Chrome 10 released [LWN.net]

Google has announced the release of version 10 of the Chrome browser. New stuff includes better JavaScript performance, GPU-accelerated video, and a number of security features including better plugin blocking and sandboxing of the Flash plugin (on Windows only, alas).

to post comments

Chrome 10 released

Posted Mar 9, 2011 17:20 UTC (Wed) by rillian (guest, #11344) [Link] (13 responses)

Looks like they have in fact removed support for mp4 playback in the <video> element. Yay!

Chrome 10 released

Posted Mar 9, 2011 17:23 UTC (Wed) by clugstj (subscriber, #4020) [Link] (12 responses)

Wow, I haven't been paying attention, but they appear to have inflated the version number to 10. Needed to get ahead of IE?

Chrome 10 released

Posted Mar 9, 2011 17:42 UTC (Wed) by Kit (guest, #55925) [Link] (10 responses)

Chrome really doesn't advertise its version number. If you go to the Chrome website, you won't see the version number _anywhere_- unlike going to the website for pretty much any other web browser. Each of Chrome's releases generally also include fairly major changes that other browsers would hold on to for 6+ months while a bunch of other features stabilize at the same time.

Chrome's development model is just extremely aggressive and fast paced, and after using it extensively for quite a while, I've never had a single stability issue with it, or even it just acting 'funky'.

Chrome 10 released

Posted Mar 9, 2011 18:49 UTC (Wed) by rillian (guest, #11344) [Link] (9 responses)

I understand Google uses Chrome's live update feature to test patches. This allows developers to get rapid feedback on performance, usability and interaction with wild pages across their user base while bounding the amount of instability introduced. They can move very quickly because of this, but the trade-off is that users give up any control over what code they're actually running.

Chrome 10 released

Posted Mar 9, 2011 19:03 UTC (Wed) by Kit (guest, #55925) [Link] (8 responses)

There's not too many users out there with _legitimate_ reason to be running older, insecure versions of web browsers (those running old and insecure versions generally do so because updating the browser requires going out of your way). Those with legitimate reason are likely also in the position to be able to compile the version of Chromium they need from source.

Chrome 10 released

Posted Mar 9, 2011 19:53 UTC (Wed) by rillian (guest, #11344) [Link] (7 responses)

My discomfort with this arrangement is nicely illustrated by your use of "legitimate" there. The ability to choose what runs on one's computer is fundamental to software freedom. That not updating might be illegitimate implies some group of people other than the user must have a say.

I can think of a number of reasons why one might not update a particular application: dependence on specific behaviour, audit status, uniformity within an organization--none of which depend on one's facility with Chromium's build system.

Certainly it may be advantageous to subscribe to software updates from any given source, and open source provides great advantage in this because many parties can provide alternate or competing formulations of the same basic software.

Open standards allow even broader alternatives. Firefox is also a web browser, and it has not implemented this kind of testing, nor do they require automatic updates--although many of their developers would benefit from it--precisely because their focus is on user control and freedom.

Chrome 10 released

Posted Mar 9, 2011 20:03 UTC (Wed) by drag (guest, #31333) [Link] (2 responses)

> I can think of a number of reasons why one might not update a particular application: dependence on specific behaviour, audit status, uniformity within an organization--none of which depend on one's facility with Chromium's build system.

Yeah, how many are those actually legit? Seems most of it is just paranoia and being stuck with crappy internal websites more then anything else.

But....

I don't know how accurate this still is, Chrome annoyingly changes stuff and with all the excitement around it and the long long beta some documentation is way out of date... but pulling from a quick 'google':

Windows: Disable auto update through a registry key. This can be deployed via active directory group policies or other sane mechanism. If your managing lots of Windows systems and are not using AD, may God help you.. cause nobody else will. (until Samba4 reaches a stable release, of course.) :-P

MacOS: You can disable it through the 'defaults' command.

Linux: Chrome just uses the regular package management system. If you can't figure out how to disable automatic updates with that then you have serious problems. :-D

Chrome 10 released

Posted Mar 9, 2011 22:40 UTC (Wed) by johndrinkwater (guest, #65840) [Link] (1 responses)

The update to Chrome 10 disabled WebGL for me.
So I can understand the reason people would want to hold on to older insecure browsers that just bloody work.

Chrome 10 released

Posted Mar 9, 2011 23:59 UTC (Wed) by drag (guest, #31333) [Link]

I am happy with the fact that I get a browser that is up to date, directly from the developers, and I do not have to wait around for months for improvements while the distros re-engineer it to fit their own conception of how software should be properly compiled and packaged...

It's important to them, bless them, but it's mostly irrelevant for my purposes.

Chrome 10 released

Posted Mar 9, 2011 20:33 UTC (Wed) by Kit (guest, #55925) [Link] (2 responses)

> Firefox is also a web browser, and it has not implemented this kind of
> testing, nor do they require automatic updates--although many of their
> developers would benefit from it--precisely because their focus is on user
> control and freedom.

