[om] A Proposal for extending OpenMath with structure sharing

[Richard Fateman]

Do we guide our design by what the least capable recipient of
any OM message might do?
The appropriate message might be "I'm too stupid to handle
valid OM messages that contain cycles"  or maybe "segmentation 
violation".  But anyone clever enough to detect
the cycle would have a shorter program just by handling it
instead of making an error.
If such a cycle makes no sense, that's not a problem. Lots
of things that program A can utter in OM may be meaningless
to program B.  Mathematica can talk trash to Maple and vice versa.


RJF



sal at dcs.st-and.ac.uk wrote:

> While I agree that there is no foundational 
> problem, it is probably appropriate to note explicitly that these objects are 
> allowed, and that many applications will not be 
> able to handle these objects and, perhaps, to give some guidance on what is 
> the appropriate error for them to raise in such circumstances.
> 
> 	Steve
> 


--
om at openmath.org  -  general discussion on OpenMath
Post public announcements to om-announce at openmath.org
Automatic list maintenance software at majordomo at openmath.org
Mail om-owner at openmath.org for assistance with any problems