Re: Linux System Engineer (100%) in Zurich
martin f krafft writes: also sprach W. Martin Borgert [EMAIL PROTECTED] [2008.11.25.2017 +0100]: I would very much appreciate, if Debian would not publish job offers that discriminate on the grounds of race, ethnic origin, disability, age, gender, sexual orientation or religion. Not only it is illegal in some countries, I find it highly inappropriate for our project. Thanks for your attention. Wolfgang, please stop this. Putting a maximum age into a job description is standard practice because a company does not want to invest time and money into a new employee for various reasons, be they simple age and thus time left to work for the company, absorptive capacity, or company culture. It may be standard practice in some places, and legal under the applicable laws in this case. However, Wolfgang is right. This kind of age limit has been illegal in the United States for 40 years. Wikipedia has a summary at http://en.wikipedia.org/wiki/Age_Discrimination_in_Employment_Act . I am rather surprised that the majority reaction to this is so different from the majority reaction to a poorly executed satire -- something which was also claimed to be discriminatory. Michael Poole -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Range Voting - the simpler better alternative to Condorcet voting
Barak A. Pearlmutter writes: The example you give is a perfect instance of the DH3 problem. You have a population of voters whose true preferences are 31x ACBXD 32x BCAXD 37x CBAXD It always amuses me how people pull out these examples close to Condorcet cycles as examples of strategy in Condorcet methods while ignoring the strategy issues in even simpler Range Voting elections that push it towards Approval-style voting. Suppose that voters have true preferences of: 60x A=60,B=40 40x B=60,A=40 It only takes six of the second group to vote B=99,A=0 to change the outcome, which is a major victory for the extreme supporters but a loss for honest moderates. Even using the full range for serious candidates does not solve this. In the five-candidate example above, if the most preferred candidate is given 99 points and the least preferred three candidates all get 0, the outcome is still quite susceptible to strategic scoring of the second place candidate. Michael Poole -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Range Voting - the simpler better alternative to Condorcet voting
Barak A. Pearlmutter writes: You also make the point 60x A=60,B=40 40x B=60,A=40 It only takes six of the second group to vote B=99,A=0 to change the outcome, which is a major victory for the extreme supporters but a loss for honest moderates. I think your point is that, with Range Voting, if *some* of the voters vote strategically while others don't, the ones who vote strategically carry extra weight in the election. (a) This is ***EXACTLY RIGHT***!!! In fact, that is the *definition* of strategic voting. If voting strategically didn't give a voter extra power to influence the election in their favour, we wouldn't call it strategic. (b) This is also true of Condorcet. Which six voters in that example have the ability to change the outcome of that Condorcet election by voting strategically rather than honestly? My point was not that strategy is possible. My point is that the obvious strategy under Range Voting is almost always optimal if you look only at one election, but that it has factionalizing effects in the longer run because it encourages artifically strong expression of opinions. Effective strategies for Condorcet methods are harder to judge, which encourages honest voting. This in turn encourages good positions from candidates, since they know more accurately what the electorate thinks is important. Unfortunately (Arrows theorem etc) there is no voting system for 2 candidates which avoids the possibility of strategic voting. And Condorcet is, unfortunately, in fact, particularly susceptible to strategic voting. In Condorcet, if all the voters vote strategically, you often get a really bad candidate winning; this is the DH3 pathology, see http://rangevoting.org/DH3.html for details. But with Range Voting, if everyone votes strategically, you get ... the honest Condorcet winner. Which isn't really so bad. In other words, if everyone in a Condorcet method has accurate knowledge of how other voters really feel *and* ignores how strategy compels others to vote, they can choose a poor strategy. This means that voters should vote honestly rather than use a poor strategy, but is really not a very interesting observation. There are similar problems in strategic Range Voting from approval cutoffs. Universal strategic Range Voting does not guarantee the honest Condorcet winner. As an addendum: in Range Voting voters are told to rate the candidates, with min/max for their least/most favoured candidates. So in your particular example with only two candidates, the actual votes would have been 60x A=99,B=0 40x A=0,B=99 which would leave no opportunity for strategic voting. But that's no great trick, since there are only two candidates. With 2 candidates, it is easy to make examples where strategic voting by Range Voting voters would make sense. Horse pocky. Where are voters told to do this? The first example on http://rangevoting.org/ talks about the Olympic scoring system. When was the last time an Olympic judge scored one competitor 0.0 and another 10.0 in one event? The How range voting works section makes no mention of it, either. I really should know better by now, but I am still surprised when Range Voting advocates throw out such bogus claims. Michael Poole -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Recompilation of ALL Debian packages ...
