Chris,
Last year a neighbor who works at a small local business asked me to
estimate replacing their old system. I went in, and the PC was ~20
years old, running Win95. Very dusty, dot matrix printer to go with
it, the whole nine yards. More data than would fit on a floppy (and no
zip software available). I took in my old PKZip floppies, but the
drive wouldn't read them. It *would* read the driver disk for an old
IOMega zip drive (100MB), which I hooked up to the parallel port (no
USB of course). Data saved. The old system was written in Paradox for
DOS.
I had an old CoCo3 for a long time, until my wife convinced me I would
never use it again. :) The first computer I ever worked on was this:
http://oldcomputers.net/hp85.html
I also did some "programming" on a Tandy T100 to run the laser light
show at the Knott's Berry Farm Good Time Theatre.
To the question it raises, hardware, OS, runtime platform; all these
things can make it hard, or impossible to keep things running, or use
them for reference. In the above Paradox case, I was able to use
DOSBox to run it. Aside from sentiment and nostalgia, businesses need
a reason to change. That reason may be risk (old systems dying),
efficiency, user happiness, competitive advantage, etc. The same
applies to new systems as well as old. How much and how fast should we
change?
I don't like churn, and I've seen so much money and time wasted
building systems that I might be considered a bit of a Luddite myself.
I have an old phone. Not a smart phone, or a fancy phone, or a new
phone. I use it as a phone, not a GPS, not a browser, not an email
client or mobile computer.
Like the people in the story, my technology works for me. You
mentioned that your main client is still heavy into IE7. We lament
this, but how much change is good, and how much is just change? In the
new era of apps and the cloud, we're coming full circle (sort of :).
Is there a sweet spot? I wish I knew.
-- Gregg