On Sun, February 22, 2009 9:32 pm, Paul Libbrecht wrote:
> Abbreviated symbols are definitely a matter of input and output and
> not matter of semantic to my taste. However the need to declare a new
> symbol often has it's own reasons... merely stating that it should be
> handled by something else is not really a good answer.
The way I see it that there would be no real need for miles_per_hour if it
didn't have a "non-standard" abbreviation, so whatever abbreviation one
had (outside semantics) would still need somesemantics to connect to.
> There's a need to declare a symbol as "the same as, if you want to
> ignore this part of the world".
> The reason I suggested OWL at the time was that it does have abilities
> to declare that two classes are to be considered the specialization or
> the  same thing provided you can see such a statement.
>
> I absolutely know that FMPs can declare that one construct is equal to
> another but the geometry of facts is different. I think the right
> question that one should be able to answer is "what to do if I meet
> symbol bla-bla and can't do anything with it".
>
> - an FMP that states that bla-bla is bla would be wrong, it would mean
> that bla-bla has no finer semantic
> - an FMP that states that bla-bla is a specialization of bla would
> allow the processor to ignore bla-bla and replace all occurrences of
> OMS-bla-bla to OMS-bla
>
> Probably some other discovery effects should be analyzed.
What we really need is a taxonomy of FMPs:-)

More seriously, we could discuss this is Grand Bend. How do you see
LandauIn fitting in here?

James Davenport
Visiting Full Professor, University of Waterloo
Otherwise:
Hebron & Medlock Professor of Information Technology and
Chairman, Powerful Computing WP, University of Bath
OpenMath Content Dictionary Editor
IMU Committee on Electronic Information and Communication

_______________________________________________
Om mailing list
[email protected]
http://openmath.org/mailman/listinfo/om

Reply via email to