KDE4's IPC

[Stefan Teleman]

On Friday 23 December 2005 10:30, Thiago Macieira wrote:
> Hello All
>
> I'd like to get the discussion on KDE4's IPC started. I'll try to separate
> my personal opinion and preference from consensus and facts (I'll post my
> opinion in a reply shortly).
>
> According to our Goals page (http://wiki.kde.org/KDE+4+Goals), our options
> for the IPC system are:
> - DBUS
> - DCE
> - DCOP

Hi.

My 0.02 on this, for what they are worth.

i have worked quite a bit with DCE in the mid 90's. DCE is a very stable, well 
documented and well understood framework. it comes with de facto integration 
with Kerberos and AFS, which is relevant to the infrastructure of large 
commercial organizations. on top of all this, it is supported by Microsoft. 
there are also CORBA bindings to DCE, which is also relevant to commercial 
organizations, which may have chosen to use CORBA instead of DCE. it is also 
fast, natively multi-threaded, and low overhead, and being written in C, it 
is compiler-friendly/neutral.

the only drawback i see is that, from a perception POV, it is no longer in the 
hype main focus. not that this is a technical drawback, but this may be open 
to (technically unfounded) criticisms.

my (very humble) suggestion would be to seriously consider it. please do not 
think even for a second that i am making this suggestion because i consider 
the other two alternatives somehow technically inferior. i am only being 
pragmatic here: a KDE IPC mechanism based on DCE would automatically make KDE 
compatible with large enterprise/corporations infrastructures.

Happy Holidays to Everyone. :-)

--Stefan

-- 
Stefan Teleman          'Nobody Expects the Spanish Inquisition'
steleman at nyc.rr.com                          -Monty Python