Thursday, February 28, 2002

Apple Visit Some folks from Apple came today to present at Chicago GSB. It was a good presentation, with a very nice turnout from both the computer science department and the business school. They also brought their new iMac, which everyone could ooh and ahh over afterwards. Nice.

Takeaways from the visit: Apple's abandoned their platform strategy and is now trying to be as good a Net citizen as it can while adding value to verticals. They're still struggling to be taken seriously by ISVs (and maybe the sales force isn't focused enough on this) and have no eyes towards the corporate market. This might be bad for Apple, as selling into MIS departments would be a big coup for them. But it's not clear how they would do this since low maintenance Macs just take power from MIS without giving them any control (at least with a Windows network, MIS guys are very needed). Locally served apps on the other hand give MIS power and low maintenance, so that might be something they would like. But why use high end hardware to display screen scrapes?

Next vertical: Silicon Graphics.

Economics of Tauzin-Dingell The House passed the Tauzin-Dingell bill that repeals 1996 legislation requiring incumbent local loop telcos to provide access to bottlenecks at "just and non-discriminatory prices." They also had to offer network elements a la carte (unbundled), allow buyers to resell, and local carriers could only enter the long-distance market if they opened the local market.

This regulation was an utter failure. It was impossible to define what a "just and non-discriminatory" price was, so there's a huge line of litigants wanting the FCC to arbitrate their greivances. Local access prices have remained the same (or risen), and most new telco entrants have gone bust. Broadband deployment remain piteous, with high prices and poor service. Given that this legislation has done no good, rolling it back shouldn't be a big deal.

The central problem is that the local loop is a natural monopoly. If the government sets access charges at marginal cost, then the local incumbent cannot cover its fixed costs and will not upgrade the network. Prices might be set low, but they will never fall. On the other hand, if the incumbent sets access charges at the monopoly price, it will have incentive to upgrade the network, but only some of this efficiency will trickle through to consumers. So prices will be higher, but they will fall.

The 1996 legislation used TELRIC pricing that tried to set access costs somewhere between the marginal cost and monopoly price. It was complicated, murky, and gave entrants and incumbents little incentive to invest in new networks (and so lower costs). It did nothing to help competition, the industry, or consumers.

Tauzen-Dingell essentially rolls back this legislation giving local loops monopoly power. It's not clear if it will pass the Senate, or in what form, but it's not bad news for consumers.

Tuesday, February 26, 2002

Usability is not customer experience Evan reviews Spolsky's UI book, quoting
Usability is not everything. If usability engineers designed a nightclub, it would be clean, quiet, brightly lit, with lots of places to sit down, plenty of bartenders, menus written in 18-point sans-serif, and easy-to-find bathrooms. But nobody would be there. They would all be down the street at Coyote Ugly pouring beer on each other.
No arguments there, but there's a difference between usability and customer experience (as Mark Hurst has been going on about for years): customer experience delivers what the customer wants, usability makes it easy to use. Customers don't want a clean, well-lit nightclub, they want dirty, dark bachanalia, so give it to them. Othertimes usability and customer experience are the same thing. But customer experience comes before usability.

Keeping actors expendible I wrote about how the recording industry cartel relied on draconian labor contracts to earn profit once their investment in the "unknown talent" had turned that commodity into a "star" (that could now demand more of the profit). The incomparably excellent Surowiecki writes about how franchise shows (like Law & Order) keep the bargaining power of actors down by not creating stars (like Frasier). And while franchise producer Dick Wolf claims the success of shows lie in the writing, I doubt writers manage to appropriate much of the value a hit franchise creates either, because faceless writers are as interchangable as ensemble actors (UK TV might be different here, as comedy writers in particular are quite well known). Thus successful producers (and studios) would keep most of the value as they own the franchise. Clearly this is a better system than the indentured servitude demanded by the recording industry.

