“Linux: If a lack of third-party plug-in support (i.e. Flash) kept you from trying out Chrome on your Linux system, then avoid no longer. The ‘early developer version’ now supports many plug-ins, and they seem to work pretty well. You’ll need to add –enable-plugins to your Chrome shortcut’s command line operation to get the ‘buggy’ plug-in support, but it seems worth the hassle, as YouTube videos are playing relatively stable and smooth. Google’s updated their alpha-level Chrome builds to include the newest start page tweaks as well, and it’s generally a browser worth checking out, even if a few standard settings and convenience items are still missing.”
Note, it should be:
–enable-plugins
With two dashes in front. They were fused into a single long dash in the Lifehacker article (maybe it was drafted in Word), and ended up as a regular single dash in the summary.
google-chrome-unstable 64bit for debian does not load 64bit flash plugin since it was compiled with the wrong ELF
hope it gets fix soon
32 bit Google Chrome does load flash, and I don’t think it’s a huge loss for the 64 bit version that it currently doesn’t: enabling plug-ins makes Google Chrome for Linux very unstable.
In my case on Kubuntu 9.04 32-bit Chrome with plugins enabled works very stable. I am surfing and watching flash videos for 2 days now and I didn’t have any crashes. I am using it right now.
Pretty stable here on 32bit ArchLinux running Chromium 3.0.197.0~svn20090801r22243 and Flash 10.0.22.87.
Actually, retracing my activities I found that what crashed my browser wasn’t Flash, but viewing HTML5 video on Youtube. It happens with plugins disabled as well, of course. So just ignore my comment.
Flash does make the browser incredibly slow, though.
> Flash does make the browser incredibly slow, though.
No change there then.
The truth is that Chrome build for 64 bits it is really a 32 bit build packaged to use 32 bits shared libraries.
Probably you have 64 bits flash plugin and you need 32 bits flash for Chrome.
Yes it works fine for me on 32 bit Ubuntu Hardy LTS, Chromium 3.0.197.0 (0) and Flash 10,0,32,18. But now I need a flashblocker as well as an adblocker plugin. So I guess I will run it without plugins enabled.
You can have an ad-blocker. You can use the extension or script version of AdSweep: http://www.adsweep.org/
Any confirmation that this actually works on Chromium 3 under Linux?
I adapted the directions the best that I could, but could not get it to work. Tried it as a user script and an extension.
chrome://extensions tells me this when I tried to load it as an extension:
“Errors
Could not load extension from ‘~/.config/chromium/User Scripts/AdSweep.crx’. Manifest is missing or invalid.
Bad version number”
EDIT: I got it to work, I put the script version in “~/.config/chromium/Default/User Scripts/” and started it with –enable-user-scripts. I didn’t put it in Default the first few times I tried. It is pretty crappy though, you see the ads when they load, then it hides them. So you aren’t saving any bandwidth or rendering time, from what I can tell, it is considerably slower.
Edited 2009-08-02 06:24 UTC
Java doesn’t work though. It bravely tried drawing a blue rectangle where the applet should appear. But that was all it did.
Hi,
Does it have a RDP plug-in. I’m planning to connect it to Windows server such as ThinServer
http://www.aikotech.com/thinserver.htm
Why not just use an RDP client then?
While we’re sort of on the topic… anyone know of a Windows Terminal Services client for Linux (doesn’t have to be a browser plugin) which handles Windows 2008 Server’s proxy via http/https capability? rdesktop does not.
i feel there is too much hype around this project and the reason is only google.
OMG chrome has plugins support now
OMG chrome looks like a gtk+ app now
chrome was nothing if it wasnt google’s. i mean, its not chrome’s goodness, its just google’s hype.
Webkit is about the most exciting development going in the browser world, due to its quality, speed, and refreshingly clean code base.
Plus Chrome and IE are the only browsers which have broken out of the old and flawed “threads are cool” mentality, and are thus the only browsers that implement a sane process model for today’s browser usage patterns. (Now, if you want to talk about hype gone pathalogical, let’s talk about threads.)
