Sunday Feb 07, 2010

Student Intersection

Every year about this time, I host some undergraduate engineering students for a two-day whirlwind tour with customers, partners and employees in New York City. It gives the students a chance to see where an engineering degree can take them, and to discover what's different between short-term college projects and complex real-world product engineering. It gives me an early glimpse of the current engineering social context and lets me what technologies are mainstream enough to be used as teaching tools.

One of our running conversations was how to tackle a problem that hadn't been seen before. What choices do you make, or constraints do you put in place, if you have to think about scale, speed, or complexity that isn't in the literature? This came up in our meeting with Hot Potato, who worry about real time and real life events, and in talking with a major sports league that provides video on demand but isn't sure how to quantify the "excitement" quotient of that video from day to day.

Borrowing a page from college days, I wrapped up our internship session with a reading list:

George Polya's "How to Solve It". The classic, rooted in mathematics and algorithms, to build up an arsenal of hard problem cracking approaches.

Simon Singh's "Fermat's Enigma". Aside from a lot of the mental action taking place at Princeton, the background on how Wiles derived his proof of Fermat's Last Theorem is great. I use this to highlight how a seemingly minor topic covered in one area becomes a major factor in another -- I had finished a podcast on elliptic curve cryptography when I read the book, and the overlap in mathematical bases was eye-opening.

The July 1997 Wired issue on scenarios, describing how a pandemic might be solved by a graphics designer and gene hacker working together. Since that issue first showed up on newstands, we've faced SARS, avian flu and H1N1 flu outbreaks. Our response mechanisms haven't gotten much better. On the other hand, both of the students had been working in an "integrated science" curriculum, where mathematics was more directly incorporated into the appropriate scientific fields. We just need to add computer science in there as well. I made the remark that one of my good friends got into computer science because she was a psychology PhD student who needed to analyze data; today the data analysis experts at social networking sites are creating work for the psychologists.

Michael Lewis' "Liar's Poker" followed by Lawrence McDonald's "Colossal Failure of Common Sense." Two views of Wall Street, from the mid-80s birth of the fixed income derivatives business to the second, third and fourth order effects of its growth that led to the demise of Bear Stearns and Lehman Brothers. What happens when mathematics isn't integrated fully into the financial engineering sciences.

Bruce Tognazzini's "Tog on Design". I've always found Tog's approach to design and user interface refreshing, and I think ideas like selective disclosure would improve much of today's popular (but badly used) software.

Cory Doctorow's short stories "Anda's Game" and "When SysAdmins Ruled The Earth." Networked business models, networked organization and networked government. Both written before the Facebook boom, and therefore more important in light of it.

Finally, each of the people we met with had some advice or guidance on life in the real world: (1) Think big and unconstrained, beccause that's what's happening to compute and storage environemnts. (2) Cross-scientific disciplines matter. No single science is isolated. (3) Watch out for "end arounds" caused by cost or time disruption (4) Stuff happens. When it does, it affects brands, reliability, user experience and customer attraction. Be ready for it. (5) It's always harder than it looks to pull the pieces together: Moore's Law hasn't applied to integration costs.

Spending time with university students is always refreshing, both to find out what they think is interesting and to see what hasn't yet registered in their curriculum. And it shows them a literal world of eating options beyond the undergraduate cafeteria and campus pizza place.

Monday Jan 04, 2010

Flash Forward and Flash Back: New Innovating@Sun Podcast

Just before the holiday break, I sat down with Lisa Noordegraf, a performance engineer in Sun's Performance Applications Engineering team, to talk about flash memory and how it's moving from the consumer space into the enterprise. Lisa and I cover where you should - and shouldn't - think of using flash memory, how you drive the reliability of a consumer technology up to enterprise quality levels, and the gamut of performance benefits that she and her team have derived from putting flash memory at various points in the memory to disk storage hierarchy.

Wednesday Sep 23, 2009

Equity in Communities

Much is written about equity, capital and networking, particularly when prefaced with "social" to ascribe some value to sites such as Facebook, Yahoo or Twitter. Conflating these terms reduces their utility in describing the problem space.

