Closed Bug 171057 Opened 22 years ago Closed 22 years ago

bad interactive performance when javascript is enabled

Categories

(Core Graveyard :: Plug-ins, defect)

x86
Windows 2000
defect
Not set
normal

Tracking

(Not tracked)

VERIFIED DUPLICATE of bug 132759

People

(Reporter: me, Assigned: rubydoo123)

References

()

Details

(Keywords: perf)

Attachments

(1 file)

User-Agent:       Mozilla/5.0 (Windows; U; Windows NT 5.0; en-US; rv:1.1) Gecko/20020826
Build Identifier: Mozilla/5.0 (Windows; U; Windows NT 5.0; en-US; rv:1.1) Gecko/20020826

Almost all pages under sina.com.cn have such problem. The page renders
correctly, but with javascript the performance is bad. Having several tabs from
it's homepage link open, one can hardly have anything done in mozilla and the
process loads constantly all the CPU idle time.

Reproducible: Always

Steps to Reproduce:
1. Go to http://www.sina.com.cn.
2. The interactive speed is not good even with the single page loaded.
3. Open some links concurrently either in new window or new tab, you'll get
really bad performance.

Actual Results:  
Browser acts very slow, loads CPU very high, could even hang mozilla forever.

Expected Results:  
Act as fast as, for example IE.
My system: WinNT4.0(SP6)  500 MHz  128 RAM

I'm having no performance problems at all with this site with 
Mozilla trunk build 2002092308 on WinNT4.0. With the main page open,
and four new tabs opened from links there, my CPU remains low.
(around 5-7%, with occasional spikes to 25%).


Note: when I load this site, I do get two alertboxes that say,

         "Redirection limit for this URL exceeded.
          Unable to load the requested page." 

Once I dismiss those, however, the site looks the same in Mozilla
as it does in IE6. I am only missing the ads that you see in IE6
in the left and right margins (that follow you when you scroll).

I assume these are missing from Mozilla because of coding errors
at the site. These errors are visible in Tools > Web Developement > 
JavaScript Console:

       Error: telelayer is not defined
       Source File: http://www.sina.com.cn/
       Line: 806

                   etc. 
                   etc.


Gary: I see your build is from 2002-08-26. Could you try the same
Mozilla trunk build (from 2002-09-23) as I have and see if you still
see the problem? WARNING: trunk builds dated after 2002-09-23
have new problems, so don't try them!


Also:

1. What are your machine specs?
2. Are you going through a proxy?
3. What version of Flash do you have? (Help > About Plug-ins)
4. Do you get the two alertboxes I mentioned above?
Component: JavaScript Engine → DOM Level 0
Keywords: perf
I don't see JavaScript functions being called at all once the page
has been loaded. To see this:

1. Load the site in Mozilla
2. Open Tools > Web Development > JavaScript Debugger
3. Click on the "Stop" button (red "X" at upper-left)
4. Nothing happens! We don't stop anywhere in JavaScript.

This shows that once the page loads, no JavaScript is being called.
If there were, the debugger would stop execution -


Based on this, reassigning to Browser-General until we can
get further information. As I mention above, builds from today
may not work at all on this site: but I think that's a different
problem affecting many different site. To test performance at 
this site, I am using Mozilla trunk binary 2002092308.

Does this site hog the CPU for anyone else with that build date?
Assignee: rogerl → asa
Component: DOM Level 0 → Browser-General
QA Contact: pschwartau → asa
One other point: I do occasionally see little child windows opening
at the lower-right of my computer screen. These are independent windows;
I can't see what's in them: they disappear almost immediately.

Is anyone else seeing these?
OK, this site uses at least 6 <object> tags for Shockwave Flash.
For example:

<object classid="clsid:D27CDB6E-AE6D-11cf-96B8-444553540000" 
codebase="http://download.macromedia.com/pub/shockwave/cabs/flash/swflash.cab#ve
rsion=5,0,0,0" WIDTH="585" HEIGHT="50">

 <param NAME=movie              
        VALUE="http://ad4.sina.com.cn/button/hanwang/fc585x50-020924.swf">
 <param NAME=quality VALUE=high>

 <embed src="http://ad4.sina.com.cn/button/hanwang/fc585x50-020924.swf" 
quality=high WIDTH="585" HEIGHT="50" TYPE="application/x-shockwave-flash" 
PLUGINSPAGE="http://www.macromedia.com/shockwave/download/index.cgi?P1_Prod_Vers
ion=ShockwaveFlash">
</embed>

</object>


Reassigning to Plug-ins for further analysis. Is this possibly
a dupe of bug 132759, "[HANG]high(100%) cpu usage with Flash: WM_USER+1 events 
given too high priority and starve UI" ? 
Assignee: asa → beppe
Component: Browser-General → Plug-ins
QA Contact: asa → shrir
Well, I now don't see such great performance problem in my mozilla
20020826 again, but my CPU still loads at about 40%-60%. I think it's
still higher than we expected. Your 5%-7% is ideal, and mine performance
well at other main portal sites in China like 163.com and sohu.com.

Actually I met the problem several days ago, and sina.com.cn may already
fixed something so we can't reproduce the browser hanging problem.

I agree with your guessing about the plug-in problem. After I disabled
Flash plug-in (move NPSWF32.dll out of the plugins dir temporarily), the
website performance so good, at 0-5% of CPU usage.

My environment:
- Desktop: DIY
  - Intel BX7+ mainboard
  - Celeron 733
  - 256M memory
  - Windows 2000 Professional SP2
- Laptop: COMPAQ Evo N400c
  (http://www.compaq.com/products/notebooks/n400/n400-subfamily.html)
  - Intel PIII 700 Mobility
  - 128M memory
  - Windows 2000 Professional SP1
- Mozilla build/plug-ins almost the same
  - Mozilla 1.1 (Gecko/20020826)
  - Shockwave Flash plug-in NPSWF32.dll, v6.0r47
  - Adobe Acrobat plug-in nppdf32.dll, v5.00 for Netscape
yes, sina was one of the reported sites for bug 132759 and that has since been
resolved. The fix landed on hte trunk on 20020926, download a version after that
and you will see the difference. marking as dup

*** This bug has been marked as a duplicate of 132759 ***
Status: UNCONFIRMED → RESOLVED
Closed: 22 years ago
Resolution: --- → DUPLICATE
v
Status: RESOLVED → VERIFIED
Product: Core → Core Graveyard
You need to log in before you can comment on or make changes to this bug.

Attachment

General

Creator:
Created:
Updated:
Size: