Re: [certi-dev] Port numbers

[Martin Spott]

Hi Eric !

"Eric Noulard" wrote:
> 2008/6/29 Martin Spott <address@hidden>:
> > Hi friendly CERTI developers  ;-)
> >
> > being in the process of preparing for a new release, would you consider
> > publishing common/typical/recommended values for the $CERTI_TCP_PORT
> > and $CERTI_UDP_PORT environment variables ?

> Those env var are used for RTIG listening port.
> However they are NOT mandatory,

Well, then I'd rephrase my question:
Would you then consider putting a statement onto the "Environment"
pages which makes clear that a default is already implemented and usage
of these variables it optional ?

> > The IANA (http://www.iana.org/assignments/port-numbers) doesn't list
> > such thing at all,
> 
> Yes true, CERTI is not [yet] widespread enough to deserve a IANA
> registration :=)
> May be the time will come....

As far as I can tell CERTI is not the sole project that aims at
implementing an OpenSource HLA RTI. To my knowledge there are the
Portico project (Java, CDDL-licensed) at:

  http://www.porticoproject.org/

and the OpenHLA (Java) at:

  http://sourceforge.net/projects/ohla/

....  apparently OpenHLA does not have a large community but at least
the -commit mailing list has signs of ongoing development.

Still CERTI seems to be the sole OpenSource HLA RTI wich is written in
a language to be compiled into 'real' binaries. Therefore I think it
might be adequate to have a try at negotiating these port numbers with
IANA as "Private Ports". This doesn't make these ports "registered"
ones, but at least it carries the use of these ports into the public.
You still could aim at negotiating registered numbers, but you might
have to choose different numbers then (1024 - 49151).

Thanks for reponding,
        Martin.
-- 
 Unix _IS_ user friendly - it's just selective about who its friends are !
--------------------------------------------------------------------------