At 3:51 PM -0400 on 7/24/99, Alain Farmer wrote:
>>Alain: OK, but what kind of decision-making process are
>>you envisionning? A voting CGI in order to establish
>>the opinion of the majority and act accordingly? This
>>could lead to Majority-Rule. Bottom-line is that we are
>>going to have to discuss the (political) issue of
>>decision-making, as part of the Collaboration section
>>of our group, before, during and after we decide which
>>licencing terms to adopt.
>
>Anthony: I'm afraid that if we vote, it'll be arbitrary. And if we make
>arbitrary decisions, we will not be liked, to say the least.
>
>Alain: For the record, I have always been for consensus rather than
>voting. Voting can be quite divisive and sooner-or-later alienates
>those that are not in the majority.
But let me point out that in a large group of people, unamity
_will_not_happen_ Never.
And the ideas thrown around with multiple copyright holders _is_ a large
group of people.
>
>Anthony: And if we have enemies, we're more likely to answer in court.
>
>Alain: I suppose youíre right, even if their claims are un-founded.
Correct. Anyone can sue anyone for anything. Of course, there has to be
some foundation for the claim -- or I just say, as the defendent, when
sued: "Motion to dismiss" or "motion for summary judgement for the defense."
>We
>have to go through the court process to determine this. Some larger
>fish might even drag a competitor through one or more of these court
>hassles, in order to immobilize or annihilate the smaller fish that,
>despite its small size, were having an impact.
Exactly. That's whats done with frivelous software patents. Only a large
corporation can afford $1.5 million to get it thrown out.
>
>>Alain: Unless he decides that he wants to contribute it
>>and we decide to integrate it into the Standard
>>Distribution. From then on, it is the OC licence
>>prevails. The author cannot change his mind afterwards
>>and retract his contribution.
>
>Anthony: We'd need a signed agreement from anyone who submits a patch.
>That'd get tedious...
>
>Alain: Not necessarily a SIGNED agreement. Software licences are
>adhered to, more often than not, by clicking on the ìagreeî button of
>the electronic version of the softwareís licence agreement.
The extent of the enforcability of the licence is usaually that you can't
copy the software -- that'd be copyright violation.
But, yes, we could have a statement in the licence about contributions to
OpenCard. The Perl Artistic, for example, makes it clear how such
contributions may be included.
<snip>
>
>Anthony: I know the FSF does it, but I'd like to get around such.
>
>Alain: What is the FSF ?
Free Software Foundation -- the people behind GNU and the GPL.
>
>>Alain: If you use his source, then you are infringing
>>his copyright but, on the other hand, if you are merely
>>engineering something similar, then that is perfectly
>>legitimate. If it wasnÌt, we would be in big trouble
>>with Apple, eh!
>
>Anthony: The reason we're not in trouble is because we've never seen
>HyperCard source. If we had, and Apple cared, they could not doubt make
>us answer in court.
>
>Alain: They would have to prove that we derived ours from theirs,
>whether we had seen their source or not. No access to their source
>makes this determination a no-brainer. How could we have derived ours
>from theirs if we NEVER got even a glimpse of their source? On the
>other hand, if we had seen their source, then this determination of
>non-infringement would have to be verified. But we are nonetheless
>rightfully allowed to use the same idea because only the EXPRESSION of
>an idea is protected by copyright.
Except that we'd be fighting an uphill battle. As long as we've never seen
the source, Apple would be fighting uphill.
>
>Anthony: I don't know who'd win, but it'd be a fight.
>
>Alain: I would NOT like to compete with such a large corporation in
>this regard.
Nor would I.
>
>Anthony: But a short patch is so easy to remember. The author of the
>patch, having become our enemy, if sufficiently made -- and if we then
>patched it with
>anything like his code -- might claim infrigement. And we would have to
>defend ourselves. In court.
>
>Alain: Experts would have a look at it before going to court to see if
>there is some reasonable basis for winning their infringement claim.
Not so. Anyone can sue anyone for any reason -- all they need is a
fairly-small filing fee.
>Large corporations might want to pursue litigation despite the weakness
>of their claim, in order to wear us down, but this nefarious strategy
>is not likely to be adopted by a small developer of relatively moderate
>means.
But think of all the threats we'd get. Do we want to sort out the threats
of real litigation from the fake ones? How could we tell? Someone would do
it.
In all likelyhood, we, upon recieving a legal threat, not wishing to go to
court, would cave to their demands.
>
>Anthony: You must understand that it's not like a novel. It's a small
>snippet, maybe a few lines long. And once we read it, we would be
>unable to implement it ourselves, because it would probably be
>infringement.
>
>Alain: Ideas are not copyright-able, as I stated above.
There are few ways to write certain things. And if the person spends hours
fixing a bug, sends us the patch, we read it, refuse it, and then patch it
ourselves, we have benifited not from his idea, but from his actual work:
It is what showed where the bug is (the major part of fixing most small
bugs!). It is what showed how to fix it.
The idea was "stop it from crashing when I do so and so." The work is the
patch, the code.
Or maybe the idea is "don't corrupt the heap by over-running array bounds."
