Re: PowerLoom's BNF usage

At 12:16 -0500 7/9/08, Robert Goldman wrote:
> Does the following correctly summarize what would happen if
> CLIF-compliance is achieved:
>
> 1.  One would have to provide translations back and forth
>     for one's dialect;

I'm not sure of the politics.  For example, it might be enough
to prove equivalence or have a demonstrable translator without
making it generally available.  Of course, I wouldn't recommend
such an approach...


> 2.  Because of 1, if one is compliant, one is automagically able to
>     translate one's code into the other dialects?

My understanding is that the dialects are all supposed to be
equivalent and mechanically translatable to each other.  So, I would
expect the translations to be stackable, etc.  I would also expect
a translation from A->B->C->B->A to conserve meaning, if not exact
syntax...


> I think part of this question is "what does it mean to provide a
> translation?"  does it mean one has to provide publicly usable
> code, or just an algorithm specification?

Again, I don't know.  Good question for the CL list (:-).


> Also, what if one has a language with, say, some second-order
> (which I am just assuming are not included in Common Logic)?
> Does one define a subset of one's language which permits
> translation back and forth?

Either that or define an extension to Common Logic, knowing that
some of it may not be usable by most recipients.  For example, CL
encodes statements but not queries or lambda expressions.  There
is nothing stopping two systems from extending CL, however, to
encode this sort of information.

-r
-- 
http://www.cfcl.com/rdm            Rich Morin
http://www.cfcl.com/rdm/resume     [email protected]
http://www.cfcl.com/rdm/weblog     +1 650-873-7841

Technical editing and writing, programming, and web development