| « July 2008 |
| 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 | 31 | | |
| | | | | | | |
| Today |

Wednesday June 25, 2008
Nice article about ZFS and Apple
I ran across a nicely-written
blog post about
ZFS today, talking about how Apple announced it will be using ZFS in an upcoming version of its
Mac OS operating system.
The article points out that the advanced but easy-to-use ZFS filesystem protects against disk failures that happen in all kinds of computers (personal computers and enterprise servers both), and how that will provide Apple with a competitive advantage over Microsoft in the PC market. Interesting take; makes sense to me.

Friday May 09, 2008
The coolest thing I saw at JavaOne 2008 was...
...the
Livescribe Smartpen, which comes with a
Java-based development kit. The pen is cool: it's a pen with ink, but it also has a tiny camera in the pen-tip that is watching what you write, 72 times a second. It records all of your writing strokes, and also has a microphone built into it so it can record what you hear, too. It's great for taking notes, in ways that is much easier to understand when you try it yourself. But I'll try to explain just part of it here.
So, suppose you're in class or a meeting, and you're taking notes as people are talking. The Pen comes with a notebook of paper with little pictures at the bottom; those pictures act as buttons to do things like start/stop recording, change volume, or select something from the pen's menu of applications. So you start recording by tapping the pen-tip on the "Start" picture at the bottom of your notebook. You just start writing your notes as people are talking, and when you're done you tap on "Stop". When you reply the audio (the pen has a little speaker in it and also a stereo headphone jack), the pen begins playing back the audio from the beginning. So far, so good.
But get this: if you tap the pen-tip on any of your notes, the pen starts playing the audio from where it was when you wrote that note! It keeps track of when you wrote each pen-stroke, so if you want to playback a particular section of audio, just write something, anything, on the paper, and during playback just tap on whatever you wrote. It's like an audio bookmark. Very cool.
The notebook paper is just paper, but it's imprinted with a tiny, almost indiscernable pattern that helps the pen figure out where on the paper it's drawing. That lets you do things like one of their sample applications, which is to make a piano out of the paper. You start the piano app and the pen asks you to draw some vertical lines on the page, then two horizontal lines to make the vertical lines into a keyboard, then two letters beneath the keyboard you drew (one to let you select an instrument, the other to let you select a rhythm pattern). Then to play the piano you just drew, you simply tap on the "keys". It's just plain fun!
And it's a Java device (Java ME), so you can write your own Java apps for the pen. Livescribe had a contest at the show this week for whoever wrote the best pen app.
It's a damned cool device, and at the show they were charging $139 for it. I don't even like writing, and I want it!
I'm not doing the thing justice; you gotta go to the website and
check out the videos and demos. Really, though, you should just check it out in person. I haven't even mentioned the stereo headphone-microphones, just another of many cool features of the pen.

Thursday May 08, 2008
Virtualization at a JavaOne Keynote
I'm watching
Douglas Fisher's keynote at
JavaOne right now, and he's showing a slide that shows how Intel wants to participate in every layer of the software stack, from BIOS to applications.
One layer he puts on his slide, above firmware and below operating systems, is virtualization. It's the first time I've seen all the major players laid out. The slide has the following:
Xen,
Sun xVM, VMware, Oracle, Citrix, Microsoft.
I think that's good for virtualization, and should make it clear that VMware actually has some work to do to remain competitive (or dominant).
The next couple of years in virtualization is going to be really interesting.

Wednesday May 07, 2008
OpenSolaris available on Amazon.com's EC2
I'm at the JavaOne conference in San Francisco this week, which actually started with something called
CommunityOne, a Sun-sponsored event for developers focusing on open source and web developer stuff.
The coolest thing I saw yesterday at CommunityOne was a talk about
Amazon Web Services and
OpenSolaris. The talk gave a great overview of Amazon's services for web developers, including their compute service, storage service, and some of the new stuff they're rolling out. It was impressive.
During the talk, Amazon announced that OpenSolaris is now available as one of the operating system platforms you can use in Amazon Web Services; you can now use an OpenSolaris AMI (Amazon Machine Image), which means people can pay ten cents an hour for running OpenSolaris-based services. If you're building web services and you like Solaris but don't want to worry about scaling, this is a great option.
Anyway, the talk was very cool, and after today's sessions at JavaOne, it looks like I've got a few more cool things to check out. It's gonna be a fun week.