Edited 2009-08-02 14:49 UTC
oh, wait, im not talking about webkit. webkit is cool and there is no doubt.
what chrome has to offer, is not really interesting. i mean it definitely was a good idea to make a browser with threads so if a tab crashes, others wont. but that that little idea/feature worths a whole new browser?
look at gmail. when started, it had many many things to offer. chrome? nothing compared to that.
also, do you really think users care about multi-thread stuff in chrome (if there was no hype around it?)
how many times you experienced a crash on firefox?
what im saying is that, if gmail was not google’s and belonged to a not really famous company, it was still going to be successful. but if chrome was another company’s project, it had no user at all.
and this multi-thread thingy is one of those OMG’s:
OMG if a tab crashes on chrome it wont cause crash in whole program.
how many times you experienced a crash on firefox?
Quite often, actually. I still use it, but I will be pretty glad when they get around to making it threaded in a way that only one tab crashes.
You probably already know this, but nspluginwrapper yields some advantages in this area even if you are running a 32 bit browser. Wrapping the plugin means it runs in its own process. When Flash crashes, the browser doesn’t. Unwrapped, the browser dies when Flash does. (Did I say Flash? I meant to say ‘the generic, anonymous plugin’.)
Edited 2009-08-02 17:20 UTC
And what other major browsers for Windows or Linux, based on Webkit, are as far along as Chrome/Chromium? None, so far as I know. MacOSX has Safari, of course.
Absolutely. If Firefox, for example, decides to follow that route, they’re going to end up with what is essentially a new browser. And yes, the feature is well worth it.
That’s ‘multiprocess’. (Actually hybrid multiprocess/multithread.) They care about stability, even if they don’t understand threads and processes.
Plenty.
But let me ask you this. Why are you taking the time and trouble to complain (in not a few words) about Chrome/Chromium getting a Page 2 story on OSAlert about a major new feature which represents a significant milestone for the open-source project? I sense an agenda of some sort.
Edited 2009-08-02 17:27 UTC
[quote]But let me ask you this. Why are you taking the time and trouble to complain (in not a few words) about Chrome/Chromium getting a Page 2 story on OSAlert about a major new feature which represents a significant milestone for the open-source project? I sense an agenda of some sort. [/quote]
i didnt meant to complain, just chatting with few people to see if they think the same and/or argue me.
about the above, well, firefox is stable to me and i personally dont think multi-thread worths that much.
firefox is stable to me and i personally dont think multi-thread worths that much.
Just try to think out of the box a little; some people have to use less stable plugins sometimes and for them any extra stability in the browser will be very welcome. As it might be very welcome for plugin developers as well. And it just so happens that FireFox ain’t all that stable for everyone. People who use their browser for research and have perhaps even dozens of tabs open will find it absolutely disastrous every time the browser dies, so to them too it’ll be god sent to be able to close only the crashed tab and not lose everything they had open.
The correct solution for buggy plugins isn’t to seperate the browser tabs into seperate processes (IMO), it’s to seperate each plugin into it’s own process so that it can’t affect the browser.
That said, I think Chrome has shown that the process per tab model can work and has interesting properties. I think (and hope) that Firefox is working on making this an optional feature, so that you can turn it on if you have the extra memory to spare or leave it off it you want to.
Also, for those who are worried about losing a dozen tabs if Firefox happens to die, worry no more. I regularly browse with > 40 tabs open at a time, and have never lost one. OK, crashes are very rare for me, but whenever they do happen (usually a windows crash rather than firefox) I just restart Firefox and it asks me whether I’d like to restore my tabs or not. I click YES, and then they’re all back.
Chrome’s hybrid multiprocess/multithread model is the best generalized solution for a complex and security sensitive application like a browser. It protects against unstable plugins and crash bugs in the browser itself. It lets users and developers clearly see the resource usage that their site/application is using. And allows the user to easily identify, and terminate, sites/apps which exhibit pathological resource consumption or other behavior. It’s just the right way to do it. Give me one good reason that my online banking session should live in the same virtual memory space as my YouTube session and OSAlert sessions. From a security standpoint, that’s insane.
Regarding the memory issue… the way Firefox hogs^Wutilizes resources to “improve” performance and “enhance” the user experience… the difference between threads and process/threads is likely to be in the noise. Especially considering the substantial memory savings of Webkit relative to Gecko.