Equity is a measure of value. Tells you how much something is worth, net net of whatever detracting, devaluing or impairing items surround it. Not just equity in the stockholder sense, but brand equity, personal equity, and

Capital is a working form of equity. Capital needs markets - networked communities - that agree on pricing, valuation, transfer, forward transactions, and membership. Capital markets may be the US Treasury auctions, or Kiva microfinance networks of affiliated microlenders.

Networking is what makes those values fungible. It's the basis for exchange of value and for disseminating values.

Assigning value to members of a community isn't a new idea: eBay has member feedback; Cory Doctorow introduced the concept of "whuffie" as a form of social contribution measurement in in Down and Out in the Magic Kingdom and that idea spawned Tara Hunt's Whuffie Factor. In a large technical community, what are the elements that create value, and what things detract from it? In particular, when you're dealing with technical documents and design patterns that share the half-life of the underlying technologies, last year's contributions have to be discounted versus updates and current interactions. On the other hand, this kind of value doesn't really fit into a market model - would you exchange two Hadoop experts for a graph theoretician? Measurement only makes sense if the semantic contexts are the same, and technologists are not necessarily players in an organizational fantasy league.

That's the set of problems that Peter Reiser and team set out to tackle when defining "community equity," a measurement of contribution and participation. Community equity rewards content publishers but equally recognizes those who interact with the content, as the commenters, redistributors and fine-tuners of content give the community heft. Can you imagine Twitter without retweeting or responses?

Peter and I talked about the state of equity measurement, code availability and what some more down to earth views of "semantic web technology" might be in the latest Innovating@Sun podcast; check out the audio stream and related links.

Friday Aug 28, 2009

Trust But Verify

Glenn Brunette is a Distinguished Engineer and leading security practitioner at Sun. He and I sat down (virtually) to talk about the Immutable Service Container project, a set of tools designed to bridge OS minimization, virtualization, and security monitoring mechanisms. An increasing number of customers are thinking about deploying applications to clouds or other virtualized environments in which they can't attest to the provenance of the underlying hardware and host operating system - ISC provides a thin layer between hypervisor and guest image that re-defines the "trust, but verify" maxim for new current generation of deployers.

Sunday Aug 09, 2009

NetBeans in a Multi-Cultural Developer World

Latest Innovating@Sun podcast is up, featuring a discussion with John Jullion-Ceccarelli (Senior Engineering Manager of the NetBeans team) centered around NetBeans in a multi-cultural world.

Years ago, it felt like there were strong forces for convergence around languages, tools, and repositories. Sun was a vibrant participant in those discussions, with "Java" often the answer preceding the question. Sometime around 2001, that attempt at compression reached the Chandrasekhar limit, at least in terms of things like WS-\* (thanks, Tim). The only things that result from such strong nuclear forces are supernovae and black holes.

Computer scientists have always invented new languages, grammars, and tools as the state of the art has evolved; new languages that fit the way data and applications will be deployed to clouds, or deployed with greater agility, are just natural evolution. NetBeans has evolved as well, no longer a "Java tool" and much more reflective of the richer state of the state, whether it's clouds, collaborative development sites, languages, or build tools.

Thursday Jul 23, 2009

Accidental Geography on Facebook

I'll admit to a certain vanity with Facebook: I've been trying to build an audience for my personal blog, using a Facebook page to import blog entries and inviting just about everyone who's a friend to follow the page. Facebook very nicely provides "insights" (analytics) on interactions with the page - number of comments, ratings, and other feedback.

Today I noticed that the "Cayman Islands" were the top country for interaction with my page. I have no friends (that I know of) who call the Cayman Islands home, so I poked around and found the page comment that generated the trend. Sure enough, it's a US-based friend vacationing in the Caymans (his public content conveys the same information). So commenting on a Facebook page creates an indirect trail to whatever IP address is reported at the time.

IP addresses are a terrible mechanism for assuring location (due to proxies, carriers, firewalls and other aggregation/translation points) but in this case, they are a fair proxy for "not at home." You could argue that if I'm using Facebook on vacation (or while traveling) I'm disclosing a signficant amount of personal information anyway, but there are many Facebook users who hide their home geographic information and by extension, might want to hide their mobile geographic information as well. The fact that source IP address trumps "home" for determining interaction sources means that Facebook is at least ignoring the intent of, if not the exact letter of, these user preferences when it comes to clouding geography.

