A429
 
can sombody explain to me how I compile a library to an absolute
address (for speed reasons) using the built-in CRLIB or Debug2 ?
I'm still trying to write that program with the gamma function, however I  need your help once more...
I want to execute the REPEAT clause while j is less than or equal to n, 
hence I have typed   ÇWHILE j n £ REPEAT 
Çrepeat codeÈ ENDÈ
The problem is that when I execute the program, it never gets to the repeat 
clause, I get the following error message: £ Error: Bad Argument 

Type
The stack looks like this:
3: 1
2:  'j'
1:  2 (This is the value of the n variable)
Alberto Reig P216rez
PS I'm using rom 1.18
==
When you use the INPUT function, your input is put on the stack as a 
Unless you do this conversion, you are comparing a number to a string, 
which produces an error.
==
Here is the entire program code:
Ç N¡ de integrales?
 INPUT OBJ? ? n
  ÇDefinicion del recinto
{ { a: Introduce el coeficiente a}
{ b: Introduce el coeficiente b}
{ ?: Introduce el coeficiente ?}
{ [Section]: Introduce el coeficiente §} }
{ } { } { } INFORMÈ
DROP OBJ? DROP ? a b ? §
  Ç 1. ? j
    ÇWHILE j n ?
      REPEAT
        ÇParametros de la integral
{ { p: Potencia de x + 1 }
{ q: Potencia de y + 1 }}
{ } { } { } INFORMÈ
DROP OBJ? DROP ? p q
        Ç a p ^ b q ^
      END
    È
  È
È
The problem now is that after the first INFORM is executed, I get the
DROP commands, the first one to get rid of the 1 that INFORM puts on level
puts in level 1.
message, as expected.  If instead I delete the second DROP, it gives me the
same message, but this time OBJ has been executed, because there is no list
in level 2, instead only the first element of the 4-element list remains on
level 1.
AFTER both it and the ? a b ? § part have been executed, but that's 

what
seems to be happening, since the second INFORM is not executed, hence it
Once again, I would really appreciate any help I can get on this, I am also
posting the program itself as an attachment, in case any of you wants to 
try
it for him/herself.
==
Hi again, sorry about the editing mess:  This is what the code should look 

like:
ÇN¡ de integrales?
INPUT OBJ? ? n
  ÇDefinicion del recinto
{ { a: Introduce el coeficiente a}
{ b: Introduce el coeficiente b}
{ ?: Introduce el coeficiente ?}
{ [Section]: Introduce el coeficiente §} }
{ } { } { } INFORMÈ
DROP OBJ? DROP ? a b ? §
  Ç 1. ? j
    Ç WHILE j n ?
      REPEAT
        Ç Parametros de la integral
{ { p: Potencia de x + 1 }
{ q: Potencia de y + 1 }}
{ } { } { } INFORM
        È DROP OBJ? DROP ? p q
        Ç a p ^ b q ^
      END
    È
  È
È
==
the latest (beta) version 1.19-5 from HP web site, but it didn't help. When
convert from.When I press arrow down button (the edit mode) the needed
number appears on the screen. But as soon as I exit from that mode, on the
stac # 0h appears - what's wrong. Maybe you (the guys who write the ROM)
havent notice it. Or maybe it something I did. Tell me, couse I need it, 
and
I can't belive that such powerful machine has a problem with converting
numbers.
Niedzwiedz
You are in HEX mode. To convert to binary integers to decimal you simply 
put
the calculator in DEC mode. Entering the command DEC will convert the 
display
to 
show all binary integers in decimal format.
We typically enter real numbers in decimal form. The HP does not natively
convert real numbers from hex to decimal and vice versa, we must convert it 

to
this command (or press the button) the real number is converted to a binary
integer, but displayed in the mode you previously select. 
Look at the top line of your display. The first annunciator indicates RAD  

DEG
or GRD and indicates whether angles are displayed in radian, degree, or 
gradian
mode. The next one will say XYZ R<Z or R<< (< = angle symbol) to indicate 

how
complex numbers are displayed: as rectangular, polar, or spherical 
coordinates.
Then comes your binary integer mode, HEX DEC OCT BIN.
If you have soft menus selected (-117 SF) you can access all these from the
soft menu easily by pressing RS 3 (BASE)
==
which the calculator can _display_ in binary, octal, decimal, or 
hexadecimal
notation.  Note that even though a Binary integer can be displayed in 

four
different notations, the internal number doesn't change.
==
that's what I observed already in the 48, though not to such
an extend as on the 49. Complex software seems always a big 
step ahead of its accompagning documentation. 
This is very bad for complex systems like the HP-calculators.
It seems to be a kind of a law that it is more easy for a 
routined programmer to progam than to document by himself 
what his program really does :-)
The problem seems also to be a linguistic. Do native English 
speaking programmers realize the advantage they posses for 
writing clear and comprehensive documention? They save the time
we not-native English programmers have to spend for learning
foreign languages. And nevertheless, their documentation is 
sometimes worse than the English writing of a Chinese, say.  
==
you have your binary word size set to 0. (Or is it 1?) The only binary
number that can be *shown* in this wordsize is #0 (Internally the
binary numbers are OK, that's why you can see them when you edit
them).
Type:
64 STWS
This sets the wordsize to 64 digits.
==
discussion of a version 1.19-5+ made available by Bernard Parisse.
Much searching around on the mentioned URLs has turned up nothing
except the 1.19-5 version I downloaded ages ago from hpcalc.org (RIP).
Recently I have notice some posters claiming to be usin 1.19-6.
So, to all you gurus out there, what is the definitive answer?
Also, to Bernard P: What is the correct URL to get at this updated
version of 1.19-5 you have been credited with (CAS updates?).
==
I forgot to put the URL of the page:
http://www.geocities.com/CapeCanaveral/Galaxy/9055/49rom.html
and the spanish version:
http://www.geocities.com/CapeCanaveral/Galaxy/9055/rom49e.html
==
however, the calc *would* both translate special characters
and compile the program if you downloaded using *Kermit* instead.
There is nothing special about the Kermit protocol itself,
but the Kermit *commands* in the HP 48/49 contain additional
embedded layers for translating and [de]compiling user programs
and objects; the use of these extra layers of processing is called
ascii transfer mode, usable for only for user programs & objects,
which can be translated into plain text for people to read;
otherwise the transfer is a binary, byte-for-byte image
of the object from internal calculator memory,
which is generally unreadable by people, much as binary PC programs
are generally equally unreadable by people.
Welcome to comp.sys.hp48, and remember,
when you see this same question posted by someone else,
it's your turn to answer it next time :)
==
Hi. I try to use bode 2.31 on my HP48. When the transfer function has poles
or zeros like 0 +- jA the program gives me an error: Infinite solution. For
example if the function is s(s+4)/(s'2 + 100) I can't plot the bode 
diagram.
==
www.hpcalc.org
There you'll find a section named metakernel, just check there for the docs
you want
If you want to know better what's MK, just install it (if you have an
expande 48GX, or download emu48 and install MK there) You'l see
==
alg216brique book from Ren216e De Graeve (english version) and I have 

test the
 examples.
Some of the commands doesn't return the output in the same order as is in 

the 
book.
The problem is the chapter 5 - Bac 00 and HP49G.
In page 112 has the 1st exercise
The goal of this exercise is to plot the curve T represented by 
M:1/2*z^2-z
with m represented by z, is a circle C of center O and radius 1. Let it be 

a
real in [-pi,pi] and m the point of C corresponding to z=e^(i*t).
1. Calculation of the coordinates of M:
with the help of EQW we enter the expression 1/2*z^2-z...
The expression is now in the command line, and we store it in the variable 

M.
Since z=e^(i*t) we type 
        SUBST(M,z=e^(i*t)) AND THE ANSWER SHOULD BE
(EXP(i*t)^2-2*EXP(i*t))/2
BUT at this time, there is an error. Yesterday at night the screen stay 
blank
for a while with the timeglass at top. After a while the screen have turned
black and only with a paper clip I have exit from this condition. Now, at 

the 
time of this message and with the flags below (the same as I attempt 
yesterday).
The screen has stayed blank with the hourglass and I exited with ON+C.
As last attempt I have cleared the memory when the TTRM message appears and 

don't have restored it and the answer was the expected: 
(EXP(i*t)^2-2*EXP(i*t))/2
Now I have restored the memory and search what is causing the error:
I have a var Z with the following:
I purged it and works again. My question is:
The calc should be return a Bad argument type or ask to purge the 
current
var. Or NOT ?
Now at last attempt I go to RPN (with Z var defined as above) and I call 
the
Help for subst and echoed the example and subst with my values. The error
continues. At this time I need to press BACKSPACE to restart the calc 
because
it hangs.
I remember that I could make the substitution in the expression directly 
instead
of it's name:
SUBST(Z^2/2-Z,Z=EXP(i*t)) and it works in ALG mode and RPN
In RPN:
2:'Z^2/2-Z'
1:Z=EXP(i*t)
SUBST 
The error is in evaluating the name.
My rom is the last rom 1.19-5 from B. Parisse 82-04-01
My flags are:
{# 1080030A05D90FF0h # 0h 
 # 800010004A00002Ch # 20h}
==
I don't know much about algebraic mode. I never used it. In RPN Mode
the following things happen:
Let's say you stored something like 'x+y^2' in 'M'.
If you type 'M' 'x=SIN(2*t)' SUBST then the HP49 returns 'M' because
it doesn't evaluate 'M' to see if something is stored in it.
If you type M 'x=SIN(2*t)' SUBST  or 'M' RCL 'x=SIN(2*t)' SUBST then
the HP49 returns 'SIN(2*t)+y^2' because x appears in argument 1.
But SUBST has never hanged my HP49.
==
I use RPN. I use algebraic mode only when I'm with people that don't 
understand RPN.
The error appers when I have a variable Z with the following contents:
In ALG mode it recalls the contents, but the Z var is a program not an 
algebraic.
and '1/2*Z^2-Z' in 'M'.
If I do 'M' 'Z=EXP(i*t)' SUBST , I get 'M' .
If I do M 'Z=EXP(i*t)' SUBST , I get '1/2*EXP(i*t)^2-EXP(i*t) .
==
was LIMIT. 
YThe first page looks like:
::
   CK2&Dispatch
   SYMSYM
   FPTR 6 1DF=^SYMLIMIT
   BINT73
   FPTR 6 1E0=^FLAGMATRIXLIMIT
;
The BINT73 is #49h and in the sysrpl objects type the 4 is an array and 9 

is
a symbolic class. With this 'discovery' I made a test:
1) I entered the following matrix:
[ [ 'X-1' '1/X' ]
  [ 'X^2+2*X-3' 'X' ] ] in the level two of the stack and 'X=1' in level 
one
and then I called the LIMIT function:
The result was a strange matrix like this:
[ '[ 0 1 ]' '[ 0 1 ]' ]
And if I press Down arrow appears a matrix 1x2 with the '[ 0 1 ]' and '[ 0 1 

]'
in column 1 and column 2 respectively, instead of [ [ 0 1 ] [ 0 1 ] ]
I'm using Rom 1.19-5 from Bernard Parisse 2/05/2001
Hi Jose,
looks to me like the HP forgot to construct the right type of object
for the user. Instead of the 2x2 matrix it gave you a 2x1 matrix where
every row is an algebraic with a 2 elements vector.
Anyway, nice to know that LIMIT also accepts a matrix and an equation
as arguments.
==
every row is an algebraic with a 2 elements vector.
It seems that there is a lot of unknown stuffs with that calculator. It's a
great and nice surprise for me too. I love more and more my 49g :)
==
How do I convert a complex number in the form 283<32degrees (where the <
represents an angle-symbol) to the form 239,998+149,967*i ??
** note:cylindrical coordinates = polar coordinates
you can also type in the command CYLIN
or to convert to cartesian use type the command RECT
The problem with this 'conversion' on the 49 is that
it completely changes the calculator display mode, for
all numbers displayed on the stack, and for all future
displays, until changed again. What is wanted is the
actual conversion function, standard on almost all of
conversion functions take a single pair (or complex
number) as argument, and return the (cylindrical) polar
representation of that number only, and the inverse.
The mode changes CYLIN or RECT, do not do that,
(see above) and have a number of side effects which
frequently are not desired.  It seems awkward to 
have to write separate little utilities to recapture
previous performance features.
But - as frequently happens - I may have missed some-
thing in the 49 with its sadly lacking documentation.
So I ask, IS there an actual conversion function, other
than the complex ARG and ABS function ? (again, not
exactly the same thing as the desired function - more
steps required.) 
correct (again the < is the angle-sign from the CHARS menu).
However, what I wanted was to be able to enter:
                RECT(238.0,£32) where the £ is some 
symbol, and get the
result: (239.998+149.967*i)
On the place of the £, I have tried using the angle-sign from the
CHARS-menu, but this gives me an Invalid Syntax with the angle-sign
highlighted as the problemcauser.
I have also tried using RECT(238.0,32) but this only gives me: (283.0 32)
which wasn't what I needed...
What am I supposed to write at the place og the £ to convert the 

polar
(283.0,<32) to the rect. (239.998+149.967*i) or (239.998,149.967)??
months) and i never had a 48, but you can divide the complex number in two 

parts
with ABS and ARG and then make an array (or vector). the other way around 

you
would have to be in cylin  mode(the calc converts automatically to whatever
format you are using, that is won't let you write a number with an angle 
while
in XYC mode) and do the reverse operations. 
probably there is a 'menu call' that you can perform (using a number call). 

I
have seen this before with menus that belonged to the 48 model (reading 
this
newsgroup). in other words, i don't know how, but you should be able to get 

the
same menu that you had in your 48. perhaps one of the experts can tell us. 

i
want to know too. :-)
[ALPHA] [RIGHT-SHIFT] [6]
mine works, but again, it will automatically change the format to XYZ
(cartesian)mode (if that is the mode you are on).
it won't let you keep the < if that is what you wanted 
[ALPHA] [RIGHT-SHIFT] [6] was what I ment by the angle-sign (or the <).
However, using this it gives me a syntax error wether I'm in polar og rect.
mode.
Does anyone know how that can be done??
Complex numbers are enclosed in parentheses, so try RECT((283.0,<32)) 
where < is the angle-sign.  RECT and CYLIN really just change the 
coordinate mode of the display -- complex numbers are always actually 
stored in rectangular coordinates.  
You can enter complex numbers using either rectangular coordinates 
_or_ with magnitude and angle, _whichever_ display mode is in effect!
They will then be displayed in the current coordinate mode.
  
