From: "Sellam Ismail"
<foo(a)siconic.com>
On Fri, 22 Nov 2002, ben franchuk wrote:
I did not say General purpose computing can't
be done, just that it is a
very small amount of memory for most user programs. It is really hard
work to have a useful programs written on the small 8 bit machines, and
fit in 32k or 48k of memory incuding the OS.
I find it rather odd that such a comment would be made in the face of tens
of thousands of applications software and games that were developed on
dozens of 8-bit computer platforms having anywhere from 4K to 64K of main
RAM where the operating system and application shared that memory space.
?
---snip---
Hi
The only thing I miss on a 8 bit machine is the lack
of space for large data sets. I've never ran out of code
space in my 15 year ( when I was doing embedded stuff ).
Then I could never understand what happened to all the
space that most of today's compilers use. I've always used
a combination of assembly and Forth for embedded stuff.
I had to work with a few applications written by others
that were in C and found it very difficult to stay
within RAM limits for even simple applications. When I first
saw languages like SmalTalk, looking at the description, I
thought that finally someone was doing something right.
They talked about the efficiency of reuse and such ( concepts
that are natural to Forth ). Then I tried it and found that
it was more bloated than the C programs I was dealing with.
The other thing that bothers me is that today we have
computers that can run several thousands of times faster
than an 8080 IMSAI. Still, it takes 5 to 10 minutes to
boot that machine. My IMSAI boots faster than I can reach
from the reset button to the keyboard. I'm told
that is
is because the machine has to figure out what the I/O
looks like on each boot ( of course, MS rejected OpenBoot
for PCI boards that could have improved the boot time ).
I find that applications written for these new and more
powerful computers that have been doubling in capabilities
every year are becoming garbaged up at a rate that is
faster than the computer get better.
I doubt there is anyone that works for MS that even knows
if 50% of the software that they release is even being
usefully run under any condition ( maybe even 80% ).
I believe that there complexity will eventually bring
them to the point that even hiring the entire graduating
class of computer science students will not fix.
Just my ramblings
Dwight