A57

>: Can anybody tell me what good will an IR upgrade do???
>Depends on which one.
>If you mean the one where the resistor in the receiver is replaced, you'll
>get double the reception range. (That brings the reliable data 
transmission
>range to a whopping 1.5 ft if both HPs are upgraded !)
>Further decreasing the resistor apparently (haven't tried it on my GX)
>brings no further increase in range. BUT it does bring increased power
>consumption.

----------------------------------------------------------------------
SnowFlakizer
By TinyWanda 
TinyWanda@aol.com
SnowFlakizer draws SnowFlakes on PICT.
These may be regulation snowflakes, or modified snowflakes.
You have the option of printing them out as a continuous stream with the
Thermal Printer, or Saving them and transferring them to your big box at a
later time.
The Snow Directory has 25 Objects in it initially...
Make     The Main Program
->P.Sz   Defines the Size of PICT
Varz     Inputs all the Variables that are Needed
Save     Saves anything without Thinking
.1to1    Defines PICT with Square Pixels for the B Printer
Ptrue    Defines PICT with Square Pixels for the A Printer
Abc      Translates Decimal Numbers to Alphabet Numbers
Bep      Makes Bepping Noises
Cpict    Moves LCD view to the Center of PICT
~Sub     Inverse of SUB
Do.?     Does something or doesn't do something
Do.F     Sets or Clears a Flag
Pxdm     Returns the Size of one Pixel
->6?     Creates a Quantumized Semi-Random Destination
->12     Plots a Kalidoscopic Array of Pixels
P->r     Polar to Real
R->p     Real to Polar
.->a     Used with Save to define auto-names
.->e     Node Memories
.->^     Flake Facets
.->$     Pixel Facets
.->d     Deviation of Angularity
.->p     Pixel Density
PPAR     Plot Parameters
Flags                 Set           Off
2   Facet Reflection  Mirrored      Asymetrical
5   Destination       Print         Save
:---------------------------:: <o>
The Objects:
:---------------------------:: <o>
Make    # 29AFh  373
Make has a double click feature which allows you to call up the Variable
assigner routine ( Varz ).
If you single click the program, it will simply begin making Snowflakes 
using
the currently defined Global Variables. 
 ( see Varz )
--------------------:: <
->P.Sz   # 6A03h  63
SnowFlakizer works best (!) when the PICT is defined as a Square, so this
program allows you to input a single Real Number to define both 
parameters...
REMEMBER! If you're going to use the line thickener, with the Thermal
printer...allow for an extra pixel to be added to the width, after 
plotting,
but before Printing!
Also; If you have an old A Thermal printer, the pixels are higher than
wider...So if you define the PICT width as 165 ( 166 for skinny line plots 
)
then set the hieght at 138 or so...???
The Ratio is .827957 of Width.
OR: (!) AND: (?) If you have a new B Thermal Printer, or are using some 
other
printer that uses the correct proportions...then change 'Ptrue' in 'Make' 
to
'.1to1'...!!!
--------------------:: <
Varz    # 52B6h  783
Varz is a separate utility to assign all the Variables that Make and some 
of
the other Programs use.
The INFORM window will ask for:
   Meta Snow Flake Variables
^  Flake Facets
   The number of arms/legs on the SnowFlake
v  Pixel Facets
   The number of sides to a pixel (?)
   The new pixels can only be added to old pixels
   at certain quantified angles to the old pixels.
   Use 6 for Regulation Snowflakes.
p  Pixel Density
   The length of a Snow Pixel
   Each Snow Pixel is actually a Line drawn from 
   an old pixel to a new pixel, this is how long that
   line is, in pixels.
e  Node Memory
   The number of recallable Pixel Nodes
   To provide continuity, new pixels are added to 
   old pixels, this is how many old pixels are 
   remembered.
a  Alpha Variable
   The first Saved Name; 1 = A
   The 'Save' program saves things by assigning a
   sequential name to those items, 
   that it's given, one at a time...
   1 starts with A, then the Alpha
   Variable is incremented to B, and so on...
   After Z, it goes to AA, then AB & so on...
d  Deviation of Rotation
   When you're plotting snowflakes, the horizontal
   lines will look 'stronger' than the diagonal lines...
   If this annoys you, you might set this variable
   to something to rotate the snowflake slightly...?
F2 Facet Reflection
   Set/1 Causes the SnowFlake arms to be plotted 
         as Reflections, like a Kalidoscope.
   Clr/0 Causes the Mandala's arms to be plotted 
         Asymetrically.
   Set/1 After the figure is plotted,
         The entire GROB will be Bolded.
   Clr/0 The plotted figure will be left alone.
F4 White on Black or vice/versa
   Set/1 After the figure is plotted,
         The GROB will be Negatized.
   Clr/0 The plotted figure will be left alone.
F5 Print or Save
   Set/1 Will print each Snowflake, 
         then begin a new one.
   Clr/0 Will Save the Snowflake in the 
         Snow Directory.
--------------------:: <
Save     # C86Bh  53
Painlessly Saves any object on 1: using the numerical value in the alpha
variable as an alphabetic name...?
if alpha = 1  then the assigned name will be 'A'
if alpha = 27 then the assigned name will be 'AA'
--------------------:: <
.1to1  # C5CBh 141.5
Reconfigures PICT so that the pixels are Square.
--------------------:: <
Ptrue  # DFB7h 115.5
Reconfigures PICT so that the pixels are 'squished' so that when the PICT 
is
recalled and Printed on the A Thermal Printer, The pixels will appear 
squared.
--------------------:: <
Abc     # 6C77h  150
Converts a real integer number to is Alphabetic equivalent.
1       = A
26      = Z
27      = AA
343821  = SNOW
2953735 = FLAKE
--------------------:: <
Bep   # 74BFh  350.5
Makes Bepping noises.
Input may either be a single Real Integer 0<b<6
or 
A list of 2 Real Integers { n 0<b<6 }
n is the number of beps
b is the kind of bep
--------------------:: <
Bold    # DA6Ch  207
Makes all the lines and areas on a given GROB thicker by one pixel. Thus a
single pixel would then be 2 by 2 pixels.
--------------------:: <
Cpict    # 1264h  73
Moves the LCD to the Center of PICT.
--------------------:: <
~Sub     # 8BFDh  66
This is the inverse (?) of SUB
So that it removes the defined sub-set, 
leaving the original, without it...!
--------------------:: <
Do.?   # A3D1h 134.5
The Object on 2: is either Evaluated or Retained if the Probability on 1: 
is
met by the PsuedoRandom Generator.
1: may be either a Real Integer, in which case it's interpretted as 1 of x
chances of being Evaluated.
or, if 1: is a fractional value, then it's interpreted to be a percentage 
of
x...? ( .25 = 1 in 4 odds )
--------------------:: <
Do.F   # 4060h  35.5
2: is either a 1 or a 0
1: is the Flag to Set or Clear
--------------------:: <
Pxdm      # 77Ah  88
Returns the pixel dimension of one pixel on PICT.
--------------------:: <
->6?     # 2478h  79
Takes an (x,y) coordinate from PICT on 1: 
and Returns a quantified pixel near it, as defined by the variables .->p &
.->$. 
Out put is actually a list of the old pixel and the new one.
 
