In terms of javascript yes it is faster but once they are both down to that level the difference is not really that noticeable or at least from what I am doing. However the difference between these two and the others listed is very noticeable.
Got a tip for us?
Let us know
Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
Significant Javascript speed improvement of firefox 3 nightly
- Thread starter clevin
- Start date
- Sort by reaction score
In terms of javascript yes it is faster but once they are both down to that level the difference is not really that noticeable or at least from what I am doing. However the difference between these two and the others listed is very noticeable.
correct, CURRENT nightlies.
a bug fix just checked in today for firefox, promises a 2 percent improvement on javascript performance
PGO also targeting firefox trunk for mac, which promises a 10-25 percent improvement on all performances
I have no other data. so there may or may not be other improvement. I will report back whenevr i got update.
yes indeed. everything is so fast nowadays, all these speed tests are running each of the script 1000000 times in order to measure statistical significant differences. when difference are within 2000ms, it really has no real world impact for anybody anymore. So eventually, I dont expect webkit nightly and firefox 3 have any significant difference on this performance.
Firefox might benefit little bit more since its UI is written in XUL.
Here is another javascript speed test
http://wd-testnet.world-direct.at/mozilla/dhtml/funo/jsTimeTest.htm
as you can see, each script has to be run 1 million times to obtain the result (I believe sunspider runs script that many times as well).
I mentioned above that, at this stage, statistical numbers represent no real world impact dealing with javascript website anymore (Unless either team can improve 200%, possible, but unlikely). So its just an normal update.
There are two bug fix landed on trunk yesterday.
https://bugzilla.mozilla.org/show_bug.cgi?id=421154
https://bugzilla.mozilla.org/show_bug.cgi?id=415455
Altho promises a 8% and a 2% win, I only observe the 2%, no sure where that 8% will or will not show up.
There are two bug fix landed on trunk yesterday.
https://bugzilla.mozilla.org/show_bug.cgi?id=421154
https://bugzilla.mozilla.org/show_bug.cgi?id=415455
Altho promises a 8% and a 2% win, I only observe the 2%, no sure where that 8% will or will not show up.
Attachments
If you truly believe the "no real world impact" reasoning, I question why you continue to post screenshots comparing the two. Plus, I think you are being disingenuous by comparing Firefox nightlys with Safari 3.0.4. You should be comparing with Webkit nightly or not using Firefox nightly.
well, if I were really comparing the firefox and safari, at this stage, I would compare firefox nightly and webkit nightly
however, My purpose is to show the "improvement" of firefox 3.
1. I can not rule out the chance that webkit will improve as well, so its not a good standard to compare against.
2. I can not use firefox 2 as well, since two copies of firefox can't be open at the same time.
3. I can not use opera 9.2 since it will not finish the test.
4. I can not use opera 9.5 for the reason same as #1.
5. I can not only list firefox's data since somebody might question if the test environment was the same.
It probably means very little, but if anybody cares, firefox 3 nightly is now officially fastest in javascript performance on all platforms. It comes in 1.02x faster than webkit 31090 (today's nightly)
Just a head up of whats to come on this topic.
Mozilla is changing their building machine from Tiger (xcode 2) to leopard (xcode 3). My test shows different improvement with different version of OSX.
Under Tiger, this change speed up sunspider by ~3%
Under leopard, this change speed up sunspider by ~8%
There might be more fix to come, but firefox 3 beta5 code freeze is tomorrow. I expect they would put more energy into spidermonkey 2.0 afterwards. or 1.9.x, we will wait and see.
PGO seems unlikely to make into mac firefox 3 at current speed.
Again, 1.02x faster is statistically, not to mention practically, insignificant. But anything can stop apple from making absurd statement is good, IMHO.
Just a head up of whats to come on this topic.
Mozilla is changing their building machine from Tiger (xcode 2) to leopard (xcode 3). My test shows different improvement with different version of OSX.
Under Tiger, this change speed up sunspider by ~3%
Under leopard, this change speed up sunspider by ~8%
There might be more fix to come, but firefox 3 beta5 code freeze is tomorrow. I expect they would put more energy into spidermonkey 2.0 afterwards. or 1.9.x, we will wait and see.
PGO seems unlikely to make into mac firefox 3 at current speed.
Again, 1.02x faster is statistically, not to mention practically, insignificant. But anything can stop apple from making absurd statement is good, IMHO.