[Biocorba-l] location of biocorba.idl

[Brad Chapman]

Hi Jason;

> The first iterations of bioperl-corba server and client code included the
> idl 'biocorba.idl'.  Do we want continue to distribute the idl with the
> implementation distributions?  

I would like to do this. It does introduce the chance for
incompatibilties, but it also makes things a lot easier for the end
user who would have to make a second download to get the IDL and put
it in the right place. This actually seems more important for the
python implementation then the perl implementation, since in python we 
do generate stub and skeleton classes, so the user never even needs to 
worry about the IDL file at all, if they don't want to.

> I can see arguments for both, but would
> like to at least have a central place where the version is controlled
> which is what I make biocorba-idls.

I agree with this completely. Do you think we could add a constant
value, IDL_VERSION or something, to the Gnome::Unknown class that
indicates the IDL version that an object comes from. This way, if
there is a problem with interoperability, a client could always check
the IDL version of the server object she has, and see if this is the
problem. 

This is kind of an informal form of versioning, but I don't know if
there is really a good standard versioning technique that is
guaranteed to work with all of the different ORB
implementations. Anyone else have good ideas?

Brad