Hello,
I just got a reject from ftpmaster (the person, not that server) for the slime package because some of a stupid packaging error on my part and because the readme says that there is no general license for the package and that you should check all files individualy. When you do that you see that:
doc/texinfo-tabulate.awk present.lisp swank-ecl.lisp swank-source-file-cache.lisp swank-source-path-parser.lisp swank.asd mkdist.sh test-all.sh test.sh
have no license information.
nregex.lisp has a (sort-of) licence without permission to distribute or anything.
Could you clear these problems up?
Groetjes, Peter
On Sun, 7 Aug 2005 15:15:47 +0200, Peter Van Eynde pvaneynd@debian.org said:
Peter> I just got a reject from ftpmaster (the person, not that server) for the slime Peter> package because some of a stupid packaging error on my part and because the Peter> readme says that there is no general license for the package and that you Peter> should check all files individualy. When you do that you see that:
Peter> doc/texinfo-tabulate.awk Peter> present.lisp Peter> swank-ecl.lisp Peter> swank-source-file-cache.lisp Peter> swank-source-path-parser.lisp Peter> swank.asd Peter> mkdist.sh Peter> test-all.sh Peter> test.sh
Peter> have no license information.
Peter> nregex.lisp has a (sort-of) licence without permission to distribute or Peter> anything.
Peter> Could you clear these problems up?
Does it also matter that some files are GPL and others are public domain?
__Martin
Il 2005-08-10, Martin Simmons martin@lispworks.com ha scritto:
On Sun, 7 Aug 2005 15:15:47 +0200, Peter Van Eynde pvaneynd@debian.org said:
Peter> Could you clear these problems up?
Does it also matter that some files are GPL and others are public domain?
That is no problem.
Groetjes, Peter
Il 2005-08-07, Peter Van Eynde pvaneynd@debian.org ha scritto:
Hello,
I just got a reject from ftpmaster (the person, not that server) for the slime package because some of a stupid packaging error on my part and because the readme says that there is no general license for the package and that you should check all files individualy. When you do that you see that:
Updated:
still without license information:
doc/texinfo-tabulate.awk nregex.lisp
nregex.lisp has a (sort-of) licence without permission to distribute or anything.
present.lisp swank-ecl.lisp swank-source-file-cache.lisp swank-source-path-parser.lisp swank.asd mkdist.sh test-all.sh test.sh
Until these files get a license I cannot add slime to the main debian distribution :-(.
Could you clear these problems up?
Groetjes, Peter
Peter Van Eynde pvaneynd@debian.org writes:
still without license information:
I added public domain notices to these:
doc/texinfo-tabulate.awk swank-source-file-cache.lisp swank-source-path-parser.lisp swank.asd mkdist.sh test-all.sh test.sh
present.lisp
please Alanr and Matthias reply and say it's okay to PD this.
swank-ecl.lisp
please Juan let us know your license of choice.
nregex.lisp
nregex.lisp has a (sort-of) licence without permission to distribute or anything.
what to do with nregexp.lisp?
Luke Gorrie luke@synap.se writes:
Peter Van Eynde pvaneynd@debian.org writes:
still without license information:
present.lisp
please Alanr and Matthias reply and say it's okay to PD this.
I place my changes to present.lisp into the public domain.
Ok by me. -Alan
On Aug 29, 2005, at 6:07 PM, Matthias Koeppe wrote:
Luke Gorrie luke@synap.se writes:
Peter Van Eynde pvaneynd@debian.org writes:
still without license information:
present.lisp
please Alanr and Matthias reply and say it's okay to PD this.
I place my changes to present.lisp into the public domain.
-- Matthias Köppe -- http://www.math.uni-magdeburg.de/~mkoeppe _______________________________________________ slime-devel site list slime-devel@common-lisp.net http://common-lisp.net/mailman/listinfo/slime-devel
OK so I added the public domain notice to present.lisp.
That leaves nregexp.lisp and swank-ecl.lisp between us and apt-get glory :-)
On 30 Aug 2005 16:08:16 +0200, Luke Gorrie luke@synap.se wrote:
That leaves nregexp.lisp and swank-ecl.lisp between us and apt-get glory :-)
I seem to remember that someone else for some other project already tried to resolve the license issues with nregex.lisp with no luck but I don't remember the details. (Was it paserve maybe?)
Cheers, Edi.
On 9217 day of my life Edi Weitz wrote:
I seem to remember that someone else for some other project already tried to resolve the license issues with nregex.lisp with no luck but I don't remember the details. (Was it paserve maybe?)
Maxima contains identical nregex.lisp.
