Am Mo., 17. Aug. 2020 um 18:10 Uhr schrieb John Cowan <xxxxxx@ccil.org>:

[snipped]

> So if I _must_ make an executive decision (and I hope not to have to) it will be to do plists.  If someone will hack out the implementation of <https://github.com/johnwcowan/r7rs-work/blob/master/PropListAPI.md> we can get it in the SRFI pipeline.

I've done so, I meaning having hacked out an implementation to
contribute to a plist SRFI. (This does not necessarily mean that I
agree to the all [snipped] arguments from Reddit.)

Before a first SRFI draft of your API is written, I have one question:
The procedure you propose mutate the argument (at least in the
"normal" case) so that the same plist can be returned. However, aren't
plists often entered as literal lists, in which case such a mutation
is not allowed?

Besides that, a typical idiom using the API will be (set! plist
(plist-frobnicate! plist ...)). I am wondering whether we should and
can simplify this.

Marc