denbeste@bbn.COM (Steven Den Beste) (12/25/87)
The original poster of this subject asked if there was a place where
multi-tasking was "essential". Many people responded with ways to use
multi-tasking, and the original poster always responded with one of two
answers:
1. There are other ways to do that that don't involve multi-tasking.
Multi-tasking therefore isn't essential for that task.
2. Only sophisticated users would want to do that. (He called them "Power
users.)
[Well, actually, there was a third answer used a couple of times:
3. Why are you being so abusive in a publicly posted article? Why didn't
you mail it to me so I could ignore it?
But that's not relevant to the subject...]
Let's get one thing straight. Multi-tasking is not "essential".
Neither are high-level languages. You can do anything in assembly language you
can do in C or BASIC, but it is less convenient. (MUCH less convenient.)
Nor are icons. People survived for years with text-only computer interfaces.
[For that matter, assemblers aren't "essential" either - you can key in machine
language directly in hex. And if you really want to get down to it, the
computer itself isn't "essential".]
I'm not suggesting that multi-tasking is as important as high level languages.
That's not the point. I _am_ saying that requiring a feature to be "essential"
to be included is a ridiculous criterion. The proper criteria are "useful"
and "convenient".
On the Mac, there are many programs which can run simultaneously. To do this,
however, they must be written in a very special way and the operating system
must be specifically informed of each.
This is certainly an approach, and it seems to work for the Mac owners.
There are certain special cases on the ST, I am sure, where more than one thing
can happen at the same time (like a print spooler, for instance) - again it is
done specially.
On a general purpose multi-tasker, ANY two programs can run simultaneously,
unless one of them is being a particularly bad citizen. Though this isn't
"essential", isn't that more convenient?
Which brings us to the second argument: Even if multi-tasking is indeed useful
and convenient, it still won't be used by the great unwashed (called typical
users in the original posting). Only "power users" would use it.
I don't quite understand why that means a feature should be omitted. Is it
being suggested that a feature should only be included in a product if EVERY
OWNER of that product will use it?
More importantly, it assumes that the users won't mature and become more
sophisticated. Though a user may not use a feature when the product is first
purchased, many times the user will grow into it as time progresses. "Power
users" all began as unsophisticated users.
However, they won't mature if the machine they are on won't let them. I used to
know quite a few owners of TRS Model I's, who were wizard Basic programmers,
but were so restricted otherwise by the machine, that they might now never
grow out of it, even if now given a better system.
In many ways, it is desirable that a machine provide some advanced features
that the new user won't immediately use, so as to provide room for maturation
and learning...
...providing it is done so that the user isn't forced to use this feature
before ready.
Which leads me to the last point: The original poster stated, but never
justified, a belief that providing multi-tasking was not merely useless but
actually undesirable. I don't see it.
If:
1. Multi-tasking doesn't add to the purchase price of the product
2. It isn't intrusive on people who don't understand or need it right now
3. It is available for those who do understand it, or those who grow into
it
Then
I state that it is an unmixed asset and should be in the product. I would
be very interested in hearing why, given these three assumptions,
multi-tasking should be omitted from a product.
If you change any of the assumptions, then you are arguing against a
strawman. All three of them are true for multi-tasking on the Amiga.
If you say "There are other ways" that is irrelevant. The point is not to
show alternatives, but to actually demonstrate that multi-tasking is a negative
thing in its own right.
If you cannot demonstrate that it is an actual danger, then it is merely
useless for most (but not all!) users. But since, by the assumptions, it
doesn't impede those users who don't use it, then its ability to attract "power
users" makes it an asset. (It attracted me - it is the main reason I bought the
Amiga.) (By the way, by so doing it represents an indirect asset for those
unsophisticated users, because the "power users" are cranking out public domain
software which the unsophisticated users can get for free. 120 880K Fish-disks
at last count.)
We therefore have three classes of users for the multi-tasking machine:
1. Those who don't understand and won't use multi-tasking, but are not impeded
by it. [If you think they are impeded, then we are out of the realm of the
theory and into fact. I state as a fact that it does not impede rank amateurs
on the Amiga.]
2. Those who don't understand it to begin with, but grow into it.
3. Those who understand it and use it from the very beginning.
Group 1 isn't bothered by its presence. Group 2 and Group 3 benefit by it. How
can it be a bad thing?
--------------------------------
Which brings us finally to the issue of the motive of the original poster.
Let's try some syllogisms, shall we?
1. The Atari ST is an ideal computer
2. The Atari ST doesn't have multi-tasking
ERGO Ideal computers don't require multi-tasking.
1. If several machines offer multi-tasking, people will grow to expect it.
2. If people grow to expect it, they will refuse to buy machines not having
it.
ERGO If several machines offer multi-tasking, the public will refuse to
buy machines not having it.
1. If several machines offer multi-tasking, the public will refuse to
buy machines not having multi-tasking.
2. The Atari ST doesn't have multi-tasking.
ERGO If several machines offer multi-tasking, Atari ST sales will suffer.
1. Anything which causes Atari ST sales to suffer is dangerous.
2. If several machines offer multi-tasking, Atari ST sales will suffer.
ERGO It is dangerous for other machines to offer multi-tasking.
Could these be the real reasons behind the original posting? Or maybe just a
deliberate attempt to stir up trouble?
--
Steven C. Den Beste, Bolt Beranek & Newman, Cambridge MA
denbeste@bbn.com(ARPA/CSNET/UUCP) harvard!bbn.com!denbeste(UUCP)
I don't think BBN cares what I think about this stuff.
And that's probably just as well.