Question: planning support for SRFI-125, SRFI-133?

Question: planning support for SRFI-125, SRFI-133?

[Ramin Honary]

Hello Guile developers:

I have come to realize that there are several finalized SRFIs
published by John Cowan and others, for example, SRFI-133 (vector
library), SRFI-125 (intermediate hash tables), SRFI-126 (R6RS-based
hash tables) that were intended to replace older SRFIs (for example
SRFI-43, and SRFI-69) when R7RS-large is supposed to be ratified.

Looking at the functionality already available in Guile, it seems like
these newer SRFIs could very easily be supported since the
functionality is already fully implemented and could be re-exported as
modules in the (srfi srfi-*) modules. And incidentally, GNU/MIT Scheme
does provide support for SRFI-125 and SRFI-133 already.

So I am wondering why this hasn't been done in Guile yet? Is it a lack
of time and human resources? Or is it somehow against the Guile
philosophy to try to implement these newer SRFIs? Or maybe you want to
wait until R7RS-large is ratified?

If it is a lack of resources, I might like to try implementing these
SRFIs in Guile and submit a patch.

Thanks for your time and hard work implementing Guile!

-- Ramin Honary

Re: Question: planning support for SRFI-125, SRFI-133?

[Taylan Kammer]

On 30.10.2023 11:24, Ramin Honary wrote:
> Hello Guile developers:
> 
> I have come to realize that there are several finalized SRFIs
> published by John Cowan and others, for example, SRFI-133 (vector
> library), SRFI-125 (intermediate hash tables), SRFI-126 (R6RS-based
> hash tables) that were intended to replace older SRFIs (for example
> SRFI-43, and SRFI-69) when R7RS-large is supposed to be ratified.
> 
> Looking at the functionality already available in Guile, it seems like
> these newer SRFIs could very easily be supported since the
> functionality is already fully implemented and could be re-exported as
> modules in the (srfi srfi-*) modules. And incidentally, GNU/MIT Scheme
> does provide support for SRFI-125 and SRFI-133 already.
> 
> So I am wondering why this hasn't been done in Guile yet? Is it a lack
> of time and human resources? Or is it somehow against the Guile
> philosophy to try to implement these newer SRFIs? Or maybe you want to
> wait until R7RS-large is ratified?
> 
> If it is a lack of resources, I might like to try implementing these
> SRFIs in Guile and submit a patch.
> 
> Thanks for your time and hard work implementing Guile!
> 
> -- Ramin Honary
> 

Note that SRFI 126 is probably *not* going to be part of R7RS-large.

(As much as it saddens me, since I'm the author.)

That being said, the reference implementation is written specifically
for Guile, and it offers some things 125 doesn't.  If some people like
126 simply as a powerful hash table library, or if they like the R6RS
hashtables API but wish it had a few more bells and whistles, then I
guess it wouldn't hurt to include SRFI 126 in Guile.  (It strictly
extends the R6RS API.)

https://github.com/scheme-requests-for-implementation/srfi-126

-- 
Taylan

Re: Question: planning support for SRFI-125, SRFI-133?

[Maxim Cournoyer]

Hi,

Ramin Honary <ramin.honary@gmail.com> writes:

> Hello Guile developers:
>
> I have come to realize that there are several finalized SRFIs
> published by John Cowan and others, for example, SRFI-133 (vector
> library), SRFI-125 (intermediate hash tables), SRFI-126 (R6RS-based
> hash tables) that were intended to replace older SRFIs (for example
> SRFI-43, and SRFI-69) when R7RS-large is supposed to be ratified.
>
> Looking at the functionality already available in Guile, it seems like
> these newer SRFIs could very easily be supported since the
> functionality is already fully implemented and could be re-exported as
> modules in the (srfi srfi-*) modules. And incidentally, GNU/MIT Scheme
> does provide support for SRFI-125 and SRFI-133 already.
>
> So I am wondering why this hasn't been done in Guile yet? Is it a lack
> of time and human resources? Or is it somehow against the Guile
> philosophy to try to implement these newer SRFIs? Or maybe you want to
> wait until R7RS-large is ratified?
>
> If it is a lack of resources, I might like to try implementing these
> SRFIs in Guile and submit a patch.
>
> Thanks for your time and hard work implementing Guile!

I think it's just a matter of someone integrating it, making sure the
test suites they provide build fine and documenting these to a minimum
in the manual.  I also need srfi-125 (and srfi-128), which are
requirements for srfi-209, which I'd like to add.  I have no experience
adding SRFIs to Guile, but it looks relatively straightforward, assuming
the existing implementation is compatible with Guile
(past commit logs are probably helpful as well,
e.g. 5f60eb6bb5af418b332101c69c88da639b117ae4, which added srfi-171).

-- 
Thanks,
Maxim