On Mon, Jul 22, 2019 at 1:13 AM Per Bothner <xxxxxx@bothner.com> wrote:
 
I agree array-map is highly desirable.  I decided to leave it out because (1) I was getting essentially no feedback or interest
on this srfi;

Well, for what it's worth I do intend to have a vote on R7RS-large arrays, with the choices being SRFI 121, SRFI 164, neither, or abstention (as usual).  I have the sense that the third option may win, because arrays are a specialized facility and most people will feel like they can't choose between one and the other.  But nevertheless they will be on the ballot.
 
FWIW, Racket supports it: https://docs.racket-lang.org/math/array_broadcasting.html
The conservative solution is to not support it.  A possible compromise is to allow
a non-array or a rank-0 array argument, but otherwise require all arguments to have
the same shape.

That's pretty much what APL does.   I would say any of the three possibilities (general broadcasting, broadcast only scalars, and no broadcasting) would be satisfactory.  Indeed, I think this is a case where anything is better than nothing.


John Cowan          http://vrici.lojban.org/~cowan        xxxxxx@ccil.org
Arise, you prisoners of Windows / Arise, you slaves of Redmond, Wash,
The day and hour soon are coming / When all the IT folks say "Gosh!"
It isn't from a clever lawsuit / That Windowsland will finally fall,
But thousands writing open source code / Like mice who nibble through a wall.
        --The Linux-nationale by Greg Baker