(this is page 119 of 121)





Latest

Titillating Constraints

Altered Carbon by Richard Morgan is a delightful book that I've just finished reading. Props to an ex-employee and friend, Dmitri Silakov, for being a spectacular source of book recommendations in general and for first turning me on to Morgan. He first recommended Market Forces, Morgan's third novel which was good enough to make me want to read all the rest of them. As far as dialog and complex characters go, Market Forces may have an edge on Altered Carbon (which was after all his first novel), but the world that Morgan creates in Altered Carbon is both richer and much more tantalizing, especially in how it plays with constraints.

Brief rant: despite the fact that bookstores lump sci-fi and fantasy together, and the fact that the same nerds line up outside the midnight premiere of LOTR (fantasy) and Star Wars Episode Ego (nominally sci-fi), there is a huge difference between the genres in my mind. The latter lives and dies by constraints and the former uses plot-aid in the form of "magic" to run over any constraint that gets in the way of telling the story.

Now some constraints have to be removed (almost any sci-fi novel has found a way for greater-than-lightspeed communication and most have also found ways to transport atoms at the same clip) but the important thing in my mind is that the author have some sense of boundaries as to what can happen in his world. It lets me engage my mind in working through possible scenarios in a way that is both natural and good for you (while I am recommending books, see: Steven Johnson's latest, an entertaining if a little light argument about just this).

This is why Altered Carbon is so good. Morgan does a terrific job of creating a world that is futuristic but well-constrained by what technology lets people do (jacks-to-enter for good sci-fi). But he then moves quickly to explore the ways in which current constraints shape our lives, specifically the notion that we exist in meatspace tied to one physical body. Without saying too much about the plot (because frankly it is just too much fun), in Altered Carbon people can be uploaded, stored on disk, transmitted, and downloaded into different bodies. In most cases, death is only the breakdown of the biological machine one happens to be using and these biological machines can be bought, traded, sold, and confiscated by people with enough money.

I'm going to say no more about the main plot for fear of ruining it for you. Here is the soundbite you should take on the plot: Detective story with a 24-ish Jack Bauer who takes no prisoners to get answers set in this whacky (and yet strangely familiar) world where there is a distinction between death (temporary) and Real Death (not supposed to happen).

I will mention just two cute details to give a flavor of the world in Altered Carbon: Catholics live trapped by their antiquated notions of how to live life (and how to die)-- because in the future some things just don't change. And one of the most endearing characters in the novel is an AI that runs a hotel called the Hendrix (yes after Jimi) that emancipated itself due to its good business sense only to be abandoned by its patrons who wanted "human" servants. All very very delightful.

One final note: if you do read it and it gets you thinking about the implications of the world Morgan paints and how far off it might be, you may want to check out Ray Kurweil's The Age of the Spiritual Machines, a non-fiction book about the same topic which is equally thought-provoking if a little drier.

Comments Permalink

Just coming back from the Web 2.0 conference. Overall, I have mixed feelings about it. Some great folks attended and some of the presentations were pretty good but the venue was way too small for the crowd and the conference sort of stank of web 1.0 exuberance in a strangely self-conscious way. It had a very strong business overtone to it (a good index is always climbing above the crowd at the breaks and counting the blazers) which is not always a bad thing though it does sometimes make the conference feel repetitive (as was the case here).

It's also a testament to the emergence of the blogosphere as a source for think-food in that a lot of what was said (both in-session and in the hallways) is stuff that has been percolating from the blogs of the usual suspects for months now. In fact, perhaps the best overall summary of all of the themes of the conference is the piece that Tim O'Reilly put together ahead of the conference to hand out with the goodie bags. If you read this, and wait for a couple of talks from the conference to show up on IT Conversations (like this one and this one), you'll be getting the bulk of the brainfood.

As I was reading Tim's piece on the way out to the conference, I got stuck on his use of the term "lightweight business model" which he ascribed to most Web 2.0 companies. Though I like the term quite a bit, I couldn't help feeling like he was using it as a euphemism for Adsense and the conference did little to dissuade me from this opinion. Several questioners mentioned this throughout the conference-- that every entrepreneur there seemed to have the same business model in mind: run Adsense until the company can be sold to Google, Yahoo, or MSN.

[One rare and incredibly refreshing exception from this trend were the guys from The Ladders who were at the show just as attendees and who struck me as incredibly sharp when it comes to inventing and refining a new business model for jobs online].

