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

List:       kde-multimedia
Subject:    MCOP
From:       Stefan Westerfeld <stefan () space ! twc ! de>
Date:       1999-11-07 23:27:48
[Download RAW message or body]

   Hi!

Replying to myself... ;)

> > Btw, I would like to note something. 
> > IMHO, we are diverging from the rest of KDE too much.
> > It seems that finally aRts will be the _only_ app in
> > KDE 2.0 that will use mico.
> 
> > Would it be possible to do as the rest of KDE ?
> > I mean, moving from mico to dcop/kidl and all that ?
> > Even if these new libs takes also 6Mb of memory, at least
> > that will be shared among every app.
> 
> The clean solution, as I see it, would be rewriting aRts to rely only on
> the Unix API and the C++ standard (that is STL).
> 
> That way, it would be portable enough to become a standard like esd.
> 
> The TODO point "transfer layer needs to be rewritten" could translate to
> "aRts should come with an own IPC mechanism, which is commited to deliver
> high performance, multimedia transfer, QoS, etc.". That would mean not
> using DCOP, but - to have a name - ACOP, and try to build highspeed
> multimedia IPC suitable for streaming videos around, sending midi events
> in time, etc.

Well I just checked in MCOP, which should be just what we talked about.
DCOP is really, really way too weak for aRts, (no typing, no interfaces,
no inheritance, streaming only with Qt marshalling, no possibility to
link against objects as library instead of communicating remotely).

And I think people wouldn't even want a fixed DCOP, because it would
probably be larger.

Speed: MCOP unfortunately only gets about 4000 synchronous invocations
per second (which is on the other hand about 2-3 times as fast as mico).

The problem seems to be TCP. If you profile 100000 invocations, you get
reported about 7 secs of CPU usage by gprof, and it takes 24 sec. I
assume that the rest of the performance gets lost somewhere deep down
in linux. (TCP packet stuff etc.)

So probably it would be good to write some shared mem based low level
communication for mcop, to achieve the full performance that would be
possible.

If anybody wants to do that (or something else) ... all help is welcome ;)

   Cu... Stefan
-- 
  -* Stefan Westerfeld, stefan@space.twc.de (PGP!), Hamburg/Germany
     KDE Developer, project infos at http://space.twc.de/~stefan/kde *-

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