Google Chrome on Windows 'completely unusable', gripe users Google's Chrome browser promises "speed, simplicity and security", but for a significant number of Windows users you can strike "speed" from that list. A thread on Google’s Chrome forum titled: “Chrome has become completely unusable, in almost every imaginable way” has attracted over 375 posts since the problem was raised in …
MacBook pro - i7, OSX mavericks, shit-loads of RAM and Chrome is bloody awful recently
Memory leaks everywhere, page rendering bugs (have to resize the window occasionally to display a new page)
I wouldn't be the first time in comouting history that more resources have made a program run worse...
MacBook Pro - i5, OSX Yosemite, plenty of RAM.
I've taken to launching Chrome only rarely as needed. I find that Chrome makes the system less stable; for example, randomly, control panels that require password input would not unlock until I quit Chrome. Why does Chrome interfere with System Preferences? I don't know. It's frustrating. And playing video in Chrome runs both cores at 100%, draining the battery very quickly and making the UI very unresponsive.
On the other hand, right now, Chrome in fullscreen on a rotated screen on Mac does correct subpixel antialiasing. No other program does this. Chrome in desktop on the rotated screen also doesn't do correct subpixel antialiasing. WTH, Apple?
I've had bad experiences with Chrome on Windows, too. My favorite period was back in May, when Google Drive crashed Google Chrome. I had to use Chrome for Flash and Firefox for Google Drive. (I also no longer install Adobe's Flash Player separately. Flash sucks.)
Windows Server 2012 R2, Windows 7 64 bit, Windows 8.1, Fedora, Ubuntu - ALL crash since about version 35 and it is getting much worse. All my machines have at least 16 to 64 GB ram and NVIDIA 760 or higher cards.
It crashes when I try to grade on eCollege
It crashes on plenty of other websites
I suspect major coding changes under the hood are causing this. I am using the default 32 bit versions. I also see Chrome sucking up plenty of CPU and RAM as well. In the process of going back to Firefox. I do see them update all the time - now on version 39 and it takes forever to apply updates.
I can say with about 75% certainty it's a problem with chrome's flash player. Right click on the "new tab" button and select "task manager' in chrome. You'll see what tabs are eating the CPU. In every case where Chrome turned to crap for me, it's been tabs with flash. Even though I have flashblock on, the fact that there's a flash element somewhere in the source makes those tabs go nuts.
But it's not consistent. It can go days without a problem. Then one day, I take the notebook out of sleep mode, and *bam*, problem's there. Exit Chrome, restart, let all the tabs reload, and 50/50 the problem will recurr. Reboot the system, restart Chrome, problem's gone again for a few days.
But always, it is the flash tabs that start this chain.
It happens quite a lot to me (use it for javascript debugging and css work) and have learnt to live with it as the dev tools are good. Two things are really irritating though.
1. It appears to be impossible to clear the file cache - clearing the cache manually rarely works entirely and the options to stop caching while the dev tools are open is flaky too. Incognito also caches files. This is bloody annoying if you are debugging js includes and Chrome is keeping hold of stale files.
2. Whoever thought "Aw Snap!" was amusing needs a knee in the nads.
Whoever thought "Aw Snap!" was amusing needs a knee in the nads.
Just had that whilst submitting my VAT return. Not amused*.
Not even using Chrome - but Chromium on Linux. IMHO its just the bloat that all good software gets destroyed with. Chrome/Chromium had it made, swift and minimal display taken from what you actually want to see. Why can't they just keep it simple?
* Re-done successfully with Firefox.
<snip>
"1. It appears to be impossible to clear the file cache - clearing the cache manually rarely works entirely and the options to stop caching while the dev tools are open is flaky too. Incognito also caches files. This is bloody annoying if you are debugging js includes and Chrome is keeping hold of stale files."
