Re: CU/ECU staying current

Jonathan Day (Jonathan.Day@umist.ac.uk)
Wed, 12 Jun 1996 17:36:04 +0000


Dennis J. Streveler wrote:
> Gosh, as much as I have trashed Cornell, and White Pine, on occasion, I must
> take firm disagreement with your approach. Of course, one _could_ still use
> WordStar for word processing and VisiCalc as your spreadsheet (as some
> actually still do) but with this approach you seriously degrade your ability
> to exploit today's technologies.
>
> I often write about falling "into a pit" with an attitude such as yours,
> just to wake up a few years down the road and find that much of the
> technological world has passed you by.

If you want to hammer a nail into a wall, you'd use a hammer, not a Pentium
processor. You use a tool for the job for which it was designed - that, after
all, was the point of getting the tool.

There are times where you /want/ the ability to transmit in B/W and /only/ in
B/W. Having a program that can do that, whistle the theme tunes to thirty TV
programs, scramble eggs and tap-dance might be more "up to date", but if that
leads to reduced stability and/or increased processor and/or memory overheads,
then it is utterly useless.

I'll give an example. One of the uses of videoconferencing that is being
proposed here is for carrying out experiments remotely. Now, here is a
situation where /rapid/ high-contrast transmission is an absolute and where
colour is of no value at all. Now, I'm not prepared to go along to those
members of staff and tell them that an all-singing, all-dancing package is
"better" than a free one that will do an identical job.

> In my opinion, if you want to play in this technological world of ours, you
> must stay current to the extent possible. That means using current processor
> chips, current operating systems, and, yes, current versions of applications.

This must be why the UK nuclear industry uses TRS-80's and Commodore PETs. :)

<snip>
> But if you want companies to develop software for you, a few bucks is a
> small price to pay for their use. If everybody had your attitude, the
> software industry would shrivel and die. There would be no incentive to
> offer new features and new capabilities.

Where those new features are important, then yes I'll agree with you, people
shouldn't go for the cheapest option just because it is the cheapest. You
should get something that does the task at hand, but within the constraints
that exist. It is that last part that's important. New capabilities DO NOT
mean better programs, nor vice versa. The two are related, but that
relationship is very complex indeed.

Imagine you had a computer with 16 megs of RAM and a 100 MHz processor. Along
comes Macrospot with a new OS that requires a minimum of 32 megs and a 200
MHz. processor to run, the only difference betwen that OS and the one you
are using is that it supports a 16384x16384 resolution monitor in truecolour.
Do you upgrade? Well, if you have the high-res monitor, you should. If you
/might/ get the high-res monitor, you should anyway. If you're using the
machine as a file-server and don't use a monitor at all, then what's the
point?

> In our case, and the reason I am active here on this LIST, is that CU
> presented a whole new opportunity for glimsing at the world. Prior to it,
> videoconferencing systems cost many THOUSANDS of dollars. Now it costs $60!
> With that kind of change in economics be sure that this technology will
> revolutionize many facets of our lives down the road. Don't you want to
> remain a part of that?

The UK Govt. cut the funding of UK Universities by 1/3, last year and look
likely to freeze it indefinitely. Getting a toothpick would probably require
approval, right now. As for the net here, well, forget it. There's a belief
that the packets are delivered by carrier pidgeon. Broadcasting in B/W is
just about possible, but full colour??? There's a possibility that distance
learning courses might be run, with 2-way communication to Africa. That's an
ideal situation for videoconferencing, but the underlying hardware to
support anything more than the absolute minimum just doesn't exist. Now,
bearing all that in mind, who is going to spend money on features they simply
can't use?