🚀 go-pugleaf

mathew@mantis.co.uk (Industrial Poet) writes:
> It is demonstrably the case that we cannot assume that the patent office will
> be competent, so *I* am not assuming it.

  Granted.  Though I wouldn't support something that amounted to an
upper-bound on royalties.  It'd have to include revamping the patent
system so that it *is* competent.

> >                                           Besides, if you could come
> > up with 999 *NON-TRIVIAL* inventions, you (according to some model of
> > the universe) deserve to reap some benefit from sharing them with us.
> 
> Right.  But what happens then is that the big companies just make sure they
> include 999 useless features which they happen to own the patent on -- that
> is, nice ideas which are useless for this particular piece of software -- so
> that they hardly pay anything to the one guy whose big idea they use.

  Well, you'll end up with people who'll either pay big $$$ so that
they can get software which has lots of fluff and people who will pay
less for software that *does what it's supposed to* without fluff.
Also, let's not forget the size and speed of an application.  You
could reclaim speed by simply having lots of inert code.

      However, I imagine the *new and improved* patent system (nips)
to involve the investigation of retail software by any number of
indepent contractors (sort of like Notary Publics) who will examine
the software to see which patents it infringes on.  Of course, they
will have a huge database with all the patents filed and approved,
plus they would have to pass an exam of some stringent measure in
order to be licensed.  In fact, I view them very much like Federal
Judges, only with training in programming (cs, whatever) AND law.
And, of course, their decision could be challenged by taking it to the
*REAL* legal system, though that would often be unecessary.  Such a
icpo (independent contractor for the patent office) would *notice*
that the inert code was *NOT* part of the working code (but simply
inert "filler" code) and, therefore, *NOT* considered applicable under
the laws.

    That would leave you with two options:

	 1) Use fluff to decrease the costs payed out to others for
	    their patents. BUT suffer on performance and size.

	 2) Use the minimum necessary to get the job done and provide
	    your estimation of reasonable functionality without having
	    to sacrifice speed and size.

Of course, again, the neat thing is - NOW, today, GNU code is
sometimes less functional (or non-existant) due to the inability to
use certain methods because they are patented - it could be more
functional at no cost, for those with access to the internet.  Bucky
Fuller wanted to see a global electricity network - I'd like to see
that too - but I'd also like to see a global internet where every
person who wanted it could hook into the network - so people wouldn't
*NEED* to buy tape copies.  The FSF could (hopefully) reap the same
money in donations (though the ol' free rider issue is a problem) as
they now get in distribution costs and not pay a *CENT* to those who
own the patents.  Seems like a big win (if you get past
free-ridership) for the FSF and anyone else who wants to write cool
software and doesn't necessarily care to charge big BUCKS for it.


> Whereas the lone inventor doesn't have 999 patents, so he's screwed.

  In what sense?  What's to keep the lone inventor from putting 999
other patented ideas into his program?  He's still going to be paying
3-5% of the selling price - each patent holder is only going to be
getting 1/1000 of 3-5% of the retail price.  I don't see him getting
screwed at all.  He won't reap as much reward as he does from the
current system, but then again he gets screwed if he tries to market
under the current system - at least he can keep those costs down in
this example.

> The net result is to reward those who have lots of useless or semi-useless
> patents and plenty of lawyers, and to encourage the inclusion of redundant
> features rather than innovative ideas.  Companies with lots of patents are
> able to use those patents to reduce the amount they pay to other people,
> hence keeping their patent fees low; small companies can't compete because
> they have to give 5% of their profits away to start with.

  I don't see this at all.  

  First, the more redundant features you include in a program the less
you pay to each individual, but not the less total, provided you don't
own the patents.  

  Second, IF you do own the patents, and you include redundant things,
there will be some  suffering in speed/size and people who care will
buy something else. 

