Gosling: Partnership with Microsoft Meaning Less and Less

Jeebus writes "At an event in Sydney this week James Gosling questioned the technical relationship between Sun Microsystems and Microsoft in light of the antitrust demands of the European Union. Gosling also talks about reverse engineering, DMCA and collaboration with Microsoft with on identity management. "

Comments

This is easy...

[AntiPasto]

... Microsoft has to let the Sun shine, or else they'd kill all their customers!

I'm confused

[PDXNerd]

Since when did a business partnership with Microsoft ever "mean" anything anyway (except decreased revenues)?

Re:I'm confused

[cduffy]

Yup, our VP of Engineering signed us up, much to IT's dismay -- as part of the contract we agreed to do a bunch of stuff we never had any intention of (like building a MS-certified product, though we're a Java shop) and agreed to let the BSA audit us at will... and in return we got a bunch of licenses (time-limited unless we renew), most of which have enough strings on them to be useful for nothing except building products that interoperate with the MS products in question. Bah!

PS - This post is a work of fiction. It's quite certainly not intended to reflect the politics, actions, etc. of any employer of mine, past or present, and any such similarity is mere coincidence.

Re:I'm confused

[FuzzyBad-Mofo]

Maybe after the FBI's Virtual Case File disaster, Microsoft Certified Partners will realize that you don't build a mission-critical application from the computer equivilant of Lincoln Logs.

EULA, DMCA and Reverse Engineering.

[sanityspeech]

From the article:

"In the past, what we'd have to do is reverse-engineering, and we had been getting into a pickle, because for open-source projects like Samba and OpenOffice, the only way to get the information was by reverse-engineering," he said."Pretty much for ALL the countries in the world, reverse-engineering was a perfectly fine thing to do."

Seeing that EULAs existed long before the DMCA came into effect, how on earth was it possible to develop a wonderful tool like SAMBA without some reverse engineering? My guess is some EULA(s) must have been violated. Surely, Microsoft could not have supported that.

IANAL, so enlightenment on this matter would be greatly appreciated.

Re:EULA, DMCA and Reverse Engineering.

[cduffy]

Let's say that I'm a UNIX guy. I don't own a copy of Microsoft Windows. I never agreed to their EULA.

I observe and reverse-engineer an over-the-wire file transfer protocol between two computers owned by my friends.

Now, tell me: How is any EULA violated? I never agreed to it in the first place, so I can't be violating it.

It's simple

[michelcultivo]

Sun dominate the Enterprise users and Microsoft dominate the low-end users with their software, one is trying to acquire knowledge from another. Very simple, even a penguin can see this.

I was there

[harikiri]

I remember James talking about the whole Microsoft/Sun collaboration. Apparently there is some confusion over what the legal agreement between the two is.

The main thing I remember him saying was that there are issues in working with MS, in that even if MS lets them have insider info on say their filesystem, they can't release this info to the Samba developers because of NDA's and IP licensing restrictions. So they have to be really careful and get signoff before they can open certain things up.

Another interesting discussion was the whole SWT vs SWING debate. James remained an advocate of Swing, and accused SWT of falling into the same traps that AWT had back in the day. From what he said, it sounded like he was saying that Swing is flexible and powerful enough to do whatever you want, but that was also its downside. An example he used was back when they were auditing Netbeans 3.6 to figure out why it was so slow. Apparently the developers had gone overboard with monitoring events, and a single drag of a window resizer would trigger thousands of events (an "event storm" he called it), which would also in turn spawn a bunch of "stormlets", small event loops (events triggering other events which trigger other events ad nauseum). Apparently this was the cause of the slowness.

One of the people who was asking a question of James asked the audience to raise their hands if they used Eclipse. I would guess that around 90% of the audience raised their hands.

When asked his opinion on the IBM vs SCO court case, his response: "I want some of what they're smoking". He didn't get asked about Sun's IP stance however.

I also have a picture that I took of the cake for the 10th anniversary of Java. It's sitting on my phone at the moment, but I saw some other attendees take snapshots too.