But the work is finding it and fixing it.
The patcher could claim, quite correctly, that we benifited not from his
idea, but from his work, and that our work relied heavily on his work of
finding the bug -- that our work is a derived work of his.
>
>Alain: Youíre right. Even if we delegate the re-ingineering of a
>refused patch to a third-party who has not seen the original source
>code, there is no guarantee that this third-party did not indirectly
>benefit from our knowledge of the source, or so the litigators could
>argue. Thus, if someone is determined to sue us, they will. There is no
>use speculating endlessly on this. There is nothing we can do about it,
>except : establish clear licencing terms, procedures, rules and so on.
Agreed. That's why I support a plain-ol Artistic licence, with one
copyright holder.
>
>Anthony: (...) Or if this happened to be someone trying to get voting
>rights on licencing decisions, we're sued.
>
>Alain: My first remark is that this is one of the reasons that voting
>is a sub-optimal strategy for decision-making. Secondly, I donít see
>why someone could expect to have full voting rights once he has
>contributed a handful of bytes.
See below.
>
>Anthony: When people contribute, they will expect to get the right to
>vote (as suggested above, by you).
>
>Alain: That was not a suggestion. It was a question.
>
>Anthony: And if they don't get it, they will be mad.
>
>Alain: There is much more to participation than voting. I suggest that
>we keep the latter only as a non-binding opinion-poll or, in extreme
>cases, referendums.
>
>Anthony: And if it was significant work -- and especially if we make
>some money off of OC -- he may sue.
>
>Alain: They would sue us because we didnít give them voting rights? Is
>there a precedent for this?
If:
1) a person is told by us that a patch will recieve them a partial
copyright in OpenCard and thus voting rights (which seems to be the
suggestion with multiple copyright holders, and the voting rights the
logical implication of it, being the only real way to make decisions under
such a situation);
2) A person submits a patch; and
3) We accept that patch -- or use something which is legally a
derived work, as I argue a re-implementation of a small thing would be,
Then:
We have entered into a contract with that person to give him a
partial copyright and voting rights.
If we don't, we've committed breach of contract, and there is _plenty_ of
precedent!
If we tell them that a patch gets the rights, and then don't give them
rights, we've committed breach of contract.
>
>Anthony: I want as little possibility of lawsuits as possible.
>
>Alain: Agreed!
>
>Anthony: Hell, we've already got to deal with the people who have their
>nice 20-year patents on algorithms. (...). These people claim they own
>a mathematical algorithm, and have exclusive rights to its use,
>regardless of independant discovery, for twenty years.
>
>Alain: The shoe is on the other foot, so to speak, eh! Wouldnít you do
>exactly the same? Make as much profit as possible by charging for the
>rights to use your invention? Let the inventor profit from his
>invention for several years before some copy-cats come along?
I believe there is nothing wrong with the patents, except:
a) They last to long. The idea of a patent is to protect the inventor,
and to not allow someone else to quickly steal his idea before he
can market it. In software, this is not 20 years. It's a year or
two, at most.
b) The patent office has failed to do its duty of accepting ONLY things
which are novel and non-obvious. It also routinly fails to do
proper searches for prior art. It instead decides to let the courts
settly it -- where it costs $1.5 million.
>
>Anthony: The patent office does not do its job, and instead expects the
>courts to do it for them. Unfortunately, getting a frivolous patent
>thrown out costs around $1.5 million.
>
>Alain: The patenting of algorithms is indeed problematic and
>controversial. And the patenting of new living organisms is also quite
>controversial, but the latter is quite off-topic, so we wonít discuss
>it on this list, save to say that I recently was told about some
>research being done on biological computers no longer based on silicon
>that will be 100,000 times faster than computers we know of today.
Yes, they've been researching it for years, now. Won't happen soon, though.
And, if you want my opinion, it won't ever be the best thing in computers
-- it's a technology based on today's ideas, expected to be the
leading-edge in 2100.
>
>Anthony: Lastly, the voting itself is a problem. Is there someone to
>tell me that I, having written Interpreter, accept a patch, I owe to
>the patch-writer equal control over Interpreter?
>
>Alain: Which is why I want to avoid voting as a binding decision-making
>tool. voting should only be consultative.
Then who makes decisions? I can answer that, but I don't think you can. The
idea of getting consensus from 40 people is beyond silly.
My answer starts now:
>
>Anthony: Or that I have to fight him for control, at all?
>
>Alain: Is the interpreter that you are working on YOUR property or the
>property of OC?
Mine.
>Will you forever have the last word on its evolution
>and its disposition?
Yes.
>At what point will the contributions of members
>other than yourself to the interpreter will be sufficient for it to be
>considered a collective work instead of the work of one very-dedicated
>individual?
Never. But do see my section on forking, below.
>
>Anthony: That is what the voting will turn into: Any person who writes
>a function gets the same rights as the person who really wrote it. Or
>do we attempt to give different voting weights? But what sillyness is
>that?! Who decides? A vote?! By whom?
>
>Alain: Democracy is not necessarily egalitarian, nor is it necessarily
>based upon majority-vote.
