Tuesday October 17, 2006 ![[Duke Thinking]](http://mediacast.sun.com/share/tmarble/Duke-Thinking.jpg)
| « November 2009 | ||||||
| Sun | Mon | Tue | Wed | Thu | Fri | Sat |
|---|---|---|---|---|---|---|
1 | 2 | 3 | 4 | 5 | 6 | 7 |
8 | 9 | 10 | 11 | 12 | 13 | 14 |
15 | 16 | 17 | 18 | 19 | 20 | 21 |
22 | 23 | 24 | 25 | 26 | 27 | 28 |
29 | 30 | |||||
| Today | ||||||
LINKS
CONTACT
Tom Marble's Weblog
template by
Helquin
Tuesday October 17, 2006
ApacheCon 2006
Sun was present in force at ApacheCon 2006 last week... we
were about 16 in all including speakers on:
- Java DB - an implementation of Apache Derby: David Van Couvering, Rick Hillegas, Francois Orsini, Olav Sandstaa, and Rebecca Hansen
- The Apache Shale Framework: Craig McClanahan
- OpenSolaris: Narayana Janga
- Apache Roller: Dave Johnson
But then, of course, I wanted to install the latest version of OpenSolaris so that I could try out Chime and jdtrace. Thanks to Gary Collins and Randy Crihfield help in our Boston lab I managed to get the T2000 to cooperate. Now I'm actually dangerous with OBP and multi-boot disk partitioning!
To really show off the power of the box, however, I needed a big workload and the obvious choice was SPECjbb2005
( SPEC(R) and the benchmark name SPECjbb(TM) are trademarks of the Standard Performance Evaluation Corporation. For the latest SPECjbb2005 benchmark results, visit http://www.spec.org/osg/jbb2005 ).
Now with SPECjbb2005 running on JDK 6 on OpenSolaris I recreated the demo we did at JavaOne (hey, who forgot to save that thing?). For fun I also enabled a "perfbar" display to show that all 32 threads were busy as well as VisualGC to peer inside the the JVM as it is running!
Monday
Of course my big topic for ApacheCon was discussing the open sourcing of the JDK. And it turns out we didn't waste any time in getting into it with Geir and Tim at dinner. Geir, of course, is passionate on the subject of open source Java and how Sun really ought to do it (but more on that later...). Geir is even more passionate about foosball and so Tim (whom I found quite cordial and reasonable on Java matters) and Dave and I kicked back while we started what would turn out to be the late night routine at Buffalo Billiards.
Tuesday
As per Murphy's Law once I had the T2000 and all the equipment assembled here... it didn't work
Fortunately my goto-guy on tricky hardware issues, Phil Wagstrom, was able to help me out (and point out that secret connector/adapter). And luckily I had brought just the right combination of serial cables to connect to the system controller and get it to behave. In the Boston lab I had the T2000 set to DHCP and I brought my own WRT54GS to setup the private net between my laptop and the T2000. But, alas, the TSA machines must have fried the ROM as the WRT wouldn't show the slightest LED
Since I had a plain hub also I configured my laptop (running
Ubuntu) to be a DHCP server and that worked great.
But I'm so bummed that now I need to get a new box for
OpenWRT!
Then I had a little more work to convince OpenSolaris to
accept my passwordless SSH key to remotely start the demo
(since we run NIS in the lab, but obviously not here).
Got the demo working, finally!
Too bad I didn't get a screen shot of the demo in the booth (shown here):
![[Sun Booth]](http://static.flickr.com/108/267055277_e24521161e_d.jpg)
After getting the demo working the payoff was to join one of my former System Engineer colleagues from the dot-com era, John Ratliff, for lunch at the Hula Hut. John is now growing the biz for Network Appliance (hey, who gave John permission to leave Sun anyway ?!!).
Later the Sun team had a fabulous dinner at Dona Emilia's (chosen by Sara Dornsife and hosted by Simon Phipps).
Wednesday
The keynote by Cliff Stoll was absolutely fantastic (well, if you're a geek that is). Cliff demonstrated how he had his 8th grade students measure the speed of light and the speed of sound. He deliberately inserted a mistake into his talk and offered a hand made Klein Bottle to the first person to identify it. Alas the answer (which I missed) was that -- the speed of light is a primary standard -- you can't measure it!
The highlight of the day was the Sun unBOF -- a party exquisitely organized by Sara at Buffalo Billiards. The main event was Geir and Tom talking about open sourcing Java:
![[The Geir and Tom Show]](http://static.flickr.com/109/268116213_9f14f2298a_d.jpg)
I began the discussion by talking about the cooperation between Sun and the Apache Software Foundation and highlighted the contribution of the Harmony project of working towards compatible innovation. I mentioned the announcements Sun has made about open sourcing Java and highlighted the goals we have for encouraging innovation, expanding platform ubiquity and maintaining the promise of Java compatibility. Geir spoke about the progress of the Harmony project and the values espoused by the Harmony Community which is built around the AL2 license. Geir went on talk about the importance of maintaining a "level playing field" for contributors and the impact of the kind of community that Sun is interested in creating.
As we delved into the history of Java and the role that major players have played in the evolution of the Java platform I asked Simon Phipps to join us on stage. As the discussion continued we had many participants come up to the mic and ask questions and share their views, including: Greg Stein, Brian Belendorf, Sam Ruby, and Ted Leung.
Our interactive conversation lasted a whopping one hour and forty five minutes! Perhaps the shots of Petrone helped facilitate libre discussion (thanks Sara)! It seems that everyone had a good time. What meant the most to me is how many Apache developers came up to me afterwards to talk about the process Sun is going through with advice, empathy and encouragement. We will need, I'm sure, many course corrections along the way and I'm counting on the Apache community to continue this dialog.
Thursday
I thoroughly enjoyed the keynote by Patrick Ball from Benetech as he discussed the application of open source software for improving human rights in the world. In discussions with many friends we agree that using our skills and technology for a higher purpose, such as Patrick has done, is extraordinarily compelling.
Later I joined a high school friend, John Robinson, for lunch at Las Manitas:
John is now an expert in semiconductor manufacturing equipment and gets to travel to the four corners of the world -- wherever there are fabs -- as part of his day job. We hadn't seen each other in 20 odd years and yet we picked up the conversation and shared stories of other classmates we've seen.
Continuing the theme of local Austin attractions several of us watched the bats fly out from underneath the bridge at sunset -- it was an amazing sight! At dinner my colleague Max (Massimo Lanfranconi) and I got to talking wine and he told me all about the Italian (Red) equivalent to French Sauternes: Amarone. This ruined us for the wine we found at a party later, but will be a great excuse to open a great bottle next time we get together!
Friday
The surprise bonus on Friday came as I was about to board my plane at the airport. Max and I were hanging out when Noel Bergman struck up a conversation and soon Cliff Schmidt joined us. Noel shared his analogy of Sun's transition to open source as a revolution in the spirit of Kuhn's Structure of Scientific Revolutions. We proceeded to have a very frank, but cordial discussion on how the Apache Software Foundation could (and would like to) play a role in the open sourcing of Java as well as how Sun can best satisfy the myriad constraints of platform compatibility, innovation and keep the business model rolling.
The takeaway from ApacheCon is that I feel we have succeeded to communicate our goals and challenges for open sourcing Java and, at the same time, we have learned a ton about how to create a successful community. Now it's just a mere matter of practical work
Posted by tmarble ( Oct 17 2006, 12:03:35 AM CDT ) Permalink
Comments:
Post a Comment:
Comments are closed for this entry.