pan pan pan

[pvaneynd]

After some consideration I must conclude that the state of the Common Lisp packages in Debian is becoming unreasonable. One of the goals of forming the pkg-common-lisp team was that I would not be a bottleneck, as RL is inflicting more and more damage to my 'Debian playtime'.

Now that Luca left I'm basically the only 'active' (for very small values of active) DD/DM left. (no hard feeling towards anybody, just loads of thanks for the work they did)

I see two alternatives:

• other people get involved, investigating bugs and sending git/darcs/whatever format patches.


• we go low impact and remove common-lisp-controller and all Common Lisp libraries, and I/we only package the lisp implementations (clisp, ecl, sbcl, cmucl and perhaps ccl) without any special changes

I don't expect the first alternative to be realistic, so unless proven wrong I'll RFA/RM all the libraries/clc on the 5th of September.

Comments

[vatine]

Unfortunately, I cannot help out. First and foremost, it look as if a MASSIVE time-eater is on the near-time horizon. Secondly, any Debian package I prepare would still need to be funneled through a DD.

clbuild

[(Anonymous)]

Providing a package for clbuild may help. I use sbcl from Debian and the rest of libraries through clbuild.

Re: clbuild

[pvaneynd]

Hmm that is a good idea. I'll investigate this... Thanks!

[(Anonymous)]

I'm really interested in preventing this: despite many packages being out of date, common-lisp-controller and the apt integration is for me a major reason to use Debian (well actually a Debian derivative) for my Common Lisp coding.

What does this involve? How much time do you spend on it? CAn you give examples of something that needs doing? I might be willing to help out.

Also, you should post this other lispy places, like comp.lang.lisp. and maybe lispforum.com as well. It's already been posted to the lisp subreddit.

Maybe I can help

[octopodial-chrome.com (from livejournal.com)]

I've really appreciated common-lisp-controller on both Fedora and Ubuntu. What exactly do you need to help pick up the slack?

Also, I don't really think that clbuild is the right approach...it downloads new code instead of letting the OS package manager handle it.

Re: Maybe I can help

[pvaneynd]

Well the problem is more or less 'death by a thousand paper-cuts'.

Updating a library package does not take a lot of time, but most don't work well with debian/watch so you have to download the image/pull the newest sources and find out.

Updating an implementation is more time consuming and complex. But there are fewer of them of course.

Reacting and triaging of bugs is a major headache. Not only do people always expect the system to DWIM (for example 508922 which I can expect), but the whole C-oriented way of working in Linux often causes problems that are difficult to fix (for example 402508). Then there are of course all the bugs I caused :-(... (for example: 535305)

Just investigating them takes a lot of time, often there is no good solution and the bug just sits there in the BTS, rotting...

Low Impact

[(Anonymous)]

Frankly, time issues aside, I wish Debian & co. (I use Ubuntu on the Desktop, with Debian running some servers) would go low impact. Common lisp controller is an annoyance and it is pretty much impossible to package most libraries in the common lisp ecosystem. It is too unstable, too readily changing, so if you are making serious use of many libraries you need to use SVN/Mercurial/Darcs/Whatever to get the source and keep it up to date.

Re: Low Impact

[pvaneynd]

Please explain in what way clc is an annoyance?

CLC

[(Anonymous)]

Just my experience:
When I began using lisp on debian (stable/sarge, at that time), I initially tried CLC. But I had to abandon it soon, because the available packages were *horribly* old, esp. SBCL and slime.
As you know, SBCL is released montly, and Slime does not even have a release cycle; unless you have a very recent version, the developers tend to ignore any help request (not that I blame them, that's just how it is).

Also, I am in no position to have an opinion on it, but It seems that the very concept of CLC is not universally appreciated; see certain posts in this thread for example:
http://groups.google.it/group/comp.lang.lisp/browse_thread/thread/d40f9d1c2464f679