Russ Allbery writes: Source-code trojans are more dangerous because people fear binaries but think that if they've compiled it, it's fine, when the only real distinction is between code that's been audited and code that hasn't. Binaries built and uploaded by a maintainer who audits the upstream code are significantly safer than uncompiled source code uploaded by a maintainer who doesn't. This compares apples (a maintainer who audits the upstream code) to oranges (one who doesn't). Even given human error, the approach to auditing a source code package is reasonably well-understood. For binary packages, it is not, but it is clear that it is much more labor-intensive. Michael Poole -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Proposal: The DFSG do not require source code for data, including firmware
Nathanael Nerode writes: If you want to amend the DFSG to state 3. Source Code The program must include source code, and must allow distribution in source code as well as compiled form. However, this requirement does not apply to firmware, defined as insert your pet exemption here. I would strongly oppose such a change, but it would be a legitimate, reasonable GR (requiring 3:1 supermajority of course). Recent history -- in particular, GR 2006-001's winning option -- suggests that broad DFSG exemptions, when treated as clarifications or interpretations of the project, are not necessarily so clear-cut about requiring a 3:1 supermajority. Michael Poole -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: irc.debian.org
Paul Johnson writes: By design, IRC encourages people to do truly obnoxious things, like spamming the channel to announce they're going away, or indicating their status with nicknames (which also spams the channel). You also get spammed on IRC whenever someone joins or leaves a channel. Jabber prevents this by providing a real presence system. Wrong. It is (bad) social habits that cause people to notify channels when they go away, not the design of IRC. (In many channels, that kind of antisocial spam is grounds for a ban.) IRC's AWAY, WHO, WHOIS, and related messages provide adequate presence notification. Michael Poole -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Retailing
Kieran Lloyd writes: Dear Sirs, I am considering selling some home made PC's on Ebay, the thing is I want to sell these pre-installed with Debian Linux. Would Debian have any argument with this? I will obviously be advertising that the Pc's have Debian installed however I will not be charging for it I would only be charging a mark up on my hardware. Please advise if this would be acceptable. It seems odd to me that no one has yet pointed out the congruences between this request and that of Nexenta. For example, GRUB and Linux are both licensed under the GPL. Both would be included with these retail systems and would be written to locate and call functions within the BIOS; that is, GRUB and Linux would be dynamically linked against the (presumably non-free) BIOS. What distinguishes this case from the GNU/Solaris CDDL-licensed standard library case? Has it simply gone unnoticed by those who campaign so hard to kill competition? Michael Poole -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Pledge To Killfile a person
Andrew Suffield writes: So your claim is that you can never object to people for working against consensus because doing so would be working against consensus. Well, that appears to deny you from being allowed to make that point, so I think your argument is self-defeating. I have no objection to saying that people are working against consensus -- provided it is done in a way that does not further impair that consensus. Your approach seemed much more likely to annoy and mislead people than to help identify where they agree or disagree. But I give up. I think it will be less hassle to killfile you rather than try to reform you to minimal sociability. Michael Poole -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Pledge To Killfile Andrew Suffield
Andrew Suffield writes: On Tue, Aug 09, 2005 at 11:09:16PM +0100, Andrew Suffield wrote: My response is simply this: it's lies. I challenge anybody who thinks otherwise to present evidence. So far (three days) we've had one person try, and give up after I explained every case. I think that says a lot for the accuracy of the accusations. It says a lot more about how much you regularly misrepresent plain writing to put your spin on issues. You think you rebutted my arguments; I think you actually illustrated that I was correct from the start. The real reason I gave up is that it is clear that neither of us is convincing the other. Descending to your flawed level of rhetoric, it is also telling that nobody else has stepped up to argue that your posts were acceptable. Michael Poole -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Pledge To Killfile Andrew Suffield