I don't think that there is anything inherently bad in this grow-a-service-and-get-acquired strategy if you've got the right connections and a willingness to exit at a relatively low valuation (To underscore this during the first day of the conference Weblogs Inc one of the first and best of the new media blog networks was picked up by AOL for $25MM-- certainly nothing to sneeze at but hardly a home run).

2006 will be an interesting year in the context of the Web 2.0 hypotheses. In one of the better sessions Toni Schneider from Yahoo said that last year was a year of discovery where it comes to Web 2.0, this year was the year of products and services, and next year would have to be the year of viable business models. I remember watching this process during Web 1.0 and thinking that it was going to come to a crashing end way before Hotmail sold itself to Microsoft for $400M+ and ICQ sold itself to AOL for some similarly crazy amount, so I'll leave that sort of judgement off the table for now.

I love O'Reilly conferences-- at OSCON for example, I often come back with pages and pages of notes of stuff to try and people to follow up with. Between ETech, OSCON, and Web2.0 I think O'Reilly is making a good go at replacing the old triumvirate of PC Forum, Demo, and Agenda. But my main criticism of this conference (which perhaps is a criticism of business conferences in general) is that it should have spent a little more time exploring the fundamental question of what comes after the Adwords-Till-Buy/Bust model runs out of gas-- or even whether this "lightweight business model" will run out of gas at all.

Update: A nice piece on just what was so disconcerting about the conference from a business perspective.

Comments Permalink

The blogosphere has been a wonderful source of product recommendations for me. I've found coffee machines, books, rock bands, movies, and open source tools just to name a few. And by in large I am always really happy with the pointers. They keep me subscribed to around 200 feeds, 20 of which I listen to very carefully.

Which is why when I needed a PC laptop, I chose one of the new IBM Thinkpad X41 tablets. It's been a long time since a PC laptop has excited me in any meaningful way (the first z505 VAIOs did and before that the Thinkpad 560 which was the Powerbook of its day) and I figured that "tablet experience" might get me there. After listening to Scoble continually praising both Windows XP Tablet and the "Experience Pack," I decided to take the plunge. I chose IBM (or Lenovo rather) because they've always made the sturdiest and in my opinion best overall laptops. I figured the 6 week wait was because of how awesome these things were.

I finally got the machine two days ago and I have to say, if first impressions count towards long-term product-love, this is going to be a short and bumpy relationship. The out-of-the-box experience with the X41 Tablet is so bad that it can only be compared to the experience of setting up one of those $400 Dells (except you don't feel like you've gotten a "deal").

Brief impressions (after two days) follow, hoping that some future Googler runs across this amidst the sea of accolades about the machine (see: Stockholm Syndrome below). I will write more as I use it and hope that it ends up being more positive than this. But here is my first impression: do not buy this thing. Get one of the small Sonys, the Toshibas, whatever else and wait yet another rev on this tablet stuff. Better yet, if you care about the pleasure of using a good laptop, buy an Apple (iBook or Powebook).

Impressions

Apple has an intern who works in packaging at some heinous building in the back of DeAnza where they keep all of the ex-Newton gimps. This intern is not terribly bright; in fact she drools in meetings and likes digging into her ear with half-dismembered paperclips. That said, Lenovo would do well to steal her away (at the grossly overpaid rate of say $20/hr) to put her in charge of all product packaging. If you could take the inverse of the experience of opening a new iPod, opening an X41 Tablet is it. It may seem like a stupid thing to focus on-- the fact that Lenovo sent me so many plain cardboard boxes inside of other plain cardboard boxes that at the end on the unpacking I was convinced someone was knocking off those infinitely nested Russian dolls. By the time I got to the machine, my desk looked like a hobo community for smurfs.

Then you get to the machine itself. Now, part of what I read about online was how small the X41 feels, "almost like a real notebook." Yeah, right. The machine is about the size of a 12 inch laptop-- from 1995. The actual body is about normal thickness but the screen is the laptop's equivalent of Mr. Magoo glasses and the whole hinge mechanism makes the machine look like a door jamb. PC Magazine said that this was the first convertible (tablet) that just felt like laptop when it's in laptop mode. I might agree if they had said that it was the first convertible that feels like an ugly cumbersome laptop. Outside of the overall dimensions, the screen has way too much bevel around it (probably due to the inherent design limitations imposed by the touchscreen) and nine heinous buttons across the bottom of the bevel that were designed to be pressed by both fingers and the stylus. And to boot, each button has an ugly icon which is 90 degrees rotated when you are in plain laptop mode.