Il 2005-08-30, Edi Weitz edi@agharta.de ha scritto:
On 30 Aug 2005 16:08:16 +0200, Luke Gorrie luke@synap.se wrote:
That leaves nregexp.lisp and swank-ecl.lisp between us and apt-get glory :-)
I seem to remember that someone else for some other project already tried to resolve the license issues with nregex.lisp with no luck but I don't remember the details. (Was it paserve maybe?)
portable-allegroserve use ppcre.
Groetjes, Peter
On Wed, 31 Aug 2005 06:26:41 +0200, Peter Van Eynde pvaneynd@debian.org wrote:
portable-allegroserve use ppcre.
But they used nregex once, didn't they? Maybe they switched because of the licensing issues?
Peter Van Eynde pvaneynd@debian.org writes:
portable-allegroserve use ppcre.
The real solution is of course to fix the cl-ppcre bug that it is split into multiple source files and less bundleable. :-)
Luke Gorrie writes:
Peter Van Eynde pvaneynd@debian.org writes:
portable-allegroserve use ppcre.
The real solution is of course to fix the cl-ppcre bug that it is split into multiple source files and less bundleable. :-)
Oh, that one's easy:
for file in packages specials util errors lexer parser regex-class \ convert optimize closures repetition-closures scanner api; \ do echo $file.lisp; \ done | xargs cat > ppcre.lisp
Maybe bundle-op should go into the collection of asdf hacks that I intend to clean up and release someday.
On 31 Aug 2005 10:05:08 +0200, Luke Gorrie luke@synap.se wrote:
The real solution is of course to fix the cl-ppcre bug that it is split into multiple source files and less bundleable. :-)
Hehe... :)
Just in case you weren't just joking: It's of course fine for me if SLIME or any other project uses CL-PPCRE but /please/ don't bundle it. In the last two weeks I got three bug reports from LW-ADD-ONS users. They all reported the same error which was due to a much too old version of CL-PPCRE. And in all three cases ASDF had used the CL-PPCRE that comes bundled with paserve... :(
Cheers, Edi.
On Mon, 2005-08-29 at 22:03 +0200, Luke Gorrie wrote:
swank-ecl.lisp
please Juan let us know your license of choice.
Sorry for the delay. I just give my code in swank-ecl.lisp to the public domain. There are however a couple of lines that were contributed by Johan Bockgard to load swank in interpreted mode, so he has to be asked as well.
Regards,
Juanjo
Hello,
After seeing a lot of progress being made I went and investigated the sources a little more in detail and I've found: (you might want to skip towards to end because there is a slight problem)
Are covered by the GPL: swank-clisp.lisp bridge.el hyperspec.el slime.el swank-openmcl.lisp tree-widget.el
Are in the public domain: mkdist.sh present.lisp swank-abcl.lisp swank-allegro.lisp swank.asd swank-backend.lisp swank-cmucl.lisp swank-gray.lisp swank.lisp swank-lispworks.lisp swank-loader.lisp swank-sbcl.lisp swank-source-file-cache.lisp swank-source-path-parser.lisp test-all.sh test.sh doc/slime.texi doc/texinfo-tabulate.awk metering.lisp nregex.lisp
BSD-like license: swank-corman.lisp (I think will pose no problem)
Are without licenses: swank-ecl.lisp ChangeLog doc/Makefile HACKING NEWS PROBLEMS README
I think placing them into the public domain would also be easiest.
Then came a surprise, and I'm at a loss to explain how I missed it. My apologies for this.
The file xref.lisp is not PD or GPL (unlike my greps showed) but has a rather obnoxious license that says:
... ;;; o No fees or compensation are charged for use, copies, or ;;; access to this software. You may charge a nominal ;;; distribution fee for the physical act of transferring a ;;; copy, but you may not charge for the program itself.
This means you cannot put it into a distribution, and makes it DFSG unfree (even with the later clauses taken into consideration). To add insult to injury it has:
;;; o Any work distributed or published that in whole or in part ;;; contains or is a derivative of this software or any part ;;; thereof is subject to the terms of this agreement. The ;;; aggregation of another unrelated program with this software ;;; or its derivative on a volume of storage or distribution ;;; medium does not bring the other program under the scope ;;; of these terms.
As this file is part of slime, all of slime is under this agreement. Talk about a viral license ;-).
But then we have:
;;; o Permission is granted to manufacturers and distributors of ;;; lisp compilers and interpreters to include this software ;;; with their distribution.
So we can distribute it, if and only if we include also a lisp compiler? Strange, but it still makes it DFSG non-free.
What are our options for this?
