Re: NetBeans ITP [was Re: CDDL]
Hi, On Sat, Dec 02, 2006, Tom Marble wrote: Once the the full JVM is available under GPL then running applications on top of it *are* compatible with any license as this was the specific rationale for adding the Classpath exception [1]. I think it can even go in contrib if it ends up being considered free, which is the place for free software which requires non-free software. It doesn't change a lot for the autobuilding of the Netbeans package which is probably arch: all, but it would mean it is in Debian proper (the non-free section is not part of Debian, it's just supported on a best-effort basis). It sounds like you are not aware that Daniel Baumann already uploaded Netbeans with Section: contrib/devel in the diff. You can watch packages pending reviewal by the Debian Ftp Masters in the NEW queue at: http://ftp-master.debian.org/new.html And Daniel Baumann explained he uploads all packages on his personal repository as well, so you can check the .diff.gz of his proposed package at: http://archive.daniel-baumann.ch/debian/packages/netbeans/5.5-1/ (where you should find debian/control with the Section:) Bye, -- Loïc Minier [EMAIL PROTECTED] You see, killbots have a preset kill limit. Knowing their weakness, I sent wave after wave of my own men at them until they reached their limit and shutdown.-- Zapp Brannigan -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Bug#386406: libmms: majormms haven't yet granted relicensing to LGPL
On Wed, Sep 13, 2006, Don Armstrong wrote: Ah; my understanding was that there were some parts (the xmms header files?) which were GPL only, and that the rest of libmms was licenced under the LGPL, which is what I retitled the bug to clarify. [If that's actually the case, as debian/copyright currently indicates, it should eventually be modified with the changes I suggested... but that's a trivial issue.] My picture is the following: the original code was written by MajorMMS, some files were contributed by Xine. The relicensing of the Xine files was publicly stated, but not the relicensing of the MajorMMS code. Let me be as unequivocal as I can be: There is no factual ground or reasonable reason to strip out the LGPL and return the work to the GPLed state. Because the LGPL can always be returned to the GPL, it is always compatible with the GPL. We as distributors do not need to reduce the flexibility we offer to those using the distribution.[1] 1: We obviously could... but it is not required. Good, I'm of this opinion as well! If upstream has released it under the LGPL, then that's good enough. Even if they've made a mistake, and parts of it are still GPLed, that's still fine because the LGPL is explicitly compatible with the GPL. All that would mean is that the copyright file has a fixable error in it because it doesn't document the actual license of parts carefully enough. Yes; it also has consequences on the actual runtime license of some dependent projects. The only questionable part comes if you actually need the permissions of the LGPL in Debian... but I assume that's not the case here. [And baring clarification from the submitter, we should assume that the relicensing was carried out with due dilligence, and that those extra permisions are actually available.] We don't need it *for Debian*, since it's DFSG-free enough to have all code in Debian GPL, however some people relying on libmms directly or indirectly might rely on the fact they believe it's LGPL, and they might be using Debian. Anyway, bug tagged as waiting for factual information, if I get some, I shall update the copyright, it's as simple as that. :) -- Loïc Minier [EMAIL PROTECTED] -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Bug#386406: libmms: majormms haven't yet granted relicensing to LGPL
assume that only the codes from xine, and the new codes added by libmms, is relicensed. They haven't said anything about relicensing majormms codes, then we can assume that majormms codes haven't yet been relicensed. You are assuming this. And upstream claims that libmms is being available as library, licensed under GNU Library General Public Licence (LGPL). Again, this is referenced in the copyright file. Everybody has the same information as me when I declared it could be uploaded LGPL. Even if relicensing the Debian package would have some consequences, I am open to this solution *if* this has some factual and recent ground. I currently didn't see any useful new piece of information, and would like to get convinced that the licensing is required (since I had already convinced myself that the LGPL was fine when I prepared the package in the first place). It won't be fine if someone use this library in proprietary (or any GPL-incompatible), and if this is not what majormms