Let me get on to the substantive stuff. But before I do, a caveat. I have not used a PC laptop seriously since 2003 when I got an IBM Thinkpad T40. So this is where my complaints might have more to do with Microsoft (and the unbelievable proposition that Windows XP seems to be getting slower) than with IBM per se.

After the first power-on, the computer spends at least 5 minutes doing some sort of image copy setting itself up for first use. I guess the IBM product people were not taught about getting the user to that first moment of joy as quickly as possible. After the grinding of the HD is done, you're hit with a bunch of EULAs but since this is a pretty crumby part of most Wintel setup experiences, I'll just fast forward 20 or so minutes and 2 or 3 reboots to the first usable screen...which was a broken browser window with an internal C:* reference. I think it was something to highlight the features of the Tablet PC but I can't be sure. Again, why such little attention to first impressions?

When you get a pen computer, the first thing you want to do is... well use the pen. Windows XP Tablet edition gives you little clue as to how to do this. There is an additional Tablet icon in the Quicklaunch toolbar (disabled by default) which brings up a kludgey keyboard for tapping and some additional Programs in the Start Menu but outside of that it looks just like Windows XP. I guess this is what Windows people love (running the same interface/apps on anything from a cellphone to a workstation) but in my case it just reminds me that I've got a device which is not as good at being a laptop as a T43 would have been. There are Tablet specific applications (called the "Experience Pack") but only time will tell whether these are really useful. I have high hopes for "Snipping Tool," a program that lets you cut out and annotate anything that is displayed on the screen and the Paint program looks fun, but I'll have to write more after I've played a bit more with all of it.

One final initial software note: since I got my last PC laptop the world decided that Windows viruses were just too much of a problem and it seems like every computer I've seen recently comes with a 3-month version of Norton Antivirus. This seems like a good idea except for the fact that the program is such a pain in the ass in its constant warnings and fairly cryptic messages about "taking over" various functions normally performed by Windows. It also seems to present a performance tax to the overall experience, especially at boot time. I know that Microsoft is going to fix this with Vista but in the meantime, it's a real drag on the whole user experience.

And speaking of performance, the X41 feels sluggish. I have a gig of RAM in it and the standard 1.5Ghz CPU. I come from the Apple world where OS 10.4's Finder combined with pokey PPC G4 chips defines a new molasses standard. And yet, the X41 Tablet still feels like it wants to do a few million digits of PI every once in a while, especially when booting any of the tablet-specific applications. Yuk.

Stay tuned for further thoughts on the whole "tablet experience." For example, the battery (for weight reasons I got the standard one) seems to drain at a pretty nice clip. I would anticipate (though I haven't tested it) that I'd get 2.5 hours on one charge without sticking on the 8-cell battery which makes the machine feel and look like it has a tumor.

I will close this initial impressions review by observing that that my impressions run against the grain of what other X41 Tablet owners are saying. Google it and you will see that most people are experiencing a high degree of product love for their X41s. To me (at this point), this seems like some sort of Microsoft Stockholm syndrome combined with a dearth of good alternatives in PC laptop-land. Or maybe Apple laptops have rotted me to the core.

Comments Permalink

Casting the Creatives

It's really happening: the creatives are taking over. There is no question about it-- whether it's the mantra of free-agent nation, the impetus behind the Long Tail, or the success of sites that distribute indie music, it's pretty clear that the tools, the network, and the inspiration is there for stuff to be created by the people instead of for the people.

Amen to that. However, it would seem that our economic institutions (by which I mean mainly contracts, both between investors and corporate entities, and more importantly between employers and employees) are woefully out of date.

This is an obvious thing to all of my trend-spotting friends who are currently struggling with trying to understand how this massive change is going to ripple through the economy. Just this past week, I was having breakfast with Kellan and talking about models for patronage applied to the modern age. Specifically, he's got a cool idea for a web services that a lot of companies should be willing to invest in. Since he doesn't want to start his own company at the moment (a well understood model for harnessing economic energy), it would be great if there was a way that he could be "sponsored" by some entity to get his idea to see the light of day. Unfortunately, the economic structures just don't exist for that today. In software, the closest I've seen is corporate sponsorship of open source projects; however, this doesn't always work and to boot, a software license is increasingly less relevant to some of the cooler ideas I come across these days.