<snip>
I have noticed that and it has lots of knock on effects, e.g. I can't logon to my Easyjet account butcause of a load of old outdated crap being stored. The dreaded Inertnet Exploder works fine as do the others but Chrome will not let go of old crap run when it last successfully logged in.
As a frequent traveler this sub optimal as one might say.
Ditto. Still nippy here. I can easily believe it's scraping at the graphics drivers tho. Little trouble on my recent Nvidia. More data needed.
I do find crashed ad-heavy pages after prolonged inactivity near recent heavy MMOFPS usage every now and then though. Used the internal bug reporter.
Wonder if there's something out there worming in.
My (very) old core i3 laptop (Win 7, 64 bit) - with a duff battery - lasts for nearly three hours using Firefox, and well under two using Chrome.
If I let my son stay up all evening watching Minecraft videos or playing Roblox (browser based game) on my i7 desktop PC (Win 8 Pro, 64 bit), the fans max out (and the bedroom turns into a sauna) using Chrome but remains relatively cool'n'quiet using Firefox.
Two different machines, two different configs, two different OS's two different browsers. The only component causing trouble is Chrome.
I've been using Chrome since it came out, piling more and more plugins on it - and never been anything less than faultless and butter-off-the-back-of-a-lubed-dolphin-smooth.
Not for one moment saying there can't be problems - but entirely disingenuous to state there's a major problem with Chrome.
I've asked for, voted for and even got features in the Chrome forums - I can only suggest if you think there's a problem you tell Google - and I've complete confidence they'll do their best to fix the issues.
I've been having a lot of extension crashes on Win7. But it's dev-m Chrome, so I just assumed it was the risk of running it.
These users could probably speed up the browser a lot by adding Adblock and Privacy Badger. As long as they don't crash every 5 minutes.
Edit: what happens is that an extension will start taking up 25% of the CPU (1 of 4 cores), and new tabs won't load. So I go to Chrome task manager, end the process, and click the box that immediately A pops up, allowing me to rerun the extension. Any unloaded content immediately starts working. Repeat frequently... the type of extension doesn't seem to matter. Adblock+, Privacy Badger, and RSS Live Links all do it.
'Maybe it only affects systems with really esoteric hardware or running unsupported versions of Windows like XP.'
Two desktop machines at work, one running XP and the other Win 7.
Over the past couple of weeks, Chrome on the XP box has been exhibiting all sorts of random weirdness, from massive memory leaks (with no tabs open) through to being selective which web sites it opens.
The Win 7 box is fine.
Interestingly, the version of Chrome on my Linux box@work has been acting up over the past couple of weeks (memory leaks etc, but I've not had the time to have a good look at it)..then again, there was the recent fubar they had with the 32bit linux version and Debian which took about a week to sort out..
Never really understood the necessity for a 64 bit browser - just how many tabs do you need open before memory becomes a limitation?
Anyway, all the browsers go through cycles of crappiness. Firefox had problems a couple of years ago. Still might happen if they go ahead poking an email client into the browser.
Latest builds of IE look pretty good, but not sure I'm allowed to say that on The Register.
There are other advantages to 64-bit but the key point here is that 64-bit Chrome is still Beta level software so far as I'm aware. I wonder how many of the complainers are the same as the one sampled in the article, i.e. they are just too dumb to realise that the beta 64-bit software is not going to be as reliable as the old 32-bit version they are used to.
IE itself is OK, but Bing Search returns poisoned domains at the top of the tree for some quite ordinary terms; also that silly little box in the top left is too small to be the main interaction with the browser. Microsoft need to fix their search returns and get rid of the truly awful, horrible MSN default page; commonest question I get asked about it is "how do I get rid of it?". Most of my users do not want to know about Kim Kardashian.
It's not just memory it hogs, it fairly hammers the CPU.
I first noticed on a laptop when the fan would kick in if I opened Chrome, pumping out heat across the desk. Close Chrome and it would immediately return to a quiet cool state. Investigations in the simple Task Manager showed Chrome using 30% just as the only app open (and on a single blank page).