Thursday April 24, 2008
The Relay: 199 Miles, 12 Runners, One Weekend.
It's been a busy couple of weeks. This last weekend, I joined 11 other runners in an event called
The Relay, a a 199-mile race from
Calistoga (the northern part of California's Wine Country) to
Santa Cruz. I was looking forward to the event before we got into it, but it was more fun than I imagined it would be.
(if you're interested in seeing a few pictures,
here are some from my camera.)
Here's how it works: each team of 12 runners divides into two vans of 6 runners each. The 199-mile race is divided into 36 legs; each runner runs 3 legs of varying distances and difficulties. For example, I was runner #6 on our team; I ran
leg 6,
leg 18, and
leg 30, for a total of 13.4 miles over 30 hours. That's one of the unique challenges of The Relay versus other distance runs: in a half or full marathon, you do all of your running in one shot; when you're done, you're done. In The Relay, when you finish one leg you've got to recover within hours to get ready for your next leg, which can be at weird times (I ran at 1:30PM, then later that night at 1:00AM, then again the next morning at 11:00AM, and that was one of the more conventional groupings of running times...imagine you're the guy who's running at 4AM for his second leg; ugh).
The reason for 2 vans per team is to reduce traffic congestion: with 200+ teams participating, the route simply can't handle 400 vans all at once, so we're split up. Van 1 starts at the starting line in Calistoga, with Runners #1 - 6. Van 1 generally follows the same route as the runner, which gave us lots of opportunity to cheer on our teammates or throw food at them as we saw fit.
The exchange between Runner #6 and #7 is also a van exchange; Van 1 is done for a few hours, and Van 2's runners are on, running legs 7-12 while Van 1's runners can rest for a few hours before driving to the next van exchange point (where leg 12 ends and leg 13 begins, several hours's worth of running later).
I ran my first leg around 1PM, which brought us into downtown Napa. We said hi to our Van 2 friends, then took off toward Sonoma to grab a nice lunch and take a short nap in the Sonoma Square park, then we headed out toward Petaluma, near the next van exchange point.
My next run, leg 18, has to be the best run of the whole relay: I started in Sausalito around 12:45AM on Saturday night, ran along the waterfront, then uphill to the Golden Gate Bridge under perfectly clear skies and a radiant full moon. Normally the bridge is closed to pedestrian traffic at night, but bridge staff was there to open the pedestrian gates for each runner as we came by; I felt part of a privileged class as I entered the bridge, running as fast as my breathing would let me go. When I handed off the bracelet at the end of my run, I was too excited to sleep. Van 1 headed to the home of one of our teammates for showers and food, and about 2 hours of sleep; we had to get back in the van to catch our next exchange point at Canada College in Redwood City at 6:30AM.
My third and last run was 3.1 miles of unrelenting climbing along Highway 9 in Saratoga. It was not easy climbing, but the idea of being almost done kept me going, and seeing the exchange point, with scores of vans and tons of people cheering, gave me a last burst to finish the run. We in Van 1 then drove toward the finish line, at the Santa Cruz boardwalk, while Van 2 runners ran the last 6 legs. They arrived at 4:00PM Sunday afternoon, 30 hours and 36 minutes after we started.
All in all, we got about 3 hours of sleep over the weekend; we all paid for that the next day. But nobody got injured, and we finished faster than our estimated time. (when you enter the race, you are required to submit an estimated time of arrival) What I like better than finishing early was how close we were to our estimated time: we were about 15 minutes off of our estimate. 15 minutes out of almost 31 hours is pretty damned close (less than 1%). The course has a Sandbaggers award, given to the teams that run more than 2 hours faster than their estimated times. The award is not a compliment; those teams get disqualified.
I think the coolest thing about the race, other than running over the Golden Gate Bridge in a full moon all by myself, was the van exchange points, where all the Van 1's and Van 2's met. With 197 teams competing, The Relay had to restrict traffic, so Van 1 was only allowed to drive on the parts of the route where Van 1 runners were; we weren't allowed to support Van 2 runners on their routes. That meant most of the time, we didn't see our Van 2 teammates...except at the van exchange points, where Runner #6 (me) handed off to the 1st runner of Van 2 (Runner #7), and when Runner #12 handed back off to Runner #1. Those van exchange points were held in places with huge parking lots, like Canada College. The energy of all those people was amazing, and the vans had some hilarious signage and team names (our team's name: The Fat Bastards, which is soooo not true).
I think being around all these other crazy runners, almost 2400 of them, was what ultimately made this event as much fun as it was. Everybody was supportive of everybody else: I would be running alone on a difficult part of my run and somebody else's van would drive by and honk in support, the runners in the van clapping and shouting out support. We did the same with other runners.
I would do it again. And I would recommend it to other runners who are looking for something less solitary than a marathon.

