On 8/17/05, Richard Fateman wrote:
> Yet another version of mset, code is in same place.
i've committed the code to share/contrib/defstruct.lisp (as before).
> diag(m):0 /* set diagonal to 0 */
> row(m,i): constant or
> col(m,i): constant
>
> I have not programmed the case of
> row(m,i): [a,b,c,...]
well, in the interest of getting to a decision on the defstruct stuff,
i would prefer that we limit the scope of modifications to just
what is needed to support defstruct for now.
i have some ideas about how matrix assignments could work,
and with defstruct in place i could try out some variations
and maybe come back with some proposal. but that can wait.
although the $mlistassign stuff is not part of defstruct,
i'd like to suggest we include it now.
so, richard, can we limit consideration to just the
stuff above line 227 ??
(setf (get '$diag 'mset_extension_operator) '$diagassign)
thanks for your work on this topic,
robert dodier