Anibal Monsalve Salazar writes: On Fri, Aug 12, 2005 at 04:32:52PM -0400, Michael Poole wrote: Andrew Suffield writes: On Tue, Aug 09, 2005 at 11:09:16PM +0100, Andrew Suffield wrote: My response is simply this: it's lies. I challenge anybody who thinks otherwise to present evidence. So far (three days) we've had one person try, and give up after I explained every case. I think that says a lot for the accuracy of the accusations. It says a lot more about how much you regularly misrepresent plain writing to put your spin on issues. You think you rebutted my arguments; I think you actually illustrated that I was correct from the start. The real reason I gave up is that it is clear that neither of us is convincing the other. Descending to your flawed level of rhetoric, What are the flaws? The flaw is that of projection: assuming that silence means everyone agrees with you. it is also telling that nobody else has stepped up to argue that your posts were acceptable. Why is that? Is it because asuffield is doing a good job defending himself? One could just as easily assume it is because asuffield has made such a spectacle of himself that no one else thinks he has a leg to stand on. Since there is no clear input, the conclusion is prone to observational bias. Michael Poole -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Pledge To Killfile a person
Sigh. I wasn't aware that common courtesy was so rare as to require explanation at length. Andrew Suffield writes: On Wed, Aug 10, 2005 at 11:08:05AM -0400, Michael Poole wrote: Andrew Suffield writes: My response is simply this: it's lies. I challenge anybody who thinks otherwise to present evidence. I sign almost all my outgoing mails; this should be easy, if it were true. Find mails from me that are little more than provocations, put-downs, and trolls. Not ones where people have interpreted it that way and I've either told them they're wrong or ignored them. Ones where it's actually true. Post references to this thread. See how many you *actually* get, out of the number of mails I send. You asked, and so a little bit of Googling produces these: http://lists.debian.org/debian-devel/2004/06/msg01598.html Looks like a perfectly justified response to me. I don't see how that could be classified as 'provocation' or 'troll', because in no sense did it encourage more discussion - it was quite clearly a statement that he was being ignored because he was just trying to start an argument. I suppose you could claim it was a 'put-down', but I claim it is a factually accurate description of the parent mail and I challenge anybody to prove otherwise. This is an example of one of the significant limitations (perhaps good, perhaps not) in Debian's current culture: A lot of people think rudeness is excused -- and not just excusable -- when it saves them future effort. The parent mail is not clearly a troll to me, and I think it is preposterous to assume something is a troll until proven otherwise. http://lists.debian.org/debian-vote/2004/06/msg00166.html I can see nothing in this mail that could be even remotely like that. Explain your claim. It was counterproductive in that it did not advance any discussion. It was hypocritical in that it accused unspecified people of being trolls, while complaining that *they* were resisting efforts to build consensus. Insults never build consensus: even when they drive away individuals who disagree, they also splinter the consensus. http://lists.debian.org/debian-legal/2005/05/msg00036.html And again. It's an appeal to accept your authority on whether someone is a troll. In the absense of supporting evidence, it's a put-down and simple provocation. http://lists.debian.org/debian-project/2005/04/msg00248.html A restatement of what other people had previously said, and I still see no way you can classify this as little more than provocations, put-downs, and trolls. An explanation of *why* telling users that GFDL docs moved to non-free is not particularly important would be useful, since your message is otherwise an out-of-hand dismissal of the idea. Likewise, insulting the survey without even one example is again put-down and provocation. Perhaps constructive criticism is an American peculiarity. perhaps http://lists.debian.org/debian-vote/2005/03/msg00378.html (which I would call devoid of useful information but heavy in insults) Insults? WTF? Precise quoting and explanations please, I see none here. One insult is the reference to a small group running around advocating knowingly putting non-free stuff into main, which is a serious charge. You were right that it would be absurd, but without support, it looks suspiciously like a strawman. Another implied insult is the distinction between the frequent posters to debian-legal: You are there because you send lots of short email, and others who are there are in your killfile. The only reason I see to mention that is to sugggest that they are not worth counting. Michael Poole -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Pledge To Killfile a person
Andrew Suffield writes: On Thu, Aug 11, 2005 at 09:50:12AM -0400, Michael Poole wrote: Sigh. I wasn't aware that common courtesy was so rare as to require explanation at length. When you start making accusations, you are obliged to back them up with explanations. Otherwise you are merely denying any right to respond, and what was that about common courtesy? You asked for a list of posts. I provided a list of posts. You wanted explanations. I provided them, and remarked as above because some of your specific questions suggested an antisocial perspective. (For example, thinking that a message reading Troll. is not provocative and moving the goal-posts on whether it is a put-down.) Besides which, none of this appears to have anything to do with courtesy. Most of it is you misrepresenting my statements. http://lists.debian.org/debian-devel/2004/06/msg01598.html Looks like a perfectly justified response to me. I don't see how that could be classified as 'provocation' or 'troll', because in no sense did it encourage more discussion - it was quite clearly a statement that