Google asks US Patent Office to rethink Oracle Java patents Google has asked the US patent office to reexamine four of the Java patents that Oracle is using to sue the company over the design of the Android operating system, according to a blog post from an American law firm. On Wednesday, a "patent reexamination alert" from the firm Westerman Hattori Daniels & Adrian announced that …
I wonder what Goolge would think if someone said that their algorithms in search were "abstract ideas or other non-statutory subject matter". Really, an formula that indexes words and sentences in a manner that makes them easy to extract based on formulas that match terms to these is an abstract idea as well....
A patent is supposed to have a technical effect. A patent that does not have a clear and well defined technical effect can be voided on technicality.
It is not specific to Oracle patents, it is a common disease in the industry. When you read a patent done in IT and Telecoms space , 70%+ of the cases you end up having no clue what it does. It has been drafted to be so vague and all-encompassing that the technical effect is no longer clear.
AFAIK, the current practice is moving towards the following: if the technical effect is not clear from the claims it is inferred from the description and the patent narrowed down to whatever is given as examples in the description.
Trying to invalidate patents on this count is quite rare though. Nobody wants to create a precedent that will invalidate half of the IPR out there including their own.
"Trying to invalidate patents on this count is quite rare though. Nobody wants to create a precedent that will invalidate half of the IPR out there including their own."
Google may get considerable advantage if they manage to nuke a huge swathe of Intellectual Property Rights (IPR) due to the fact they have relatively little to lose by doing so; they hold relatively few patents, and seem to rely on trade secrecy. This statement may seem surprising as Googles algorithms are well known, and a lot of their source code is open, but the exact finely honed implementation is kept secret.
Google also tries to be fast moving, in tems of getting stuff to market, although I don't know whether they will be able to keep that up.
I've not seen the actual patent (can't be bothered quite honestly) but the formatted presentation (via the link in the article) makes it look like Oracle think it's a patent on the code (work) not the algorithm (method)? Looks like they're on dodgy ground to me (for what that's worth) ... either way it shows the stupidity of trying to patent code.
Yes, I'm biased.
"On Wednesday, Google also asked the court if it could file a summary judgement in an effort to have the copyright claims dismissed. Android's Dalvik virtual machine is based in part on code from Project Harmony, an open source implementation of Java that Oracle has not granted a license for use on mobile devices. ®"
Didn't El Reg run an article which had Apache deny any connection to the code and Apache?
"both expressive material and copyright headers". Mountain View called these omissions "significant elements and features"."
Last I checked, copyright headers and other comments in a source code document have absolutely no bearing on the actual code. Sorry Google, they didn't copyright a book complete with commentary, they copyrighted the code, which is suspiciously similar/same. Best bet: prove they both came from the same source (a "how to code" book perhaps?).
It get's even funnier. 2 of the files Oracle claim are infringing copyright were written by *Google* then contributed under an open licence to Java *from* Android!
Removing copyright notices that identify Google as the copyright holder is about as stupid a mistake as it's possible to make. Did they forget Google are in the information business, that sleight of word processor couldn't be hidden?
This really is closer to the SCO fiasco than even I believed, the same not even knowing what they own before storming into court. Really need a popcorn icon ;)
If the removed notices contain a permissive licence binding on the complainant it blows their fairytale out of the water.
If they contain valid copyright notices that prove someone else owns the code, that's perjury.
Right now we're just waiting to find out which it is for each specific sample...
mathematical algorithms was brought to an end ? Would requiring all users of mathematical analysis to pay for a licence to Brook Taylor Limited have encouraged or discouraged discovery in both that area of human endeavour or, say, physics ? I may be naive, but I submit that the latter is obviously the case. However, for some strange reason, I very much doubt that the US Patent Office will become the agent of necessary change in its own practices....
Henri
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2024