Sun's hardware and software is good A while ago I claimed that Sun was the Apple of the Unix world because they tied hardware to software. My most excellent friend BE writes in with these comments:
on tying OS to hardware: i don't really consider apple or microsoft as contenders in the serious computing category, and for this reason i think your likening of them to sun is not really on point. here's the difference: apple and microsoft are in the business of providing pleasant UIs for mundane tasks (email, internet, word processing). or at least, they should be. redmond's every effort to build a serious big-iron platform has fallen flat. the only guys buying NT servers are people with lots of money and no expertise, or medium/small enterprises. this is fine, as microsoft can fill a niche in this particular market. however, they have no hold on the buyers in the serious unix market that Sun, IBM, and possibly HP hold, along with others like SCO and EMC. so, sun is *not* the apple of the unix world because its hardware addiction is voluntary and extremely beneficial. (apple's is no longer by choice, and it will eventually kill them).

running, configuring, and maintaining solaris 8 on x86 vs SPARC is a world of difference, and, as someone who does just this on both platforms for a living and has taken a bit of time to understand the differences (Solaris Internals: Core Kernel Architecture -- a must-read if you're going to comment heavily on hardware "addiction" vs platform independence vis a vis sun), i'll readily attest to the fact that the SPARC architecture has been tightly integrated to some of the most important and advanced features in the solaris system. in other words, when you control the chip architecture and the OS, you can hook them together at key points in ways that let you do hard things more easily. moreover, the fact that both are under one roof has allowed sun to radically alter both to fit the changing needs of modern computing (this will also come out in a close reading of the book above, and i'd be happy to comment in more detail if you're interested).

IBM has done much of the same thing with CCS and other platforms with varying degrees of success; EMC holds a lock on the big-iron disk solution because it uses its own hardware, with its own version of SCSI protocol to make stuff happen faster. did this make it hard for sun to port their OS? well, the source is open, so you can decide for yourself. but, in my reading, the answer is a pretty firm "no." that's why they were able to put solaris on x86 in the first place. that they have decided to put a hiatus on multiplatform support does not surprise me either -- in a downturn, you want to circle the wagons around your best performers instead of diverting precious capital towards laggards.

from an administrative and development standpoint, there is no comparison between something like Solaris on SPARC, BSD on anything, or other linuxes mixed and matched with various filesystems, memory managment strategies, and hardware drivers. microsoft can't ever touch this hold on "serious computing" for the same reason that they're in trouble with the law: they don't play nice with the other kids. what i mean is that microsoft doesn't benefit from things like all of the myriad standard unix utilities, the standards-based filesystem architecture (which allows folks like veritas and open source developers to build truly great products), and, soon, they will not benefit from the rich contributions in the field of networking (once .NET becomes an addled, stumbling reality, in my preditions, at least).

they get none of the 40 years experience and cumulative development that the unix world has because they refuse to play nice -- SAMBA is a great example of this (MS would do very well to help the project, as someone like sun or hp surely would; instead, they work hard on crippling it). as a result, microsoft gets to win the desktop war (for now), since the unix community never concerned itself with this and no one else played the game as well as they did in the early years. however, for all that being big helps them in their fight, it isn't the only factor or even the most important one as they saddle up for the race against the ibms and suns of the world.

The key argument here is that Sun owning both the upstream (hardware) and downstream (OS) business allows them to capture all the value from investments that benefit both parties, so there won't be the underinvestment you'd see if these were split. Apple, meanwhile, failed to invest in innovations that created value from both software *and* hardware, turning that asset into a liability. Apple has changed its position on this, getting out of the kernal business (Darwin) and developing the iPod, which delivers a better experience because it runs on combined hardware and software. I don't use the Sun platform and don't know how they use hardware/OS integration to their benefit, but presumably Solaris Internals could set me straight there.

Monday, February 25, 2002

Microsoft shares source I read Apple by Jim Carlton over the weekend, and was struck by how clueless Jean Louis Gasee was about technology strategy (and the importance of application barriers to entry). Moreover, Gasee seemed to have learned nothing in his time at Apple, because Be failed for the same reason Apple struggled--no apps.

I was also struck by how Apple just could not shake the hardware addiction and move to a software only company, even though Billg himself urged Apple to do exactly that in the 80s. Microsoft is stuck in a similar conundrum, addicted to OS while it's really an application company (and seeing GNU/Linux eat the OS market around it). Bizarrely, they've decided to share source with some integrators, instead of more sensibly offering .NET applications on all platforms.

