[Linux-aus] Re: Software patents: Munich has put Linux project on hold!

[Anthony Towns]

On Fri, Aug 13, 2004 at 08:32:50AM +1000, Con Zymaris wrote:
> On Fri, Aug 13, 2004 at 12:52:10AM +1000, Anthony Towns wrote:
> > On Thu, Aug 12, 2004 at 09:48:34PM +1000, Con Zymaris wrote:
> > > thanks for putting your thoughts down so we can better understand the 
> > > ideas behind your past comments.
> > Hrm, it seems more like a cross-examination at this point than trying to
> > understand, but oh well.
> Crooss-examination is a form of understanding.

Err, no, cross-examination is a form of argument. Not that there's anything
wrong with that.

> > > On Thu, Aug 12, 2004 at 06:07:45PM +1000, Anthony Towns wrote:
> > > > > 1) Should software patents exist? If so, why?
> > > > I don't know. [...]
> > > Can you envisage software technology advancing 'just as fast' without 
> > > patents?
> > Of course. I can also envisage software technology not advancing anywhere
> > near as effectively too. It might advance much faster too. If I was sure
> > which was the case, I wouldn't have said "I don't know".
> As mentioned, lack of patents has not stopped advancements in
> architectural design, above and beyond any underlying civil engineering
> technology. 

And as mentioned, unless you're arguing that all patents are unnecessary
(which you might be, but if so you're just wasting your time and mine),
you need to also demonstrate that software's more like "architecture"
and less like "engineering".

> To my understanding, there were no software patents for the first 40 years
> of the software industry. It didn't appear stop innovation. Quite the
> reverse. These was more innovation in software engineering then than now.

Again, you can say things like that all you like, but just saying it doesn't
make it true. It's quite easy to say the opposite, too:

} To my understanding, there were no software patents for the first 40 years
} of the software industry. Did that have any effect on innovation? Certainly!
} There is far more innovation in software engineering now than then.

I'd actually be very surprised if there was less innovation now than then:
there's a lot more money and people working on software these days than
there was 20-60 years ago, and thanks to hardware improvements there's
just a lot more that software can conceivably do -- and no matter how
bad you think patents might be, there's not really much chance of them
being so bad as to overcome all that.

But in any case, if you're going to make that argument, you need convincing
metrics to support it, not just empty rhetoric.

> > Huh? Patents are there to reward innovation, and thus encourage it. If
> > they can do that, but still not satisfy you as being "just" or "morally
> > defensible" then you should go talk philosophy with RMS or someone,
> > because that just doesn't interest me.
> Copyright is there to do the same thing, and it doesn't trample on other 
> IP holders' rights.

Uh, "IP rights" are what legislation says, nothing more, nothing
less. If you think there's some sort of natural "IP holders' rights"
to get unjustly trampled on, that's great, but it's just not a debate
I care about at all.

> > defense, you'll need to look elsewhere. Pointing at other fields isn't
> > very helpful unless you're going to be arguing that no patents should be
> > offered anywhere, for pharmaceuticals or for better mousetraps or for
> > anything; otherwise you're already acknowledging fields are different,
> > so why should we assume that software's more like architecture rather
> > than drugs?
> If pharmaceuticals could be 'copyrighted' then they too shouldn't need 
> patents either. But they can't be, so they need patent protection. 

Pharmaceuticals could be copyrighted though, you could fairly easily say
"you're not allowed to make copies of this drug, you have to develop your
own drug from scratch". You could even make it substantively different --
make it a capsule instead of a pill, use different inert ingredients,
change the colour, vary the proportions somewhat, etc. 

And, again, you might say "well! great! let's not have drug patents!" Most
people won't though -- copyright protection doesn't achieve the same thing
patent protection does, and patent protection has long been considered
necessary for inventions. You can wave your hand and try to dismiss that
all you like, but that's not an actual argument.

> More importantly, there isn't a large case history of pharmaceutical
> firms being awarded bogus patents, as there is in the software and
> business process space. This is a killer difference.

