Thursday Oct 28, 2004

Great Minds Think Alike (Fools Seldom Differ?)

It must be the lunar eclipse. I'm agreeing with a lot of statements being made in the industry. Thought a few of them were worth noting.

Sun and Microsoft agree that software pricing should not differentiate between single core and multi-core cpu's. Which is why we are moving all Sun's software offerings to a single, common, per "socket" pricing format, instead. We're counting sockets, not cores. But truth be told, even that's going to be tough to do going forward. Audits are painful for everyone.

Longer run, what better meter for usage than the number of employees you have (as listed in your SEC filings)? PeopleSoft has it right - if all employees will use your products, price on a per employee basis - all businesses count employees (even tiny businesses). And Sun's software is of the nature that it's used not only by all employees, but we tend to serve businesses that extend our software to their customer bases - employees and customers both authenticate to directories, view the web through web servers, run applications against app servers (especially at companies that rely on the internet). So why bother with two prices - Sun actually waives pricing for external usage - pay $100/employee, and external deployment is free of charge. Free. Free free free. (We're becoming great friends with procurement officers the world over :) (And by the way, if you're fewer than 100 employees, the products are just plain free - for internal and external usage.)

But back to the top level point, like I said, pricing by the core in a cpu is nonsensical - especially with Sun's whomping 32-thread systems on the way (and what's beyond that, I'm not even sure we'll know how to count). Speaking of which...

Sun and Intel agree that multi-core computing is the wave of the future, and that their gigahertz race is dead. Granted we had that viewpoint approximately 3 years ahead of Mr. Barrett, but we're glad we're finally in agreement. If the world of web services is multi-threaded, and your operating system eats threads for lunch, your microprocessor may as well be optimized for multi-core/multi-thread workloads. And speaking of web services...

Sun and the Apache Software Foundation and JBoss agree on integration standards for web services. Two beacons in the open source software community just signed on with Sun, and a host of other partners (including SAP, Oracle, Novell, Sonic, others), to the JCP's Specification Request 208 (known affectionately as 'JSR208').

What's remarkable is that just last week, IBM pulled out of JSR 208 - out of a fear their proprietary approach would be cannibalized by an open standard. IBM, love to have you back... sure would hate to see your approach to Service Oriented Architectures garner a PROPRIETARY warning label, while the open standards and open source community leave you behind.

Remember, customers are in charge. And no matter how distasteful they find per-core software pricing, they find being locked in even worse.

Share this post  del.icio.us | digg.com | slashdot.org | technorati.com | reddit | facebook | stumbleupon

Thursday Oct 21, 2004

On the Importance of Constituencies

One of my favorite stories at Sun, and one of the most illuminating about our business, surrounds a series of discussions I had with an industry analyst at JavaOne two years ago. He started by saying:

"My daughter loves Java."

What? The only folks I'd really met that loved Java technology didn't bear much resemblance to a 12 year old girl. So I was a little skeptical, and asked "why?"

He told me, when he came home from work, he emptied his pockets, and his daughter wasted no time in grabbing his phone to start playing games. Which were primarily written to Java, displaying the logo as they launched. She'd come to recognize the best games, and the new ones her friends talked about buying from their carrier, were written with Java. I'd never thought, frankly, about that particular audience - but that's partially what led one of our teams to create java.com (which I assure you, talks to a non-traditional Sun demographic).

Now, I had a parallel set of interactions at this year's JavaOne, at which a bunch of friends joined us for a discussion on the open sourcing of Java. Among the luminaries present was Brian Behlendorf, who opened his statements by asking what I'm sure he felt was a question with a popular answer, "How many of you work on an open source project?"

I expected to see a flurry of hands, and I'm sure he did, too.

Neither of us saw hands go up.

The community represented at JavaOne either worked within the Java Community, or were developers with other issues on their minds (like their day jobs). Interesting.

Building on the 'unexpected diversity of audiences' theme, I was keynoting a CIO event in Cincinatti a few weeks back. The event was attended by a cross section of American companies, from retailers to pharmaceutical companies, logistics and airlines. Toward the end of my prepared remarks, I started previewing the open sourcing of Solaris (and our Red Hat upgrade programs, just for fun). One of the CIO's stopped me to ask, "why are you open sourcing Solaris? The last thing I want is more source code." My response, "No offense intended, but you're not my target demographic. It's your developers, and they'd love the ability to see/evolve the source."

