MIG->Corba

[Mridul Jain]

hi,
As I was preparing to work on the topic - Corba
replacement for MIG ;
Markus and Roland adviced that I post the topic on
bug-hurd@gnu.org so as
to get input from others.I waited for some time to do
some homework and
was discussing with Eric Edie from utah about some
issues and he was of great 
help.Now I think is the time to discuss some of the
issues etc on the list
itself.

I have been able to make/generate some of the Hurd
Server(existing) code IIOP compliant i.e
start with MIG IDL and produce stubs that communicate
via IIOP.
Now the reason why I am thinking along the IIOP as
communication protocol is that It is the Corba
standard.If I want a "TRUE" Corba
implementation in GNUMach/HURD then the protocol
should 
be IIOP compliant.Also if we have strict CORBA
implementation we can write the
HURD servers in other languages too.
There are also a bunch of other reasons which can be
discussed in subsequent mails.
Although I had a bit of success with some Hurd Server
Eric had said that there might be problems 
in some cases as there are many MIG IDL notions that
don't map well 
onto IIOP, e.g., the various flavors of port rights. 
It would take some 
effort to emulate these features while avoiding Mach
IPC.

Could someone give a pointer to "steps to compile a
HURD server"??

There is another approach i.e take the *.IDL Corba
file in case you are making  a new 
Hurd Server and then generate ipc compliant code which
is much more easier.
But the problem is that it is not iiop compliant and
lose the many benefits there.

But we can discuss the pros and cons etc.

Mridul.








__________________________________________________
Do You Yahoo!?
Yahoo! Auctions - Buy the things you want at great prices.
http://auctions.yahoo.com/