(this is page 85 of 121)





Latest

Both Matt Webb and the Economist this week are ruminating on the economics of the software industry. Matt has a interesting post which basically argues that there are other more compelling revenue models for building enduser software than the straight-up license (a lesson which I learned well while book-enabling the iPhoto application while the rest of the consumer image management space was cratering). He covers ads, subscription fees, and some other models, making a whole bunch of interesting observations along the way.

It was the piece in the Economist though that really made me laugh. Especially this juicy bit on why the SaaS (software as service) business model just isn't as good as its progenitor:

Vendors of conventional enterprise software made a killing by requiring customers to pay a high licensing fee upfront and then charging them for maintenance. Web-based firms, by contrast, have to make do with subscription fees.

We need to stop taking for granted that just because software vendors were able to maintain obscenely high margins for the last 30 years, it means that this should serve as a baseline for all software related businesses going forward. In the consumer space specifically, Microsoft's ability to extract rent for that layer of value seems to me to have been a historical accident— one that made Bill Gates very wealthy in the ensuing 30 years— but not necessarily one that can be relied on going forward. I know less about the enterprise space, but would be shocked if the same microeconomic force of marginal revenue trending towards zero wasn't in full swing in the age of the Internet, open source, and software-on-demand.

Instead we have to get used to saying that 22-30% margins are actually a great achievement, and that the era of crazy absurd software profits (though not growth, as Apple is showing), is now behind us. And the bonus is that in the process we get to repurpose old business models (atoms!) and invent new ones.

Comments Permalink

Now with comment juice

I used to think that blogs were interesting because they represented an individual's raw voice unfiltered by editors and undisturbed by distractions. Thus I saw the introduction of comments into blogs as a weird adaptation from the land of forums. Pings and trackbacks seemed like a much better way for people to carry on a conversation in the blogosphere.

Unfortunately my characteristically Betamax instinct was on the money yet again, as it turns out that most of the people I want to carry on a conversation with aren't interested in the distributed computing showcases. When combined with the constant chiding that I take in the office for a) writing ridiculous things in this blog and b) not providing a vehicle for talking back, I've finally been worn down to a nubble. As of tonight, this blog is now commentable.

As always, I have to give credit to the wonderful Django framework (and it's sparsely documented Freecomment sub-framework), along with the very useful comment-utils (from James at the B-List) which provides Akismet filtering and basic moderation in a nice and transparent way. Trackbacks may not have won the day, but it's nice to see that web services like Akismet are now used as simply and plainly as one might have used a third-party library two years ago.

So if you're an RSS-only reader, hit the V key and come visit, stay for a while, and leave a comment or two.

Just not the guy who used to comment on my last blog— I don't know what V-ONE-AGRA is, but I don't think I need it...

Comments Permalink

Annoying robotic flies

Continuing on the hardware theme for the week, I was recently captivated by Adrian Bowyer's presentation at POPTech last year on self-replicating 3-D printers. Still very early on, these 3-D printers are essentially glorified glue guns mounted on harnesses that can use them to squirt out three-dimensional objects one layer at a time. The really neat hack to his work however is that the printers can "print" most of their own parts thus placing the devices themselves on an exponential, evolution-like curve where the random mutation role is played by people tinkering with the printers to make them better while the evolutionary cycles are fueled by these same people posting the improvements back to the Internet.

Imagine what technologies like these would do to another project I found this weekend at the Harvard microbotics lab, a robotic fly which can be manufactured cheaply and is almost exactly the size of a real fly (see the video).

How annoying is it going to be when robotic flies are capable of self-replicating and evolving on their own?

Comments Permalink

Hacking hardware

This is CES week which, ever since we were a struggling startup looking to stay alive in this space back at Memora, gives me the willies. Unlike most of the computer tradeshows, CES manages to feel polished and scummy at the same time— with vendors pimping their wares to a pulsating Las Vegas gluttony which never ceases.

This morning the New York Times had a piece which gives is me hope that consumer electronics may be getting some of that good old hacker ethic. The piece covers an open Tivo-like device from a company called Neuros that is "hacker friendly."

More significantly, companies like Chumby Industries and Bug Labs are going one step further by selling kits instead of finished products that can be used by folks to extend the reach of the net into the physical world.

I worry a little that the promise of hackability is a lot more powerful than what actually gets done. In the case of the PC for example, it was the shift from "kit computers" like the Apple I to the productized Apple ][, with its pre-assembled good looks and relative ease of use, that readied the entire industry for all of the software innovation that followed*. In my own case, I've dabbled with a number of Linux appliances over the years only to be disappointed by either too many bugs, or a lack of extensibility. My Chumby for instance, seems to be the net-connected version of a VCR clock, spending most of its days flashing the widget equivalent of 12:00AM (though I am hoping the official release will make this better).

That said, I'm still hoping that these kits will result in some useful devices that free us from thinking that the Internet is something we consume with a keyboard and a mouse.

* See the very awesome book, Fire in the Valley

Comments Permalink

The OLPC nightmare

The Economist (my favorite rag) has a scorching review of the OLPC XO that unfortunately is 100% right in all of its criticisms. I feel terrible for all of the people that I helped convince to pitch $400 down the drain on this thing; it is slow, buggy, and almost entirely unusable by people with advanced degrees, never mind small children. That said, I don't feel nearly as bad for all of the 1st world kids who've had their Christmases ruined by this plastic paperweight as I do about the kids in the third world whose governments are going to be bamboozled into buying XOs.

I remember having written about the OLPC before so I went back into the archives to see my very first post which was on Negroponte's announcement of the project at the 2005 D Conference. Apparently I was bamboozled as well and should have stuck with my first instinct.

The OLPC's software is just horrendous. The Etoys environment which I was so excited for is too cramped, too slow, and not at all simplified from Squeak Smalltalk. The browser is almost unusable, and the music application has such incredible latency between doing anything and hearing sound that you're more capable of controlling the noise the machine makes by waving your cellphone in front of it. Just about the only decent thing for a 5 year-old to do with it is to play the memory game or to practice typing in a word processor that makes Word 5 for the Mac appear supersonic (and tends to crash).

Why are they not using their OLPC?In fact just about the only thing you can do well with the laptop is drop it— which is good because I promise that after a couple of hours with it, you'll want to test its resilience to impact.

The OLPC foundation would do better using its clout to buy Eee PCs and tweaking what is a much more usable system overall. Or even getting old laptops reconditioned and re-imaged with stripped down versions of their original operating systems.

As always, Fake Steve says it better than I could.

Comments Permalink