[comp.sys.handhelds] Should I take the plunge?

tldavis@athena.mit.edu (Timothy L. Davis) (06/25/91)

Since my 10yr old HP41C has taken on an intermittent inability to count,
I've been ogling the 48s/48sx.  At the risk of posing a FAQ, could someone
please describe to me the pros and cons of the HP48 "Saturn" architecture?
It seems to be much better than "Synthetic Programming" on the HP41.

I have a few specific questions:

1)  How fast is it?  Both in program execution and keyboard/display I/O.
    I've heard a few distressing things about lost keystrokes, etc., and
    slow display updating.  I want to see my numbers as they are entered!
    I had previously doubled the speed of my 41 by changing a timing cap.
    Is this possible or desirable with the 48?  How fast in comparison?

2)  Is it truly the platform of the future?  I'd like to use it for
    engineering problems (I'm doing biomedical research, often in the lab
    away from a desktop computer, possibly in the clinic as well), and I want
    it to work well for me until 2001, as my HP41 worked from 1981-1991.
    If there are serious flaws, or a new handheld in the works, I'd like
    to know.  I can settle for an HP42 or whatever for awhile, until the
    next one is ready.

3)  What ROM cards are/will be available?  Is RAM at a premium?  I could
    save $75 by going with the new 48s.

4)  What's the cheapest price?  Elek-Tek has HP48sx for $249, HP48s for $175.
    (They are mail-order from Chicago.  Were reputable when I bought my last
     round of calculator equipment, 1981-84, phone (800) 395-1000.)

As an aside: 
Assuming that I get positive responses, I'll have a (slightly flakey
modified, with original parts etc.) HP41C for sale, and in addition a
like-new Card Reader, Advantage Pac, Clock/Timer Module, Quad Ram
(makes it a 41CV), Games Pac, Timer solutions book, and about a Quad
Ram full of slick integer/fraction/quadradic arithmetic, timing,
note-pad, numerical integration, and other software which I wrote
myself, including a 7-line GCF algorithm (Euler's Method).  $630 new, now
$200 o.b.o. takes the works.

Thanks, 

Tim Davis (tldavis@athena.mit.edu)