[0x12]

Technology is a means to an end. Once you can reach the end with the tech that you have there is no need to upgrade any further.

Upgrading is an investment in both time and funds, both can be spent better if you are already able to do what you want to do.

Most of us could get through the day just fine on a 486/33 if we had to. In fact, we'd probably write better software if we did.

[barrkel]

You could get by fine on such a machine, so long as you were writing software for about 1993. But writing software for 2011? My source tree would just about fit on a disk that could be connected to such a machine, but it would take eons of page swapping to compile with 8MB of RAM.

Software does a lot more stuff these days, with richer media, more layers of abstraction and more reuse of code. We mock Iris as a poor knockoff of Siri, but the thing has text-to-speech and speech-to-text linked in to it, and was put together in 8 hours! Try doing that on your 486/33, with the tools of the era. You'd be lucky to have a sound card (MPC was a standard to encourage OEMs of the time to include them, along with a CD drive, if you recall) and a color console I/O library, much less a network connection good enough for cloud speech recognition.

My point? People today are forever bitching about how little we've progressed, how software development is still the same, but I can't agree. I remember what software development used to be like. It was great in that everything was small enough to be fully understandable, but that's all that was great about it - otherwise it sucked, deeply.

[Blunt]

150% agree. I have a Pentium III class machine we still use to run our company email. It's been rock solid for so long, there has never been a need to upgrade it. Should the hardware finally fail, we have a few spares kicking around on stand-by and a build script that builds the software and an rsynced version of the mailbox store (cyrus IMAP). I guess, just another data point to your point.

[cstross]

Comments like this make me wish I could occasionally grant more than a +1 upvote ...

It's so true. Sometimes we need more power to make a given job possible, but as often as not we could accomplish it with far less and all that throwing extra hardware at a software problem achieves is to allow sloppy coding to proliferate.

[adrianN]

Generally it's a pretty good deal to trade hardware requirements for programmer productivity.

[InclinedPlane]

I assume this means you're using a 486/33 right now, since it helps you write better software.

[icandoitbetter]

>Upgrading is an investment in both time and funds, both can be spent better if you are already able to do what you want to do.

You are operating under the assumption that the usual way of dealing with technology is that you have some goal, and then you search for the right tool to achieve that goal. I disagree with that assumption. Most of the times you don't know that you want a specific functionality until the technology comes that implements it. Then you realize that you wanted it 'all along.' You are shaped by your environment more than you shape it.

Your time is indeed better spent doing things other than upgrading, the reason being that the software business just hasn't been able to keep up with the advances in hardware. Even though we have so much computational power available to us, we still haven't learned how to handle complexity and make more interesting, flexible interfaces. There can be innovation in document publishing software, even though we haven't seen any for such a long time. And when that innovation comes, I'm sure you will start wanting features from your word processor that you couldn't even imagine were possible.