Question about learning Python

[Dennis Lee Bieber]

On Thu, 8 Sep 2022 07:42:19 +1200, dn <PythonList at DancesWithMice.info>
declaimed the following:

>TSRs? Now that was an ugly period of history! (trying to make a
>single-process operating system do multi-processing - only to find that
>many program[me]s assumed they had full use and undisputed control of
>the computer. Happy days...)
>

	I laughed when M$ MSDOS (2?) introduced TSRs... My TRS-80 running
L(S)DOS had similar things at least a year earlier. And these were
/run-time/ loadable. They called them "filters" (and device drivers were
also an option). Key-click was one such -- though it also showed some
quirks (like... If the processor was really busy, the key-board driver
would buffer key-strokes, but the filter activated when an application
/read/ the key-board). Filter to control printer formatting, a JobLog
filter, Key-Stroke Multiply filter (I never used it, but it apparently uses
a table of special keys and expands them to longer strings). Commands to
load device drivers (or remove them!). Could even change the number of
cylinders for a floppy drive -- My drives were "loose" enough to allow my
to add 2 cylinders.

>On the other hand, one can start too 'high' or too 'modern'. Like the
>person enthusing about MSFT's and AWS' programming AIs, thinking that
>such tools will replace programmers (one of the aims of the COBOL
>language back in the 1960s). His short-form description spoke volumes:
>'it saves anyone from having to look-up Stack Overflow any more' - a
>'blind' cut-and-paste prospect that saves the 'author' from the
>difficulties of 'learning stuff'; until it is time to, um, learn-stuff -
>to know why one needs to learn-stuff BEFORE taking from SO/AI.

	I once worked with someone whose idea of programming was to find
examples of working code, and cut&paste snippets to make an application. We
needed to control four devices via GPIB... She wrote four short programs
and a DCL script to run them in sequence.

	Problem: each program had to issue a GPIB initialization command before
it could continue to talk to any device. But each initialization command
would UNDO any configuration the previous program had set up! We discovered
THAT when TDY to the remote site (and she'd already returned home). I had
to debug the situation, and rewrite the four programs into a single
consolidate program over a weekend (and I'd never worked with GPIB before
this -- my task was the post processing of the data that was collected
after the GPIB chain had been set up for data collection).

	I think I'm not giving away any secrets these days, but this was a
Quick Response Contract for a proof of concept -- that a geo-bird could
catch over the horizon GPS signals, and thereby improve the satellite
tracking data for ephemeris generation. Had to be over-the-horizon as
geo-birds are above GPS, and GPS aims signals down to earth. We had to
catch side-lobes as GPS birds were high enough to avoid
ionosphere/troposphere effects.



-- 
	Wulfraed                 Dennis Lee Bieber         AF6VN
	wlfraed at ix.netcom.com    http://wlfraed.microdiversity.freeddns.org/