[prev in list] [next in list] [prev in thread] [next in thread] 

List:       kde-devel
Subject:    Re: low performance of kio
From:       mosfet <mosfet () mandrakesoft ! com>
Date:       2000-04-21 20:51:06
[Download RAW message or body]

Lotzi Boloni wrote:
> 
> > Just to re-iterate, I was *not* trying to say konqueror *IS* slower than
> > file manager X.  I was trying to say *if* konqueror is slower than most
> > other file managers out there, that's bad.  I haven't done any benchmarks
> > myself, so I purposely did not draw any conclusions.
> 
>   Ok, a quick test shows
> 
>   642MB CD image file, copied from one harddrive to another:
> 
> konqi: 5 min 40 sec
> mc: 5 min 32 sec.
> 
>   So I think that there is nothing to be discussed about the copying
> issue. We all agree that if a KDE program is 10 times slower than another
> program that's not a good thing. It is not the case.
> 

Yep, I think part of the problem is that while it may seem obvious to
some of us that things like copying wouldn't require data to be
transfered between the client and the server, it's not to everyone and
people have been lumping all types of KIO transfer together. In fact kio
file's copy routine seems quite nice :) The few seconds discrepancy here
is probably mostly due to getting the io process going and updating the
UI, and is very much worth it considering it is non-blocking.

This is cool and I can understand how some of the developers get
frustrated when people start knocking that part of the code when the
only real issue is with getting and putting data between the client and
server with local files (which is the only thing I have been talking
about, and the only thing to show bad timings). This is totally
unrelated to copying. Waldo, if you read the thread you would of seen
that was what *I* was talking about, it was timed, and there was some
good productive discussion on it. Check out the archive if you don't
have the full thread. Again, I didn't mean to offend anyone and most of
KIO does rock. 

Anyways, both David and Coolo said they understand the points that were
relevant and have ideas about how to make it better, so it's probably
not useful to go into this further. No sense beating a dead horse, and
the KIO developers can be trusted to do the right thing IMHO - whatever
that may be.

>         Lotzi

-- 
Daniel M. Duley - Unix developer & sys admin.
http://www.mosfet.org - The place for KDE development news.
mosfet@mandrakesoft.com
mosfet@kde.org

[prev in list] [next in list] [prev in thread] [next in thread]