Re: Documentation Standards was Re: [ox-en] UserLinux
- From: "Niall Douglas" <s_fsfeurope2 nedprod.com>
- Date: Mon, 08 Dec 2003 04:13:40 -0000
On 7 Dec 2003 at 11:51, Russell McOrmond wrote:
Software is very, very different from other types of creative work -
it's a direct engineering solution. Not plans for a solution, an
*actual* solution.
I've had this debate in various pro-patent forums and I disagree with
it here just as much as I disagree with it there. Software may exist
within the field of engineering, but it is also an expression of public
policy, of art, and of things in many other fields. Yes there are
those from the early days that did not differentiate hardware from
software that consider software to either be an expression of math or
or engineering, but we need to get past that limited way of looking at
software.
Software is no more simply an engineering solution than Robert's
rules or an act of parliament is.
Of course it isn't, but it is *primarily* an engineering solution.
Anyway, all those pro-patent idiots just don't seem to twig that
irrespective of whether software can be represented in hardware, it
doesn't automatically mean the same laws should apply. They're two
very different things - one is infinitely malleable, the other is
not. One can be copied for zero cost, the other can not. One is far
more reusable, the other is not etc. etc. - therefore, it's obvious
to a labotimised chimpanzee that it requires a separate legal
framework.
Anyone advocating software patents is making a severe error in
logical typing or they're just really ignorant. They should leave
important decisions like this to people who actually spend time
around software and know it.
of technical knowledge just to operate a computer, you need way way
more to design software plus there is very little tolerance for
error - - a comma in the wrong place, and your software isn't going
to work. There is more give for error in designing a jet engine.
This is a discussion of the limits of early languages used to create
software, not an expression that is inherently true of all software
creation.
I'm very sure it will always be true, just as tiny mistakes in the
mathematics behind string theory are a serious problem. So long as
computer languages have algorithms, this will be true.
I also challenge you to speak to a musician to see if those
instruments
are as forgiving as you seem to believe, or that the knowledge is so
much easier. I truly believe the issue is access to tools (and part
of that relates to breaking past the limits of early languages used to
express software) and literacy.
One man's music is another man's noise.
Why is one type of knowledge worker considered uniquely different
than any and all other types of knowledge workers?
My point was that programmers seem to invest much more in the
opinions of elite programmers than correlates in other professions.
If some top violinist says that all music must be free, would a
sizeable minority of all musicians totally buy into that mantra?
I take the view that all copyright & patent law needs throwing away
and redesigning from the ground up.
The important question then becomes: what is the ground? What is
your starting point for these laws? I have offered mine, and I agree
with article 19 and 27 of the UN UDHR.
Copyright will inevitably become unenforceable with time. It has been
made obsolete by the march of technology.
I have realized that the changes I would make to these laws are not
all that different than the origins of the existing laws. They simply
need to be modernized coming from first principles, not abolished.
How can you possibly enforce information scarcity in a world where
information can be copied for zero cost?
Patents should be limited to processes involving the manipulation of
nature to create a tangible rivalrous product. Where the process is
involved in the creation of intangible non-rivalrous goods, such as
information processes (software, business models, methods of
organizing people, governance structures, etc) then patent law simply
does not apply.
I'd go much further. Patents need a different lifespan for each
industry - vacumn cleaners are different to diamond cutting - 20
years as TRIPS mandates is just foolish. Patents for business should
cost 100x more than for private individuals. Patents should have a
real test for prior art and business applications should pay for
private application reviews. The patent system should also become
part of the UN because large parts of Asia just ignore it, feeling
they get no benefit from it.
However, I'm a realist - patent theory in the 18th century just
doesn't apply in a world of big business. Even if a large company
violates my patent they usually have to pay me rather than stop their
own production. Furthermore, globalisation means that production in
the west is artificially ludicrously expensive and individuals and
small companies can't outsource to the third world, so any form of
mass production must now be done by large companies - therefore, much
of the original use of patents is pretty pointless - they're just
tools of economic warfare nowadays.
I'd doubt if the world would change much if they were simply done
away with.
The ability to cheaply communicate information is unrelated to
whether or not governments should offer some sort of limited temporary
monopoly to the moral and material rewards or the creator/inventor.
Not if everyone ignores the government.
- music copyright infringement online is solved by changing the
business model. Don't charge for non-commercial private communication
of music, but treat that as advertisement for commercial communication
(media sales, concerts, other products/services, etc). Don't promote
using expensive commercial air-time, but receive payment from
commercial air-time and promote peer-to-peer.
When people's mobile phones get beefy enough, never ever more need to
pay to listen or watch anything ever again especially with adverts.
You just fetch it from your (technically illegal) private encrypted
file swapping group over your 8Mbit connection.
- software copyright infringement is also solved by changing business
models away from royalty-based business models. FLOSS business models
are ideal, and limit the copyright infringement cases to cases against
"software pirates" like the SCO group. Proprietary software business
models would also work if on a subscription service like many companies
are moving to where it is per-computer-per-year or
per-employee-per-year, not per-copy. Proprietary software business
models probably don't make sense for the home computer market at all.
There's a reason why even Microsoft haven't tried to rent Windows out
to clients and that's because it's deeply, deeply unpopular. That
won't change.
The free software revenue model simply outsources software production
to the third world. Bye bye jobs for western programmers.
Local support companies will remain eg; RedHat Europe's main purpose
is installation support. But as you'll note from RedHat Europe, its
employees don't do much programming - support is not programming.
Cheers,
Niall
_______________________
http://www.oekonux.org/