[Biojava-l] Re: biocorba idl

[Matthew Pocock]

Brad,

Brad Chapman wrote:

> </snip>
> Thanks for getting this together! I've been trying to get it working
> so I can work on connecting biopython with biojava and I can get
> sooooo far, but not quite all the way. The server started up with no
> problems, but then when I run the client code I get the following:
>
> Exception in thread "main" org.omg.CORBA.COMM_FAILURE:   minor code:
> 1 completed: Maybe
>         at com.sun.CORBA.iiop.IIOPConnection.purge_calls(Compiled Code)
>         at com.sun.CORBA.iiop.ReaderThread.run(Compiled Code)
>

>From the org.omg.CORBA.COMM_FAILURE docs it says:
1 Unable to connect to the host and port specified in the object reference,
or in the object reference obtained after location/object forward.

My first plan of attack would be to hack the server so that it writes the
IOR to a file, and the client so that it reads the IOR from the file.
Failing that, you may be able to get the name-server thingy going. The IDL
tutorial on the sun web site should tell you all that you need to know.
Here are a couple of embarasing mistakes I made during debugging:

1) Run the server. Kill the server. Run the client.
  Result - lots of exceptions
2) Run the server in bg. Run the client. Paste in the IOR (but part of it
was overwritten by my dumb shell).
  Result - different lot of exceptions about stuff & things
3) Run the server. Run the client. Paste in the IOR from the previous
server
  Result - different lot of exceptions about how an object could not be
found in a table

I hope this helps.

>
> I'm not sure why I'm getting this. It could be a platform specific
> problem--I'm currently running the Linux Blackdown Port RC4 of JDK 1.2
> under FreeBSD (using Linux compatibility libraries). So this is kind
> of hackish, but 1.2 hasn't been ported to FreeBSD yet, so it is my
> only choice. However, I'm not that experienced with Java and thought I
> would run it by everyone to see if anyone had any ideas to help me
> out...

>
>     By the way, I just checked out biojava-live today and noticed
> during compile that another copy of biojava-live is in src/org/acedb.
> Just thought I would point that out so it doesn't cause any more
> trouble.

This didn't check out for me from the read/write CVS repository when I made
a clean check out yesterday evening. I blew this skeleton away last week.
Which CVS server did you connect to? If this directory does appear for you
then it is safe to rm -rf it away.

>
>     Thanks again for all the excellent code!

Cheers.

>
>
> Brad
>

Matthew

>
> _______________________________________________
> Biojava-l mailing list  -  Biojava-l@biojava.org
> http://biojava.org/mailman/listinfo/biojava-l