ESP Wiki is looking for moderators and active contributors!

Talk:Oracle v. Google (2010, USA)

Revision as of 15:59, 11 September 2010 by Ciaran (talk | contribs) (OpenJDK: Yes, unfortunately. The scope of the OIN coverage isn't very clear. Maybe it would cover all the features that were in 1.1.10. So if someone had patents on features that are new in 2.)

OpenJDK

I don't quite understand what the author means with this information:

-8<- Rumor was the OpenJDK was only for desktop distribution, and Java Micro Edition (JavaME) was not under the GPL license. Thus we can have a debate about whether or not the desktop OpenJDK can actually be used on a mobile platform such as Android. (Of course, the counter-argument is that hardware keeps getting more powerful, with the smartphones of today having the ~same CPU power as a desktop from 10 years ago, which was after the introduction of Java on the desktop in 1996...so what actually constitutes "mobile"?) -8<-

Android has a 'normal' Java implementation (bytecode is incompatible, sourcecode is compatible. Which is actually the point where all the fuss is about). Dalvik doesn't implement any of the Java ME API's. Also Java ME isn't a different Java distribution, its just a set API's for mobile devices on top of Java. There is no Desktop, Mobile or Server Java, they are just API's on top of basic Java. The OpenJDK implements the SE (Standard Edition) set of API's (Dalvik implements a large set of the SE API's, not all of them) . Android provides its own set of API's for mobile devices on top SE.

The biggest problem (IMHO) for using OpenJDK would be that since the bytecode of Dalvik is incompatible with the rest of the world (Hotspot, OpenJDK, BEA JRockit, etc. etc.) everyone would have to recompile there programs and redistribute them (This is also the real reason for the lawsuit).

KermitTheFragger 11:55, 11 September 2010 (EDT)

Some of that section will have to be cleaned up. I'm not sure that documenting rumours has any value for anti-swpat campaigns. Anyway...
JavaME is distributed by Oracle under GPLv2 as "PhotoME": Java_and_patents#What_exactly_was_distributed_under_GPLv2
I've never written Java, so don't know much about the differences between the various editions. If you see errors or stuff that needs to be more explicit, do jump in and edit the page! Ciaran 12:08, 11 September 2010 (EDT)
I vote for removeing the entire section about 'Rumor was the OpenJDK was...' since Java ME does not have anything to do with this lawsuit.
I also have a remark about this section: "The test suites are only available to JCP members and have a restriction that limits the use of the tested software to desktops."
The TCK's (ie. 'Technology Compatibility Kit', the test suits) are are available for all standards in the JCP, including all non-desktop like: Java EE (Enterprise), JDBC, JavaMail, etc. The problem (among things) is Dalvik produces incompatible bytecode and therefor fails the TCK for Java SE.
This whole lawsuit doesn't have as much to do with Java as most people think (Or I got it wrong, which is also possible ;-). Strictly speaking Dalvik is not a Java implementation because it fails the TCK's. Oracle could also use patents like "Controlling Access To A Resource" on Mono or Microsoft for their .Net implementation since their inner workings resemble those of Java.
--KermitTheFragger 12:37, 11 September 2010 (EDT)
(quick reply) Those patents might apply to Mono, but Mono is covered by the OIN licence that Oracle signed, so Oracle can't sue Mono users. Ciaran 12:46, 11 September 2010 (EDT)
(Also an quick reply ;-) My knowledge about OIN is limited, so I wonder; Why can Oracle sue Google (who is an OIN licensee) over Android but wouldn't be able to sue Novell over Mono over the same patent ?
--KermitTheFragger 13:45, 11 September 2010 (EDT)
Because Android/Dalvik isn't on OIN's list of approved software: http://www.openinventionnetwork.com/pat_linuxdefpop.html
GNU Classpath is on the approved list, so if Google built on top of GNU Classpath, they might have been protected. Ciaran 14:35, 11 September 2010 (EDT)
Mono is listed as version 1.1.10 in that list, that release is over 5 years old. Does that mean that the patent pledge doesn't cover the 2.X branche ? --KermitTheFragger 15:42, 11 September 2010 (EDT)
Yes, unfortunately. The scope of the OIN coverage isn't very clear. Maybe it would cover all the features that were in 1.1.10. So if someone had patents on features that are new in 2.X, they couldn't count on OIN for protection. Ciaran 15:59, 11 September 2010 (EDT)

Unusual patent number

It's because it's a reissued patent. The patent number is 05367685.

Thanks! I'll update the article. Ciaran 10:22, 13 August 2010 (UTC)

All trivial patents?

The quoted claims seem to all be completely trivial.

Shouldn't Google be able to get them all invalidated easily? (even without prior art)

Invalidating is never easy. For one example, see Amazon's 1-click patent. The review process took five full years, and the end result was that it was narrowed but upheld.
But let's see what we can gather anyway - it could be useful for this case or for a future case. Ciaran 14:24, 13 August 2010 (UTC)

Actually, looking deeper, the problem is that the parts quoted here (the first claim, which the patent then expands upon) don't seem to very representative of the actual patents, which do have more specific detail.

You should probably remove the quotes and either reproduce the full patent, write a summary or just have a link.

Private / Protected

From the page:

Prior art: This is C++ private / protected.

No, those apply at the level of individual members - not classes.

Whoever made that comment regarding not classes is incorrect. Private/protected can be applied to members AND classes in C++. As well as C# and probably some other languages.

Sublicensing and free software licenses

The part that reads "Further, this grant does not include the right to sublicense. That is incompatible with any free software licence and means that only meeting all of the six onerous requirements would actually grant a patent license" needs some expansion or clarification as to why not allowing sublicensing is incompatible with free software licenses.

Consider that GPLv3 explicitly prohibits sublicensing (last sentence of section 2, "Conveying under any other circumstances is permitted solely under the conditions stated below. Sublicensing is not allowed; section 10 makes it unnecessary."). Section 10 is the section that says when you convey the GPLv3 work, the recipient receives a license from the original licensor.

That seems to indicate that it is OK in a copyright license to prohibit sublicensing yet still be compatible with free software. I don't see why it would be different in a patent license.

That's a good question. I'll trying to find an answer. (Or if someone else has the answer, I would be interested to hear it.) Ciaran 17:56, 17 August 2010 (UTC)