Closed Bug 57947 Opened 24 years ago Closed 23 years ago

Mozilla loads page 2x slower than nn, ie

Categories

(SeaMonkey :: General, defect, P2)

x86
All

Tracking

(Not tracked)

RESOLVED WORKSFORME

People

(Reporter: ezh, Assigned: cathleennscp)

References

()

Details

(Keywords: perf)

Load the URL in mozilla. There's 2 files script and noscript. Load the script file. Look how many time it takes until the pic is displayed and the . Now do it in NN. It's about 2* quicker. Now load the noscript file. Mozilla is faster or about the same with NN. build 2000102504-mtunk and NN4.76 on 3Mbit cable modem (the server is on my cable provider if this matters)
over to javascript engine. Updating summary slightly from "Mozilla loads page 2x than nn, ie" to "Mozilla loads page 2x slower than nn, ie". Eugene, could you reduce this testcase to determine what part of hte script is slowing us down?
Assignee: asa → rogerl
Component: Browser-General → Javascript Engine
QA Contact: doronr → pschwartau
Summary: Mozilla loads page 2x than nn, ie → Mozilla loads page 2x slower than nn, ie
I'm not a programmer... :( I exclude one and then another and put them there. Links follows.
Keywords: perf
Could this be related to Bug 54743?
http://www.menelon.ee/netscape/perf/script1.html I removed this part and the speed had increased. <script language="javascript"> u="u040.82.spylog.com";d=document;nv=navigator;na=nv.appName;t="";p=1; sz=" width=1 height=1 "; hl=history.length;d.cookie="b=b";c=0; bv=Math.round(parseFloat(nv.appVersion)*100); if (d.cookie) c=1;n=(na.substring(0,2)=="Mi")?0:1; if((n==0)||(bv >= 300)){rn=Math.random();t=(new Date()).getTimezoneOffset();} else {rn=0;} z="p="+p+"&rn="+rn+"&t="+t+"&c="+c+"&hl="+hl+"&tl=0&ls=0&ln=0"; if (self != top) { fr=1;} else { fr=0;} r1=escape(d.referrer);r=escape(parent.document.referrer); sl="1.0";h=0; </script>
No, I'm not sure about my previous comment... But I do belive it is the dup of bug 54743 - when I removed the whole end script section it was defenitly faster... If so, then 90% of .ru sites are affected, coz they use this counter... www.spylog.ru - very powerful.
No, it doesnt' seem to be related to the JavaScript block at all. Method: I got "script.html" from the URL, saved it locally, and put in timing code: At the top of <SCRIPT> block: start = new Date(); At the bottom of <SCRIPT> block: end = new Date(); time = end.getTime() - start.getTime(); alert(time); I then closed all instances of NN4.7 and Mozilla (MN6 branch binary 2000102609) I deleted my disk cache, opened the browser, and loaded this local file. The results were identical on WinNT for both browsers: 15 milliseconds. I repeated this experiment many times to be sure (deleting disk cache each time) So the script is executing in exactly the same time. Note however, there are other differences between the "noscript.html" and "script.html" files at the given URL. For example, the latter contains these lines not found in the former: <BASE HREF="http://www.menelon.ee/netscape/perf/"> <? require("../phpAds/config.inc.php3"); require("../phpAds/view.inc.php3"); require("../phpAds/acl.inc.php3"); ?> Could this have something to do with it? One final note: I also deleted the <SCRIPT> block from "script.html" and ran a comparison between NN4.7 and Mozilla. Mozilla loaded it much slower than NN4.7. Reassigning to Ed for help on where this bug should go. It's definitely a performance issue, but not involving JS Engine -
Assignee: rogerl → kandrot
Component: Javascript Engine → Browser-General
OS: other → All
Setting default QA contact -
QA Contact: pschwartau → doronr
Look at this http://www.mdmbank.com/default.htm Mozilla load this page for 10.8 sec for me, but NN 4.76 does it with 5. Pics are loaded very lazily... We have here again the www.spylog.ru counter + mouseover JS. I try to test some mare pages with www.spylog.ru counters.
http://www.anet.ee/ - contains also the www.spylog.ru counter and does not even load all pics.
Severity: normal → major
Priority: P3 → P2
Also look at bug 59552. They are probably related.
I will research this problem, then determine if I can fix it, or found out who should.
Status: NEW → ASSIGNED
Target Milestone: --- → mozilla1.0
reassigning kandrot perf bugs.
Assignee: kandrot → cathleen
Status: ASSIGNED → NEW
Target Milestone: mozilla1.0 → ---
Just tried on my win2k box, and I do not see a slowness of loading those pages, comparing to nn4.75 and ie5. Marking this bug as WORKSFORME. Please reopen if problem still exists.
Status: NEW → RESOLVED
Closed: 23 years ago
Resolution: --- → WORKSFORME
Product: Browser → Seamonkey
You need to log in before you can comment on or make changes to this bug.