Discuss JaegerMonkey and bugs here

[RyanVM]

@sunspider its not really faster
263ms vs 288ms nightly some days ago....

As would be expecting judging by the numbers on AWFY...

[mat--]

From my testing methodJIT for chrome causes memory leaks with either ABP or Noscript.

[Bluefang]

What does methodjit.chrome means?
Why is it set to false and what will it bring? (performance?)
I didn't understand things well.
Excellent work, love to see more improvements.

"chrome" is the name of the Firefox interface.

So javascript.options.methodjit.chrome enables JM for Firefox its self.
And javascript.options.methodjit.content enables JM for web pages.

[_Dexter_]

javascript.options.methodjit.chrome enabled causes problems with FEBE 7 (beta version for Fx4) and LastPass.

[pal-moz]

delete

[neo86]

http://xplsv.com/prods/demos/xplsv_orsotheysay/ (push , to see the fps counter smooth as butter)

Not seeing that. My computer isn't that impressive (Axthlon II X4 620), but it's only "smooth as butter" in Google Chrome.

[Omega X]

Can anyone with a Netflix account can confirm that the arrows flow the content in the wrong direction?

[gnwaa]

Can anyone with a Netflix account can confirm that the arrows flow the content in the wrong direction?

Confirmed.

Actually, both arrows scroll the content towards the right, so only one arrow is broken (the right one).

[Omega X]

IT works with MethodJIT disabled.

Filed:
https://bugzilla.mozilla.org/show_bug.cgi?id=595585

EDIT:
The "Add To Queue" button is also having issues.

[neo86]

http://xplsv.com/prods/demos/xplsv_orsotheysay/ (push , to see the fps counter smooth as butter)

Not seeing that. My computer isn't that impressive (Axthlon II X4 620), but it's only "smooth as butter" in Google Chrome.

Seems to be an issue with D3D9 layers actually. Without accelerated layers I get ~7 fps instead of 2-3. Still half as fast as Chrome, but not as bad I guess.

[WonderCsabo]

I've tested JaegerMonkey, too.

Minefield, D2D, layers OFF:

SS: 462.8ms
V8: 1880

Minefield, D2D, layers ON:

SS: 473.4ms
V8: 1300

Other browsers:

Opera 10.61:
SS: 377.4ms
V8: 3222

Chrome 7.0.517.0 dev:
SS: 343.6ms
V8: 5307

[sciguyryan]

I can't believe everyone is still testing against the botched V8 test in an event. The test is crappy and doesn't even deserve to be called a test, since it is all messed up (there have been numerous discussions about making a corrected version of it on Bugzilla).

I have one more test idea for you, try turning off TM for content and see how that affects it.

I tested this with the JS NES tool, and it works quite well with JM. MUCH better than TM alone.

[WonderCsabo]

I can't believe everyone is still testing against the botched V8 test in an event. The test is crappy and doesn't even deserve to be called a test, since it is all messed up (there have been numerous discussions about making a corrected version of it on Bugzilla).

I have one more test idea for you, try turning off TM for content and see how that affects it.

I tested this with the JS NES tool, and it works quite well with JM. MUCH better than TM alone.

Everybody tests V8, because it's on AWFY, too...

On SS, i get 20ms better with JM only...

BTW: how can is get the ms value on V8 test instead of points?

[phuzi0n]

http://xplsv.com/prods/demos/xplsv_orsotheysay/ (push , to see the fps counter smooth as butter)

Not seeing that. My computer isn't that impressive (Axthlon II X4 620), but it's only "smooth as butter" in Google Chrome.

Seems to be an issue with D3D9 layers actually. Without accelerated layers I get ~7 fps instead of 2-3. Still half as fast as Chrome, but not as bad I guess.

20-30 FPS in chrome but only 1-2 FPS in Firefox JM+TM. I tried enabling/disabling varying combinations of D2D/DW, D3D layers, and trace JIT content but I couldn't find any combination that worked above 2 FPS. Athlon X4 635 + Radeon HD4650.

[sciguyryan]

Yeah, they may test it on AWFY but that's only because nobody has come up with another test to replace it.