Re: Another try at subclass

> Michael Kifer wrote:
> > 
> > What is BLD++? Your proposal made no sense to me. You said that BLD should
> > be BLD without classification and BLD++ should include it. In the above,
> > however, you are saying that BLD should include everything that has already
> > been specified. So, what do you mean by BLD++ now?
> 
> When I wrote "keep the BLD document complete with all that is already 
> specified", I meant, "keep the document complete...", not BLD the 
> dialect; sorry for the confusion.
> 
> The idea is to keep everything in a single document, and then define 
> Core and one or more flavour of "BLD" as subsets of what is specified in 
> the document (e.g. a dialect that contains no named argument Uniterms 
> nor classification and that we could call BLD; another that contains 
> everything specified in the document and that we could call BLD++)..

What would be the justification for this verbal acrobatics? Why not, for
example, BLD-- and BLD? :-)


> My understanding was that this would satisfying your sense of a profile, 
> without the indirection problem that could complicate extensions (and it 
> also seem that we have very different things in mind when we say 
> extensions, but that is another subject).

The BLD-- and BLD would satisfy my sense of a profile, etc., much more. :-)



	--michael  


> Christian
> 
> 

Received on Tuesday, 18 December 2007 20:30:07 UTC