--------------------:: <
->12    # AC4Ch  272
Is used by Make, and requires a stack configuration unique to that
application...! (?)
4:   The Derived Position of the Pixel Set
3:   The Remembered Node Pixel Set ( List )
2:   The Pair of Pixels to be Plotted
1:   Dup of 2:
If the second pixel of the pair to be plotted falls on a black pixel, than 
no
pixels are plotted, and the originating pixel is deleted from the Node 
List!
--------------------:: <
P->r   # 83B2h  48.5
Polar to Real
( radius, angle ) is converted to ( x, y )
--------------------:: <
R->p   # FCA7h  48.5
Real to Polar
( x, y ) is converted to ( radius, angle )
--------------------:: <
:----------------------------------------:: o ::
The Snow Directory
:----------------------------------------:: o ::
Name:         Snow
CheckSum:  # E64Ch
Size:         3270
:-------------------:: o :
%%HP: T(3)A(D)F(.);
DIR
  Make
    << .2 WAIT KEY
      << DROP Varz
      >> IFT :&:
Cpict EVAL
      DO
        IFERR { }
RAND .38 * RAND .38
* ->V2 + ERASE
          DO DUP 1
DUP 4 ROLLD GET ->6?
DUP ->12 2 GET DUP 3
ROLLD SWAP + DUP
SIZE 183->Ge >= :: TAIL
IFT
          UNTIL
SWAP ABS PPAR 2 GET
RE 183->254 - >=
          END
        THEN
        END RAND 60
* CEIL { 3 } + Bep
PICT RCL 3 FS? :&:
Bold IFT 4 FS? ::
NEG IFT 5 FS?
        << PR1 CR
        >> 'Save'
IFTE CLEAR
      UNTIL KEY
      END DROP
    >>
  ->P183Sz
    <<
PICT Size As A Square
 INPUT OBJ-> R->B
DUP PDIM
    >>
  Varz
    <<
META SNOW FLAKE VARIABLES
{ { GD:
FLAKE FACETS } {
Ga:
ALPHA VARIABLE }
{ F2
1:FACET REFLECTION
} { .V:
PIXEL FACETS } {
240:
DEVIATION OF ROTATION
} { F3
1:THICKER LINES }
{ 254:
PIXEL DENSITY } {
} { F4
1:WHITE ON BLACK
} { Ge:
NODE MEMORY } { }
{ F5
1:PRINT OR 0:SAVE
} } { 3 3 } { 6 1 1
6 0 1 3 1 38 1 }
183->GD 183->Ga 2 FS? 183->167
183->240 3 FS? 183->254 Pxdm
ABS / 0 RND 4 FS?
183->Ge 5 FS? 10 ->LIST
INFORM
      << EVAL 5 Do183F
9 ROLLD 4 Do183F -1 1
DUP2 XRNG YRNG 5
FS? 'Ptrue' '1831to1'
IFTE Pxdm ABS * 8
ROLLD 3 Do183F ROT 2
Do183F 6 ->LIST { 183->254
183->Ge 183->GD 183->Ga 183->167 183->240
} STO
      >>
      << { 20 4 }
Bep
Uh-Oh!!!
Run Away, Run Away!!!
DOERR
      >> IFTE
    >>
  Save
    << 183->Ga Abc OBJ->
STO '183->Ga' 1 STO+
    >>
  1831to1
    << { # 1d # 0d }
PX->C { # 0d # 1d }
PX->C - V-> DUP2 MAX
OVER ==
      << SWAP / *W
      >>
      << / *H
      >> IFTE
    >>
  Ptrue
    << :&: Pxdm EVAL
OBJ-> / DUP .83788 <
      << INV .83788
* *W
      >>
      << 1.1936 * *H
      >> IFTE
    >>
  Abc
    << 
      WHILE OVER 26
>
      REPEAT OVER
26 SWAP 1 - SWAP
MOD 1 + DUP 64 +
CHR ROT + 3 ROLLD -
26 / SWAP
      END SWAP 64 +
CHR SWAP +
    >>
  Bep
    << DUP TYPE 5 ==
      << EVAL
      >>
      << 1 SWAP
      >> IFTE -> Ge GD
      << { { 20 15
.5 } { 100 20 .1 }
{ 2000 500 .02 } {
4000 1000 .005 } {
8000 10 .04 } } GD
GET EVAL -> a b c
        << RCLF -56
CF 1 Ge
          START
RAND a * b + c BEEP
          NEXT STOF
        >>
      >>
    >>
  Bold
    << -> $
      << $ SIZE 1 +
SWAP 1 + SWAP BLANK
{ # 0d # 0d } $
REPL { # 1d # 0d }
$ GOR { # 1d # 1d }
$ GOR { # 0d # 1d }
$ GOR
      >>
    >>
  Cpict
    << PICT SIZE 64
- 2 / SWAP 131 - 2
/ SWAP 2 ->LIST
PVIEW
    >>
  ~Sub
    << ROT DUP SIZE
OVER 1 6 ROLL 1 -
SUB ROT 4 ROLL 1 +
4 ROLL SUB +
    >>
  Do183?
    <<
2: DoThis
1: Chance
.2 WAIT KEY
      << DROP 
INPUT OBJ->
      >> :: DROP
IFTE DUP 1 < :: INV
IFT RAND * CEIL 1
== :: EVAL :: DROP
IFTE
    >>
  Do183F
    << SWAP :: SF ::
CF IFTE
    >>
  Pxdm
    << { # 1d # 0d }
PX->C { # 0d # 1d }
PX->C -
    >>
  ->6?
    << DUP 183->254 RAND
183->167 * CEIL 360 183->167
/ * ->V2 P->r + 2
->LIST
    >>
  ->12
    << { LINE TLINE
} RAND 2 * CEIL GET
-> |v
      << DUP 2 GET
PIX?
        << ROT 4
ROLL DUP ~Sub 3
ROLLD
        >>
        << 4 ROLL
DROP -> Ga
          << 0 2 FS?
            FOR 191 0
359
              FOR x-
Ga 1
<< R->p 191 :: CONJ IFT
V-> x- + 183->240 + ->V2
P->r
>> DOLIST EVAL |v
EVAL 360 183->GD /
              STEP
            NEXT
          >>
        >> IFTE
      >>
    >>
  P->r
    << V-> -16 SF ->V2
-16 CF
    >>
  R->p
    << -16 SF V-> -16
CF R->C
    >>
  183->Ga 1
  183->Ge 26
  183->GD 6
  183->167 6
  183->240 7
  183->254
5.17395205737E-2
  PPAR { (-1,-1)
(1,1) X 0 (0,0)
FUNCTION Y }
END
A57

====
While most of the speculation about the successor to the HP48 focuses on
things like processor speed, screen size, etc...it seems to me that that 
most
important question is 'What will the new programming environment be?'
For starters, if the rumors of a different processor are true, then we face
the prospect of learning a whole new assembly language. Also, what changes
will be made to SysRPL? Would it be unchanged, an upgrade like VC++5.0 to
6.0, or a total reworking like C to C++?
I think secretiveness on this matter is not beneficial to HP or to users. 
The
purpose of industrial secrecy is to prevent competitors from 'scooping' 
your
product. But within the market niche of the HP48, there are no competitors.
Everyone would be better served in this case if HP were to follow the lead 
in
Intel and keep the public informed of the direction of development so that 
we
can plan and prepare for the use of the new features.
Anyone care to comment?
http://www.dejanews.com/       Search, Read, Discuss, or Start Your Own    

>I think secretiveness on this matter is not beneficial to HP or to users.
The
>purpose of industrial secrecy is to prevent competitors from 'scooping'
your
>product. But within the market niche of the HP48, there are no 
competitors.
>Everyone would be better served in this case if HP were to follow the lead
in
>Intel and keep the public informed of the direction of development so that
we
>can plan and prepare for the use of the new features.
Keeping secrets is probably not beneficial to the user - yes, but it
obviously keeps your competitors somewhat at bay, helps you move the last 
of
the old stock, lets you get it right before release, keeps with HP
tradition, what else?
I think HP definitely has 48 competitors.
IMO Intel's public education is necessary to evangelize, to help open
development of systems and peripherals, and help corporations with
purchasing plans. Things move so fast in their world that the delay would 
be
intolerable.
BTW -- what about a StrongArm + Wince device? Is there any consensus among
the speculators on the CPU and OS?
----------------------------------------------------------------------
>BTW -- what about a StrongArm + Wince device? Is there any consensus among
>the speculators on the CPU and OS?
>
Why would anyone care what the speculators think?  HP is developiung the
device, and if they've decided what platform to base it on by watching this 
NG,
they need better engineers.  (Note: HP decidedly is NOT having us make 
these
decisions for them, I'm not impugning their engineers.)  If a company is
tight-lipped about it's next product, why do people always go nuts about 
what
some random, uninformed  person thinks MIGHT be coming, instead of just 
finding
out from the company when the product is released?
Not to be bitter, but I had more than enough of this in the course of the 
long,
slow death of the Amiga...

I asked about consensus because there may be some clear favorite for CPU or
storage. Others know more about state of the art low power stuff than I do.
I agree that it makes little difference and gets somewhat tiring to read
posts about stuff that will probably never happen but ....  It is fun to
speculate about what one is familiar with, to guess at what might be, and
even to see how far off you were.
I enjoy it because it puts you in the perspective of a person who is 
sitting
before a blank canvas and assigned to best fill it. You have a palette of
technologies to blend while balancing a set of constraints. How would you 
do
it?
>>BTW -- what about a StrongArm + Wince device? Is there any consensus 
among
>>the speculators on the CPU and OS?
>>
>
>Why would anyone care what the speculators think?  HP is developiung the
>device, and if they've decided what platform to base it on by watching 
this
NG,
>they need better engineers.  (Note: HP decidedly is NOT having us make
these
>decisions for them, I'm not impugning their engineers.)  If a company is
>tight-lipped about it's next product, why do people always go nuts about
what
>some random, uninformed  person thinks MIGHT be coming, instead of just
finding
>out from the company when the product is released?
>
>Not to be bitter, but I had more than enough of this in the course of the
long,
>slow death of the Amiga...
        
Twisting your post (but not its meaning) slightly:
[big snip]
] Anyone care to comment?
] ...if HP were to follow the lead in Intel...
Sorry, I have to disagree with this.  Following Intel is not something I
would *ever* do willingly; it's just not healthy to have the whole world
hanging on the coat-tails of one manufacturer like that.  It is incumbent 
on
all free thinkers to remain independent and to choose the most sensible,
rather than the most popular route.:-)
Phew---that's better!