Perhaps some software running in the background interfering with it..
Antivirus?
Trusteer Rapport? (I see this cause issues frequently)
Bad drivers?
The last major issue I had with Chrome was due to the error reporting module.. Would stop virtual machines from being able to allocate enough memory.
Both. Hideous performance on client browsers (Chrome and IE) often down to Norton / Avast toolbar and / or Trusteer Crapport. I always ditch them, then no worries. As regards ad-heavy sites, I often employ various popular HOSTS builds (to avoid malware rather than ads), but boy does it improve performance better than ad-blocking BHOs (apologies - kind of said something similar elsewhere today but it's relevant to the topic)
It also seems like every update makes the omnibar less and less useful. I still don't know how to remove unwanted entries from its suggestion list. Time was when typing 'th' would bring up 'theregister'. Now I have to type 'ther' before it offers that as the default. I've seen articles saying that some combination of shift keys with delete will remove items but it's never worked for me.
And yes, I've had multi-core CPU machines jam up while Chrome fannies around doing who knows what. It's still mostly okay but over the past year or two it does seem to have gone down hill.
It will depend on how much history you have that starts with 'th' already. I pops up for me when I just type 't'. To delete an entry use the keyboard arrow keys to scroll down to the entry you want and then press shift+delete, don't hover over it with the mouse.
To clear all omnibox data just press ctrl+shift+del while in Chrome and clear browsing history from the beginning of time.
but every morning when I come into work after resuming my laptop from standby, it's hung up, used up all my windows socket handles and I need to end all the Chrome processes and start again.
Something has definately gone badly wrong since about Chrome 36 on Windows.
With the addition of the 'background' programs now loaded by default the probability of system instability increases dramatically. I've noticed a degradation in performance even after blocking all the background nonsense I never wanted or needed. Comparing the code to something like Opera you can literally see the differences in architecture.
This post has been deleted by its author
I find it's Google Sync that's the problem. I go into my google settings on their web page and it says I'm syncing something ridiculous like 64,000 bookmarks. I clear it and it keeps gradually increasing even though I have added no new bookmarks.
Windows is nothing compared to Chrome on Android though, it'd freeze for 20 seconds every time I opened it. In the end I just disabled Chrome sync and that fixed it.
Only takes so long before X browser ( indeed any type of app ) has more and more bolt ons made and the original source tree is expanded and expanded with stuff is was never originally designed to have attached. The compiled software just gets slower and more bogged down. Witness Firefox, lightning fast and the best browser when it first arrived, these days it's tired and need to go on a serious diet.
I still prefer Chrome over the others, never had an serious problems and indeed found it to be to usually the best browser for getting nasty bloated web apps working properly, where IE or FF would collapse. I think Google need to stop tinkering and leave it alone.
"Only takes so long before X browser ( indeed any type of app ) has more and more bolt ons made and the original source tree is expanded and expanded with stuff is was never originally designed to have attached. The compiled software just gets slower and more bogged down. Witness Firefox, lightning fast and the best browser when it first arrived, these days it's tired and need to go on a serious diet.
I still prefer Chrome over the others, never had an serious problems and indeed found it to be to usually the best browser for getting nasty bloated web apps working properly, where IE or FF would collapse. I think Google need to stop tinkering and leave it alone."
I think you've hit several nails well and truly on the head there. The effort to turn a browser into an app execution environment is really screwing up the whole basic idea of a browser. At the same time the user experience of web apps is awful; they're the worst ever for stupid things (unmovable 'dialog' boxes, no such thing as two apps in view at the same time, etc), and they're slow and often buggy as hell.
Windows 8's Metro environment is arguably akin to a polished up web app look and feel (one app in view at any one time, big clunky unmoveable things, a restricted GUI widget set, etc). That's not done so great. What makes anyone think that a web apps are ever going to be as good as that, let alone better? One of the better web apps I've seen it the web view dished up by modern Exchange servers. It's OK at best, but it's certainly no where near as nice to use as Outlook.
