Monday November 16, 2009

I'm back in lovely Antwerp for Devoxx. The purgatory we're in over the situation with Oracle has it's pluses and minuses. On the plus side, I don't have to do a keynote.... Steve Harris from Oracle get's that job. I will be doing a talk, but I'll be concentrating on the store we're in the process of launching. The hard part is that the only questions that anyone will be asking are the ones that neither Steve nor I can answer: until the acquisition clears the EC competition commission and closes, we're required to be mostly silent about the future. We're pretty much limited to the official statements.

I'll be spending the end of the week at Cartes, talking with the SmartCard crowd. Between the 10th anniversary of GlobalPlatform (12 years of JavaCard!) and the release of JavaCard 3, it's a big year for us in the SmartCard world. Permalink Comments [5]

Tuesday November 03, 2009

Put an accountant, a lawyer, an MBA and a software engineer together into a room... Sounds like the lead-in to a bad joke, but it's the exercise that the Java Store team has been living through for the past several months. At the PayPal conference today Eric Klein did an announcement and demo of the next phase in the Java Store's development. We've been working with PayPal on this for some time, using their new PayPal X platform. It always amazes me how complex it is to deal with all the details of global finance. And even so, the store today only handles US issues. But the framework is in place to go global as fast as the lawyers and accountants can work through the details — but it'll take a while. There's a new client application for shopping in the store, and a new warehouse site for developers to upload products.

Even with the current US-only restriction, that's about 65 million desktops for a target market. Please check it out, kick the tires, let us know what you think: we'd like to get it out of beta and do a real large scale consumer launch as soon as we can. Permalink Comments [7]

Thursday October 29, 2009

Yet Another Happy Birthday Intertubes!! Today marks 40 years of the internet, although there's some debate as to the actual date. I consider myself a latecomer: I didn't get my first real internet email address until 1977, C410JG40@CMUA. I was "jag" on various Unix systems before then, but it wasn't until 1977 that the ARPAnet and email really took over my life. I soon realized that the only real-world friendships I kept up with were folks that I could send email to. I disappeared from my brother and sister's lives until they got email addresses 20 years later. Of course now it's gotten to the point where restaurants don't exist unless they're on OpenTable :-) I wrote the original Unix Emacs in 1978 and because of that by sometime in 1980 I had the unusual distinction of having login IDs on every non-military host on the ARPAnet (I kept track - it had become sport).

When I joined Sun in 1984 (yikes! Has it been that long?) one of the big attractions was Sun's position on networking: every machine had a network connection. At the time, that was considered pretty weird. I had known Bill and Andy for years. They had both tried to get me to join Sun at the beginning in 82, but I foolishly didn't. "The Network is the Computer" was a pretty odd tagline at the time, and it didn't make sense to most people, but it had geeky appeal. A lot could be done when you tied machines together: from harnessing the compute power of clusters of machines, accessing filesystems remotely via NFS, remote graphics access, and a whole lot more.

But the network had more than mere geek appeal for me. It felt important in a world-changing kind of way. This crystalized when I read Robert Axelrod's 1984 book The Evolution of Cooperation. This is a reasonably accessible game theory book that discusses experiments involving the Prisoners Dilemma. I won't repeat what you can find by reading the Wikipedia article (which I urge you to do; read the book too). But by the end of the book, there is a strong conclusion that as the frequency of interaction increases, the optimal strategy shifts from hostility to cooperation. This really appealed to my 60's peacenik leanings because it suggested that all you needed to do was to get people communicating, and inevitably peace would break out. It doesn't matter much the form the communication takes, all is good.

At that point in my life, the network went from a geeky toy to a moral good. I don't know how to express how thrilled I am at how it has all played out. From the effects of the network in keeping the news flowing during the 1991 Soviet coup d'état attempt, to FaceBook, Twitter, Blogging and all of todays social media.

These days I think that Sun should shorten its tagline from "The Network is the Computer" to simply "The Network Is".

