Does anyone know the current status of the CLOS metaobject
protocol within the X3J13 committee? We have heard that it
was rejected, but it is included in the CLOS implementations
in Lucid Common Lisp and TI's Explorer Common Lisp.
Mark Adler
AI Applications
DEC
(508) 490-9174
·····@aiag.enet.dec.com
In article <·····@shlump.nac.dec.com> ·····@aiag.enet.dec.com writes:
> Does anyone know the current status of the CLOS metaobject protocol
> within the X3J13 committee? We have heard that it was rejected, but
> it is included in the CLOS implementations in Lucid Common Lisp and
> TI's Explorer Common Lisp.
It was not rejected; the CLOS designers decided that they could not
produce a completely satisfactory spec in time for the X3J13 draft and
withdrew their intention to standardise that part of CLOS at this
time. This was formally announced at the January 1989 meeting in
Kauai.
Since PCL is based on the metaobject protocol (or visa-versa :-)),
most PCL-based CLOS implementations will include at least a varient of
the same protocol. TI's implementation is not PCL-based, but it's
(prime?) author was a member of the CLOS subcommittee.
--
dan
In real life: Dan Pierson, Encore Computer Corporation, Research
UUCP: {talcott,linus,necis,decvax}!encore!pierson
Internet: ·······@encore.com
In article <·····@shlump.nac.dec.com> ·····@aiag.enet.dec.com () writes:
>Does anyone know the current status of the CLOS metaobject
>protocol within the X3J13 committee? We have heard that it
>was rejected, but it is included in the CLOS implementations
>in Lucid Common Lisp and TI's Explorer Common Lisp.
The metaobject protocol couldn't have been "rejected", since it was never
proposed for adoption. The designers still haven't finished designing it.
The only decision regarding the metaobject protocol that X3J13 made was
that we didn't need to hold up the standard for it.
Many CLOS implementations include *a* metaobject protocol. Most are
presumably based on the draft specs that have been published within X3J13
from time to time. If the CLOS developers from those vendors have also
been involved in the metaobject protocol design then they may even be more
up-to-date than the specs. But there will undoubtedly have to be changes
made when the final metaobject protocol is adopted.
--
Barry Margolin, Thinking Machines Corp.
······@think.com
{uunet,harvard}!think!barmar
In article <·····@think.Think.COM> ······@think.com (Barry Margolin) writes:
>Many CLOS implementations include *a* metaobject protocol. Most are
>presumably based on the draft specs that have been published within X3J13
>from time to time. If the CLOS developers from those vendors have also
>been involved in the metaobject protocol design then they may even be more
>up-to-date than the specs. But there will undoubtedly have to be changes
>made when the final metaobject protocol is adopted.
You are supposing that it will be finished and then approved.
I think there's something to be said for implementations that
don't have an elaborate metaobject protocol. I also think
there are other things related to Common Lisp (eg, foreign
function interface) that it is more important to standardize.