The Next America

Peerflix vs. Netflix

I have become increasingly intrigued with PeerFlix lately. They're like Netflix but with no central inventory. You send your movies directly to someone else on request in exchange for points. Then you use the points to get more movies. Net cost is $1.37/trade. They sucked me in last month on a weblogger promo -- they gave me two free movies with no strings attached. This gave me a starter inventory to trade from. I then posted a bunch of old movies I had. Some were pretty good, some weren't (many were freebies I got via rebate when I bought a DVD player in 1999). But I could live without them. Surprisingly, people wanted them.

The best part was trading the pan/scan version of a movie I love (Chicago) for the widescreen version. Best $1.37 I've spent.

The worst part is the emails. Since its a distributed system, if you request a specific movie they email multiple people asking them to send the movie. The first person that commits to sending it within 24 hours closes the deal. However, if you're not sitting directly on top of your email this means you end up getting lots of requests that someone else has already filled. Including requests for some atrocious movies (like Rambo) that I am eagerly looking forward to saying goodbye to.

I'm cancelling Netflix at the end of the month (5 year subscriber). The selection at peerflix isn't as good. Its about the same speed since you get mailed the next movie as soon as you commit to send a movie (like Netflix did before 2001).  But you can make up for selection by having a bigger "inventory". If you want 5 movies instead of 4, just pay $10 at Amazon for a used DVD and your inventory increases. After $20 on Amazon, the two free movies, and some trades, I have a 6 movie inventory. And my new cost drops to $5/month from $20.

Peerflix parallels Ebay. In the early days, discounters were king, with online auction sites run by discounters selling from their own inventory. Then Ebay cut out the middleman, settling for a few bucks on each transaction. Netflix does a great job, but the $1.37 may crush them in the long term. Peerflix can clearly make real profits with this model.

Trojan.Vundo & VirtuMonde

This is the nastiest worm/spyware I've come across. And McAfee and Norton are helpless.

I found it in the wild this weekend (my sister's computer). This is a worm since it spreads through an IE browser vulnerability. I suspect it also spreads through "older" worm infestations. XPSP2 had been installed months before on the machine which closed the door the worm supposedly travels through.

The observable symptom is that internet browsing is very slow (though functional). There are also popups, including repeatable ones to WinFirewall and WinAntiVirus whenever Mcafee or Symantec are accessed. Symantec has a scrubber -- while it seemed to eliminate several files, the end result was totally ineffective, probably because it was seeking on previous behavior. AdAware detected it (and gave me its name), but was unable to fight it.

Hijaak This is a very useful tool for manually fighting it. The primary indicator is a Browser Helper Object or BHO (which is pushing the ads) that shouldn't be there and is immediately restored whenever the registry entry is deleted. Also, you will see processes that do not correspond to an actual file on the hard drive, which the worm uses to masquerade itself. The nastiest part of the worm, which Symantec documents, is that it randomly generates the name of the worm (even worse, it uses typical Microsoft filename terminology). This makes it very hard to troubleshoot via google, since the suspicious process can't be directly traced to Virtumonde. One key indicator, however, in Hijaak This is that the processes are marked with stars (thus loading them in safe mode). Almost no legitimate process should be so marked. The worm process uses about 12MB of memory, which is also a stable indicator. One of the files is commonly named BKINST.exe or BKINIT.exe.

One of the trickier parts is deleting the main body of the worm. Its continually busy (tied to an active process), and the process is restarted immediately on kill (two processes "protect" each other). A way I discovered is to attempt to delete the file, then immediately kill the process. The delete will execute before the process can restart.

When the son of VirtuMonde uses the name of legitimate "helper" applications and drivers, it will be impossible to find and kill.  VirtuMonde demonstrated to me that we have nearly lost the war on worms and spyware. I am becoming more and more intrigued with the concept of read-only operating systems.

Brooks Brothers is a Spam Source

I use a custom email domain for "one-off" email addresses, normally for ecommerce sites. So I've been getting a lot of SPAM lately on the custom address for best buy. I get all mail to all addresses in the domain so it isn't a dictionary attack. Why did Brooks Brothers sell my email address to spammers?

[Edit: Apologies to Best Buy, whom I originally identified as the source.]

Whizspark

Marc Canter pointed out Whizspark which has revealed their (relatively novel) method of monetizing social networks -- use them as promotional vehicles for events. I'm impressed. There are certain people I know who are event "hubs" -- I'd be more than happy to get email from them about good events they're aware of. They get incentivized by free stuff to build and maintain a network of contacts. Even better, this service could potentially be used to see who else is going to be there -- a definite value add, especially in meeting friends of a friend. There are a lot of events that are only fun if you know a certain number of people there. An ingenious idea would be tracking the invitations as they spread so overlapping social networks would only receive one email (from the first attempt to send).