First off, what do you mean by 'this kind of testing'? Chrome's "testing" situation is similar to that of Debian than anything else: Stable vs Stable channel, Testing vs Beta channel, Unstable vs Dev Channel, features move down into Stable from Dev, via Beta (and there's also another Canary channel added a while back, closer to nightly builds).

Also, if you don't want Chrome to automatically update, just install the Standalone version: http://www.google.com/chrome/eula.html?standalone=1

In addition, Firefox also has an auto-updater- it's just less user friendly (it'll interrupt your work multiple times, and won't work if you're not an administrator/root... and last I checked, it wouldn't even automatically notify you of updates in those situations!).

Chrome 10 released

Posted Mar 9, 2011 21:14 UTC (Wed) by rillian (guest, #11344) [Link]

If the differential testing is opt in, and there's a way to turn off the auto updater, then I misunderstood the situation with Chrome. Apologies.

I also see I generally failed to make my point. There's a common itch to want to just take over management of everything for the convenience and benefit of one's users. And that's a great impulse while it really is beneficial. But like any situation were many people give power over themselves to a small group, it can easily become abusive. To the extent that we're Free Software advocates, it behooves us to watch such situations carefully and make sure real choice is in fact easy and available.

Chrome 10 released

Posted Mar 11, 2011 3:29 UTC (Fri) by jonabbey (guest, #2736) [Link]

It can still work if you're not an admin/root, assuming you have exploded the firefox.tar.gz in your own directory with your own user privileges.

Chrome 10 released

Posted Mar 9, 2011 23:46 UTC (Wed) by robert_s (subscriber, #42402) [Link]

"I can think of a number of reasons why one might not update a particular application: dependence on specific behaviour, audit status, uniformity within an organization--none of which depend on one's facility with Chromium's build system."

The biggest one I've found is testing. I discovered a bug in a webapp of mine with chrome which I didn't remember seeing before. So I needed to find out whether this was a bug in my code or a recently introduced bug in chrome. Of course, all the windows test machines I had access to had magically updated themselves to the latest chrome version, and I couldn't find any real way of testing against an older chrome.

I won't touch Chrome on any of my (real) systems because of its behaviour. Google treat the user's PC like another one of their websites. Theirs to do whatever they want with whenever they want.

Life with transparently self-updating applications is like living in a state of permanent amnesia.

Chrome 10 released

Posted Mar 9, 2011 22:37 UTC (Wed) by einstein (subscriber, #2052) [Link]

> they appear to have inflated the version number to 10.

Nope, the reason it's at version 10 is that it's the 10th iteration of the browser. There are no gaps in the succession of version numbers. 9 was stable and 10 was beta, now 10 is stable and 11 is beta, and so on.

Improved JavaScript performance

Posted Mar 9, 2011 17:33 UTC (Wed) by abacus (guest, #49001) [Link] (10 responses)

Impressive that JavaScript performance has been improved further. Last time I checked Chrome 9 finished the SunSpider benchmark already three times faster than Firefox 3.6.15.

Improved JavaScript performance

Posted Mar 9, 2011 18:27 UTC (Wed) by kripkenstein (guest, #43281) [Link] (8 responses)

> Impressive that JavaScript performance has been improved further. Last time I checked Chrome 9 finished the SunSpider benchmark already three times faster than Firefox 3.6.15.

Actually the SunSpider benchmark is a bad example. Both FF4 and IE9, which will be released this month, are better at that benchmark than Chrome (both due to features Chrome doesn't have: Tracing JIT in FF4, DCE in IE9).

Chrome dominates in Google's own V8 benchmark, though. And it is impressive that Chrome has improved even further in that specific benchmark.

Improved JavaScript performance

Posted Mar 9, 2011 18:37 UTC (Wed) by pranith (subscriber, #53092) [Link] (7 responses)

I just compared the beta of Firefox 4.0 and Chrome 10 in Sunspider benchmark and chrome beats ff4.0 by a whisker. (chrome is 1.056 times faster)

This being 4.0b13pre(2011-03-08), I do not expect much to change when ff4.0 final is released.

Improved JavaScript performance

Posted Mar 10, 2011 16:23 UTC (Thu) by tuos (guest, #43318) [Link] (6 responses)

Nitpicking..

1.056 times faster == more than twice as fast
0.056 times faster == faster by a whisker

So, the second one, right?

Improved JavaScript performance

Posted Mar 10, 2011 18:25 UTC (Thu) by alecs1 (guest, #46699) [Link] (5 responses)

I really think you're wrong. I'm not a native English speaker, but I've never heard "one time as fast", instead "two/three times as fast". That's a multiplication, not adding of percent (like in 100% percent faster). It holds in Romanian, it surely holds in my conversation with native English speakers.

