You are not logged in.
Does anyone know of a way of enabling TraceMonkey on Arch64? A patch to backport, or config option?
Why wait? Why not just compile a bin32-firefox-pgo?
Last edited by DevoidOfWindows (2009-07-02 04:33:28)
Offline
Does anyone know of a way of enabling TraceMonkey on Arch64? A patch to backport, or config option?
A
have to change javascript.options.jit.content to true in about:config, but looks like its not fully supported for 64 yet.
Offline
And while I am a big fan of non-patented standards, etc., Ogg Theora is... well... not good, not even decent. Compared to x264-encoded H.264, it's a step back a couple decades (two, to be exact).
...snip...
No doubt. I wonder if the Mozilla people actually fell for the Vorbis/Theora patent free bit or if they came to their senses and decided not to care?
Anyway, this release does feel somewhat quicker, but more importantly to me, it feels a lot more fluid.
Offline
Ranguvar wrote:Does anyone know of a way of enabling TraceMonkey on Arch64? A patch to backport, or config option?
DevoidOfWindows wrote:Why wait? Why not just compile a bin32-firefox-pgo?
Because I like my 64-bit browser with its 64-bit plugins, I don't want a crapload more 32-bit libs on my system, and I find major applications being bin32- creates problems more often than not. I knew I could use a bin32 ![]()
have to change javascript.options.jit.content to true in about:config, but looks like its not fully supported for 64 yet.
I had that set to true by default, but I had javascript.options.jit.chrome (which controls using TraceMonkey for Firefox itself (basically), not the pages) set to false (I then made it true).
Don't know TraceMonkey is on or off, really, I have to do some tests. It does feel faster than FF 3.0, but I'm sure TraceMonkey isn't the only speed improvement.
No doubt. I wonder if the Mozilla people actually fell for the Vorbis/Theora patent free bit or if they came to their senses and decided not to care?
Came to their senses? Not trying to derail the thread, but I believe pushing like this for non-patented software is a great idea (and may encourage more unpatented R&D). I was just trying to contribute to general knowledge about video codecs
And yeah, they chose Theora because it was unpatented. They might have run into trouble using a patented codec, maybe not, but the main reason was no patents for Theora (though as I've said, this isn't 100% certain, unlike MPEG-1). Oh, and by the way, Vorbis kicks ass (really). HE-AACv2 can beat it, but only at reeeaaally low bitrates. Otherwise, Vorbis is top dog for lossy audio.
http://www.0xdeadbeef.com/weblog/?p=977
Offline
This is what the wiki says about tracemonkey https://wiki.mozilla.org/JavaScript:TraceMonkey
In my case javascript.options.jit.chrome is set to false (default value) and javascript.options.jit.content is set to true (default value) so I guess tracemonkey is enabled for pages rendering (unless something fails and it falls back to the older method).
Edit:
I forgot to say but FF3.5 does feel more fluid and pages do seem to "hang" less, this on x86_64.
Last edited by R00KIE (2009-07-02 08:11:04)
R00KIE
Tm90aGluZyB0byBzZWUgaGVyZSwgbW92ZSBhbG9uZy4K
Offline
Hi, today I updated my laptop and I saw that firefox 3.5 was present! My system is an x86_64..Everything went fine during the update, but now I can start firefox only as root.
If I try as my user, it doesn't gives no output from the terminal...
I guess could be a problem with permissions, do you have any idea??
Thanks a lot
Offline
@marconato - can you post your /etc/pacman.d/mirrorlist for me? I just updated and didn't pull down the 3.5 package at all yet it is present on the arch package search:
http://www.archlinux.org/packages/?arch … =&limit=50
x86_64 Extra firefox-i18n 3.5-1 Language packs for Firefox jgc 2009-07-01
x86_64 Extra firefox 3.5-1 Standalone web browser from mozilla.org jgc 2009-07-01
x86_64 Extra totem-plugin 2.26.3-1 Totem mozilla/firefox plugin Orphan 2009-07-01
EDIT: nevermind... dunno what the problem was w/ it but after changing servers I have a whole lot of packages to update suddenly!
Targets (67): apr-1.3.5-1 readline-6.0.00-1 unixodbc-2.2.14-2 apr-util-1.3.7-1 bash-4.0.024-1
bash-completion-1.0-2 brasero-2.26.3-1 cairo-1.8.8-1 gnome-keyring-2.26.3-1
sqlite3-3.6.15-2 heimdal-1.2.1-5 libgcrypt-1.4.4-2 gnutls-2.8.1-1 libxml2-2.7.3-2
libsoup-2.26.3-1 libgweather-2.26.2.1-1 evolution-data-server-2.26.3-1
gnome-desktop-2.26.3-1 cheese-2.26.3-1 device-mapper-1.02.32-2 python-2.6.2-3
libxt-1.0.6-1 xulrunner-1.9.1-1 epiphany-2.26.3-1 evolution-exchange-2.26.3-1
firefox-3.5-1 jack-audio-connection-kit-0.116.2-2 fluidsynth-1.0.9-1
gcalctool-5.26.3-1 gedit-2.26.3-1 gnome-menus-2.26.2-1 libwnck-2.26.2-1
gnome-panel-2.26.3-1 gucharmap-2.26.3.1-1 gnome-applets-2.26.3-1 guile-1.8.6-2
gnome-games-2.26.3-1 gnome-mag-0.15.7-1 gnome-python-extras-2.25.3-2
gnome-session-2.26.2-1 gnome-terminal-2.26.3.1-1 gnome-themes-2.26.3.1-1 gnupg-1.4.9-4
gnupg2-2.0.12-1 gphoto2-2.4.5-2 gtkhtml-3.26.3-1 hamster-applet-2.26.3-1
libepc-0.3.10-1 libthai-0.1.12-1 libxaw-1.0.6-1 lua-5.1.4-4 lvm2-2.02.47-2
mkvtoolnix-2.9.7-1 mousetweaks-2.26.3-1 ntp-4.2.4p7-3 openjdk6-1.5-3 orca-2.26.3-1
pango-1.24.4-1 parted-1.8.8-3 postgresql-libs-8.3.7-2 qt-4.5.2-1 ruby-1.8.7_p173-2
smbclient-3.3.6-2 totem-2.26.3-1 wpa_supplicant-0.6.9-2 xz-utils-4.999.8beta-4
yelp-2.26.0-3
Total Download Size: 259.57 MB
Total Installed Size: 791.96 MBLast edited by graysky (2009-07-03 10:38:25)
Offline
I was using the swiss repo I think. After that, I changed to a french one, and I got a big list of updates, but not as yours..