author want, he may come out to claim his right, and Debian (including libmms) may be sued. It won't be fine for the consumers of libmms based stacks *outside of Debian*. As far as I know, Debian does not ship packages incompatible with a GPL-ed libmms. I'm not deliberately making a promess of a LGPL libmms library in order to trick propreitary application authors relying on Debian to have sane licenses. I am using what I believe is the appropriate license for the libmms library in Debian and the upstream license. I already suggested what you could do in order to make me change my mind on the status of the GPL versus LGPL licensing of libmms; please refer to my earlier messages if you truly want to clarify this (upstream) issue. For the record, the decision to license as LGPL was taken during the discussion of the ITP, with James Troup (I think he reviewed libmms for Ubuntu and he is a Debian ftpmaster); see Debian #330355 for details: http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=330355 Bye, -- Loïc Minier [EMAIL PROTECTED] -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Bug#386406: libmms: majormms haven't yet granted relicensing to LGPL
severity 386406 important tags 386406 + moreinfo retitle 386406 Some bits might not have been properly relicensed, but please provide/gather more information stop Hi, On Tue, Sep 12, 2006, Don Armstrong wrote: Loic: The only actual (minor) issue here is that the package doesn't reference the GPL and include the copyright+licensing statement of the parts that are GPL licensed. It would also be useful to include the actual copyright statements of the parts which you have included the LGPLed license statement. [Adjusting the severity accordingly; feel free to override.] The upstream tarball is said to be LGPL, and was relicensed as such after the message the OP links to. This is the result of an upstream relicensing work, which was announced publicly and followed by a tarball release. My understanding is that the submitter claims that this relicensing is incomplete and/or erroneous, but doesn't provide any further information than references to messages before the relicensing happened. Even if relicensing the Debian package would have some consequences, I am open to this solution *if* this has some factual and recent ground. I currently didn't see any useful new piece of information, and would like to get convinced that the licensing is required (since I had already convinced myself that the LGPL was fine when I prepared the package in the first place). Perhaps the best way to achieve this is for the OP to clarify the conditions of the relicensing with upstream. Before anyone suggest I should be the one doing this, please recall that I already did it when uploading to Debian, and there's no new piece of information in the report. Thanks for discussion, -- Loïc Minier [EMAIL PROTECTED] -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Bug#386406: libmms: majormms haven't yet granted relicensing to LGPL
Hi, On Mon, Sep 11, 2006, Anon Sricharoenchai wrote: If there's one upstream that extract and modify dll files from Windows Media Player and relicense those dll to some freeware license, then this upstream software can be packaged into Debian/non-free? Although it is apparently that Microsoft doesn't give any privilege for them to relicense? I think you want the obvious answer: no. And in this case, we would likely question: where do the DLL come from (where's the code); but you're taking a not too good example since: 1) it involves non-free 2) it involves binary only code. In libmms, we're talking about source code which has a public history, and we're shipping it in main. If Debian redistribute this package in freeware license, Debian will risk to get sued by Microsoft? Yes. Or Debian can claim that, hey, it's not our responsibility, we got the valid license from the upstream author, if you (Microsoft) get loss, you should sue the upstream author to get compensate for all of the loss resulted by such relicensing? We can always argue that we did a mistake, and try our best to repair it. Can debian relicense this package back to GPL to avoid this conflict? Sure, as soon as this is proven to be LGPL or not LGPL one way or the other, I will act accordingly. According to Section 3 of LGPL, Debian or anyone can relicense LGPL software into the ordinary GPL. Debian can decide to relicense this libmms to GPL to avoid licensing problem with the original author of majormms, until the upstream author of libmms have cleared the problem about their license. This is just an interim solution. First, I won't relicense any application that is currently LGPL as GPL in Debian only following the claim of someone that wants me to do so. I think we should follow the upstream license whenever possible. Beside, some stacks (such as GStreamer) benefit from the fact that this is LGPL and *not* GPL. So, no, I don't intend to just relicense it as GPL if this is not required. Bye, -- Loïc Minier [EMAIL PROTECTED] -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Bug#386406: libmms: majormms haven't yet granted relicensing to LGPL