Just remember to enclose in parentheses and separate the parts with 
a comma for rectangular input or the angle sign for polar. 
Remember also to set the correct angle mode: degrees, radians, 
or grads, when entering angles.
In RPN mode you can separate the x and y values with a space instead 
of a comma, but you need the comma in algebraic mode on the hp49.  
Since complex numbers can be added, multiplied,etc just like 
ordinary numbers (no matter what the display mode) there is less 
need to switch modes than in early hp calculators where the parts 
had to be manipulated to do calculations.  You can also use the 
MODE menu and select Rectangular or Polar to change the display 
of numbers on the history stack.
If some other time you want to enter rectangular components
and see polar form displayed, do a CYLIN command to change
the display mode and enter rectangular coordinates, e.g.
  (3 4)
Now you *automatically* see the polar form on the stack.
all conversion has become *automatic* !!
During input, the mode (polar or rectangular) of your input
is indicated by whether you include or omit the <) symbol.
During display, the polar or rectangular display of
complex numbers (and 2D or 3D vectors) is determined by
commands, and shown on the screen by the presence or absence
of the same <) symbol.  The display mode commands are:
RECT, CYLIN, or SPHERE
(either of the latter is polar, differing only in 3D vector displays)
If you'd like the angle symbol to be assigned to, say, the
left-shifted zero key, to make it handier to type polar input,
then the one-time setup would be:
128 CHR 102.2 ASN
To make persistent user key assignments take effect, you must
also press left-shift USER until USR shows at the top of the screen
(but it won't show if you completely turned off the status header)
To get '3+2i' instead of (3,2) see also:
 has used HP calcs for so long that I'm sure he knows
exactly when the transition began to occur from calculators
whose registers or stack levels could contain only
plain numbers (now called real numbers)
to calcs whose stack levels could contain other objects as well.
For me, my HP15C is the first such calc I encountered;
in default mode, each stack level contains a real number
(or a matrix, but we're not interested in that just now),
two separate stack levels to operate, because
complex numbers each require two components.
In complex mode (SF 8), the story changes, because each
stack level now contains *both* parts of a complex number,
work with both parts of the complex object on the *same* stack level,
which is just as it is in the HP48/49.
only form in which the internal math functions accept complex numbers,
the final output of math operations in polar form, once again.
If those two steps could be done automatically for us,
then we wouldn't need those extra commands, after all.
By the time the HP48 and HP49 come along, that's exactly what
has happened -- we now enter or display both components
at the same time, on the same stack level,
and the <) symbol has been introduced,
which distinguishes polar coordinates from rectangular
at all times, both while entering and displaying.
All conversion now occurs automatically,
during input (compiler) and output (decompiler),
so all manual conversion is obsolete;
the unambiguous <) symbol also removes all doubt
as to which form we are dealing with at any time,
so we even can't trip ourselves up by forgetting.
But old habits die hard; it's like when the steel braces
are removed from the polio victim's legs -- even though
he can now walk without them, he keeps looking around
for something that he's so used to that at first,
it bothers him that they are missing.
Still, there's the paradox that a GUI, although more
user friendly, may also slow us down, if we are well
accustomed to using older commands and raw data arguments.
(and why shouldn't everyone have what he wants?)
well, these have only been posted a few dozen times before,
and if you'd just bookmark these for once, J.R.,
you'd have no more excuse for complaining about their loss ;-)
Two-level input, two-level output:
To preserve original flag state, if desired,
insert RCLF UNROT at beginning and ROT STOF at end.
One-level input, one-level output:
[assuming that polar flag -16 is now never set]
The fabulous HP15C:
http://www.hpmuseum.org/hp15.htm
http://www.finseth.com/~fin/hpdata/hp15c.html
The Titan was too expensive for me:
http://www.hpmuseum.org/hp71.htm
http://www.finseth.com/~fin/hpdata/hp71b.html
Actually, it was previous NON-performance features which
from early calculators was the ability to distinguish the type
of coordinates which were being manually entered (now detectable
by the presence or absence of the <) symbol between the values),
and there was also no inherent ability to display objects in
alternate ways, making conversion (actually changing the
original internal data) necessary, whereas now we leave the
internal objects alone, and just display them the way we want.
We are thus now protected from making the mistake of using
a complex number when it is in the wrong, converted form
(internal functions of complex numbers all require the
rectangular coordinate form, and will deliver meaningless results
if you convert to polar and then try to add/subtract/multiply/etc.)
That's what used to happen in HP15C etc., for example, if the user
forgot which way he had last left some complex stack value, and
kept on calculating anyway (maybe that's why the bridge collapsed :)
In the HP48 and HP49, any complex number object can exist
in only one form, by definition, and that is in rectangular
coordinate form.  When you input a complex value, you indicate
by presence or absence of <) which type of coordinates you
are feeding in, and no matter which it is, the compiler builds
an internal object containing rectangular coordinates.
When you view results, you also have your choice of which way
you want to *see* them; there is no internal conversion
of the object at all -- just a convenient *presentation*
to show you the type of coordinates you want to see,
while the internal object is never endangered by being
changed to values which are incorrect for further processing.
You may say that you can't display the whole stack showing half
the complex values in rectangular form and the other half of
the complex values in polar form all at the same time;
this is certainly true, but you can certainly display
the values yourself in text form instead, specifying
the presentation mode differently for different values.
You can also toggle the display mode at will (more easily on the
the continuing customizing ability of the 49, you can still make it
act like a 48 by re-defining some key).  Do you engineers really
feel that you need to switch *display* modes that often?
Now, don't confuse this with what you remember as how you had
to convert each polar input value separately, perform your
calculations, and then convert each output value separately;
with the HP48/49, in this same situation, you would never need
to convert anything -- you would simply type entries with
a <) symbol when entering polar coordinates, and you would
do *nothing* after your calculations, because you could leave
the display in its polar presentation mode all the time,
if that's the form in which you need to get your results.
For many people, this is actually easier than before,
but convincing them is often as difficult as telling anyone
long addicted to algebraic mode that RPN might be useful :)
It is also much safer than before, because you never destroy the
correct internal values of complex numbers, just to display them,
as you had to do on any older calc, so you can always continue
calculating with previous results, without ever fearing that
you may have forgotten which form you had left them in.
When IBM wanted to win government computer system contracts,
they once influenced the specifications to require COBOL
as the source language, thereby eliminating
some of the more modern competition :)
The more modern way of calculating with actual complex-number objects,
instead of pairs of real numbers simulating each complex number,
changes the rules on the playing field.  If you keep insisting
on the old ground rules, you'll keep getting old-style computing,
with all its extra burdens, including making more mistakes, as above.
ARG and ABS are used to isolate *single* polar components of
a complex value; this is exactly analogous to RE and IM being used
to isolate *single* rectangular components of a complex value.
out components corresponding to the current display mode --
now wasn't that a mighty clever idea?
correspond to the current display mode.  Brilliant!
the commands above to do the trick, and also preserve the
which utilize two stack levels, but to use the one-level version,
set flag -19 to get a complex number object type result, or
clear flag -19 to get a vector result with identical components.
The double-parentheses did the trick!!
It's a lot of time since a new official (non beta) rom for hp 49  hasn't
been released. Is the 1.18 the lastest?
Yes it is. Probably there'll be more beta roms, but there won't be any
official rom any more (unless HP changes and decides to prepare new
ones......)
You should change to 1.19-6 it's the latest one and it's stable
See the previous thread Beta ROM 1.19 for exact adress where you can 
find
this. The very latest beta is an improved 1.19.5, in which B. Parisse has
apparently corrected a few CAS bugs.
However it is NOT 1.19-6
It is 1.19-5, with the CAS fixes for 1.19-6.  Which means that the system
and the rest are the same from 1.19-5.  JYA will probably be releasing a
complete 1.19-6 in the next couple weeks.
I have created a page about the ROM upgrade of the HP49G. It gives
detailed instructions, step by step, showing images of both the
calculator and the PC.
I would like the users of this group to visit the page and send me
comments or corrections. I'm thinking to make an improvement to the
page, so good ideas would be apreciated.
The page is at http://www.geocities.com/CapeCanaveral/Galaxy/9055/49rom.
html
The correct adress is:
http://www.geocities.com/jcrmaster/49rom.html
for the english version, and
http://www.geocities.com/jcrmaster/rom49e.html
       I am going to ask this question,even though it has probably
been asked before. 
I know that MES is available to users of the HP49,but the instructions
at hpcalc.org seem overly complicated to a person who knows zilch
about ML and SYSRPL programming, etc.
Has the MES been brought to the surface, so to speak, in the latest
ROM version, so that idiots like myself can access it more easily?
If idiots like me can access the MES, then anybody can. ;-)
Now, how to use it. Put the a list with the equqtions of the system
you want to solve in 'EQ'. Now type: 116 MENU . This brings you to the
MES menu. Press the menu key MINIT. Then press menu key MSOLV. You get
a menu with the variables for which you can give initial values or
solve.
takes you in turn to the MES commands menu (menu 116).
There is a parallel between using the standard numeric solver menu
and the MES solver menu, as follows (this looks best in a standard
newsreader constant-width font, such as Courier):
One equation                 Multiple equations
____________________         ______________________
'one equation' STEQ          { equation list } STEQ
Stores the initial equation or list into reserved variable 'EQ'
                              MINIT
                              Copies 'EQ' into 'Mpar' (Library Data)
30 MENU                       MSOLVR
Creates menu from 'EQ'        Creates menu from 'Mpar'
SOLVR menu: near page 18-6    MSOLVR menu: near page 25-6
Generally that's all you will need, but other advanced
programmable MES commands are discussed in Chapter 25 of the
User's Guide, and/or the HP48G Advanced Users Reference (AUR):
http://www.hpcalc.org/hp48/docs/misc/hp48gug.zip
http://www.hpcalc.org/hp48/docs/programming/hp48gaur.zip
Note that MES can not numerically solve arbitrary equation sets
for all variables; there must always be a sequence
of individual equations in the list for which
only one variable at a time remains unknown (this stepwise
solution method can also lead to inconsistent solutions),
but for all the equation sets in the built-in HP48G series
Equation Library, it all works like a charm,
fortunately for Physics :)
For more demanding applications, turn to the MSLV [49 CAS]
command, or to Sune Bredahl's Solvesys library:
http://www.hpcalc.org/search.php?query=solvesys&title=1
  Knocks, Rocks, and Clocks :)
A Fountain of Time:
http://www.boulder.nist.gov/timefreq/cesium/fountain.htm
The Dawn of Time:
http://www.boulder.nist.gov/timefreq/ofm/images/people/meekhof.html
help, we now have some information about the HP49 browser. However, as it 

is
still insufficient, I am asking here (to M Brebisson or perhaps other 
member
of the development team) for more information on dynamically-changing 
browsing
box. Something like the built-in CAT, that changes menus for each command.
In my case, what I would like to do is to display some string in the 
display
whenever an element is selected (in a way that this displayed string does 

not
affect the choose box). I suppose this can be done, I just don't know how.
  
ftp://ftp.math.fu-berlin.de/pub/usr/raut/HP49/
Since I was asked by several people on how to 
translate ACC to other languages I created a
packet ACCtrans.zip on my site with detailed
instructions. You'll need only the latest
version of OT49 from 
ftp://ftp.math.fu-berlin.de/pub/usr/raut/H49/
Note that the present ACC on hpcalc.org is a
self-decompressing code object which can be 
hacked only by specialists. Moreover, it is
based on the US-calendar which differs from
the European calendar look. Thus, ACCtrans is
based on a normal calendar in which the week
starts with work (Monday) and not with Sunday.
 
PS. Although it is obvious that ACC is the 
smallest and yet most comfortable tool of its
kind, it isn't tagged with the famous GetIt
on hpcalc.org.
 ftp://ftp.math.fu-berlin.de/pub/usr/raut/HP49/
I tried it and I enjoyed the display.
Analog clock is better then digital, you have the right feeling only
by a snap shot.
Nice to pape from one month to the next or back.
I will have a look to your detailed description.
bytes is faster and smaller. It haven't the analogue clock but I don't
care about it (the 49 have a nice clock). I've recently read a similar
post about TETRI Vs Jytrix or asking somebody that he must name your
Keyman (his fault not to name it).
Sorry Wolfgang, but I think that you don't need such a self promotion
or competition because your programs are really luxury objets for
itself. They are not only useful but well programed (how many other
programers optimized his progs for GC?), holistic considering the calc
(you take EVERYTHING in account to be consistent) and you put a *LOT* of
common sense in it. 
So please don't worry about trivial things and keep giving us all the
features that designers forget :)
By the way (and making relax :)... Why you take out the pack option in
the optns menu in KEYMAN? Isn't it useful?
However I don't agree with ACC being a so useful tool. It may be a
unique tool (because none other have this two specific features) but
since I tried it I found that Calendar49 (by R. Dominguez
http://www.hpcalc.org/hp49/utils/time/calen49.zip) who takes just 606
bytes is faster and smaller. It haven't the analogue clock but I don't
care about it (the 49 have a nice clock). I've recently read a similar
post about TETRI Vs Jytrix or asking somebody that he must name your
Keyman (his fault not to name it).
The ACCfr.zip will be available, soon.
Required tools for building ACC in other languages:
OT49, ACCtrans and ACCtrans.txt
Keyman 2001.7 was very usefull (but is not required)
All is well discribed in the ACCtrans.txt.
I used 3tog to exchagange the week day GROB in the stack.
Sys~ enables to edit the back-toggled program for changing the
month names and it is need to compile back.
Compressing tools are very usefull to reduce drastically the size and avoid 

misuse.
It was my first trial, editing a SysRPL program.
 I'm in recording as a part time pro, but haven't seen anything. It would 

be
nice if you could get some
real time spectrum analyzer interface. I think in one of the Grapevine
publications, there was an I/O
interface, sort of  like the TI   CBR interface, that would allow one to
collect data and put it in some
graphing form or something. I'll see if I can find it and post it or email
me personally. If you have been
in audio long, most of the stuff is done on computers.
http://www.hp.com/calculators/techsupport/bm_manuals.html#m48g
it is part number:00048-90136 (at bottom of page)
call 1800-227-8164
costs $25.00 (includes shipping)
also there is a good userRpl manual at www.hpcalc.org (maybe the manual is 