Sorry this is a little haphazard. I didn't really take notes. :)

Had me then dropped me

[KarmaBlackballed]

The last paragraph of the news story ends like this: In the hour-and-a-half session, Gosling answered many questions on a range of topics, including Eclipse and other Java IDEs (integrated development environments), DVD technology, security in Microsoft's .Net platform, the future of embedded software and more.

Only problem is the author thinks that's all we care to know about that. Sorta like writing "yadayadayada".

No need to actually report what his answers were. (Guess only an extreme geek like myself cares to hear what he said about these obscure technical topics.)

More is Really Less

[value_added]

Gosling offers a bit of insight when he says:

Reverse-engineering in the United States is now "legal for stuff, except stuff doing digital rights management," or DRM, he said. "So what has been happening is folks like Microsoft have been putting DRM into everything. DRM has been put into places you wouldn't think would make a whole lot of sense, like the document format being wrapped in DRM stuff...Under the sheets, the major justification is to make reverse-engineering illegal."

Bill Gates, on the other hand, offers a very different (albeit hardly suprising) point of view in a recent NY Times article.

``Over the years, our industry has tried many approaches to come to grips with the heterogeneity of software,'' Gates said, ``But the solution that has proven consistently effective -- and the one that yields the greatest success for developers today -- is a strong commitment to interoperability.''

Microsoft is also facing competition from Linux and other software that can be copied and modified freely. Proponents of such software say its flexible distribution makes it easier to design to work with other software.

Gates argued, however, that open source software encourages the proliferation of different software systems, making it harder to integrate them with other proprietary systems.

Many Microsoft products already work with other non-Microsoft products, and the company will build more interoperability into the design of its products, Gates said.

So, there you have it. Things are fine, and getting better.

I was there and asked him a couple of questions

[sonamchauhan]

I was there at this event and asked James Gosling a couple of questions.

"You spoke earlier about Jython and Ruby -- how Sun does not want to "choose" on the de-facto scripting language for Java.
Will Sun follow the lead of .NET - and now Perl6 - in supporting multiple languages that compile and run within the same virtual machine?"

I impression I got about his answer was: No, Sun won't publicly support multiple languages compiling to the JVM like Microsoft does in .NET (though he did not say this explicitly).

He reiterated the JVM did support multiple languages (the examples he gave were Fortran and Lisp) compiling to Java bytecode and running in the JVM. He said that the JVM architecture has constraints due to which languages like C/C++ cannot run in the JVM efficiently or safely. He said Microsoft actually made a big deal about their support for 'Managed C++' in .NET. He poked fun at this - saying due their support for pointers, etc, their Managed C++ implementation had security security holes "big enough to drive several trucks through".

"Follow up question: Will the JVM architecture ever change? The Parrot/Perl6 folks talk about how their new Register-Based VM architecture is inherently superior to stack based VMs. Any comments?" [Java uses a stack based VM ]
His answer boiled down to: "The Perl guys are wrong". He mentioned a few other complex points to justify this. An interesting thing he mentioned was that an early development version of the JVM used a register based VM "that no one other than me saw", and that he changed Java over to a stack based VM since the register based one "sucked so badly".

At the end of the event, the hosts (Sun Australia I think) brought out a big cake to celebrate the 10th birthday of Java. Gosling said that the day (Wednesday 2/2/05) was "uncomfortably close to the 10th anniversary of the first release of the JVM". The audience gave three hip-hip-hurrahs.

Doesn't Make Sense...

[Comatose51]

Usually pacts with the devil take on greater significance as you get closer and closer to death (*cough* Sun *cough*)...

News to who?

[gnovos]

I woked at a company a few years ago that was a newly minted "Microsoft Partner". Microsoft came in one day and we all had a big meeting. The rep told us, point blank, that they were developing the same software as us, but were a little behind. The deal was this: We'll liscense them all our software at a n unfathomably great rate, they'll promote our company for two years to thier other partners, and then they'll release thier own version of the software at that point, having built in all the improvements that they can from examining our software, and undercut us.

If we don't agree to the terms, they'll release thier software now and compete with us directly withotu the two-year gap.

So basically, that was life as a Microsoft Partner.