Nah!  The next HP calculator will be an PDA type machine (similar to the
E-11) and run WinCE (or maybe linux?).  You will be able to program in VB
x.x.  The calculator 'system' will be in ROM and will be updated
periodically (knowing HP - at megabucks/update).
Brian
<73tfv0$ia5$1@nnrp1.dejanews.com>...
>While most of the speculation about the successor to the HP48 focuses on
>things like processor speed, screen size, etc...it seems to me that that
most
>important question is 'What will the new programming environment be?'
>
>For starters, if the rumors of a different processor are true, then we 
face
>the prospect of learning a whole new assembly language. Also, what changes
>will be made to SysRPL? Would it be unchanged, an upgrade like VC++5.0 to
>6.0, or a total reworking like C to C++?
>
>I think secretiveness on this matter is not beneficial to HP or to users.
The
>purpose of industrial secrecy is to prevent competitors from 'scooping'
your
>product. But within the market niche of the HP48, there are no 
competitors.
>Everyone would be better served in this case if HP were to follow the lead
in
>Intel and keep the public informed of the direction of development so that
we
>can plan and prepare for the use of the new features.
>
>Anyone care to comment?
>
>http://www.dejanews.com/       Search, Read, Discuss, or Start Your Own
----------------------------------------------------------------------
====
> While most of the speculation about the successor to the HP48 focuses on
> things like processor speed, screen size, etc...it seems to me that that 
most
> important question is 'What will the new programming environment be?'
>
> For starters, if the rumors of a different processor are true, then we 
face
> the prospect of learning a whole new assembly language. Also, what 
changes
> will be made to SysRPL? Would it be unchanged, an upgrade like VC++5.0 to
> 6.0, or a total reworking like C to C++?
>
> I think secretiveness on this matter is not beneficial to HP or to users. 
The
> purpose of industrial secrecy is to prevent competitors from 'scooping' 
your
> product. But within the market niche of the HP48, there are no 
competitors.
> Everyone would be better served in this case if HP were to follow the lead 
in
> Intel and keep the public informed of the direction of development so that 
we
> can plan and prepare for the use of the new features.
>
> Anyone care to comment?
>
> http://www.dejanews.com/       Search, Read, Discuss, or Start Your Own
>
Wendell: You raise a very interesting question: what information, if any,
should HP release regarding a new machine? You believe HP machines are so
specialized within their market niche that there is no competition, and
details regarding various aspects, i.e., programming, should be given out.
Believe me, as an HP nut (I bought my first HP-35 in 1973) I'd love to know 
a
lot myself. But the reality is that if HP has some truly revolutionary new
technology that it will use in the new machine, it will be patented, and 
this
is definitely proprietary. Leaks may occur from time to time, and they have
to be judged as to the source of the leak. But on the whole, it is unwise 
to
release information before a product is ready. I would much rather have a 
new
machine sprung on me, rather than a repeat of the ridiculous release of the
TI-89. TI hyped that to death for over 6 months before it hit the market,
then it was almost impossible to buy one for the first two months. If HP
really has something fantastic to offer, they're the first ones to try and
get it to market as soon as possible. Besides, don't you want HP to get it
right? If HP can't make a new machine to beat the TI-89, they might as well
get out of the calculator business. I want a machine that will beat the
living daylights out of the TI-89. The occasional interesting leak aside, 
I'm
willing to wait for HP to get it right and just present me with it one day 
in
a great news release that it's available at local retailers. - J.C. 
Randerson
http://www.dejanews.com/       Search, Read, Discuss, or Start Your Own    
----------------------------------------------------------------------
====
> 
> While most of the speculation about the successor to the HP48 focuses on
> things like processor speed, screen size, etc...it seems to me that that 
most
> important question is 'What will the new programming environment be?'
> 
> For starters, if the rumors of a different processor are true, then we 
face
> the prospect of learning a whole new assembly language. Also, what 
changes
I think that depends.  Sometimes the differences between assemblies are
pronounced, such as x86 and Saturn.  However, generally, assembly
languages are quite similar.
> will be made to SysRPL? Would it be unchanged, an upgrade like VC++5.0 to
> 6.0, or a total reworking like C to C++?
VC++5.0 and 6.0 is primarily compiler enhancements, not language
changes...right?  Or maybe it had something to do with STL?  I couldn't
say.
C++ isn't exactly a total reworking of c...  If you know c, its not
particularly difficult to pick up c++.  By the way, I hope HP implements
c++ or maybe java into the HP48 in addition to RPL.  RPL is a good
language, but I have to admit RPL isn't always the best language to work
with.  For smaller programs it is VERY fast, but for large complex
programs, I'd prefer some other language, such as c++. Perhaps c++ isn't
completely practical for a calculator, but I don't think Java is too far
off the wall...afterall isn't Java's motto write once, run everywhere or
something like that?
-Jim
----------------------------------------------------------------------
>Perhaps c++ isn't
>completely practical for a calculator, but I don't think Java is too far
>off the wall...afterall isn't Java's motto write once, run everywhere or
>something like that?
>
A calculator based on Java - VERY cool idea!  It'd be a great way to 
justify
the time learning to program the calc - It's job training!    Something 
like
that could pry my 48GX from my hands without them being cold and stiff. 
The problem is, if the calc was based on Java, wouldn't it be easy to rip 
out
the ROMs and have the complete new HP running on (per the slogan)  
everything? 
Michael
----------------------------------------------------------------------
====
> For starters, if the rumors of a different processor are true, then we 
face
> the prospect of learning a whole new assembly language. Also, what 
changes
Well, I know several assembly languages. Once you have learnt one,
learning
another one is not difficult.
> will be made to SysRPL? Would it be unchanged, an upgrade like VC++5.0 to
> 6.0, or a total reworking like C to C++?
C++ added several things to C, but most C programs compile fine with a
C++ compiler. If SysRPL is kept, the major difference may be the size of
the
pointers, which are 20 bits on the 48, but would surely become a
multiple of 
8 bits. The adresses would then also change, but I don't see any reason
why
the base concept (as described in rplman.doc) would change.
> I think secretiveness on this matter is not beneficial to HP or to users. 
The
> purpose of industrial secrecy is to prevent competitors from 'scooping' 
your
> product. But within the market niche of the HP48, there are no 
competitors.
Well, It's no use telling the users what the calculator will look like
if they
don't know themselves. I'm sure they are most concerned about user
features
(color/b&w/gray LCD, memory, power consumtion) than the programming
language.
> Everyone would be better served in this case if HP were to follow the lead 
in
> Intel and keep the public informed of the direction of development so that 
we
Ha, Intel *never* told the public what they were doing. They did all
their best
to hide informations, especially with Appendix H. If you want the whole
story,
try http://www.x86.org
 
T.A.
adress.
----------------------------------------------------------------------
Me gustaria saber si en este grupo hay mas gente de ingenieros de Bilbao, y
si es asi ver sus paginas,... y crear alguna p.87gina especifica de 
nuestra
escuela.
----------------------------------------------------------------------
====
        Is there any program to compile C language to HP's RPL language, 
or any other that can work on the HP48GX? If it exist, can you send it to
                  _________________________________________
                 /         T'ha escrit en Marc.              
                 |         Te ha esrito Marc.                |
                 |         You've been writen by Marc.       |
                 |                   si05614@salleurl.edu      |
                 |                   www.salleurl.edu/~si05614 |           
                 _________________________________________/
                         /
                        /
                       /
                      
                    /
                  -(oo)-   ->my mascot.
                    ][
----------------------------------------------------------------------
====
Is there an emulator of Motorola MC68000 for hp48? i've searched
hp48.ml.org, but i haven't found it
thanx for help and excuse me for my bad english :-)
----------------------------------------------------------------------
====
JA> Jazz is an assembler, not a breaklib program. Therefore you will
JA> have to use some other tools like:
JA> LIBEX, or Mika's HACK.
JA> Unfortunately, you must install the library you want to break
JA> before you can work on it.
This is not true for all library splitters. The splitter from Mika's Hack
library takes a lib as argument (it doesn't have to be installed) and 
breaks
it. LIBEX takes the number of an installed lib and breaks it.
cu
Balazs Fischer
Balazs.Fischer@studbox.uni-stuttgart.de
Fingerprint: 8F EF C0 BA 22 B4 15 B3  56 27 07 04 37 7B D6 00
... Don't say during sex: Hurry up! This room rents by the Hour!
----------------------------------------------------------------------
====
> Just have a look on the Library structure and you will see at the 
beginning
> an offset to the config object, from which you can easily find the config
> object.
Where is that information? I have not found at hpcalc how is a lib
inside of it. I have no information about HP's data, I mean: I don't
know what is inside a lib, except commands stored into something with a
CRC at the other end.
Where is information about Jazz entry points? There are a lot when I do
use EC to see some, but I don't know or understand 99 % of it.
I have searched for one week, at hpcalc and in other places, info about
the HP pointers, those stored by 8xxxx and I didn't found.
-Gilb-
00 E2 22 76 58 1F 72 BA  EA 94 78 F3 A2 63 E0 94
----------------------------------------------------------------------
====
>Where is information about Jazz entry points? There are a lot when I do
>use EC to see some, but I don't know or understand 99 % of it.
1. http://www.engr.uvic.ca/~aschoorl/
   Low left corner, entries sorted by subject and by address.
2. http://www.hpcalc.org/docs/programming/
   a) Library File Format
   b) RAMGX
   c) RPLMan from Goodies Disk 4
   d) SASM Docs
   e) Saturn Processor Reference 1.0
   f) Programming in System RPL
-- 
---
    --> Mika Heiskanen   mheiskan@gamma.hut.fi   
http://www.hut.fi/~mheiskan
----------------------------------------------------------------------
====
Does anybody know how can I protect the programs that I do. i've heard
that some programs change the vars and the other commands for something
like EXTERNAL but I'm not quite sure.
If anybody wants a C compiler for programing in C and compiling to RPL, I
See ya!!!
  *************************************************************************
  * Francesc Esplugas i Mart.92 *                                           
*
  *   si05772@salleurl.edu    *                                           *
  *   madferit@lettera.net    *                                           *
  *   frankie@lettera.net     *                                           * 