Saturday, February 23, 2002

Love I absolutely adore my new iPod. *Adore*. Am currently listening to The Orb: Adventures Beyond the Ultraworld.

Friday, February 22, 2002

Limited vs. Unlimited recap Just to look at the economic calculation of limited vs. unlimited once again: treating copyright as a serious of annuities let's people have a sensible discussion over how long copyright should be before it becomes, in essence, "unlimited" (and hence, unconstitutional). If people just argue over length in years, it's very difficult to agree on anything -- is 100 years more or less justified than 50? By how much?

But by recasting the question as "how close in % terms can you get to 'unlimited' and still count as 'limited'" leads to more intelligent conversations. You've got to be on someone's payroll to claim that 99% of unlimited equals limited, but people can disagree whether the level should be set at 95%, 50%, or 30%.

At 5% discount rate:
95% of unlimited is 60 yrs
50% of unlimited is 14 yrs
30% of unlimited is 7 yrs

At 7% discount rate:
95% of unlimited is 43 yrs
50% of unlimited is 10 yrs
30% of unlimited is 7 yrs

Thursday, February 21, 2002

Limited vs. Unlimited The Constitution states that Congress only has the ability to set copyright to a limited term. But how long does "limited" have to be until it becomes, in essence, "unlimited"?

The way to work this out is to think of copyright payments as an annuity, pick a discount rate, and see how close every limited term's net present value is to the unlimited term's net present value. People can have an intelligent conversation about how close a limited term needs to get to an unlimited term in % terms until it becomes, in essence, unlimited.

The formula for the net present value of an annuity is (1 - (1 + discount rate) ^ -years)/discount rate.

At 5% discount rate:
10 year duration is 39% to unlimited term
20 year duration is 62% to unlimited term
50 year duration is 91% to unlimited term
100 year duration is 99% to unlimited term

At 7% discount rate:
10 year duration is 49% to unlimited term
20 year duration is 74% to unlimited term
50 year duration is 97% to unlimited term
100 year duration is 100% to unlimited term

Scalia, being a strict constructionist, will almost certainly rule that copyright extensions cannot be applied retroactively. If he takes his Chicago Law & Economics approach to defining when "limited" becomes "unlimited", we should see copyright being restricted to 30-50 years. Nice.

Wednesday, February 20, 2002

The copyright tax Jim Valenti asked Lawrence Lessig "who does it harm if Mickey Mouse is under copyright for 1,000 yrs?" Lessig struggled to answer, but here's some math that points us in the right direction.

Let's look at music sales. This year, $763M CDs were sold in the US, average price $15. Price elasticity for CDs (in New Zealand--all I could find on google) is -0.5. Assume it's the same in the US. Marginal cost for music (a la Napster) is $0 (note--this is marginal cost, not average variable cost, so chill out about up front costs).

You do the math and come up with the following graph:

What does this mean? If music was available at marginal cost, total good to society would be $25,740M, all of it going to consumers. As things stand, society only gets $22,890, split evenly between producers and consumers. There's a dead weight loss (i.e. money that's burnt to collect this "tax") of $2,857M. This means that market for music is $2,857M smaller than it would be in an efficient market. This is equivalent to a copyright tax rate of about 12% (income tax efficiency is about 20% I think). So, for CDs, you can claim each additional year of copyright costs society $2,857M in dead weight loss.

But this is just for a single year. Copyright currently lasts for authors life plus 70 years (and counting). So, let's treat this as a 140 year annuity of $2,857M at a 5% discount rate. Calculating the present value gives you: 19.978 * $2,857M = $57,078M. This is the present value of the cost to society of a 140 year copyright tax. Coming back to Valenti, if this was increased to 1000 years, it would cost society approximately the same amount, since costs that far in the future are so deeply discounted they fall to zero (i.e. present value multiplier at 1000 years = present value multiplier at infinite years = 20).