Groetjes, Peter
Peter Van Eynde pvaneynd@debian.org writes:
The file xref.lisp is not PD or GPL (unlike my greps showed) but has a rather obnoxious license that says:
You may try asking Mark Kantrowitz whether he is willing to distribute xref.lisp under a revised license. He did this for MK:DEFSYSTEM, which is now part of CLOCC (see src/defsystem-3.x/defsystem.lisp in the CLOCC distribution). If you have trouble contacting Mark, Marco Antoniotti, who took care of the MK:DEFSYSTEM license change, may be able to help.
Paolo
Il 2005-09-19, Paolo Amoroso amoroso@mclink.it ha scritto:
Peter Van Eynde pvaneynd@debian.org writes:
The file xref.lisp is not PD or GPL (unlike my greps showed) but has a rather obnoxious license that says:
You may try asking Mark Kantrowitz whether he is willing to distribute xref.lisp under a revised license. He did this for MK:DEFSYSTEM,
I contacted him and he is willing to allow xref to be distributed with slime, but is not willing to change the licence as he did for MK:DEFSYSTEM.
This is not 'free enough' for Debian so either we have to put slime into non-free or replace xref.lisp. Or a more able person could talk to him ;-).
Groetjes, Peter
Il 2005-09-23, Peter Van Eynde pvaneynd@debian.org ha scritto:
This is not 'free enough' for Debian so either we have to put slime into non-free or replace xref.lisp. Or a more able person could talk to him ;-).
If nobody reacts before 07:00:00 30/09/2005 UTC I will upload slime as a non-free package.
Groetjes, Peter
Peter Van Eynde pvaneynd@debian.org writes:
I contacted him and he is willing to allow xref to be distributed with slime, but is not willing to change the licence as he did for MK:DEFSYSTEM.
This is not 'free enough' for Debian so either we have to put slime into non-free or replace xref.lisp. Or a more able person could talk to him ;-).
I also emailed Mark and I just got the same response. He is happy with the idea of our bundling xref.lisp but he hasn't released it in a way that's suitable for Debian.
So we have three competing goals:
- Getting SLIME into debian proper. - Preventing Debian & non-Debian SLIME from having different features. - Making cross-referencing work out-of-the-box for CLISP users.
I rank them in the order listed and agree with Peter's suggestion that we stop bundling xref.lisp and just leave hooks for it instead. I don't like removing useful features but I'd like to see a full SLIME in Debian for easy access by newbies.
Major objections?
Cheers, Luke
P.S. Witness the invincible arm of Debian sending programs with awkward licenses hurtling into limbo! *WHOOSH*!
On 24 Oct 2005 01:54:13 +0200, luke@synap.se wrote:
- Getting SLIME into debian proper.
- Preventing Debian & non-Debian SLIME from having different features.
- Making cross-referencing work out-of-the-box for CLISP users.
I rank them in the order listed and agree with Peter's suggestion that we stop bundling xref.lisp and just leave hooks for it instead. I don't like removing useful features but I'd like to see a full SLIME in Debian for easy access by newbies.
Major objections?
Is there any indication that something 'bad' is happening? As near as I can tell, anyone interested in common lisp hears about slime pretty quick. The general suggestion is to run out of CVS in fact. Unless debian is tracking CVS closely, what slime a distro has is immaterial. Anyone serious quickly dumps a distro'ed slime rather fast (for instance, I had a cron cvs fetch running the second day of slime usage). Just a note, if you think that debian offers the best support of common lisp, Gentoo passed it some time ago in terms of offered packages.
Frankly, this looks like the old debian strongarm, you may recall the 'interesting conversations' that occurred at the birth of debian. Since the majority of lisp users I saw with computers at ILC were running either windows (i.e. probably clisp) or macs [only one linux user encountered], and the typical post in c.l.l mentions clisp more often than not, I suspect the downside here is deeper than imagined. Perhaps floating this idea on the clisp mailing lists might be something to consider.
P.S. Witness the invincible arm of Debian sending programs with awkward licenses hurtling into limbo! *WHOOSH*!
And what does it leave in their place? How many linux real-time plotter programs exist?
* Luke Gorrie [2005-10-24 01:54+0200] writes:
Major objections?
I've no objections, but a question.
How difficult would it be to write something like xref from scratch with a BSD-style license? Could somebody who has looked at the xref.lisp source still do that?
On Tue, 25 Oct 2005 12:19:56 +0200, heller@common-lisp.net wrote:
How difficult would it be to write something like xref from scratch with a BSD-style license? Could somebody who has looked at the xref.lisp source still do that?
