Hi folks,

I've been watching the operation of the 0.93.2
05/01/04 release, and have noticed that the thruput
and remaining time for both uploads and downloads is
often very inaccurate on my system. It's a
dual-processor Solaris 8 system, running on a pair of
Tualatin chips @ 1GHZ, using the gtk+-2.2.4/glib-2.2.3
libs.

For instance, I have seen an individual upload
indicate 0.0k/sec on it's status line, while the
percentage complete jumps 2-3% compared to the
previous update, indicating almost a 27K/sec rate for
that particular file. This is a intermittant problem,
because the next status update often reflects close to
what I think the upload was actually doing for that
subsequent update interval. The problem then repeats
after one or two further update intervals (i.e., speed
shows 0.0, but percentage complete climbs a
significant amount.) 

I've also noticed that on a multi-source download, the
individual source status line speed and remaining time
for a chunk does not match either the aggregate speed
and remaining time or the speed shown in the left-hand
status panel, when it's the only download active.
Often, the remaining time is hours different from the
total of the individual source times added together!

Am I expecting too much accuracy from the status
displays? Or has the recent code changes affected the
way that the status information is collected or
displayed?

I'd consider this a minor problem, not something that
would be considered high priority, as it does not
interfere with the operation of uploads or downloads. 

But it would be nice to have reasonably accurate
information on transfers. For the download I
mentioned, the aggregate remaining time is around 10
hours and 30 minutes, while the individual values
total out to only about 7 hours. I felt that a
difference of over 30% between the two numbers should
be reported to the list, as it is certainly a
significant difference. 

I know that there is a large area of new code in this
release, and it's possible that some fine-tuning of
the code might be required. Hopefully, reports of
issues like this one will help in that process. 

Regards,
--Carl

ps: glad to assist any interested party to test or
gather information on this problem. Just drop a note
if I can help, or can provide further information. 


__________________________________
Do you Yahoo!?
Yahoo! Hotjobs: Enter the "Signing Bonus" Sweepstakes
http://hotjobs.sweepstakes.yahoo.com/signingbonus


-------------------------------------------------------
This SF.net email is sponsored by: Perforce Software.
Perforce is the Fast Software Configuration Management System offering
advanced branching capabilities and atomic changes on 50+ platforms.
Free Eval! http://www.perforce.com/perforce/loadprog.html
_______________________________________________
Gtk-gnutella-devel mailing list
[EMAIL PROTECTED]
https://lists.sourceforge.net/lists/listinfo/gtk-gnutella-devel

Reply via email to