So from a financial standpoint, the length of copyright currently might as well be infinite, as it costs society approximately that much. This clearly goes against the "for limited duration" section in the constitution. So when someone asks you "who does extending copyright hurt?" you can answer "everyone, to a tune of about $3B a year just for music." You could also add that the economic length of copyright is currently essentially infinite. Any argument for extending copyright further is about control, not incentive to produce, since it's impossible to increase the incentive to produce any more.
Link to this article

Tuesday, February 19, 2002

Eldard v Ashcroft It looks like Lessig and Co's petition to declare the Mickey Mouse Copyright Act (Sonny Bono Copyright Extension Act) unconstitutional might go back to the Supreme Court on appeal.

Lessig & Co. argued that a retroactive extension of copyright did nothing to "promote the Progress of Science and useful Arts"--which if course is 100% correct. The Supremes disagreed, arguing there are "no First Amendment rights to use the copyrighted works of others" (which misses the point). They also declared that the deprivation to society of keeping works out of the public domain is "threatened as much 'for works not yet created as for extant works on which the copyrights are about to expire'" essentially overlooking the economic difference between longer copyright and retroactive copyright extension. This is just plain wrong.

There's too much Disney money behind perrenial copyright for Congress to do the right thing, so the Supreme Court is it. I'll also post an economic analysis of copyright term to put a dollar value on extension Sometime Soon.

Correction Larry Stanton points out that the DC Circuit negged the original motion. This is the first time the Supremes are getting into the picture. Thanks Larry!

Monday, February 18, 2002

Server consolidation Two major items on every CTO's mind: server consolidation and reducing cost. The former is anaethema to Microsoft, who can't shake its NT (aka server .NET) habit which is hoping shrinkwrapped high-volume code running on cheap x86s will handle the latter. IBM is moving downstream with its baby mainframe running everything from GNU/Linux to zOS/e. Easy to administer, cheap apps running on x86s (or comparably cheap hardware) will win in the low-medium size corporate space.

Quick links Some quick links on IP issues today (NYTimes, subs required):

There is no standard payment mechanism for Internet radio. Copyright hoarders want to require new and different rates for different codecs and qualities. AOL is on both sides of this issue.

The recording biz is about advances, not royalties. Therefore, pressplay and musicnet don't open new revenue streams for artists. But if audiences paid the advances directly, artists could release on Napster with no problems.

Instead of filing patents, some firms are making info public so they can't be blocked by patents in the future. This is a clear example of how patents are being used to block innovation, not commercialize it.

Tureblog argues that Spolsky was advocating managing expectations by controlling UI mockups, not making bad software. That's fine, but it's really not good enough. There's too much bad software in the world to justify less painful production of more bad software.

Sunday, February 17, 2002

Cringely on crack Cringely advocates Apple offering Mac OS X on x86 architecture, arguing that software is higher margin than hardware, and this would increase Apple's profits without damaging their hardware sales.

I say Cringely is on crack, and Apple is too. Cringely is dead wrong saying that OS X on cheap x86 architecture would not cut into Apple's hardware sales--the PowerMac G4 in particular would be utterly decimated. And Apple can't shake it's hardware addiction, it *has* to sell boxes to manage revenue and profit in the near (1-2 year) term. Sun also has this hardware addiction, making it the Apple of the Unix world. Similarly, Microsoft has an OS addiction, meaning its new application servers *must* run on Windows NT (gussied up as .NET servers). This makes them supremely vulnerable to a vendor who just makes applications that run on commodity OS (GNU/Linux) running on commodity hardware (x86). Microsoft's OS habit is the greatest threat to its .NET strategy.

Ugly advice update My friend RE responded to a recent post lambasting Spolsky's cynical attitude towards producing good software by writing:
I think that he goes too far in writing off the possibility of real customer-developer cooperation - although I believe that it is possible, I agree with him that it is RARE. I think he's a little unfair to XP on this point - XP explicitly calls for an involved customer who is a (the) major project stakeholder, who is ultimately responsible for its success or failure, definitely not some dweeb from accounts payable.