  *************************************************************************
----------------------------------------------------------------------
hi all
i just bought a hp 48 gx calculator  and i am looking for civil engineering
programs
any1 can help me?
thanks in advance
----------------------------------------------------------------------
====
Do you know how to make a program for the azimut. I've got problems with if
else end,....
----------------------------------------------------------------------
====
Is it possible to do Cubic or Quartic regressions on the HP. In STAT,
FITDATA, there are linear and quadratic regressions, but I need Cubic. I
know the TI's can do it. That's what the rest of my class has. I am
trying to defend the HP, but this would be a real bummer if it can't do
it. Please help. If there is an additional program to do it, I would be
open to that as well.
----------------------------------------------------------------------
====
Keep asking. Yes there is stuff like that. Think there is something on
Horn's disk.
You'll probably get a reply from someone on a specific program which will
handle different degrees of fit.
>Is it possible to do Cubic or Quartic regressions on the HP. In STAT,
>FITDATA, there are linear and quadratic regressions, but I need Cubic. I
>know the TI's can do it. That's what the rest of my class has. I am
>trying to defend the HP, but this would be a real bummer if it can't do
>it. Please help. If there is an additional program to do it, I would be
>open to that as well.
>
----------------------------------------------------------------------
>Keep asking. Yes there is stuff like that. Think there is something on
>Horn's disk.
>
>You'll probably get a reply from someone on a specific program which will
>handle different degrees of fit.
There are several programs to fit polynomial. In most of them you can
specify the order you want to fit too. I use sometimes PREGR which you
can find at the HP48 Software Archive (math section - I'm not sure
which subsection there)
Find your best fit there ;-)
Greetings from Cologne
Peter
E-Mail: karpfenteich@gmx.de
_______________________________
Do you know the great Frequently Asked Questions?
http://www.engr.uvic.ca/~aschoorl/faq/
and the superb HP48 Software Archive?
http://www.hpcalc.org
to look for *old* HP48 postings see
http://www.dejanews.com
----------------------------------------------------------------------
====
I will need to get one withuing a week or so, december 12 at latest, or
i will just buy one from a company, however, i woyuld prefer to buy a
used one, if anyone out there has a used 48g+, or 48gx, let me know.
----------------------------------------------------------------------
====
  I have an HP 48GX that I bought a year ago and I love it.  But, has 
anyone
out there has heard any rumors of HP putting out a new graphing calculator 
in
the near future?  Correct me if i'm wrong, but it's my understanding that
even the GX came out years ago, maybe as early as '93?  The HP Calculator
webpage has no info of such an event on the horizon.  Just curious, cuz 4 
Mhz
  JDog the Rocket Scientist
Knowledge is Power - Sir Francis Bacon
http://www.dejanews.com/       Search, Read, Discuss, or Start Your Own    
----------------------------------------------------------------------
====
Nor will they.  It is suicidal for any manufacturer to advertise a 'future'
product until that product is ready to ship.  The moment HP announces
anything about a new calculator, sales of the 48GX will come to a 
screeching
halt.  Ask Adam Osborne about that mistake - he made it famous with the
Osborne 1
Brian Denley
>  I have an HP 48GX that I bought a year ago and I love it.  But, has
anyone
>out there has heard any rumors of HP putting out a new graphing calculator
in
>the near future?  Correct me if i'm wrong, but it's my understanding that
>even the GX came out years ago, maybe as early as '93? The HP Calculator
>webpage has no info of such an event on the horizon.  Just curious, cuz 4
Mhz
>
>  JDog the Rocket Scientist
>
>Knowledge is Power - Sir Francis Bacon
>
>http://www.dejanews.com/       Search, Read, Discuss, or Start Your Own
----------------------------------------------------------------------
====
> Nor will they.  It is suicidal for any manufacturer to advertise a 
'future'
> product until that product is ready to ship.
It seems that TI doesn't subscribe to that line of thinking...
--
Aaron.
----------------------------------------------------------------------
====
Has TI advertised what their next calculator will be?
Brian
>
>
>
>> Nor will they.  It is suicidal for any manufacturer to advertise a
'future'
>> product until that product is ready to ship.
>
>It seems that TI doesn't subscribe to that line of thinking...
>
>--
>Aaron.
>
----------------------------------------------------------------------
====
TI just recently started selling the TI-89 (Sept 98), so they're some
ways from having a new calculator ready.
TI usually announces new calculators about 6 mos in advance.
Sometimes more, and there have been at least one that they never
did sell.
This strategy makes sense when you are not the technological
leader - such as IBM with the System 360 Model 91 to compete
with the Control Data 6600. IBM pulled them back, but managed
to convince some customers to hold off bying CDC 6600's, and
later bought IBM machines anyways. CDC sued, yelling that
it was an unfair sales tactic, and won (although it was some 7-10
years later).
John Edry
>
>Has TI advertised what their next calculator will be?
>Brian
>
>>
>>
>>> Nor will they.  It is suicidal for any manufacturer to advertise a
>'future'
>>> product until that product is ready to ship.
>>
>>It seems that TI doesn't subscribe to that line of thinking...
>>--
>>Aaron.
----------------------------------------------------------------------
I'm sorry in advance for asking such a stupid and obvious question but....
How would I go about transfering a text file from my computer to my 48g
(using hp48 explorer) and reading that text on the calculator? all I really
want to do is to type stuff in on my computer and be able to read it on my
calculator. are there any programs that I need? are there any programs that
would make this easier? thanks in advance
Dave
----------------------------------------------------------------------
transferring a text file to the HP48 is covered in the manual.
If there's a poblem, try the FAQ from A. Schoorl.
There a are some fast text viewers for use on the HP48.
They can be found on http://www.hpcalc.org under
text editors and viewers.
Hope this helps
Greetings
Raymond
----------------------------------------------------------------------
====
I have seen a lot of speculation about the follow-up to the HP48. Most of 
it
concerns physical things like processor speed or screen size. That seems to
be entirely missing the most important aspect of the upgrade. The number 
one
question should be 'How will programming for it be different from for the
HP48?'
Consider that there very likely will be a different processor. That means
programmers will have to learn an entirely new assembly language. Also, it
seems likely that there will be some modification of SysRPL, but we don't
know if it will be an incremental change, like from VC++5.0 to 6.0, or a
paradigm shift, like from C to C++.
I think HP's secrecy in this matter is detrimental to all concerned. The
purpose of industrial secrecy is the prevent competitors from 'scooping' 
your
technology. But within the HP48's market niche, there are no competitors!
Also, unlike with other calculators, HP48 users rely heavily on third party
software. In that sense, the HP48 market is more like the personal computer
market. The interests of HP, the developers, and the users would be best
served if HP would keep the public informed of the direction of new
developments, as does Intel.
Anyone care to comment?
http://www.dejanews.com/       Search, Read, Discuss, or Start Your Own    
----------------------------------------------------------------------
====
I'm writing a tool using ML that takes a lib as argument, and pushes on
the stack a string, with each lib's command, and its size. Like:
I don't know how is a lib's structure inside. I need to read the lib,
and find each command's start address, and search for its end (or read
its length, if it's coded too) and put in my output string the name of
the command, and the size found.
-Gilb-
00 E2 22 76 58 1F 72 BA  EA 94 78 F3 A2 63 E0 94
----------------------------------------------------------------------
====
Library:
 Prologue                      5 nibbles = 02B40
 lenght without prologe   5 nibbles
 Number Characters first name   2 nibbles
 char 1 name library             2
 ....
 char n name library         2
 Lib number                  3 nibbles
 offset to HASH TABLE       5
 offset message      5
 offset Link               5
 offset config         5
 HASH TABLE
 MESSAGE ARRAY
 LINK TABLE
 Type Xlib1                     1 or 3
 Number of lib of Xlib1   3
 Number of command of Xlib1     3
 Object(s) Xlib1           
 ....
 Type Xlibn
 Idem
 
 bye
> 
> I'm writing a tool using ML that takes a lib as argument, and pushes on
> the stack a string, with each lib's command, and its size. Like:
> 
> I don't know how is a lib's structure inside. I need to read the lib,
> and find each command's start address, and search for its end (or read
> its length, if it's coded too) and put in my output string the name of
> the command, and the size found.
> 
> 
> -Gilb-
> 00 E2 22 76 58 1F 72 BA  EA 94 78 F3 A2 63 E0 94
> 
> 
        ~+                   .
                     .          *
 ^               *       +              Saturn: More than a simply CPU?
           '                  |
       ()    .-.,=``=.    - o -       Of course, it's a planet!
             '=/_            |
          *   |  '=._    |
                    `=./`,        '
            .   '=.__.=' `='      *
   +                         +
        O      *        '       .
----------------------------------------------------------------------
have you tried the <-LIB-> package ?
There's a command inside which does what you want;-)
It's name is LIBp.
<-LIB-> can be found at hpcalc.org , of course.
You can rewrite it in ML, although I think it's fast enough..
Greetings
Raymond
----------------------------------------------------------------------
====
> have you tried the <-LIB-> package ?
> There's a command inside which does what you want;-)
> It's name is LIBp.
I don't care if there is already something that does it. I'm learning ML
and HP's objects structure. It's not by using tools that do what I
want that I will learn.
 