Hi, On Thu, Sep 07, 2006, Anon Sricharoenchai wrote: According to, http://marc.theaimsgroup.com/?l=xine-develm=107261185004445w=2 , since they can't find where to contact the author of majormms, either in majormms website, http://www.geocities.com/majormms/ , and its source, http://www.geocities.com/majormms/mms_client-0.0.3.tar.gz , it still can't ensure that the author of majormms will allow relicensing to LGPL or not. The URL you mention is also present in the copyright file, this is no new information. Upstream publicly announced the LGPL relicensing. They explained part of the process they followed and what they got relicensed. I even challenged this state of fact by verifying part of this relicensing and by contacting the Ubuntu ftpmaster who reviewed the source package for Ubuntu and found it to be LGPL. Now, what is the basis of your claims that this might not be LGPL? What new information did you find? And why are you discussing this with Debian? This seems like an upstream problem to me and something to discuss with upstream. I think you can't blame Debian for trusting its own upstream software distributors to a certain extent. Can debian relicense this package back to GPL to avoid this conflict? Sure, as soon as this is proven to be LGPL or not LGPL one way or the other, I will act accordingly. My current position is to trust upstream that this is LGPL, I've verified what I could of this information when I uploaded libmms to Debian. New information may change this position. Bye, -- Loïc Minier [EMAIL PROTECTED] -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: LGPL module linked with a GPL lib
Hi, On Mon, Jul 25, 2005, Jeff Licquia wrote: From the GPL: Activities other than copying, distribution and modification are not covered by this License; they are outside its scope. The act of running the Program is not restricted... So the particular details of how things are distributed in memory while running aren't directly relevant. Modification and distribution are what matters, and it's clear from looking at the packages that GStreamer is distributed in Debian in conjunction with GPLed bits in a manner that's more than mere aggregation. I'm not sure to understand: you mean that since some LGPL GStreamer plugins are shipped in Debian along with GPL packages and they can play together means that the whole is GPLed? Would it be ok to have a copyright file along these lines: The source code for all plugins in the GStreamer Plugins source package is licensed under the LGPL, however some plugins are built with the help of header files from GPL libraries, and will be linked to GPL libraries when loaded in memory. Thus, using these plugins will switch their license to GPL, and you can only use them in applications with a license compatible with the GPL. You should have received a copy of the GPL and LGPL licenses ... Is a list of plugins necessary? I guess it's up to the interested person to check, nowadays it's relatively easy with tags and Debian's copyright files, and I don't want to maintain such a list. Bye, -- Loïc Minier [EMAIL PROTECTED] Come, your destiny awaits! -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: LGPL module linked with a GPL lib
Hi, I agree with most of what you said, except I'd like clarification on this part: On Sun, Jul 24, 2005, Jeff Licquia wrote: The copyright of the rest of GStreamer depends on how it's distributed. In Debian, it's clear that GStreamer is distributed with MAD support, which makes its effective license the GPL. However, someone interested in distributing proprietary plugins or apps for GStreamer (as part of a derivative, for example) could do so by removing the GPL plugins from the distribution before adding the non-free bits. This wouldn't even require a recompile to do. GStreamer's build process builds separate binaries for the various plugins, these are then dlopened when requested. I would personnally think that installing only Debian's GStreamer packages that are linked to LGPL libraries doesn't make your GStreamer installation / packages GPL (that is the build process has nothing to do with the resulting packages). I would even thing that installing GStreamer plugins packages which link to GPL libraries don't make your installation nor your running GStreamer applications GPL (that is only dlopening() something GPL makes the whole program in memory GPL, while it remains in memory). Is that correct? Bye, -- Loïc Minier [EMAIL PROTECTED] Come, your destiny awaits! -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