I'm surprised he didn't mention is incremental development. I consider it a to be a Scientific Fact that projects that lack it are doomed, doomed, doomed. it's also a useful tactic for dealing with a less-than-well-involved customer - if you can at least get the customer's attention once a month, every time you complete an increment and start a new one, you can keep steering the project in the direction that the customer wants/needs. (no surprise that this is a core XP practice as well).

So, Archipelago asked me to propose a solution instead of just ranting. Here goes:
1) Extreme programming brings users into the development phase and builds incrementally, making it a good model for coding off behavior. Joel dismisses XP as being "too extreme"--but I think it's piteous that including users in developing the tools they're going to be using is considered "extreme" at all.

2) Develop use cases using Creative Goodstyle customer experience methodology. XP uses use cases as roadmaps for development. Non-directive, open ended, contextual user testing is a good way to nail the mission critical feature set and control scope.

3) Create an economy that values productivity. This part is tricky--the real reason technology is so hard to use is because ease-of-use often isn't critical for business success (Microsoft beat Apple for strategic reasons even though Macs have always been easier to use). Websites, particularly e-commerce, is an area where usability is critical to success--which is clear when you look at the current winners (Google, Amazon, Yahoo!). Software-as-a-service (NOT as rental) will also win in this economy. The main goal of is to understand why technology is so bad by looking at the underlying economics.

Thursday, February 14, 2002

Ugly advice Joel Spolsky, who I usually think a lot of, has written a cycnical and damaging account of how to deal with clients during a software project. While I agree managers can be inept, and clients can have no clue as to what they're doing, the mission of a consultant is to help their clients be more successful. In the case of software development, this means doing whatever it takes to produce a product that their customers can use and drives the clients' business forward. This may mean bringing customers into the design cycle, managing expectations, and being jerked around, but the solution is not to pander to corporate myopia and produce software that helps neither the business nor its customers. Do demos and screenshots seduce people? Sure. But at the end of the day, somebody's business is going to rely on what you produce and I beleive it is unethical to knowingly create a product that doesn't improve the life of the end customer (if you don't know how you're merely incompetant).

Razorfish, Scient, Viant and all their ilk did business the Spolsky way, focusing on flashy demos and misleading screenshots that turned the web into a nightmare of sites that were slow, hard to navigate, and benefited neither customers nor businesses. Creative Good has been harping on about this for years.

If you want to find the real iceberg in the technology world, it's that the 1% of people who create technology are utterly clueless at figuring out what 99% of people actually want. The technology world is a wasteland of pointlessly hard to use products that don't benefit anyone and were expensive to produce. Joel might be OK with this--I'm not.

Wednesday, February 13, 2002

Windows in the Enterprise I read Breaking Windows yesterday, which recounts how Bill Gates' unyeilding attitude towards the Justice Department caused its current legal woes. The book ends with a sunny chapter on Web Services, and how Microsoft is excited about a bright future for the controversial .NET.

The truth of the matter is that .NET is far from nailed, and we'll probably see egregious things like Hailstorm be rejected by a suspicious market. Moreover, 0 utility plus 0 bandwidth conspire to make .NET services essentially useless for home users (where Apple offers the best system, bar none). Basically, people don't need Word to be a server. So that leaves the enterprise, where Microsoft has fierce competitors, where they offer little of competitive value (except for "richer client experience", which enterprises don't care about--remember, they were quite happy using windows), and where they have no ISV support (all of whom prefer GNU/Linux).

Tuesday, February 12, 2002

Artist Rights It's articles like this that made me stop reading Slate. That and the site's terrible redesign. The article critiques the Recording Artists Coalition's arguing that it's proposed labor law changes (7 year ceiling on recording contracts, weakening work for hire clause) will limit the recording industries ability to make profit and so reduce their investments in new talent. But for those of us who frown on indentured servitude, the recording industry should not be allowed to own its labor supply any more than Hollywood could under the studio system. The fundamental problem is that artists are in very weak bargaining positions before being picked up by a label, and in a much stronger position after (if) they become hits. So who should appropriate the profit from hit artists? Draconian labor contracts benefit the recording companies, as they can force artists to work for low wages even if they're wildly popular. Weaker labor contracts mean that artists get to keep most of their value if they become hits, and studios will earn lower profits unless they find better and/or cheaper ways to find profitable acts. Every other industry manages fine with more limited power of their labor suppliers, so I see no reason why the music biz should be any different.

