Embedded systems OS/FS
Trying to pick some brains before I get up on stage and make a fool of myself again (Intel Developers Conference). I'm told the audience will be a mix of both SW and HW developers, with the HW folks doing a mix of embedded devices and chip/circuit designs. Question I've got: how does software licensing, free/OS or otherwise, effect the hardware market. My read is that some licenses, notably the GNU L/GPL, may have their source availability requirements triggered by the physical distribution of media (HW) on which the software is embedded, etched, or otherwise fixed. The primary statuatory provisions of copyright (in the US) are of the reproduction, making derivative works, distributing, performing, and displaying of copies. The GPL specifically restricts itself to "copying, distribution and modification" (section 0). An instance in which this would matter: ACME Mfg. creates printers. They incorporate GPLd code 'gnuprint' into their product 'acmeprint', creating a derived work 'gnuacmeprint' of the two programs. In distributing the printers to wholesalers and eventually customers, does ACME trigger the GNU GPL's source distribution and relicensing requirements? To whom does the source obligation apply -- wholesalers, final customers, or both? My read is that yes, ACME does. The code to 'gnuacmeprint' must be licensed under the GPL, and the terms of 3(a) or 3(b) of the GNU GPL must be met. I'm not sure that the wholesalers would have a solid claim to code, the end customer certainly would. Other takers? -- Karsten M. Self [EMAIL PROTECTED] http://www.netcom.com/~kmself Evangelist, Opensales, Inc.http://www.opensales.org What part of "Gestalt" don't you understand? Debian GNU/Linux rocks! http://gestalt-system.sourceforge.net/K5: http://www.kuro5hin.org GPG fingerprint: F932 8B25 5FDD 2528 D595 DC61 3847 889F 55F2 B9B0 PGP signature
Re: RMS on OpenMotif
On Mon, 21 Aug 2000, David Johnson wrote: will not be defined by the Windows98 user, but by the typical user of Unix (since that is the platform for Motif), who would have a radically different perception of what an OS is. If you're going to get into legal terms, then no, Unix does not have relevance here, since we're talking about free operating systems, none of which are Unix(tm). Yes, they work the same, but no, they're not Unix(tm). Since clearly Motif runs on them, Motif runs on more than Unix, so Unix definitions are insufficiently broad. Aren't there versions of Motif that run under Windows, too? We can argue this all day (I've gotten into arguments about definitions before, too, so I *know* we can argue this all day). My point is that we *can* argue it, and we can both make reasonable arguments, at which it comes down to with which a judge agrees. Don't leave it up to the lawyers and judges and jurors involved in a particular case -- just define the terms. Matthew Weigel Programmer/Student [EMAIL PROTECTED]
Re: Plan 9 license
On Mon, 21 Aug 2000, David Johnson wrote:
I'm not certain I agree with that, myself. Its requirement that
licensees choose between licensing Plan 9 and being able to protect
their intellectual property is particularly onerous. The right of Bell
Labs to demand private source is also unacceptable.
I don't see anything in the OSD forbidding the demand of private
derivations. This clause is certainly pretty poor, but it doesn't go
against the *letter* of the OSD.
Read what I said again. I'm saying it should be in the OSD, because the
clauses are "particularly onerous" and "unacceptable." I would argue the OSD
was not, is not, and never will be an airtight document whose letter you can
follow, but whose spirit you can ignore; the letter of the definition was
largely thrown together, in an attempt to encapsulate the spirit. Others
can say, "here, approximately, is what Open Source is," but the OSI and the
open source community should focus on the spirit, and when inconsistencies
arise, correct the letter, and not say "oh, we forgot to mention that
freedom, so we'll just roll over and ignore that it was ever important to
us." Yes, that means people submitting new licenses may be caught by
surprise; but OSI Certification should not be an assembly line production,
but a discussion (a discussion including the license's authors, so that
questions and explanations of intent, that might clarify a clause or get a
clause changed, can happen).
To quote from the page in question, "More precisely, it refers to four
kinds of freedom, for the users of the software... A program is free
software if users have all of these freedoms. " This seems to sum it up
for me. The rest is just commentary.
Yes, four *kinds* of freedom -- to be explained further in the document;
think of it as an explanation and statement of intent. The document is not
the four bullet points, and the clarifications and explanations included
discuss those four freedoms as RMS sees them. In order to have those four
kinds of freedom, there are preconditions, clarifications, and explanations
involved as well -- as discussed later in the document.
Later on, he does make some clarifications. One of them (I found two)
is: "but what they can and must do is refuse to impose them [export
control regulations] as conditions of use of the program". This may
apply to the Plan 9 license.
I think it's pretty clear it does ("can and must"). It would also be a part
of freedom #2, the freedom to redistribute copies so you can help your
neighbor. I think it's also pretty clear you were not reading very closely;
In order for these freedoms to be real, they must be irrevocable as
long as you do nothing wrong; if the developer of the software has
the power to revoke the license, even though you have not given
cause, the software is not free.
This goes back to the IP suit clause in the Plan 9 license, and is also
clearly a requirement for being free software -- you must have the four
freedoms, and the four freedoms must be real.
But whether this clarification is part of
the Free Software definition is, in my mind, debatable, as it still
meets the four necessary and sufficient freedoms stated at the top of
the page.