Friday Apr 17, 2009

Cloud APIs: Call'em As You See (and Extend) Them

Tim Bray and I sat down (albeit 2,600 miles apart) to talk about the Sun Cloud APIs in all of their RESTful grace. We got into why a Creative Commons license makes sense for an API, why the top-level API set is so small, how and why a cloud deployer might want to expand the APIs, and what lessons Tim learned from slogging through more developer documentation than is considered healthy, even by Canadian standards.

Hockey was not discussed.

Transcripts and pointer to the audio are on the Innovating@Sun blog or click to play away below:

Thursday Feb 26, 2009

Reshaping Services Industries

I'm participating in a technology roundtable for one of the services industries this week; it's a closed-door session with some pretty heated discussion. The economy is definitely hurting this segment and one of the recurrent themes is that personal (low volume) customers are going to be the growth engine, not business (high volume) users as companies cut back on their spending. Without disclosing industries or players, here are the themes I presented as ways to reshape the services experience for a broader range of current and potential customers:

1. Open up interfaces to what you do. It's a time-honored tradition to think of a business' data as its very own. But what if that data can create new ways to think about the service, new market segments, or new demand for the service? Web mashups are interesting if they have useful data sources to draw from; this doesn't mean that services companies should expose personal data but they can provide interfaces to widely useful, logistics, location, or inventory information. The New York Times has done exactly this with the Times APIs, a set of services that let you search Times content by keyword, context, or specific parameters like dates and by-lines. It doesn't create new newspapers; it creates new ways of using the news that hopefully driver readers back to the Times site for additional information or context. It's not sufficient just to think about consumers as the endpoints of a transaction. Which brings me to...

2. Determine the value of social networks. In the words of Tim O'Reilly, it's not the network radix, it's what gets shared. What experiences are valued, or not valued? What are people saying about you? What does a Twitter search on a hash tag or keyword associated with your service turn up? Simple example: the 3rd generation of my family's cousins have a group on Facebook; every time we think about having a reunion somewhere we end up discussing air fare, hotels, meeting rooms, babysitting, photo sharing and mass quantities of food (it is my family, after all). All of those services businesses could find a pre-aggregated demand pool if they'd build a "Book Your Family Gig" application in Facebook.

3. Create an inbound channel. More informally, this is "listen" but it's the corollary to using the social network to get a pulse on what people think of your service. What are the limits on elasticity, choice, price, and user generated content that demonstrates new uses or specific value add to your service? Far too much services marketing is outbound only; I'm awash in email promotions, coupons, special codes, and one-time offers. But very rarely am I asked what I'd pay for something, or what my trade-off points are. Even more direct: just watch Twitter for "brand name #fail" and see where the exceptions are happening. Reach out and address, listen indirectly, because that creates....

4. Pleasant inconsistency drives loyalty. Give -- don't offer -- me a better seat, better room, nicer car, or 9 holes on the course that I'm going to bludgeon with a golf club that I can't control (this is in fact how my golf badness started). The first time I experience something out of the ordinary flow, I'm likely to decide whether or not I'm willing to use it again, pay a premium for it, or arrange my brand loyalties so that I'll get the benefits of being a repeat customer. Obviously, bad inconsistency, or consistently bad experiences, drive loyalty to other players.

All of these conversations have to happen with input from multiple demographics: Millennials, Gen Y, Gen X, Boomers and now Gloomers, crossed with various degrees of online experience, social networking utility and trust. If you don't figure out how to meet the demands of the Millennials, you won't be in a position to sell, serve or employ them in 5-10 years when they are riding the economic recovery as the core of the spending and working public.