Edited 2009-08-02 18:45 UTC
I don’t know why this myth about firefox being a resource hog persists, when it’s clearly no longer true. For example, I tested out Chrome using the same tabs I currently have opened in my Firefox session. (66 total, and Chrome still is unusable with that many because the tabs become so small that even the icons are removed).
Result:
Chromium 2.0.172.39
Private: 591,452k
Shared: 4,953k
Total: 596,405k
Virtual Private: 751,708k
Virtual Mapped: 437,980k
Firefox 3.5.1
Private: 403,940k
Shared: 11,344k
Total: 415,284k
Virtual Private: 409,612k
Virtual Mapped: 28,556k
That’s a pretty clear win for Firefox, and it’s typical of what I’ve seen. (Memory stats taken from the very cool Chrome about:Memory tab)
One thing Chrome is very good at is remaining very responsive even while under load. That’s the biggest issue I have with Firefox, is that sometime when it’s working on one tab the others slow down. That kind of issue is very noticeable by the users, and is a big win for the Chrome process model.
Edited 2009-08-02 19:13 UTC
Because it’s true. Even the difference between Firefox and Epiphany (which also uses Gecko) becomes quite clear when you have 70 users running 70 simultaneous browser sessions on one machine. And the difference is measured in *gigabytes*. The FF devs have a bad case of singleuseritis. And in the case of our servers, a terminal one, as we got rid of it.
I’ve not actually tested Chromium on our live system. But preliminary testing suggests that it would do as well or better than Epiphany for memory consumption. Although Epiphany does have the shared memory advantage in that match, as it is a Gnome app.
Edit: I should probably elaborate on my “singleuseritis” comment. There are way too many places in FF (the word “permeated” comes to mind) where it checks to see how much total system RAM is available, and allocates resources based upon that. Apparently, it never even occurred to the devs that anyone would (gasp!) run more than one instance on the same machine. 70 Firefox sessions each see 8GB of RAM, lick their chops, and dive in. It’s like your system RAM was being attacked by a school of piranha.
Edited 2009-08-02 19:29 UTC
Sure, but 99.9% of people are single users. So if you mean Firefox is a resource hog for your specific very rare situation, Ok, i agree. But let’s try to not pretend that your situation is typical.
By the way, it looks like my memory stats above were actually unfair. I gave chrome a while to rest and purge all it’s caches after loading the pages, and now it’s memory is down to within a few MBs of Firefox. So they actually seem to be about on par on my system.
Running corporate desktops as terminals is rare??? Are you serious?
OK. I have a total of between 80 and 90 browser users running this way. (And we’re small fish compared to larger companies which might have thousands, or more, users running as terminals.) How many browser users do you have running your way? (And I do want an answer to that.)
In any case, basing resource allocation off of total system RAM is insane even for a single user system. What if that user wanted to (gasp!) run an application like an office suite along side Firefox? FF assumes that it is the *only* application on the system. Which makes one seriously wonder about the FF devs’ judgment and/or mindset.
Also, FF is the *only* browser which acts in this insane way. So let’s try not to pretend that FF’s dubious policies are typical.
Edited 2009-08-02 19:58 UTC
Yes. I am. A million terminals like that is small potatoes to the worldwide market.
Well, I believe the latest stats showed FF at about 25% of the market – I think practically all of that is single-user windows systems, since FF isn’t used much on corporate systems, Linux only makes up a fraction of the market, and most OSX users stick with Safari.
If you want a more specific example, I can tell you about the company I work for. We have about 500 users running IE on their desktops, not through any terminals. Plus a handful running miscellaneous other browsers from their desktop. Maybe 1 or 2 running IE remotely through remote desktop at any one time. I’d call that a rare case – in fact I think if we could disable that functionality completely, we wouldn’t have more than a handful of complaints.
I didn’t, I just said yours weren’t either. Clearly you aren’t the target user they are trying to go after. I’m not saying that’s the right decision, just a fact. In fact, i think it’s pretty crazy that they’ve never created an msi installer for corporate use – but they apparently don’t care enough about corporate use to do so.
Edited 2009-08-02 21:07 UTC
Your answer to the portion of my questions which you chose to answer can be summed up as:
“Your use case, which only affects a few million people, doesn’t matter.”