LGPL module linked with a GPL lib
Hi, The GStreamer suite ships a lot of plugins which are dlopened() when needed. Some of them link with GPL libraries. I received a bug report (#317129) to change the copyright files of libgstreamer0.8-0 and gstreamer0.8-mad to GPL. The upstream README mentions the situation, so I think I will mention it in the README.Debian with the next upload, but is the copyright supposed to reflect this? Does the whole distribution switch to GPL? I believe not, but would like a confirmation. Bye, -- Loïc Minier [EMAIL PROTECTED] Come, your destiny awaits! -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Patents on encoders in Europe
Hi, With the recent clarifications on software patents in Europe, would it be possible to distribute encoders packages from Europe? My current understanding is that the algorithm can be patented, but a pure software implementation is not violating such a patent. Is that correct? Bye, -- Loïc Minier [EMAIL PROTECTED] Come, your destiny awaits! -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Bug#289856: mdnsresponder: Wrong license
Marco d'Itri [EMAIL PROTECTED] - Fri, Jan 21, 2005: Do you suggest removing from the archive all packages whose licenses impose uncommon restrictions or just this one? In this software the problem is two folds, some parts of the software are clearly free, and some other parts are a fork of some code under APSL 2. Which leaves two options: rewriting the non-free part, or splitting in two packages, one for contrib and one for non-free. I think some software was already built on the libs provided by this package[1], so it is not trivial at all (indirect dependencies make some packages depend on it via the shlibs mecanism). Which makes me wonder whether APSL 2 is acceptable for non-free? Regards, [1] bee% grep-available -FDepends libhowl -sPackage Package: gnome-terminal Package: gnome-gv Package: nautilus-cd-burner Package: libgnomeui-0 Package: galeon Package: libhowl-dev Package: howl-utils Package: libgnomevfs2-common Package: gnome-games Package: epiphany-browser Package: gnome-pilot Package: gnome-session Package: libgnomevfs2-0 Package: libgnomevfs2-dev Package: totem-xine -- Loïc Minier [EMAIL PROTECTED] Neutral President: I have no strong feelings one way or the other. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: firmware status for eagle-usb-*
Martin Braure de Calignon [EMAIL PROTECTED] - Sat, Oct 09, 2004: I wanted to know if the binary files in the eagle-usb-{utils,data,source} package are free. When I get the source of the package (apt-get source), there is a LICENSE file in the root directory which says that the package is GPL. I've discussed this with the maintainer of eagle-usb-* packages a while ago, and he explained to me that Sagem (upstream for the linux driver and firmware) had released the tarball in two versions, the first one being clearly unclear about the licensing of binary files. The second release of Sagem, which was done thanks to the efforts of Pierre Machard to clarify the situation with Sagem, includes a clear statement that all files in the packages are under the GPL, even the binary firmware. [ Pierre Machard was kind enough to send me the mails he exchanged with people at Sagem, but since they're in french, I'm afraid you wouldn't want me to forward them to the list. You seem to be a native french speaker, so I might send them offlist -- if you want. ] If you wonder wether the sources of the firmware are distributed, I think they aren't. But I think this is a different program we are talking about. The program act as a firmware to drive the [EMAIL PROTECTED] 800 modem could have been put as a program under the GPL, but I think the firmware data itself is under GPL for its use in the program Sagem drivers for the [EMAIL PROTECTED] 800 modem -- ie the linux driver and utilities. You can redistribute the binary firmware freely under the terms of the GPL, as you would with a logo for example. Regards, [ I'd prefer Pierre Machard discussing this himself, but he still needs some rest. I wish he'll be back real soon now! For now, I'm afraid he doesn't follow lists too much. ] -- Loïc Minier [EMAIL PROTECTED]
Re: firmware status for eagle-usb-*
