Path: utzoo!utgpu!jarvis.csri.toronto.edu!rutgers!usc!snorkelwacker!bloom-beacon!eru!luth!sunic!mcsun!hp4nl!kunivv1!ge
From: ge@kunivv1.sci.kun.nl (Ge' Weijers)
Newsgroups: comp.sys.handhelds
Subject: Re: HP-48 Comparison Test
Keywords: HP-48SX, calculators, portable computers, handhelds, HP
Message-ID: <1172@kunivv1.sci.kun.nl>
Date: 13 Mar 90 09:37:38 GMT
References: <39230@apple.Apple.COM> <326@cbnewsb.ATT.COM>
Distribution: comp.sys.handhelds,comp.sys.hp
Organization: University of Nijmegen, The Netherlands.
Lines: 50

neal@druhi.ATT.COM (Neal D. McBurnett) writes:

>In article <39230@apple.Apple.COM>, dan@Apple.COM (Dan Allen) writes:

>Re: calendar
>> How does the 48SX fare?  Well, not too bad.  It has date arithmetic; it has 
>> time arithmetic, and HMS conversions.  It has about an 8400 year calendar.

>I assume you are referring to the fact that dates between 15 Oct 1582 and
>31 Dec 9999 are valid.  I think it is a shame that they didn't put in code
>to handle the Julian calendar which was used before that fateful date
>of Gregorian calendar reform - it's not hard!  It would also be
>useful to have accurate historical dates back before the year 1, although
>this would require a modification to the numerical representation.

This is tricky. The Julian calendar was abolished in 1582 in only a part
of the area using it at that time. In Russia is was abolished after the 
revolution (the februari revolution happened in march).

>As a sidelight, the calculator says that the year 4000 will be a leap year,
>but I think that it should not be (assuming that our civilization still
>exists and hasn't yet adopted a more sane calendar, changed the orbit
>of the earth, etc.).

As far as I know 4000 is a leap year the number being divisible by 4.

>What was the "EXACT" feature of the 75C/71B?

I suspect it is something similar to a function on the HP41 time module.
You can correct/calibrate the module by resetting the time after a preferably
long interval. The module calculates its own drift, and compensates for it.

>I also look forward to the gurus figuring out how to change the time
>rate of the "date" and "time" functions.  So far mine seems to lose
>about a half second a day.  I may have to revert to the hand-crafted
>routines (based on the internal 8192-tick-per-second clock) I use on my
>hp28S to get an accurate time (e.g. for timing occultations of stars).

As a work-around I'd suggest running a program automatically in the wee hours
of night that corrects the drift by adding 1/2 a second to the time. If there
is no function available to do this in one step do a garbage collection first
(MEM on the 28) otherwise the drift will become much larger once in a blue moon.

Ge' Weijers


Ge' Weijers                                    Internet/UUCP: ge@cs.kun.nl
Faculty of Mathematics and Computer Science,   (uunet.uu.net!cs.kun.nl!ge)
University of Nijmegen, Toernooiveld 1         
6525 ED Nijmegen, the Netherlands              tel. +3180612483 (UTC-2)