Improved JavaScript performance

Posted Mar 10, 2011 21:32 UTC (Thu) by nybble41 (subscriber, #55106) [Link] (4 responses)

tuos is correct. Translated into math, if X is the original speed, then:

"Y times/percent as fast (as X)" => Y * X
"Y times/percent faster (than X)" => X + (Y * X)

So the original phrase "1.056 times faster" would actually mean slightly over twice the speed (X + (1.056 * X) = 2.056 * X). The correct phrase is probably "1.056 times as fast" or "0.056 times faster" (or "5.6% faster").

Improved JavaScript performance

Posted Mar 14, 2011 21:37 UTC (Mon) by bronson (subscriber, #4806) [Link]

So when people say "it's two times faster!" you claim that they mean it's 1+(2*1)=3X as fast?

That's clearly wrong. Here's lots of evidence: http://www.google.com/search?q=two+times+faster

You're trying to apply mathematical rigor to the English language.

Improved JavaScript performance

Posted Mar 22, 2011 15:44 UTC (Tue) by nye (guest, #51576) [Link] (2 responses)

>So the original phrase "1.056 times faster" would actually mean slightly over twice the speed

I'd bet both my kidneys that if you actually try talking like this in real life, you would be misunderstood 100 percent of the time. Because you're completely wrong, of course.

Improved JavaScript performance

Posted Mar 22, 2011 20:52 UTC (Tue) by nybble41 (subscriber, #55106) [Link] (1 responses)

I agree that many other people, perhaps most, would misunderstand. However, that is merely because they are used to authors using the phrases incorrectly. Some sources (e.g. <http://www.theslot.com/times.html>) go so far as to say "it's safe to assume that a writer is using the 'times more' phrasing erroneously". It is safest to simply word the statement another way, like "faster by a factor of 1.056", if that is what one actually means.

The plain and precise meaning of "X times faster" is "X times more speed", which reduces to "(original speed multiplied by X) more speed" => "(original speed multiplied by X) plus (original speed)" => "(X plus one) times the original speed" => "(X plus one) times as fast".

Consider that "50% faster" means "50% more than the original speed", not "50% of the original speed", and "50%" is identical to "0.5 times". Why should "105.6% faster" be interpreted any differently?

Obviously English is not a prescriptive language, and words can change their meaning over time; however, I would hate to see a useful phrase like "X times more" ruined in this way when we already have a perfectly good way to express the intended concept, "X times as much".

Improved JavaScript performance

Posted Mar 22, 2011 21:08 UTC (Tue) by bronson (subscriber, #4806) [Link]

> I would hate to see a useful phrase like "X times more" ruined in this way

Um, hello? Look at the link in my reply above. It's long since ruined and, I'm sorry to say, your one-man effort on LWN isn't going to change anything.

Improved JavaScript performance

Posted Mar 10, 2011 5:53 UTC (Thu) by rilder (guest, #59804) [Link]

You are comparing a very old version of firefox to latest version of Chrome. The result, hence is invalid. If you really want to compare, compare with the betas of Firefox 4 with releases of Chrome. I am sure if Firefox development versioned their betas, their version number will also be like 10 or 15.

Chrome 10 released

Posted Mar 9, 2011 22:19 UTC (Wed) by lab (guest, #51153) [Link] (2 responses)

Using it now on my Debian SID. Very nice, very polished, VERY fast.

Chrome 10 released

Posted Mar 15, 2011 6:44 UTC (Tue) by daniel (guest, #3181) [Link] (1 responses)

"Using it now on my Debian SID. Very nice, very polished, VERY fast."

And very lacking of any way to save all tabs as a bookmarks folder. Which makes it nothing more than an interesting toy for me. Compounded by forgetting how to restore itself when it crashes in various situations. So I use chrome for "throwaway" browsing and any serious search or research happens in Firefox. Not as wizzy in some respects but much more functional.

Had a chat with one of the Chrome devs, and this is not not perceived as a serious issue.

Chrome 10 released

Posted Mar 15, 2011 9:40 UTC (Tue) by jezuch (subscriber, #52988) [Link]

> Had a chat with one of the Chrome devs, and this is not not perceived as a serious issue.

Which one? The tabs-to-bookmarks-folder one, or the restore-broken-sometimes one? I agree that the former is a pain, but I rarely use bookmarks these days... (Weird, I know.) As for the latter, the only case it bit me was when it crashed while displaying the "do you want to restore?" page. (An oversight, perhaps?) Do you know of other scenarios, or is it random for you?

Chrome 10 released

Posted Mar 10, 2011 3:20 UTC (Thu) by clarkemaryland (guest, #73497) [Link] (1 responses)

google is a spy/nsa company

ok fools google is a spy/nsa company

never used to be but theyre infiltrated

gmail gspy gserver do the math you fools

my buddy at nsa told me goolge is g

do the math you fools

tor is nsa also

tor is nsa :)

Chrome 10 released

Posted Mar 10, 2011 13:18 UTC (Thu) by mpr22 (subscriber, #60784) [Link]

Nice to know. Have you heard the latest news? The Pope is a member of the Roman Catholic Church.