he was being ignored because he was just trying to start an argument. I suppose you could claim it was a 'put-down', but I claim it is a factually accurate description of the parent mail and I challenge anybody to prove otherwise. This is an example of one of the significant limitations (perhaps good, perhaps not) in Debian's current culture: A lot of people think rudeness is excused -- and not just excusable -- when it saves them future effort. Then I don't think you've got any grounds to accuse me of it specifically, and not any of the others. Regardless of whether or not it happens to be my belief (it isn't). How else should I consider a mail that simply declares Troll.? Do you think it is not rude? Or was the point of the brevity something besides saving yourself the effort of justifying the judgment? Whether you consciously had that motivation or not, it comes up in most of the semiannual Women in Debian flame wars. The parent mail is not clearly a troll to me, and I think it is preposterous to assume something is a troll until proven otherwise. The parent mail is an instance of argumentum ad hominem and a claim to authority, combined with a straw man, on a subject which is tangential to the one under discussion, which is written in a clearly antagonistic manner and adds no new information or valid arguments. If that's not a troll, then what is? When you held yourself up as knowing the security mechanisms used by every CA, it was entirely appropriate to puncture that impression. If you insist that martin f krafft's post was a troll, your post which he responded to was equally a troll. And why are you attacking me instead? I am attacking you because you asked me to explain my selection of posts to illustrate when you send things that are little more than provocations, put-downs and trolls. If you did not want me to elaborate on why your posts qualify, you should not have asked. http://lists.debian.org/debian-vote/2004/06/msg00166.html I can see nothing in this mail that could be even remotely like that. Explain your claim. It was counterproductive in that it did not advance any discussion. Interesting definition. I have not encoutered this one before. It is an example of counterproductive, not a definition. Inflammatory remarks which do not also illuminate are little more than provocations, put-downs and trolls, and such remarks are only one way to be counterproductive in one's discourse. I assumed that their counterproductive nature was why you chose that description when asking people to find objectionable posts of yours. It was hypocritical in that it accused unspecified people of being trolls, while complaining that *they* were resisting efforts to build consensus. It provided a simple mechanism by which people could judge for themselves, which was the whole point. I find that far more honourable than listing names and accusations. I do not see how it is hypocritical. Insults never build consensus: even when they drive away individuals who disagree, they also splinter the consensus. This statement appears disconnected from the rest of the paragraph; if it was meant to be a point, please restate it. It was an elaboration of why the complaint about consensus-building was hypocritical: because the post itself worked against consensus. Michael Poole -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Pledge To Killfile a person
MJ Ray writes: Michael Poole [EMAIL PROTECTED] wrote: [...] How else should I consider a mail that simply declares Troll.? Do you think it is not rude? Or was the point of the brevity something besides saving yourself the effort of justifying the judgment? I thought that pro-active anti-troll interventions that might help forestall vulnerable groups being harassed were suggested as Best Practice by several debian contributors? I would not be surprised. As I said before, I think it is an aspect of a wider behavior pattern -- one that sees frequent objections. Personally, I consider them obnoxious and counter-productive, but few were willing to support that position last time I told it to a debian list. I'm surprised to see such strong objection now. Is it because troll alerts are daft or because of who posted it? They're daft in general, but such an extreme is very rare, and I only cited it because it was topical. When there's a reasoned explanation of why something is unhelpful, I look more forgivingly on it. More generally, email argument over what is improper tends not to get far, and nobody's behavior had crossed my personal threshold for starting a meta-argument. I may not have agreed with your position earlier for a number of reasons; the time I want to spend contributing to Debian discussions is a function of how busy I am with other activities. Michael Poole -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Pledge To Killfile a person
