Closed Bug 47627 Opened 24 years ago Closed 24 years ago

Initial and relaunch times degraded by 10x on Win98

Categories

(Derivatives Graveyard :: Netscape 6, defect, P3)

Product:
Derivatives Graveyard
Component:
Netscape 6
Platform:
x86
Windows 98
Type:
defect

Tracking

(Not tracked)

Status:
VERIFIED MOVED

People

(Reporter: chriss, Assigned: jband_mozilla)

Details

(Keywords: perf) 

I'm running a win98 PII/300Mhz/160MB RAM laptop.
I've installed PR1 and PR2 (Aug 4 M17 Build) in separate directories.

Steps to reproduce 
1. reboot computer
2. launch Netscape 6 PR1 and then quit
3. launch Netscape 6 M17 and then quit
4. relaunch PR1 and measure time for Nav window to appear
5. relaunch M17 and measure time for Nav window to appear

I get 8 sec for PR1 and 80 sec for PR2; therefore start time performance has 
degraded by 10x on relaunch. 

Initial start times are also bad. I will reboot and take another set of 
measurements.
I measured the first time relaunch (after OS reboot). Here are the times I got:

PR1 17 sec
PR2 90 sec

So initial launch takes approx 5x longer and relaunch 10x longer.
Another data point is that I ran the same test on Win2000 desktop with what I 
believe is a slower processor than my laptop and M17 start performance was much 
much faster. So this appears to be a Win98 problem.

Adding [perf] keyword.
Keywords: perf
This is commercial specific...moving to Derivatives Product for a move to 
Bugscape.
Assignee: asa → leger
Component: Browser-General → Netscape 6
Product: Browser → Derivatives
QA Contact: doronr → leger
Version: other → unspecified
reassign this to paw for proper person this should go to.
Assignee: leger → paw
I deleted mozver.dat file and did a reinstall.
The launch time problem went away. Unfortunately I didn't save my copy of 
mozver.dat.
Warren can you take a look at this or assign it to someone who can.  Thanks.

Also changing QA Contact
Assignee: paw → warren
QA Contact: leger → paw
Maybe jband has some cycles to look at this.
Assignee: warren → jband
Maybe Dan can offer a clue on what might be in mozver.dat that could cause such 
a load time difference. Also, doesn't this mozver.dat wrinkle bring into doubt 
the assertion that this was a Win98 specififc problem rather than a problem 
caused by an old mozver.dat on that machine (which happened to be running 
Win98)?
dveditz told me that he can't think of any rationale by which any version of 
mozver.dat could cause such a thing. 

I'll try to reproduce this and see what happens.
Status: NEW → ASSIGNED
Bug moved to http://bugscape.netscape.com/.

If the move succeeded, paw@netscape.com will recieve a mail
containing the number of the new bug in the other database.
If all went well,  please mark this bug verified, and paste
in a link to the new bug. Otherwise, reopen this bug.
Status: ASSIGNED → RESOLVED
Closed: 24 years ago
Resolution: --- → MOVED
The move did not work.  Shiva is check to find out why.  Adding shiva to the cc
list.
This bug is assigned to me. Why are you moving it?
Per Jan leger's comments in the bug.  This bug is commericial only, so it was to
be moved to bugscape.  Jan no longer works at netscape, so I inherited this task
of moving commerical only bugs to bugscape. Your ownership is not affected the
bug just lives in a different bug database.
Asa or Paul, could you verify that this was, indeed, MOVED?  Thanks.
I can't find it in bugscape. I tried to move a while ago but it failed, so I
have left the bug here.

Bug moved to http://bugscape.netscape.com/.

If the move succeeded, asa@mozilla.org will recieve a mail
containing the number of the new bug in the other database.
If all went well,  please mark this bug verified, and paste
in a link to the new bug. Otherwise, reopen this bug.
Closed: 24 years ago → 24 years ago
verified.  This is now bugscape http://bugscape.netscape.com//show_bug.cgi?id=2677
Status: RESOLVED → VERIFIED
You need to log in before you can comment on or make changes to this bug.