My biggest disappointment with the internet is that it seems that the "killer app" that makes the economics of the Internet work is advertising... Permalink Comments [8]

Monday October 26, 2009

The JavaCard team have been cranking away. Development on the 3.0 version is finally (almost) finished, and it's pretty amazing. Java Card 3 is available in two Editions.

Classic Edition

This is the same as Java Card 2 with some enhancements/bug fixes. It is almost 10 years young and is the most popular platform for the SIM and ID markets.

Connected Edition

This is the next generation Java Card technology:

• JDK6 Compatible VM: Except for floats, it support class file version 50.
• Full Java Language support: Java Card 2 has restrictions on the language itself. But JC3 has no limits. You can use all language features like annotations, enhanced for-loops etc... (except floating point)
• Rich API: This is mixture of CLDC, GCF, Servlet, JavaCard2 API, Sockets, Threads, Transactions ...
• Three application models and two library models, which makes it possible to have virtually any kind of secure application on JC3:
  • Servlets, extended-Applets, Classic-Applets
  • Extension-Library and Classic-Library
• Servlet Container with Servlet 2.5 support.
• HTTP and HTTPS interface: No need for special client programming. Use any web client to reach JC3.
• Still tiny(!!):24K RAM, 128K EEPROM, 512K ROM with a 32 bit processor
• It is not just "Card" any more: With the newly added USB interface this technology can go beyond Smart Cards into devices like secure USB tokens, Secure Personal Databases, Embedded Servers, WebDAV compliant thumb drives and more.

Last but not least, there is a Netbeans Plugin for easy development. Nightly builds of NetBeans point to the latest JavaCard plugin.

Tuesday October 20, 2009

I ripped the little demo map browser component out of my Oracle OpenWorld slides and moved it to kenai as a new project called OSMBrowser. Not very polished, more of a starting point for someone motivated to play :-)

Thanks to the crew at the Open Street Map project for a nice database and tile server. A Thing of Beauty.

Update: I fixed the busted .jnlp file, so it can be run. Permalink Comments [9]

Friday October 16, 2009

Several folks have asked for copies of my slides from Oracle Openworld. Unfortunately, there's no printable form of them, since I did them as a JavaFX app. You'll find them at http://fxslideshowtest.kenai.com which will launch the app (with all it's rather large images) via JNLP. If you're curious about the sources, they're on kenai. The code is pretty ugly: I just slapped it together. I'm not proud :-) The code for the map browsing component is in there too. It uses the tile server from openstreetmap.org (Click and drag with the mouse to move, scroll wheel to zoom). I kinda like the map browsing component, so I'm cleaning up the source and I'll push it out to kenai as a separate project sometime soon.

I would have published it earlier in the week, except that there's a bug on Snow Leopard where webstarted apps that use fullscreen get a bus error :-( Works fine in fullscreen as a regular desktop app, but something weird happens when webstarting. Works fine on all other platforms. I gave up trying to figure out what was going on and just dropped the fullscreen startup. Permalink Comments [10]

Thursday October 08, 2009

Next week should be real interesting: I'm spending a big chunk of it at Oracle's OpenWorld conference. I'll be helping out Scott at his keynote. As usual, I expect Scott to be fun :-). I'm doing a talk the next day down the road at the Hilton where Oracle is holding their Develop conference. Sun is leading a number of sessions on enterprise topics. I'm going to try to stretch people's minds about what "enterprise" means. Permalink Comments [3]

Thursday September 10, 2009

It's decades late, but a lovely gesture none-the-less: Gordon Brown, the Prime Minister of Britain has formally apologized for the treatment of Alan Turing. It's hard to overstate his impact on the latter half of the 20th century, and the discipline of computer science in particular. Think of what your life would have been like without him: a longer and more devastating second world war; and the development of computers delayed by many years. How much better could the world have been if his life had not been cut short by thoughtless intolerance? For details, as always, consult wikipedia. Permalink Comments [3]

Saturday August 29, 2009

I don't know what it is about Apple and NFS, but they keep moving things around. The new UI to NFS mounting is much nicer than it was before, but it's now in a totally different place: the Disk Utility. But if you use a lot of NFS file systems, it's a pain to have to mount them one by one: ignoring the UI and using the /net automount filesystem is far more convenient. Just use the file name /net/hostname/path and you don't have to mess with any mounting, it just happens by automagic. I wrote a blog entry about this a long time ago.

