[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

where is srfi-17 going?



I have gotten no support for my proposal.  While the srfi process
allows for srfis that are controversal and don't have consensus, at
this stage it seems kind of pointless to pursue it.  I have no
particular desire to push for something most people dislike, at least
in this context.  So if anybody out there thinks the generalized set!
is a good idea, now is the time to speak up.

There has been slightly more support for a generalization of set!
under a different name.  That is not what I had in mind, but if
requested, I can modify the proposal to use a different name.
Otherwise, I'd just as soon drop the whole thing.

One reason for using a single name is that I'm interested in
experimenting with alternative syntaxes, including use of infix
operators.  One approach is to use pre-processor.  It would translate
`x := y' to `(set! x y)'.  If we have "generalized set!", the same
translation also can translate `(car x) := y' or `x.car := y' to
`(set! (car x) y)'.  You might argue that using the same operator for
variable assignment as for field assignment is still a bad idea, but I
believe the context of the idea makes a difference.  If a language
uses infix operators you want to pick a small set of general
operators, while when using identifiers you have more luxury to use
many of them.   Also, if you use an operator (i.e. `x[y]') for the
value-extraction operation, it seems strange to use function
suntax for the value-setting operation.

This is not meant to be an argument for using set! for both
purposes, only to provide some background.
-- 
	--Per Bothner
per@xxxxxxxxxxx   http://www.bothner.com/~per/