Monday April 14, 2008
I got married on Facebook!
Here's an update to last week's entry about me getting engaged on Facebook. (I am so online, it's not even funny)
I got an email message from Facebook; here is the Subject line of that message:
<name withheld to protect the innocent> said that you two are married...
That's quite a claim. I just had to read on; here is what the message said:
<name-of-person-requesting-Facebook-marriage> said on Facebook that
you two are married. We need you to confirm that you are, in fact,
married to <name-of-future-wife>.
To confirm this relationship request, follow the link below:
When I logged into Facebook, it notified me that You have a relationship request from <said-person>. Clicking on the link brought me to The Big Facebook Moment:
You have a request from <still-not-going-to-tell-you-her-name> (Silicon Valley, CA) to add her as your spouse... Would you like to confirm your relationship with <she-is-mine-go-get-your-own-Facebook-wife>?
then the two buttons Confirm and Ignore.
Geez, this is more involved than our wedding vows.
What happens when I press Confirm? It's anticlimactic, really: Facebook congratulates us by saying You are now in a relationship with <name-of-Facebook-and-real-life-wife>.
What? That's it? I don't even get, like, a coupon good for 20% off dinner at a nice, quiet restaurant in (Silicon Valley, CA)? Apparently, being married is supposed to be satisfying enough.
Turns out it is.

Tuesday April 08, 2008
I got engaged on Facebook!
It's true, sort of:
Facebook is correctly
showing me as engaged, and I couldn't be happier. But Facebook put a scare into one of my friends; let me back up and tell you the story.
I got engaged about six months ago, and in December my fiancee decided to update our friendship status on Facebook. The way you do that is to update your profile to say "I am now engaged to so-and-so." Simple enough; my fiancee updated her relationship status, and I got a message from Facebook that said:
<your fiancee> said on Facebook that you two are engaged.
We need you to confirm that you are, in fact, engaged to <your fiancee>.
To confirm this relationship request, follow the link below.
I suppose that's a good thing: I don't want just anybody to declare they are engaged to me and have Facebook shout it to the world without my approval. It seemed a little silly to me, but I went into Facebook and confirmed that, yes I am engaged to the person who updated our relationship status.
But I don't check Facebook incessantly, and there was a lag between the time my fiancee made the relationship-status-upgrade request
* and when I confirmed. Meanwhile, one of my close friends logged into her Facebook account and saw a status update message about me: it told her that "<my fiancee> and George Drapeau ended their relationship." She immediately sent me a panicked message (through Facebook, of course) to ask if I neglected to tell her something. I assured her that we were okay, it was just a Facebook thing.
I guess Facebook is taking the safe route: until you confirm a relationship change, you are in relationship-limbo. Maybe this is the Facebook version of two-phase commit (no pun intended, for you computer geeks out there).
I tell ya though: I'm a little bit nervous about changing my status this Saturday to "Married". Does this mean we'll have to break up first? And what if somebody else swoops in with a relationship update request to one of us while we're waiting for the confirmation to go through? This is turning into the most stressful part of the wedding planning; sheesh!
* In the future, marriage proposals will sound like this: "If I updated my Facebook relationship status to say that I am engaged with you, would you click the 'Confirm' button?" "Yes, I would be happy to confirm lol"
Sigh.

Tuesday February 05, 2008
Why did MySQL change its mind and join Sun?
I just read this cool posting by Marten Mickos of MySQL about why they changed their mind and agreed to be acquired by Sun.
Here is his posting; this should be some interesting times coming up for Sun specifically, and the enterprise computing world in general, over the next couple of years.
Powered by ScribeFire.