An Hp48SX manual (both since at that time Hp documented their
calculators
instead of being CHEAP) is available via CD from the hp.museum.
This was a two volume set and the 2nd volume was a programmers
reference.  Quite a bit was changed between the s and G series, but
over 90% remained the same.
You can buy just a single cd Hp calc manual set (make sure you specify
the correct CD) for $15-$20, but you could buy the whole 5 CD set for
$35.
http://www.hpmuseum.org/software/swcd.htm, and a HP 48SX Scientific
Expandable Quick Reference Guide (pocket guide).  The HP 48
Programmer's Reference Manual was purchased separately, and, as far as
Ditto.  Without it, it's very difficult to program the
thing.  It's necessary if you want to do any programming.
It's a wonder that they can sell this thing without a
programmer's manual and still call it programmable.
I have also one of the old HP49 and its serial port works well with most
systems. But I have now trouble with a notebook.
So I called the support line for germany and the guy told me that he
wants to make the setup procedure for the HPComm software. I told him
that I measured the serial port pins with an oscillocope and that I read
about the bug here. He told me that this is not official HP (I know
that) and when the calc works with another 48 (it does) than there is no
hardware problem of the calc. Then he asked me for the SN of the product
and told me again that he wants to make the setup procedure with me.
I told him I do not want to do that and hung up.
delievered to the same supporter?) and I will see what will go on.
He remarked in the phone call that he had another person asking him for
the same problem. Is that person one of us here? If yes do you still
have the old calc or did you manage to get a new one?
I find perfectly normal that when calling the technical support the guy you
have on the phone wants to make sure that it's not a problem on your
settings.. As in this kind of problem, 99% comes from a user settings
problem.
To hang up on somebody is not really the most constructive way to resolve
problems...
I agree with you. But I was not able to reproduce it at phone because I
am not at home now. I can try to get my cable here and try to install it
and than call him again.
Remember I told him (and I did ;-) that I measured it with a
oscilliscope and saw bad signals (instead of -4,5 V I got slowing down
signals).
Sorry there seemed to be a expression problem I did of course not simply
hang up. I told him that I was not satified with his answer and said
good bye before I closed the connection ;-)
It seems you have bought the wrong calculator for your needs.
However, even if you«re used to a CAS/AOS calculator, it is 

very
easy to adapt to HP«s RPL (Reverse Polish Lisp), which is 
basically
an object oriented approach to RPN (Reverse Polish Notation) aka
UPN (Umgekehrt Polnische Notation). It will come in handy, and
you will soon find it much more natural to use and superior to an
in-line notation.
The original documentation (in particular the rather buggy German
issue) might not be the best place to start with to understand the
basic concepts, if you have never made yourself familiar with
the fundamental stack concepts found in almost any programming
language. Look out for text books on RPL, RPN/UPN, Forth, or Lisp
in the university«s library, and once you understand the idea 

of objects
on different stack levels, you will soon discover, what a bright
yet simple idea this actually is. I am quite sure this will also
give you a deeper insight into computers and programming as is.
In my opinion, CAS versus RPL is similar to the GUI versus CLI
paradigmn: At the first glance, the first appears to be easier and
more intuitive, but the more advanced you become, the more you
will discover all the unnatural restrictions by concept, and the
more you will miss the power of the latter.
I also sometimes use simple calculators for trivial maths or
easy hex calcs, but if you are facing non-trivial problems, you
will have a hard time trying to solve them with a non-symbolic,
non-RPL based calculator. It is possible, but...
One of the things you will learn in your studies is that
it is most important to find the right algorithm and data
representation for a given problem. The concepts of RPL
are definitely a very good choice for many problems.
There is an endless lists of things on the plus side, but
only very few things, I miss in or don«t like about the
HP48G(X). And they are minor in nature and have nothing
to do with the underlaying concept, but are just due to the
old hardware technology used in the calculator.
(in no particular order)
- slow user interface response especially when working
  with large objects on the stack
  (it somehow reminds me of early ZX81 times, where all the
  variable areas and even display data was dynamically
  assigned and permanently moved around in memory to
  allow it to work within just 1 Kb)
- quite huge power consumption in comparison to other
  LCD calculators I use or used (TI-54, TI-56, TI-82, Astor 1
  and Sharp PC1403H). Using state-of-the-art hardware
  technology it could either consume much less power,
  or be much faster.
- no native support for the ISO 28601/8601 international
  date format (ccyy-mm-dd)
- no Euro currency in the symbol set - I once suggested to add
  this at code point 160 (Anyway, it is very excuseable, as the
  symbol was not defined before 1997)
- no advanced interface port for external data devices (except
  for slow 3-wire serial communications) and no means to buffer
  the internal (rechargeable) batteries by connecting the calculator
  to an external power supply.
- by default, no good protection for the display, that is,
  no hard case (but see FAQ for solutions)
- no flashable firmware
- For faster typing with less finger force I would prefer a softer
  key click similar to the keyboard feeling on my TI calculators.
  Otherwise, the HP48 keyboard is very good in quality.
Some of these issues might have been solved with the HP49G
already, but from what I have heard, there was a decrease in
manufacturing quality, so i will happily stick with the HP48GX
for now.
I'd like to add, that for those of us, who have memory cards in their
4xGX, the Metakernel available on http://www.hpcalc.org + Erable +
ALG48 add tremendously to the overall performance. A much faster
equation writer, access to character design and some sort of CAS.
If you like to stick with your 48GX and consider to add memory cards,
have a look at http://www.cynox.de, you could buy a 128Kb and a 512Kb
card for about 90.-Euro. 
FEM48 and FEM49 version 4.2 have been released
SED48 and SED49 version 1.2 have been released
available at:
http://home.hccnet.nl/c.lugtmeier/
or use the direct URL's:
http://home.hccnet.nl/c.lugtmeier/HP48Fil/Fem48/FEM4849v42.zip
http://home.hccnet.nl/c.lugtmeier/HP48Fil/SED48/SED4849v12.zip
FEM49 and SED49 are ported versions of FEM48 and SED48 (porting done by
Edwin C227rdoba)
Main Features of FEM48/49:
- extremely easy to use due to excellent interface,
- supports frames, trusses and beams
- beam analysis (numerical and/or graphical)
- data can be entered in three (!) ways
    1: using a structure wizard for standard structures
    2: using input prompts for structure data
    3: assemble data arrays and store them from the stack
- text file (string) output of input and calculation results
- (un)deformed structure plots with optional displaying of supports, node
numbers, members numbers
- file management
- support for BZ compressor for saved files
- configurable matrix viewer/editor (choose your own matrix editor i.e.
MATRIX or EDITB with Metakernel)
- configurable stringviewer (i.e. VV or EDITB with Metakernel)
- optional auto-rounding of calculation results
- optional GAUSS solve routine for low memory conditions
- links to an external section database (SED48 v1.2 or higher)
- completely programmable (write your own batch files)
- moving loads generator is available
some minor (but important) differences :
5 : the form title (same way : a string).
4 : a liste of variables (global or local).
3 : the format (same way : a number or a list of two numbers).
2 : liste of reset values (all that match).
1 : an equation or a program.
When called, SOLFORM will look though the path for recall the variables
(level four) value for initialisation (if a variable do not exist, it is
created and its value is set to the reset value) : numerical solvers inform
(Polynome, System, Diff Eq, Finance) do so... Of course the othe known
syntax could be used for each variable (aka {var, text, allowed_types})
When the form is set, the first page of the menu should be :
A : EDIT (as usual),
C : CHOOSE (like numerical solvers built-in apps -but for 2nd key),
B : blank (    ),
D : blank
E : blank (instead of CANCEL)
F : SOLVE (istead of OK)
The second page of the menu will stay as it use to be
 RESET,CALC,TYPES,blank,CANCEL,OK)
Right. When SOLVE is pressed, the equation (or program) is solved (or
evalueted) for the enlighted field (aka variable) using the CAS SOLVE (not
the usual ROOT) , like what is done with the apps 'Solve Equations'...
(results are also sent to the stack, tagged with the variable)
It is just an idea like another, but what do think about it ?
Hi Gildas,
hmmm, good idea, but isn't it already possible through programming?
Now that we have InformBuilder it should be not so difficult to do
that. Or did I misunderstand you?
It is possible. Instead of SOLVe, we have the OK key. When pressed, I check
for the field left blanck (NOVAL) and solve for it (if too many fields, 
then
error message) and cycle...
Not really difficult. I just sugest hp give us a command to do it in User
RPL. I find it usefull because I often use this kinda thing to provide an
interface to my own progs (they often deal with solvings...)
Oh, I see. So everybody could solve something but with a nice
interface.
Saying interface the old numeric solver menu comes in my mind. You
know, the one that is now menu 30. Well, you can put your equation or
list of equations or list of equations + menu with programs and
everything else in the variable 'EQ' and then use 30 MENU to access
the old solver menu, which does everything that you want. (Except for
a title). Now it perhaps doesn't look so nice, but it works fine and
it can be used from within programs.
Recently I posted here asking for places to buy the 48G+ in London.
helpful.
But since then the parameters have changed. I found that German
company named Cynox (from http://www.hpcalc.org/buying.php). Their
prices seem outstanding. They ship to Greece (and claim it will take
about 10 days for the package to arrive). They even offer cash on
delivery sales which is great (I don`t have a credit card after all).
Now, here are the (somewhat obvious) questions:
1. Are they a reliable vendor? It is a sort of too good to be true
thing...
2. Their product range includes 48G+ units with added memory cards. I
have  ©.  Are these upgrades dependable or prone to 
failure? Should
I face any problems?
3. Will this void the warranty?
4. Cynox`es site and the buying guide at hpcalc.org both state that a
a 48G+ with 256KB  is slightly (3-5 EUR) more expensive that a unit
with 512KB! Am I losing something or there`s something wrong here?
 
I just got a 256k ram card for my 48gx and was wondering if it's at all
possible to install meta kernel on it.  It seems that it requires port 1
memory to install, and the docs that came with the card (and the calc) say
that it won't work in port 1 (and I'd rather not plug it in to try if I'm
not sure it'll work).  But the MK install.txt says it requires a 128, 256,
or 512k card, so maybe someone knows something I don't since when I try
and grab the file I get out of memory errors (it's like 2k too big to fit
without merged memory)
 
other one will stay hidden.
switched between banks when switched on and off), however, in my personal
experience, I've been using such card in port one for onve 2 years without 

a
single problem.
installation manual, it's pretty straightforward.
Cool, worked great.  I can't believe how much faster it really is...but
out of curiosity is there any way to install it with the card using port 1
and then moving it to the other slot so I can get to my other 128k? Not
too important at the moment, but I tend to install a lot of math/engr libs
that take a lot of space, and it hurts to know I have this extra 128k just
sitting there with no address lines talking to it :(  I tried moving it to
see what'd happen, but the most I could get it to do is show the splash
screen and drop me at my old 4-line stack.
If you plug your card in port 1 , there's no way you can see the extra
128KB.
But if you put your card in port 2 then you will see it.
The MK will not run in port 2, so just fill the extra bank, when you're
finished put your card back in port 1, and that's it...
renaming will not work but there is a nice program called XNView
that will convert almost any graphical file type into another
(including hp48/49 grobs)
it also makes grayscales (4/8/16/32)
i got that program, and i converted the bmps, but when i load it on to
the calc, its not in greyscale and theres about 6 images of the bmp i
wanted but all different..
 This is probably because the program is set in grayscale mode. You can
either use a grayscale viewer (Gdream 32 for images with 5 planes) or tell
XNview to make B&W conversion.
 
http://zap.to/hsimpson
 
Is there any simplier program (to use) than Digitician (which can 
reduce
Karnaugh maps by Quine-McCluskey's method and second program gives a truth
table from any type of boolean expression).
Why  Digitician  can't reduce this expresion: ABC' + A'C + B'C, which 

is
same as AB XOR C.
 I just got a new hp49g (today) and, according to the FAQ, I got an old
one. It came with software version 1.10. Also, the serial number
starts with ID94, which according to the FAQ, is old.
Should I return this and tell them I want a new one? I got it from
 Well, you don't have the early run with a buggy serial port. You have an
improved keyboard and so on. Just upgrade your calculator to ROM 1.19-5 and
you'll be as good as anybody else here :)
 
        My HP49G (bought in Spain two weeks ago) is S/N ID93600011.
It came with ROM 1.18. I am a bit confused, as I have heard of units 
manufactrured in China, etc.
        Is this a current unit? I know the 1.18 ROM is the latest version, 

but I 
am not sure about the serial number. Is this the unit with the improved 
keyboard and screen, or is it one of the bad units?
       The cited part is correct but perhaps the words could be better.
I would change (those with serial number under ID94...) to (those
with serial number below ID94...) for better understanding.
 
I've got one of the older ones (ID939...) but never had any problems. 
When it was new, the keys stuck a little but that changed with using it.
I just wonder why the X-key looks like someone touched it with an
solder iron at the bottom edge. *g*
And - yes - the display cover has now some scratches - but it is in use
and I can still read everything sharp and clear.
I even never had any problem with uploading a new flash ROM with
different computers.
 
 
The same goes for me (ID93303564).
But then I am not a very advanced user.
 
http://www.hpcalc.org/hp48/docs/opening/
Linux Registered User: 202 170
 
In the case of these calcs, performing a Reset does not even
clear memory, so it is not risky to perform;
its usual worst effect is to clear the stack,
although if you performed a reset during execution, say while
user memory was being re-arranged to store or purge something,
then you could actually manage to corrupt memory.
The need for a reset, however, is commonly due to an errant program
which may have corrupted memory, in which case any damage
has already been done, before ever performing the reset.
My Wizards have never actually needed a reset, but my HP's have;
perhaps it speaks of great confidence that no reset would ever
likely be needed by HP calc users, because ROM was so bug-free
(they just didn't anticipate all these SysRPL/ML coders,
who need to leave the paper clip in the hole at all times :)
 The reset switch is formed from the aluminum ground shield
 that lies along the plastic back case of your HP48.
 A piece of that shield is formed into a spring contact.
 When you press on that contact through the hole,
 the contact touches a large pad on the circuit board,
 grounding the Yorke IC Reset pin.
 If the spring contact has broken,
 you could still actuate the Reset pin
 by inserting a metal widget (re-bent paper clip) through the hole
 to touch the large pad on the circuit board.  Ground the paper clip
 widget using a wire to the negative battery terminal.
 Conversely, you could remove the batteries and use an Ohm Meter to
 test whether the spring contact is missing by connecting the meter
 between the negative battery terminal and the paper clip widget.
 
