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

List:       kde-core-devel
Subject:    Re: Fixing kio_http
From:       Alexander Neundorf <alexander.neundorf () rz ! tu-ilmenau ! de>
Date:       2001-06-29 13:20:52
[Download RAW message or body]

On Friday 29 June 2001 14:34, Dawit Alemayehu wrote:
> On Thursday 28 June 2001 22:26, Rik Hemsley wrote:
> > #if Dawit Alemayehu
> >
> > > BTW, shouldn't that be Fixing the scheduler.  Poor kio_http gets it
> > > from everyone even though it did not do anything wrong :-))))))
> > >
> > > kio_http now feels that it has to unionize to protect its interests
> > >
> > > :-))))
> >
> > Hehe, well, the limit should really be determined by the slaves...
> > Perhaps there could be a .desktop entry saying 'MaxConcurrentConnections'
> > or something ?
>
> Uhh... this is not possible because of the design of KIO.  Each slave has
> no idea about any other that is running.  This is the reason somethings
> like the cookiejar are independent processes that run in their own space. 
> So no, this is the responsibility of either the application or in our case
> the scheduler.  

Well, doesn't KProtocolManager (or how the name was) read the .desktop files 
of all ioslaves ?
Wouldn't it be possible that they could contain the mentioned field, and the 
scheduler could get this information from the KProtocolManager ?

Of course, it is already quite long ago that I looked at this stuff, so maybe 
some things have changed in the meantime.

Bye
Alex

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