Recent Android builds have been pretty bad too.
They've got nothing on the new version of the Android GMail app. The new Lolipop like interface is awful, its a memory hog and performance is so bad it hangs for several minutes trying to display a short email. It's just an email how effin hard can it be!
I'm not denying your experience but just to put another point of view. Since getting the Lollipop OTA for my Nexus 4 the whole interface is way quicker. The new Gmail loads a lot quick and e-mails appear almost instantly, that's with 5 accounts including, a Exchange, Hotmail and Gmail.
Memory usage is also hardly any. I'm currently using about 500MB out of 2GB with about 15 apps running.
Overall since Lollipop I think the whole interface is far better, much easier to use, looks better and runs a lot faster with less resources.
As I said, this is just my use case, but so far it has breathed new life into my N4.
As I wouldn't touch that Google spyware. Makes their WiFi Street view slurp look tame. They are NOT an altruistic Tech company, but a company making money from Adverts. They are a parasite. Look at their attitude to other people's media (Books, Music, Video and how they run YouTube).
Google's Web services used by many websites, Chrome and Android seriously makes the EU cookie concerns look daft.
about once ever 3-6 months, I have to purge Chromes cache, then its snappy again for the next few months. history -> clear browsing data.. -> uncheck passwords and form data if you want to keep them, its the browsing history and files you need to clear, and clear 'from the beginning of time'.
My wife has to use FaceBook for work and that gobbles RAM and CPU usage massively. I think it may be Flash rather than the browser itself. I had to open FB (In FireFox) at work today and it quickly gobbled up almost all my memory!
I've noticed Chrome seems to have a process entry for each open tab. 50MB each for simple pages!
"It is well known that Google makes little use of Windows internally. In 2010 it was reported that Google was forbidding internal use of Windows. “We’re not doing any more Windows. It is a security effort,” said an unnamed employee."
Yeah, because Android is well known to be the most secure OS. Ever.
Bollocks.
It's a slightly crazy notion. For better or worse a very large fraction of Google's customers use Windows one way or another. Google not using the Windows means that there's a good chance they'll screw up and not notice before they annoy a large number of customers. Their own staff are a valuable beta testing resource, yet they're not using them for this purpose on one of the largest platforms out there.
Oh, it seems that that's just happened.
Perhaps, but that real work doesn't seem to involve making Android work properly, or ensuring that one of the most important tools (Chrome) used to entice punters to their ad display stream (imaginatively called Google's "Services") works on one of the most common platforms out there.
Without those basics in place and working properly, what other real work is there in Google?!
It's almost as if they're becoming a bit Apple-esque; there's no point making things properly anymore, punters will just use it or buy it anyway. Why waste the money getting rid of the bugs that clearly aren't putting off the billions of people who give them money (even if only indirectly) every single day of the year?
Not just me then. I kind of thought it had something to do with HTML5 video and facebook or G+ notification checks. So background javascript then. Win7, 32 bit. I was getting the "Aw-snap" about once a day, and then more recently I'd walk away to make a coffee and come back to find Chrome had crashed completely.
This stuff comes and goes. Frequent hidden upgrades are both a blessing and a curse. You do get quick fixes, but you also get quick borkens. Methinks their regression testing could be a little more rigorous.
I've been hearing complaints all year from XP to Win 8 users and have experienced problems as well.
Some examples:
- does not work seamlessly with all Google service itself (Seriously, WTH?)
- page/tab crashes
- outright crashes (have they been taking lessons from FireFox?)
- erratic speed (I will give the benefit of the doubt here to erratic ISP connections)
- user preferences directory corrupts easily (this may be the culprit of everything else right here)
The user preference default directory seems to be at the heart of this issue. I've had to copy, move, delete old directory and then move the new copy back to fix problems several times. Each time, it worked perfectly for a few weeks.