I have opened my hp48sx and found nothing that I can repair. I
checked a big capacitor and it seemed alright.
One funny thing. The hole marked R as in reset had no button
underneath it. Instead, the button was placed at the top, right
rubber foot, below the smallest hole (unmarked). The hard reset
did no good, it just turned itself off most of the time.
Dept. of Computer Engineering      http://www.ce.chalmers.se/~thomasl/
 Once probed that they are true, i could transfer them to my g.48;
I work solving problems of ecology and biochemistry.
Please let me know if you found something like that.
 
Maybe you could use my Universal Equation Browser library.
This can be called by your own library and provides an interface akin the
built-in Equation Browser.
So, you built your own equations, grobs and text for variables, assemble
them as a library and voila....
piece of paper:
1) Press Shift-number with Lib
2) Input Y values in {}
3) Input X values in {}
4) Press F-button with word Integ above it (they were using the 
integral
rule program
hehehe, and they brought in 2 calculators into the exam because they could
grasp the concept of RPN (or they didn't realise that they could do it in
algebraic with '')
Sad thing is that they are going to sell of the calculators once they
graduate.
Ironic thing is that most of them failed the subject! hahaha.. And there
were lots of beeping from their calculators during the exam. They didn't
know how to turn off the error beeps.
 so I ended up writing my own in UserRPL in 1 night) Oh, and I got 100% on
that exam http://www.curriculum.org/occ/resources/10hp38g.htm     (English)
http://www.centre-leadership.on.ca/R1013Fr.htm          (French)
The OCC (Ontario Curriculum Centre) describes the HP 38G this way:
Graphing calculator with infrared and serial ports, built-in Aplets,
rechargeable batteries, hard plastic cover
And later, under Manufacturer's Descriptions, it's stated that:
Every HP 38G has rechargeable batteries and an English or French 
manual.
 
You can't. What's the library?
 
Have you seen http://www.hp-sources.com/rom.asp?id=12 ?
ROM pour HP49G Les ROM HP48SX/GX, ROM post216e le N216ant. means 
ROM 
for
HP49G. ROMs HP48SX/GX, ROM posted by 'N216ant'
    It seems that these ROMs can be used on the 49 to emulate the 48,
but I didn't try them.. maybe 'Jose Miguel Garc222a Moncho' would
experiment and tell us whether they work properly :-)
 
these 48 Series ROM's are to be used with Emu48,
which is capable of emulating either the 48 Series or the 49G,
and so on.
These ROM's are not for the real 49G, of course.
 
 
When trying :
LIMIT( x / ((log(x))^3 + 2x) , infinite) (tending to infinite)
It shows an error message:
Operator not implemented (SERIES)
and ? is given as result.
The correct result should be 1/2.
Is this a CAS bug or I am doing sthg wrong?
 
Kernel 2.4.1
http://xie121.infovia.xtec.es/~rblasco
Un222os, hermanos linuxeros
 
What's the link between CAS and RPN ??
One is a math symbolic package, the other one a notation. You can use the
2nd one with the first one ..
Jean-Yves
Probably he was referring to EQW.........
 
http://xie121.infovia.xtec.es/~rblasco
Un222os, hermanos linuxeros
 
My guess is that you typed 'log' instead of LOG (N.B. LOG is the
logarithm to the base 10), or you forgot 'x=infinity'
Try:
        LIMIT(X/(LOG(X)^3+2*X),X=00), where '00' is the infinity symbol.
 5
It seems that I didn't get it right :-)
Please read the following quote from Steen's reply, it contains the right
answer :
The operator that is not implemented in SERIES is LOG.
You have to convert the LOG to LN(X)/LN(10) with for example EXPLN. Then 
the
calc returns 1/2.
 
 
ERROR. (Non-Aprox Mode). But nothing bad happens when trying COS(100)
I think 40G wasn't revised at all, it has really stupid bugs like this
which could have been solved if anyone had tried it a pair fo times.
I know it's impossible to make free-bug sotware, so I accept errors when
calculating difficult limits, etc... but I think this things can't be on
a calculator either it's cheap, expensive, Casio, TI or HP, it doesn't
matter..
 
Yes it is, but you are french so you should know it :-)
The french really wanted the angle system to be based on 10
so they defined a perpendicular angle to be 100 grad
then they took the meridian passing through Paris and
defined 1/100 of 1 grad to be 1 meter
My point is: let's be grateful for the grad because without it we
would still be in the dark ages with all those feet inches and
furlongs.
--
This message was written entirely with recycled electrons
Pivo
Does HP launch software without checking it????
I think Casio does, my $8 Casio FX-82 does not have this kind of
problems.......
 
or any illegal or copyrighted postings.
 
Don't try this on a 1.19-4 HP49; it will crash the calculator, and the
TTRM
will fail. 
 
To my annoyance, there is no manual for the CAS and
I could not find one on the web.
I did find the 49g CAS manual.
What differences are there between the two CAS'?
In other words, can I use the 49g CAS manual for the 40g CAS?
Martin Cohen
mjcohen@acm.org
In the early ROM versions of the HP 49G the CAS was the same as the one 
used
in the HP 40G. If you want, if have a complete manual of the HP 40G
including the CAS in french. If you're interested, mail me and I'll send it
to you.
 
Cynox never includes CAS manual, but I bought my 40G in spain and I got
it.....
try here:
Ren216e de Graeve has written the CAS manual for the 40G
http://www-fourier.ujf-grenoble.fr/~degraeve/
The 49's is more complete and has more commands, and as it can be
updated it has less errors.
Some orders may be the same, but some others not, like domain, which
doesn't exist on 40G (try it and you'll see what happens)
 Unfortunately, I do not understand French, and Parisse's site
(http://www-fourier.ujf-grenoble.fr/~parisse/english.html)
also only has a French version, although an English version
is promised soon.
Hello
There is an English version available, and the link to Renee De Graeve's
site (where you will find the English manual) is provided.
 
No. 40G has a light version of 49G CAS, there're some functions which 

don't
exist im 40 but they are in 49.
Also 40G CAS can't be updated so I has bugs and they can't be corrected
 
As fas as I know, hp40g has CAS from ROM 1.19-1 for hp49g.. with 
his features.. and his bugs...
'Cause hp40 has no FlashROM it can't be modified, that's the 
reason hp40g CAS is not exactly the same as ROM 1.19-5 CAS for 
hp49g.
 
P.S:I need a good Math software for my hp J525 ... emu49 is 
really slooow..How about GP/Pari+C interpreter for it? This exams 
are killing me...
 s.
http://www.hp.com/calculators/france/prod/casenglish.zip
 
I think Hp40G CAS has some restrictions, it's not complete. Just try 
Domain(x) and you'll get an
error message: Not available on the HP40G.
So it was reduced
 
And if hp49 and hp40 don't react the same way, 49 CAS manual may help on
understanding the purpose and the syntax of the CAS cmd. That's the reason
why i sugest the use of 49 CAS manual (as alternative)
 
Have the CAS of 40G changed during the production? If I want to buy a 40G 

(or a 39G) do I have to be
aware of the serial no. to be sure that the CAS is the newest (or the one 

with the fewest bugs)?...
 
As far as I know it hasn't been changed. Don't worry about serial numbers, 

no rainbow effect, no reverse
keys, no damaged ports..........
  
http://www.etsiig.uniovi.es/asociaciones/clubusu
P.S: There is a version for iPaq using Linux.. Linux in a 
StrongARM PDA machine.. Was this one of the features of the 
rumoured new Linux cheap PDA from HP ????
P.S: If we could get this for our PocketPC's, this would be 
really Xpander Math software.. don't you think it? (CAS, 
plotfunc,..)
 
 
Please note that despite the fact that the linux version is better 
than the windows one, the xcas application is still in alpha stage.
Note that I did not test it (I do not have an ipaq), I just
compiled for ARM using handhelds.org facilities.
 
Yes.
Linux version:
 http://www.hpcalc.org/hp48/pc/programming/hp48xgcc-1.0.1-linux.zip
Just search gcc in hpcalc.org
HP48XGCC 1.0.1 for Unix
1278KB
 hp48xgcc is a C cross-compiler. It is a program that runs on Unix and
translates C code into objects that can
 be executed on the HP48. It is based on (well, that's an
understatement) the GNU C Compiler (gcc), therefore
 it supports the complete C language and not just a subset.
 
navigation module through DaVinci Technologies. I am sure that I will be
able to get my money back through the credit card insurance, however I am
very anxious to trace a supplier of the original or equivilent module, as
this was the reason for the purchase of the calculator.
    I believe that the module was originally distributed by Sparcom, but I
cannot trace them either. However I believe that the software was written 

by
a third party. If anyone knows the author of the software, or if he should
by chance read this himself, I would very much like to communicate with 
him.
 
It was written by Thomas Metcalf at the Hawaii Institute of Navigation.
There is no HP49C version.
by
him.
 
I have looked in various FAQS and the manuals and I am stumped
I have XVIEW installed and it is library 1600
Now I wish to install a quantum physics library and it is also library 1600
As you can imagine when I try to save it in port 0 it says object in use.
1)How can I change the name of the quantum physics library to something 
else, like library 1390 or
whatever?
2) Is this the proper way to solve this dillemma?
 
Since you didn't mention what calc you have (48 /49)
I shall discuss the 49, the 48 is similar but with some differences (between 

brackets)
1. Explode the lib into a directory
    Requires special tools; download LibMan from hpcalc.org
    (48 Hack 9.4a from the same site)
    Read the docs for further instructions
2. Change the number in the variable $ROMID to the libnumber you wish
    to give the lib. (48 same)
3. store the number 1 in the variable $CONFIG
4. rebuild the lib with CRLIB or the tools you have downloaded
    (48 use tools)
Be careful with libs that use a message table
or $CONFIG 's that are different from the standard
 
There is a library name change utility called LIBID that simplifies the 
process to:
L1: old lib no.
L2: new lib no.
The download link is  http://www.hpcalc.org/hp49/programming/libs/libid.zip 

.  I believe it works on
either HP48 or 49.
 
L1: old lib
L2: new lib no.
 
out. Is there some way of adjusting this? Does HP offer any sort of spec on
this? There is nothing in the manuals. If it is possible to get inside it
and adjust a trimmer I could probably do it. I am in the time and frequency
business and have all sorts of references and counters.
 
http://www.faqs.org/faqs/hp/hp48-faq/part1/
See 4.16 of this Web page. It says that the clock is tied to the 4MHz 
timing
pulses. A software approach may help.  You could set up regular clock
corrections through a time based program, once per day or once per hour. 
This
worked with a BASIC program on a PC. It may work on the HP48. The above Web
page refers to a speedup module.
 
interesting....I have found mine to be very accurate....lose maybe a minute
a year......must be that 1992 production time frame.  :)  Damn, just
realized that my 48GX is almost TEN years old.  I am so proud of it!
 
The HP48's clocks are almost all very inaccurate.  
Some have lucked out and gotten an instrument
with an 'accurate' clock.   HP released a machine-
code adjuster (software) for the 48SX which of
course crashes on the G/GX models.
I have (recently!) downloaded an old program
which works well on my 48GX - called, strangely
enough, 'CLOCKadj' - by (1991) a person who had
email at DEC: 'edp@jareth.enet.dec.com' which
surely is no longer accurate.  I use it, it
is quite accurate (for me, perhaps not for you).
It works by updating each night via an alarm,
after you let it observe the average clock
rate for a while (days? weeks? - longer the
better, no doubt), then calculates and sets
an alarm which updates daily to correct for
the error.  
The program is actually contained in 'Article
5590 of comp.sys.handhelds', which is basically
a text document with downloadable code inter-
spersed.  I can send it to you if you wish -
again it is just plain text - or if you are
knowledgeable in ways of usenet you may be
able to find it (I'm not).  Or it may be still
on hpcalc.org, or on a Joe Horn disk.
It can be done - tho if you are 'in the time
...business' it may still not provide the
accuracy you need, since the calculator is
still subject to temp variations, etc.
John Chaffer
-----------------------
It may also have to do with your climate.  I understand the circuit
has considerable temperature dependence
-- 
Nate Eldredge
neldredge@hmc.edu
===
Subject: Re: Clock in 48GX
available on hpcalc.org.  It's called TimeKeeper, and is a copy of a
post by Eric Postpischil, which has an ASCII formatted directory object
(all UserRPL) embedded in it.  As posted it makes the correction by a
repeat control alarm at 4 A.M., but it has provisions for changing the
scheduling of the alarm, and you can execute the correction program at
any time without waiting for the alarm to come due.  This method assumes
that the clock rate is constant; avoiding temperature changes should
help to keep it constant.  The (same?) program is also available as
TIMEKEEP on Joe Horn's Goodies Disk 9, which can also be downloaded
from hpcalc.org.
If you can set up a serial connection to one of your reference
timepieces, you might be able to get the current time from one of them
as an aid to setting the clock.
 
an earlier version, dated 1990.  I wish I had seen that
one some time back, myself.  I'm sure it is functionally
equivalent, if not identical.
I just checked my HP-48 clock after several weeks of
inattention - but with this routine running in the
background, so to speak - it is within 2 sec of USNO
clock.  That's good enough for me.
 
After taking a second look at TimeKeeper on hpcalc.org, I found that
it's corrupted, in particular the program delimiters are missing or
changed to semicolons. My guess is that this is because it was
uploaded from the calculator in translate mode 2 instead of translate
mode 3, and probably somewhere along the way the high order bit was
stripped from the characters.
TIMEKEEP in Goodies Disk 9 UTILS directory seems to be fine, and
it's dated 1994 instead of 1990.
 
