Browser War: FF3 vs. IE8 vs. G-Chrome!?

[Pikarya]

http://www.cnn.com/2008/TECH/ptech/0....ap/index.html

Most interesting part of the story;

When playing a YouTube video, Firefox 3 took up 95 percent of the CPU time on a three-year old laptop running Windows XP.

Chrome came in at 60 percent -- still too much. Especially since Google owns YouTube! You'd think it could make its browser work well with that site in particular.

Internet Explorer barely broke a sweat, taking up just a few percent.

When I told each browser to load eight pages, some of which were heavy with Flash and graphics, Firefox took 17 seconds and ended with a continuous CPU load of 50 percent. That means it took up half of my available processing power, even if I wasn't looking at any of the pages.

Chrome loaded them the fastest, at 12 seconds, and ended with a CPU load of about 40 percent.

Internet Explorer 8 took 13 seconds to load, but ended with no CPU load at all.

Opinions?

[Xno Kappa]

Can anyone personally test this? I'm not too tech-savvy but this sounds like some kind of Micro$oft propaganda.

[Correction]

Nice.

Test Browser performance by reading how well a 3rd-party Plugin (Flash) performs on a site.

Nice.

[octopus]

As a very unscientific test last night, I was encoding some anime into x264 (so my CPU is 100%), Chrome got very slow when the page has any kind of flash (ads) on the pages. But when the encoding was over, Chrome was extremely fast again.

I think this has to do with what Google mentioned about plugins running in low priority mode...

[The Blackrose]

All it does is cement the thought process that Flash needs to die in a fucking fire, along with webmasters who use it primarily to design their sites.

One of the biggest resource hogs ever.

[Myst]

it's a fucking beta.

[octopus]

http://lifehacker.com/5044668/beta-b...ich-is-fastest

[Zosi]

If I had to guess, I'd say that IE was hiding the CPU usage somehow. Flash plugins are binary blobs, they have very little to do with the browser outside of the plugin API, and Flash performance (and CPU use) should be identical on different browsers unless there's something very fishy going on.

[Katlan]

it's a fucking beta.

This, can't really test Chrome vs other browsers, at this point, when the words "beta" are written all over the package. People will obviously benchmark Chrome vs it's competition, but not like it matters since "beta" does not = "final client" product

[Miltani]

oh great pple forget there's a browser called opera out there

[Correction]

oh great pple forget there's a browser called opera out there

A 2% market share and terrible branding will do that to a product.

[VZX]

Using G-chrome now
It's fast, but when I see the children processes VMem size...

[Mizango]

oh great pple forget there's a browser called opera out there

I mentioned in the other thread I use Opera, not forgotten, I think the cpu savvy people here like to experiment with different browsers. I have a different one basically for all of my boxes with FF as the main always.

[Mizango]

Apparently Chrome uses an older version of WebKit that hasn't been repaired or upgraded yet. For you Safari users you may recall a before the latest patch when it would freeze and crash just out of the blue for NO reason? Its essentially is using that same shitty Webkit as Safari did back when they were encountering the crash issues and it apparently has some security flaws, thats to be expected with a new browser and all but it seems like there are people intentionally taking shots at it and trying to compromise security. There have been reports that apparently Chrome allows certain executable files to be automatically downloaded and ran to the user's computer without the user being prompted or even asked for permission to download said file(s).

Not trying to scare the masses, I am just putting that out there for anyone that isn't computer Savvy and thinking of replacing Firefox on day 2 with this. I know that sounds like a stretch but in this business you never assume everyone knows what you know. I've been playing around with it for the past 2 days at work and on a non-FF box I have at home that I use as dumb terminal.

I will post links and more things I find as I mess around with it more.


Edit: Here are security article links.

http://www.readwriteweb.com/archives...gle_chrome.php

http://www.eweek.com/c/a/Security/Go...y-Flaw/?kc=rss

[Kaylia]

Nice.

Test Browser performance by reading how well a 3rd-party Plugin (Flash) performs on a site.

Nice.

3rd party plugin maybe..but half of the website use flash (ok, probably less, but it's still something that should be considered). Now, I call bullshit on the result, hard to say what but it sound terribly biased.

[Haruk]

meh, right now chrome has a lot of bugs, but again, it's beta, and not like chrome 8 beta, or chrome firefox beta, it has a long way to go. But I definitely can see chrome eventually becoming the new firefox

[Siegtaru]

Apparently Chrome uses an older version of WebKit that hasn't been repaired or upgraded yet. For you Safari users you may recall a before the latest patch when it would freeze and crash just out of the blue for NO reason? Its essentially is using that same shitty Webkit as Safari did back when they were encountering the crash issues and it apparently has some security flaws, thats to be expected with a new browser and all but it seems like there are people intentionally taking shots at it and trying to compromise security. There have been reports that apparently Chrome allows certain executable files to be automatically downloaded and ran to the user's computer without the user being prompted or even asked for permission to download said file(s).

Not trying to scare the masses, I am just putting that out there for anyone that isn't computer Savvy and thinking of replacing Firefox on day 2 with this. I know that sounds like a stretch but in this business you never assume everyone knows what you know. I've been playing around with it for the past 2 days at work and on a non-FF box I have at home that I use as dumb terminal.

I will post links and more things I find as I mess around with it more.


Edit: Here are security article links.

http://www.readwriteweb.com/archives...gle_chrome.php

http://www.eweek.com/c/a/Security/Go...y-Flaw/?kc=rss

This right here (including the articles) are enough to make me and I hope anyone, computer savvy (which I am) or not, to not go near Chrome and stick to your FF or IE or whatever until they at least patch it. I wouldn't want to have this shit on my computer if I played FFXI on my PC (which I don't, thank God) with all of these supposed hacking applications infiltrating players through frequented sites.

I don't want to have to wipe my computer again thanks to some crappy trojan, basically. ^^; Firefox onry

[Khamsin]

Why couldn't they sandbox plugins/flash content, by the way?

I can sandbox the entire browser itself, and any exploits used by a flash security exploit would be contained. Why can't they sandbox the plugin, and if it wants to write something or send something, it can pop up a box telling you what it's doing and asks if you want to trust it?

[MisterBob]

Why couldn't they sandbox plugins/flash content, by the way?

I can sandbox the entire browser itself, and any exploits used by a flash security exploit would be contained. Why can't they sandbox the plugin, and if it wants to write something or send something, it can pop up a box telling you what it's doing and asks if you want to trust it?

They said they've been working on it, but so far they haven't been able to really get this type of concept to work. The problem is that the plugins often require higher level access on the machine due to the way they have been developed. Without someone creating a clone of certain aspects of your machine, there's no way to sandbox them without getting the actual plugin developers to change their code.

[Khamsin]

They said they've been working on it, but so far they haven't been able to really get this type of concept to work. The problem is that the plugins often require higher level access on the machine due to the way they have been developed. Without someone creating a clone of certain aspects of your machine, there's no way to sandbox them without getting the actual plugin developers to change their code.

That's what a regular sandbox does when a sandboxed program wants to write or modify something, but the size of the file that legit plugins have access to are fairly small and can be cloned without issue. The only ones that would cause problems are certain plugins that you intend to have access to your computer as a whole, like the kaspersky scanner, etc. although even that doesn't write, only reads.

If I can take IE and run it through Sandboxie, what's preventing Chrome (which they're writing from scratch) from doing it for a tab containing a plugin?