would be valuable. I'll check FB-side reports too, see if there's a pattern.
|
|
#
?
Apr 14, 2014 20:21
|
|
|
What causes this poo poo?
|
|
#
?
Apr 15, 2014 09:48
|
|
|
Riso posted:What causes this poo poo? Generally a missing font, but since the glyph in question is in the Unicode private use area, the website should have loaded an appropriate font. Looking up the code it seems that old iOS/OSX versions used it to display a cherry blossom emoji, before using a proper codepoint: 🌸. (Also, if you use NoScript or similar, many website use third parties to provide custom webfonts so you should unblock them.) Grim Up North fucked around with this message at 10:31 on Apr 15, 2014 |
|
#
?
Apr 15, 2014 10:27
|
|
|
Third party fonts. Well that explains it, thanks.
|
|
#
?
Apr 15, 2014 10:53
|
|
|
|
| # ? May 16, 2024 23:23 |
|
|
Subjunctive posted:Knowing if a clean profile helps would be valuable. It isn't.
|
|
#
?
Apr 16, 2014 04:01
|
|