I'm sorry, where's the large case history of bogus software patents
being awarded in Australia again?

> This avoids the key question. There _are_ sometimes legitimate reasons why
> drivers need to go over the speed limit. This is why there are no speed
> chokes on cars. Similarly, there are legitimate reasons why DVD owners
> will want deCSS to watch DVDs. 

That's nice. It's not an example of a legitimate reason to want to
illegally copy a DVD though -- playing DVDs doesn't involve illegal
copying, either because there's no copy made in material form (current
law), or because any copies made are incidental (FTA implementation).

I spoke exclusively about technology that prevents illegal copying for
a reason.

> If drivers are caught going over the speed
> limit, they should be charged. If DVD owners are caught infringing
> copyright, then they should be similarly charged.  Neither process should
> be choked off automatically by the machinery.
> 
> Agreed?

No. I meant exactly what I said in the previous mail; these are considered
opinions.

> > If we could come up with a technology that did nothing other than prevent
> > illegal copying, it'd be a good thing. We can't do that, of course,
> > but we might be able to come close to it, and sometimes close enough
> > can be good enough.
> This isn't good enough.  Not if it has serious on-flow repercussions.

If it does nothing other than prevent illegal copying, it doesn't have
serious on-flow repercussions by definition. If it comes "close enough",
then it has repercussions, but they're not serious, again by definition.

> > > Why should the 'circumvention' aspect of these devices take legal 
> > > precedence over the non-circumvention aspect? 
> > I'm not saying it should; I'm saying it _does_. (IANAL, TINLA)
> Irrespective of whether this is implemented or cemented by the FTA or by a
> preceeding law, 

You can't talk meaningfully about the implications of laws without
knowing what they are.

> my question is: shouldn't effort be made to ammend this?  

And you can't judge the worth of a law, or know if or how it's worth
amending, until you've considered the implications of that law.

And the Digital Agenda Review, the JSCT inquiry into the FTA, and
the Labor members of the Senate Select ctte have all recommended that
exceptions be made for (at least) interoperable software for playing DVDs.
That question's been answered, and the answer's "yes".

> Shouldn't legitimate uses be recognised? Or are you saying that bad laws 
> can't be or shouldn't be changed?

Legitimate uses are already recognised -- you can currently obtain and
use circumvention devices for certain "permitted purposes". Not all uses
that are, IMO, legitimate uses fall under "permitted purposes" though,
but that's okay because everyone wants to change that.

(In particular, you can't decide whether a law is bad, or how bad it
is, without understanding the law. Understanding the law, and accepting
that it *is* the law, is the *first* step in getting it changed, it's
not contradictory with that goal. But seriously, "are you saying that
bad laws can't be or shouldn't be changed"? WTF? Do you /really/ imagine
there's /any/ chance that I think that?)

There's a large leap from "more legitimate uses should be allowed"
to "circumvention devices shouldn't be regulated" though.

And look, don't get me wrong: I'm quite confident that leap *can* actually
be made, but we haven't demonstrated that and we've not even found any
evidence for it beyond our own gut feelings. Getting all pompous about
how "No code should be illegal to write. Period." doesn't help with
that, and worse, moves the debate into a religious one where there's no
point trying to prove anything, rather than making it a scientific one,
where we've got some chance of a rational outcome, and we don't have to
concern ourselves overly with the gut fears of the movie/record industry.

I'll be away skiing from tomorrow, so might not see any further posts
'til I get back in a week.

Cheers,
aj

-- 
Anthony Towns <aj at humbug.org.au> <http://azure.humbug.org.au/~aj/>
Don't assume I speak for anyone but myself. GPG signed mail preferred.

``Like the ski resort of girls looking for husbands and husbands looking
  for girls, the situation is not as symmetrical as it might seem.''
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 197 bytes
Desc: Digital signature
Url : http://lists.linux.org.au/pipermail/linux-aus/attachments/20040813/c73d9581/attachment-0001.pgp