[MOBY-l] New GO service at BDGP.

Mark Wilkinson markw at illuminae.com
Mon Aug 12 20:36:25 UTC 2002


Paul Gordon wrote:

>As one of the points of MOBY is to create a chain of evidence so to
>speak, I guess I was thinking it would be nice to be able to distinguish
>between a successful transaction and a non-successful transaction
>
I wouldn't say that it is "a point" of the MOBY project - leaving a 
trail of bubble-gum wrappers is much more of a Client-side issue.  MOBY 
is a simple solution to a simple problem, and we have to be careful what 
additional things we hang on it to prevent it from morphing into a 
monstrosity.  KISS  :-)

Nevertheless, you may be right that this is something worth dealing with 
explicitly.  Perhaps in order to keep the "flow" of object-in/object-out 
that exists currently it would be reasonable to return to you the object 
that you sent in the event of a successful transaction that has no 
inherent return value?  I'd prefer to avoid creating ad hoc message 
objects (since presumably these can not be used as input to services)... 
or perhaps a single "OK" message object as a partner to an error message 
object would suffice?  Or a generic message object with arbitrary 
service-defined content that is meant to be human-readable?

Hmmm.... I'm on holiday, so I'm not thinking too deeply about it right 
now  ;-)

Cheers Paul!

M




More information about the moby-l mailing list