What is happening that the top three browser's, well only 3 actually, quality control has gone to shit and they have become unstable on even old Windows? Windows itself? Again?
What version are we talking about, specifically? the article is pretty vague.
Chrome's memory footprint has grown sharply over the couple of last years, and especially memory leaks seem to be getting worse and worse. On my lowly 32-bit XP machine (yeah... shoot me) I used to be able to keep Chrome open with 4-5 windows, each with half a dozen tabs, for a week or so before my machine started paging/swapping like hell. Now I can hardly spend a full day without my laptop flatlining. Things have really started to get much worse around version 32 IMHO.
That's one of the reasons why I tend to stay far behind Chrome releases (yeah, shoot me again).
I don't use many extensions so not being able to access the store any more doesn't bother me.
Please don't reply "use decent hardware". Not the point here. Seems to me that 3 1/2 gigs of RAM should still be enough nowadays to have ~25 browser tabs open, but maybe I'm just too old school. When I was a young engineer 8 people could compile Cobol on ASCII terminals hooked to a 6809-based machine with 1 meg of RAM... (on UniFLEX if you want to know)
Have it deployed as the default (and only) web browser for an entire school - from 3 year olds to the admin staff via the teaching staff and gap-year / PGCE students.
We deploy on ordinary business hardware, nothing special, with 4Gb on Windows 8.
The only problems I have are:
Sometimes the javascript engine (I presume) keeps running instead of dying. You have to kill all Chrome processes and the "EventEd" thing in task manager before it will let you run Chrome as that user again. The kids reboot. The staff have permission to use task manager and know to kill it.
Weird error messages that are too easy to confuse between "The Internet is down" and "This website doesn't work". I get a lot of calls and have to keep saying "Does a Google search work? Yes? Then it's the website.".
I've disabled installation of extensions with Group Policy, so that's not an issue. All our banking and payroll that says it needs IE? Turns out it's lying (though I am monitoring the NPAPI situation carefully). Integrates with the staff Google accounts nicely. Loads quickly. Browses everything we want to browse, etc. Picks up auto-proxy settings first time (more than I can say for most programs, don't even get me started on iPads and Macs...).
So, sort out that EventID thing that stops me spawning new (visible) Chrome instances when there are old (invisible) ones, and smarten up the grey-screen error messages a bit and I'll have nothing to complain about.
'I get a lot of calls and have to keep saying "Does a Google search work? Yes? Then it's the website.".'
Advise them to try the following website. It might save you some calls:
Chrome's "Use hardware (GPU) acceleration when available" setting was causing problems on my ThinkPad. the GPU driver crashed occasionally, sometimes several times a day, particularly while using Chrome to watch videos on Youtube, Vimeo, etc. fortunately, the GPU driver would recover automatically after a few seconds, but the disruption was annoying to tsay the least. the crashing problems stopped after i disabled the GPU acceleration option. Chrome is also a greedy memory hog. i've seen some chrome.exe processes gobble up more than 2GB of memory, and that was just ONE TAB in the browser.
however, the latest 64bit build (version 39.0.2171.71 m) now seems stable on my machine. i've re-enabled GPU acceleration again, and no crashes so far.
I could have sworn that parts of the slowdown occurred because they globally disabled some accelerated math functions that in particular have huge impact on SSL/TLS optimisations, as well as that of the chrome pepper flash player plugin. I think that removing the compilation switch for SSE and possibly SSE2 was one of them (or at least that I've noticed on an older machine) so a whole bunch of relatively slow machines (because they don't have AVX/SSE3(x)/SSE4(x)) have become extremely slow. The only problem is that I've tried to find confirmation on this but have not tracked down through either release notes, or bug tracker, and I can't remember where I read this.
It was also roughly at the same time of heartbleed and some client side SSL/TLS safety checks were included.