The safe way would be a "Chinese Wall". Have someone that has looked at the code describe it (at high level, certainly not line-by-line) to someone that has never seen the code.
On Tue, 25 Oct 2005, GP lisper wrote:
The safe way would be a "Chinese Wall". Have someone that has looked at the code describe it (at high level, certainly not line-by-line) to someone that has never seen the code.
As far as I can tell this is quite unecessary. As long as you can make a case that the reimplementation is an independent effort, not derivative.
I would personally feel quite comfortable reimplementing xref.lisp (not that I have the time to do it) despite having read it. I would even feel comfortable to _refer_ to it while writing another implementation: "I wonder how it dealt with LET?". If the original author had a problem with that I they could sue me.
IANAL, but I don't let that bother me.
Cheers,
-- Nikodemus Schemer: "Buddha is small, clean, and serious." Lispnik: "Buddha is big, has hairy armpits, and laughs."
Apart from that, I think that ALBERT could be yet another alternative source. I was never able to quite follow through ALBERT's code, and I had problems with it, but then again I didn't put too much time into it.
Cheers -- Marco
On Oct 25, 2005, at 3:16 PM, Nikodemus Siivola wrote:
On Tue, 25 Oct 2005, GP lisper wrote:
The safe way would be a "Chinese Wall". Have someone that has looked at the code describe it (at high level, certainly not line-by-line) to someone that has never seen the code.
As far as I can tell this is quite unecessary. As long as you can make a case that the reimplementation is an independent effort, not derivative.
I would personally feel quite comfortable reimplementing xref.lisp (not that I have the time to do it) despite having read it. I would even feel comfortable to _refer_ to it while writing another implementation: "I wonder how it dealt with LET?". If the original author had a problem with that I they could sue me.
IANAL, but I don't let that bother me.
Cheers,
-- Nikodemus Schemer: "Buddha is small, clean, and serious." Lispnik: "Buddha is big, has hairy armpits, and laughs." _______________________________________________ slime-devel site list slime-devel@common-lisp.net http://common-lisp.net/mailman/listinfo/slime-devel
-- Marco Antoniotti http://bioinformatics.nyu.edu/~marcoxa NYU Courant Bioinformatics Group tel. +1 - 212 - 998 3488 715 Broadway 10th FL fax. +1 - 212 - 998 3484 New York, NY, 10003, U.S.A.
On Tue, 25 Oct 2005 22:16:04 +0300 (EEST), tsiivola@cc.hut.fi wrote:
On Tue, 25 Oct 2005, GP lisper wrote:
The safe way would be a "Chinese Wall". Have someone that has looked at the code describe it (at high level, certainly not line-by-line) to someone that has never seen the code.
As far as I can tell this is quite unecessary. As long as you can make a case that the reimplementation is an independent effort, not derivative.
The purpose of the technique is to eliminate any question of source code infringement, i.e. avoid the 'make the case'. This whole question is raised due to the picky nature of Debian licensing desires, and the desire not to cut down slime to meet debian requirements.
If the original author had a problem with that I they could sue me.
--- GP lisper spambait@CloudDancer.com wrote:
On Tue, 25 Oct 2005 12:19:56 +0200, heller@common-lisp.net wrote:
How difficult would it be to write something like xref from scratch with a BSD-style license? Could somebody who has looked at the xref.lisp source still do that?
The safe way would be a "Chinese Wall". Have someone that has looked at the code describe it (at high level, certainly not line-by-line) to someone that has never seen the code.
Also of use might be the xref implementation in cmucl, which I don't think has any license issues. Of course it's probably very cmucl specific, but perhaps it would be a useful reference.
CY
__________________________________ Yahoo! Mail - PC Magazine Editors' Choice 2005 http://mail.yahoo.com
So who is really using XREF with CLISP for WHO-CALLS etc? Anyone?
IMO the really nice feature of XREF is PRINT-CALLER-TREES which shows you the subroutine tree structure of a program. I will be surprised if WHO-CALLS via SLIME is popular since XREF is mostly batch-mode and won't typically have an update to date database to make impromptu queries to. (There is no hook into the compiler for updates).
XREF lovers please speak up if I am mistaken :-)
P.S. Getting SLIME into Debian proper is absolutely a Very Good Thing and I am very grateful to Peter for his patience!
You could of course add a few lines that download and install xref.lisp on request. I'm under the impression that packaging an installer instead of the program itself is a time-honored technique in Debian. ;-)
Cheers,
-- Nikodemus Schemer: "Buddha is small, clean, and serious." Lispnik: "Buddha is big, has hairy armpits, and laughs."