To "future proof" or not?

[FSCDiablo]

It's a term used quite often in the hardware scene. In theory it sounds like a smart decision, but in practice how does it usually work out for you?

For me personally, when I think about future proofing (or have attempted to in the past) it usually just ends up pointless. Almost invariably when the "future" finally comes around there is a better upgrade option in front of me.

Should I by xyz board with a lower speed processot with the idea of putting in a faster processor on down the road? Down the road some other new pic count is out and the newer procs aren't supported on my old board.

Should I buy a SLI board and 1 6800GT today, because it's what I can afford? Down the road will probably be a newer single card solution faster than the dual 6800's AND support new features, AND at a good price. Why did I buy SLI again?

Should I buy 2 sticks of 512MB RAM now, and the 2nd pair down the road when I might have more money? Whoops, a new chipset came out and I'll want to save for a new proc and mobo so I can't sink money on my old setup.

on, and on, and on...

How many of you followed through with your prior future proofing? For me I see it as poor planning now, and when I get asked to help people with thier hardware needs I build a machine than fits todays needs and do not get hung up on future what if's. Build the right tool for the job at hand and save your dollars today instead of spending more on maybe's.

 

[imagine]

In the tech industry, it's rare that it turns out very well since the progression is so fast and completely non-linear. If things just moved in one direction then you could just buy the biggest and the baddest thing that you could and expect it to work exceedingly well for the better part of a decade. Instead, though, you see things like operating system's developing 64-bit technology, so a 32-bit CPU, even if it is very good and very "fast," just isn't going to cut it.

So no, there isn't a very good way to 'future proof' much of anything.