Wednesday Jan 28, 2009

WordCamp NYC 2009

We are going to host and sponsor another WordCamp, once again at Sun's NYC office. The date is October 24, 2009, the place is 101 Park Avenue, and the details are being worked out. Last year we had Matt Mullenweg as a keynote, and some spectacular speakers covering everything from building a personal brand to writing code. Where else do PHP, MySQL and HTML hackers congregate with budding journalists, photographers and other micro voices? Video from previous WordCamps is now part of, a great way to fill the 9 months it's going to take to gestate our next show.

Early shout out to the GSE Divas for their continued support and efforts in making these happen.

Tuesday Nov 25, 2008

Facebook and the Wrong Definition of Productivity

I'm convinced that "productivity" is a dumb word. It presumes some magic metric for how people create value in the workplace, and that metric is usually, inexorably tied to a clocking problem. Work faster, work harder, work more hours - and my favorite - waste fewer hours! I hear Tock's admonishments ringing in the back of my head every time I see the red flag of Facebook notifications. The open question: is Facebook the new Solitaire?

In short, Facebook is a valuable business tool provided you treat it as a context creation vehicle and not actual work product (for most people; Sun has people whose primary work product is created via Facebook and that's because they're primarily recharging employee workplaces). If you spend hours a day creating goofy groups and inviting random friends, or searching for the transitive closure of your friend(friend(cousin(high school buddy))) relationships, then you probably do have a time management problem. But the problem with casting any activity as a "bad use of hours" infers that there's some sorting and prioritization of hours that belong to your employer versus your friends, family or co-workers. Whose hours are they, anyway? The subject is at the heart of what is typically called "work/life balance", but I've more recently heard simplified to "life balance". That's the right emphasis -- when you're always connected, always thinking, frequently Tweeting to inform your e-crew (and self-selected marketing bots), there's no thick-drawn line separating the two. While Facebook, LinkedIn, Twitter, and their social relatives are heavily colored by non-work relationships and content, they can improve rather than impair productivity.


Productivity is about reducing problems of time, geography, and knowledge. Who can help me with a problem, where is the information (or the expert) that I need, how quickly can I find the best answer? There's a negative side the scale as well - what's impairing progress, what if I take the first answer that later creates a myriad of issues, what if I follow a bad link (people, website, or driving directions)? Building -- and maintaining -- relationships drives the positive side of productivity, because they help you navigate to a suitable win more quickly. Actually paying attention to the content traversing those social graphs sometimes addresses the productivity impairments.

This was brought home, literally, last week when I twittered that I was "pissed off." Immediate Facebook status comments echoed Journalism 101: What upset me? Why? When? My status wasn't work-derived, only the after effects of a bad conversation with someone who sent me a very incorrect bill, but without the context it was an attention-grabber. I even got a ping from my boss, who readily admits that he follows me on Facebook as a way of managing me. There's probably some obtuse managerial treatise in that statement, but his outreach kind of snapped me to attention and quite honestly -- got me back to work. Well within the usual hour long damping factor needed to get productive again after such aggravation. Whose hour was it? No need to debate ownership: it was a useful chunk of time.

Friday Nov 21, 2008

Communications Developers

I was talking this week to a company that builds communications technology. More than one, actually, and their definition of "communications" is as broad as it should be in a world of wireless, digital, social, and voice flavored bit streams. We had a great sidebar on who they believed their developers are: their in-house team? An outsourced team hired to complete a project? Customers? Integration partners? And I couldn't resist bringing up: What about open source communities?

Somewhere in this story is a moral about balancing protein and caffeine before afternoon meetings, but I missed it. I suggested that perhaps the Asterix community was a good leading indicator of how complex communications systems would be built in the future. For the reaction I got, I might have suggested the other Asterix as a source of modern technology insight.

But let's face facts: the hacker culture has thrived on being able to play with phones since day one. Taking a platform that was in-band and invisible for so long and making it a developer play is just a natural evolution -- one driven by the growth in power, performance, and real-time capabilities of general purpose operating systems, availability of general purpose hardware on which to run them, and application level communications software. Put another way: Any comment about software not being "ready" or "capable" has been proven untrue over time.