However, there is a huge problem with this: OS X does a phenominal amount of file locking (some would say, needlessly so) and has always been really sensitive to the configuration of locking on the NFS servers. So much so that if you randomly pick an NFS server in a large enterprise, true success is pretty unlikely. It'll succeed, but you'll keep getting messages indicating that the lock server is down, followed quickly by another message that the lock server is back up again. Even if you do get the NFS server tuned precisely the way that OS X wants it, performance sucks because of all the lock/unlock protocol requests that fly across the network. They clearly did something in Snow Leopard to aggravate this problem: it's now nasty enough to make NFS almost useless for me.

Fortunately, there is a fix: just turn off network locking. You can do it by adding the "nolocks,locallocks" options in the advanced options field of the Disk Utility NFS mounting UI, but this is painful if you do a lot of them, and doesn't help at all with /net. You can edit /etc/auto_master to add these options to the /net entry, but it doesn't affect other mounts - however I do recommend deleting the hidefromfinder option in auto_master. If you want to fix every automount, edit /etc/autofs.conf and search for the line that starts with AUTOMOUNTD_MNTOPTS=. These options get applied on every mount. Add nolocks,locallocks and your world will be faster and happier after you reboot. Permalink Comments [14]

Thursday August 06, 2009

[cross-posted from the JVM summit mailing list] This is a reminder for the 2009 JVM Language Summit to be held at Sun's Santa Clara campus on September 16-18, 2009. Registration is now open for speaker submissions (presentations and workshops) and general attendance. More information is available at http://jvmlangsummit.com.

The JVM Language Summit is an open technical collaboration among language designers, compiler writers, tool builders, runtime engineers, and VM architects. We will share our experiences as creators of programming languages for the JVM and of the JVM itself. We also welcome non-JVM developers on similar technologies to attend or speak on their runtime, VM, or language of choice.

The format this year will be slightly different from last year. Attendees told us that the most successful interactions came from small discussion groups rather than prepared lectures, so we've divided the schedule equally between traditional presentations (we're limiting most presentations to 30 minutes this year) and "Workshops". Workshops are informal, facilitated discussion groups among smaller, self-selected participants, and should enable "deeper dives" into the subject matter. There will also be impromptu "lightning talks".

We encourage speakers to submit both a presentation and a workshop; we will arrange to schedule the presentation before the workshop, so that the presentation can spark people's interest and the workshop will allow those who are really interested to go deeper into the subject area. Workshop facilitators may, but are not expected to, prepare presentation materials, but they should come prepared to guide a deep technical discussion. We will be finalizing the program by August 19.

The Summit is being organized by the JVM Engineering team; no managers or marketers involved! So bring your slide rules and be prepared for some seriously geeky discussions.

The registration page is now open at: http://registration.jvmlangsummit.com/

If you have any questions, send inquiries to inquire AT jvmlangsummit.com.

We hope to see you in September! Permalink

Wednesday July 29, 2009

It's been a really weird summer, mostly in a good way. Easily the weirdest summer of my professional career. A lot of stuff has been going on, much of which would make good blog material, but none of it can be blogged about. The Oracle transition can't happen fast enough.

I have spent some time on a silly but fun project: moving the Duke artwork from duke.dev.java.net to kenai.com/projects/duke. Along the way I added a bunch of new artwork collected from assorted places. Duke's new home uses a new kenai feature that's in early alpha testing: the ability to associate a WebDAV filesystem with a project. In this case, it's http://duke.kenai.com, or via webdav at https://kenai.com/website/duke

Among the new artwork is the giant billboard/banner that appeared at Moscone for the last JavaOne. The "gigantic" image is 28000x6400: don't bother downloading it unless you have a really big printer.

