🚀 go-pugleaf

In article <1991Mar5.052803.22186@Solbourne.COM> imp@Solbourne.COM (Warner Losh) writes:
>In article <878@nddsun1.sps.mot.com> waters@nddsun1.sps.mot.com (Mike
>Waters) writes: 
>>See above. I have already described how to do a patent search for under
>>$10, and most of that (here at least) is for parking fees!
>
>Part of the problem is the shear scope of what would have to be
>searched for.  Do I have to search for something as simple as linked
>lists?  Or fancy XOR techniques?  Or using include files?  Or binary
>trees?  Or hashing?

I guess it depends just what the problem is that you are trying to
solve! If I am trying to write code to solve some significant problem
(circuit design or software) I have always included a day or two at a
library doing a literature/patent search, at worst I get a bunch of
ideas that don't work for some reason, at best I find something close to
what I'm trying to do. I also know if I MAY have a patent problem.

If I plan to market/sell the end result there are a bunch more things
like trademarks that need to be looked at. There is enough hassle in
actually selling "things" this way that about 90% of my ideas have been
simply written up and published in a magazine. The "author incentive
award" plus any payment from the magazine is generally more than I would
have cleared by selling it. I haven't been smart/lucky enough to develop
the next Space Invaders yet, (i.e. make a fortune from it).

>I may use a dozen or two techniques in any given program.  Even at $10
>a pop, this can get expensive.  Also, I can't do this all the time
>from just the abstract of patents.  While it takes little tine to grep
>through all the patents for XOR or hash or whatever, it takes a fair
>amount of time to figure out just what exactly is being claimed by a
>patent.  It also may be claimed in prose rather than using the current
>technical term.

THis is what is often called a "technical education". You get it by
studying the field in which you want to become an expert. Most people
keep a notebook on things that are relevant to their interests so they
don't have to keep redoing things.

>Have you ever read a patent?  

Every day. Currently I write patents for a living.

>They are very hard to understand w/o studying them for extended 
>periods of time.  

Depends on the patent. Some are some aren't.

>I feel that this time is
>better spent writing and testing code rather than checking out n
>patents that may, on the surface, appear to cover what I want to do.

That is your choice, but that choice has consequences.

>They are written to be as obscure as possible w/o causing the patent
>to be rejected.

Nonsense. The "Background" and "Summary of the Invention" are pure
unadulterated selling jobs. The writer WANTS you to use the patent (for
obvious reasons :-) ). THe "Abstract" and the "Claims" tend to be
obscure, but only because of legal constraints on their structure. For
example each claim MUST be a single sentence; which leads to some very
strange sentences; each sentence being separated by a semicolan to make
as much sense as possible.

>Maybe the hardest thing for me to grasp is how so many obvious things
>are patented.  Do you do a search for 2+2=4?  Then why would you do
>one for XOR?  After all, it is just a special mathematcal function.  I
>know that it is patented, but what about the other couple of dozen
>things that I learned in school/on the job.....

If it truly is "obvious" you cant patent it! Just because someone claims
that something is obvious doesn't make it so!

-- 
           *Mike Waters    AA4MW/7  waters@nddsun1.sps.mot.com *
If entropy is increasing, where is it coming from?