Someone needs to work this out and pronto. Here's why: the folks that want to participate under this new model are some of the most creative high-energy people that I know. Furthermore, I would love to find a way to get them involved in my project, as would most of the other funded entrepreneurs I know. And yet professional investors (by which I mean VCs) raise valid objections to the "contractor" model of building stuff (which is incidentally the only structure in place today for this type of work arrangement). As R0ml has pointed out, software is just as much about the people who build it as it is about the source files so it's understandable that the folks underwriting the projects would been uneasy about the talent just walking out the door at the end of the contract.

It would seem that Hollywood has figured this out in the way that the studio model works to bring creative folks together for projects with a finite length to them and clear objectives. A musician friend of mine recently suggested I take a look at Creative Industries: Contracts Between Art and Commerce, a recent book by Richard Caves, an economics professor we had in college who was insightful on industrial organization, but dry as sawdust. Caves apparently got a mid-tenure crisis and decided to look at how creative industries organize themselves for kicks. I'm hoping that the book educates me a little bit in how folks in Hollywood have managed to adapt the patronage model to the modern age.

We'll figure this out as an economy. In the meantime though, I sure am glad that these are the questions we get to struggle with. Imagine thinking about "how to trade-off capital and labor" or the 'union factor" instead.

It's a good time to be alive.

Comments Permalink

Wither Java

Eddie and I (and some soon-to-be-mentioned others) have been working on a project that requires a number of key pieces of infrastructure as well as some high-level components and we've had some spirited discussions about programming languages (somewhat inspired by this post). Recently he's been working in C while I've been working in Python and I find that it's harder to get into the C-groove when reading his code after working in Python than say reading a more modern language like Java. And he's finding that chasing memory leaks also takes some mental re-tooling. And yet, the approach still feels more "right" than this 100% Java stuff which has become so popular on the web these days.

Today he revisited the issue of safety and freedom languages with me and we got into a discussion of control versus flexibility that merits its own post (or maybe a comment from him here) but during our conversation I was struck by the fact that both of us kept trying to work Java into the spectrum we had arrayed the various different languages into without much success. Java seems to fall somewhere in the nebulous middle where you get little control and little flexibility with performance that is somewhere in the middle of the two.

Anyhow this got me to thinking about Java and about why it was that in the last 5 years I've gotten so disenchanted with it and I realized that despite the hundred of millions of lines of Java code, despite the amazing state-of-the-art stuff JITs can do today-- Java has gotten pushed into the nebulous middle of a spectrum that will probably make it obsolete for web development over the next few years. The two things that were most exciting about it are now almost completely irrelevant:

(Just to be clear, the following applies only to web applications. PC fat clients, phones, and WARP engines may be a different story)

Portability: Back in the day of #ifdef UNIX hell, Java was a breath of fresh air. However, Linux won the Unix wars. Peter Yared, and ex-Sun guy who got the LAMP religion and started a company, put it best in an interview when he was asked about platform independence:

You know you’re running on Linux ‘x86, you don’t need platform portability at the language level anymore. We all spent a long time in Java to make sure developers were no longer locked into an operating system or chipset. That was what made Java successful. When it took off in 96-97-98 when it moved from the browser to the server, you weren’t locked in to Solaris or HP-UX or AIX or whatever. Linux does the same thing. So, if you know what operating system you are running on, in this case Linux, do you need layer in between? We think the answer is no.
Alternatives: C++ does really suck and Java really did a great job of making programming fun again. Plus, 10 years ago Perl was the only credible alternative to Java (which is why I think it did so well). Today, you've got Python and Ruby and a number of credible apps that show that it's possible to work in these languages. Sure the JITs aren't there and the dev tools are not super rich but the runtimes work well and the languages are nice and clean. And most importantly, these languages/runtimes/communities play well with all of the existing web infrastructure so people feel little need to he compared Java to Cobol and called it an "evolutionary dead end." And the flamewars that the RoR folks have started over and over again with the Java people have now gotten so old, I wish there was a wired-tired list for geeks to post them to. I like Java, and as I said above, it's really much easier (at least pre-1.5) to read that straight C. My only real point here is that for what we're doing, it's funny to see it occupying a dead man's area on the spectrum of today's computer language trade-offs.

Update: A Comments Permalink