Thread: RE: [HACKERS] Readline use in trouble?
Bruce wrote: [snip] > > Although I haven't been paying close attention to the Ghostscript > > situation, I suspect that the real story is either that the readline > > interface code that someone contributed to Ghostscript was contributed > > with GPL terms already attached to it, or that Aladdin is concerned > > Oh, that is an interesting issue that I never considered. Reminds us we > can't use GPL code. That was why when I merged Adrian's and my JDBC drivers for inclusion, I used Adrians as the core, and re-wrote the additions I made to mine to it, as mine was GPL'ed, and his wasn't. Just a thought: How does this affect anything placed in the contrib directory? If someone writes a tool under the GPL, can it be included under the src/contrib directory, or would we fall foul just because it's included with our source? I don't think we have a problem with the CD distribution, as they are clearly separate, but contrib is not that clear cut. Peter -- Peter Mount Enterprise Support Maidstone Borough Council Any views stated are my own, and not those of Maidstone Borough Council.
Peter Mount <petermount@it.maidstone.gov.uk> writes: > Just a thought: How does this affect anything placed in the contrib > directory? If someone writes a tool under the GPL, can it be included > under the src/contrib directory, or would we fall foul just because it's > included with our source? Good question. The GPL contains a clause to the effect that "mere aggregation" of a GPL'd piece of code in a source distribution with unrelated pieces of code is OK, even if those other pieces of code are not GPL'd. But the contrib directory is not exactly unrelated to the main Postgres distribution, so I'm not sure that we can point to this clause to justify putting a GPL'd program in contrib. It'd be a gray area... I'd be inclined to say "if you want to put your tool under GPL, fine, but then distributing it is up to you". We don't need to be taking any legal risks on this point. A safe policy is that everything distributed by the Postgres group has to carry the same BSD license. regards, tom lane
On Wed, 20 Oct 1999, Tom Lane wrote: > Peter Mount <petermount@it.maidstone.gov.uk> writes: > > Just a thought: How does this affect anything placed in the contrib > > directory? If someone writes a tool under the GPL, can it be included > > under the src/contrib directory, or would we fall foul just because it's > > included with our source? > > Good question. The GPL contains a clause to the effect that "mere > aggregation" of a GPL'd piece of code in a source distribution with > unrelated pieces of code is OK, even if those other pieces of code > are not GPL'd. But the contrib directory is not exactly unrelated > to the main Postgres distribution, so I'm not sure that we can point > to this clause to justify putting a GPL'd program in contrib. It'd > be a gray area... Items in the contrib section aren't required for the use of PostgreSQL, however PostgreSQL *is* required to use those items. So shouldn't the items in contrib have to change to a Berkeley style license? :) I mean it's only fair! Vince. -- ========================================================================== Vince Vielhaber -- KA8CSH email: vev@michvhf.com flame-mail: /dev/null # include <std/disclaimers.h> Have you seenhttp://www.pop4.net? Online Campground Directory http://www.camping-usa.com Online Giftshop Superstore http://www.cloudninegifts.com ==========================================================================
Good question. The GPL contains a clause to the effect that "mere aggregation" of a GPL'd piece of code in a sourcedistribution with unrelated pieces of code is OK, even if those other pieces of code are not GPL'd. But the contribdirectory is not exactly unrelated to the main Postgres distribution, so I'm not sure that we can point to thisclause to justify putting a GPL'd program in contrib. It'd be a gray area... The problem only comes if I, for example, want to distribute all of postgresql (contrib included) in a non-source (i.e., proprietary) way. That is fine if contrib includes no GPL code; if it does, I need to distribute the code for that portion only. Thus, if we want to maintain as broad a potential as possible (including non-source distributions) we need to encourage adoption of the BSD license for all source. To make it easier for those distributing postgresql to keep track of this stuff, perhaps we need a gnu (or gpl) directory (like or under contrib) in which would go GPL code. Then it would be crystal clear which portion of the code has which restrictions. It would also be clear that this is an aggregation. This is the mechanism used by NetBSD for their code tree, which does include some gnu software. Still, encouraging non-GPL contrib stuff is a good thing in order to maintain future options, because GPL contrib code _cannot_ be added to the main tree without affecting the distribution of the entire thing. Cheers, Brook
> Good question. The GPL contains a clause to the effect that "mere > aggregation" of a GPL'd piece of code in a source distribution with > unrelated pieces of code is OK, even if those other pieces of code > are not GPL'd. But the contrib directory is not exactly unrelated > to the main Postgres distribution, so I'm not sure that we can point > to this clause to justify putting a GPL'd program in contrib. It'd > be a gray area... > > The problem only comes if I, for example, want to distribute all of > postgresql (contrib included) in a non-source (i.e., proprietary) way. > That is fine if contrib includes no GPL code; if it does, I need to > distribute the code for that portion only. Thus, if we want to > maintain as broad a potential as possible (including non-source > distributions) we need to encourage adoption of the BSD license for > all source. But Alladin Ghostscript is distributed in source form. This GPL legal stuff is a terrible hassle. -- Bruce Momjian | http://www.op.net/~candle maillist@candle.pha.pa.us | (610) 853-3000+ If your life is a hard drive, | 830 Blythe Avenue + Christ can be your backup. | Drexel Hill, Pennsylvania19026