Monday, February 11, 2002

Web services I asked Redhat CEO Mathew Szulick where he thought GNU/Linux was going in the enterprise, and he said "operating systems follow applications." Three days later, Google releases it's search appliance and with it, the clearest example of an application server. Application servers:
  • Deploy over LANs because of security and bandwidth constraints
  • Simply provide a service while playing well with others
  • Run on GNU/Linux
Very nice. I wish I had info on pricing.

Apologies to Miguel In case it wasn't clear earlier, I'm issuing an apology for my alarmist comments regarding GNOME and .NET. As RMS pointed out, adding C# compatibility to GNOME is fine (good luck), but GNOME itself is not being ported to .NET APIs. Another upshot is that no one really understands what .NET is, which is unsurprising given how many different initiatives have been packed under that umbrella term. Also, to state the obvious, no one beleives Microsoft's descriptions of .NET either, as the ISV community is convinced Microsoft has sneaky tricks up its sleeve. A single company that bundles OS with applications is not a safe platform for ISVs (and by extensions, end customers).

Patent madness In 1976, BT managed to patent hyperlinks. It stumbled upon this patent about a year ago, and now is trying to collect a toll every time something blue and underlined is clicked. This case illustrates how software patent law fails to generate innovation, while BT had they patent they did nothing to promote the Web. Instead, they're trying to use their IP to extort money from everyone else. I don't know if their appeal will be successful, but this case provides good ammunition for those discussing how patent law should be fixed.

Sunday, February 10, 2002

Cyberposium day 3 Dean Kamen spoke about the Segway in what was probably the best and most peculiar keynote of the weekend. He kept the Segway secret for so long was not because of the hype, but because he didn't want people to think of it as a scooter. He genuinely hopes it will change the way cities are planned by easing the 1-10 mile commute, and so address problems of congestion and pollution. Kamen was disdainful towards the avarice and greed characteristic of MBAs, and urged students to focus their efforts on the big social problems.

Saturday, February 09, 2002

Cyberposium day 2 Tom Seibel canceled his speech this morning, so Matthew Szulik took over instead. I was thrilled at this, because I was more interested in what Szulik had to say than Seibel. He delivered a great address, focusing on the importance of choice, and how that forces technical integrity and high standards. This is right on--locking customers in destroys incentive to improve software in a meaningful way and shifts the vendors focus to driving upgrades and further entrenchment. The complete stagnation of not just the Office suite, but the Microsoft desktop itself is a good example of this dynamic. Red Hat's problem is capturing enough of the value it creates to remain viable and grow. They invest 20% of revenue (not profits) into R&D that feeds back into the Open Source community and I don't know how they plan to recoup this investment.

I also attended a panel on Web Services with Sun, Microsoft, IBM, and Accenture (important in the integration space). Much of the discussion focused around standards creation, with IBM commenting that collaborative development usually isn't very successful here, they're usually driven by some small group just declaring them and rolling them out. This fits well with Winer's SOAP experience, (SOAP was mentioned often, Userland never). It's also the story behind HTML. Simple, open standards are useful to developers.

Microsoft was the the villain in the group, and I don't think their cause is helped when they cite kereberos was a way in which they're embracing open standards (seeing as how it involved abrupt proprietary extensions). Steven Lewis (the Microsoft representative) also spoke heresy when he said Microsoft did not care about Hailstorm and was thinking of joining the Liberty Alliance. We'll see what happens.

(Had a worthwhile hallway conversation with James Adamczyk of Accenture (who sold mustard online for a while) about the semantic problems of webservices. Essentially, interop does not happen because of business reasons, not technology reasons. XML might have solved some syntatic niggles, but the real business value lies in semantics, and whether or not there's a market for interoperability. Often there might not be.)