The thread tying these stories together is simple - there is no single definition of 'user' that encompasses the diversity of the constituencies we serve, or our means of doing so. From 12 year old girls, to system administrators, to CIO's, to naval officers, the folks on groklaw, to sell-side analysts (one of whom, I'm flattered to say, recently quoted my blog without attribution - I'm not holding my breath for a royalty check). Sun's audiences are as diverse as those watching TV or listening to the radio. (Which makes perfect sense - what is the 'net but the logical evolution of broadcast media.)

But looked at in reverse, what about serving a single constituency, a Java developer? Note that with the Tiger release of J2SE, the newest NetBeans gathering momentum (and Eclipse converts), and the unveiling of Java Creator, each product uses a different development and licensing model, appropriate to its objectives. J2SE is the result of an extraordinary collaboration between a vibrant and inclusive community, the most pervasive on the net (just go check out who belongs to the Java Community Process). NetBeans is the product of a traditionally defined open source community, churning out enhancements under a vastly different governance model. And then there's Java Studio Creator, built by Sun, just by Sun, as a means of driving to market a Java development tool for fans seeking an open, cross-platform alternative to Visual Basic.

Three different products, three different styles of building and delivering innovation. One fundamental audience.

Which all goes to say, there are those that persist in trying to draw the industry as filled with binary extremes. (Mr. Epstein (and his peers on the opposite side), I have a gift for you.) I choose to see it differently - the network reaches a market so broad, there can never be one definition, one product or one market.

Nor will there be one opportunity. It'll all depend upon the constituency you're serving.


Share this post  del.icio.us | digg.com | slashdot.org | technorati.com | reddit | facebook | stumbleupon

Wednesday Oct 20, 2004

Customers are in Charge

If you ever doubted my assertion that customers are in charge, try reading this. (And this just for color.)

I believe Sun now has the industry's most complete line-up of Liberty enabled Identity Offerings - and it's obvious that the Liberty standard is in demand by customers seeking to secure and simplify identity interoperability (identity being the most fundamental of all web services). It's also obvious IBM had to follow the market, or lose revenue. Customers spoke.

What's up next? We're signing up more ISV's for Solaris 10 every day. On SPARC and AMD Opteron. More and more customers are seeing huge performance, utilization and expense savings - on Dell hardware, on Sun's Opteron systems, as much as Sun's SPARC hardware.

So when you talk to your ISV's and partners, be sure to ask for their Solaris 10 shipdate - it's the single biggest improvement we've ever delivered in a commercial operating system. And gives you an opportunity to run all your existing Solaris apps and all your linux apps unmodified, get logical partitioning (driving massive utilization improvements), fault management and extreme performance improvements - for less than the price of "free".

And with Oracle, BEA, Veritas, Sybase, CA and Siebel - and more to be announced on November 15th - signed up and committed to shipping on Solaris 10/x86 and SPARC - there's only one ISV that's not listening to customers...

And finally, speaking of innovation to solve real world problems, I made my donation, did you? Talk about helping customers solve real-world problems...

Share this post  del.icio.us | digg.com | slashdot.org | technorati.com | reddit | facebook | stumbleupon

Monday Oct 18, 2004

Putting Sun's Kodak Settlement in Context

There seems to be some confusion surrounding Sun's settlement with Kodak (which clouded what otherwise would've been very positive Q1 news last week, including (non-GAAP) positive operating income, margin improvement, etc.) - so I thought I'd provide some clarity (see www.sun.com/investors for full financials and GAAP/non-GAAP comparisons).

First things first: nothing's changed. In terms of patents, and intellectual property, nothing whatsoever about the Kodak settlement suggests Sun's position on IP has changed. Moreover, the settlement was all about ensuring what Kodak was doing to Sun, they could do to no one else in the Java community.

But let's get the issues out on the table.

First, this settlement is proof we're committed to protecting our customers and communities.

How? As a result of this settlement, the Java community, our customers, licensees and shareholders, have now been freed from worrying about Kodak's litigation. We paid $92M (not $82M) to protect our constituents from their lawyers. Do we believe their patents are invalid? Yes (just crack a Smalltalk textbook). Was it worth having this suit hang over our heads, no. Absolutely not. That's why we settled - not to validate Kodak, not to validate those patents, but to let our customers and employees and stockholders focus on market opportunity, not litigation.

Second, nothing's changed in our policy regarding intellectual property.

