>For Speedometer, these optimizations have resulted in a 10% improvement since August 2024.
The score is going from 42.8 to 53.4. That is 25% speed increase. I am not sure how that 10% figure came from.
And 25% improvement on top of an already insanely well optimised browser is quite an amazing achievement. Apple M4 Safari is something close to 48 / 50.
And Chrome is now great with memory too. They went through a few years of Firefox Memshrink like cycle. I still dont like the interface, but in terms of multi tab browsing it is a lot better. Probably not Firefox few thousands tabs great but definitely very useable for up to a few hundred tabs.
Safari continues to be the worst browser. Or may be good for Mobile and a few tabs usage only.
hedora · 1d ago
I have to use chrome at work for compatibility testing, and the enshitification is only rivaled by win 11.
For instance, if you use “login with google” with a work account on a third party website, it now displays a dark pattern dialog saying you are allowing your employer to monitor 100% of your browsing activity, install crapware, download your history, bookmarks, etc.
It’s unclear if “etc” includes passwords, totp and passkeys.
Anyway, clicking cancel seems to opt out but let the authentication flow complete.
I don’t understand why people are switching to this garbage. There’s ~zero cost to switching web browsers.
tekla · 1d ago
There is cost. 99% of Web Devs only test on Chrome. I've been on several sites related to healthcare, engineering, govt sites, that do not work on Firefox.
karmakaze · 1d ago
Several of hundreds/thousands of sites you visit? So FF only works 99.x% of the time.
peterlada · 1d ago
As one of the monitizable subject I am eternally indented to my trillionaire corporate overlords.
Seriously, break up Google already.
dtgm93 · 1d ago
This is just measuring the speed at which chrome superficially is finished rendering and displaying content?
Actual cpu/memory performance of a given task, or some real world measure of a program's speed, efficiency, reactiveness, etc isn't indicated... I would be more interested in a Bloatscore metric!
msdz · 1d ago
> Actual cpu/memory performance of a given task, or some real world measure
Take a look at what the Speedometer 3 test suite includes.
The score is going from 42.8 to 53.4. That is 25% speed increase. I am not sure how that 10% figure came from.
And 25% improvement on top of an already insanely well optimised browser is quite an amazing achievement. Apple M4 Safari is something close to 48 / 50.
And Chrome is now great with memory too. They went through a few years of Firefox Memshrink like cycle. I still dont like the interface, but in terms of multi tab browsing it is a lot better. Probably not Firefox few thousands tabs great but definitely very useable for up to a few hundred tabs.
Safari continues to be the worst browser. Or may be good for Mobile and a few tabs usage only.
For instance, if you use “login with google” with a work account on a third party website, it now displays a dark pattern dialog saying you are allowing your employer to monitor 100% of your browsing activity, install crapware, download your history, bookmarks, etc.
It’s unclear if “etc” includes passwords, totp and passkeys.
Anyway, clicking cancel seems to opt out but let the authentication flow complete.
I don’t understand why people are switching to this garbage. There’s ~zero cost to switching web browsers.
Seriously, break up Google already.
Actual cpu/memory performance of a given task, or some real world measure of a program's speed, efficiency, reactiveness, etc isn't indicated... I would be more interested in a Bloatscore metric!
Take a look at what the Speedometer 3 test suite includes.
I still chuckled at Bloatscore, though.