It's all spinning wheels and self-doubt until the first pot of coffee.

Upgrades versus Antiques

From Greenmonk: The Blog - Cherish The AIR? Just because you can doesn’t mean you should:

...the core idea that should be upheld by companies like Apple should be about making things better and less often. Making things that will be able to evolve, be upgraded, be adaptable, hackable and more fun to use for longer so that as a customer I don’t think that I’m buying version 3.4 of something that will only be as good as it’s last press release. I want to buy “the” quintessential Apple product and cherish it for years, like people would cherish a vintage car.

From De-Scribed: The upgrade treadmill is wearing us out:

Replace "quintessential Apple product" with pretty much "any product" (use your imagination) and you start to get at a shift in both attitude and culture that would solve a lot of problems. Get off the upgrade treadmill. There are already too many others.

When's the last time you upgraded your watch to one with a more reliable chronometer? Bought a new pen with a better ink formulation or superior delivery mechanism? You may, of course, be an exception—but there are a lot of other personal artifacts that have reached similar plateaus for most, where the classic vintage version is every bit as useful as the modern iteration.

I keep wondering when personal computers will hit a flat spot in Moore's Law. Or, failing that, when the capabilities and usability of such a machine meet and surpass any reasonable owner's practical demands to the point where fashion and style become the selling features. I think we're almost there—actually, we're probably there now for most people who aren't me.

Beyond the constantly growing demands of gaming, it seems like most personal computers now can do most of what anyone wants them to do. Even for me, this last-gen MacBook Pro feels like the best computer I've owned to date, and strangely enough I'm hesitant to think about upgrades for awhile this time. I like the product design, performance is very good, and I've got the memory maxed out to the point where I can run several copies of Windows in Parallels—which, itself, I think is a very weird requirement, yet still satisfied by this machine.

Storage media can hold amazing amounts of personal photos and video, and will reach near-incomprehensible levels pretty soon. Screen resolution is just about at a point where most people don't care if it gets better, and the same goes for sound. Barring any sort of Moore's Law for finding new uses for personal computers that require significantly more power, demands for functionality will likely be completely met by even the lowest end laptop soon.

So, how long until the first new personal computer arrives that turns out to be the old beater I still use 15 years from now, or sell to a kid down the block who just got his web learner's permit?

Archived Comments

  • For the most part, I'm already there.

    I have several PCs around the house in daily use that are at least 8 years old. Drives may have been upgraded and RAM added here and there, but essentially they're still the same. Some run Win2K (there's something that hit a sweet spot and only got worse thereafter) while others run up-to-date Ubuntu or Debian. Whichever, they're perfectly usable for web browsing, document editing, email, and even development.

    I'm not claiming my daily workhorse laptop is quite such an antique, but it must be a couple of years old and I have no urge to upgrade it yet.

  • I'm running a Power Mac G5 that's a few years old now but, with the latest OS and software revisions, is still perfectly adequate for my computing needs. Sure, I'd love to upgrade to a top-of-the-line Mac Pro but I simply can't justify the expense when the G5 is such a great "antique".

    I also still use my Apple MessagePad 2100 every day. It's an extremely reliable and well-featured PDA and I see no need to replace it.

    Finally we have an old P4 PC running XP. It's real slow but the kids use it for homework and web browsing and it'll stay for while yet.

    So I agree, we may have reached that sweet spot already.

  • I'm just now upgrading my main desktop from a 12 year old AMD-586 (150MHz, 32M RAM, something like 17G drive space) to a dual-core P4 with a gig of RAM and 150G of drive space (I think---it's large compared to what I'm used to).

    Why haven't I upgraded before? Because the AMD system met my needs for that long. I do mostly C development, and it's been a good workhorse for all that time. And now that computers have pretty much settled down, my new machine should be good for at least a decade if not more.

  • The problem with screen resolution isn't so much about DPI as it is about size--this is why HDTV is, in its own way, the wrong idea. Users are far more liable to pick a larger display as being more high-quality than they are to pick a higher-resolution display.

    Apart from that, I think we'll see a lot more throwing away of the various hacks and optimization that let us do what we wanted on lesser hardware. Smooth video playback is a solved problem? Let's funnel low-quality video through a proprietary browser plugin that chews so much CPU that you can't watch it even in a tiny window without skipping and juddering! Mmm, Flash. You make me party like it's 1999. Of course, using something like youtube-dl and playing the video with a dedicated player works fine. Perhaps the Flash folks are part of a massive conspiracy to keep us juicing up our hardware, eh?

    I do think that some components can't be reasonably manufactured to last for fifteen years of continuous use, like power supplies and hard drives, but that's more of a corner case.

    And if you'd like to consider a field moving from a tradition of finely-manufactured heirloom objects that you could leave to your grandkids, to the upgrade treadmill of the eighteen-month semiconductor manufacturing cycle, consider photography. Yes, the marginal cost of a digital photo is nil, but instead of buying new film, you have to buy a new camera every year or two. People who work in the field have written on this.

    I have a (roughly?) five year old DSLR, and I can't tell if it feels ancient because I'm used to electronics aging like that, or because it was one of the first non-professional (i.e., affordable by mere mortals) models. If I could afford it, I'd have probably bought a new one by now.

  • Unlike pens and watches, computers ("computing infrastructure") are still expanding to fill cultural niches, still breaking out into the wave of creative destruction. Phones and computers and smart dust and sensor-laden clothes are all still vying for resources like weed species colonizing an empty continent.

    But like weed species, they have limited lifetimes. They're depending on r-selection, still. Not K-selection, which is what you want. I expect r-selection is part of the marketing culture surrounding all this tech.

    And now I think about it, I wonder if maybe it works the other way around: you can't "market" something that's long-lived, especially something you interact with communicatively and increasingly like a peer (unlike your pen or your watch).

    I think it's time for a new shade tree in the back yard. That cow is looking kindof slow compared to these new cows they have, maybe we should replace it. I was thinking of a new left hand; what do you think of yours?

    That'll be the real Andy Clark moment, by the way. And the counterpoint to the one you're waiting for.