Andrew Suffield writes: My response is simply this: it's lies. I challenge anybody who thinks otherwise to present evidence. I sign almost all my outgoing mails; this should be easy, if it were true. Find mails from me that are little more than provocations, put-downs, and trolls. Not ones where people have interpreted it that way and I've either told them they're wrong or ignored them. Ones where it's actually true. Post references to this thread. See how many you *actually* get, out of the number of mails I send. You asked, and so a little bit of Googling produces these: http://lists.debian.org/debian-devel/2004/06/msg01598.html http://lists.debian.org/debian-vote/2004/06/msg00166.html http://lists.debian.org/debian-legal/2005/05/msg00036.html http://lists.debian.org/debian-project/2005/04/msg00248.html perhaps http://lists.debian.org/debian-vote/2005/03/msg00378.html (which I would call devoid of useful information but heavy in insults) Lately, I try to not make note of people behaving badly; there are too many instances for worrying over it to be healthy, and most people balance it out over time. Even though I personally tend to ignore Andrew Suffield, I think an organized effort to killfile _anyone_ is a misguided application of social pressure. If the alleged misbehavior is not so systematic that everyone can see it, it is not bad enough to warrant ostracism. Michael Poole -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Poll results: User views on the FDL issue
Marty writes: Invariant sections are perfect example of a restriction that enhances the rights of the author (copyright holder) at the expense of the end user, but does so in a way that promotes sharing of information as opposed to hoarding. This is a rather curious contention. How do invariant sections (by themselves) promote sharing of information? The FSF largely uses them to preach free software, but others might use them to preach a disagreeable agenda, or one that is illegal to promote in certain jurisdictions. Users in those jurisdictions would be limited in how they can use or distribute the work, simply because the author injected a diatribe that does not pertain to the main body of the work[1]. It is rather short-sighted to encourage a significant limitation in freedom because no author has yet abused that limitation. Michael Poole [1]- To a first approximation, under the GFDL, invariant sections must be off-topic for the work as a whole. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: non-free but distributable packages and kernel firmware
Sven Luther writes: Hello, In the wake of the discussion about solving the murky non-free binary firmware blobs in our linux kernel issue, there was a proposition to split out the firmware blobs from the main drivers, and distribute them in a separate but canonical place from ftp.kernel.org. As thus i was wondering if, together with the volatile effort, it would not be time for us to split the non-free archive into two parts, namely : 1) non-free-but-freely-distributable Which would hold all the files which are freely distributable, but fail one of the freely modificable criterias of the DFSG. 2) rest of non-free Which would include all but the ones in the first part, and impose some furter restriction on distribution. This has the strong smell of ranking some DFSG criteria above others in importance. If you want this kind of distinction, I think a less discriminatory way would be to flag (internally or on a central web site somewhere) each package in non-free according to which parts of DFSG it fails. Michael Poole -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: documentation x executable code
Peter Vandenabeele writes: On Thu, Jan 06, 2005 at 05:31:31AM -0500, Glenn Maynard wrote: On Thu, Jan 06, 2005 at 10:56:57AM +0100, Peter Vandenabeele wrote: An interesting consequence of this proposal is that a Copy-Exact of the GPL License could not longer go into main (as it is essentially one large invariant section. I quote from GPL: Everyone is permitted to copy and distribute verbatim copies of this license document, but changing it is not allowed.) This is false. I suggest you read the rest of the thread to see why, or any of the other three or four hundred times people have tried to convince us that Free Software is hopeless and we should just give up by claiming that license documents can't go in main. I am fairly new to this list, but I have followed this thread for some time now. I did not understand why a document with invariant sections cannot be part of Free/main (in the Debian context) and the GPL license which states that it only allows verbatim copies can be. An invariant section is an integral part of the documentation; by the GFDL's definition, it is otherwise irrelevant content. The license is legally required metadata: The copyright owner provides a particular license to users, and those users must know exactly what that license is. No redistributor may alter that license's text and pass it off as applying to the original software. Michael Poole
Re: Theo de Raadt On Firmware Activism
Andrew Pollock writes: On Wed, Nov 03, 2004 at 10:30:10PM +, Martin Michlmayr - Debian Project Leader wrote: http://kerneltrap.org/node/view/4118 Kudos to Theo for OpenBSD getting out and poking the vendors. My concern is that for all their effort, and potential flow on benefits to Linux, it won't be considered good enough for Debian because of the current stance on firmware, and source code to it... Where did we get up to with that anyway, binary blobs are out, end of story? Binary-only firmware must go into non-free: users (and Debian itself) do not get the freedoms identified in the DFSG and cannot correct bugs or deficiencies in the firmware. Firmware that includes source but which cannot be built within Debian might be able to go into contrib; I am not aware of any firmware like that. There was a recent debate on debian-legal about the status of drivers for devices that can or must use firmware, as opposed to the firmware itself; this is relevant when the driver provides a mechanism to load external firmware. Some argue that these drivers can go into main on the basis of the abstract interface to the device; some argue that the drivers must go into contrib on the basis that the device is not functional without the firmware, and the driver is not functional without the device. Michael Poole
