Tim Oren's Due Diligence

The Roving Eye: Power Sources, The Future of Tanning, Amazing Mazes

Aluminum smelters and data centers. Are alike in needed abundant and reliable electrical power. So the Columbia River valley is growing a crop of server farms. The Economist article notes that virtualization technology can be applied to migrate processing to where the juice is cheaper, as well as optimize the number of servers powered up to handle the given workload. Indeed, virtualization management startups are being reflagged as 'green' as fast as the PPT decks and web sites can be rewritten. Remapping the network connections, storage and other resources used by virtualized processes could sink any savings into a sea of management overhead if not optimized as well.

RIP, Robert Asprin. John Scalzi reports the passing of the well-known fantasy author. His Thieves' World was one of the more enjoyable multi-author sword and sorcery creations. Some few us also remember him as "Yang the Nauseating" in days gone by. I can think of far worse ways to go than on a couch, reading an SF novel.

All that time video-gaming was not wasted. At least if you want to join the Army or Marines and use the real thing. Designers of weapons and other interactive systems can now take facility with game controllers, computers, and networks for granted. We've come a long ways from the days when we tracked down (still abundant) naive users to try out our latest designs.

Hope for the pasty white? Perhaps for future generations: While I was paying attention to other things, the genetics behind human skin pigmentation were figured out. Seems that crosses between the melanin endowed and those less so 'average down', towards the paleface end. Not to fear, given the worldwide genome pool, a few more generations of 747s, migration and out-marriage, and everyone will come with a decent base tan pre-installed. It will be a more boring but perhaps more peaceful world, and fewer engineers will have to brave carcinoma to lose their hacker's pallor.

Speaking of microloans for mobiles. Here's a report by Kevin Kelly on a talk by Iqbal Quadir, founder of the original loans for village and family mobile phones program in Bangladesh.

Concrete Spaghetti. You think the MacArthur Maze is a mess? Check out these feats of civil engineering in Japan.

The Roving Eye: Someone Poke Fake Steve, Spybots Invade England!, Genetic Nannyism

Reality sets in at Facebook: Low entry barrier, low CPM. Even Fake Steve has noticed that most Facebook apps are fluff. More pointedly, developers are seeing very low effective CPM rates. CPM being a rather obsolete concept, I'd prefer to see data presented in revenue-per-user terms, as here. But it's clear that the open, easy, widely adopted Facebook is struggling to monetize, while the smaller, focused, and closed LinkedIn is claiming much higher effective rates (see interesting discussion here.) I'm still picking Facebook to be the Pointcast story of this decade.

Winning the War with Rhino Snot. Letting the troops (as versus the Pentagon) name your products can have interesting results. This company looks like it's self-supporting, but it would be fun to see its CEO try to get through a VC pitch without cracking a grin. It would definitely earn a place on the 'best name' honor roll.

Better see what's hiding in the closet. Following the precedent of DARPA's robotic Grand Challenge, Great Britain's Ministry of Defense (MoD) is mounting its own competition for automatons, with a twist. This time the robots are for surveillance purposes, meant to spy out snipers, IEDs and armed vehicles and soldiers in a village sized trial area. The task specific element of the challenge ups the ante from DARPA's trials, which were mostly about success in navigation. The MoD's competition also allows cooperating teams of bots, which could be of different types. One team anticipates coordinating a team of flying and earthborne bots. The MoD trial will take place in August.

The best things IBM ever made... were those clicky, battleship-weight PC keyboards. Dan's Data sings the praises of the 'buckling spring' design. (Via Derek Miller). Having spent many hours banging on an awesome converted Selectric used as the console on a IBM 1800 way back when, the PC keyboard was a welcome relief from the mushy action of the VT series and other 'glass TTYs' of the time. Almost worth having to put up with segment registers. The economics of the PC industry and weight requirements for portability put these things into the museum, but I still miss 'em.

