From fahree at gmail.com  Mon Jun 11 23:13:09 2012
From: fahree at gmail.com (=?ISO-8859-1?Q?Far=E9?=)
Date: Mon, 11 Jun 2012 19:13:09 -0400
Subject: [cl-containers-devel] cl-containers
Message-ID: <CAN7nBXfcf1cP_B0BVjN+VLyEG3CNqZjW-F8-=Ew-6Fho5yMqxg@mail.gmail.com>

Dear Gary and cl-containers developers,

Eric O'Connor and I are a bit sick of the situation
regarding data-structure libraries in CL.
There are tens of implementation of any given implemented data-structure,
most of them of less-than-stellar quality,
all the while plenty of useful data-structures are not implemented at all.
And it's a huge pain any time one wants to find a library
(though the pain is somewhat alleviated by quicklisp).

Therefore, Eric and I are considering doing some work of consolidation
around CL datastructure libraries.
Our goal is to have one library that provides all the interfaces
(generic functions, macros) to all the datastructures that anyone may want,
then separate libraries to implement these interfaces.
For compatibility, these could be distributed as part of the same tarball.

Separately (or not), we'd develop some library for pure datastructures
that would try to maintain some kind of
API coherence with the stateful variants.

Your cl-containers looks like
the best-factored of the existing datastructure libraries,
as far as stateful algorithms go. Are you still maintaining it?
Are you accepting patches and/or co-maintainers who would perform some surgery
on the library to split it into interface and implementations,
and add implementations to missing datastructures (such as fibonacci heaps),
and add thread-safe variants of various datastructures, etc.
All the while we'd seek to preserve backwards compatibility.

??? ? Fran?ois-Ren? ?VB Rideau ?Reflection&Cybernethics? http://fare.tunes.org
Cubans bemoan the three failures of the revolution: breakfast, lunch and dinner



From whalliburton at gmail.com  Mon Jun 11 23:42:20 2012
From: whalliburton at gmail.com (William Halliburton)
Date: Mon, 11 Jun 2012 16:42:20 -0700
Subject: [cl-containers-devel] cl-containers
In-Reply-To: <CAN7nBXfcf1cP_B0BVjN+VLyEG3CNqZjW-F8-=Ew-6Fho5yMqxg@mail.gmail.com>
References: <CAN7nBXfcf1cP_B0BVjN+VLyEG3CNqZjW-F8-=Ew-6Fho5yMqxg@mail.gmail.com>
Message-ID: <CAHPzxA0rN6WEjGUUOZnA_9pU65zNi0TXYmE_sz3cb_dJEzgaHg@mail.gmail.com>

Yay for, "The One Container Library to Rule Them All".

I'd be up to helping as needed.

On Mon, Jun 11, 2012 at 4:13 PM, Far? <fahree at gmail.com> wrote:

> Dear Gary and cl-containers developers,
>
> Eric O'Connor and I are a bit sick of the situation
> regarding data-structure libraries in CL.
> There are tens of implementation of any given implemented data-structure,
> most of them of less-than-stellar quality,
> all the while plenty of useful data-structures are not implemented at all.
> And it's a huge pain any time one wants to find a library
> (though the pain is somewhat alleviated by quicklisp).
>
> Therefore, Eric and I are considering doing some work of consolidation
> around CL datastructure libraries.
> Our goal is to have one library that provides all the interfaces
> (generic functions, macros) to all the datastructures that anyone may want,
> then separate libraries to implement these interfaces.
> For compatibility, these could be distributed as part of the same tarball.
>
> Separately (or not), we'd develop some library for pure datastructures
> that would try to maintain some kind of
> API coherence with the stateful variants.
>
> Your cl-containers looks like
> the best-factored of the existing datastructure libraries,
> as far as stateful algorithms go. Are you still maintaining it?
> Are you accepting patches and/or co-maintainers who would perform some
> surgery
> on the library to split it into interface and implementations,
> and add implementations to missing datastructures (such as fibonacci
> heaps),
> and add thread-safe variants of various datastructures, etc.
> All the while we'd seek to preserve backwards compatibility.
>
> ??? ? Fran?ois-Ren? ?VB Rideau ?Reflection&Cybernethics?
> http://fare.tunes.org
> Cubans bemoan the three failures of the revolution: breakfast, lunch and
> dinner
>
> _______________________________________________
> cl-containers-devel mailing list
> cl-containers-devel at common-lisp.net
> http://lists.common-lisp.net/cgi-bin/mailman/listinfo/cl-containers-devel
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://mailman.common-lisp.net/pipermail/cl-containers-devel/attachments/20120611/72b1da6b/attachment.html>

From gwking at metabang.com  Wed Jun 13 03:59:58 2012
From: gwking at metabang.com (Gary King)
Date: Tue, 12 Jun 2012 23:59:58 -0400
Subject: [cl-containers-devel] cl-containers
In-Reply-To: <CAN7nBXfcf1cP_B0BVjN+VLyEG3CNqZjW-F8-=Ew-6Fho5yMqxg@mail.gmail.com>
References: <CAN7nBXfcf1cP_B0BVjN+VLyEG3CNqZjW-F8-=Ew-6Fho5yMqxg@mail.gmail.com>
Message-ID: <99A46D64-D778-433E-81EA-D6E7A36E84C9@metabang.com>

> Dear Gary and cl-containers developers,

-- snip --

> Separately (or not), we'd develop some library for pure datastructures
> that would try to maintain some kind of
> API coherence with the stateful variants.

