ORBit C++ folding in ...

[Michael Meeks <michael ximian com>]

Hi Murray,

On 10 Dec 2001, Murray Cumming wrote:
> I suspect that it won't be that simple to convert C++ code that uses
> STL containers to Glib stuff, and I'm certain that it won't be as
> readable or robust. This is with the usual caveat that I'm not
> actually very familiar with the OrbitC++ code.

        It uses STL somewhat extensively, and AFAICS is an unreadable
nightmare :-) but then - I don't read much (if any) C++ code.

> What are the advantages to having no C++ code in ORBit?

        Well - mostly a maintainance issue really - I want to be able to
alter the IDL compiler innards and have some chance of updating the C++
compiler in a safe and consistant fashion.

        It's not that I want no C++ code in ORBit2 - that's not possible,
we'll need to have a least a few bits I'm sure [ and a conditional build
etc. etc. ]

> I'll do what I can. I was already resigned to having to fix up
> OrbitC++ for ORBIT2 even though I haven't been involved in it so far.

	That's really great. Of course - I don't expect the IDL compiler
to be ported to C instantly - or as a pre-condition to merging; it's just
something that I might do myself if / when bored, and I'd like to agree
that direction first.

	So - it seems Maciej doesn't like our 'pretty big addition' - so
we'll do the work on a branch; the 'cpp' branch - I'll add the files from
ORBit-cpp in sensiblish places, and see what happens there;

	Is that ok Sam ? 'cvs -z3 upd -Pd ORBit2 -r cpp' to get the code.

	I'm optimistic that we can get this working quickly, and given
it's neccessity for complete bindings - push it into Gnome 2.0 at some
stage [ it's purely an API addition ] - and get some more reasoned debate
from other release team members :-)

	Regards,

		Michael.

-- 
 mmeeks@gnu.org  <><, Pseudo Engineer, itinerant idiot