Re: Patent clauses in licenses
Andrew Suffield writes: On Thu, Sep 23, 2004 at 08:17:27AM -0400, Glenn Maynard wrote: On Thu, Sep 23, 2004 at 12:25:21PM +0100, Andrew Suffield wrote: And again, I don't believe the freedom to prosecute with patent accusations is an important freedom to protect, any more than freedom to take my software proprietary. I think it's valid and legitimate for a free license to restrict this freedom. Same old bogus comparison; you never *had* the freedom to take the software proprietary, so you can't protect it. You *did* have the freedom to prosecute with patent accusations. By that line of reasoning, you never had the freedom to use my software while at the same time alleging that it violates your patents You can't combine things of different type like that. Computer programs, by their nature, combine both copyrightable and patentable elements. Neither Glenn nor patent lawsuit termination clauses were the first to combine them. Michael Poole
Re: Patent clauses in licenses
MJ Ray writes: On 2004-09-24 13:37:42 +0100 Michael Poole [EMAIL PROTECTED] wrote: Computer programs, by their nature, combine both copyrightable and patentable elements. Mathematical relationships are discoveries, not inventions. Similarly algorithms: the right one existed, but you just didn't know it. Expressions of algorithms are covered by copyright. There is not necessarily any construction or fabrication, either real or imagined. So computer programs, by their nature, contain no patentable elements under English law. Your (claimed) law is not everyone's law. It's silly to ignore the case of the US (and apparently Australia[1], most of Asia[2, pg 25], and others), since a significant number of mirrors, developers, and users are there. A number of other courts, including the German Supreme Court[2, pg 13], have held that computer programs can infringe patents. [1]- http://old.linux.org.au/papers/no-patents.html [2]- http://www.slwk.com/CM/PhoneSeminars/asia.pdf Michael Poole
Re: Patent clauses in licenses
MJ Ray writes: On 2004-09-24 15:49:12 +0100 Michael Poole [EMAIL PROTECTED] wrote: [...] patents covering programs is not a problem specific to the USA. [...] Indeed, but that's neither global nor natural. The post I first replied to seemed to be claiming it was natural that programs are patentable. You repeatedly making a different point (many countries have this bug) doesn't contradict my point (programs not naturally patentable). You seem to think I don't read parts snipped (I do), but I think you're not understanding what you're quoting. Patents themselves are not natural. They are defined by a particular legal system. (Copy rights are not natural, either, but they are treated more consistently than are patents.) When I said that computer programs include patentable elements in my earlier mail I meant simply that people can and have received patents that cover the operation of certain computer programs. There is a European patent (EP0482154) on the IDEA cipher. That, and its US equivalent, prevent a lot of people from writing computer programs that use IDEA. The question is not whether a particular program can be patented, but whether a particular program (whether by simply existing or through its execution) can infringe a patent. See also http://www.cl.cam.ac.uk/~mgk25/stallman-patents.html . Maybe the European IDEA patent has been ruled unenforceable for computer programs. Maybe similar cases in Japan, South Korea, etc, have that effect in those countries. Maybe your point is not related to the question of whether a program can be made non-free by a patent. However, I have not seen anything to support those possibilities. The reason I mention non-freeness is because that is what this thread is about: When someone asserts a patent claim, which -- if any -- granted rights may be terminated by a free license? The broadest termination I have seen supported[1] is termination of both copy and patent grants to a person who claims in court that the program infringes a patent. Your argument seems to be that such a court claim either is or should be legal nonsense; perhaps that is the case in the UK, but the European patent on IDEA and the German Supreme Court case cited earlier suggest that the rest of Europe differs. The US and Asia certainly have different laws. To claim that Debian should make policy based on the most permissive laws is contrary to both prior practice and common sense; for example, -legal requires new licenses to specifically license users to execute or use software, since not all copyright regimes have the US's exemption for that. [1]- I intentionally set aside section 4 of GPLv2 and the non-Debian support for termination when one sues the software author(s). Michael Poole
Re: Patent clauses in licenses
MJ Ray writes: However, do we agree that a free copyright licence may not terminate when someone asserts a copyright claim? Suppose Alice publishes Asoft under the GPL. Bob files a lawsuit claiming that Asoft infringes his copyright on his proprietary program Bsoft. Does that breach section 4 of the GPL, terminating his license to modify, copy, and so forth, the remainder of Asoft? I think it does, although I am open to arguments as to why Bob would retain a license to Asoft. Some have suggsted that Bob gets a new license every time he gets a copy of Asoft from someone else. I disagree with that because that interpretation opens a significant loophole in the GPL's protection and because getting a new license would not remedy previous instances of copyright infringement. Michael Poole