No. If you want to say "necessary and sufficient," then look at it like a
theorem, and you'll see again what I'm saying. Each of those conditions
has, on its own, necessary and sufficient conditions, all of which must be
satisfied for the condition -- the freedom listed in the bullet point -- to
be present. You can't simply wave your hand, say "you have the freedom to
redistribute copies to help your neighbor, except you have to restrict this
freedom according to US export law," and have the statement "you have the
freedom to redistribute copies to help your neighbor" unconditionally.
The only one I can see that is open to interpretation is this clause:
You should also have the freedom to make modifications and use them
privately in your own work or play, without even mentioning that
they exist.
The use of "should" rather than "must" causes it to appear that this is
optional, but strongly encouraged. This freedom would fall under
o The freedom to study how the program works, and adapt it to your
needs
However, if you have "the freedom to study how the program works," should
that not imply that your licensed ability to study how the program works --
which includes modifications as a matter of experimentation -- must not be
limited by notification to third parties? And if no third parties (the
licensor included) need be notified for private copies, how can private
modifications be required to be submitted to a third party?
I'll say it one last time, and then I'll try not argue this point with you
again. Without reading the entire document, you are not grasping what the
four kinds of freedom described are and
RE: Embedded systems OS/FS
Copyright protection subsists in original works of authoriship "fixed in any tangible medium of expression." It makes no difference that the tangible medium is an embedded device in hardware. The licensee of GPL code must honor the license terms regardless of whether he embeds the licensed code in a derivative work on a floppy diskette or on a chip. Once the final manufactured product (e.g., a printer) is distributed from the factory, the distributor or owner (e.g., the wholesaler or end customer) of that printer can sell it without triggering any further obligations to the original licensor. /Larry Rosen -Original Message- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED]] Sent: Tuesday, August 22, 2000 2:29 AM To: License-Discuss list Subject: Embedded systems OS/FS Trying to pick some brains before I get up on stage and make a fool of myself again (Intel Developers Conference). I'm told the audience will be a mix of both SW and HW developers, with the HW folks doing a mix of embedded devices and chip/circuit designs. Question I've got: how does software licensing, free/OS or otherwise, effect the hardware market. My read is that some licenses, notably the GNU L/GPL, may have their source availability requirements triggered by the physical distribution of media (HW) on which the software is embedded, etched, or otherwise fixed. The primary statuatory provisions of copyright (in the US) are of the reproduction, making derivative works, distributing, performing, and displaying of copies. The GPL specifically restricts itself to "copying, distribution and modification" (section 0). An instance in which this would matter: ACME Mfg. creates printers. They incorporate GPLd code 'gnuprint' into their product 'acmeprint', creating a derived work 'gnuacmeprint' of the two programs. In distributing the printers to wholesalers and eventually customers, does ACME trigger the GNU GPL's source distribution and relicensing requirements? To whom does the source obligation apply -- wholesalers, final customers, or both? My read is that yes, ACME does. The code to 'gnuacmeprint' must be licensed under the GPL, and the terms of 3(a) or 3(b) of the GNU GPL must be met. I'm not sure that the wholesalers would have a solid claim to code, the end customer certainly would. Other takers? -- Karsten M. Self [EMAIL PROTECTED] http://www.netcom.com/~kmself Evangelist, Opensales, Inc.http://www.opensales.org What part of "Gestalt" don't you understand? Debian GNU/Linux rocks! http://gestalt-system.sourceforge.net/K5: http://www.kuro5hin.org GPG fingerprint: F932 8B25 5FDD 2528 D595 DC61 3847 889F 55F2 B9B0
Re: Plan 9 license
On Tue, 22 Aug 2000, David Johnson wrote: RMS claims that the Artistic License is not free. His reasoning seems to be that it is vague. If vagueness disqualifies a license from being free, then people should know it right up front. It's not unfree (according to RMS) because it's vague per se. It's unfree (according to RMS) because it is too vague to *tell* whether it meets the requirements of a free license. Specifically, clause 5 says you can charge a "reasonable" fee for distributing the software (an undefined term), but not for the software itself. That could be interpreted to infringe the freedom 2. By suing Lucent over IP, you *have* given cause for revocation. This clause is better applied to other licenses, like the APSL, where a license can be revoked through no action on the user's part. It isn't just Lucent, it's *any* contributor to Plan 9, and the IP suit can be about anything whatever. If somebody abuses your GPLed software, and you sue, and the perpetrator turns out to be a Plan 9 contribute, *pip* goes your right to modify or distribute Plan 9. -- John Cowan [EMAIL PROTECTED] "[O]n the whole I'd rather make love than shoot guns [...]" --Eric Raymond
Artistic License (was Plan 9)
On Tue, 22 Aug 2000, Rick Moen wrote:
DFSG provision #1 ("Free Distribution") begins: "The license of a
Debian component may not restrict any party from selling or giving away
the software as a component of an aggregate software distribution
containing programs from several different sources."
Quickly grabbing my copy of the OSD to both see if it says the same
thing, and to keep this on topic :-) I find that yes, it essentially
says the same thing, so let's steer this on topic...
The AL clearly meets this point. So what's the controversy? Why is
there a comment in the OSD about this clause having to be put in to
satisfy the AL? Why is so obnoxious about allowing people to aggregate
AL software with non-AL software? If the aggregation stuff were not in
the definition, would not the AL still meet it?
Yes, I've heard the argument that one could write a 5 line proprietary
package, aggregate with AL software, then sell the lot for
"unreasonable" prices. But one can do the same with GPL software,
can't they? What am I missing here?
--
David Johnson
_
http://www.usermode.org