In some cases where I've deliberately crashed chrome, I have gone from a usable 100 tabs across 10-15 windows under Chrome 30 and re-opened in 60 secs to a completely unusable mess that requires closing multiple windows to get down to a total of 15 tabs capable of running in less than 5 minutes.
People who have AES-NI or similar will have next to no problem (if there is any will be un-noticable because of the operation speed) as the slow down on non-AES-NI capable machines. This CPU capabilities flag as well as that of maths accelerations should be noticeable in the problem reports - that is if they are collecting it. This may explain why some recent intel chips fail, whilst older ones work as stated in the story..
It should be relatively simple to determine, go to portableapps, get all versions from 30 onwards, load up several tabs each of facebook, gmail, twitter, tumblr, pinterest, etc amongst others with mind numbing amounts of ads and flash and crash it (kill chrome.exe from task manager it should be the process without any flags/switches if you have command line column active.) Restart and open chrome task manager (Shift-Esc). If you are really committed some debug/profile software may help.
SSE3 is pretty old hat now, you'd have to have a PC about 10 years old to be lacking it. And even then a CPU core is still a pretty mighty computational beast. I shouldn't think that running SSL/TLS algorithms with only x86/x87 op codes isn't going to be as slow as Chrome appears to be for various users.
Installed it on my XP box (yeah, yeah) late last year, and it always took a while to launch and get up to speed, but once up and running it was generally OK (even on bloated pig websites like Fecebook). The project that spurred the Chrome install ended, and I stopped using it but recently I had to fire it up and it took over ten minutes (wall clock time) to get from launch to a useable state; two windows (one regular, one incognito) opened a half-dozen separate processes, and my machine was more or less locked up for the duration. Just to go to one website that refuses to work with Opera or Firefox. I had heard that once upon a time Chrome was a lean, mean, snappy browsing machine, but I am not seeing this at all. I always clear all browser history "since the beginning of time" on exit, only have two extensions (ghostery and a no-script thingy) installed, so I am a bit flummoxed. Forced ten-minute breaks should be welcome but are mostly inconvenient. Pah.
yeh, it's not as nippy as it used to be. Being at the bleeding edge of standards support comes at a price, I guess, For me, tt was pausing for 10 seconds whenever I opened a link in a new tab. Very annoying.
Since switching to Linux, however, I've noticed no problems at all.
For the most part I find chrome nice and snappy ... unless I leave a tab open on Facebook for too long (or any site with ties to it, which include far too many). Looking into it in task manager, one chrome process can be commonly found to consume 500M to over 1G of RAM for just one tab .... kill that process and invariably the FB tab changes to a crash report screen, all other tabs ok, unless they tie into FB somehow, then they crash too. Weird part is I have plenty of ram (16G), but when FB is taking up so much memory, the system as a whole seems far less snappy. How many of you people having problems, do as I often do, and have a facebook window backgrounded in a tab? Have you kept an eye on its memory usage crawling up over time? I've attributed this to shitty programming on Facebooks part (seems to be a nasty memory leak), and not a chrome issue, although I would love it if chrome allowed you to set a per tab memory limit to mitigate it. I havent really used other browsers in years, so no idea if it is just on chrome that FB has memory leak problems, but I'd suspect it affected other browsers too.
Runs very slowly on a HP Envy AllIn1 running 8.1 patched up to date.
Intel i5, 8GB RAM.
Google searches are the worst of all, I sit looking at a blank screen for many seconds.
Have just turned off hardware acceleration as suggested in an earlier post, will see what happens.
For my personal browsing I use firefox and love it. Because I've lately been doing web app development targeted at android devices (sigh), I pretty much have to use chrome for off-device development (using webstorm/ripple) to minimize weirdness once I move to the device for integration testing.
I can't stand the thing. It has a mind of its own for clearing cache (sometimes it just goes all Bartelby on you and "would rather not"), builds will randomly break plugins (and there's no equivalent of the extended stability firefox path), and it's just plain slow. Admittedly, that's emulating the device browser fairly well, so good job I guess?