Re: Patent clauses in licenses
MJ Ray writes: On 2004-09-24 14:23:34 +0100 Michael Poole [EMAIL PROTECTED] wrote: Your (claimed) law is not everyone's law. It's silly to ignore the case of the US [...] Similarly, it's silly to assert that US law is everyone's law and that computer programs are naturally patentable. It's silly that US law permits patenting of discoveries. I'd not be surprised if the US grants patents on looking at particular stars. You must be smoking the same thing(s) that Andrew Suffield is. The rest of my mail (which you snipped, saving a whopping four lines of text) pointed out that patents covering programs is not a problem specific to the USA. Ignoring evidence you do not like is neither convincing nor helpful. Michael Poole
Re: Patent clauses in licenses
MJ Ray writes: On 2004-09-21 10:21:58 +0100 Glenn Maynard [EMAIL PROTECTED] wrote: On Mon, Sep 20, 2004 at 03:02:49PM +0100, MJ Ray wrote: It is bad patent law which favours patent owners. It is fine to use copyright licences to correct copyright law, but using copyright licences to correct non-copyright law - be it patent law, gun control law or nuclear technology laws - is not. Why? What freedoms does this protect? Respectively: the freedom to prosecute with and defend yourself against patent accusations; the freedom to bear arms; and the freedom to use nuclear technology. Of course, not all jurisdictions allow those freedoms, but that's determined by laws, not by copyright licences. This argument over-simplifies the case: No putatively free license has included a waiver of patent claims, just termination of patent and/or copyright license if you assert those claims. The interesting case (that some argue is free) is when your license terminates when you claim the software infringes your patent. This is a self-protection measure for the software; while someone might try to construct similar anti-gun or anti-nuclear license terms, they would have to be so narrow as to be both silly and pointless. Why should copyright not be used to protect free software from patent abuse, just as it's used to protect against software hoarding? Mainly because most possible uses have unpleasant side-effects in some cases. Software hoarding is a description of a copyright-based problem, if you are referring to rms's Why Software Should Be Free. It seems just to use copyright to solve it. Why should we use copyright against patent law, instead of encouraging patent-afflicted developers to find ways to use patenting against itself? Is there a way to use patenting against itself? For all the times I have seen someone suggest that, I have yet to see a good way to do that. The biggest problem is that you can infringe a patent without knowing it, and trying to not infringe patents is impractical. (Most people cannot afford to hire a patent attorney to review their use of free software against extant patents. Reviewing patents yourself opens the possibility of willful infringement if you are wrong about what is covered. In both cases, valid patents may be pending but not public when you do your review.) Setting aside the question of practicality, I know of three ways to reduce the chance that I will successfully be sued for infringing someone's patent by writing or using free software: (A) establish prior art to refute possible patents; (B) hold for myself broad patents to use defensively; or (C) change the law so fewer issued patents cover software. (A) is impractical because of the number of ideas that are awarded patents, and the relatively small differences that suffice to make them non-obvoius. The one-year US grace period on prior art also makes this hard. (B) is impractical because of the cost. It is undesirable because it encourages large numbers of broad patents that may not even be practiced by the inventor. (C) may be practical, and people are working to do that. If they succeed, most of the license termination clauses will have little or no legal effect. Similarly, why should copyright not be used to protect free software use from gun abuse and nuclear technology abuse? No one has tried. We have so far considered terms of software that people want to include in Debian (including how to fix non-free licenses), so I do not see good reason to debate what would make such clauses free. Michael Poole
Re: Patent clauses in licenses
Andrew Suffield writes: On Sun, Sep 19, 2004 at 02:41:03PM +0100, Martin Michlmayr wrote: * Matthew Garrett [EMAIL PROTECTED] [2004-09-17 10:05]: The GPL does much the same. If someone distributes GPLed software without complying with section 3 (which gives you various ways in which you have to make source code available to the recipient), then they lose the right to use that GPLed software. We have various licenses that terminate if you do something wrong - we've just come to the conclusion that it's acceptable that people not be allowed to do that thing. In the past, we've accepted various compromises on freedom because they help free software. I agree with this reasoning and think that we should treat at least Any patent action against the licensor connected to the licensed work as free. I'd like to hear more possible