Well, actually, the opening program delimiters, ASCII 171d, were
changed to + (ASCII 43d) characters, and the closing program
delimiters, ASCII 187d, were changed to ; (ASCII 59d) characters.
Another good reason for using translate mode 3.
Same except for the program delimiters are ok and the variables are in
a different order within the directory.
 
math as well. The problem I have, is that my math textbook has a
section in it for Calculator Spotlight. Unfortunatly, it spotlights
the ti 82 or 83. Most of the simple commands (i.e. fractions,
equations etc...) I have figured out the way the HP 48 G/GX does it. I
am coming upon more though that I can't seem to find, like the ti has
a table function for determing the coordinates on a graph. Is there
something in the 48 that does the same? I appoligize if this is RTFM,
but I have and, can't seem to find it. Is there a total comparrision
chart or something? I do not want to get a ti.
See Table 1.92 by Aaron Wallace at www.hpcalc.org.  If you're doing 
graphing,
get his power plot also.
Bill
alternate E-dress wtstorey@ieee.org.no.spam.please 
(Use the obvious)
greedy, but I am hoping for a chart of: TI does this    HP does it
this way
Maybe a book on RPN would help??
 
The name of the file is actually 'accu_clock.txt',
as I downloaded it from some site in UK recently.
I don't see it on any of Joe Horn's disks, looking
quickly.  It may also be on hpcalc.org,  
 
Same old trig table man and know that I can interpret
sine, cosine, tangent for a given x... ie:, but what about the
cotangent of  52.18 should net a 0.77289 according to the
table method. When I punch in 52.18 in degrees and press
for cotangent. Is there a different way to calculate cotangent
on the 48gx?
 
Unfortunately there is no single keystroke or command for cotangent.
You may want to learn a little programming while accomplishing something
useful for you, namely a menu with secant, cosecant, and cotangent. You can
at your option even make it take dms (hms) input, convert to dec.deg, then
output desired value, saving lots of keystrokes.
 
Try Libman from http://www.hpcalc.org/
 
 
Why aren't the components of curl simplified (it's just horrible to see
x^(1-1/2)))?
I get so long components that i can't see them on screen and they aren't
writen in matrix. Why?
How can i simplify curl? (and why is 1/sqr(x) simplified to sqr(x)/x)
 
I don't know! probably to gain some speed..!! sometimes automatic
simplification might be undesirable anyway :-)
Just use VIEW from the TOOLS menu (you're using the HP94G, right?)
use EXPAND . You might need  EXP2POW to simplify something like
switch to USR mode, and you're done!! a quicker access but probably a 
longer
execution time because of EXPAND..)
The form 'sqrt(x)/x' is the called the normal form for the CAS..we 
don't
like square roots in the denominator neither XROOTs: This helps the
calculator to identify 0!!
Please, read the following qoute from Parisse's reply in [Re: New approach
to square root bug ( was Big BUG (1.19-5))]
Not the most desired result you would like, but it's a normal form that
helps the calculator decide whether an algebraic expression is 0.
Normal forms are sometimes much more complicated than what you would
expect
For example take
         (x^1000-1)/(x-1)
the normal form is (after simplification by the GCD)
         x^999+x^998+...+1
looks much more complicated.
Another example is
        1/(sqrt(2)+sqrt(3)+sqrt(5)+sqrt(7))
where the normal form avoids sqrt in the denominator and is much more
complicated
In fact, mastering any CAS requires really time and a minimal understanding
how they work. [...]
Hello all,
I have written a small tool, CstLIB, allowing a 
complete customization of the libmenu, the menu
you see if pressing rightshift 2 (LIB). This is 
useful if there are many libraries in the ports, as
the titles in the libmenu can be reordered and split
into categories. First the game libs, say :-)
You have only to create a file *LPAR*. This is a list
containing the romids (= numbers) of libs in the order
you prefer in the libmenu. You may also omit a romid
of a library whose commands are not used directly and
hence only bothers in the libmen. For users of Emacs
this concerns extable, for instance. Omit also romids
of libs having an extra key for invoke, like Keyman. 
Finally, you may also have different LPAR in different
directories so that the libmenu depends on the context.
If LPAR is empty, libmenu shows the port entries only,
as does the PORT directory on the HP48.  
The program of CstLIB should directly be assigned to
the LIB-key or to righshift-hold LIB (if only the name 
were assigned the key wouldm't work in edit mode).
Clearly, CstLIB applies only in USR mode. Hence, access
to the default libmenu is granted at any moment by just
leaving the USR mode. 
If a library is purged and its romid is still in LPAR
the LIB-key errors. I could have made it also in such
a way that then the default menu is set. CstLIB is on
my site below. Let's first see the reaction :-)
- Wolfgang  
ftp://ftp.math.fu-berlin.de/pub/usr/raut/HP49/tools/
PS. This tool was promised answering to a mail of JHM.
 
I made the same thing but without the LPAR variable in
a directory, instead it stores the libmenu in the hidden directory
which makes it twice as fast as CstLIB and the buildin menu.
To make it more user friendly, it works with a list of titles
instead of libid's
Of course my program is larger than Wolfgangs and is
only available via E-mail [ hpcalc  is down :-(    ].
So if you want one send an email to
p.f.geelhoed@student.tn.tudelft.nl
Pending the report of my beta testers you will be sent the
beta3 version
 
 
Module. They charged my credit card  but no goods arrived. After several
months they said that the order has been lost in transit, and that they
would sent a new shipment. Still nothing arrived, and now they will not
answer Emails , Faxes, or the telephone,  and their Voice Mail is full,
presumably with messages from angry customers.
    Does anyone know anything about this company? Do they still exist?
    Also I am still looking for a Celestial Navigation module for the 48GX.
Does anyone know of any other companies that do one?
    Cosmo Little  Falmouth UK
I also have an order long-pending with Megha's company,
thru another vendor.  One explanation was that there
was a death in the family, but that was some time ago,
and various promises have been made - none of them
fulfilled.  I don't know what to think, now.
 
Here is their new web address.  Could be that the lost email went to an old
account.
http://www.dvtg.com/
 
 
store a Global Variable as part of the program.  When creating a library
processing by the STO command.  However, I do not know how to declare a
Global Name using Debug2.  I appreciate your guidance.
If it helps, when using CRLIB on the 49, a file called $VAR is used to list
Global Variables.  I just do not know how to do this same operation using
Debug2.
Please respond through this message board or directly to me at:
AccuGrowth@MindSpring.com
Storing:
  ' ID GlobalNamedVariable STO
Retrieving
  ID GlobalNamedVariable 
or
  ' ID GlobalNamedVariable @
Pieter
I appreciate the response.  However, I need a little more information than
what is listed below.  Because I am building a library program with Debug2,
I need to know how to pass to the MAKE LIBRARY program that I am using a
GlobalNamedVariable.  Otherwise, references are not updated correctly and 

my
program fails.  In CRLIB it is done by declaring a list of global names in
$VAR.  I just do not know how to accomplish the same action in Debug2.
I appreciate some insight.  Respond through this newsgroup or directly to 

me
at AccuGrowth@MindSpring.com
 
subprograms get executed just as a single command.
Try with this:
 
the program and executes it step by step in debug mode, while SST
executes the whole sub-program in one step. Useful if you know that
your sub-program works fine and you don't want to debug it.
 
enter in subroutine, for example:
Ç ... CMD0 Ç CMD1 CMD2 CMD3 
È È
When you reach the command CMD0 and press SST appears in the top line 
Ç CMD1 CMD2 CMD3 È, if you press SST if will 

run as if you desn't have run
 
I like this as well :) but don't you think that it would better if this
information is displayed in a 'pop-up' window rather than leaving the
current page to see the 'details' page?
 do you can tell me the difference between the HP48G and the HP48S?
I am searching the whole Internet for it.
Even on the HP-Site I did not find anything.
 
The 48 version works on both S and G series (tested on ROM revisions J, P
and R). The 49 library has been developed and tested on version 1.19-5,
hacking late)
It's not a simulation but a migration. The code has actually been ported
(not a trivial process this time). Feel free to explore it: you'll find
that this demo is not a succession of compressed grobs displayed; a macro
language has been implemented (by Feng Yuan or Jim Donnelly?), even with
functions not finally used!
I would like to dedicate these libraries to those folks in comp.sys.hp48
who are or have been interested in the HP 38G.
 
PS: Quick instructions: To quit press any key. To pause press and hold down
ON. When ready press and release B (F2) and then release ON.
 
The display of the OLD 49 is like a mirror. i removed it and would like now
to cut out the blank part. But for that i would buy a new screen-cover.
Where can I buy it?
 
Yes, that's right - it's a flag that we didn't put in STK.  This was
to avoid having to add an extra page or scroller.  See the manual for
details:
   http://www.engr.uvic.ca/~aschoorl/java/
 
My old HP-42s quite naturally applies any single-
number function to any real matrix (or any complex
matrix, for that matter) as a matter of course.
This is important in certain mathematical en-
deavors.
Apparently, HP abandoned this functionality with
the 48/49 series - it does not work on my 48SX,
48GX, or on the 49G, no matter what flags I set.
However, there are two functions that make it
easy to write a short routine to do the same
thing - namely, ALX (and/or ALM), and DOLIST.
Here is such a routine (no doubt equivalent has
been posted, in some archive somewhere..)
good name - program DOMAT
This short routine expects a numerical matrix on
stack level 2, a program containing a function
(or the name of such an existing program) on stack
level 1.  It kicks out the matrix with the function
applied to each entry.  (amat and fnc are 'local
variables'.)
Example:
  2: [[ 2.000  3.000 ]
      [ 0.000  -1.000 ]]
Execute DOMAT (routine above), with result (Fix 3):
  1: [[ 7.389 20.086 ]
      [ 1.000 0.368 ]]
I suspect it will work on complex numbers too, tho
have not tried it.
I've discovered after buying Urroz' books that there
really are some pretty powerful things in the HP-49G.
Too bad many of them will remain beyond the reach
of most users, for lack of good, easy-to-use,
supporting documentation from HP.
 
Well, of course, this doesn't work for the 48's,
since there is no ALX or ALM - which appears
to exist only in the CAS for the 49.  These two
functions are indeed pretty neat.
Ah well.  So, I don't know how to do this on
the 48 series, not that it matters.
JRC
-------------------------------------
.... there are two functions that make it
(which fails on the 48's, of course - jrc)
 .
Instead of executing your DOMAT, try the built-in MAP command, it does just
the same.
 
are there any functions on the 48 that perform the same task as the
program below or the function 'MAP' on the 49?  I'm doing state-space
control and I don't want TI owners having the advantage on this...
AFAIK there isn't, but you can make your own, the basics could be:
2.- Use the data in level one to apply a program to every data in the
matrix, which is now in the stack.
I'm too lazy to make it, but I'm sure someone will come up with a complete
program. But you should make your own, so in the future you can come up 
with
your own solutions. That's the beauty of the HPs, if it's not built-in you
can program it!
 
I was trying to follow the examples given at http://www.area48.com/integ
ral/index.html for doing double and triple integrals, using their
examples, but they gave an error like undefined name even though I
input the equation as written. What am I doing wrong here?
 
I don't exactly know what you're trying to do, but you could use the
equation writer (EQW): It's much faster and easier :)
If doing improper integration, make sure that the integral is convergent
(finite), also check that the integrand is continuous over the integration
domain.
 I get exactly the same result as you - regardless of the
setting of
the general solution flag.  I cannot find any flag setting
which
will make this work.
This is why so many of us are so disgusted with the complete
lack of decent documentation.  It makes the machine nearly
useless.
 IT WORKS WHEN IN COMPLEX MODE. 
SOMETIMES THE CALCULATOR WILL ASK TO CHANGE TO COMPLEX, BUT IN THIS CASE IT
DOESN'T ?? 
THERE ARE A LOT OF TYPING ERRORS IN THE TWO BOOKS BUT OVERALL THESE ARE THE 

BEST
2 BOOKS YOU WILL FIND AROUND. 
HAVE FUN
 
    As I don't have the book, I don't know what equation you're trying to
solve.. could you post it to this NG please :)
    By the way, which ROM version are you using? If not 1.19-5 then I 
strongly
recommend that you do the upgrade; as with this version you get online help 

and a much powerful CAS!!
 
Great quotes all, the only problem is the previous poster is correct.
Documentation is very thin on the 49. I would hope Hewlett Packards' ACO
(If it still exists) would actually find a good writer to pen a
progessive text manual for this machine.
It is a fantastic calculator- even with its' quirks, but imho it would
be much better accepted with a good old fashioned tech manual. I think
there are many users who feel this way. The users guides' HP put out are
a joke. Just for the cost of this equipment, you would think a book is
in order.The calc has been out since '99 for crying out loud.
  I know this is not really the same topic, but the documentation issue 
seems
to have reared it's head again.
I think the HP49 is a fantastic machine but I do agree that the
documentation is pathetic and HP obviously do not intend doing anything
about it.
At work, I recently had to purchase a new digital storage scope for the
electronics workshop. Needless to say it was not an HP!  HP may not care,
but I at least had the satisfaction of supporting a rival company who also
manufactures excellent equipment! (And it even came with decent manuals and
operating instructions!) For the price, I could have bought about 40 
HP49G's
We also have a number of different printers at work and you can't help
noticing that the HP inkjet printers don't quite seem to be as reliable as
some of their competitor's printers....
I can't help getting the nagging feeling that HP today is not the same
company that brought us excellent products like the HP-41 and HP-48. HP
should also realise that they have may very capable competitors and their
reputation for supplying some of the best equipment available is rapidly
declining
 
STO Error: Undefined XLIB name.
This happens no matter which library I attempt to access. Any ideas
what I am doing wrong?
 