That's great!

> Your cl-containers looks like
> the best-factored of the existing datastructure libraries,
> as far as stateful algorithms go.

why thank you!

> Are you still maintaining it?

in theory, yes... in reality, not so much. life interferes.

> Are you accepting patches and/or co-maintainers who would perform some surgery

co-maintainers would be wonderful. I'd be happy to hand the whole thing off to people that have time and motivation to push it forward.

> on the library to split it into interface and implementations,
> and add implementations to missing datastructures (such as fibonacci heaps),
> and add thread-safe variants of various datastructures, etc.
> All the while we'd seek to preserve backwards compatibility.


All of that would be great. I don't know the best mechanism(s) but if you want me to give you super-powers on github or whatever, let me know.


--
Gary Warren King, metabang.com 
Cell: (413) 559 8738
Fax: (206) 338-4052
gwkkwg on Skype * garethsan on AIM * gwking on twitter




From fahree at gmail.com  Wed Jun 13 16:16:24 2012
From: fahree at gmail.com (=?ISO-8859-1?Q?Far=E9?=)
Date: Wed, 13 Jun 2012 12:16:24 -0400
Subject: [cl-containers-devel] cl-containers
In-Reply-To: <99A46D64-D778-433E-81EA-D6E7A36E84C9@metabang.com>
References: <CAN7nBXfcf1cP_B0BVjN+VLyEG3CNqZjW-F8-=Ew-6Fho5yMqxg@mail.gmail.com>
	<99A46D64-D778-433E-81EA-D6E7A36E84C9@metabang.com>
Message-ID: <CAN7nBXd-bX7ELAwDAWLQHr1wpuKGsCHNd4pWcz0d2Y-rM-j0Ew@mail.gmail.com>

>> Separately (or not), we'd develop some library for pure datastructures
>> that would try to maintain some kind of
>> API coherence with the stateful variants.
>
> That's great!
>
I moved my pure data-structures out of fare-utils into a new library:
	https://github.com/fare/lisp-interface-library

>> Your cl-containers looks like
>> the best-factored of the existing datastructure libraries,
>> as far as stateful algorithms go.
>
> why thank you!
>
No, thank you!

>> Are you still maintaining it?
>
> in theory, yes... in reality, not so much. life interferes.
>
>> Are you accepting patches and/or co-maintainers who would perform some surgery
>
> co-maintainers would be wonderful. I'd be happy to hand the whole thing off
> to people that have time and motivation to push it forward.
>
Great! Let's do it.

>> on the library to split it into interface and implementations,
>> and add implementations to missing datastructures (such as fibonacci heaps),
>> and add thread-safe variants of various datastructures, etc.
>> All the while we'd seek to preserve backwards compatibility.
>
> All of that would be great.
>
More controversial perhaps:
Eric and I plan to export interfaces and
encourage other data-structure implementers to use them.

Before we do that, though, and if you give us permission,
we are considering maybe renaming a few gf's.
Of course, we'd leave trampoline defun's to preserve the old name;
but other implementers would specialize methods on the newly-named gf's.
We will discuss the names on the list, and
won't push to master without review.
We will probably get inspiration from how things are done
in the rest of the lisp or non-lisp worlds:
scheme, clojure, java, ocaml, etc.

> I don't know the best mechanism(s) but
> if you want me to give you super-powers on github or whatever, let me know.
>
Can you add fare and oconnore as committers on github?

??? ? Fran?ois-Ren? ?VB Rideau ?Reflection&Cybernethics? http://fare.tunes.org
My opinions may have changed, but not the fact that I am right.



From gwking at metabang.com  Fri Jun 15 16:50:13 2012
From: gwking at metabang.com (Gary King)
Date: Fri, 15 Jun 2012 12:50:13 -0400
Subject: [cl-containers-devel] cl-containers
In-Reply-To: <CAN7nBXd-bX7ELAwDAWLQHr1wpuKGsCHNd4pWcz0d2Y-rM-j0Ew@mail.gmail.com>
References: <CAN7nBXfcf1cP_B0BVjN+VLyEG3CNqZjW-F8-=Ew-6Fho5yMqxg@mail.gmail.com>
	<99A46D64-D778-433E-81EA-D6E7A36E84C9@metabang.com>
	<CAN7nBXd-bX7ELAwDAWLQHr1wpuKGsCHNd4pWcz0d2Y-rM-j0Ew@mail.gmail.com>
Message-ID: <3B1D45F9-E7CF-4CEA-B72E-99220D5BC935@metabang.com>

Hey,

> More controversial perhaps:
> Eric and I plan to export interfaces and
> encourage other data-structure implementers to use them.
> 
> Before we do that, though, and if you give us permission,
> we are considering maybe renaming a few gf's.

You have my permission. Making the API easier to extend and use would be great.

> We will probably get inspiration from how things are done
> in the rest of the lisp or non-lisp worlds:
> scheme, clojure, java, ocaml, etc.

excellent.

>> 
> Can you add fare and oconnore as committers on github?


done

Wish I had more time and energy to contribute and am really happy that you two do!
--
Gary Warren King, metabang.com 
Cell: (413) 559 8738
Fax: (206) 338-4052
gwkkwg on Skype * garethsan on AIM * gwking on twitter