  Third, 3-5% of the retail is a small enough amount that it shouldn't
put a serious dent in *ANYONE'S* pocket.  If you sell your software
for $39.95, then you'll owe me (assuming you use my 999 patented ideas
and 1 of your own) a grand total of $1.99 per copy (assuming 5%) or
$1.19 (assuming 3%).  Now, the same software would only cost me $0.002
per copy (assuming 5%) or $0.002 (assuming 3%) - both figures are
rounded up.  So, while I do get a big advantage over you, get an even
bigger advantage if I just hoard my 999 patented ideas and do without
your 1 patented idea.

> In other words, the situation is no better than it is now.  Big companies
> win, individuals and small companies lose.

  Not true.  In the case which seems more likely assume you write
software which uses 15 patented methods, you own 14 of them, I own 1.
You sell your software for $199.95.  It'll cost you $0.67 per copy in
royalties to me.  (note that this is MORE than the average recording
artist makes off of their records, tapes, etc.  they tend to make
somewhere between $0.02 $0.50, as far as I know).  Now, it'll cost me
$9.33 to sell the same software for the same price.    Still a large
discrepancy in the cost, but notice what happens when my ownership of
patents drops to 0.  Same software, all patents owned by you, my cost
is $10.00 a copy.  (The same $0.67 you were paying to me before).
Now, if I sell millions of copies, then I'll probably end up paying
you more than it would have cost to license the stuff in the first
place - but there's nothing to keep me from doing that under this
system either.

> Of course, it might force people to use GNU methods of software distribution,
> since they might be better off charging for service rather than for the
> product in order to avoid the patent fees.

  Not necessarily.  If you have the same 14 patents, I own 1.  Cost of
software - a modest $39.95.  My price in royalties to you:  $1.86 per
copy sold.  That's not a whole lot, but if it's too much, then I'll
simply try to arrange to license the patents from you for a set fee.
As long as you don't have a run-away bestseller,  you can probably end
up *saving* money by not licensing.

> The problem with this is that it's much like the situation we have now; big
> companies like IBM can do what they like, because they have so many patents
> that you can't make anything without infringeing one of them.  But the small
> companies can't enter the market, because they have no patents to trade in
> order to let them get the permission to use the vital patents owned by the
> big companies.

  There's a big difference.  In today's system, small companies can go
broke trying to see if they are infringing upon someone's patent and
then trying to license it.  Under the new system, access to the
patents would be easy (just consult your icpo) and costs would be kept
down.

  Of course, the new system is not a *TOTAL* solution - there isn't a
single *TOTAL* solution which will make everyone happy.  But this is a
good start down the road to a reasonable compromise!

> We shouldn't be giving big companies an economic advantage over small
> companies AT ALL.

  Again, you're approaching this from a different angle.  You talk
about removal of patents, I say "fine - take 'em away."  I agree with
the notion that there aren't any ideas that are *SO* obscure it's
*NECESSARY* to patent them in order to get them out in the open.  I
think that many people will rediscover them.

      However, here, now, today, many people will not agree to this.
Here, now, today, there is a chance to make a *GOOD* change in the
patenting process (one that cuts costs on finding out what patents
there are and whether or not you infringe, one that will make
standards for patenting stricter ) - ultimately, I would like to see
the patenting of software ideas go down the drain.  But I think the
best way to get there is to walk down the road one-step-at-a-time.
I'd like our first step to be controlling the patent system which is
obviously out of control.  Once it's under control, though, you might
find that many people who support you now - won't.

     I'm a member of the LPF, won't you join too?


----------------------------------|++++++++++++++++++++++++++++++++++++++++
| "He divines remedies against injuries;   | "Words are drugs."           |
|  he knows how to turn serious accidents  |     -Antero Alli             |
|  to his own advantage; whatever does not |                              |
|  kill him makes him stronger."           | "Culture is for bacteria."   |
|                   - Friedrich Nietzsche  |     - Christopher Hyatt      |
-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-
|"When a man commits himself to anything, fully realising that he is not  |
| only choosing what he will be, but is thereby at the same time a legis- |
| lator deciding for the whole of mankind - in such a moment a man cannot |
| escape from the sense of complete and utter responsibility."            |
|            - Jean Paul Sartre  on the meaning of "anguish"              |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+