Bruce Momjian <maillist@candle.pha.pa.us> writes: >> That is fine if contrib includes no GPL code; if it does, I need to >> distribute the code for that portion only. Thus, if we want to >> maintain as broad a potential as possible (including non-source >> distributions) we need to encourage adoption of the BSD license for >> all source. > But Alladin Ghostscript is distributed in source form. But not *only* in source form. Aladdin make their living by selling Ghostscript to printer manufacturers and so forth. The printer makers are not about to ship out printers with copies of source code, nor even with notices explaining where to get the printer source code. If they obtained Ghostscript under GPL then they'd have to make not only the PS interpreter source available, but probably the entire firmware for the printer (it's a derived work, no?) and they are certainly not about to do that. So they pay Aladdin for the rights to use Ghostscript with a commercial license instead of GPL. In the same way, if we distributed Postgres under GPL, it would not be possible to sell proprietary systems that use Postgres as a component. That is, in fact, exactly what the GPL is designed to prevent. But it doesn't strike me as something we want for Postgres. We'd be cutting off too much of the potential "market" of Postgres users. (Not only would we lose companies who had an immediate interest in selling DBMS-based code, but also those who had any thought of possibly doing so in the future; that could be a lot of people.) regards, tom lane
> Bruce Momjian <maillist@candle.pha.pa.us> writes: > >> That is fine if contrib includes no GPL code; if it does, I need to > >> distribute the code for that portion only. Thus, if we want to > >> maintain as broad a potential as possible (including non-source > >> distributions) we need to encourage adoption of the BSD license for > >> all source. > > > But Alladin Ghostscript is distributed in source form. > > But not *only* in source form. Aladdin make their living by selling > Ghostscript to printer manufacturers and so forth. The printer makers > are not about to ship out printers with copies of source code, nor > even with notices explaining where to get the printer source code. > If they obtained Ghostscript under GPL then they'd have to make not > only the PS interpreter source available, but probably the entire > firmware for the printer (it's a derived work, no?) and they are > certainly not about to do that. So they pay Aladdin for the rights > to use Ghostscript with a commercial license instead of GPL. Oh, I didn't realize they had a binary-only distribution that was _different_ from the source distribution. -- Bruce Momjian | http://www.op.net/~candle maillist@candle.pha.pa.us | (610) 853-3000+ If your life is a hard drive, | 830 Blythe Avenue + Christ can be your backup. | Drexel Hill, Pennsylvania19026
Vince Vielhaber wrote: > Items in the contrib section aren't required for the use of PostgreSQL, > however PostgreSQL *is* required to use those items. So shouldn't the > items in contrib have to change to a Berkeley style license? :) > > I mean it's only fair! I know of at least two items in contrib that are required to run the regression tests -- which, arguably, make PostgreSQL require those two components (autoinc and refint). And, the GPL is not fair. It is highly restrictive to programmer freedom in ways (and promotes code freedom in others -- for many things it makes sense). It's not called the 'GNU Public Virus' without merit. Lots of great software has been GPL'd -- and that's fine. But PostgreSQL is not -- and if PostgreSQL wants to remain BSD'd, then GPL'd code is a real sticky mess that's best left alone. I am not against either of these two licenses -- but the known issues of dealing with them have to be understood, or problems may arise. JMHO, of course. -- Lamar Owen WGCR Internet Radio 1 Peter 4:11
Oh, I didn't realize they had a binary-only distribution that was _different_ from the source distribution. DIFFERENT is not relevant. I could today ship a binary verion of postgresql in its present form as a proprietary product with no source code. No license problems arise from doing so. Allowing GPL code into the base system causes the problem. As just said, this is a good thing from the point of view of encouraging participation and commercial success. As long as the open source version of postgresql remains a well-designed, solid product it behooves any commercial distributor to aid in its maintenance rather than take on the whole thing. Ideally, they will contribute any fixes they make so that all can benefit and perhaps more importantly so they don't have to maintain the separate fixes any longer. Cheers, Brook
> Oh, I didn't realize they had a binary-only distribution that was > _different_ from the source distribution. > > DIFFERENT is not relevant. I could today ship a binary verion of > postgresql in its present form as a proprietary product with no source > code. No license problems arise from doing so. Allowing GPL code > into the base system causes the problem. Does GPL require the source to be included, or just available for free? -- Bruce Momjian | http://www.op.net/~candle maillist@candle.pha.pa.us | (610) 853-3000+ If your life is a hard drive, | 830 Blythe Avenue + Christ can be your backup. | Drexel Hill, Pennsylvania19026
Does GPL require the source to be included, or just available for free? Pretty sure just a pointer to location is good enough. The catch is that it has to be the exact version shipped and there is a time limit (3 years?) for availability, so pointing to the gnu ftp site probably doesn't work. Cheers, Brook
On Wed, 20 Oct 1999, Bruce Momjian wrote: > Does GPL require the source to be included, or just available for free? Require sources to be available. It is enough to distribute a binary and provide a pointer to sources. (There are some obscured words that the pointer should be available for general public...) Oleg. ---- Oleg Broytmann http://members.xoom.com/phd2/ phd2@earthling.net Programmers don't die, they justGOSUB without RETURN.
Lamar Owen wrote: > > Vince Vielhaber wrote: > > Items in the contrib section aren't required for the use of PostgreSQL, > > however PostgreSQL *is* required to use those items. So shouldn't the > > items in contrib have to change to a Berkeley style license? :) > > > > I mean it's only fair! > > I know of at least two items in contrib that are required to run the > regression tests -- which, arguably, make PostgreSQL require those two > components (autoinc and refint). Well, I made them... so you know what copyright is... -:) Vadim