Like, an empty lib of number 600 gives that:
04B20                Prologue
22000                Size
00852                #258h = 600
0000000000000000000  ?
030A9                CRC
(I kept all data reversed)
And if I give it a name of 4 characters, like Vide (which means
Empty in french) I get that:
04B20                Prologue
C2000                Size
40                   #04 = lib's name size
65964656             The name
04                   #04 = lib's name size
00000000000000       ? 
0BCEC                CRC
I have looked into MAKEROM.DOC and the lib structure is not explained,
but there is something about a hash thing. Then, I looked into
USRLIB.DOC, but it only explains the $var things to use. I have finally
searched into RPLMAN.DOC, and a lot of objects are described, but not
the library one. Besides being a very interesting file, nothing in
there.
Can someone tell me how is a lib inside? I would love some text and
simple diagrams; I love reading hex bins..
-Gilb-
00 E2 22 76 58 1F 72 BA  EA 94 78 F3 A2 63 E0 94
----------------------------------------------------------------------
====
This is a copy of a posting by Mika Heiskanen :
/-- copy start ---/
        Library Format
        ==============
The library data starts with a header containing
        1. Library prolog
        2. Library size field
        3. Title
        4. Library number
        5. Link to optional hash table
        6. Link to optional message table
        7. Link to optional link table
        8. Link to optional configuration object
Next comes the main body of the library, containing all the programs in the
library. The optional objects pointed to in the header are also contained 
in
this part. Depending on how the library was built these optional objects 
can
be in various positions (in any position when the library is explicitly
created), but usually either at the start or the end of the main body.
Last comes a four nibble CRC, which is a checksum for all the data between
the prolog and the CRC itself. Note in particular that the prolog itself is
not included in the CRC.
The explicit form of the main parts is
ASSEMBLE
        CON(5)  =DOLIB
        REL(5)  EndOfLibrary
        CON(2)  title_chars     
        NIBASC  'title'           or    CON(2)  0   if there is no title
        CON(2)  title_chars     /
        CON(3)  romid
        REL(5)  hash_table      
        REL(5)  mesg_table        or 0 if the object does not exist
        REL(5)  link_table
        REL(5)  conf_object     /
        NIBHEX  ...             main body
        CON(4)  CRC             filled in by linker or special program
EndOfLibrary
RPL
Details to note are:
        o The title length may be indicated once or twice, depending
          on the title length itself. This means one must usually
          use conditional code to skip the title, I myself call
          the respective (unsupported) ROM subroutines:
                =TRAVERSE+      EQU #8400       skip title forwards
                =TRAVERSE-      EQU #841E       skip title backwards
                C=0     A
        o A zero link field indicates the respective object and all
          properties implied therein are not in use. This does not
          necessarily imply the respective object does not exist,
          as can be wittnessed by simply changing an existing library
          with a nonzero link.
        o The pointee of the link can be outside of the library itself
          as indicated by the address range specified by the address
          of the prolog and the size field of the prolog iff the library
          is compiled to a fixed address. This is done for some of the
          internal libraries to save precious uncovered ROM space (SX).
        o If the pointee in a S/SX is a system binary, the value
          of the binary implies the true address of the respective
          object.
        o If the pointee in a G/GX is an access pointer, the address
          part and access parts indicate the true address and port
          of the respective object, respectively (uh).
The message table is expected to be a 1-dimensional array of strings.
The error numbers of the messages is formed by multiplying the library
number by 256 and adding the array index to that. This number is expected
to take 5 nibbles, meaning one shouldn't have more than 255 error messages.
The link table is expected to be a hex string, whose body contains links to
the objects in the main body. Any object which is to be accessed by ROMPTR
calls must be in this library in order to be found. As can be deduced from
the following example, the number of accessible commands in a library can 
in
principle be calculated by dividing the size field by 5 and substracting
1. However some people seem to think this makes a nice protection against
splitting a library, and thus poke a very large value over the true size
field. This makes many library splitters think there are more links than
there truly are, and thus the splitter will end up trying to extract
objects from wherever the data after the true body ends points to, usually
resulting in a crash.
ASSEMBLE
LinkTab CON(5)  =DOHSTR
        REL(5)  EndOfLinks
        REL(5)  Command1
        REL(5)  Command2
        ...
        REL(5)  CommandN
EndOfLinks
RPL
The hash table is expected to be a hex string, whose body contains a
        o forward link table based on the length of a name
        o collection of name length-name-command number triplets
        o backward link table indexed by command number
The number of forward links is 16, indicating the maximum size of a command
name. The number of backward links can be smaller than the number of
commands, which implies all the remaining links are interpreted to be zero,
which in turn implies the command has no name.
Whenever the internal compiler encounters a name it does not recognize it
traverses the current path. For each library attached to the directories in
the current path it checks if the respective library has a hash table, and
for those that do it checks whether the respective forward link for the 
known
length of the name is nonzero. If so it jumps forward to that address, and
starts comparing the name against the name length-name parts of the
triplets. If a match is found then the compiler forms a ROMPTR object from
the known library number (attached to directory) and command number (third
slot in the matching triplet). The outer loop continues until a match is
found, or all attached libraries in the current path have been checked, in
which case the compiler decides the name is a global identifier.
Whenever the internal decompiler encounters a ROMPTR, it traverses the
internal ROMPTAB containing the library numbers and addresses of each
library. For each library it checks for a hash table, then skips to the
backward link part, then to the link for the known index. If this link
exists (e.g. its calculated address is before the known end of the hex
string) and the link is nonzero, the decompiler jumps backwards to the
triplet it points to, and forms a string from the name length-name pair.
The compiler is used for example when typing DUP on the command line, the
decompiler when typing { DUP } and the stack displays the list. DUP is 
simply
a command in the internal library number 2.
ASSEMBLE
HashTab CON(5)  =DOHSTR
        REL(5)  EndHashTab
        REL(5)  Triplets1               
        REL(5)  Triplets2                 0, if no names of the given size
        ...                               exist
        REL(5)  Triplets16              /
        REL(5)  backlinks               offset past the triplets
        CON(2)  characters              
        NIBASC  'name'                    triplets in no particular order
        CON(3)  command number          /
        ...
backlinks
        CON(5)  (*)-(triplet1)          *postive* backward offset to
                                        the triplet for command number 1
        ...                             a zero indicates no name for 
command
        CON(5)  (*)-(tripletN)          last command which has a name
                                        (or we could also have zero links)
EndHashTab
RPL
The library menu displayer assumes that all user level commands have lower
command numbers than the smallest command number for the internal
subroutines. Since internal subroutines are not expected to have names in 
the
name table, the menu displayer will assume it is supposed to display names
from the hash table until no name is found. This can mean the end of the 
hash
table, or a zero link in the backward links. Some programs utilize this to
give 'hidden' commands by putting a dummy command with no name at a 
specific
position in the link table (e.g. with a given command number). Any user 
level
command with a command number higher than the dummy's will not be displayed
in the menu, but since the compiler uses only the forward link part it will
find the names of these hidden commands, and since the decompiler jumps
directly to the indexed backward link it too will find the names of the
hidden commands. The difference comes from trying to find a specific 
triplet
vs. traversing the triplets in the order of command numbers.
Aside from the above tables, the main body usually contains only valid
objects, although any data can be put in the main body.
A program in the main body is considered a user level command if
   o It is a pointee in the link table
   o It has a triplet in the hash table
   o It has property flags and the main body of the respective ROMPTR
     stored in front of it.
   o Is followed by pointers or objects implied by the preceding
     property flags.
If any of above is not true, any attempt to interpret the program as a
user level command may result in unexpected behaviour.
Storing the body of the respective ROMPTR in front of the program serves to
give quick access to the respective library and command numbers (as opposed
to searching all library link tables for a match) so that the expected 
CK<n>
type command at the start of the program can save the library and command
numbers as the 'current command'. This information is then used to 
construct
informative error messages (the name of the command causing the error).
The property flags in front of the 6-nibble header indicate, as expected,
what properties the command has. This allows easy and flexible addition of
new commands without having to reprogram any function using these 
properties.
The property flags and the respective objects expected after the body of 
