En réponse à John Cowan :

>Technically, a translation in the U.S. cannot be made even for purely
>private use, because it is a derivative work, and the making (not merely
>the distributing or displaying) of a derivative work is protected by
>the copyright.

Really? I thought private use was outside the scope of copyright. After 
all, you *are* allowed to make a copy of a works for private use. 
Everything I know about American copyright law is that private use is an 

>   In practice, of course, there is no particular way
>for the copyright owner to even know that such a private translation exists.

Which is the reason, like for private copies, why they are actually allowed.

>The name "Klingon" is trademarked; a trademark must actually be used in
>trade, in this case to sell the dictionary and various other things.
>The words of Klingon are not trademarked, and probably cannot be
>copyrighted either, at least in the U.S.

I thought they were already, and that it was a problem for personal works 
made in Klingon.

>So do I, but that doesn't mean they won't happen.  The U.S. patent
>commissioner has gone on record to the effect that a legal argument
>used in court could be patented by the law firm, provided it was
>sufficiently novel (not very) and useful (no, you don't have to win).

Patents are really going out of control. Time for the US to stop it or they 
are gonna paralyse their own economy under an impossible justice system.

>Not unless the language is used in commerce somehow.  The inventor of
>Loglan tried to trademark it as a name for grammars and dictionaries
>(not as a language per se), but the USPTO ruled that it was a generic
>name for logical languages generally.

But that's only because of the obvious origin of the name. If Loglan had 
been called "Milthannezic", I think the trademark would have been possible.

> > [S]imply put, you can only patent
> > something truly original - the "no prior art" argument -, something which
> > is not based on simply using the laws of physics with little innovation -
> > you cannot patent a new shape of catapult -, and something which is *not*
> > the only way to implement a certain action - a patent mustn't lead to an
> > absolute monolopy. There *must* be possible alternatives, although they
> > needn't have been developed yet -.
>Patent law is country-specific: these things may be true in France or the
>EU, but are definitely not true in the U.S.  New shapes of industrial widgets
>are patented all the time.

Actually, you are wrong here: I was talking only about US patent law, as I 
was explained it by American lawyers. The fact that new widgets are 
patented doesn't say that they are patentable, only that your patent 
commissioner has been very bad at following its own rules (under the 
influence of a lot of money, for sure :)) ). None of those patents would 
ever survive a law suit. And their point is not to win law suits. The 
software patent holders are big companies with more than enough money to 
put on lawsuits, even lawsuits with no value, and to drag them long enough 
to ruin the defendant and end the lawsuit without a verdict. The American 
justice system is extremely expensive, and companies know it. Patents are 
used not as a way to win lawsuits, but to threaten to ruin competitors by 
dragging them into frivolous trials. You should read for a good 
introduction on those things.

Note that just today, the holder of the JPEG patent has announced that it 
was suing 31 companies for illegal use of their patent, despite the fact 
that the patent expires in the 6th of October. The opinion of various US 
lawyers I read about was that it won't work, and one of their arguments was 
that the algorithms the JPEG format is based on were published already more 
than a year prior to the deposition of the patent, making the patent 
invalid through the "prior art" argument. That JPEG was patented anyway 
doesn't say anything about the validity of the patent, only about the 
brokenness of your patent system.

>In the U.S. you can collect the entire damages from whoever revealed the
>secret (provided they have it, of course).

But how do you calculate such damage? :) And if the robber is unsolvable, 
what do you do?

>IANAL, TINLA (this is not legal advice).

Same for me. But I read a lot about US IP laws in the last few months, do I 
can say I have some knowledge of them :) .

Christophe Grandsire.

You need a straight mind to invent a twisted conlang.