Brian T Rice
Fri Dec 13 13:14:01 2002
So I presume there are no objections or strong opinions either way...
On Mon, 9 Dec 2002, Brian T Rice wrote:
> I have made some content changes at http://tunes.org/new/HLL/ .
So here's the $10 million dollar question: does anyone here think that
that description of the HLL is concrete enough? Am I hand-waving too much
in some place that you're interested in?
The meta-object library required to demonstrate all the concepts needs to
be specified, and I will formulate something for this soon. The
object-attribute associations have a known generalization framework
concept, but I've never seen a language really employ that concept except
in a limited scale in AP5 (http://www.ap5.com). If there are better
examples, please reply with them.
The other meta-level frameworks are not so clearly worked out in research,
and at some point there will need to be some explicit influence of the
rewrite and Maude ideas on all of this.
I will make this disclaimer up front: Slate as is can't easily support all
of the MOP-based features that TUNES specifies, but a lot of them can be
made. I also don't see Slate's syntax as particularly preferrable in a
general sense, but then our intention is to build interface tools that
abstract over syntax and allow building other syntaxes. Slate's keyword
syntax ultimately has a strong effect on naming of functions, so this is
one upper bound.
On a final note, I have started uploading some improvements to the
Interfaces subproject area, so people with a bone to pick should let me
know. Speaking of disputes, I need someone to take a survey of the mailing
list archives and try to collect important ideas that never made it to
Tunes' front page. Do Fare, Tril, Jecel, Billy, or anyone else have some
recollections about where to look?
Brian T. Rice
LOGOS Research and Development