|
Authored by: Anonymous on Monday, June 11 2012 @ 02:15 PM EDT |
Thank you.
I do agree that there is some merit to software patents,
in the abstract.
Your proposed solutions appear to be likely to, long-
term, reduce the problem of new, counterproductive, software
patents being granted and enforced.
In terms of additional suggestions, your solutions do not
mitigate the following problems:
(1) Uncertainty/risk. Based on observation of software
patents used in lawsuits, roughly 90% do not seem to be
valid. This creates a substantial risk tax for anyone
starting a business. It is also counter-intuitive, because
presuming that a patent is valid because it is granted when
90% of them are not seems ridiculous. (I am either being
unfair or too generous - and since I'm not sure - am
probably not too far wrong.)
(2) Cost. The most cost-effective way to invalidate patents
appears to be via lawsuits. However, most people do not
have enough money to fend off a lawsuit full of invalid
patents. Therefore, most small defendant will pay up for
infringement of arguably invalid patents. This also seems
ridiculous.
(3) Duration: The software industry is genuinely different
in that it takes much less time and money to get an idea to
market - so allowing 20 year monopolies stemming from patent
grant date is probably suboptimal.
(4) Overhang/multiplicity - there are about 20 years left on
what is essentially a gigantic pile of festering, ridiculous
overbroad software patents. This pile is a substantial tax
on innovation because _anything_ anyone can develop probably
infringes on 20odd patents.
I suggest in addition:
(1) Eliminating the presumption of validity unless the
patent has passed a 'validity trial'. That presumption, for
currently granted software patents, fails the laugh test.
(2) Allowing patent owners to initiate a very public
validity trial. (~2M USD, much less than any single
infringement trial) And offering modest bounties (~1M USD
total) for submitted briefs leading to invalidation. It is
simply unjust to ask a small business owner to front 2-3M
USD to fight off an infringement suit based on a probably
invalid patent, particularly when that business owner could
easily be infringing 20+ patents.
(3)Shortening software patent terms to 12 years from patent
submission. 20 (really 30 years because of USPTO delays)
monopolies make sense in manufacturing-based industries.
(can take 5-10 years just to build a factory...add 3-5 years
for R&D and 5 years to start turning a profit...) They
don't make sense for software-based industries. (2-3 years
for R&D, 2-3 years to turn a profit...)
(4) Restricting injunctive relief to patents which grant
substantial fractions of product value - and fairly
restrictive royalty amounts based on fractions of product
value. If a product incorporates 300 patents, each of which
is eligible for injunctive relief, it becomes really, really
easy for disputes to escalate until nothing is accomplished.
--Erwin[ Reply to This | Parent | # ]
|
|
Authored by: albert on Thursday, June 14 2012 @ 02:00 PM EDT |
Patenting the configuration of the gate arrays is a software
patent! Programmable gate arrays just save time and money by
eliminating the need to assemble discrete components. If the
gate array controls a patented process, then why patent it?
If it doesn't, then I can duplicate the functionality in a
general purpose computer, so what good is the patent?[ Reply to This | Parent | # ]
|
|
|
|
|