I know, I know....I should really buy the AUR but what can I say...I am a
poor, poor college student and it broke my bank buying my new HP 48G+ (GX
costed too much for me).  I was searching around the web for a web version
and I found http://www.contrib.andrew.cmu.edu/~drury/hp48gaur/ which was
good but not really complete.  I found some dead links to supposedly 
full
versions of the AUR.  Does anyone know or have the full version of AUR that
Oh forgot to add....I found the AUR for 49G but not sure if it will do much
good for my 48G+ =P
   Since I cannot upload to hpcalc.org for unknown
reasons, the new version is for the time being available at
ftp://zon.astro.uva.nl/users/dominik/emacs102.zip
Most important changes:
- The CHOOSE box shown by RPLCPL now gives
  direct access to decompiled code (via Nosy) and to stack diagrams (via
  SDIAG) for the diplayed enties.  Thus you can check out the commands
  before completing them into the commandline.
- The command EDOB (formally DOB, renamed because of a conflict with
  Jazz49) now defaults to an excursion inot Nosy to show the extracted
  and decompiled code.  Press the key for EDOB long in order to get a
  copy of it into a secondary editor.
  The CQIF interface is not affected by this change.
- Carsten
 
Emacs 1.03 has just been uploaded to hpcalc.  Until it shows up there,
you can get it from
ftp://zon.astro.uva.nl/users/dominik/emacs103.zip
Changes
-------
- Denis Martinez' LOCATOR has been integrated into Emacs.  In a large
  programming project distributed over many files, LOCATOR allows you
  to find definitions of functions and constants in other files
  directly, without leaving the editor.  Currently geared towards ML
  projects, but we may include support for SysRPL projects in the
  donating his code.
- It is now very easy to repeat the previous search (LONG F..ind).
- EDOB defaults again to loading decompiled code into another editor.
  LONG EDOB excurses into Nosy.  Thus, the corresponding change from
  1.02 has been reversed.  Sorry for the trouble.
- Several small bug fixes and improvements.
- SDIAG has *not* changed from 1.02
- Carsten and Pivo
 I have a HP48G and I recently downloaded Emu48. I downloaded programs and
libraries. However, I cannot load them neither run them on EMU48. Can
someone explain the procedure to me.
 
placed
on the stack. Then you have to either store it as a var or in a port (if
it's a library). Of course, if it's a library, you have to perform a
WarmStart for the library to appear in the Library menu (Right+Click the ON
button then Left+Click the C button).
 You have three ways to do that :
. drag and drop the object to load from your explorer to the screen of 
emu48
. use 'Edit / Load object' from emu48
. use the command line 'DDE48' (available on 
http://privat.swol.de/ChristophGiesselink/emu48.htm)
In each of these cases, the result is that the object is put on the level 1 

of
the stack (and not in a variable as it is done when you download it to your 

real
HP-48 via the serial link).
Then you just have to do as usual :
. for a library to put in port 0 : 0 STO ON-C
. for an executable to put in the variable MyExec : 'MyExec' STO
 
Does anyone know the keyboard key for the RS & LS on the 48 emulator.  Can
seem to find them
 
I supose you use 48 ROM
LS = Insert
RS = Supr
49 ROM
LS = Shift
RS = Ctrl
 
I am wondering how to install EMU48 on my Windows computer in such a
way, that it will be able to emulate all different calculators
(48,49,39,40).  I don't seem to be able to do it.  When I run a
certain emulation, there seems to be a global customization which
causes other emu48 binaries also to try to launch that same calc.
How are other people doing this?
 
me.
with
That's the public meaning, but that's wrong. Only all models use the same
general settings like COM port or speed options.
Yah, very easy.
Some general statements:
Of course it's impossible that two or more program instances run the same
file with write access. So every Emu48 needs his own state file instance.
Shared files, like ROMs and port2 files in the HP48 are specially treated.
The first instance get write access (i.e. HP49 ROM) all others get only 
read
access on this shared file.
The trick is using the command line arguments of Emu48.
For every emulation type HP38G, HP39G, HP48SX, HP48GX, HP48G I create a
seperate state file. I save them in a data directory inside the Emu48
directory.
Then I create a link on the Windows desktop or in a folder (the most used
one are on the desktop, the others are in a folder) to Emu48, as name for
the link I use i.e. the emulated calculator (HP48GX).
Next I select the property dialog of the created link and goto the link
folder.
In the first step I change the working directory to my state file 
directory.
In the second step I append the wanted state file name behind the Emu48
call. If you don't want to use the HP48 port2 file defined in the Emu48.ini
file you can add the wanted port2 file as second calling argument. The 
port2
file must be in the Emu48 directory, otherwise you have to use the full 
path
name as 2nd argument.
Now close the property dialog. Repeat the steps with all other files you
want.
Finally you can call all instances without getting the message This file 

is
missing, ....
But have you ever tried to OK all messages until you get the Choose your
KML script? Cancel this message and call the File/Open dialog and load 

a
non used Emu48 state file. You get another Emu48 instance.
Hope this helps. It's so easy when you know how. ;-)
 
Sorry for asking a FAQ.  And thank you for the quote...
 but the top bottons are not displayed, and I have and empty place 
down.. It looks like the KML is moved up.. Could anyone tell me 
what happen?
About the buggy emulator I have written the author (Leo Bueno, 
who is from a city near me in Spain), but I have no answer. Other 
think I saw is, that bigger the emu screen is, slower the 
emulation gets (MKernel logo can be viewd moving from left to 
right when it disappears after every warm start).
Could the main author of the project, Christoph Gie§elink, do 
something to get a simpler and littler emu for PocketPC ?? (I 
think you don't want nothing to the in this way, do you?)
I would like to use software designed for math for this great 
machine, but Giac-Xcas, C++ CAS from B.Parisse, is hard to 
compile for PocketPC, and GP-Pari PocketPC version is under 
development (I promise to tell you the news about this). Any 
ideas?
 
http://www.etsiig.uniovi.es/asociaciones/clubusu
  
i also installed the winCE (vers. 2.1) emu in a philips nino510 and it is
worthless. it freezes the pda all the time (try to graph something!), also 

it
would be imposible to use it if you hadn't use the real calculator before
(everything is so small, you can't see anything) I have seen a simplified
version of the bitmap file, but i wasn't able to get it to work :-(. it was
probably made for other setup.
the real calculator with 4Mhz is at least 30 times faster than the emu with 

a
115 MHZ pda(don't take my estimates seriously), so even if i had a nice 
bitmap,
i wouldn't use it. it is not practical, but at least it looks cool.
says...
(http://desktop-os.tripod.com/jornada.htm) and in trying to modify the
bitmap that came with the 720 version, I find a problem.
I took the bitmap, brought it into Photoshop and cleaned it up so that
it is more legible. When I save it back (as a bitmap) and then pull it
into the emulator, having changed the KML script to reflect the new
file name, it appears as totally black (except for the LCD portion).
When you click on where a button is located, the button location
inverts and the function is performed, but my question is why is the
graphic all black?
 
There currently is none... However, for a slightly less powerful, yet
nonetheless truly awesome (and RPN-based, of course) PalmOS calculator, try
RPN 2.57.... It is simple, yet amazing.
 
I think JYA posted once that the emulation would not be much faster than 
the
real thing due to slow processor
 
You might want to check out MathUPro, I think its even better and it sports
the familiar HP-like f and g funtion keys!!
 
 
I have a Handspring Visor Prism (the colorversion). 
The Visors have an expansionslot called springport which uses
expansionboards/modules which follow the springboard-standard (an open
standard). 
I think a HP49G in such a board would be cool. Than you could made an
overclocked version because you already have clock in the Visor. 
You could also let the Visor's processor take care of the screen and let
the Saturn take care of the mathematics. Maybe it is even possible to
make the two processor cooperate even more.
You have IrDa-port in the visor, and you could let the calculator use
it. 
You could also use the Visor's cradle to connect the calculator to a
computer. 
Finally the only thing I would miss is a real keypad and functionbuttons
like on a real calculator, but that problem could be solved if HP sold
the calculatormodule together with a simple keypad/board to connect to
the visor. You can already get normal keyboard to the Visor called
StowAway-keyboard.
If you like the idea please tell HP about it :-)
 
It's not specifically an HP calculator, but pcalc has an RPN mode that 
works much like an HP calculator:
 
There was also HPCalc by Mike Wessler (I think), try a search...
-- 
 
   see http://www.madrau.com/html/others.html They have an application 
called tpm48. I've never used it but it seems to be what you're looking 
for. I've been using CalcWare link from Sparcom corporation, but I'm not 
sure they still support the product.
  The FreeBSD ports collection lists an HP-48sx emulator as well, and with 
a
bit of fiddling with the config files, you should be able to get the source
to compile on Darwin.  www.freebsd.org/ports/emulators and look down the
list for x48-0.4.1.
 
  You'll find it at http://www.emulation.net/
--  
 
  DEPTH
  #0=?SEMI
  DUP
CODE
GOSBVL PopASavptr
D0=A 
D0+10
A=DAT0  B 
LC 40
A=A-C  B 
GOSBVL GETPTRFALSE
A=DAT0  A 
D0+5
PC=(A)  
*true
GOSBVL GETPTRTRUE
A=DAT0  A 
D0+5
PC=(A)  
ENDCODE
  ITE
  ::
    CDR$
    ::

    SWAP&$
    
;
@
    &$
    xASM
    EVAL
  ;
  ::
    palparse
    case
    EVAL
    LEN$
    SWAPROT
    CMD_PLUS
    #1-DUP
    CMD_STO_FIN
    SWAP#-
    CMD_STO_DEBUT
    SYNTAXERR
  ;
;
 
UserRPL actually does almost anything you need if you restrict
yourself to the legal objects.  LEN$ is what SIZE does for strings.
While this is an interesting example for alphaENTER, I would hesitate
to weaken safety in the UserRPL world in this way.  For the purpose of
access to such SysRPL words I have my own private library where I put
in small commands with proper argument checking.  For your example,
something like
:: CK1&Dispatch 
   str :: LEN$ UNCOERCE ; 
;
would do the trick.  Also, check lib 256 which already gives access to
a few SysRPL words in this protected way.
- Carsten
There are some things that would be nice.  I don't own an HP49, but I know
something that I'd like to have on my HP48 would be an UNROT or some
other systemRPL commands that have no userRPL equivalent.  Another would
be COERCE.
 
I have found this to be a pretty good solution to this problem as well.
I just made a utils library for the HP48 with all the system RPL stuff 

I
need in there as visible commands.  Each time I think of a new one I 
may
need, I simply decompile the library and add another command.  I've found
this is better than memorizing entries or doing COERCE ASS on my HP48.
--
Aaron.
All variables means all variables which would appear
in the *numeric*solver* (30 MENU or Solve equation form).
which does not examine valid programs to locate variables,
but instead errors or returns variable names containing programs
(so that your clean solver would then purge the programs!)
This is not only the fastest way to do the job, but also
does not require storing the expression or equation into 'EQ'
(although if the equation is already there, then do RCEQ first).
For special lists that are valid for 30 MENU
(e.g. multiple equations, equations with user-specified menus),
only the first equation in the list is processed,
because the internal functions can not handle the entire list.
Please back up memory before experimenting with these.
HP49 only [this is new]:
I hope that this was/is a stable reference
(it is a reference by number, not a ROM address)
HP48 only:
This was an unsupported but stable ROM address.
(and Joe Horn GD#10 for FINDVARS)
-----------------------------------------------------------
or any illegal or copyrighted postings.
The code shown by Nosy for the above
(and for ROMPTR 0A9 01E and ROMPTR 0A7 006,
either of which can be used to create the SOLVR menu)
is the same in HP48 and HP49; the latter functions,
used in 30 MENU, eventually call the above entry point
(the one which finds all variable names for any object
which the numeric solver accepts as an equation).
Repeating the HP48-only version:
(Unsupported but stable)
 
I have installed the Equation Library in my 49g and when I attempt to
use it, I get the folllowing error message:
STO Error: Undefined XLIB name.
This happens no matter which library I attempt to access. Any ideas
what I am doing wrong?
 
Is there an equivalent to the HP48 equation library for the HP49g to
download and install? 
Are there library functions for sec, cosec etc for HP49g?
 
Look at http://www.hpcalc.org/hp49/math/misc/ .
I'd recommend EQLIB 1.0
Yes there is. I've mailed you one.
 
I'm interested in your trig lib also.
 
 
...sending it your way now.
 
have to do in order to resolve this system of equation with the HP48
V = -aRI1 + RI3 -RI1
V  = -aRI1 - RI2
I  = I1 + I2
I  = I1 + I3
I1, I2, I3 and V are unknown and R, a, I are known. What can I do?
 
Press Rshift SOLVE UP UP. Insert coefficent matrix in Field A. Insert 
matrix
costants in field B.Move, if necessary, in field X: and press solve.
Pag 18-12 Italian Manual.
If you don't have the manual take a look at hpcalc in documentation 
section.
 
When you build an Identity matrix using erables IDENT cmd why do you get # 

1
d and # 0 d instead of 1 and 0 ? I can't find anything in the docs about
identity matrix, but there must be a way of making it produce the correct
matrix in the first place. Any Ideas?
In the HP49 I type n IDN ang get an nxn identity matrix. If I remember
well, this was also possible on the HP48GX without erable. ( Or am I
already experiencing the consequences of age ? ;-) )
Nick
Does anyone out there have a program to calculate gaussian error
function (erf) for the 41c?  I have the math module, if that is needed.
I do NOT have a stat module.
 
Mi hp48 G (32k) fails when i do the esd test.
port 1 and 2 no exist in the G model , this is the reason of the fail
when i resume the test??
The number in the test is always the same. esd xxxx but then displays
fail xxxxx.
 
Nothing to worry about, it also used to fail on my calc ( before I had
two ram cards plugged in ). This is due to a bug in the ESD test. You
can find a corrected version as well as an explanation of this
behavior in : http://www.hpcalc.org/hp48/utils/misc/esd.zip .
                               
Remove the random permutation of HPRULES! from my e-mail address
before replying.
 
 
Only works in Slot 2.
Go to:
http://cgi.ebay.com/aw-cgi/eBayISAPI.dll?MfcISAPICommand=ViewItem&item=12423

  
Chotkeh EIT-FE Engineering Equation Card for HP 48SX, 48GX
This card is designed specifically for the Engineer in 
Training/Fundamentals
of Engineering examination.  It includes sections for Dynamics, 
Electricity,
Engineering Economics, Fluid Mechanics, Gas Properties, Mathematics,
Mechanics of Materials, and Thermodynamics.  Very useful for the exam or
undergraduate engineering classes.
Go to:
http://cgi.ebay.com/aw-cgi/eBayISAPI.dll?MfcISAPICommand=ViewItem&item=12423

 
Although it varies from state to state they are not banned in most
places, in particular NY. I researched this thoroughly before taking
both the EIT and PE exams. I even called the state department of
 
(the factory service manual contains the part numbers, the schematic and a
troubleshooting guide.it is used by factory techs for repair.)
 
None of which are orderable outside HP.
Which has very little of interest.  You can trace it out yourself.
Probably not, since they don't repair them.
The few HP calculator service manuals I've managed to get my hands on
have been scanned and are on the HP Museum CD set.  If anyone wants to
make other HP calculator service manuals available to be scanned, send
me email.
 
[*] They did offer service manuals for some of the HP-98xx desktop
calculators, which were made by a different HP division, but those
service manuals didn't contain schematics or component information.)
 t :
I'am sure of none !!
Did you select cells before ??
cell ?
 
