Re: bld vs. rif

> 
> kifer@cs.sunysb.edu (Michael Kifer) writes:
> > 
> > > 
> > > 
> > > It seems to me that many of the things we could disagree about in the
> > > short term are difficult to settle because they are based on ideas about
> > > what "RIF" will be, rather than just what "BLD" will be.
> > > 
> > > I think we might be able to get consensus on a lot of issues with a
> > > certain caveat, however.  Something like this:
> > > 
> > >      The design of BLD expressed in this document is (except where noted
> > >      in the document) deemed by the Working Group to be stable and is
> > >      not likely to be changed without new information.  The group has,
> > >      however, not yet designed a way for dialects to fit together to
> > >      form a coherent greater RIF.  It is fairly likely that as it does
> > >      so, the Working Group will discover new information which will
> > >      cause changes in BLD.
> > 
> > Right. This is why we need another dialect, like production rules, to get going.
> 
> Yeah, the question is how to make decisions between now and then...
> 
> So are you comfortable with this approach and this text?  Making
> decisions at F2F7 and for the next BLD draft without trying too hard to
> think into the future, because we're allowed to backtrack if necessary?
> This isn't actually a change in normal WG policy (where Resolutions can
> be revisited if there is new information), but it seems helpful to make
> clear our expectation here.

I am fine with this text. (As an aside, making a cut and ruling out
backtracking at this stage would be unwise, I think.)

However, if BLD becomes a recommendation then backtracking would be hard.
So, I assume you mean backtracking while we are still working on the spec.
Right?


	--michael  



> > > [ In prolog terms, I'm saying we should be clear that we're not doing a
> > > cut after the design of BLD; we may need to backtrack and come up with a
> > > new BLD in order to find a suitable all-of-RIF.  :-)   We want to print
> > > out the current BLD solution, though.... ]
> > > 
> > >     -- Sandro
> > > 
> > > 
> > 
> > 
>

Received on Tuesday, 25 September 2007 19:59:52 UTC