And that, sadly, is fairly typical for the Firefox camp. Complaints all across the board… and all ignored. Only positive feedback is acknowledged.
I suspect that Chrome and other Webkit browsers are going to eat Mozilla Corp’s lunch over the next two years, not in small part due to that attitude. Note that Chrome/Chromium *already* has advantages over FF, despite FF’s years-long head start, for single user *and* terminal server use. And the evening is still very, very young. (And, of course, almost all of Mozilla’s revenue comes from Google, but that’s a different discussion.)
I note that you completely ignore my points regarding the obvious insanity, even on single user machines, of allocating resources as if FF was the only application installed. Which is simply the extension of the developers’ arrogant presumption into the behavior of the product itself.
I hereby predict that by Aug 2011 Mozilla devs won’t have the luxury of being so cavalier about losing a few million users.
You’re taking things out of context. I think i was fairly clear in saying that the use case you’re talking about is the exception rather than the rule. I never said it didn’t matter at all, except to mention that mozilla developers don’t care about it much, which is something you apparently agree with. My point, is that it doesn’t matter as much as the single user case which makes up the vast majority of use. Your posts seem to suggest the opposite – that the single user case doesn’t matter at all and only the multi-user system you’ve got setup is worth talking about at all. I respectfully disagree.
I think they’ll catch up in terms of market share as well, because they’ve simply got a better architecture and cleaner codebase, which should make future improvements faster. That said, I think you’re sorely underestimating the power of branding here. Lots of people stick with crap just because they’re familiar with the name, and Firefox got to the point it has today by using a massive word-of-mouth campaign. Consider that more people know what Firefox is then Chrome, even after Google’s been advertising it on their search page for months, and you see what a big mountain they still have to climb.
You’re right, that has nothing to do with this. So I’m not sure why you brought it up exactly…
Well, I ignored them because I have mixed feelings. I do think they’re not optimal, and yet I don’t think they’re as crazy as you obviously do. Probably because they have never actually given me any problems. In fact, I’ve never been able to use up the amount of RAM I’ve got on this computer (only 3.5GB) so it doesn’t affect me at all. And as I mentioned, comparing FF to Chrome resulted in equivalent memory usage for me, so the different allocation algorithms seem to have come to basically the same conclusion.
I still suspect that Mozilla’s market share will be growing by then, at the expense of IE. It’s competitors will likely have grown by larger amounts, though.
The point you seem to keep missing is that by making their determinations of memory allocation less pathological, they could easily handle the multiuser situation well without hurting grandma in the slightest. You know. Like every other browser out there that is not Firefox does? This is not some hard decision requiring a difficult judgment call regarding the relative importance of different users. This is a matter of fixing a totally brain dead and nonsensical algorithm for determining allocations which causes the pathological behavior that makes FF unsuitable for millions of desktop seats. If they want to make it self-tuning, then even using free memory (under Linux, Free + Cache + Buffer) instead of total system memory would be somewhat less insane.
I think, however, that we do agree that they don’t care. I just want to be very clear about just what it *is* that they don’t care about, that other browser makers do.
BTW, the Google revenue issue *is* very relevant… but we could have a whole other long thread about it. If and when that money dries up, Mozilla Corp is going to be looking for another source. (Rather desperately, I should think.) And I’ll bet corporate terminal rollouts won’t get such a cold shoulder then. Attention will inevitably turn to support money. And grandmas don’t pay for support contracts, as a general rule.
Edited 2009-08-02 23:03 UTC
I’m not missing that at all, I’m just saying it’s not as important as you think it is. I agree it would be nice if they did so.
It’s impossible to see into the future, but I rather imagine that if Google cut off funding right now, MS would be rather ecstatic to take over. In fact, I think Mozilla could probably set their prices quite a bit higher in exchange for using Bing instead of Google for defaults. Sure, MS may not be really happy about Firefox, but they’re pouring billions of dollars into Bing with the hope of competing with Google in the online advertisements business. The money they’ve thrown at IE lately is all rather minor in comparison and I don’t think they’d let that stand in the way. Of course, there are probably quite a few Mozilla contributors who would get upset and it would cause issues. But I don’t think they’re quite as reliant on Google as most people think.
