Google Developer Testifies That Java Memo Was Misinterpreted

benfrog writes with a piece that appeared in yesterday's Wall Street Journal about the in-progress legal battle between Oracle and Google over Java: "Ex-Sun and current Google employee Tim Lindholm testified that it was "not what he meant" when asked about the smoking gun email (included here (PDF)) that essentially said that Google needed to get a license for Java because all the alternatives 'suck[ed].' He went on in 'brief but tense testimony' to claim that his day-to-day involvement with Android was limited."

Comments

Nice foresight

[ghn]

On Fri, Feb 20, 2009 at 7:59AM, Dave Sobota wrote:

2. Can you spell out the risk of us relying on Sun for support in more detail? I thought Java was largely opensourced anyway -- so I don't understand why we'd be so worried if Sun went bankrupt, was sold to an unfriendly company or just decided to act erratically with respect to Java. Is it that we are concerned about the parts that are not opensourced (e.g., test suite) --that Sun might jack up the license fees or just stop licensing those altogether?

p29 of the exhibit

Re:Pretty clear to me

[symbolset]

It's about both patent and copyright. They're down to only one patent though. The actual copyright is minimal, so Oracle's trying to make the silly argument that they have a copyright on the API. Not only is such a thing impossible, but they can't produce a "work" that's infringed nor an exemplar of a reproduction.

I'm going to agree with the grandparent. One employee urging the company buy something to solve a certain problem is not proof the company stole it. If the company decides to achieve its goals in a different way strategically because the object of their desire is not for sale (which is the case here) this also doesn't mean that they stole it.

Re:Obvious implementation

[sco08y]

That "copied" code is a joke.

If you have the function prototype per the java doc and you test the limit cases to see what exceptions are thrown, how else can you implement that array range check in a non-trivial way?

That's just lawyer bullshit.

My first thought, too, but it's a private method, not implementing any API.

What's funny is that they're highlighting this trivial function while inadvertently showing that Android uses TimSort, whereas the Java standard library which, according to the docs, uses a tuned quicksort, adapted from Jon L. Bentley and M. Douglas McIlroy's "Engineering a Sort Function".

He ain't a lawyer

[sirwired]

I read it as: Whatever kind of license we need to run Android, we should get one. As somebody who was not, himself, actually involved in any kind of licensing negotiations, laws, etc., he didn't have the least flipping clue WHAT that might entail.

Imagine my boss seeing emacs for the first time and saying "Holy $hit! this emacs is awesome! SirWired, go buy whatever we need to run it." That isn't any kind of admission that running emacs requires paying somebody; just a statement that he wants it.

Google's position is that no license from Sun was in fact needed, and that Eric Lindstrom was not the person that had anything to do with that determination

Re:This e-mail was years after Google started Andr

[Chibi Merrow]

Actually they are trying to use the email to show that Google were knowingly breaching the license, not that google knew ahead of time that they were in the wrong.

I don't think I understand the subtle difference in what I said and what you said... So let me try to be clear by pretending to know what I'm talking about for a second: Oracle is trying to use the e-mail to show evidence of WILLFUL INFRINGEMENT.

And it could be evidence of that, except: 1) it was written after Oracle started threatening, 2) it was written AFTER Sun's CEO publicly congratulated Google on releasing Android and promised Sun would support the Android platform 3) it was written by someone in Google not working on Android 4) it was saved as a draft but not sent.

Oracle are dicks, but that is only a mild improvement on the twats that are google.

Are you trying to say Oracle is less evil than Google? Are you at all familiar with Oracle?

Honestly even if Google is in the wrong here (which I strongly disagree with), I think it's vitally important Oracle lose this case, as the legal theory they are trying to push (that public APIs are subject to copyright and licensing) is absolute INSANITY and would be the death knell to a large percentage of the software industry.

Though it would make developing on proprietary systems significantly more painful than Free/Open systems like Linux, so that would make some people around here very happy...

Re:Oh come on

[BillX]

Looking at the various email threads, it reads to me like Google decided a Java-compatible platform was their best option, and licensing one from Sun (they already had one, called Java) was the easiest path to getting the job done. Another email in the exhibit mentions the possibility of clean-rooming their own Java-compatible platform, but basically saying it would be a giant PITA and not cost-effective vs. just licensing Sun's existing one. I don't see the Java license discussions as construing "proof of knowledge of infringement" (etc.), or belief that it was the only legal path forward, only that at least one guy believed it was the easiest path forward.