There was a time when Chrome had the competition beat on performance and stability, but they've totally pissed that away.
In my own little unscientific test, I had 3 tabs open in Chrome and 3 tabs open in IE. Chrome had spawned 11 separate services, which consumed well over 800 MB memory (physical or virtual). IE on the other hand, only generated 3 threads and consumed just a fuzz over 275 MB of memory. Why? I could understand the differences if I were running different applets in each tab, but in order to be objective, I opened the same pages in both browsers. I'm not a programmer so I genuinely don't understand why the huge difference in resources consumed. What's more, I performed the exact same test on Firefox, it only had one service running, which consumed just a smidge over 250 MB of RAM.
The other thing I didn't understand is the cpu would bounce on any of the chrome occurrences, even when the application was moved to the background. It's quite obvious to me chrome has become ungodly slow compared to the other two.
I know a lot of exceptionally brilliant people frequent these boards, so could one of you explain these subtle nuances?
There is absolutely nothing wrong with CLANG and LLVM, in fact the opposite - it's great software. However, it's not "bug compatible" with gcc, so it's indeed possible that switching compilers has uncovered misbehavior that they could get away with in GCC.
It doesn't really matter what the problem is, they need to fix it pronto or they'll lose all their product users overnight - there is very little friction to move to another browser like firefox.
> There is absolutely nothing wrong with CLANG and LLVM, in fact the opposite - it's great software.
Here's a few tidbits of wisdom accumulated through many years of working on compilers:
1. There is no such thing as a bug-free compiler. All compiler have bugs.
2. A mature, responsible and experienced software engineering team knows this.
3. Clang/LLVM - great, GCC - buggy - please stop The Bullshit. This isn't a horse race or a beauty contest. Both compilers have bugs.
4. If and when an experienced and professional software engineering team switches compilers for their product, they don't just release said product - which used to be built with a different compiler - simply because "it builds" and "it seems to run OK". That's for amateurs. First you continue to provide builds of product with the "old" and known-to-work compiler. You then provide additional, experimental builds of the same exact product, on the same exact source code base, but built with the new compiler. You let your users know ahead of time that the "new" compiler builds are E-X-P-E-R-I-M-E-N-T-A-L, in big black bold letters.
That's what mature, experienced and knowledgeable software engineering teams do. Evidently, that's not everyone.
Having said all that - and probably having fallen on deaf ears: Clang/LLVM is known to have some serious bugs affecting the alignment of certain types of x86_64 instructions. That is a bug that GCC does not have. The Google Chrome team was aware of the existence of these bugs in x86_64 code emitted by Clang/LLVM. But, they chose to release their product anyway.
So. no. It's not a matter of "they got away with things in GCC that Clang didn't let them get away with". It's very likely a matter of the currently unresolved x86_64 instruction alignment bugs in Clang/LLVM that do not exist in GCC.
I'll take a wager on it.
I stopped users installing it some time back as I saw some strange system wide affects that spoke of "More than a browser".
I think it is either being clever with systems calls not releasing resources or just plain "getting involved" in processes it has not legitimate need to.
I'm not even sure Firefox has gone over to the Chrome prefered "Ignore local DNS" model too on the latest, seems less willing to look at local sites, grabbing google DNS results before local. I could be wrong, still prefer Firefox over Chrome right now and have for the last year or two.
I know.. and what is it with people's obsessions on defragging? Isn't the filesystem for Windows fixed yet?
It wasn't broken in that respect anyway. It just didn't choose to be as anal about fragmentation as, say, HFS or HFS+. Applications can reserve space ahead of time if they want contiguous room to grow but in most cases it isn't needed. HDD performance and capacity has helped make it less of an issue at the hardware level but most of all Windows itself has a dynamic cache that on most versions grows to occupy all available RAM. When you can have a 1GB or more of cache disk fragmentation isn't really a problem.