We still believe IP is what keeps our industry alive, and we still believe its the most durable asset created by Sun. And a $92M Kodak settlement is just that, a settlement to get this behind us. So that we can get back to creating IP for a networked world, and stop flying back and forth to Rochester, NY. Headquarters of the Kodak Corporation. Home to most of their employees, and an economy reliant upon the revival of Kodak for an economic turnaround. Did I mention the trial was heard by a Rochester jury?

Third, patents still matter.

And moreover, our settlement with Kodak shows exactly why - when confronted by an aggressor, we want the ability to battle back. To settle for $92M, not for a billion. But bear in mind, we've shied away from being a patent aggressor. We've always, except when confronted by clear and well established transgressions, relied upon innovation as our competitive weapon, never litigation. (And we've always been open to new ideas.)

Now, with all that said, I'd like to make a few more points.

Are software patents without controversy?

No, not at all. Some systemic changes need to be made. But the assertion that patent coverage should be eliminated from the software industry makes no sense.

There's this great irony: spend time with Sun's microprocessor team, and guess what - they're all writing software. Spend time with our data center switching team - they're all writing software, too. Viewed simplistically, computing hardware is software burned into and onto physical things. And over time, more and more routine software elements end up in hardware, for acceleration or optimization. SSL accelerators, JVM on a chip, you name it. So, where do you draw the line on patents? Firmware? FPGA's? Silicon? Systems?

Yes, we should reform the system to stop granting spurious patents. And yes, we should disincent spurious litigation. That's what we're advocating in Europe, before the legislature there settles on the worst rather than the best of US practice. Eliminating patent coverage to benefit copyright-only companies is imbalanced - but safeguards are needed if the system is to protect innovation rather than allow predators to tax standards. Safeguards, I agree, that aren't in place today.

Finally, don't believe those that are playing the community.

What do I mean? For IBM to claim, "as an ally that believes in the positive power that the Linux community is having on collaborative innovation, I can assure you we have no intention of asserting our patents against the [GNU]Linux kernel..." is a throwaway comment. Communities and kernels and individuals without assets are not useful targets for IBM-style corporate litigation. IBM can't get the billions they do in IP licensing by suing Groklaw's readership (no offense, PJ), or even Linus himself - so their comment is a shirtsleeve off their vest. For the real money, IBM would have to go after corporations.

So how about IBM starts naming the companies to whom they'll issue their patent amnesty? Let's start with Red Hat. If IBM's not going to sue Red Hat, then shouldn't IBM state that in writing? MySQL clearly spawned from collaborative innovation, will IBM grant patent amnesty to them, too? In writing? IBM, demonstrate your commitment to the community, start documenting the companies to whom you're issuing patent amnesty. Love to see the list.

Comments about "we won't sue the kernel" or "we won't sue the community" show the cynicism with which IBM is pursuing its marketing initiatives. They're playing the community like a violin, not protecting it.

Novell's announcements? They got it right - whether a product is open or closed source is a sideshow. What matters is whether a company will stand behind its products. As we did in settling with Kodak. We not only protected our products, we protected our customers. Further, we protected the Java community from litigation. But even Novell's policy has limitations - they'll only protect their enterprise servers, not their developer communities. Will they protect JBoss? MySQL? I doubt it (and I wonder why they referenced them in their press releases?). But no company's coffers are deep enough to protect the universe (read the last two sentences in David Berlind's cogent analysis).

The net of which is, investing in intellectual property is how Sun invests in our future. Protecting our customers is a part of our commitment to them, to stand behind what we do, in open source, closed source, community source, binary, streamed formats or otherwise. That was the strategy before the Kodak settlement, that is the strategy after the Kodak settlement. Don't believe the conspiracy theorists - or the companies that don't stand behind their IP - who say this was all about setting up Kodak to attack the rest of the planet.

That's as spurious as the litigation.

_________________

Given the volume of emails folks have been sending to me (misaddressed) regarding the invalidation of Kodak's patents, please use jonathan.i.schwartz@sun.com. Thanks for the help, much appreciated.


Share this post  del.icio.us | digg.com | slashdot.org | technorati.com | reddit | facebook | stumbleupon

Tuesday Oct 12, 2004

Happy Birthday OpenOffice.org!!

OpenOffice.org celebrated its fourth birthday today. Over tens of millions of downloads, OpenOffice.org received an incredible endorsement -- the European Union recommended OpenOffice.org as the basis for standard file formats and document interchange. Chalk one up for open standards. The competition's on the run.