Josh Triplett [EMAIL PROTECTED] - Mon, Oct 18, 2004: I don't believe you can. In order to distribute software under the GPL, we must provide the preferred form for modification of that software, which is the source. From your description, it sounds like such source exists but is not being distributed. This means that we do not have the preferred form for modification available, so we cannot make it available to others, which means we can't satisfy our obligations under the GPL, and therefore we cannot distribute the software at all. sillyWhat makes you think there's a program in firmware-xyz.bin?/silly I think you should re-read my mail, I'm talking about two different programs: a linux driver, with some utilities, with some data you're usually interested to send (for example a logo to display on a LCD) but that only affect a non-Debian system; *AND* a firmware program, that might or might not be a program (what if it is a configuration file?). Now, we both know the firmware is certainly a program too, but it is not the same program: for a Debian system running the drivers, this is mere data. I know this is an ambiguous position, and I'm sorry I have to take it, but think for a minute the modem would be running on a ROM and not a RAM, or on a Flash! We wouldn't have to send that binary blob that we don't need to worry about when it's already there. Regards, -- Loïc Minier [EMAIL PROTECTED]
Re: firmware status for eagle-usb-*
[ Stop Cc:ing me please, I read this mailing list. ] Josh Triplett [EMAIL PROTECTED] - Mon, Oct 18, 2004: This argument has been made before, and the clear consensus is that firmware is software; this is even clearer than the situation over documents and other data, which were also decided (on a project-wide basis) to be software. There is still occasional argument on the topic related to the status of the driver if Debian doesn't distribute the firmware at all, but there is a strong consensus against including the binary firmware image without source in Debian. I'm interested, would you please provide some pointers on those discussions? My question to you is this: does the author of the firmware edit the binary directly, or do they have some other format? If the latter, it shouldn't be in Debian main, and if it is GPLed, we can't legally distribute it at all. Once again, I think Debian won't be building or running the program that is hidden in the binary blob. I don't see how that relates to the program included in Debian. The binary blob is needed as well as you need to talk I'm not sure if I'm parsing this correctly. Are you saying that some modems don't need the firmware to be provided? If that is the case, then the driver has some useful functionality without the non-free software, so it can go to main (without the firmware); then you would just need to figure out the legal details of supplying the firmware (such as the GPL-without-source issue). No I'm saying that if the modem would work without the need to send a firmware first, or if you could send a firmware if you wish but that is not mandatory, then nobody would complain about the abscence of the sources of a program that runs on the modem. I am comparing the usefulness of the source code of firmware: suddenly, when the firmware has not a permanent storage on the device, its source needs to be GPL too! How logical is this? Most recent computers have flashable BIOSes, or flashable VGA card BIOSes, and nobody went asking for the source of these programs. Nobody is interested in modifying the source of such programs, and if they were, _this_ would be a project on its own. Regards, -- Loïc Minier [EMAIL PROTECTED]
Re: firmware status for eagle-usb-*
Loïc Minier [EMAIL PROTECTED] - Tue, Oct 19, 2004: The binary blob is needed as well as you need to talk Sorry, copy-paste problem, forget that half-sentence. -- Loïc Minier [EMAIL PROTECTED]
Re: firmware status for eagle-usb-*
Don Armstrong [EMAIL PROTECTED] - Mon, Oct 18, 2004: No sourcecode bits: http://people.debian.org/~terpstra/thread/20021106.222149.24f92b22.en.html Quite interesting, although related to code running on the host, most of the thread is interesting. In the context of DSP Binaries: http://people.debian.org/~terpstra/thread/20030922.064726.2833dd35.en.html Interesting, and particularly this: http://people.debian.org/~terpstra/message/20030924.002544.1f7d5160.en.html (where it is stated that if the source has been lost, binary is the preferred form of modification, which could eventually match the Sagem case) And the incredibly gargantuan keep non-free proposal thread: http://people.debian.org/~terpstra/thread/20040129.052350.5b5e7192.en.html It's 3am, I won't read that one! ;) And finally: http://www.debian.org/vote/2004/vote_003 and http://www.debian.org/vote/2004/vote_004 Hmm that's a strong commitment of Debian, and I can't argue against that. Anyway, my initial goal was only to report some history of Sagem with this driver, and stating that Pierre Machard cann't reply to this right now. I'm really sorry I re-started a long-discussed troll again, and I'm sad Debian won't provide support for a lot of hardware in a close future. Regards, -- Loïc Minier [EMAIL PROTECTED]