Other highlights: DRM folks see privacy as being the major growth markets for these systems, allowing sensitive documents like medical records to be passed around online. In the entertainment sector, they primarily see DRM as a mechanism for perfect price discrimination (although they did not use that term) which would transfer all surplus from producer to consumer, perhaps increasing efficiency in the process. If this space is competitive, bare MP3s just like those used today will win by offering maximum surplus, and all the DRM systems will be worth squat. Had a good conversation with Jorge Contreras, Jr. a lawyer for Hale and Dorr (who were on the right side of the challenge to the Michey Mouse Copyright Act) who beleives the DMCA will be weakened.

Oh yeah, it's hard to sell big expensive software in a downturn. Some firms, like Juniper, bring in very beta products, get access to source in case the company goes under, negotiate cheap terms, and has excellent scripters glue everything together with Perl. Sounds like a good way of building a robust, heterogeneous IT environment.

Friday, February 08, 2002

Cyberposium day 1 I attended opening ceremonies at Harvard Business School's Cyberposium today. I'm told it was a tepid affair compared to previous years, but I thought it was still a great turnout for a student run tech conference. The opening panel was with Esther Dyson, with Phil Terry mediating. I struggled to get much out of Dyson's comments, but it's always great to see Phil working the crowd. Phil was my boss at Creative Good, he's a good guy, and the firm does good work.

Then there was a panel featuring the founders of Lycos, Bowstreet, Third Age, and the CEO of Red Hat, Mathew Szulik. Szulik was very personable, and was on top of important issues like DRM, SSSCA, and what Open Source is really all about (which Dyson did not mention). Mind you, he also made worrying comments like "We change business plans hourly" but overall he struck me as being the smartest guy on the panel (except for perhaps the Lycos fellow, Bob Davis, who sold when the selling was good).

One thing that surprised me was how focused the group was on culture and people. While culture and people are fine, my time at Chicago is really driving home the importance of sound economics--you essentially want some form of monopoly position you can use to expropriate economic rents from others, while making sure those rents don't get expropriated from you in turn. Many great "culture stories" (like the media darling, South West) can be better explained by microeconomic analysis. But the media's not dumb, they know what sells copy, and CEOs aren't dumb either, they know what journalists need to put in their copy so it sells.

Fair use thumbnails The frequently overturned 9th circuit court has ruled that thumbnails of pictures are OK under fair use, but displaying fullsize images through framing and linking harmed the market for original photographs, and was not allowed. But this is only true though if the original photos are available for sale. If the originals are not commercially available, or tied to commercial gain, then they should be linkable and frameable.

Thursday, February 07, 2002

Whither broadband? Microsoft will not be able to deliver applications (or experiences) that are any good over a network unless low-latency, high-bandwidth networks become widely available. And given the state of cable and DSL, that's not going to happen anytime soon. This poses a key threat to .NET service revenue taking off within the next five years, which means Microsoft will have to continue relying on costly Windows and Office XP upgrades (which just drives customers into the arms of GNU/Linux anyways).

I met the Director of supply chain from Juniper earlier today who talked about how telcos were the bread-and-butter of their core business, and how broadband was going to drive demand for their new edge-network products. Given that telcos are toast, and broadband isn't going to happen anytime soon, I'd say the company is in serious trouble from a revenue growth standpoint. The markets figured this out a while ago, stocks trading at 13.50, down from over 100 a year ago. While operationally Juniper is very advanced, I don't know strategically how they're going to get out of this position.

Wednesday, February 06, 2002

Update on GNOME/.NET Great thread on slashdot critiquing Miguel's lame explanation for Mono .NET support. Alan Cox himself puts it well: Be assured that the day [Microsoft] decide[s] you are a nuisance the [.NET] VM will acquire a patented neat feature that kills you off. Just ask the Samba people. Or Kereberos.

Also, if the pixie-dust magic of a runtime environment converting all languages into compatible bytecode sounds too good to be true: it is. Repeat after me--there is No Silver Bullet.

But if you don't like reading 15 year old refutations, here's a sanity check from the Java Lobby. They get the essential point--C# is a runtime environment, or operating system middleware, that binds applications (or "experiences") to a single platform: Windows. This is equivelent to saying it's an Internet fork, splitting the world of networked bits between Microsoft, and everyone else (i.e. *nix platforms). Half-baked cross platform language compatibility, which is really the only sort of compatibility you can expect, cripples languages, for exactly that reason.