We saw it happen with Linux, MySQL, and Drupal, and now we'll see it happen in the classic embedded spaces as well. All you need to do is follow the hobbyist space to sense the edges of the market: Linux on Linksys in the hackable router market is a perfect example, and it leads to creative applications of open platforms, like turning the Internet upside down. Go ahead and laugh, but there's an entire suite of access control, identity management, and auditing applications waiting to be built in slightly less user-inhibiting ways.

For true star power not related to the star (asterisk) key on your phone (calling by name, not by value), check out Sun's own Brian Aker discussing how he builds (current tense, as in work in progress) his own PBX - his musing on Asterix are even funnier than the Francophone Asterix. [note: Four letter words, Pecha Kucha style, mash-ups, and Brian's love life. Simultaneously.]

If you aren't thinking broadly about where your developers will come from, they'll surprise you and possibly your product plans.

Sunday Oct 19, 2008

Newt Gingrich, Light Bulbs and Market Disruption

I was digging through old paper files this weekend and found a note I had scribbled on a hotel phone message notepad that read simply "Newt Gingrich and candles." Truth be told, I was doing a periodic office purge and happened to find a manilla folder of clippings, notes and sketches I had for a random book idea back in 1998. The Newt-onian mechanics actually sent me on a Talmudic scholar-like Google quest for the original quote (also verified here) which was along the lines of "If Thomas Edison invented the light bulb today, [the news] would report it as 'Candle making industry threatened'". Put the quote in historic context: Gingrich was railing against CBS News at the time, he was reacting to stories about how the budding Internet phenomenon was going to upset retail, and he was looking for a sound byte.

Political contexts often remain invariant over time.

I used that quote frequently when talking about Java, networking, and the .com boom of the late 90s and how the ability to directly interact with customers, consumers and potential users of your product completely reshaped distribution, and that any company that chose to ignore those factors would become less interesting. It is, and was, a matter of perspective on massive market disruption: opportunity or threat.

Market contexts often remain invariant over time.

I won't whitewash the current financial crisis: the markets are a mess, and we probably won't know the true extent of the mess until the cyclic graph problem known as counter-party risk is sorted out, assets priced or auctioned, and some senses of liquidity restored to the credit markets. But actually traversing these graphs and identifying the risks highlights the opportunity, rather than the threat - we have a chance, as technologists, to create the infrastructure, tools, and data management mechanisms underlying any improved transparency in the financial markets. In the words of my ever-cranky buddy Geoff, if you want to understand the risk of a zloty-denominated futures contract on the price of oil in emerging markets to be delivered to Eastern Europe, you better have transparent market information for all of the intermediaries and end points.

The information exists, but it's not evenly distributed or acknowledged. That is, by the way, what William Gibson says about the future. With the .com boom, we had the good part first, then the correction; I'd prefer to think about the current credit market crisis as the correction preceding the next build-out of financial market infrastructure. Compute grids used for Monte Carlo simulation of interest rate and credit quality future-looking scenarios, often used to price mortgage-backed securities, will not be the sole banking infrastructure that rebuilds confidence and velocity in the credit markets. Those scientific HPC systems are going to be supplemented by data-rich, "commercial HPC" installations to examine systemic, structural and market risk based on transaction data, market data, and multiple sources of public data: less prediction and more counting.

Thinking anything less lets financial candle-makers extinguish the possibilities enabled by the electric light bulb.

Sunday Oct 05, 2008

WordCamp NY

What would make 130 bloggers, mySQL hackers, PHP coders, writers, journalists, photographers, and other assorted geeks come into a midtown New York office building on a rainy Sunday morning when the Big Apple is more traffic impaired than usual?

WordCamp NY.

And it was worth every bit of logistical challenge -- even for those of us who just got to hear WordPress creator Matt Mullenweg go apolitical about the "State of the Word," including an update on WordPress 2.7, how plugins become part of the core, how the WordPress community is thriving, and why he does what he does (and he's becoming more of a softspoken Jimmy Wales and less of a PHP hacker elite, and I think that's A Good Thing).