Sun's contribution of StarOffice to the open source community was one of the single largest open sourcing efforts in the history of the industry. Ever. Granted, I haven't done a line count in a while, but it's a massive codebase, managed and evolved by a global community of folks interested in enhancing, evolving, innovating and localizing the fastest growing office productivity suite on the market. Millions of lines of code managed by a global team, committed to affordability, universal access, and security.

No matter what the cynics say, since its inception, Sun has been a believer in and contributor to open source communities - after all, we were built on an open source operating system. Open source is in our blood, not just our press releases and billboards.

Congratulations, folks. No one can possibly doubt that we're making a difference.

Share this post  del.icio.us | digg.com | slashdot.org | technorati.com | reddit | facebook | stumbleupon

Monday Oct 11, 2004

Kangaroo is Delicious (and the end of my political career)

For the past few days, I've been in Australia meeting with customers, and hosting along with our sales leadership our annual SunRise event (a training/celebratory event for our most committed folks). Other than the challenge my flight faced on the way over (we ran short of fuel and had to land in Brisbane to get more), it's been a really enjoyable trip, meeting with customers, press and analysts.

I was at a customer dinner last night, hosting a series of CIO's and CTO's from Australia's largest businesses. As always, I find their stories and challenges fascinating - the CTO for a leading financial institution told me "we're doing more transactions on-line today than we do in all 1,000+ branches we have in the world combined." Wow. I met with a leader in government whose organization delivers live traffic video to citizens wanting traffic reports. On mobile handsets. And this was a government agency (the US should take notice). He was more facile with issues of content distribution and rights management than most technologists.

I've also had a chance to meet with a few of HP's largest customers - many of whom seem to share my anxiety about the fate of HP/UX, and many, courtesy of HP (and Computerworld Australia), have read my perspectives on the challenge HP faces without an operating system. I love the internet, communication can be so efficient (stay tuned here if you'd like to read the cease and desist letter, and our formal response clarifying HP's strategy to HP's staffers and their entire user community).

The title of my speech at a local chamber of commerce was "Why I Believe the Dot Com Bubble Was Just A Proof Of Concept." Where I reviewed some of the more interesting trends we see in the industry - the commoditization of the market bringing massive new volumes, the explosive growth of market opportunities (and share values), the realization that Henry Ford's approach to mass production (you can have any color, so long as it's black) is the right approach for the bulk of business processes; and the need for businesses to take PC security as seriously as Telstra takes handset security. It was an exceptionally friendly audience, all of whom agreed that the 'net has become way more central to business today than ever before - and that a new wave of value will accrue to those that invest in R&D and intellectual property.

Now, on a personal divergence to what is an otherwise very business oriented blog...

I've been to a ton of places in my line of work, and so it's not without some experience that I make the following recommendation: if you haven't visited Australia, you should. It's a gorgeous country. As a Californian, I don't make that statement lightly, either. We Californians tend to be a bit parsimonious in doling out the word 'gorgeous.' But Sydney's got the climate of Seattle, the cityscape of San Francisco, the endless vistas of Texas, and a population with an addictive sense of humor. What more could you ask for. Food, and wine? Bad news, California, Australia's got both.

And because I don't want it to come out on some tech tabloid tell-all, I would like to inform everyone that reads my blog that I did, in fact, taste kangaroo meat at a luncheon yesterday. I feel bad saying that, I hope my neices and nephews don't find out about it, but I tasted it. And I know this will likely disqualify me from public office at some point, but I need also confess, I enjoyed it. It even paired well with a good shiraz.

There, I said it. Secret's out.

Share this post  del.icio.us | digg.com | slashdot.org | technorati.com | reddit | facebook | stumbleupon

Wednesday Oct 06, 2004

Calling out the Cavalry

Interesting read.

For the first time ever, it feels like the desktop and office productivity marketplace is becoming competitive. Really competitive.

And the good news about the internet is we can send copies of StarOffice and OpenOffice to far more places than they can send those awkward men on horses.

Share this post  del.icio.us | digg.com | slashdot.org | technorati.com | reddit | facebook | stumbleupon

Tuesday Oct 05, 2004

Java Technology in Politics

The perfect gift for your favorite fund raiser.

Share this post  del.icio.us | digg.com | slashdot.org | technorati.com | reddit | facebook | stumbleupon