Democracy not based on majority-rule? What kind of game are you playing
here with definitions? Democracy, by definition, is majority rule (or
representitive majority rule)!
>
>Anthony: Can we reduce another person's share -- as splitting it with
>another party would -- without his consent; is that not theft?
>
>Alain: Minorities have rights. We cannot arbitrarily take anything away
>from them that is rightfully theirs.
Can we do so non-arbitrarily?
>Alain: One person as the Copyright Holder for the whole OC
>Distribution? That would be silly, unless the person is a corporation
Why is it silly? Go look for the answer to these questions:
Who is the sole copyright holder of Perl?
Who owns Linux?
>or something. But I suspect that you are suggesting that EACH component
>of the OC Distribution, your interpreter for example, would have 1 to 3
>authors cited.
I don't mind giving "special thanks" to people who I want to thank. But not
copyrights.
>
>Anthony: I will not enter into any arrangement that may endanger my
>right to and rights in Interpreter -- by democracy, dictatorship, or
>otherwise.
>
>Alain: Why have you chosen the Open Source strategy instead of forming
>your own entreprise and copyrighting your software in the traditional
>manner?
Two reasons:
1) With OpenSource, I get a free development team to port my software
to weird platforms, to develop fixes, to add features.
2) With OpenSource, I get a much larger market penetration and
guess who winds up selling more OpenCard stacks and Interpreter
support.
>
>Anthony: If someone wishes to give me a patch, fine.
>
>Alain: What if someone contributes something much more substantial than
>a mere patch? What if someone wishes to improve the interpreter but in
>a manner or style that you donít particularly like?
Fork!
>
>Anthony: But all they get is the guarantee of the Artistic that they
>may distribute the program and its sources, and the chance that I may
>integrate and maintain their code for them. And if anyone does not like
>what I do with Interpreter, and proves unable to convince me through
>reason to act otherwise, I'll tell him one thing: "Fork!"
>
>Alain: I am uncomfortable with the idea of collaborating on a huge
>project, for several years (or indefinitely), investing time and money
>and brain-power into it, if one of the key elements (the interpreter)
>is forever controlled by one particular member that insists that he
>invariably has the Last Word.
I don't have the last word. You do. The last work is "fork." You can rename
it something else, and then it's all your to do with as you please. You
don't have to get my approval, you just have to follow the very liberal
Perl Artistic.
Think about the power of forking -- and then you'll see why it's so important.
I think your uncomfortness "if one of the key elements...is forever
controlled by one particular member that insists that he invariably has the
Last Word" is based on your belief that I am the last word if I insist on
sole ownership of Interpreter, I am the Last Word. I'm the Last Word on the
official version of a certain product called "Interpreter." But if Alain
Farmer wishes, Alain Farmer can be the Last Word on a certain product
called "AlainInterp" if he simply releases the source code.
>
>Anthony: All the authors sign an agreement with the publisher to allow
>the publisher to have it published. And the publisher would be obliged,
>no doubt, to track down an author were it contractually obligated to
>pay him royalties.
>
>Alain: There are no royalties to pay out because we are an Open Source
>initiative, so locating authors should not be so imperative. They only
>need to be contacted to negotiate licencing terms that are not already
>provided for. This would be the case if each separate component had a
>distinct set of authors, but would be avoided altogether if everything
>is under one licence.
If there were many copyright holders, how do you decide about "making other
arrangements with the copyright holder"?
>
>Anthony: But three authors (you seldom have more than that) is _far_
>better than 300.
>
>Alain: And one collective licence is far better than hundreds.
There would not be hundreds. The Artistic would let you licence OC as you
choose -- you obvoisly are using a different name AND you are releasing
source.
>Anthony: I guarantee you this. Most people will _not_ want email
>because they happened to contribute to OC. And if we require them to
>allow us to send mail, many will not contribute.
>
>Alain: Participation in this list would be voluntary. Those who do not
>wish to participate can not-suscribe, those who no longer wish to
>participate can un-suscribe, or those who move away without updating
>their coordinates will be implicitly unsuscribed.
But, then how do we obtain their permission?
>
>Anthony: I will say that I will have no part in the mailing, because if
>I did, I would have to deal with my ISP, which has a tough policy on
>SPAM (thankfully!). And presently, I'm on good grounds with Rosie [the
>Erol's Abuse Guy], and don't intend to change that.
>
>Alain: Your ISP prevents you from participating in a mailing list where
>licencing terms of OC would be discussed??
I thaught we were talking about a list to which one is subscribed because
he holds a partial copyright to OpenCard because he contributed a patch.
At the time, I thaught it would be mandatory. I still don't see how it
would not be.
>Alain: Clearly it would have to be voluntary.
How? Wait... I see, below:
>Alain: ...If they donít participate in
>the deliberations, however, they will have to accept its conclusions.
If they own part of it and have rights to it, how can we make them accept
its conclusions? Either they own part of the copyright or they don't.
Either we accept majority rule or we don't.
I say they don't on the former and that thus the latter does not matter
(ak! that rimes!).