Tuesday January 29, 2008
Intel and Sun Partnership 1 year old today (plus, Blackbox availability)
I just ran into an interesting
blog post by an Intel employee who discusses the engineering accomplishments made by the one-year-old-today Intel + Sun collaboration to improve Solaris on Intel.
The post is pretty cool! I was impressed with the variety of improvement projects going on, and what's really cool to me is that it's happening in
OpenSolaris. This looks like really nice work.
Oh, and on a separate note: Sun's "Project Blackbox" is now officially available as a product you can order. The product name is the
Sun Modular Datacenter, or
SunMD. That thing is damned cool! Apparently, the Stanford Linear Accelerator Center (SLAC) thought it was so cool that they ordered a second one.
Powered by ScribeFire.

Monday January 14, 2008
SPARC Servers Win 2007 Product of the Year
I just saw
this article that announced the Sun T5120 and T5220 servers winning Small Server of the Year for 2007.
Both server products use the UltraSPARC T2 8-core, 8-thread-per-core processor, run fast (benchmark-winning fast) at less than 100W of power dissipation, which is a big deal with high-end processors.
Nice going, Systems Group.
Powered by ScribeFire.

Friday January 11, 2008
Interesting articles about music publishing, decision markets
There's
an article in the
December 2007 issue of Wired Magazine, written by David Byrne (you may know him as the front man for the band "Talking Heads"). He wrote a great piece about the music business today and your best options for prosperity if you are an artist.
The article gives some numbers about how the pie is divided from a CD purchase, and lays out various options for artists who wish to publish their music. I think he does a great job of separating the business of distributing content from the goal of selling a transcendental experience to an audience (i.e., music takes us to a place that we want to return to again and again; music is the vehicle for taking us there, but the "there" is what we are trying to purchase).
The reason I mention it here is partly because it's interesting, but also because his line of reasoning reminds me a little of how Sun thinks about
software, specificallly, pricing. In the enterprise software market, you'll see software priced by the CPU, which is becoming a less relevant measure of how much value the software provides the customer. It's confusing. Sun has taken that argument off the table: we don't charge for you to use the software. You can simply download it and begin using it right away, so the product has the opportunity to show its value to you immediately, with no money out of a customer's pocket. You pay for service, which is directly related to customer value. To me, the relation to David Byrne's article is that Sun is stepping out of the industry's "conventional wisdom" (I'm not a big fan of this term, which is why it's in quotes) and trying to think clearly about how to connect our revenue to what a customer really wants. Software on its own doesn't solve a problem; it's a delivery system for solving somebody's problem, so we're lowering the cost of the delivery system.
And while I'm on the subject of interesting articles, here's a quick one: it's
a Google blog entry about decision markets and an internal study they did. What I found interesting was that the strongest predictor of success for in their study of information flows at Google was physical proximity. Bottom line: the best way to simplify communication was to put team members physically close to each other.
I'm a supporter of Sun's flexible work policy, Open Work, which allows thousands of Sun employees to forgo an office and work from home. But I also believe that there are negative effects of a distributed work group, and those effects are real but difficult to quantify.
If you have to pick one of these articles to read, read the David Byrne one. It's well-written: clear, and simple.
Powered by ScribeFire.