Any plans to port the rest of the 48 modules to the 49 or are they incuded
in the FEM49 module.
 
Yes, FEM49 includes all modules of the FEM48 family to date.
 
I take it you use FEM49.
FEM48 has no problem here.
FEM49 (checked with emulator) seems to have no problem either.....
So, I would like to see this bug confirmed by someone (with an actual
HP49).
Caspar
--
 
You are right, this bug in the custom browser is introduced because you use
minifont. I had not realised this before (otherwise it wouldn't exist ;)
The SysRPL code uses the DISPN entry, on the HP49 this is relative because
of the minifont. I will inform Edwin and see what can be done about this.
 
It's a minor glitch in a great program.
For the time being which flags do I need to change to get the menu to
display correctly.  There are so many of them and I haven't touched them 
for
a long time.
I use the minifont because I'm greedy and want to see as much info as
possible on that little screen.
 tried a lot of them) for hp48 dedicated in particular to engineering
students: INT48(by Jeremy Laughery). It's a very complete collection
of tables in pretty print with a large number of
integrals,Fourier,Laplace,Zeta transforms,trigonometry,and so
on...(quite easy things but sometimes,or during test,memory can
fail...). I wanted to use it also on my hp49,so I tried to convert its
code with HPCONV (from hpcalc.org) but,though I had no problem with
other programs,in this case it asks something about entry points to be
entered(equivalent to number 778D...???probably something which the
program doesn't know)and then aborts conversion. Is there anything
EASY I can do?
Are you interested in this kind of things?(do you know other complete
formularies like that?)
Ah,unfortunately at the time I'm absolutely not skilled in
 
there is also something similar for the HP49. Try Integrals Sym/Num
1.0 at http://www.hpcalc.org/hp49/math/symbolic/
Hope it helps a little,
 
3) How do I disassemble a program using menu 256?
4) Why the command DOMAIN do not calculate X^2/3, |x|, etc...?
Now Area48.com has a free space for your classifieds.
It is an automated CGI page that permit you post and 
announce about HP48/49, other calculators, classes, books, cables, etc...
It is free... and the best,
you dont need password nor register.
See it at 
www.area48.com/classifieds/
  
took a statistics class last year that required a TI83+. I was
determined that any future classes I would use a HP48 and purchased
one. Why is it so slow? Part of it I am still reading the manual an
occasionally go into the wrong menu. However, I find I am entering
equations faster than the HP48 can receive the data. Add to the fact
that the screen is not nearly as clear at an angle. It has been
drilled into me that real engineers only use HP calculators; I
currently use HP32SII at work so I don&#8217;t even like having the
TI83 on my desk.
The 48 is built so that you can customize your work in it. You can
write a small program (don't say i don't want to learn to program
it. why did you buy a goddamn programmable for anyway?) that will
make fewer the steps that you do to solve a problem... put usual
commands in CST, use USR more, etcetc.
I find it hard to believe that the calc seen as one of the best ever,
doesn't fit you. It is a powerful calc, but it is only as powerful as
the way you use it. If you just use the + - * / , that's all it is...
a 4function calc.
learn it, love it, hug it. then it will love you too and give you
answers faster.
 
If you have an older 48S series, they are noticably slower than a G
series.  Both can be outrun by really fast key entry, especially the S
series.  And I also own the Ti-83 and it is out of the box, the best
stat calc available (The newer 49G may have a better stat package, but
Ti' own CAS calc the Ti-89 has somewhat less capability out of the box
than the lowly Ti-83). However it is still ti-trash and great,
comparable stat progs are available to raise your 48 way above the
ti-83 (aside from its built in stats, the Ti-83 is inferior to
everything except MAYBE a 38G).
I suspect you have the older S series and the new G series would solve
your problem with a better screen and a 50% faster response.  If you
do have a G series and you have these problems, The new 49G isn't any
better (and may even be worse, not the same type of keyboard quality
(don't bother comment on this, I am stating it is not the traditional
keyboard, not inferior, just not typical Hp old timer keys, but newer
hard rubber and not the same tactil feel. and not the traditional
enter key placement). You really don't have any alternatives, other
than switch to a brand with a faster key entry buffer.  None of the
other calc brands offer all the practical features of the HP line, but
maybe you don't need them.  The Ti-89 (won't feel any better than the
Ti-83) has a free download to turn it into an RPN calc.  It has a
faster key entry and is faster than an Hp49 for most uses (aside from
serious number crunching of Matrices).  It is a very powerful calc,
and would be enough for over 95% of the users on this board.  The
other 5% need the Hp49 and the rest of us don't want 2nd best and be
looked down upon (if truth be told).
I use a 48G at work only because of it large conversion library and I
can load up my calc with any software I might fancy to use (well up to
30K).  I actually paid less for my discontinued 48G than my 32s, so in
my own value, it is worth less and therefore is more expendable as
well.  My48s doesn't have as good of screen as the G series and the
key input is slower.  This is due directly to the 2 MHz speed of the S
series vs. 4 Mhz of the G series.
I hope my ramblings are of some value.
eq. lib. make this superior to even the 48sx unless you have some
specialized card.  The 48G+ sells for $79 at office max if they still
have them.
 
But it is - it has much faster gui and key response.
Which does crash once in a while, since it's not written in the supported
TI-BASIC.
The HP49G is actually faster than the TI89 for maybe 80% of all tasks, but
the gui feels much faster on the TI89. The HP49G has maybe 60% more
functionality than the TI89.
This eq lib is ported to the HP49G as well - it's free.
 
This is true enough to keep the lawyers happy but...
Having both on my desk I would say that for keying in formulae there isn't 

much
difference between the two machines; the faster display routines of the 49
more-or-less make up for the poor keyboard feel (if the keyboard gives a 
poor
response you need the best feedback you can get) and the functionality is
better. 
Wolfgang Rautenberg's program UNITMAN 3.2001 which re-shapes 
the Units menus for convenient use.  
Later on I noticed the same program was listed for the '49G, but this 
time with the GET-IT! label added to it.  As it would seem this program 
was written first for the '48 (I think), and later ported to the '49G, it 
seems very much illogical not to have awarded this deserving program 
the GET-IT! label for the '48. 
May I therefore kindly request that the GET-IT! label be added for the 
'48 version as well.  
I would also ask Wolfgang Rautenberg if it would be possible 
somehow to include a simple method of rearranging the ORDER of 
menu labels for the various categories of units.  Apparently the only 
method at present is to successively purge/add new menu items, 
which is very tedious and time consuming.  
 
In a new edition of the 49-Unitman it will be very easy
to rearrange the units in a chosen Unit-Dir. This includes 
purging an uninteresting unit as a special case.
Moreover, it will be possible to rearrange the Unit-Dirs menu
on the UNITS key. Maybe someone wants to have the electricity
directory ELEC on the first page, etc. Finally, one may also
*directly* rename a Unit-Dir, so that every user can choose 
his favorite name for a category of units. In other words, it 
will be possible to customize UNITS in such a way that it looses
its old-fashioneed american style. An european engineer can 
easily adapt the whole UNITS system to his needs. The present 
UNITS system, overloaded by funny and strange looking US-units, 
is a main reason why the 48/49 never became popular among 
european students of engineering.
I could do that also for the 48-Unitman, but I somehow lost 
inclination to do so. Since you are the only person who ever 
mentioned the 48-Unitman in this NG, I have the feeling there 
are not many 48-users interested in a perfect UNITS system.
 
 
That's wrong. There are still many 48 users! :-)
Yes, we may tend to be more quit reading/posting in this newsgroup
now. But when most messages refer to the HP49 that's no wonder! 
And yes: HP48 users are still interested in improved programs. unitman
is one of the really nice programs which where new in the last two
years!  I like to use unitman48 and would welcome further development
for the HP48 :-)
 
the 49-Unitman vers. 6.2001 (only on my site at present)
has everything you can imagine. For instance, Unit menus 
(maybe mixed with user-defined units) can be reordered
and even renamed in one step. And the whole UNITS system
can be rearranged (to get the unit-dir MONEY to the first
place instead of LENG, say :-)
GET-IT in hpcalc.org, I personally do not pay much 
attention to it because I was too often disappointed. 
I single person's opinion on software quality cannot
be objective. And as is well known, it is difficult in  
general to objectively measure a software's quality. 
Clearly, there are a few points one could agree upon. 
The ratio USE/SIZE should be as large as possible. For 
instance, a tetris game with 25 KB for the 48 or 49 can,
IMHO, not be classified as the best one if there are 
other tetris games with the same comfort (sounds etc) with
only 2.5 KB. But there are many other judgment points...
If Eric honoured the 49-Unitman with his GET-IT but
not the 48-Unitman, he may have his reason. Clearly,
I cannot consider this to be a motive for speeding up 
any updating of the 48-Unitman. But I would probably 
do it faster if more people would ask for it :-)
 
This would be a reason to have a download counter on hpcalc.org.  Then you
could easily see which programs are most downloaded by users.. kind of a
voting system.  Perhaps this is how Eric determines which programs
should be denoted with a GET IT?  I would guess it wuld take quite a 
bit
of time to demo every program uploaded to hpcalc.org.
 
 
Wow! Where may I ask is this documented? My docs say that solve will only 

eat a
single equation or expression.
Are there any other simultaneous equation -aware commands?
 
CAS and mathematics with the HP49G 
http://www-fourier.ujf-grenoble.fr/~degraeve/
 
Mailandnews.com is a bit slow.
Newsranger.com is pretty fast and neat,
but has no searching, nor archives.
This message is being posted from Google;
it might show up at the above places *before*
showing up on Google itself,
but better late than never :)
 
 
(then to come back you would press [CONT]), but what if you press EQW again 

(and
write sth)?
and then again [ALPHA] [L.S] [ON] (puts you back in the stack)? and so and 

so.
it creates a short of list of of HALTS modes one after the other lined up in 

a
stack.
where is the info of the different EQWs stored?
can i jump between them (as oposed to finish one after the other)or have
indirect access?
what is my limit as the number of HALTS modes?
imagine the open possibilities, if i were able to jump around between 6
different halt modes, i would virtually have 6 different working areas( 
that
could be programmed to the 6 F top keys!). now combine that with the click 

and
paste i would have a heck of a system (imagine a program that utilizes that
too). now, how can i jump?  hmm... ho well i'll go to sleep and i'll jump in 

my
dreams.
says...
Hi rcobo!
Hmmm, it also works with only [L.S] [ON]. I assume this interrupts
temporarily the EQW in order to be able to do something on stack,
which you may want to take on the EQW. I think you can chain this
operations but you can only create a second EQW window from the first
one, then a third from the second and so on. Then you can go back
sequentially from the last to the first, But you can't go from for
example from the third to the first, without encountering the second.
It is a little bit like sub-programs:
MAIN
   SUB1
      SUB2
         SUB3
         ENDSUB3
      ENDSUB2
   ENDSUB1
ENDMAIN
But the stack contains *all* your inputs, which I find very nice :-)
The max. number of those operations is limited by the available
memory.
Nick
BTW, have you noticed that you can do the same both with internal 
and with USER RPL Inform's
It's kind of interesting that with the menu key CALC, available 
on any inform when you press NEXT, you can do exactly the same, 
and it behaves the same way. Interesting, uhm??
Gregorio
gregorio@rpn48.zzn.com
http://www.rpn48.zzn.com
Oh yes, it is. Can we then assume that the mechanism is the same in both 
cases?
Nick
J.Manrique
CdU de la ETSIG
http://www.etsiig.uniovi.es/asociaciones/clubusu
thank's for this great machine.
and for this Two year of help us.
get a new one. Not being interested in expansions I have to choose
between the 48G+ and the new 49 (about which I don't know much).
What are currently retail prices (in Italy), and which of the two do
you suggest? I only saw some pics of the 49 and don't like much the
look, and those soft keys... :-)
 
I have both the HP48GX and the HP49G. Now colours, keys and the like
are subjects to personal gusto, so I can't speak about facts. Fact is
that the HP49G has the best CAS for calcs for the time being. (Not
only algebra, but also many other areas of higher math is included)
You will be surpriced!
 
If you liked your 48GX, you will like the 48G+.  No expansion ports
the 49G has much more math capability out of the box.  However it
doesn't come with built in quations, has a different type of keyboard,
and the enter key is in the lower right corner (where it is also an
equals key by default until you switch into RPN Mode)
If you are still in school or are taking some classes that require
some tough
math, buy the 49G.  It does have lots of nifty new features.  If you
are out of school and just need a handy number cruncher that you can
load up special applications, the 48G+ will do everything that you
ask.  (actually you can expand a 48GX past the 49 with memory, and the
basic CAS system of the 49
evolved from the 48G series software that is available for free)  The
49's software has been optimized, but for most use's the 48G series
software would
much more than a 49G.
 
I think I'll go for a 48G+, I am too much used to it and use it almost
always to do everyday calculations. For heavy stuff I use a laptop.
 
If you're used to the '48, then you're used to the '49. That's it - the '49
will do exactlly what the '48 will do, in the exact same way.
 
key... And it costs 170 Û, while the 49 costs 270 
Û (all including VAT
& shipping, in Italy). I really don't need the extra speed and
features of the 49, I am graduating in engineering and for hard
calculus I always use a laptop.
 
You get used to this very fast - I did, and never use my HP48GX anymore.
If you order the HP49G from Digitalis, it just cost 179 Û + 
shipment - this
is maybe 20-25 Û. Send a mail to info@dynatech-online.de for 

details. This
time.
Steen
ship to Italy (otherwise I'll contact some german friends). Do you
know what the SpeedUpMAX is?
 
