>
> Apologies for not responding at once. I tried to write down an obvious
> (seemingly) example of why 251 would be better than 245, but when I
> invested some more time into it, it became less obvious.

I responded quickly because I was trying every single day for a decade
at $WORK to avoid the kind of code that SRFI-245 is allowing, and not
forbidding. Tho, what should go into the standard is a different story.

Similarly, SRFI is not the R7RS bis that is worked together. It is the
good, and sometime left to the appreciation of the editor rules of SRFI
process that makes it competing SRFI a possibility. Such as the famous
SRFI-200 vs. SRFI-204 that was quite a marvel of a convo, and specification.

Note: I read the 'finite game, and infinite games'.

> In fact, while the benefit of running some type checks and debugging
> printing at the beginning of the function is unquestionable, I am not
> even sure that letrec* is the correct model for the following
> definitions. Maybe lectrec (without a *) would be a better choice.
>
> I need to think more about it, I am sorry.
>
> In any case, I certainly don't like something like this:
>
> (let ((x 0)
> (define (set-x!) (set! x 1))
> (long-long-long function)
> (long-long-long function)
> (long-long-long function)
> (long-long-long function)
> (long-long-long function)
> (long-long-long function)
> (long-long-long function)
> (long-long-long function)
> (long-long-long function)
> (long-long-long function)
> (long-long-long function)
> (long-long-long function)
> (long-long-long function)
> (long-long-long function)
> (long-long-long function)
> (long-long-long function)
> (define x 1))
>
> where set-x! would be mutating the bottom x if the bottom define is
> present, or the top one if it is not.

That is more or less what a PhD told me happens when teaching programming
to non programmers, also happens in notebooks all the time [tinker FTW].

> Yeah, don't write code this badly, but "don't do something" is a bad
> advice in programming.

My point remain zigzag reading that is required to understand the above
snippet is very painful. Nobody shall do that, except newbies that do not
know better, and casuals that want to get things done. That is an opportunity to
paraphrase arcfide along the line 'code by newbies, does not look like code by
more experienced programmers, and, should not be the same.'

I think quoting macro cons artists is "deceptive" [for lake of better word].
It is like a child saying to a parent: there are stains on the stains should,
I still wash it? The macro systems are what they are until a proof of work of
vau, but the chairs are busy doing the common singleton union of something is
already united. /r8rs pro out

The work of R6RS contributors, and aficionados, is NOT below awesome. The work
of R7RS bis is interesting, but like I said somewhere else, that is not what
I am aiming, with my scheming, at this time. AND I hope it does not entitle me to
stay silent on the mailing list!

Note: interleaving sounds, and etymology, grounds the concepts much more poetically
than intermingling, based on my Western Europe culture.

At the end of the day, literally, I think SRFI-245 is better, because it gives more
freedom of expression, and more freedom to make, all-in-all, harmless mistakes, that
can be fixed with an automated review too. The large bonus is: it makes the implementation easier.

Regarding the better text, it is somewhat nostalgia, and educational. SRFI-245 mention letrec*
and it is less obvious that it is too... a call to investigate.

Note: We can keep at bay, one or more of: a) Scheme is failure, b) failing, and c) not meant
to succeed. It will be a more please experience. Along, obviously clear political attacks, and
that induce sadness, misunderstand, and chaos.