the
command are explained in more detail in the entries.srt document. A simple
example is that a library with a command number less than 1792 (#700h) and
with a property flag field 8 is interpreted to be a regular command with no
properties, and with a 3-nibble zero property field to be a regular 
function
rarely used (alternative IF-THEN-ELSE-END constructs etc).
/-- copy end -----/
For more info on the 'property field' see the entries.srt file, also by 
Mika.
If YOU are able to sort it all out, please tell us!  I never got further 
than
the standard CON(1) 8 property field.
--
Werner Huysegoms
remove the x before replying
http://www.dejanews.com/       Search, Read, Discuss, or Start Your Own    
----------------------------------------------------------------------
I need a new calculator. I was considering the TI89, or one of the HP48s. I
need something that will express equations symbolically, and in terms of
variables. I also need it to be as user friendly as possible. The big
problem is that I learned on a TI. When I was playing around with one of
the HPs today I couldn't even figure out how to do some of the most basic
things. Any opinions and or sugestios would be welcom.
                                      -Chris
----------------------------------------------------------------------
====
> I need a new calculator. I was considering the TI89, or one of the HP48s. 
I
> need something that will express equations symbolically, and in terms of
> variables. I also need it to be as user friendly as possible. The big
> problem is that I learned on a TI. When I was playing around with one of
> the HPs today I couldn't even figure out how to do some of the most basic
> things. Any opinions and or sugestios would be welcom.
>                               -Chris
>
You learned on the TI that's way it's easier. I also learned on a algebraic
type calculator, but now sense I use the hp a lot more I found that I
couldn't ever go back to the algebraic type. I now think that it's easier 
to
use RPN. Not just in the graphic calculators either, I recently purchased 
the
hp32sII because it has RPN instead of the TI's algebraic method. As far as
symbolics, with the programs that I downloaded from the internet, my hpgx 
is
now just as powerful if not more powerful than the TI92. I know once you
start using the hp you will find that it is the best way to calcute any
number. When I play around with the TI's I have trouble, not with the hp's.
The hp's also have big following on the internet, where you can get free
software and a lot of help, this group for one. When ever I have a problem
using my hp I can always get an answer right here.
http://www.dejanews.com/       Search, Read, Discuss, or Start Your Own    
----------------------------------------------------------------------
Sender: khuber@isis.visi.com
Hal> I need a new calculator. I was considering the TI89, or one of
Hal> the HP48s. I need something that will express equations
Hal> symbolically, and in terms of variables. I also need it to be as
Hal> user friendly as possible. The big problem is that I learned on a
Hal> TI. When I was playing around with one of the HPs today I
Hal> couldn't even figure out how to do some of the most basic
Hal> things. Any opinions and or sugestios would be welcom.  -Chris
I think you should go for the TI-89.  Especially if user friendly is
important to you -- a HP will have a steep learning curve.
HPs are great for engineering type numerical calculations, and if you
get Erable, EqStack/Java or some of the other addins they are pretty
competitive for symbolics.  However, easy to use they ain't.  If I
haven't done a particular type of calculation for a while, I find
myself consulting the manual.
With the TI-92/TI-89 history mechanism, infix notation suddenly
becomes more useful than RPN in my opinion.  With HP's RPN it is
impossible to recall past operations because only results are on the
stack.  Obviously you can define functions or write programs.
For example, on my TI-92 I can have
        3*x^2+2*y
        3*x^3+2*y^4
Then I decide that I want to change the first equation to 3*x^2+2*z.
On the TI I can just cursor up in the history to recall that
calculation.  It is all very natural on the TI because the TI always
uses infix.
There is no way to do this on the HP using RPN because the stack only
holds results: 3 x SQR * 2 y * + becomes a numeric value if x and y
are defined and the SQR * and + operations are lost.  Yes you can
leave '3*x^2+2*y' on the HP stack, but infix notation is extra work
and you'd have to remember to DUP all the time before you eval.  The
advantages of RPN are lost when you use quoted infix -- when you do
symbolic work on the HP you generally use infix directly.  
If you need to do sophisticated symbolic work, I would recommend
purchasing Mathematica which makes both the TI and HP look quite
primitive.
-Kevin
----------------------------------------------------------------------
a good way could be working through some of the examples given
in the manuals of the HP48, so you may get a feeling for the
user interface and how to use it.
Another source of information could be the FAQ by A. Schoorl.
You will see that, after a short period, it's very easy to use;-)
It's different from other user interfces for calculators, but it's
worth giving a try.
Greetings
Raymond
----------------------------------------------------------------------
====
The following elisp performs the following tasks:
  1. Removes asm-mode from auto-mode-alist. This because it is traditional
     to use .s for SRPL source files.
  2. Defines a new interactive function called fyslab-font-lock.
     The command merely redefines various font-lock attributes, most
     importantly the colours. However, the new colours work best
     on pale backgrounds. IMHO the default font-lock colours suck.
  3. Defines saturn-mode, which defines the font-lock regexps for
     system RPL and Saturn assembly.
  4. Adds saturn-mode to auto-mode-alist.
To use the elisp simply add it to the end of your .emacs (or _emacs).
Note that I personally prefer to use the colours in all modes, mainly
because the colours actually don't suck in all modes anymore.
There are some obvious problems with the regular expressions, most notably
the colorization is used in comments too. There are also some problems
with precedence. However, I do not have sufficient knowledge of either
elisp or regexps to know how to fix either. If somebody does know how
to fix the problems, please post the fixes for others to enjoy.
Summary of colorization:
Keywords:
        #1+_ONE_DO DUP#0_DO toLEN_DO ZERO_DO ONE_DO DO
        NOT_WHILE DUP#0<>WHILE WHILE NOT_UNTIL #0=UNTIL UNTIL
        BEGIN AGAIN REPEAT
        DROPLOOP SWAPLOOP STOPLOOP +LOOP LOOP
        ERRSET ERRTRAP
        CODE ENDCODE ASSEMBLE RPL
        DOBIND DUP1LAMBIND 1LAMBIND
        1GETABND ABND
        CACHE DUMP 1ABNDSWAP
        CK&DISPATCH0 CK&DISPATCH1
        CK1&Dispatch ... CK5&Dispatch
Comments:
        ( comment )
        * [-=* #]+ comment               ( I do use f.ex. * ------------- 
... )
Labels etc:
        xROMID xxx
        ROMID xxx
        NAMELESS label
        DEFINE label xxx
        NAME label
        NULLNAME label
        xNAME label
        sNAME label
        tNAME label
        LABEL label
        LOCALNAME label
        LOCALLABEL label
        EXTERNAL label
        LOCAL label
Named NULLLAM delimiters:
        {{ }}
Assembly language:
        label    EQU  xxx
        label    ABS  xxx
                 ABASE     xxx
        label    ALLOC        xxx
        label    REL  xxx
        =label
                 STITLE    text
                 MESSAGE   text
                 INCLUDE   filename
                 RDSYMB    filename
                 CHARMAP   filename
                 SETFLAG   flagname
                 CLRFLAG   flagname
        label    IF
        label    IFANYCARRY
        label    IFCARRYCLR
        label    IFCARRYSET
        label    IFNDEF
        label    IFDEF
        label    IFEQ
        label    IFGE
        label    IFGT
        label    IFLE
        label    IFLT
        label    IFNE
        label    IFNEG
        label    IFNOPC
        label    IFOPC
        label    IFNZ
        label    IFPASS1
        label    IFPASS2
        label    IFPOS
        label    IFREACHED
        label    IFZER
        label    ELSE
        label    ENDIF
        label    MACRO
                 EXITM
        label    ENDM
With warning-face (bright red):
        [ignored] BUG text
        [whitespace] END
---------- CLIP ---------- CLIP ---------- CLIP ---------- CLIP ----------
; Remove asm-mode, it interferes with Saturn code
; There must be an easier way to do this..
(defvar blah auto-mode-alist)
(setq auto-mode-alist ())
(while blah
  (if (not (eq (cdr (car blah)) 'asm-mode))
     (setq auto-mode-alist (append auto-mode-alist (list (car blah)))))
  (setq blah (cdr blah)))
; ----------------------------------------------------------------------
; FYSLAB font-lock settings - suitable only for pale color backgrounds
; such as white and #F0E0C0
; ----------------------------------------------------------------------
; Note: face arguments: foreground,background,boldp,italicp,underlinep
(defun fyslab-font-lock()
  (interactive)
  (setq font-lock-maximum-size nil)
  (setq font-lock-maximum-decoration '((latex-mode 2) (t nil)))
  (setq font-lock-face-attributes
       '((font-lock-comment-face Firebrick)
         (font-lock-string-face  Sienna)
         (font-lock-keyword-face ForestGreen nil t)
         (font-lock-builtin-face Orchid)
         (font-lock-function-name-face Blue nil t)
         (font-lock-variable-name-face Purple nil t)
         (font-lock-type-face BlueViolet nil t)
         (font-lock-reference-face MediumBlue)
         (font-lock-warning-face Red nil t)))
  (turn-on-font-lock))
(fyslab-font-lock)
(global-font-lock-mode t)
; ----------------------------------------------------------------------
; Saturn editing mode
; ----------------------------------------------------------------------
(defun saturn-mode()
  (interactive)
  (setq saturn-font-lock-keywords
   '(t
     (( [^)]+ )
        (0 font-lock-comment-face))
     (^*[*+ =#-]*
        (0 font-lock-comment-face))
     ([ 
tn]((#1+_ONE|DUP#0|toLEN|ZERO|ONE)?_?DO
|BEGIN|(NOT_|DUP#0<>)?WHILE|(NOT_|#0=)?UNTI
L|AGAIN|REPEAT|(DROP|SWAP|STOP|+)?LOOP
|ERR(SET|TRAP))>
        (0 font-lock-keyword-face nil))
     (<((END)?CODE|ASSEMBLE|RPL)>
        (0 font-lock-keyword-face nil))
     