First principles: Sun was a sponsor of this event, which was held in our midtown NY office, and the obvious question (asked by a few people) was "Why Sun?" In my opening, caffeine-addled comments, I tried to outline three reasons:

(1) mySQL, the database on which WordPress is built, is part of Sun, and therefore WordPress users are indirectly Sun software users, whether by intention or default

(2) WordCamp epitomizes the sharing of good ideas. Sun has always believed in building on the ideas of others and sharing our ideas with others, whether through standards work (think NFS), open source projects like GlassFish and OpenSolaris or direct employee participation in a variety of smaller open source projects like noodling around with WordPress plug-ins.

(3) The digital divide is an information one. Bloggers help bridge it daily. Every great consumer technology - cell phones, printers, GPS systems, digital cameras - had its start as a very high end, professional (read: sometimes military) version which came down in price and up in accessibility, empowering an entire host of users to apply the technology in new ways. WordPress gives the power of a rich publishing system (not just word processing, but producing something with a vibe) to individuals.

But in the words of Marty Debergi, 3 minutes was enough of my yacking - Matt Mullenweg provided data points that had me scribbling as fast as I was nodding and remembering to keep my mouth closed.

- WordPress has 3 core developers and about 90 contributors. It defines "good community" and "global" - the core team lives in England. And Florida. And Vancouver - one each.

- WordPress has hit the big time - it powers the NFL blogs and CNN's blogs, the latter of which provided the best up to date Hurricane Ike information available. With over 230M daily page hits, aggregate content is a "significant percent" of Internet viewership.

- There have been over 11M downloads of WordPress, with 2.38M new blogs this year, 35.8M new posts and a run rate of about 4M posts a month. In larger units of measure, that's the equivalent of two full English Wikipedias a month.

- There have been 18 WordCamps this year, with 9 more planned. Matt showed pictures covering the spectrum from a hotel venue that surrounded an outdoor pool to a restuarant to a formal lecture hall. My note to other-selfs: local organization is key; without Jonathan Dingman driving this in NY, there would not have been a WordCamp, and his able team of volunteers made this fly. Thanks, guys.

- WordPress, like OpenOffice and Firefox (nice company!) is a widespread consumer open source phenomenon. I had to think about this - but it's a very strong statement about the permeation of "that open source stuff" into not only the broad market, but the next broad market that matters: the people who have grown up empowered by consumer technology.

- Great distribution enables great customization. With an average of 4.96 plugins per blog, but a very long tail of plugins available, nearly every WordPress instance is unique. If I was looking for a powerful statement about mySQL, that was one that fulfilled the wish in an unexpected way. The underlying database schema, database tables, and interfaces are the same through (nearly all) of that dynamic range of plugins, but the combinatorial possibilities are measured in scientific notation. This isn't a case of powerful being good -- as Matt later commented, simple is good. Simple but solid enough to do the job worked -- which is why they chose mySQL and PHP to power WordPress.

- Akismet, the anti-comment spam tool, remains the most popular plugin, attesting to the widespread nature and depth of the issue. Matt's take: Spammers are clever, and if AI is going to be invented, spammers will do it first.

Nothing like hard data to make my hastily scribbled notes look good =- and get us thinking about WordCamp NJ this summer.

Friday Sep 12, 2008

AHRQ interview posted

The phone interview that sparked my long rant on innovation frameworks has been published online by the Agency for Healthcare Research and Quality. There is, as always, a backstory to the front story: Lise Rybowski, a principal at AHRQ, is (was) a classmate of mine who found me on Facebook. Intersecting circles of trust sometimes produce more than cute Venn diagrams.

Wednesday Sep 03, 2008

Word Camp NY

WordCamp NY is now in full-bore {planning, scheduling, registration} mode. I'm providing local support in terms of Sun office space and some logistics, and I'm looking forward to hearing about everything from CMS to spam reduction. Thanks to the GSE Communications Divas for working the details (and using "hygroscopic" in a sentence!)

Top question I'm asked when I mention this: What does WordPress have to do with Sun? Primary: WordPress is written in PHP and uses mySQL as a content and metadata storage engine. Every WordPress author is indirectly a Sun product consumer. Secondary: Wordcamp is about blogging style and structure, and that's directly aligned with the culture of transparency and public discourse that we've been living for several years.


Hal Stern's thoughts on software, services, cloud computing, security, privacy, and data management


« April 2014