On Fri, 20 Jun 2025 23:48:34 +0000, Lawrence D'Oliveiro wrote:Please note that "no patent disclosing the idea" means that all Knuth-era textbook sw algorithms can now be implemented in VLSI and then patented.
On Fri, 20 Jun 2025 18:07:43 -0000 (UTC), Thomas Koenig wrote:No, you just have to explain the idea to the examiner in such a way
>Lack of inventive step is more thorny.>
Also, you don’t actually have to prove an idea works before getting a
patent on it.
that [s]he::
a) thoroughly understands the idea
b) can find no patent disclosing that idea
c) verifies the claims match the evidentiary text
d) finds no essentially identical claims in other patents.
Being documented in a 20-year old uni text book is not enough to stop the US patent office from allowing your patent.
I have personally reviewed 4 different, supposedly valid, patents that fell into this category, along with 4 others where the "patent step" was non-existent, i.e. obviously the only reasonable way for anyone "skilled in the art" to sole that particular problem.
Terje
--
- <Terje.Mathisen at tmsw.no>
"almost all programming can be viewed as an exercise in caching"
Les messages affichés proviennent d'usenet.