The problem: the site is useless unless you register. The functionality I presume is based solely on Marc Canter's description. You can't even see how many people in the system are near your zipcode without a username and password. And forget examining the model of exactly how listed events are distributing prizes based on effective promotion using the system. I no longer register for sites based on the premise of value -- they require actual value to be shown in advance. Show me that there are 100 people within 50 miles of me (and there is a reasonable default such as "allow anonymous contact within 100 miles" -- I for one would be glad to get location-based anonymous email as long as there are restrictions such as one email per event) and I'm sold.

Especially with social networking, assume the customer base has been burned. We will not spam our friends. We do not trust you in the beginning. Make our lives easier and we will begin to build a relationship. "Small" events are a great market -- most are desperately in need of tools for "professionalization". But I won't give my info until I see value (and I refuse to waste my time putting in fake info. The folk music concert I'm throwing next weekend (Liz Reed, Bristol, RI, $10, money to the Northfield Conference) has made me painfully aware of the difficulties in small event promotion.

These guys may have the solution, but they're doing a weak job of showing it to potential customers.

Shopping Engine Scams

I've been upgrading my stereo. Looking for two very specific pieces of gear from mainstream manufacturers. $800 from a dealer (b&m or online). $450 from the very shady grey market. Various intermediate points.

A perfect time to use shopping engines. Very helpful. But spending several hours with them, red flags keep going off. The commentary on the cheapest stores (and the shadiest) is schizophrenic. Horror stories, each compensated by four or so very positive experiences. Each positive experience mentions specific equipment (good) but very little other detail. Except it seems perfectly tailored to defend against the negative reviews by highlighting qualities in dispute (responsiveness, timeliness of shipment).

I've heard some rumors that there are some shops in India which provide this type of service. Even if the shopping engine were to limit feedback to "real" customers, there is no way of preventing a vendor from qualifying the false buyers by completely ignoring their actual orders. This is where FOAF, weblogs, and anything else that gives us context is so important.

I didn't buy from the cheapest source. And I was quite careful to use American Express.

CIOs and Technical Background

Jerry Gregoire has a great article in CIO about the vanishing IT department. His first warning is well-taken about outsourcing -- an organization that chooses to outsource will only receive mediocrity. This may be completely acceptable but it is a long-term decision with enormous consequences. If you need lower than average cost, above-average reliability, true flexibility, or cutting edge technology, outsourcing is unlikely to give you what you are looking for. Which isn't to say that a high-performance IT shop shouldn't use contractors. But there is a huge difference between tactical outsourcing (specific "hard" or "easy" jobs with well-defined SLAs) and strategic outsourcing.

But his second warning was unusual. CIOs should have a technical background in IT. I think that this is not so much due to the contribution they will make, but due to the ripple effects across the organization. Without it, hiring and mentoring the best technical people will be difficult. 75% of an IT leadership team should have a strong technical background. Without it, they will become increasingly captive to large system integration vendors or the status quo. It is very appealing to be told to sign a check and the latest, most modern technology will be deployed to your users. At least 25% of the presentation is normally a lie (usuals: user impact, cost, schedule, technical risk). Without a gut feel for technical risk, a CIO must be extraordinarily savvy to compensate. That's not to say that they aren't some extremely good nontechnical CIOs. But they must specifically focus on building technical leaders and listening to them on key decisions. When push comes to shove, that is a difficult focus to maintain. A nontechnical CIO can work well when an overly technical organization needs shaking up or when better business focus is needed (or as a safe pair of hands as technical leaders are developed for the next level). But if technology is a key differentiator of a company, a CIO with a technical background is essential. Much like CFOs should have a CPA/MBA.

The problem is how to measure a technical background. A PhD/Masters doesn't necessarily indicate the right skillset. Programming talent isn't necessary. MCSEs and the like are useless as an identifier. The good certs (for someone ten years from now): CCDP (understands network fundamentals and design), PMP, MBA, CISSP, ??. Someday soon, a major business school will create a year-long course for CIOs with these as prerequisites (I'd suggest M.S. Information Engineering). And the graduates will be stunningly successful.
The courses would cover contract law, negotiation, future technologies, IT execution, business process modelling, and quality measurement systems. Harvard's five year PhD program in IT Management is a possibility -- but it focuses on late 20s people with 5 years and a hankering to teach in business schools rather than early 40s ready to leap into the executive ranks.

Tivo as Remote Control Solution

One surprise from my new SD-H400 Toshiba TIVO/DVD hybrid was that it solved the remote control issue. Two years ago I invested in a Sony touchscreen remote (AV2100). After programming, it was great. A little sluggish when controlling the VCR. But all in all, very wife and visitor friendly.

