[Biocorba-l] ORBit hassle

[Jason Stajich]

I have a weird ORBit bug (running on RH Linux 7.0 with ORBit from an rpm
0.5.7-3). This also seemed to appear on my ( now dead ) FreeBSD laptop but
I don't remember which version I had installed.

You can duplicate the bug if you checkout the bioperl-corba-server cvs
module from cvs.bioperl.org, have 2 windows, run
[window 1]
% perl servers/simpleseq.pl < t/multifa.seq

[window 2]
% perl testclients/getseq.pl

You should get something like this:
...
sequence 1,6 is gctcgg
incomplete message received at testclients/getseq.pl line 27.
Error demarshalling result at testclients/getseq.pl line 27.
incomplete message received at testclients/getseq.pl line 27.
Error demarshalling result at testclients/getseq.pl line 27.
incomplete message received at testclients/getseq.pl line 27.
Error demarshalling result at testclients/getseq.pl line 27.
incomplete message received at testclients/getseq.pl line 27.
Error demarshalling result at testclients/getseq.pl line 27.
sequence 1,11 is gctcggactct
...

This means that when sub_seq is called with the parameters 1,7 -> 1,10
there is a demarshalling problem.  This feel like some sort of word
boundary error but I've not got the real inclination to debug ORBit
internals to track it down.

Maybe this is a CORBA::ORBit bug (running 0.4.3)?

Any hints?

-jason

-- 
Jason Stajich
Duke University
jason@cgt.mc.duke.edu