Re: MIG->Corba

[Brian May <bam@debian.org>]

>>>>> "Mridul" == Mridul Jain <gnuindian@yahoo.com> writes:

    Mridul> hi, As I was preparing to work on the topic - Corba
    Mridul> replacement for MIG ; Markus and Roland adviced that I

I for one, would be really interested in any moves to make Hurd use
CORBA, since CORBA is a well known standard. (sorry, I can't remember
off-hand: what is MIG? Mach IPC something?)

    Mridul> post the topic on bug-hurd@gnu.org so as to get input from

(I am not subscribed to that mailing list though)

    Mridul> I have been able to make/generate some of the Hurd
    Mridul> Server(existing) code IIOP compliant i.e start with MIG
    Mridul> IDL and produce stubs that communicate via IIOP.  Now the
    Mridul> reason why I am thinking along the IIOP as communication

I think (from what I have seen on the linux-kernel development mailing
list) that the power of CORBA is that you can use any communications
protocol, not just IIOP.

So, while IIOP might be the best for inter-computer communication,
something else might be more optimal for interprocess communication.

However, IIOP probably is a good starting point.

    Mridul> protocol is that It is the Corba standard.If I want a
    Mridul> "TRUE" Corba implementation in GNUMach/HURD then the
    Mridul> protocol should be IIOP compliant.Also if we have strict
    Mridul> CORBA implementation we can write the HURD servers in
    Mridul> other languages too.  There are also a bunch of other

Agreed, I think this is an advantage.

Multi-language support might already be possible with MIG(?), but
since CORBA is more standard, it is likely to be available for more
languages.

    Mridul> reasons which can be discussed in subsequent mails.
    Mridul> Although I had a bit of success with some Hurd Server Eric
    Mridul> had said that there might be problems in some cases as
    Mridul> there are many MIG IDL notions that don't map well onto
    Mridul> IIOP, e.g., the various flavors of port rights. It would
    Mridul> take some effort to emulate these features while avoiding
    Mridul> Mach IPC.

I am not really familiar with these issues, but I think it would still
be worth it even if it led to short term breakage and/or
incompatibilities.

If you haven't done so already, I would recommend you look at the
ORBit implementation of CORBA. It aims to be a small, efficient,
implementation that has been ported to the Linux kernel (some people
were pushing for CORBA to be used instead of the conventional /proc/
text mode files).
-- 
Brian May <bam@debian.org>