What happened to 'Know Thyself'? Our self-appointed guardians in New York and California are on watch, making sure you can't get your genome analyzed at your own expense. Because it raises big concerns: 'What will patients do with this information? " Gee, would you suppose that might be their own business? if you wanted to freeze private investment into this area, you couldn't have picked a better way than this kind of statist fear-mongering. And maybe that's the point.

What Google Should Buy -plus!- Fun With R

The Experiment

A couple of weeks ago, I mentioned a survey experiment that I was attempting using a new Google Docs interface, and promised a report on the experience. Here it is, along with some thoughts on implications for Google.

The Docs spreadsheet itself is pretty easy to figure out. (I did not read the documentation - an experienced user shouldn't need it for an entry level program). It takes a bit to get used to the somewhat modal interface required when using a browser to substitute for 'normal' menus in Excel or the like. However, it's fairly easy to find the common operations by just knocking around a bit. It would be nice to have a quick 'what do you want to do' internal search options - without dancing paper clips. I give the initial experience an A-.

Go As Business Metaphor

Business writing and speech are loaded with metaphors drawn from sports and games - home runs, gambits, huddles, assists, game plans, and on and on. At the risk of being overly obscure, I'm going to drag in another source: The ancient Asian game of Go.

About twenty years ago I was working at Apple and had a cube in the De Anza 3 building. One of the campus cafés was downstairs, and I soon discovered there was regular lunch time Go playing group. I had learned the rudiments of the game in college, due to a friend working on a very early attempt at a computerized player. So I was drawn in as a kibitzer and soon active player. I never made it past somewhat decent amateur status before I was (inevitably) re'orged and moved to another office. But my period of play overlapped with a time when I was observing and learning(?) a lot about high tech business in general and software and platform competition in particular. Inevitably the two experiences grew together, and provided some metaphors that have proved interesting over the years, and seem appropriate to current issues.

Likely most people have at least seen a go board and stones (if not, try the Wikipedia article). Very generally speaking, the object of the game is to play groups of stones in order to surround as much space as possible. A group of stones without enough space within will die, and one with too much space inside may be vulnerable to the opponent trying to build his own viable group within that space.

Basic business metaphors can be something like this: The whole board is the available market. The stones played down are investment and expense to go after the market. A space surrounded is the net margin achieved by a product or company (the group). Too little space, and you're squeezed out of the market. Too much margin, and you are inviting competition to jump into your space.

So much for the basics, but further metaphors can be derived from tactical situations and styles of play. Of these, the one that keeps coming to mind lately is the issue of light vs. heavy plays (karui vs. omoi). These can refer to attempts to extend a group or jump into new territory. Light plays are probes. If the opponent attacks strongly, a stone or small group can be sacrificed to gain advantage elsewhere. Heavy is playing so large a group that it cannot be sacrificed, or fighting over an initial probe that would be better abandoned. This may let a more flexible opponent gain the net advantage.

Of course, such metaphors must be incomplete. The real board is multi-dimensional, and shifts in size and shape. There are usually multiple players, and people are not white and black stones. The board seldom starts empty. The customers get a vote on the moves. But the comparison can sometimes be instructive:

Given the description of light and heavy above (or found on the links), what style would you say Google, Microsoft and Yahoo are each playing? What type of play is best for a startup?

Is Vista Microsoft's Micro Channel?

OK, all of you fellow Old Farts who understand the title skip the next three paragraphs. For you newbs, here's some back story:

Once upon a time, there was this box called an IBM PC AT, which featured a kickin' 80286 processor, a substantial upgrade to the 8088 in the original PC. The two machines shared a common bus architecture, which allowed the use of plug-in cards to expand functionality. Back then you needed an expansion to do just about anything, including run the display. Trouble was, IBM had openly published the specs for that bus, as well as the rest of the original PC, and a lot of 3rd parties had started punching out expansion cards to compete with mother IBM. Even worse, there were 'cloners' who were copying the entire machine, and selling against Big Blue. This was not Good For Business and something needed to change.