Yeah. I thought about that. Mozilla could make a deal with Microsoft. And users everywhere would surely flock to help Microsoft with every search they made. (Everybody loves Microsoft and Bing!) That would make the current FF user base very happy, I’m sure. Maybe they could start including the Moonlight plugin by default, complete with a complimentary Mono runtime built right into the browser, and make some special marketing arrangement with Novell, to further enhance their image in the community.
And please get real. Mozilla Corp would miss Google’s tens of millions (annually) very badly, because it represents the overwhelming majority of the revenue stream that allows them the luxury of ignoring, and turning away, in fact, millions of potential business desktop users. They are every bit as dependent upon Google as people think.
Edited 2009-08-02 23:45 UTC
That’s the conventional wisdom of someone who reads OSAlert, I’m sure. But really, Mozilla has always been focused on the windows platform primarily. Mac support has come a very distant second, and they barely even acknowledge linux still exists. How many windows users are really that anti-microsoft? I’m sure any kind of deal like that would cause lots of anger, recriminations, and general chaos. But if they had no alternative, I think it would be doable.
Well, yeah. It would be doable for the Pope to go Protestant, as well. But there would be repercussions and consequences. I wonder how many of those techies who turned the masses on to FF in the first place would start recommending something else? (There are both philosophical and technical reasons to prefer Webkit.) I wonder how many of those little “Take Back the Web!” links to getfirefox.com would disappear and/or be replaced by something else?
Much of FF’s popularity has come from their “David” image as they’ve gone up against Goliath. I doubt that their setting up house together would do much for Firefox’s popularity.
Fortunately, the OSS web browser market is finally getting going. So it wouldn’t be a complete catastrophe. I’d just hate to see us end up back in a one OSS browser/renderer market again because of short-sightedness, arrogance, and overconfidence on the part of Mozilla Corp.
Edit: I guess I should also note that my professional life gives me a little wider perspective than just reading OSAlert. As is no doubt true of many people here. Try to keep that in mind.
Edited 2009-08-03 14:20 UTC
Two things:
a) Just separating plugins into processes is not enough for stability. Yes, Flash causes a lot of Firefox crashes, but Firefox itself causes a lot of Firefox crashes.
b) Firefox memory consumption is atrocious once you start actually doing stuff with it. It’s worst of all on Linux where the OS overcommits memory and then swaps out the entire f–king desktop stack when Firefox starts churning through memory, making it impossible to even open xkill or a shell to kill Firefox off. Not plugins, but Firefox itself that does this. All it takes is one bad JavaScript site.
Just opening a bunch of pages is not a good criteria. Open/Close many pages/tabs, watch some videos on youtube, submit some forms, play a flash game, check your gmail, etc, etc.
The memory fragmentation in Firefox is unavoidable, when closing a tab in Chrome – all memory will be reclaimed, not so with Firefox.
It’s still possible that Firefox will fare better in memory usage (as having a process per page comes at a cost). But Chrome is much snappier. I still use Firefox on my PC(love those extensions), but on my anemic eeepc – Chrome is sooo much faster…
The fact that it loads way faster than Firefox and handles more stuff more promptly does count a lot to me. Firefox starts feeling heavy rather fast no matter if I use it on my Macbook or on my netbook. Chromium flies by comparison (and I am not talking about HTML rendering speed, but about the way the chrome -hehe- reacts to user input).
“If a lack of third-party plug-in support (i.e. Flash) kept you from trying out Chrome on your Linux system, then avoid no longer.”
Nah… in my case, it was more along the lines of daily 12MB “updates,” extreme bugginess, incompleteness, and in general its obvious alpha status that hits you upside the head like a sledgehammer every time you run it…
Chromium was such a bad experience, I might just even bypass even the betas and *maybe* finally give it a chance once it hits its first official release. And even then, I seriously doubt it’ll become a browser I use on a daily (or even weekly) basis. I’m in no hurry really, considering Google Chrome on Windows didn’t exactly impress me from a user interface perspective. Under the hood, though, I admit that it makes some nice changes that would be great to have (processes per tab/tab separation and process monitor, specifically).
Edited 2009-08-03 01:01 UTC
So to build a browser on linux, I now need to install google build tools?
Totally unreasonable,