Then I bought the Tivo. I got the integrated one since I needed a progressive scan DVD player anyway. Unexpectedly it is the holy grail of convergence. It handles DVD and VCR functions. It uses an IR blaster to handle Cable TV. AND it controls the key functions of the receiver (power and volume). AND I no longer need to switch receiver inputs since it is handling DVD, VCR, and CATV. The Tivo remote control is light, easy to program, and wife-friendly.

So the RM-AV2100 has been put aside as unnecessary. The only commonly used key the Tivo remote does not support is TV aspect ratio. A few programmable keys and it would be the killer remote for many users.

Smart remotes are key to widespread acceptance of advanced AV technology. But they are fundamentally limited since 2-way communication can't be assumed. Better is the serial port control now becoming routine on high-end gear. But even better would be 802.11b for devices where latency is acceptable (amplifiers, TVs). Receivers should integrate total system control (IR blasting, remote control programming, IR to serial conversion). For a couple years they had lost video, but component upconversion is rewinning the battle (I'm looking forward to my new Denon AVR-2506 to actually watch progressive DVDs -- a TV with only one component input is a mess, but controlling multiple TV inputs can be even worse).

Since Tivo is looking for manufacturing partners they need to make their software extensible with plugins. I, for one, would be very interested in writing a serial control plugin that would automatically control inputs and aspect ratios on my TV. Tivo should position themselves at the "operating system" level of AV automation. Other than extensibility, Tivo is a better mousetrap for people interested in HTPCs. They just need to take the next step.

Tablet PCs and DRM

TeleRead picked up my tablet user experience entry from yesterday.

DRM is a big problem for Ebooks -- and Tablet PCs. I don't read very much except for newspapers and periodicals. So when I do read, it doesn't bother me to restrict my intake to out of copyright classics (Dickens is a favorite) or cutting-edge people like Lawrence Lessig and Cory Doctorow. I think there is potential in ePeriodicals since I mainly throw them away when I'm finished. I'd gladly exchange long-term access and searching for the limitations of the eBook physical format.

But for mainstream novels, the equation is different. I mainly receive them as gifts, and often regift them when done (I don't loan from my personal library -- I give -- if I really want it again I'll ask for it back or buy another on Ebay). I have books that have been sitting in my to-read pile for 10 years. EBooks without DRM give me something more -- guaranteed retention, ability to regift, ease of searching, better user experience. I'm willing to pay full price. Ebooks with DRM need to come with a substantial price break. I just can't trust that they'll be on my to-read pile 10 years from now.

I think the long-term solution will be DRM on initial release with an open version escrowed by a third party for release 12-18 months later. The open version could have advertising (especially for new books by the same author) and serialization (to "imprint" my name on the book -- much like libraries of old).

But I can't give my mother (an avid reader) a Tablet PC -- even though it would be great for her aging eyes -- until DRM is solved. She would gladly pay a fair price (remove the middlemen and paper handling costs and there is plenty of profit for authors, editors, and publishers) for easy-to-use eBooks. Note to publishers: the buyers of books are not the Kazaa type. And no DRM solution I've seen so far is going to meet her requirements for ease of use, sharing, and long-term retention. DRM publishers are marketing to me (lucky to buy two books a year, tech-savvy) instead of her (75 books a year, tech clueless).

A Tablet PC Gripe

For all of Microsoft's vaunted user interface research, why is it so hard to enter a URL on a tablet PC? The handwriting recognizer insists on adding spaces everywhere. The recognizer needs to be context sensitive (if you're in the URL bar, force text into a valid URL). Edit: Apparently this is fixed in Lonestar (XP Service Pack 2).

Tablet PCs

So the new thing to do is bash Microsoft on Tablet PCs. I recently demoed one at work -- but returned it at the end of two weeks. It was an Acer Travelmate C300. Beautiful screen, but just too heavy. In fact, it reminded me of a pre-hard drive NEC laptop I purchased in college for $100 (and ran Qedit on -- I sold it two years later for $75). And not much relevance to my workstyle. While it is cool that you can annotate powerpoint presentations, frankly no one is interested in my annotations.

But then I found the killer application (thanks Cory Doctorow). Ebooks. A tablet PC is truly a better mousetrap. I think my reading productivity is up at least 10%. Easier on the eyes, easier to hold, easier to turn pages, less weight and volume if you carry more than one book. So when my wife needed a new laptop last week, I purchased her the Fujitsu T3010. The two pound difference is extremely noticeable (though there is no DVD or floppy). It also works quite well for playing online poker and reading the New York Times and Slashdot in tablet mode. The integrated 802.11g is outstanding (integrated reception is much, much better than add-on PCMCIA cards). A great "utility": the factory drive image is saved to another partition and additional user images can also be created and stored. (Right now, I'm trying to prep her old laptop for sale and can't find the original CDs -- so I am especially appreciative).

A tablet PC for the home is a luxury. But it sure is nice.