(And come to think of it, the camera I'm using now has more on-board storage with its meagre little flash 128mb card than the 486 class PC with its 120mb hard drive that I was using in 1993.)
A few further anecdotes. Last year I spent almost $2000 on my photo-habit. At least 75% of that was spent on developing proofs. During the first 4 months of this year, I've only spent $500. That breaks down into $300 on a digital camera, memory and batteries and $200 having various prints made. This will definitely drop over the remaining 8 months because I won't need to buy another digital camera and memory and I've figured out how to get gift quality1 digital photo-prints out of my inkjet. My big win; I don't need to spend another nickel on my habit - save what the inkjet consumables are going to cost. Free? Of course not, but at least this iteration of my habit isn't going to cost me another $2000 this year.
The Personal Computer is finally becoming multi-purpose useable. People are finally capable of doing *new* things with their computers that were previously impossible, or difficult, to do at home. Developing photographs is a great example. The investment is roughly the same - a well-equipped darkroom will never be cheap, but digital is much more efficient and useable - and affordable. But everyone knows this, I'm not saying anything drastically new.
There is something else that this draws into relief. Much more subtle, but quite dramatic in its impact. Devices are getting smarter and becoming much more discrete - and amazingly useful. Take a look at the new generation of photoprinters for example. Completely self-contained units that don't even need to be connected to a computer to be useful. Just feed them data and *bam* they spit out amazingly high quality prints.
I've been pushing the following thought for a while now in meatspace - let's see what the internet thinks of it. Its not really mine - think of it as a synthesis of many opinions...
The popularity of the home computer is simply a direct result of a limited utility backlash against single-purpose, analog-focused units - the typewriter, VCR, "hi-fi", microwave and the rest. Each highly useful in their own way, but ultimately limited in function. A typewriter is just as bad at facilitating complex calculations as a VCR is at assisting real time-shifting. Both can do it, but it never ends up being a wholly satisfactory experience. Manufacturers that want to capture the hearts and minds of the home will need to fix this. I'd even go so far as to venture that their future success is tied to the emergence of the digital equivalencies of these formerly unconnected, non-calculating units. The good news is that we're starting to see them all ready; IP enabled, hard drive based home audio units, Tivo's, MP3 players, digital cameras, home security systems, smarter cars. Each of these, in their own way, allow us to move beyond our reliance on the centralized home computer paradigm and into a realm of higher function computing that really makes sense in our day-to-day lives.
This all means that we're in for a lot of change over the next few years. Not evolution, but outright change. First up, expect a showdown between Little Iron and Big Iron. The multi-purpose desktop computer needs a massive amounts of storage and massive amounts of MIPS to live up to its destiny as the ultimate convergence of all home based data processing devices - the hi-fi, the typewriter, the daytimer, the calendar on the refrigerator, the rangefinder in the closet and so on. Bringing brains to these home based data processing devices doesn't require the same level of technology. Its much cheaper and low-tech to give a camera digital processing capabilities - storage, calculation and the like (and while we're at it, why not give it an IP backplane so that it can talk to other newly smart devices.) Using portable Little Iron to replace portable Analog Stuff seems to be a lot closer to what customers are expecting - despite what the current market leaders have been pushing.
I don't need an Onyx 3000 in my garage to take care of the data processing requirements of my family. Sounds like an obvious statement to make, but this is where the market leaders want us to go. Big Iron heating the garage and dumb little poorly connected units that barely network the homestead together2. There will be no information superhighway, let it go - its over. Its dead Jim.
What I need are much smart single function devices that can talk to one another and enable the discrete but free flow of relevant information between my devices. I need a way to create and store a set of mail filters in such a way that I can access them from any mail client that I'm signed on to. I need a way to search my image library, from any computer in the world, by a series of keywords that bootstraps a slideshow of relevant images. I need a way to listen to my music collection whereever I am using whatever digital music enabled device that I'm closest too (including my car and those cheap airplane headphones.)
I'm betting on Little Iron, but mainly because it offers what I need and what provides me with a more exciting view of tomorrow's network. If Little Iron wins, device identity, ad hoc device interaction and "find-me/follow-me" services become critical aspects of the infrastructure. The technical emphasis will shift from managing human-device interactions and towards device-device interaction. Interface design will become more about physical ergnomics than about overcoming KMM input limitations.
What does that leave us with? A lot of internetting between devices. Human interaction with network resources will almost certainly take a back seat to device transactions. In a lot of respects, the network will melt into the background - "there" will be "here" and everything will be at your fingertips.
We might see Mark Anderson' s AORTA before we thought.
1A "gift quality print" is a print from an inkjet made on photo-quality paper that, when put into a reasonably priced frame, is of sufficient quality to give as a gift. I made the benchmark and the definition up. Regardless of its origin, it is a reasonable measure of the current state of the art in home photographic technology. "Megapixels" and "dots per inch" lack the context necessary to be relevant to the average user.
2USB and Firewire are about the absolutely worst thing to happen to home network since...well, ever. IP is cheap and configurable and gives users a lot more options down the road. We don't need a new networking standard. Make it all go away. It is all too complicated and doesn't work all that great compared to an intelligent device setting at the end of an ethernet or local area wi-fi connection.