So the simple answer is: Defragging is mostly just a user amuser on a modern desktop. On a server it might be more important but I doubt it really makes much difference.
I've been a Google evangelist since I got my e-mail account in 1997. And I'm about to break up with her. There's not a single Google device, app or service that I use that doesn't have one, "Are you f'in kidding me?", primitive usability issue. Here's a brief list of the Google services I use...
> Chrome - Would up and down scroll arrows kill you? Why yes I did want to arbitrarily jump to another section of a very long web page for no good reason.
> Google drive + 1TB storage THAT I PAY FOR - Can't upload to it. NOT SYNC; just *(^%^ upload.
> GMail - The archive vs delete debacle, usability on a PC like getting to your contacts
> Google+ - Let me type this quick while it's up. Oh look there's all my friends from I'm high school... no that's that other site :-(
> Google music (try and get a simple text list of your music by artist)
> Support - The community HELL of https://support.google.com/ that is one atom better than a bulletin board from 1996
I've not forgotten to take my meds or anything. Maybe I'm just manstruating. But the Google storage experience from last night has just put me over the edge. I'd PAY GOOD money for their services if they just worked like you'd expect them to.
It's not me Google. It's you.
Always makes me smile when users throw tantrums about problems with free software.
Nobody seems to be taking a methodical approach to diagnosing this problem. Maybe if some of the people experiencing this issue actually tried to carry out some diagnostic work rather than simply complaining they'd get to the root of it.
I used Netscape until Mozilla then have used Internet Explorer ever since with few issues, and a lot of great browsing. On the iPad though I use Chrome. It's awful: slow buggy, crashes - worse than Windows 95 days. The only reason I use Chrome over Safari is because Safari is terrible.
I use chrome on my iPad. Been irritated with slow scrolling through pages particularly a problem for several weeks now that have been driving me to frustration at times. Read the article and was having terrible stutter when scrolling up and down the page so just tried viewing the same pages using Safari and it is many times better. Totally unscientific experiment of course but.....
This is the third time in a month where Chrome has just failed with "Aww Snap" and no matter what troubleshooting I do, I have to un-install it, reboot and re-install it before it plays nice.
This is Chrome 64 on a Core i7 2600K, 16GB's, 512GB SSD, 2TB storage HDD and Win7 x64 Enterprise so not a slow machine by any means.
I've found a few issues..
Firstly, blocking flash helps a lot.
However, my main issue is that the omnibox becomes unusable after a few weeks use.Deleting the contents of Chrome\User Data\Default\Cache regularly seems the only solution. I've seen lots of issues raised for this dating back many months yet the problem has still not been solved for me.
Two identical 16GB machines with Ubuntu 14.04.1
The one that was upgraded from 12.04 is fine (it's what I'm using right now). The new install crashed with the Aww snap nonsense. Turning off hardware acceleration reduced that (and fixed a Google Maps display issue) but flash still crashes regularly. No problem when it's just an ad but a pita with some other stuff.
Similar hardware with Win7Pro doesn't exhibit any problems.
But 8GB W7p Laptop Chrome becomes slow to unresponsive usually after just a few minutes use. Finally a 1GB netbook WinXP has never been able to run Chrome although my now defunct 4GB XP and 8GB Vista desktops did without apparent problem.
I started using Chrome because the "bloated" resource hogging of Firefox was getting to be an issue. But liking to use the same bowser on all regularly used machines Chrome has become untenable might mean it's back to Firefox.
Wah! who knew? I have been unimpressed with the childish "Snap", "Nope" etc in Chrome for a while and just last week it went completely potty on a win7 machine i5 + a stonkabyte of disk and ram. Even after a few reboots to tame the bleedin' thing... dumped it and went back to FF. That said I have it on a number of other machines and it is fine... bit bloaty (huge working sets) but meh. So very hit n miss
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2025
