[lnkForumImage]
TotalShareware - Download Free Software
Confronta i prezzi di migliaia di prodotti.
Asp Forum
 Home | Login | Register | Search 


 

Forums >

comp.lang.lisp

Proposal for Lisp Request For Implementation process

Michael Maul

7/26/2015 4:20:00 AM

Common lisp libraries in general lacks a consistent interface for just about anything. One thing we might borrow from the Scheme community is the SRFI, for lisp the LRFI. (for those not familiar with the SRFI http://srfi.schemers.org/srfi-pr...). A SRFI is specification for a feature or functionality in the spirit of internet RFCs that is reviewed and approved. Of course that would require a organization to manage the LRFI process. If this process is adopted by the community however it could be very powerful. Given specification library implementers could publish libraries that conform to a given LRFI. if the LRFI would work on depends on community adoption. If it did work then it would grow a set of libraries that conform to a defined set of specifications an perhaps certain conventions and code guidelines.. I see value in this, an would support it, an be willing to be part of the process. .

The rapid growth of the libraries on Quicklisp definitely shows that the community want to contribute, if the contributions can be guided by a process then I think library quality will be increased, the number of libraries that provide redundant functionality would be reduced and the community as a whole would benefit.
5 Answers

Matthew Carter

7/26/2015 6:20:00 AM

0

Michael Maul <mike.maul@gmail.com> writes:

> Common lisp libraries in general lacks a consistent interface for just
> about anything. One thing we might borrow from the Scheme community is
> the SRFI, for lisp the LRFI. (for those not familiar with the SRFI
> http://srfi.schemers.org/srfi-pr...). A SRFI is specification
> for a feature or functionality in the spirit of internet RFCs that is
> reviewed and approved. Of course that would require a organization to
> manage the LRFI process. If this process is adopted by the community
> however it could be very powerful. Given specification library
> implementers could publish libraries that conform to a given LRFI. if
> the LRFI would work on depends on community adoption. If it did work
> then it would grow a set of libraries that conform to a defined set of
> specifications an perhaps certain conventions and code guidelines. I
> see value in this, an would support it, an be willing to be part of
> the process. .
>
> The rapid growth of the libraries on Quicklisp definitely shows that
> the community want to contribute, if the contributions can be guided
> by a process then I think library quality will be increased, the
> number of libraries that provide redundant functionality would be
> reduced and the community as a whole would benefit.

I think we need more Quicklisp libraries, not fewer. Having a
requirement like this would (imo) greatly reduce the number of incoming
libraries.

Where does scheme host a Quicklisp equivalent? We're up to around 1300
libraries in Quicklisp now - does scheme have many?

--
Matthew Carter (m@ahungry.com)
http://a...

Marco Antoniotti

7/26/2015 12:06:00 PM

0

On Sunday, July 26, 2015 at 6:19:44 AM UTC+2, Michael Maul wrote:
> Common lisp libraries in general lacks a consistent interface for just about anything. One thing we might borrow from the Scheme community is the SRFI, for lisp the LRFI. (for those not familiar with the SRFI http://srfi.schemers.org/srfi-pr...). A SRFI is specification for a feature or functionality in the spirit of internet RFCs that is reviewed and approved. Of course that would require a organization to manage the LRFI process. If this process is adopted by the community however it could be very powerful. Given specification library implementers could publish libraries that conform to a given LRFI. if the LRFI would work on depends on community adoption. If it did work then it would grow a set of libraries that conform to a defined set of specifications an perhaps certain conventions and code guidelines. I see value in this, an would support it, an be willing to be part of the process. .
>
> The rapid growth of the libraries on Quicklisp definitely shows that the community want to contribute, if the contributions can be guided by a process then I think library quality will be increased, the number of libraries that provide redundant functionality would be reduced and the community as a whole would benefit.

It is already in place: CDR is your friend.

https://common-lisp.net/pr...

Note that the requirements are somewhat looser than SRFI (which will not forbid things like the "Scheme with Python Syntax", or the equivalent spec for a IRSABOL function).

Contribute!

Cheers
--
MA

Kaz Kylheku

7/26/2015 2:08:00 PM

0

On 2015-07-26, Michael Maul <mike.maul@gmail.com> wrote:
> Common lisp libraries in general lacks a consistent interface for just about
> anything. One thing we might borrow from the Scheme community is the SRFI,
> for lisp the LRFI.

This is really only useful for features that require development at the
implementation level.

Something which just portable code that you can load into numerous Lisp
implementations, it doesn't require this cumbersome approach.

For LRFI to be meaningful, you have to rekindle the CL standardization process.
Someone has to manage these LRFI's, and basically the same group of people
decide which make it to the next revision of the language.

Marco Antoniotti

7/26/2015 2:49:00 PM

0

On Sunday, July 26, 2015 at 4:08:37 PM UTC+2, Kaz Kylheku wrote:
> On 2015-07-26, Michael Maul <mike.maul@gmail.com> wrote:
> > Common lisp libraries in general lacks a consistent interface for just about
> > anything. One thing we might borrow from the Scheme community is the SRFI,
> > for lisp the LRFI.
>
> This is really only useful for features that require development at the
> implementation level.
>
> Something which just portable code that you can load into numerous Lisp
> implementations, it doesn't require this cumbersome approach.

Define "portable code" 3:)

> For LRFI to be meaningful, you have to rekindle the CL standardization process.
> Someone has to manage these LRFI's, and basically the same group of people
> decide which make it to the next revision of the language.

The SRFI does not define a process by which something makes into the R^{n+k}RS. But it is a good thing. Inasmuch as the specs are well written and comprehensive. Something that is not easy to do.

Cheers
--
MA

Zach Beane

7/26/2015 6:38:00 PM

0

Matthew Carter <m@ahungry.com> writes:

> Where does scheme host a Quicklisp equivalent? We're up to around 1300
> libraries in Quicklisp now - does scheme have many?

Scheme Now! is kind of like that. http://snow.iro.umo...

I really like that Scheme Now is shortened to "snow", and the repo is
called a "snowfort", and the package files are called "snowballs". Super
cute.

Zach