🚀 go-pugleaf

In article <1992Apr15.033850.16536@uwm.edu> markh@csd4.csd.uwm.edu (Mark William Hopkins) writes:
>If it's being run on a DOS, the problem is likely the DOS.  System calls are
>usually critical regions in DOS.  If two processes happen to be concurrently
>accessing the same critical region, they could destroy the integrity of the DOS.
>
>That might be the problem.

In article <1992Apr15.235445.9454@sequent.com> furballs@sequent.com writes:
>1) You can call South Mountain Software (201) 762-6965 and order
>Hold Everything, which is a TSR library to spawn context switching
>on the fly under TSR's, or
>
>2) You can build into your TSR the ability to save the sate of the
>machine and DOs itself and handle the context switch, plus all the
>associated house keeping needed to juggle multiple threads.
>
>Having done both, I can recommend #1 as the greatest time saver and
>least stress inducer. These guys have already done the home work.
>Don't reinvent the wheel.

Isn't "don't reinvent the wheel" where you ask last semester's student to take
your exam for you because "they've already done it before"?

I'd recommend #2.  Doing #1 totally defeats the purpose, which is to get
experience in this and become an expert.  There's no substitute for learning
or thinking through a problem and this one provides a substantial learning
experience which will so significantly expand a programmer's personal
repertoire as to suddenly make an entire field of programming accessible.
All competent programmers need to master this (process control, synchronization,
multitasking) or something equivalent and do something like this at least once.
There's other items that need to be done at least once: 1 Editor, 1 Compiler or
Assembler, 1 Database program, 1 Real-time/Control program, and several other
areas.

Time is cheap.  Knowledge, experience, and expertise is the real currency.
It's an excellent investment.

On a diversion...

There's also an interesting phenomenon that not too many people are aware of:
when you know enough about enough in a diverse enough range of areas, a
threshold is crossed where everything else becomes dramatically more
accessible and transparent.  At that point all of a sudden everything you know
in one field applies as critical problem-solving insight in every other field
you know.  It doesn't even matter anymore whether you know something or not
because the learning curve on learning itself will have been rounded.  That's
the point from where "don't reinvent the wheel" starts to look like a joke.

Maybe not everyone wants to build up to that threshold, but it's definitely
not a waste of time.  In purely selfish terms, it will lead to substantial
profitability in a world like today's.  And today's world badly needs more
people like this.