I believe it's overclocking of the processor to 6 MHz (standard is 4 MHz).
The module comes with a switch, so you can switch it off, or else you
wouldn't be able to use the serial port (both the serial port and the clock
goes haywire when you turn the module on).
 
and even some functionality, if you examine it closely.
Now, for those who always wanted a really powerful CAS,
the 49 is a vastly superior development than its predecessor,
and is worth adjusting to, but for those who really don't care
much about the CAS, and want what the 48GX had already perfected,
satisfaction with a 49 is not guaranteed, and there are some people
to whom one ought not to exaggerate claims of being exactly the same
It is substantially, but not totally, compatible in programming;
it is similar in manual operation, but there have been many
adjustments in that category, some of which may be welcome
and some not, according to the taste and objectives
of the individual user.
When it comes to algebraic functions, even the original HP48S/G
functions have sometimes changed, and not always to everyone's
advantage.  If you want a little example to illustrate this,
you are most welcome to take my HP48G UserRPL triangle solver
program TRI1 and port it to the HP49G, for the benefit of
whoever might want to keep using it on this new model
http://www.hpcalc.org/hp48/math/misc/tri1.zip
u
u
Do not attempt to load and run any of the above in an HP49 as-is,
because the SYSEVAL addresses are for HP48 only (may crash HP49
unless adjusted), and one other slight adjustment is also necessary.
For experimental purposes, you can actually just delete the first
and last instances of #XXXXX SYSEVAL with the editor,
and follow the instructions in the shortest post found above
to eliminate the remaining pair of SYSEVALs in the middle (or just
then you will have a pure UserRPL program to play with,
and if you get it to work, we can later add back the original
'Here' and 'Hidden' commands which I have posted elsewhere,
which will transfer the triangle variables back to the
hidden directory again, instead of using your current directory.
their respective HP49 or HP48, and let the other folks
enjoy their own preferences as well.
 
In some ways, maybe. It was not really untrue for me though.
I always typed commands in, and seldom used the menus.
This is not the case in most commands, and as you have to set up a certain
number of flags, to get HP48 functionality (RPN mode, soft menus etc.), you
could also set silent mode on. This will get rid of the demands.
Yes. What I really meant was that the transition would be pretty smooth -
when I got the HP49G, I could play along in all the directions previously
used on my HP48, and new functionality was intuitively implemented.
 
I also had no/few problems going from the 48 to 49. Perhaps because I
did not use much of the 48 stuff on the 49. But what I used on both
calcs works the same way on both calcs. (With some precautions)
 
 
That's right, I had forgotten about the RAD-mode-blues. This is 50% a 

flaw in the HP49G in my opinion.
 
I think the best option is to download EMU48 and the ROM of the calcs
you want to buy(EMU48 emulates HP38, 39, 40, 48* and 49)
Then you can play all the way with them (the emulator is like the calc
without the keyboard). Even you can test any program you like and set
it for real calculator speed.
Go to the real shop just to test if you like the keyboard and look of
the calc and that's all!
 
how i can transform the x (9x1) solution from decimal to irrational?
i have try to push : ==Èq  and Ç== symbol q  

 but doesn't work.
how transform a vector from decimal to irrazional?
 
IF THE RESULT IS A VECTOR (i.e. of the form [ x1 x2 ... xn ] -- more
likely):
1. Convert the vector to a list:
        { '1/4' '9/20' '1/3' }
If you get notoriously long fractions, press UNDO and try something
like  10  FIX  before attempting stage 2 again.
IF THE RESULT IS A MATRIX (i.e. of the form [[ x1 ] [ x2 ] ... [ xn ]]:
1. If you're on an HP 48, turn the matrix into a list of lists. With the
matrix on L1 do:
If you get notoriously long fractions, press UNDO and try something
like  10  FIX  before attempting stage 2 again.
 
 
For example:
[ .5  .25  .125 ]
AXL
{ .5  .25  .125 }
{ 1/2  1/4  1/8 }
AXL
[ 1/2  1/4  1/8 ]
 
This is probably very easy but i'm newbie and i still can't program so i
please for your help.
I would like to do a program in which i would enter variables and variable
that i'd like to calculate. To be more concrete: i have this equation (this
is spherical triangle rule - i use it for astronomy class)
Sin(h)=Sin(f)*Sin(d)+Cos(f)*Cos(d)*Cos(H)
I know h,d,H and looking for f (and all other combinations - i have given
h,H,f and looking for d). So i'd like to enter h,d,H (and f as unknown) and
program would calculate f.
I stored this equation as variable, but it takes long time to enter all
variabels and substitute them. Another problem is that you have to 
calculate
f and d numericaly.
 
 
documentation with the HP49, even with a disclaimer that it is old or
unsupported documentation, it would be a great help to newbies, and can't 

be
terribly expensive (can it??)
Store this equation in a variable called 'EQ' and the press
RS 7 (NUMSOLV)
Enter the variables on the screen you know, then press SOLVE for the one 
you
don't!
Dennis
PS Get all the HP48 documentation at www.hpcalc.org you can. It will all be
useful at sometime or another to you!
I have a program I use for slope staking I had written a few years ago. As
of now you have to enter the elevation by hand along with the distance.
What I would like to do is incorporate our Sokkia SDL30 with has a
comminications port to send the data straight to the 48.  Of course it 
would
only be the elevations since the distances we use are measured by hand.
So what happens is the 48 asks for a reference elevation.  You input that
elevation and take you backsight rod reading.  Typically it is a positive
rod reading say 2.254 for example.  It takes that reference elevation and
adds the rod reading to come with a height.  If my reference elevations is
125.254 and my reading is 2.254 my height would be 127.508.
Now that you have your height, you then take more readings (foresights).
To my question,
How hard would it be to add the commications to our current program?  I 
have
the information below and our code for the slope staking routine is in 
ascii
form.
Measurement codes to initiate a measurement on a Sokkia Total Station
Standard Measurement commands are as folows: Start Measurement LM
      Stop measurement LT
The data returned has the following format:
  LM_-9.9999,999.999CR-LF
  Data ID, Staff Reading, Distance
  [_] signifies space (20H).
Sample program written in Qbasic:
'[Measurement request]
OPEN COM1:1200,N,8,1 FOR RANDOM AS #1 'open commport
PRINT #1, LM 'start measurement
LINE INPUT #1, d$ 'get output SDL
PRINT d$ 'print output SDL
PRINT #1, LT 'stop measurement
CLOSE #1 'close commport
 
Every time I try to simplify a quadratic extensions by using the ALG48 
version
something I need to set in the modes or something? Is the only library I 
need
installed the ALG48.lib.    Any help would be appreciated.   
 
hidden line removed 3D plotting. It is possible to easily adjust the
viewpoint so that you can rotate the graph.
amazed.
 
 
I'm just curious: what method did you used to acheive fast hidden-line
removal?
 
Have a look at the source code.. and try to understand :-)
 when plotting, for each pixel column, plotting is clipped between the
minimum and maximum values. After plotting a pixel, minimum or maximum
is adjusted.
Sounds like the horizon line or floating horizon algorithm to me.
Here is an explanation if you want more information :
http://www.idi.ntnu.no/~grafisk2/visible.pdf. ( Most of the info in
this doc seems to have come right from Computer Graphics : Principles
and Practice by Foley and van Dam. If you're interested in the theory
behind computer graphics I highly recommend this book. )
 
 
Drag the file and drop it on the emulator and then its on the stack.  
Follow
your calculator instructions from there.  
Hint : to press ON-C, press and hold the ON key while dragging the mouse 

off.
ON should stay illuminated.  Then click the C. 
Hope this helps
Bill
Bill
alternate E-dress wtstorey@ieee.org.no.spam.please 
(Use the obvious)
of the stack
Ricardo
--
Linux Registered User: 202 170
Kernel 2.4.1
http://xie121.infovia.xtec.es/~rblasco
Un222os, hermanos linuxeros
and download it into the calculator using a serial cable.
However, I am using an emulator for HP48G, its called EMU48.EXE,
works great, and I can use it for math and stuffs.  But I am
very interested in downloading some games found on the web into
this emulator , just to see how it is.  Maybe, I can download
rom from those cards, electrical engineering etc. etc.
Could someone please show me info on how I can download games/card rom
into the emulator?  Or perhaps just give me step by step if its
 
I just downloaded a game called HP Chess v1.0 and unzipped it.  I got 

the
following files:
hpchess.doc, hpchess.inf and hpchess.lib.  So according your instruction, I
launched the
emulator by clicking on emu48.exe from my windows explorer.  It now launches 

the
HP48
emulator.  You are tell me to drag and drop from my windows explorer the 
file
hpchess.lib right onto the emulator that is already running.
I see the stack shows:
1:  Library 1720:  HpC...
That seems right so far.   I am not sure what to do next.  I don't have the 

manual for this calculator.  So how do you run this thing on the stack?
 
Hi all,
What is the easiest and fastest way to hide a directory in VAR menu ?
 
try at hpcalc.org, there are some problems that can do this...
 
I think that www.hpcalc.org has some more links.
Keep in mind that a 48 cable is useful if you've got the right adaptor.
Look at this page for information:
http://www.jps.net/joehorn/cable.htm
Of course you may want to buy a cable.
The following site sells it quite cheap (not sure if proper for the 38):
http://www.samsoncables.com/catalog/prodDetail.cfm?Prod_ID=194
 
first of all I just forgot to switch to exact mode. The HP will not
differentiate then. Now as I got that working I still got a problem because
I use two independet variables x and y, that have the same index I.
The formula looks like this
Sum(from v=1 to N) [ y(v) - (a*x(v)^2 + b*x(v) + c) ]
Now I want to get the first derivate of this, differentiating for a (then 

b,
then c)
If I do not enter the indices, the HP thinks x and y are just factors and
eliminates the Sum of course.
Now how can I tell it to leave the x and y int the sum?
P.S.
It should look like
Sum(from v=1 to N) [ y(v) - (a*x(v)^2 + b*x(v) + c) ] =
= Sum(v=1 to N) [ y(v) - a*x(v)^2 - b*x(v) - c ] =
= Sum( y(v) ) - Sum( a*x(v)^2 ) - Sum( b*x(v) ) - Sum( c )
Then, when differentiating for a:    da / f(x) = Sum( x(v)^2 )
for b:                                           db / f(x) = Sum( x(v) )
for c:                                           dc / f(x) = N
How can I get that?
 
I think you should read the answers to your first question before
posting a new one that is more or less the same like the old one. It
sometimes helps. ;-)
Now what you want is not only differentiation of a sum, but also
distribition of the operator E over the sub-expressions of the sum. (
E(a+b)=E(a)+E(b) where E as always is the sum operator.) This is not
build-in in the HP49. But don't worry, you can achieve it easily. (You
did not gave your bucks in vain ;-) )
STOre the following program in the variable DSUM:
<<
  << 'DERIV(
     2 OVER SIZE 1 - SUB +
     , +
     2 OVER SIZE 1 - SUB +
     )' +
     { 'DERIV(E(&n=&a,&N,&F),&V)' 'E(&n=&a,&N,DERIV(&F,&V))' }
     |^MATCH DROP EXPAND
(Where |^MATCH is the up-arrow-MATCH command. It is the sixth command
beginning from the end of the command catalog.)
Now put your sum on the stack. Put the differentiation variable on the
stack.
Press DSUM and wait a couple of seconds, (about 12 seconds on my HP49
with your sum and 'a' as differentiation variable.) and here you are.
Note that for derivatives the command MATCH works best when they are
in a form that contains DERIV. You get troubles when you want to use
this command with derivatives in the forrm d1f(var). That is why this
program builds the appropriate DERIV-form using a string that is
puts the sum and the integration variable on the stack and then uses
the DERIV command, then the result is a derivative in  the
d1f(var)-form and the following MATCH command causes troubles.
Also note that the variables of the form '&n', '&a' and so on, are
extremely useful for pattern replacement. That means, your sum can
have any summation variable (n, m and so on) any lower and uppel
summation limit, any algebraic in the sum and any differentiation
variable. (The program wouzld also work for the sum of
'c*exp(x(n))+d*sin(y(n))' differentiated with respect to 'd'.)
 
it is unbelievable, what possibilities are hidden in this thing,
right? It doesn't only gives a large number of tools to handle your
problems, it gives you also tools for building new tools that are not
built-in. I sometimes think, after so many years I have only scratched
the surface and the real stuff still waits to be discovered.
 
I will assume that it doesn't matter which formula is to be applied when 
x = 2.
or
'f(x) = IFTE(x < 2, x+1,x+2)'    will give f(2) = 4.
 
Oops!!
You're right.. it seems that something has gone wrong with me :-)
 
  How does this combination compare to an hp 49G?
  The 100LX was a heck of a nice machine.  Qwerty
keyboard with separate numeric keypad; good tactile
feedback, built in Lotus 1-2-3, note taking, etc., and
the extra screen space (CGA comp.) would be nice.
I'm thinking this could be my ideal do-everything, take-
along box.  I'm assuming it would be slower than a
49G, and mathematically less profound, but more
convenient in other respects.  Eg., I really need the
spreadsheet and organizer functions, and I could load
in some old DOS programs.
  The 100lx is a *lot* faster.  However they are two very very different 
beasts.
Having owned both, I wouldn't even consider a 100lx as a calculator.  
Things
like solving N equations of N unknowns or just plain math are cake w/ the 

49G
due to it's far superior user interface.  However, things like solving 
symbolic
math are pretty hopeless on the 48/9.  The 49/9 is a 4 bit machine while 
the
100lx has a 16 bit processor albeit with an 8 bit bus.
Get both and use them for what they're intended for.
HP 95LX  $81
HP 100X  $85
HP 100LX $107
HP 200LX $325
HP 200LX $217.50
HP 200LX + 6MB flash  $182.50
All mint used.  So prices are quirky, but it appears to me that a 100LX 

could
be the best deal.  I doubt I'd ever use Quicken, and I'll probably want a 

flash
Occurs to me that bidders assume the 200LX is twice as good as a 100LX.  And 

if
they search the net, they find 200LX supported and going for a pretty penny 

on
the PalmTop site.
Rick Wenger
Is there some reason I *must* have the 200LX, that I don't know about?
    I've been using the 200LX/DERIVE combo for years--(started out with a 

95LX
and DERIVE 2.4) and I think this is a *wonderful* combination, but it
doesn't seem, IMHO, to be quite the same as using a calculator.  I think
DERIVE is easier to use for soving an equation for an unknown and for
integration and differentiation, but the 48/49 is infinitely better at 
using
units, if that's important to you, and fast and dirty number crunching. I
usually wind up carrying them both--what a geek!!
 
small compared to present versions. I still have 21 MB free. If one could 

only
carry one thing, this is my choice. the only problem is that the hp-48 
emulator
for DOS uses expanded memory and this is not available for a 200LX.
C 
use some of your C: memory for expanded memory.  And drivers are
available.
This is where the 200lx has an advantage if you get the 2 or 4 meg
models.  Or you can get the Thaddeus memory upgrade and use some of
that for expanded memory.  That can be done with either the 100lx
or the 200lx, but it's fairly expensive.  They also have a double
speed upgrade that a lot of people use and seems problem free.  But
I still worry so I don't use it.