Chrome 6: Early peek under the hood | Hacker News Reader

Chrome 6: Early peek under the hood

geek.com

64 pointsthomas21 comments

http://img.deepgamers.com/1278097072.png A friend of mine has benchmarked the beta of Chrome along with several other browsers on his own PC. It's pretty impressive except for that weird box that gets cut out from the acid3 test.

why the need for such a large picture to show such little info?

Why only prereleases for IE and Chrome? Also that HTML5 test is bogus as its weights are completely arbitrary, counts things that were rejected by the W3C like Web SQL Database, and does not verify correctness (e.g., Chrome's implementation of WebSockets does not match the current draft.)

The HTML5 test doesn't include rejected features in its score. It gives a score based on actual HTML5, and gives a separate number of "bonus points" based on related technologies that make the browser more compatible with others.

One thing I do wish is that they'd stop mucking about with the security model in places that weren't important. Ever since they stopped allowing local xml files to render with stylesheets, my life has grown much harder.

Does the Chrome team think they need to reach version number parity with the other browsers? At least from a UI standpoint, the current version of Chrome that I'm using (5.0.375.99 Beta) is not enough different from the first Beta to warrant 4 full version numbers. About the only features that I can think of that are visibly new are the themes and the extension system. Did themes come with v4?

Also, if there are good reasons to warrant full version number change, then why are those versions installing automatically? A full version change is not an "update". It should be much more explicit to the user that a new full version is available.

Version number parity is kind of amusing and silly because version numbers are arbitrary, right? But if version numbers are arbitrary, then on what basis can one say that a higher number is not warranted?

if version numbers are arbitrary, then on what basis can one say that a higher number is not warranted?

I'm just really angry that they're changing meaningless numbers as if they have meaning. They're such idiots for thinking the numbers matter enough to change them.

But if they're meaningless anyway, what does it matter if they change them?

Yes, that was the joke. Thanks for explaining.

I re-read your comment and I can see it now, but I missed the humor first time around.

They are changing them to push updates. Your browser won't update unless a version with a higher number is available, so they have to increase the number.

Most companies bundle updates (example: windows service packs) which means they only need to increase version number a few times. Chrome decided to update continuously instead.

The number changes are not meaningless; there is change in the background, you just don't see it.

No. I'm not involved with Google or Chrome but I follow a few tickets.

4, 5, and 6 are just milestones for project planning. Version 6 is about feature parity across platforms. They've been cranking on code quite intensely and therefore hitting the milestones quick.

This last v5 to v6 change was a little surprising as the browser skin changed, but there are no major features that make v6 wildly different from v5. It's just a continuum of tiny updates silently getting applied in the background.

It has nothing to do with Firefox 4 or IE 9.

4,5,6 are just milestones for features.

Branches (6.0.477.0) = 477 are self explanatory.

"Missing" or arbitrary branch numbers are mostly due to failed ideas.

It seems the chrome team simply doesn't do "releases". Chrome evolves over time, and every so often they bump up the release number, but they're not working to a release in the same way as Photoshop or Firefox. This reduces the significance of the release number, but that's just a product of the differing development model.

I'm pretty sure we are just being exposed to the internal developer versions, which don't have the arbitrary stupidity of public-facing versions. All the numbers mean is 'this build is more recent than the build you have on your machine'.

The Chrome team uses version numbers in a very developer centric way. Any time they want to make a major breaking change they bump the major version number, irrespective of any marketing needs. This has the added benefit, as was intended with version numbering from the get-go, of making it dead-easy to determine which major features are in which release.

Wonder how many sites search for 6 in the user agent to filter I.E. 6 or to set a custom css file

None worth visiting I'm sure.

That just wouldn't work on so many levels.. So many version strings of, i assume, every browser have the number 6 in them.

Maybe I'm not understanding you?

  <!--[if IE 6]>
    <link rel="stylesheet" type="text/css" href="ie6.css" />
  <![endif]-->

The standard way to show a custom stylesheet to IE6 only is to use the IE6 conditional, which is ignored by all browsers except IE6, because it's semantically a comment and other IE browsers will ignore it, because they don't match the version number.

Reply on news.ycombinator.com