L.D. Best wrote:
>
> Think about it ....
>
> When I ran my bookkeeping business using a Texas Instruments 16-bit
> Professional Computer with 256K memory & dual 320K FDDs, I could sit
> down, place a diskette in the A drive, turn on the computer, and before
> I could light a cigarette [which wasn't a mortal sin back then], the
> non-compiled BASIC program was waiting for me to provide input. In
> those days no one mentioned MHz ... you simply used the computer.
>
> Today, with a Pentium 233MHz boasting 128Mb SDRAM and using a Fast SCSI
> HDD with track-to-track scan times too fast to measure in any meaningful
> way, I can stop to turn the computer on as I head for the kitchen, pour
> myself a huge mug of coffee, refill the water in the coffee pot, get
> some milk to make the coffee bearable, take the coffee back to my desk
> and sit down and *wait* for the first screen of config.sys to appear on
> the monitor.
Thought of a cigarette scares the breath out of me! Staying away from tobacco
smoke is literally a matter of life and breath with me.
I remember hearing that, with computer applications keeping up with hardware
improvements, the increasing software bloat was such that newer computers were
no faster than the old computers.
from Clarence Verge:
>Ahhh. Those WERE good old days. Back before Windows. Back when the OS was
>written in 100% ASM. Back when 640k *WAS* enough. Back before a torrent of
>crap written in "C" hit the shelves. Back before some twits decided that
>the most successful and copied computer every made should be run like the
>SECOND most successful (but not copied) computer.
>
>I've been here before.
>Why do you have a 233Mhz computer with 128kb ram ?
>Because of the "C" programming language.
>Why is it STILL not fast enough ?
>Because of the "C" programming language.
C is the language for professional programmers because of all the sneaky
pitfalls. C gives the programmer plenty or rope to hang him- or herself.
Pointers are rather devious in C, easier to understand in assembler. Ada, with
strong type checking, is much less sneaky.
Now I think Visual C++ is big, along with IBM's attempt to compete with Visual
Age C++. I've never got into visual programming, even dBASE IV for DOS had it.
Visual programming simplifies putting a program together for the human
programmer, but automated generation of code would tend to be more bloated than
more direct coding. Throw more hardware at it! Faster CPUs and bigger hard
drives, and the need to get a software application ready before the competitor
does, make programmers less likely to care about efficient coding.