scenarios what Any patent action against the licensor might mean in reality, such as Nathanael's IBM example. I think such possible scenarios/examples are a good way to think about the implications of these clauses. Here's a scenario for you: Company A releases a piece of software that includes this clause in its license. Company B releases a modified version of this software, that includes an extra feature. Company A has no interest or use in the piece of software created by company B; furthermore it desires to eliminate this version. Company A sues company B alleging that the extra feature in the modified version infringes some of its patents. Company A no longer has a license to the modified version, which it didn't want anyway, so it is not concerned about this. Company B cannot make counterclaims from its defensive patent portfolio, because that would invoke the termination clause and kill its modified version. Company B has no practical defence against this lawsuit, so the modified version is killed. They have been effectively trapped in a double-bind. I just pulled that one out of the air. There are countless more like it. All you are accomplishing is to permit copyright holders more control over their software; this cannot be a good thing. Trying to game the legal system *doesn't work*. Company B's defensive claims also affect all other users of the original software -- now that they attempt to enforce their patent rights, no other users can assume themselves to be safe. Why is Company B's self-defense more important to free software than unrelated users? This is inevitable from first principles; significant arbitrary restrictions are non-free. You will always be able to find ways to abuse them to gain arbitrary degrees of control over the software. One could claim that allowing others to access a program over the network includes sufficient transfer of copyrighted material to trigger section 3 of the GPL. Debian has rejected explicit external deployment clauses in the past, but accepts the GPL despite this possibility. I just pulled that one out of the air. There are countless more like it. Making up corner cases is not particularly useful. Michael Poole
Re: Patent clauses in licenses
John Hasler writes: Michael Poole writes: Company B's defensive claims also affect all other users of the original software -- now that they attempt to enforce their patent rights, no other users can assume themselves to be safe. Why do you assume that company B's claims must have to do with the original software, or even with software at all? Martin's original mail had the qualifier connected to the licensed work in one place and not the other; my mail was addressed case where I think there is significant disagreement on -legal: the case when that kind of qualifier is used. I agree with Andrew (and, from what I can tell, most of -legal) that license termination for a patent lawsuit unrelated to the licensed software is non-free. I cannot find offhand any license that is quite so broad. I suspect -- but have no strong opinion yet -- that it is also non-free if the termination is limited to software patents against the software's author(s). Some licenses that do that are the Apple Public Source License and IBM Public License. A more generally productive (and acceptable) way to fix that problem is to lobby against software patents. Michael Poole
Re: On the uselessness of Debian trademarks.
Nathanael Nerode writes: MJ Ray wrote: On 2004-05-07 14:20:37 +0100 Stephen Frost [EMAIL PROTECTED] wrote: Uh, or they use the Debian trademark for something that's not Debian at all.. That's not necessairly claiming it as backing or endorsement from Debian. If it's software, it seems illegal anyway. But it's illegal because it's a trademark violation. Not for any other reason! Unless I'm very much mistaken. If I am someone will have to point out the laws under which it's illegal. Note that a trademark doesn't have to be registered to exist. Not being a lawyer, I'm not sure how widely you can claim trademark rights on a logo. I know that for plain text, you can safely use words for different companies in different fields (see, e.g. Apple Computer vs Apple Records and their dispute). That is one reason that the USPTO asks for field of use when registering a copyright. A trademark does not have to be registered to get common law protection, but protections for an unregistered trademark are almost useless: You will not get costs and attorney's fees in a suit for common law trademark infringement, only in a suit for Lanham Act (registered) trademark infringement. Michael
Re: On the uselessness of Debian trademarks.
Stephen Frost writes: * MJ Ray ([EMAIL PROTECTED]) wrote: Not only is it not very Debian, but accurate use of the Debian mark to refer to our Debian doesn't look like something we can stop with trademarking in the UK: I don't get it. Doesn't this mean, also, that in the UK people *could* sell shirts with the Coke logo on them? In which case it would seem to me that the reasons above for having a trademark in the UK would be perfectly legit and very reasonable and enforceable, and their intended use? I doubt it -- selling shirts would be a commercial purpose outside identifying goods or services as those of the proprietor or a licensee. I do not think it would even be safe to sell shirts that say Buy your Coke at Joe's Convenience Mart. Since you see income from selling the shirts, it could be argued as using the mark in commerce. Michael Poole