Another interesting one is the artwork for a tshirt that was never produced that was intended for an End of an Era party for Sun employees at Great America. The party never happened. The artwork is kinda fun: it features Duke having a beer with a character based on a Sun100 workstation - sun's first computer. If you look at the high resolution version, you'll see that the Sun100 is running OpenSolaris :-) If you want a shirt or mug made with it, I put it up on cafepress. Permalink Comments [8]

Monday June 15, 2009

I'll be spending next week in Zurich at Jazoon'09. They've got a great lineup of technical sessions to pump your head full of all the latest everything. The lineup of speakers is pretty impressive.

PS. Several folks have asked why I disappeared from JavaOne so quickly after my Toy Show keynote: my youngest daughter's school was holding a talent show at noon, and I had to be there. Work is not at the top of life's priority list :-) Permalink Comments [1]

Monday June 08, 2009

This was another amazing JavaOne. It was also the weirdest: between the Oracle situation, the global meltdown, and the financial situation, it was very different. Early on, we were really concerned (==nearly paniced) that no one would show up. Almost every company that usually sends a crowd of people to JavaOne had travel restrictions that meant that few could attend. From what we had seen from other conferences, we feared the worst. And yet, attendance ended up being about 15,000. Very respectable!

The press, of course, was weird this year: we did few interviews, mostly because there was only one topic they wanted to ask about, and there was nothing we could say.

My favorite part of JavaOne this year was the standing ovation that Scott got on Monday. It was a remarkably emotional moment for everyone. To all of you who contributed to that thundering applause: Thanks!

Our lineup of new technologies was great: JDK7 is looking wonderful; JavaFX reached another release milestone and is maturing nicely, along with it's tools; glassfish V3 is becoming totally wonderful; NetBeans 6.7+kenai is a killer combination; and the store is on it's way to being an amazing addition.

The fun part, for me, is the Toy Show. It's not about strategic initiatives, business issues, corporate relations, or new releases. It's all about jaw-dropping surprise. Everything from the Mifos (MIcroFinance, Open Source) project: bringing millions out of poverty. To Manuel Tijerino's jukebox architected to give indie musicians broad exposure and a source of revenue independent of the studios. To teaching kids through BlueJ and robotics competitions. And saving lives by doing computationally intensive image matching in an EE server with a JavaFX front end to analyse cancer biopsies. And the total victory of technology and style over common sense that is Neil Young's LincVolt. And a whole lot more. Permalink Comments [4]

Monday June 01, 2009

JavaOne is finally beginning! Getting ready for this one has been an incredible amount of work for everyone involved. The lineup of sessions is superb. I had a *really* hard time picking the Duke's choice winners (they all deserved to win). The engineering teams have been doing grand acts of heroism to get a pile of software releases ready. The demos have come together beautifully. Some are a little too close to the bleeding edge, but that's what makes them fun. The big surprise for everyone is attendance: given what we had been hearing about attendance at other conferences, we braced for the worst.... But we've done rather nicely. I'm *so* looking forward to the Toy Show on Friday. There's some jaw-dropping stuff to see. Permalink Comments [4]

Wednesday April 29, 2009

One of the many interesting things we've been working on lately is the Developer Cloud. There are two major components to it. One is the cloud infrastructure itself: Kenai a much-more-than-a-forge collection of developer facilities that allows you to assemble project areas from a selection of services that range from several SCM systems, bug management systems, wikis and forums. There's a lot more stuff in the pipeline for Kenai, but it's pretty impressive already. The other major component is the tool support that is showing up in NetBeans 6.7. The beta was just released. I've been running the "dailies" for quite a while now. It's real impressive: there's a lot more to NB 6.7 than the developer cloud, but the cloud support is the standout feature. We're just beginning, but it's already transformed the way I work.

If you're one of those neolithic vi users, clinging to the banging-together of rocks: Stop! Go to NetBeans now! Permalink Comments [3]