Further update RMS writes explaining what really went on: he was saying he didn't want the GNOME license changed, didn't beleive Miguel would change the license, and is fine with free software supporting as many languages as possible, including C#. Also, a friend of mine beleives the CLR is less lousy than the above suggests, and says the true test will come when it tries to handle the ramshakle C++. We shall see.

All your *movie* bits are belong to us Over a year ago I did some webwork for Madstone, a company in New York looking to digitize cinema projection. They also were into supporting indie directors, a pointless distraction in my opinion, which seems to be born out by the fact they've yet to produce any product or any service.

Microsoft's now offering the technology behind such distribution, the main goal being to convince Hollywood to move all its bits onto Microsoft DRM networks. Although the paucity of bandwidth (which will not get better ANY time soon) protects movies from being Napsterized, MPAA's gonna hook up with Microsoft anyway. This will reduce their profits, but effect customers little. All the bottlenecks in movie distribution have to do with cinema consolidation (and economics), not network infrastructure.

Tuesday, February 05, 2002

End of GNOME I guess it's official. Miguel's announcement to base GNOME on .NET has finished that desktop as a viable strategy. .NET is a mostly closed set of APIs Microsoft is rolling out to extend its Windows platform to non-Windows computers (making them effectively Windows computers in the process). Either Gnome developers will revolt, and the project will fail, or Microsoft will shift the APIs from under the successful project and so cripple it. Read all about it on Slashdot. Miguel posted his response a day later, and it's pretty lame. It ignores how Microsoft will tie .NET infrastructure with .NET applications, and how the functionality he seeks already exists with Java. Despite what Redmond wants people to think, .NET is still very much pie-in-the-sky.

Oh, and the Palm OS is finished too. Their new OS is saddled with technical gee-gaws that cost lots of money to produce, but confer no benefit on the end user, and probably hurt the quality of the experience. Since they can't hand out $400 with each unit (the way Microsoft does with its Pocket PC) their new product will tank in the market. Unable to amortize away the high development costs, the company will fold, probably taking the OS with it. And don't count on Palm putting the new OS on non-PC devices--embedded GNU/Linux is already there.

Monday, February 04, 2002

Trustworthy Computing I had a conversation with someone at Microsoft today about trustworthy computing. She assured me that the February Morotorium on new features was true, and that Microsoft was dedicated to this new vision of Trustworthy Computing. It's going to take more than a month of refactoring to get there though. Microsoft is going to have to change the way it fundamentally does business. So let's think about some catalytic mechanisms the company could adopt that would force real change:

1) Support immediate, full disclosure of vulnerabilities with exploits.
Microsoft can't hide from it's own bad code

2) Both Microsoft and independent developers to get all APIs of the same, open documentation.
No secret APIs. Microsoft must build better products based on superior development processes, not coercion

3) Open file formats.
People should use Microsoft applications because they're better, not because they're trapped by secret code

4) Split Applications from OS
Again, people should use Microsoft because they want to, not because they've been coerced into it.

Trust is having confidence that your supplier has your best interests at heart. Unfortunately, in a closed platform world, this is *only* true until the vendor has built up an installed base. After that point, a supplier's economic imperitive is to harvest monopoly rents, not to better serve customer needs. If Microsoft is serious about building trust, it must abandon platform economics. Kind of like Red Hat

Test Moved winterspeak to new servers over the weekend. Test blog.

Friday, February 01, 2002

GNOME moves to .NET I had lunch with someone who worked at Microsoft over the summer, and he was pretty candid about how the company openly, and routinely, plans to put competitors out of business and take over entire markets. .NET is their strategy to do this to the Internet, so I think Miguel de Icaza is dead wrong in porting GNOME to this platform. You can already do all .NET stuff with RPC and Perl, it's just a way for Microsoft to move developers (and applications) onto its API's which it will then change when appropriate. Miguel seems like a smart enough guy, so I'm pretty sure he's figured this out. Maybe he's cut a deal with Redmond.