Thursday January 03, 2008
An inspirational leader
Hi folks, and welcome to 2008. I hope this year goes as well for you as it is already starting for me.
This blog entry isn't really about Sun stuff. It's sort of about USC stuff (my
alma mater), but really it's about a particular man at USC who just amazes more the more I learn about him. His name is Pete Carroll; he's the head football coach at USC and if you follow college football at all, you know what he's done for the USC football program over the past seven years he's been head coach there. USC is in its golden age of football, and that's saying something considering that 'SC is one of the top college football programs of all time.
I was watching the Rose Bowl a couple of days ago (we'll not dwell on how USC rolled over Illinois; it was a bad matchup but this is not about how badly the BCS picked its matchups), and during the telecast, Brent Musberger mentioned an article in
Los Angeles Magazine about Pete Carroll. I don't see a whole lot of material about him, so I decided to check out what the article was about. I was expecting a light, fairly content-free article about how great a coach he is. Nope, the article was much more interesting.
Want proof?
Read it for yourself. (or
here if the first link doesn't work for you)
The guy is amazing. He is relentlessly upbeat. He sets crazy, lofty goals for himself (although he doesn't like the word "goals"; he has tasks that he either completes, does not complete, or quits), and he spends a lot of time thinking and teaching about how to get into a space where you are solely motivated by joy and learning, where fear is off the table, and how to stay in that space. It's not just about football; it's about how he wants to live his life.
I'm doing a poor job of describing it; better you read the article (or skim, at least) yourself. The thing is, you can see it all come out when you watch TV broadcasts of his team's games. When the camera is following him on the sidelines, often times he's laughing; the guy is just having a blast out there on the field, and he's totally jacked up to be playing the game, living life.
He's one in a million, Pete Carroll, and just an inspirational human being.
Happy New Year, everybody!
Powered by ScribeFire.

Monday October 15, 2007
Get more pink dots!
It's been a while since I've posted here; I suppose it's because I've been spending more time working on internal stuff that I just can't talk about on this blog. Sorry about that; I've missed writing here.
Today, though, I'm at an internal gathering at Sun about open source. It just started, but already I've gotten something about of it. What did I get? The four words in the title of this posting: get more pink dots!
That was the direction given to us today by this guy, our CEO. He was talking about our business model, addressing questions Sun employees have about how we expect to make money from free software.
First, the pink dots: this site shows registrations of the Solaris operating system. Jonathan pointed out that many of the dots are in places where we don't have sales reps; these are customers that were found by our software. Let me say that a different way: the software is available for any use for free. Customers found and downloaded the software; they registered the software with Sun. Thus, the software caused Sun to learn about new customers. They may not be paying customers yet, but they're new leads to Sun.
Jonathan said the model is pretty simple: there are lots of people who will never pay for software, and there are some people who absolutely must pay for software. Our job is to get our software, for free, to as many of the former group as possible (software developers), because they eventually go to work for the latter group (enterprises that cannot afford the risk of unsupported software; they pay large amounts of money for support). If we want to grow, the job of our product groups is to get products to as many people as possible and not worry about the revenue. Just get more pink dots. Each of those dots is a sales lead; we have a sales force that can follow up on the leads, try to convert these nice customers into paying, nice customers.
If we spend our time asking the sales force to help us get more customers, that will prove inefficient. The sales force is motivated to focus on its existing, known customers since cold-calling is difficult and inefficient compared to talking to known customers. We can best help the sales force by giving them opportunities to sell support contracts. The more dots we have, the more leads the sales force has, generated in a more efficient way than spending money on cold-calling.
I thought that made things pretty clear. The software product groups don't have to worry about why we're giving away our products; the goal is to increase usage and stop worrying about money! That's somebody else's job, and for the kinds of customers we have, there are plenty of opportunities to sell support. It's what they want anyway; there's just too much at stake for them to avoid paying for support.
Works for me, and it was an interesting start to the day's open source gathering.

Wednesday September 12, 2007
Tons of news for Sun and its ISV partners
Wow, it's been a crazy week, and it's still only Wednesday. Here's what I'm talking about:
Monday:
IBM joins OpenOffice.org, saying it will start contributing code to the project. This move comes from the Lotus division of IBM, so you'd think this was a move with Sun to position against Microsoft (specifically, Office). And maybe it is, but then this morning, this happened:
Wednesday:
Microsoft and Sun announce an expansion of their strategic alliance. Turns out Sun is going to become an OEM of Windows Server, but what I thought was the cool part was our new efforts to make Solaris optimized as a guest OS in Microsoft Virtual Server, and Windows to be optimized as a guest in Sun's virtualization technologies.
All of this just goes to show how complex corporate partnerships can be: on the same week that Sun is involved with a major partnership that's competitive to Microsoft, Sun also announces a major new initiative with Microsoft where both companies benefit.
It also shows me how broad Sun's technology portfolio is; just look at the ISVs (Lotus, Windows products) we're announcing with. That's crazy. (the good kind of crazy, though).
Oh, and by the way, we also announced more ISV stuff this week:
Wednesday:
Sun acquires most of the assets of Cluster File Systems, including the
Lustre FS. I've been hearing about these guys for what's got to be over a year; they keep coming up in the high performance computing (HPC) area where Sun is gaining strength. Sounds like a good strategic move to me. Others might argue "What? Yet another file system? What about that ZFS stuff?" But my understanding of Clustre FS is that it is built specifically to be a parallel FS, meaning it's built to deliver data through more than one channel at a time, streaming massive bandwidth of data to HPC clusters that need it. ZFS isn't exactly about that; it solves a different (but compatible) set of problems.
So, cool week for Sun.
Powered by ScribeFire.

