|
Authored by: Anonymous on Friday, May 18 2012 @ 09:28 AM EDT |
Exactly! Even before Oracle broaden any claims, the patent
appears to be describing a linker (static as well as
dynamic, since it can be done only once so long as the final
result "depends on the actual installation".)
The whole thing shows the absurdity of software patents:
whatever you patent is only ever going to be an algorithm,
and discussing, analysing and coming up with new ones is the
essence of the profession. It's as absurd, and counter-
productive, as granting patents on mathematical formulae.
Everyone knows software patents are overly broad, and
companies make them as broad as possible, so that they can
threaten suits like this one, after being granted a patent
to either a description of existing ideas, or incredibly
obvious non-ideas (like the Stacker patent, or "connecting
over a network to perform a transaction^W purchase".)
There's no peer review at all, and patent examiners aren't
software engineers. The continual broadening (which is
natural commercially, and technically given the nature of
software) should have led to the whole thing collapsing
years ago. As it is, patents are used by companies to put up
artificial barriers-to-entry via cross-licensing of what are
known to be dud patents. After all, the only way to
challenge them is through the legal process, right?
Totally distorting the free-market and making a sick joke
out of the USPO.
Ranjit.[ Reply to This | Parent | # ]
|
|
|
|
|