multiple-values and chain-and Dr. Arne Babenhauserheide (24 Aug 2020 11:51 UTC)
Re: multiple-values and chain-and Marc Nieper-Wißkirchen (24 Aug 2020 12:02 UTC)
Re: multiple-values and chain-and Dr. Arne Babenhauserheide (24 Aug 2020 14:03 UTC)
Re: multiple-values and chain-and Marc Nieper-Wißkirchen (24 Aug 2020 14:13 UTC)
Re: multiple-values and chain-and Dr. Arne Babenhauserheide (29 Aug 2020 11:38 UTC)
Re: multiple-values and chain-and Adam Nelson (28 Aug 2020 02:47 UTC)

multiple-values and chain-and Dr. Arne Babenhauserheide 24 Aug 2020 11:51 UTC
Hi,

The SRFI looks pretty neat, and I like the multiple values support a
lot.

Points I’m wondering about:

Can I re-order multiple values?
-------------------------------

I can think of quite a few reasons for using re-ordering. Did I miss
this is the discussion?

chain (floor/ 3 2) (fraction-of-the-rest _ _) ?

Why doesn’t chain-and support multiple values?
----------------------------------------------

It could use the first value as control-value. If the first value is #f,
the chain ends.

(I see the reasoning for chain-when)

Nest seems misplaced
--------------------

While I see the utility of a nest macro, it seems out of place to me for
three reasons:

- it functions quite differently from chain and I can see usecases both
  for chain without nest and for nest without chain.

- it seems strange that nest nests steps in the opposite order of chain.
  I would have expected nest and chain from the same SRFI to also
  operate in the same order.

- it does not support multiple values.

Would it be possible to move nest into a separate SRFI?

If that doesn’t seem viable, could nest and nest-reverse be swapped so
people using this srfi don’t have to keep two different concepts in mind?

Best wishes,
Arne
--
Unpolitisch sein
heißt politisch sein
ohne es zu merken