When the 80286 generation rolled over to 80386, IBM tried to make that change. They ripped the standard PC bus out of the design, and substituted something called the Micro Channel Architecture (MCA), which was of course proprietary and patented by IBM. That would be the end of the license-free 3rd party add-ons, and of truly 'IBM compatible' PC clones.

In spite of some claimed architectural benefits of the MCA, the business intent was abundantly clear to both vendors and customers, and the market rejected it. The 'cloners' built 80386 boxes around the PC bus architecture that ran the PC-DOS of the day without modification. IBM's MCA based PCs entered the market late and ran a poor second to the cloners' machines. IBM was eventually forced to withdraw MCA from the market. Big Blue suffered an enormous erosion of PC market share and permanently lost strategic control of the PC platform to Microsoft, which happily supplied software to the cloners.

I'm not suggesting there is a complete parallel to Microsoft's situation with Vista, but there are a lot of similarities: A product introduced for strategic advantage at least as much as customer benefit. A workable alternative from the same vendor that remains in the market and still has wide support. Platform level competitors emerging as an alternative solution. Let me count the ways:

The Roving Eye: Trusting Wikipedia, Airportectomy, More Bad Debt

Wikipedia and Trust. When should you take Wikipedia's word on something? My own rule of thumb is that either the matter should be objectively checkable or the person or event in question should be long past and noncontroversial. Anything politically sensitive, either currently or as a plot point backstopping a current political narrative, should be regarded with deep suspicion. In spite of process changes at Wikipedia, you still get things like this, where a biased critic of an organization ends up editing its Wikipedia page. Community sourcing has its place, but this points out one of its hard limits: It cannot cope with politically charged matters. It either turns into yet another debating venue, or an echo chamber dominated by whichever side has the most time to kill. Update: another case in point.

LISP-head! So it turns out that Brad Feld is an old time parenthesis junky. I never got into the life-as-a-linked-list style. Back in the day a lot of my work was instrumentation, control and low level graphics, all necessarily real-time, so assembler and C were the weapons of choice. I still have this residual tic whenever I read about something that I know will be real-time implemented in an interpreted language: "What'll happen when the GC runs!!?" Oh yeah, incremental collection, twenty years of Moore's Law - guess it'll be OK. (Via Ole Eichhorn.)

A Successful Airportectomy. Or maybe I should say transplant. Over the weekend, I replaced my slowly dying Apple 'Snow' Airport with an Airport Extreme base station, and the wireless network situation has improved tremendously. Installation was mostly straightforward, but with a few security related glitches that invoked the cussing reflex. Just in case Google brings others performing the same operation here: As of 3/25/08, the interface and help documentation of the Airport Utility don't correspond in at least a couple of places. If you want to do MAC filtering, the doco say there's a separate menu entry. But it doesn't exist - instead use the Timed Access mode and set the times to Unlimited. When setting up a wireless distribution system (WDS), the doco says to use the 'add' button to authorize more wireless nodes, but that's actually on a separate tab in the interface. Also, if you reasonably decide to give the new network the same name as the old, and also decide to conscientiously change the network password at the same time, it will work great on your admin machine. But it will cause all the other client machines to be kicked off the network without any diagnostic, necessitating a bunch of spelunking in the obscure keychain interface. So don't do that. Once through these hazards, everything worked great, and I rounded off the install by hooking up my surround system and an old USB printer to the network via an Airport Express. Worked first time, and I've now got a stable 'four bars' throughout the house.

Go State! My wife and I are both alums of Michigan State University, and have been rooting for their men's basketball team in the NCAA tournament. They are through two rounds and into the Sweet Sixteen. If both State and Stanford win their next game, they will end up head-to-head, making for a real case of divided loyalty. Gotta go with the old school tie there.