(<((DO|(DUP)?1LAM)?BIND|(1GET)?ABND
|CACHE|DUMP|1ABNDSWAP)>
        (0 font-lock-keyword-face t))
     (({{|}})
        (0 font-lock-keyword-face))
     
(<(CK&DISPATCH(0|1)|CK(0|1|2|3|
4|5)&Dispatch)>
        (0 font-lock-keyword-face))
     (^=?([^ tn]+)[ t]+(=[ 
tn]+|(EQU|=|AB(S|ASE)|ALLOC|REL)
>)
        (1 font-lock-variable-name-face nil)
        (2 font-lock-keyword-face nil))
     (^=([^ tn]+)
        (0 font-lock-function-name-face nil))
     (<(S?TITLE|MESSAGE)>[ t]*([^n]*)
        (1 font-lock-keyword-face nil)
        (2 font-lock-string-face nil))
     
(<(INCLUDE|RDSYMB|CHARMAP|SETFLAG|CLRFLAG)>
[ t]*([^ tn]*)
        (1 font-lock-keyword-face nil)
        (2 font-lock-variable-name-face nil))
     (<(x?ROMID)>[ t]*([^ tn]*)
        (1 font-lock-keyword-face nil)
        (2 font-lock-constant-face nil))
     
(<(NAMELESS|DEFINE|NAME|NULLNAME|xNAME|sNAME
|tNAME|LABEL|LOCALNAME|LOCALLABEL|EXTERNAL|LOCAL|DEF
INE)>[ t]*([^ tn]+)?
        (1 font-lock-keyword-face nil)
        (2 font-lock-function-name-face nil))
     (^([^ tn]*)[ 
t]+(IF(ANYCARRY|CARRY(CLR|SET)|N?DEF|EQ
|(G|L)(E|T)|NEG?|N?OPC|NZ|PASS(1
|2)|POS|REACHED|ZER)?|ELSE|ENDIF)>
        (1 font-lock-function-name-face nil)
        (2 font-lock-keyword-face nil))
     (^([^ tn]+)[ 
t]*(MACRO|ENDM|EXITM)>
        (1 font-lock-function-name-face nil)
        (2 font-lock-keyword-face nil))
     (<BUG[^n]*
        (0 font-lock-warning-face))
     (^[ t]+END>
        (0 font-lock-warning-face))
    ))
  (setq font-lock-defaults '(saturn-font-lock-keywords nil t))
  (font-lock-mode)
)
(setq auto-mode-alist (append (list '(.s' . saturn-mode)) 
auto-mode-alist))
---------- CLIP ---------- CLIP ---------- CLIP ---------- CLIP ----------
-- 
---
    --> Mika Heiskanen   mheiskan@gamma.hut.fi   
http://www.hut.fi/~mheiskan
----------------------------------------------------------------------
====
ok, the problem is the lack of use of makerom. I can not send attachments
confirmed with DOS batch files, so i made a dirty fix to work. mass 
generates
__tmp.a and __build.mn everytime it runs, so you only have to change mass.
Here you have to change three files, if you make changes in version an
library title.
----mass.bat----
rplcomp jazz __jazz.a __jazz.ext
makerom __build.mn __build.m
sasm -EH __tmp
mv __tmp.l __listing
mv __tmp.o lib
lbcrc lib
----end of mass.bat
----__tmp.a-------
        NIBASC   /HPHP48-X/
        INCLUDE  SupRomEntr.a
        INCLUDE  __head.a
        SETLIST  INCLUDE
        INCLUDE  __jazz.a
        CLRLIST INCLUDE
        INCLUDE  __hash.a
        INCLUDE  __end.a
-------end of __tmp.a------
-------__build.mn-------
TITLE Jazz Library
OUTPUT __xx.o
LLIST __xx.lr
CONFIGURE xxCONFIG
MESSAGE xxMESSAGE
NAME JAZZ V6.6 by Fin 25.05.98
ROMPHEAD __head.a
REL __jazz.o
TABLE __hash.a
FINISH __end.a
END
---------end of __build.mn------
>I would like to assemble a middle-weight version of Jazz 6.6,
>which would contain all the normal Jazz 6.6 commands except for
>the entry table manipulation commands, namely:
>
>EA
>RTAB
>DTAB
>RTB->
>->RTB
>->DTB
>
>So I downloaded the Jazz 6.6 sources zip from Mika's page, with
>the intent to assemble my own version using the GNU Tools 2.1.9
>on my Win95 PC.
>
>Here's where things get strange: with _no_ changes to any of the
>source files, I attempted to assemble Jazz 6.6 (as a test run; if
>everything was working properly, the output file should be an exact
>duplicate of the Jazz library file).
>
>The output file was slightly different in size (by only 609 bytes),
>but more importantly it caused my calc to lock into the WS-splash
>(i.e., endless warmstart cycle), so obviously I did something wrong!
>
>Here's what I did, exactly:
>
>1. Unzip the jsrc66.zip file to the directory C:Program Filesgtools
>   [ The subdirectory structure of the .zip file was preserved ]
>2. Using Joe Horn's make.bat file (which has always worked w/ the
>   HPTOOLS & GNU Tools in the past), I did:
>
>   C:Program Filesgtoolssrc> make jazz
>
>3. Wait for about 2 minutes; output file was called jazz. There were
>   no warnings or error messages during the process.
>
>Here is the make.bat file, by Joe Horn (thanks, Joe!)
>
>if not exist %1.s exit
>rplcomp %1.s %1.a
>if exist %1 del %1
>sasm -N %1
>echo SE ENTRIES.o > %1.m
>echo RE %1.o >> %1.m
>sload -H -N -o %1 %1.m
>for %%f in (%1.a %1.m %1.o) do del %%f
>
>So, what am I doing wrong? It must be some simple switch or omission
>on my part; the files are very close in size on PC:
>
>jazz.lib (As distributed in jazzv66.zip)        71,946
> jazz     (As assembled by me :(                 71,337
>
> TIA,
> katz
> --
> modified to thwart spammers. To reply, please send message to:
> jmkatz-72 at worldnet dot att dot net ( at=@ dot=. )
http://www.dejanews.com/       Search, Read, Discuss, or Start Your Own    
----------------------------------------------------------------------
====
       I have inherited a hp 48sx from my brother. Unfortunately with no
manual. Is there any other place on the net where a manual exists or where
the basic functions of this machine can be found?
----------------------------------------------------------------------
====
Gracias, Mario...
En nombre de todos los usuarios de HP hispanoparlantes, un mill.97n de 
gracias
por haber dedicado tanto de tu tiempo a compartir con nosotros tus
conocimientos sobre programaci.97n, desinteresadamente.
Eres un pionero, Mario. Dios te bendiga.
http://www.dejanews.com/       Search, Read, Discuss, or Start Your Own    
----------------------------------------------------------------------
====
My boss is in need of some hp42s calculators and, as you all wil know,
they aren't made anymore. My question is, does anyone of you accidently
have a redundant hp42s for sale? (for about the price of a hp32s)
----------------------------------------------------------------------
Joris asks:
...ion is, does anyone of you accidently
have a redundant hp42s for sale? (for about the price of a hp32s)
--------------:: o
right...like there's ANYONE out there that's thinking...
'gee, i wish i could find someone to unload this loathsome HP42 on...i'd be
willing to just GIVE it to someone for a dollar or two...'
???
!!!
----------- :: o
.---..-..-..-..-..-..-.  .  .-. .-.  .-..-..-..   .-.  
`| |'| || .` | >  /   / / / /    | .` || | ) /    
 `-' `-'`-'`-' `-'    `. ^ .' `--^--'`-'`-'`-'' `--^--'
Join The BabyNous Cult : The Friendly Neighborhood Cult
----------------------------------------------------------------------
====
In message <19981202171438.11078.00001369@ng25.aol.com>
> Joris asks:
> ...ion is, does anyone of you accidently
> have a redundant hp42s for sale? (for about the price of a hp32s)
> --------------:: o
> 
> right...like there's ANYONE out there that's thinking...
> 'gee, i wish i could find someone to unload this loathsome HP42 on...i'd 
be
> willing to just GIVE it to someone for a dollar or two...'
The problem is to find somebody who has this type of calculator and my
boss i willing to pay for it, since he can earn money using them. I
think there could be people around owning a machine like it and not
using it, but not eager to go through the hassle of sending it to a
complete stranger. That's why the earnable money could be a usefull
stimulance.
----------------------------------------------------------------------
====
--------------B5015F3DDE257AE2A6A20313
> My boss is in need of some hp42s calculators and, as you all wil know,
> they aren't made anymore. My question is, does anyone of you accidently
> have a redundant hp42s for sale? (for about the price of a hp32s)
 They are very hard to find today.  They normally sell for over $200 if
you can find them.  I recently had a few extra ones which I sold for $200
each and I could have sold dozens more if only I had them.
You might try ebay - they sometimes go on auction there.
Jerry
--
=====================================================================
=  Jerry Petrey - Consultant Software Engineer  - Member Team Ada   =
=                 Lockheed Martin                 Member Team Forth =
=====================================================================
--------------B5015F3DDE257AE2A6A20313
<HTML>
<BLOCKQUOTE TYPE=CITE>My boss is in need of some hp42s calculators and,
as you all wil know,
<BR>they aren't made anymore. My question is, does anyone of you accidently
<BR>have a redundant hp42s for sale? (for about the price of a 
hp32s)</BLOCKQUOTE>
&nbsp;They are very hard to find today.&nbsp; They normally sell for over
$200 if
<BR>you can find them.&nbsp; I recently had a few extra ones which I sold
for $200
<BR>each and I could have sold dozens more if only I had them.
<BR>You might try ebay - they sometimes go on auction there.
<P>Jerry
<PRE>--&nbsp;
=====================================================================
=&nbsp; Jerry Petrey - Consultant Software Engineer&nbsp; - Member Team 
Ada&nbsp;&nbsp; =
=&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nb
sp;&nbsp;&nbsp;&nbsp; Lockheed 
Martin&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp
;&nbsp;&nbsp;&nbsp;&nbsp; Member Team Forth =
=====================================================================</PRE>
&nbsp;</HTML>
--------------B5015F3DDE257AE2A6A20313--
----------------------------------------------------------------------
Bought a 48G over the weekend.
too complicated for me.
I wanted to return it but I lost the manual.
I was able to get a photocopied version.
Never used so Brand new.
looking for approx $90
bought it for $120
--
Enrico Ng <ng@usa.net>
----------------------------------------------------------------------
====
I think i have an extra manual for Hp48  but it's in Spanish.....in the
other hand i cand change it to a friend to get an english version of it...
if you're still interested on return it, and want the manual write-me
 > Bought a 48G over the weekend.
> too complicated for me.
> I wanted to return it but I lost the manual.
> I was able to get a photocopied version.
> Never used so Brand new.
> 
> looking for approx $90
> 
> bought it for $120
> 
> --
> Enrico Ng <ng@usa.net>
> 
> 
> 
> 
> 
                 _______     ________    ________     __    __     
                /______   /_______  /_______   /_  /_   

               / / ___  /  / / ____  / / / ______/  / / /_/ / /  
              / / /_/ /  / / / / / / / / /_____  / / /__/ /   
             / / _____/  / / /_/ / /  /_____   / / / ____  /     
            / / /       / / /__/ /  /____/  / / / / / / /   
            /_/        /______ /   /_______/  /_/  /_/
         
