Am So., 16. Aug. 2020 um 12:24 Uhr schrieb Lassi Kortela <xxxxxx@lassi.io>:

> It's a good question whether `write` should stick to RnRS syntax or use
> extensions from SRFIs. We had the same discussion with SRFI 169
> (underscores in numbers) and decided that `write` shouldn't use the SRFI.

If the implementation is supposed to be an implementation of R7RS, it
*mustn't* use any special syntax not defined in R7RS when writing
objects that have a datum representation in R7RS.

Thus the paragraph cited by Lassi should be removed.

(i have had my share of compatibility problems with implementations
that chose to use custom write syntax, like special character names
not defined in the R7RS.)

The reader and the writer are not symmetric in this regard. A reader
may accept a wider set of syntax, a writer may use a smaller set of
syntax than standardized to remain compatible to R7RS.

Marc