More On Bad Debt. I'm not the only one taking notice of the unfunded pension liabilities at the state and municipal level. The Weekly Standard features California public employee pensions follies, starting with the recent near-bankruptcy of Bay Area suburb Vallejo. A lot of this material is from the California Foundation for Fiscal Responsibility, which is backing a public pensions reform ballot initiative for the state. The politician's handouts to the public employees' unions are starting to come due, just as the boomers head into retirement. It's not going to be pretty.

Rant: 'Social graph' considered harmful

(Minor backstory: Blog-buddy Marc 'Armed Liberal' Danziger posted at Winds of Change on the 'social graph' and Google kerfuffle and then asked for comment behind the scenes. Which I provided in a rushed e-mail, and then let the topic simmer through the weekend, distilling several fractions of bogosity from the raw material on offer. Which I now offer, from most trivial to consequential.)

'Social graph' is an unneeded and misleading coinage. Dave Winer lays this out well, in his inimitable style. A graph as referenced is an abstract math and computer science concept, a class of representations for - among other things - networks. Real people schmoozing at parties - real or virtual - are networking, not 'graphing' (that's something you try to beat out of Excel). For the sophisticates, there are well established fields like social network analysis and the virtual business category social network service which the incumbent term invokes. For either, 'social graph' is needless use of obscure terminology, perhaps in service of trumping up a category with no good reason to exist.

The Google story is highly suspect. So what actually happened here? A deliberate workaround on NDAs signed by attendees at a Google technology briefing, with vague mentions of various internal services and external partners that might or not be implicated in an attempt at more 'openness' in the SNS space, and some stringing together of names of folks who might or might not be involved in the project(s). If that's something you'll base a business judgement on, I suggest a job at CBS, not an SNS startup.

Google's immediate impact in SNS is limited. I recommend Danny Sullivan's well-stated comment to the above story. It's not like Orkut and iGoogle have been burning up the world of late. If Google gets too free and loose with their data, most folks won't even have to opt-out, because they've never opted-in. Google's incentive is to come up with something that will attract new community members, not keep them away.

...unless Google's making an evil-stupid move. The Googlers tend to be chaotic but clueful, so I doubt this, but it could happen. The evil-stupid thing would be to reverse engineer social networks out of people's GMail content and use them for marketing purposes. I'm talking address books and message headers, even content. This is known to be possible. It's also so far away from the originally intended use of that information that it's almost certain to create instant and explosive blow back.

Open social network representation is (partly) counterproductive. That includes both SixApart and associated efforts in that direction. If you're concerned about privacy - and SixApart is at least using that rhetoric - then why the heck do you want to make it easier for representations of social networks to be mobilized and shared? Wouldn't you like the shadow of the past to die along with a SNS that might be abusing your confidence? Oh, that's right, we're all (users and SNS providers) going to sign up for OpenID before that happens...

But it doesn't matter, because it's DOA anyway. For the non-technical: Horse, barn. Cat, bag. For the others: FOAF, client-side certs, digital wallets. What do these have in common, other than being dead meat? They all are/were bets that J. Random User wants and will use more secure and capable privacy and portability mechanisms if they are put into their hands. J. User continues to sign up for GMail, Facebook, etc. without any of these mechanisms and complexities. SixApart has one platform, LiveJournal, that is in the SNS space by a stretch and is a fading star. I love you guys (see the URL for this blog), but this is an overreach. Wake me up when someone with a user base and growth curve buys in. (Oh, yeah - SNS' are hit driven businesses, have you noticed?)

There is no singular social network (or graph). SixApart says the right words: "Everyone has many social graphs, and they shouldn't always be connected">. danah boyd nailed this one some time ago (PDF warning). Social networks aren't fragmented only because of technology issues, but deliberately so, as a means of controlling one's own social presentation, hiding and controlling information, even constructing identity. Before working on collapsing all those social affordances into one data structure, you might consider whether that's something the individual will appreciate. And even think about whether then attempting to market based on that unified view is going to be appreciated, or resented.