Monday July 30, 2007
Closure.
This blog entry has absolutely nothing to do with Sun Microsystems (although if you look at our full
Fiscal Year 2007 results which are being announced as I type this, you'll see that Sun has good news to report today, coming in way above our projections of 4% operating margin in our fourth quarter). But enough of Sun; I'm here to write about me.
One of the goals I set for my life was to complete a marathon before the age of 40. I did that, in March of 2002, at the Los Angeles Marathon (race results
here, search for bib number 10465, or my first and last name). I completed the race so strictly speaking, I accomplished my goal. But I had injured myself during my last long training runs, so I ran more slowly than I thought I was capable of (4:31:21). Worse, it was a full eighteen months before my knees felt well enough for me to run on a regular basis. That really sucked, and scared me a bit, since running is how I hoped to stay healthy for a long life ahead.
I exited that marathon in 2002 with a lack of closure. I had finished, but not as I thought I could, and physically damaged as well. I felt I could do better.
I changed my training regimen to include strength training (Sun's
Fitness Center has been a great employee benefit for me), being better about my post-run stretching routine, and importantly, adding significant hill and trail runs to my regular run schedule. I run four times a week most weeks, and at least two of those runs are in the hills. My long run alternates between a run on flat ground and a long run in the hills of the Bay Area Peninsula.
I entered this year's San Francisco Marathon with two goals:
- Finish, without injury
- Finish in under 4 hours (a 9:10-per-mile pace)
Weather conditions were good, and as you'd expect for San Francisco in the middle of summer: foggy and cool (it's like the city was built for runners; geez). The first nine miles were a bit annoying: it took that long for the crowd of runners to thin out, and for the first 9 miles I spent most of my attention trying to avoid stepping on other runners and dealing with runners who were bumping into me. After leaving the Golden Gate Bridge (yep, they closed half the car lanes so we could run over and back across the bridge; how cool is that?), there was more room for us to spread out, and the run became truly enjoyable. I was able to settle into a nice, steady pace, and felt no injuries nagging me. The hill training helped, too: every time we came into a climb, it was simply no problem.
I didn't hit the wall until maybe mile 24, but by the time I saw the 24-mile marker sign I realized that I was going to finish under the 4-hour mark. That gave me a huge psychological boost, so I decided to push myself to keep going, even though my legs were weakening and stiffening.
I finished in 3:52:35; I did it! I was sore and tired the rest of the day, but not injured. I feel a little sore today, but I can walk up and down stairs just fine, and I'll go for a short recovery run later today (well, "shuffle" is more like it; just something to get blood flowing into my muscles).
I don't have to run anymore marathons in my life if I don't want to; I am at peace with myself. More importantly, I'm confident again that I'll be able to run for the rest of my life, which I hope will last many, many years.
Here are my marathon-by-the-numbers results:
| Bib: |
5451 |
| Name: |
George Drapeau |
| Gender: |
M |
|
|
|
|
| Place Overall: |
915 |
out of 4250 |
| Men: |
766 |
out of 2805 |
| M 40-44: |
119 |
out of 417 |
| AgeGrade: |
57.10% |
Place: 1128 |
| FINISH: |
3:52:35 |
pace: 8:53 |
| 7.5 Miles: |
1:06:44 |
pace: 8:54 |
| Half: |
1:54:34 |
pace: 8:45 |
| 20.7 Miles: |
3:04:29 |
pace: 8:55 |
| Chip Time: |
3:52:35 |
|
| Gun Time: |
4:18:38 |
|
Powered by ScribeFire.