Now you can compare your Chromium browser with that other Chromium browser using Speedometer 3.0

The latest Speedometer, a benchmark measuring web responsiveness, is out in version 3.0. However, some browser makers have questioned how helpful such benchmarking is in the era of shared rendering engines. Speedometer is a tool to help browser makers spot regressions and improve performance. Its intention is to simulate what …

[Korev]

So does it also benchmark a zillion tracking cookies and tediously unselecting all ticks on GDPR dialogues?

[IGotOut]

The good news is that that's being sorted. Websites will be only allowed to display what is effectively Opt-In or Opt-Out. No going through scumbag media outlets that you have to visit 50 different pages to opt out.

[Patrician]

I came across a website that if you wanted to refuse cookies you had to individually disable nearly 400 different ones. Needless to say the browser back button was used instead.

[zipityzi]

It doesn’t test CPU JS execution?

> However, "Using a different computer with a different amount of memory, a different processor or a different operating system makes the biggest difference of all, and these tests do not cover any of that."

That is an unexpected comment. Shouldn’t Speedometer 3.0 also include the JS execution time, which would depend on the CPU?

[cyberdemon]

Re: It doesn’t test CPU JS execution?

It should.. The article's point is that this is a circle-jerk benchmark that only tests Chromium against itself. It is deliberately agnostic of hardware and conveniently not applicable to any non-Chromium browser..

[Andrew Hodgkinson]

Re: It doesn’t test CPU JS execution?

Untrue. The source article is bad; the author doesn't appear to understand how browsers work or are built.

Chrome and Safari have completely different JavaScript execution engines (V8 vs JavaScriptCore). That's part of why they've been trying to trade blows on performance all this time. Chrome on macOS historically had bad GPU acceleration support too, so even though the WebKit-based engines might assess the document markup at the same rate, the *painting* rate could wildly vary.

Just because Chrome might be doing OK on that front today does *not* mean that Google can be trusted to maintain or not otherwise break that tomorrow.

Moreover, Firefox still exists and plenty of people use it.

(Edited to add that "conveniently not applicable to any non-Chromium browser" is essentially all but gibberish and certainly nonsense; it tests various automated operations using a collection of popular JS frameworks just as it always has and, when I ran 3.0 a few minutes ago, gave Firefox 22.6 and Safari 22.1 - so Firefox "won").

[Grogan]

Re: It doesn’t test CPU JS execution?

No reason the author should know this unless intimate with the firefox build system, but browserbench's speedometer is one of the tests used for profiling the build. Benchmarks are good for profiling because they run through a lot of functions of the browser you wouldn't ordinarily hit just by running the software.

[Grogan]

Well, I use benchmarks to compare my build results. Did my optimization flags and build method show any improvement, am I just spinning my wheels, or have I degraded performance. I run through several, and the browserbench suite is one of them.

Benchmarks aren't useful for comparing different browsers in itself... try it on the same machine, with the same resolution/window size, then you're comparing the performance of Chrome vs. Mozilla vs. the other chrome etc.

It would be silly to use the benchmarks across different systems unless you're comparing that (I didn't specifically, but I sure noted the difference in browser benchmark scores on my new rig, for example)

[Grogan]

Now that I've tried it, what immediately comes to mind, with the change to this test, is that the scores are meaningless to me at this time. I have to establish a baseline again :-)

[chroot]

Firefox won

Here too: Chromium (no extensions) 13.8; Firefox (new profile) 18.6.