An analysis about a generic desktop application configuration management system

[Philip Van Hoof]

On Mon, 2005-04-11 at 23:23 +0100, Richard Moore wrote:
 
> > So it seems likely that D-BUS will be acceptable to everyone, while there
> > are a few reasons why DCOP might not be.
> 
> IIRC DBUS currently lacks some of the features of DCOP. That said, maybe
> it will.


Which ones? The one great feature that D-Conf will need of DCOP
(signals) is also available in D-BUS.

I know about transactions (DCOP has DCOPClientTransaction) and an IDL
compiler (DCOP has dcopidl).

The IDL compiler would be useful for D-Conf. There's a few people
working on an IDL compiler for D-BUS. I'm confident that by the time
development of D-Conf starts, this IDL-compiler will be ready.

However, I fail to see how the transactions are a requirement.

So, D-BUS fulfills the requirements. 


And given the fact that D-Conf shouldn't be an X11-only technology, and
DCOP seems to depend on many X11-related libs, D-BUS is a better
candidate. (but DO correct me about the dependencies of DCOP if I'm
incorrect here. I don't know whether DCOP can be compiled different than
on my standard Ubuntu Hoary).

freax at lort:~/cvs/gnome/evolution $ ldd /usr/lib/libDCOP.so.4
                libqt-mt.so.3 => /usr/lib/libqt-mt.so.3 (0xb78fc000)
        libpng12.so.0 => /usr/lib/libpng12.so.0 (0xb78d8000)
        libz.so.1 => /usr/lib/libz.so.1 (0xb78c6000)
        libXext.so.6 => /usr/X11R6/lib/libXext.so.6 (0xb78b9000)
        libX11.so.6 => /usr/X11R6/lib/libX11.so.6 (0xb77f4000)
        libSM.so.6 => /usr/X11R6/lib/libSM.so.6 (0xb77eb000)
        libICE.so.6 => /usr/X11R6/lib/libICE.so.6 (0xb77d3000)
        libpthread.so.0 => /lib/tls/i686/cmov/libpthread.so.0
(0xb77c3000)
        libstdc++.so.5 => /usr/lib/libstdc++.so.5 (0xb7708000)
        libm.so.6 => /lib/tls/i686/cmov/libm.so.6 (0xb76e7000)
        libc.so.6 => /lib/tls/i686/cmov/libc.so.6 (0xb75ba000)
        libgcc_s.so.1 => /lib/libgcc_s.so.1 (0xb75af000)
        libfontconfig.so.1 => /usr/lib/libfontconfig.so.1 (0xb7588000)
        libaudio.so.2 => /usr/lib/libaudio.so.2 (0xb7573000)
        libXt.so.6 => /usr/X11R6/lib/libXt.so.6 (0xb7523000)
        libXrender.so.1 => /usr/lib/libXrender.so.1 (0xb751b000)
        libXrandr.so.2 => /usr/X11R6/lib/libXrandr.so.2 (0xb7517000)
        libXcursor.so.1 => /usr/lib/libXcursor.so.1 (0xb750e000)
        libXinerama.so.1 => /usr/X11R6/lib/libXinerama.so.1 (0xb750b000)
        libXft.so.2 => /usr/lib/libXft.so.2 (0xb74f9000)
        libfreetype.so.6 => /usr/lib/libfreetype.so.6 (0xb748b000)
        libdl.so.2 => /lib/tls/i686/cmov/libdl.so.2 (0xb7488000)
        /lib/ld-linux.so.2 => /lib/ld-linux.so.2 (0x80000000)
        libexpat.so.1 => /usr/lib/libexpat.so.1 (0xb7468000)
freax at lort:~/cvs/gnome/evolution $ ldd /usr/lib/libdbus-1.so
                libnsl.so.1 => /lib/tls/i686/cmov/libnsl.so.1
(0xb7f9f000)
        libc.so.6 => /lib/tls/i686/cmov/libc.so.6 (0xb7e72000)
        /lib/ld-linux.so.2 => /lib/ld-linux.so.2 (0x80000000)
freax at lort:~/cvs/gnome/evolution $


If it's a necessity to create an analysis about DCOP, ORBit-2 and D-BUS
so that we know which one is the best choice, then I'll do that. I
will. 

However. Given the fact that D-BUS was heavily influenced by the DCOP
system I'm assuming it will eventually have all the important features
of DCOP. And given the fact that also many GNOME developers like using
D-BUS, we have a perfect candidate for an interprocess communication
system.


But I am open for ideas, suggestions and mindsharing. If DCOP is better,
then please prove it. Don't just tell us: Use DCOP or we wont accept it.
Thats unfair. And why would we, as the opensource / free software
community, play the game unfair with each other? I don't see a reason.



-- 
Philip Van Hoof, Software Developer @ Cronos
home: me at pvanhoof dot be
gnome: pvanhoof at gnome dot org
work: philip dot vanhoof at cronos dot be
junk: philip dot vanhoof at gmail dot com
http://www.pvanhoof.be/