Withdrawal Daphne Preston-Kendal (08 Feb 2025 08:12 UTC)
Re: Withdrawal John Cowan (08 Feb 2025 19:38 UTC)
Re: Withdrawal Daphne Preston-Kendal (09 Feb 2025 10:10 UTC)
Re: Withdrawal Arthur A. Gleckler (08 Feb 2025 23:29 UTC)
Re: Withdrawal Daphne Preston-Kendal (09 Feb 2025 10:33 UTC)
Re: Withdrawal Arthur A. Gleckler (09 Feb 2025 20:53 UTC)
Re: Withdrawal Daphne Preston-Kendal (16 Feb 2025 09:15 UTC)
Re: Withdrawal Arthur A. Gleckler (17 Feb 2025 14:03 UTC)

Re: Withdrawal Daphne Preston-Kendal 16 Feb 2025 09:15 UTC

On 9 Feb 2025, at 21:53, Arthur A. Gleckler <xxxxxx@speechcode.com> wrote:
>
> On Sun, Feb 9, 2025 at 2:33 AM Daphne Preston-Kendal <xxxxxx@nonceword.org> wrote:
>>   Indeed. If you hadn’t heard, I’m working with them on a capability-based sublanguage for untrusted Scheme code (which is why this kind of thinking is turning up more often in my Scheme writing lately, e.g. in SRFI 259 as well).
>
> Cool.  Where can I read more?

There’s a teensy bit of information at <https://spritelyproject.org/#oaken> but there will be more in the next few weeks. I’ll send you a pointer when there’s more to see.

> Using order only will preclude a pattern that I've used many times: I pick the order of arguments to the constructor to follow some clear pattern, but alphabetize the list of field definitions so that it's easy to find what I'm looking for later.  Also, if they have to appear in the same order, what's the point of listing them in the constructor definition?  They can be implicit.

So you want to do something like this?:

(define-record-type Person
  (make-person name age height) person?
  (age person-age)
  (height person-height)
  (name person-name)
)

To be honest, I wasn’t aware that this sort of thing was allowed by R7RS small anyway. I do see the utility, though.

By the way, I just discovered this thread from the SRFI 9 mailing list in which Olin suggested two very sensible modifications to the original syntax of SRFI 9 that would have offered, at least, a significant help in the extensibility issue.
<https://srfi-email.schemers.org/srfi-9/msg/2777121/>

Kelsey said no, more or less explicitly reasoning that future SRFIs would invent incompatible record-type definition syntaxes anyway, if they wanted to extend the functionality!
<https://srfi-email.schemers.org/srfi-9/msg/2777161/>

With the benefit of hindsight, Olin was obviously right. SRFI 9 is one of the more brittle syntactic designs in the history of Scheme, and people obviously do want to extend it compatibly. (Not least because it got the blessing of R7RS small, but SRFI 99 predates that.) But it’s easy to say such things with hindsight.

Daphne