what sort of system are you on? I don't see any difference on my p2 400 256 ram
It was running on a 400 MHz P3 with 256MB RAM Linux Kernel 2.2 Machine displayed as X11 remote over a local 100 MBit LAN on a 400Mz P3 128 MB RAM also under Linux Kernel 2.2. (The remote Display should not cause the described effect, but it might be)
Assignee: asa → rogerl
QA Contact: doronr → pschwartau
This operates slowly for me too. I'm on a K6-2 300Mhz with 256Mb ram, building from the tip with --disable-debug --enable-optimize=-O2. The menus react approximately 2x slower than when in Nav4.7. Marking new. I suspect this can be linked to various speed problems, some XP and come Linux-specific.
Status: UNCONFIRMED → NEW
Ever confirmed: true
Summary: Java script menues function very slow → Java script menus function very slow
page appears down, please attach a testcase or something. either bug looks like a reasonable dupe given i can't actually see the page...
Assignee: rogerl → jst
QA Contact: pschwartau → desale
interestingly enough, this seems to be slower for me on win32 (2001052320) than 4.76. On linux I did not notice any difference. I wonder if we have any "DHTML is slow" meta bug.
Doron: yes, we're blocking it :).
There's nothing we can do about this now, marking Future for now...
Target Milestone: --- → Future
Another good example: http://www.likno.com/
Moving over pulldown menus of both http://www.likno.com/ and http://eulox.net works fine with build 2001102503, Win98, K6-III/400, 192MB. Is this time to close the bug? :-)
www.eulox.net works fast here (P3/500). marking WFM reporter: please reopen, if you disagree :-)
Status: NEW → RESOLVED
Last Resolved: 17 years ago
Resolution: --- → WORKSFORME
Wholeheartedly disagree. Did you compile your builds --include-speed or something? I just tested as follows: * Mozilla build taken from the tip today at 0800GMT, with gcc2.95.4, using the following options: ac_add_options --disable-debug ac_add_options --enable-optimize=-O2 ac_add_options --enable-elf-dynstr-gc * Netscape 4.76 Both on Debian Sid Linux, kernel 2.4.9; machine is AMD k6-2 300Mhz, 256MB RAM, 512MB swap. Plenty of memory free during testing. In Netscape 4, menus reacted 'instantly.' No noticable delay between my switching between entries and the final drawing on screen. In Mozilla, each menu change took 1-2 seconds, very often a glide over multiple entries caused only a few of the entries to be lit up, i.e. several were skipped. Very much less than what most users are used to and expect. Repening. *PLEASE* when you test performance stuff use both Linux (gcc can't do compiler optimisations as Mac/Windows ones can due to patents) and a two year old box minimum (300Mhz at most). I'm also requesting mozilla1.0 consideration; not vital but that depends what our priorities are.
Status: RESOLVED → REOPENED
Keywords: qawanted → mozilla1.0
Resolution: WORKSFORME → ---
I condirm that the problem hasn't been resolved on Win32 either !! On my K6-2-300, 256 Mo Ram, Win98SE and Build 201102403, I have the same slowliness James experimented with Linux. When you put your mouse on eulox.com DHTML links, there is a 1 second delay before the menu appears. With NS4.78, this is instantaneous. CPU usage when moving the mouse over the DHTML menu links is between 50 and 60%. Using NS4.78, it is between 15 and 20%. likno.com DHTML menus are so slow that they are amost unusable.
OS: Linux → All
Hardware: PC → All
Dimitrios : I agree that my CPU is not very powerful compared to today's machines, but I would not change the name of the bug because the problem is not that it is slow on older machines but that it is slow compared to IE6 on the same machine. Big cpu usage for a sluggish performance is not a normal behaviour. Furthermore, I am no programmer, but I guess that it could also affect the performance of Gecko on PDAs or any other devices which have even slower CPUs.
URL no longer works
URL is now dead. Resolving as wontfix.
Status: REOPENED → RESOLVED
Last Resolved: 17 years ago → 17 years ago
Resolution: --- → WONTFIX
You need to log in before you can comment on or make changes to this bug.