And you'll never get it out of people anyway.. But let's suppose that in this new transparent society we won't care if our significant others, our minister or rabbi, and some banker all see the same social network. Still won't work. Listen to the Old Fart for a moment here. Back in the day - before there was a real Web - academics sat around talking about the rhetoric of hypertext and how we'd all encode our links. Then we built the real thing - and it didn't happen. Take a look around you - how many typed links or even consistently tagged pages do you see? Real people don't do that.

Now you'll likely admit that marking up Web links is not an emotional topic or something discussed at parties for most of us (unless you're working at Google). Interpersonal relationships are both - see 'gossip'. What the heck makes anyone think we'll get all the nuance of social networks externalized - truthfully and completely and consistently, mind you - if people won't do it for something simple and unloaded like the Web? Doesn't even pass the grin test. So, we're right back into having to reverse-engineer semantics from minimal and partial representations (a la PageRank) but in an incredibly loaded setting. Plenty of step-on-your-crank potential for both incumbent and wannabe SNS'. With the properly perverse attitude, it should be entertaining to watch. Pass the popcorn?

So that's where 'cut-and-paste' came from!

A Newsweek piece on the Boomers and computers from Steven Levy - one of the few technology writers with a sense of the deep history of the field - provides a perfect example of what I've called reality creep:

"...I (born 1951) told these twentysomethings that there was a time when people wrote on machines called typewriters, beginning at the beginning and plowing through until the end, at which point they would mark up the manuscript with pen or pencil for the next run through the typewriter. If there was a need to recast a couple of sentences or even an entire paragraph, you would type on a new sheet of paper, cut the new text from the page with scissors and use Elmer's glue to paste it over the original not-so-hot lines. "Oh!" said one of the Googlers, of 1980s vintage. "So that's where 'cut-and-paste' came from!""

Those of us with a techie background has long borrowed computing concepts into real life: I'm multitasking, you're thrashing, my buffer is overflowing, I need to ping her... and so on. We're now well down the road to where using a concept from the virtual to describe the real is a commonplace in the general population. The Googlers of Levy's quotes are natives in that land.

Vacationing in Earthsea

Nearly 20 years ago, I was a 'guest artist' at a summer arts program at Humboldt State University on California's North Coast. The first and last time I've held the title, which amounted to coder-at-large for one of the first experiments in what was then called hypermedia fiction. One of the other artists, more in the traditional definition, was writer Vonda McIntyre. In an evening bull session she mentioned that fellow fantasy author Ursula LeGuin loved the North Coast area around Trinidad and had used it as the model for the seascape of her Earthsea series of books.

So, by that definition, we've spent the last week vacationing in a little bit of Earthsea. The wild blackberries are ripe, the salmon are running up on the Klamath, and a little front blew out the fog for the last two days. Life is good! Now on to Mendo for a few days before returning home.

Another 'Almost PC' Bites The Dust

...as Palm kills the Foleo three months after its introduction. Bringing out a product that is too similar in price, functionality and even form factor to the contemporary vision of the PC - without being one - has been the kiss of death for over two decades. It started with 'near clones' from the likes of DEC and AT&T, but it was IBM itself that provided the paradigm case with the PCjr.

From then to now, there's been a regular tide of similar errors by product managers wanting to attack an emergent market and trying to reach their design and BOM targets by cutting away functionality from the perceived 'standard' platform. A recurrent folly of wannabe multimedia PCs, TV PCs, network computers, e-mail terminals and what have you, and all dead. Palm is just the latest example of forgetting the tech past, and repeating it. In spite of the dwindling strategic importance of the so-called 'desktop', its image with customers is still so strong that it acts as a sort of tech marketing black hole, ripping apart competing visions that venture within its event horizon . (I expect the '$100 laptop' will ultimately end up in the black hole as well.)

Palm's cutting the Foleo is a sign the new investors and management team are showing some discipline. RHIT this isn't the only project to have been canned to free up resources and respond to the iPhone. Let's hope there's still some design magic left there. If Palm goes dark, and then produces nothing but an iPhone clone to restore the fading Treo brand, they're dead as well.