What repo are you using?
Offline
For me firefox (I use 3.6.a1pre) is even faster than uzbl and midori at sunspider benchmark
firefox-hg-pgo:
Total: 2194.8ms +/- 5.7%
uzbl:
Total: 2428.0ms +/- 2.0%
midori:
Total: 2297.8ms +/- 2.6%SIC!
CPU is Xeon X3360 @ 8.5x400=3.40 GHz
Arch x64_86
Firefox 3.0.11 (Sunspider 0.9) Total: 2515.6ms +/- 7.5%
Firefox 3.5 (Sunspider 0.9) Total: 2279.4ms +/- 5.1%
Offline
I was using the swiss repo I think. After that, I changed to a french one, and I got a big list of updates, but not as yours..
What repo are you using?
$ cat /etc/pacman.d/mirrorlist
# Server list generated by rankmirrors on 2009-07-03
# United States
Server = http://mirror.cs.vt.edu/pub/ArchLinux/$repo/os/x86_64
Server = http://mirrors.easynews.com/linux/archlinux/$repo/os/x86_64
Server = http://mirror.umoss.org/archlinux/$repo/os/x86_64
Server = http://www.gtlib.gatech.edu/pub/linux/distributions/archlinux/$repo/os/x86_64
Server = http://mirror.archlinux.com.ve/$repo/os/x86_64
Server = ftp://ftp.archlinux.org/$repo/os/x86_64Don't forget to do a pacman -Syy when you modify this file.
Offline
What about firefox 3.5 in the i686 repo?
Offline
What about firefox 3.5 in the i686 repo?
No offense, but did you even bother searching for this before posting?
Offline
Haha, I am in a hurry and yes i used the forum search, maybe I should visit the oculist.
But thank you.
Offline
============================================
RESULTS (means and 95% confidence intervals)
--------------------------------------------
Total: 2133.0ms +/- 1.3%
--------------------------------------------3d: 285.2ms +/- 4.7%
cube: 90.2ms +/- 7.0%
morph: 63.2ms +/- 2.6%
raytrace: 131.8ms +/- 6.6%access: 293.0ms +/- 2.7%
binary-trees: 84.2ms +/- 10.9%
fannkuch: 123.6ms +/- 3.7%
nbody: 62.4ms +/- 4.8%
nsieve: 22.8ms +/- 8.1%bitops: 60.0ms +/- 2.9%
3bit-bits-in-byte: 2.0ms +/- 0.0%
bits-in-byte: 10.8ms +/- 5.1%
bitwise-and: 2.8ms +/- 19.9%
nsieve-bits: 44.4ms +/- 4.2%controlflow: 74.0ms +/- 3.4%
recursive: 74.0ms +/- 3.4%crypto: 109.6ms +/- 3.7%
aes: 63.4ms +/- 7.4%
md5: 30.8ms +/- 5.3%
sha1: 15.4ms +/- 4.4%date: 279.8ms +/- 3.4%
format-tofte: 150.4ms +/- 6.2%
format-xparb: 129.4ms +/- 2.6%math: 117.0ms +/- 13.1%
cordic: 71.8ms +/- 16.7%
partial-sums: 31.8ms +/- 12.2%
spectral-norm: 13.4ms +/- 5.1%regexp: 170.0ms +/- 12.2%
dna: 170.0ms +/- 12.2%string: 744.4ms +/- 5.8%
base64: 33.8ms +/- 3.1%
fasta: 163.2ms +/- 15.2%
tagcloud: 204.0ms +/- 7.4%
unpack-code: 269.2ms +/- 19.1%
validate-input: 74.2ms +/- 8.5%
PGO: Compiling...
Last edited by DevoidOfWindows (2009-07-03 17:43:30)
Offline
User of 64 bits vote for this bug
https://bugzilla.mozilla.org/show_bug.cgi?id=489146
To enable TM on 64 bits systems
Thanks
Offline
Voted.
Offline
Edit: Never mind. It seems to be a GTK issue.
---------
Anybody know why FF stopped noting Alt Gr (after update, that is)? For example: I have to press Alt Gr + 8 to get [ but FF doesn't respond in any way.
In other apps things work fine.
Last edited by peabrain (2009-07-09 15:24:51)
Offline