       _______________________________________________________   
      /______________________________________________________
     / /                                                      /
   / / * IO.SYS file not found. (A)bort,(R)etiry,(P)anic.   /
  / / * Pilot virtual:   EC-LCV                            /
  /______________________________________________________/
----------------------------------------------------------------------
>Bought a 48G over the weekend.
>too complicated for me.
>I wanted to return it but I lost the manual.
>I was able to get a photocopied version.
>Never used so Brand new.
>
>looking for approx $90
>
>bought it for $120
You lost a huge manual in a weekend?
But you were able to find one to photocopy?
$90? -- sells new around here for $99.
----------------------------------------------------------------------
====
Where I am, you can buy a full-warranty 48G, with new manual, for around
$80.
>>Bought a 48G over the weekend.
>>too complicated for me.
>>I wanted to return it but I lost the manual.
>>I was able to get a photocopied version.
>>Never used so Brand new.
>>
>>looking for approx $90
>>
>>bought it for $120
>
>You lost a huge manual in a weekend?
>But you were able to find one to photocopy?
>$90? -- sells new around here for $99.
----------------------------------------------------------------------
====
huge manual?
its only about 1cm thick from what i can remeber.
I photocopied my friends manual.
and I said approx $90 which means within $10.
it is a brand new calculator anyways.
--
--
Enrico Ng <ng@usa.net>
WVHS Internet
http://www.ipsd.org/wvhs
>>Bought a 48G over the weekend.
>>too complicated for me.
>>I wanted to return it but I lost the manual.
>>I was able to get a photocopied version.
>>Never used so Brand new.
>>
>>looking for approx $90
>>
>>bought it for $120
>
>You lost a huge manual in a weekend?
>But you were able to find one to photocopy?
>$90? -- sells new around here for $99.
----------------------------------------------------------------------
> blinking cursor (Like TED) in ML, using FNT1. I've tried this in SysRPL
but
> it's too slow because of the blink and the entry at the same time, so I
> think ML would be a better (faster) solution.
> Can someone help me ?
>What should the Code do? Show an entry line? And where?
>And what should it to after the user typed in something? Put the string
>on stack?
The code I'm thinking off is a general prupose entry line to be inserted in
SysRpl programs
it takes three arguments from the stack:
4:
3:  Text String                          ---> Entry Label
2:  X position in ABUFF
1:  Y position in ABUFF
After the value is typed (ex. real number) it will be put on the stack.
I'm trying to rewrite some useful engineering UsrRpl programs in SysRpl and
all the values are to be entered with this small entry line with FNT1, for 
a
better management of the display area. I don't know any ML to make the
routine or/and the blinking cursor. If you know any code for this routine,
Paulo Tavares
----------------------------------------------------------------------
====
> The code I'm thinking off is a general prupose entry line to be inserted
> in SysRpl programs it takes three arguments from the stack:
> 
> 4: 3:  Text String                          ---> Entry Label 2:  X
> position in ABUFF 1:  Y position in ABUFF
What is ABUFF?
> After the value is typed (ex. real number) it will be put on the stack.
The code will read data from keyboard, and according to it, push a real,
float, or any other value onto stack. It can only allow strict real of
float values too (easier to code ;-)
> I'm trying to rewrite some useful engineering UsrRpl programs in SysRpl
> and all the values are to be entered with this small entry line with 
FNT1,
> for a better management of the display area. I don't know any ML to make
> the routine or/and the blinking cursor. If you know any code for this
I don't know how to use UFL from ML, I'm going to look at it.
Doing a blinking cursor is easy. The bigger part is probably to choose
what to push onto stack according to what is typed.
-Gilb-
00 E2 22 76 58 1F 72 BA  EA 94 78 F3 A2 63 E0 94
----------------------------------------------------------------------
<1djexvz.1t9yk2516qnneaN@tntrasp18-157.abo.wanadoo.fr>...
-----Original Message-----
small fonts and allowing only real float values would be a great help.
Paulo Tavares
----------------------------------------------------------------------
====
> 
> I don't know if coding without UFL is easier. I got it with Jazz, but 
> I have no file explaining how to use UFL.. I got programs that do, but
> I don't know how my own ML ones could do.
Did you get the ACCESS.DIR ? It contains several ML programs which
use the UFL. They are small enough to be easily understood by
someone with your ML-knowledge in a matter of minutes.
The ACCESS.DIR is available along with the UFL from Andre's page.
David
-- 
David Hanon          phone : 32-2-650.55.29 
ULB - CP 231         fax   : 32-2-650.57.67
Belgium              WWW   : http://poseidon.ulb.ac.be/groupe/dh/
----------------------------------------------------------------------
====
> Did you get the ACCESS.DIR ? It contains several ML programs which
> use the UFL. They are small enough to be easily understood by
> someone with your ML-knowledge in a matter of minutes.
> 
> The ACCESS.DIR is available along with the UFL from Andre's page.
The UFL v1.02 can be found here:
http://www.engr.uvic.ca/~aschoorl/ufl/
HTH,
katz
-- 
-----------+
----------------------------------------------------------------------
I have put a new, unused HP-12C Financial Calculator up for auction at ebay. 

The minimum bid is $5.00, and there is no reserve.  If interested, you may
visit the auction at:
http://cgi.ebay.com/aw-cgi/eBayISAPI.dll?ViewItem&item=46013029
+-------------------------+
 Stephen J Thomas
 SJThomasOD@aol.com
----------------------------------------------------------------------
Stupid pet tricks with RAND (but will they fool The Great Randi ?)
To warm up, before the main act, we might review how to
give the RAND function a random spin, so that it
can be started off in a randomized place, instead of
always producing the same sequence after the calc memory
has been cleared (as in ON+A+F+Yes, but not after mere warmstarts):
HP recommends 0 RDZ to set a time-dependent starting value for RAND;
however, this then only uses the low-order 20 bits of TICKS, converted
to an integer, as its actual argument.  Said value not only re-cycles
itself every two minutes or so, but also tends to leave the last six
digits of the first subsequent RAND output almost always the same,
which I think would make Bob Wheeler very angry :)
With proper choice of an RDZ argument, however, you can create up to
1E14 different possible starting values for RAND, so you can do
much better than 0 RDZ if you want a much more varied time-dependent
starting value; the following are based upon the fact that for
arguments > 1, RDZ utilizes all 12 mantissa digits and the final
two digits of the exponent (power of 10) of the argument:
Suggestions for an RDZ0 function (to replace 0 RDZ):
Steve VanDevender suggested using TICKS mod 1E14:
<< RCLF 8 SQ STWS TICKS 2 ALOG DUP2 / DUP 4 ROLLD * - B->R ALOG
    SWAP 6 ALOG SQ DUP2 / * - B->R * RDZ STOF >>
I use a highly non-linear generator of 14 digits:
<< DATE FACT TIME EXP * 0 RDZ RAND 2 ALOG * ALOG * RDZ >>
The 100-year internal clock is only about 2.5E13 ticks, so a 14-digit
seed generator can never go through all possible starting values.
Since the internal clock is updated 8192 times per second,
there is also no way to obtain the same result twice,
even if you repeat RDZ0 as rapidly as possible, in a running program.
Since RAND always maintains an odd number in its internal
15-digit mantissa value, RAND can never produce a result
exactly equal to zero; truncation (rather than rounding) to a
12-digit result also ensures that RAND never outputs exactly 1.
The smallest and largest possible outputs from RAND
are therefore 1E-15 and .999999999999
In case you think that you will never live to see the rare event
of catching these values appearing as output, however, try this:
5.13684117130E47   RDZ  RAND RAND RAND RAND ==> .000000000000001
4.39513938817E18   RDZ  RAND RAND           ==> .999999999999999
Another peculiarity of RAND is that once you have generated some
random outputs, there seems to be no way to go backwards
and replay them, to produce the same effects once again.
But have no fear, The Great Randi was here, and left us a
reverse RAND function, below, which runs RAND backwards
from any given point, to wherever you want it to back up to,
after which it will once again play out the exact same hand
when going forwards with RAND again, as it had done earlier.
The HP15C calculator had one more random-number function than does
the HP48, and that function was recall current random number,
which delivered again the *same* output value as did the
most recent RAND.  One of the purposes of this function was
to let you *save* the exact current position in the output
sequence of RAND, so that you could *interrupt* some process or
simulation, say, then use RAND for something else for a while,
and then later resume it for the interrupted program, exactly
where it had left off, by supplying the saved value to the HP15C's
store random number function, which was comparable to the
HP48's RDZ (set random number seed), except that there was
no clock in the HP15C, so that you could not randomly start
the random number generator by using the date and time, other than
by looking at your own clock, and digitizing its output manually.
It was possible for the HP15C to resume an interrupted RAND sequence
exactly where it was, because the internally saved random number
mantissa was the same 10-digit length as was available in all
user-accessible storage registers and the stack; the HP48
does *not* permit you to do this, however, because the
internally saved mantissa is 15 digits long, but it gets
truncated to 12 digits when RAND outputs a value, and then
once again, there are only 12 user-specified digits in
real numbers that you can offer as an argument to RDZ;
to add one more coup de grace to all this, RDZ also always
forces the 15th internal mantissa digit to be a 1,
no matter what else you might have liked it to be.
However, new functions are supplied below for the HP48 which can
both recall and restore the current internal RAND value to its full
15-digit precision, using a Long Real (extended precision) object,
which happens to contain a 15-digit mantissa, so now you can do
on an HP48 the same things that you used to be able to do
many years earlier, on the good old sturdy HP15C, which was the first
HP calc with complex-valued functions and matrices, to accompany
the already old hat numeric Solve and Integration functions.
For you young folk, that model was made back when there were
all-metal keyboards which didn't wear out, because quality
electronic devices were not made as throw-aways.
Come to think of it, that was only about fifteen or so years ago,
which demonstrates how technology has so quickly progressed, to
make things which now are already obsolete or dead when shipped :)
So, with much ado about nothing much,
here are the HP48 functions for reverse RAND
and recall/store current random number
(plus an extra free bonus function, not even advertised above):
Oh, I should first mention that since these are written in ML
by an idiot, using unsupported entry points and a wild guess
that they ought to work as well on an S/SX, you might want to
back up your memory before trying them (the Federal Trade Commission
requires me to make these statements, which would be unnecessary
if only I were selling tobacco or genetically engineered foods :)
@ RANPREV:        ==> previous_rand_value               S/SX/G/GX
CCD20E80008FB9760051B2CFA216F14613615AEAC097C713F335380333005999
0AFAAF1AFC3F30832198329935906600A70A0E59FBF10C54F158EAFC2ECC90DA
0CCB9155FAD48DD32A25CFE @ 73.5 bytes, #EFC5h
@ RANRCL:         ==> Long Real [unnormalized]        S/SX/G/GX
CCD20640008FB97601B2CFA216F14613615AEAC0AF805D0CC959908D532A2344
03008D04F01F4B5 @ 37.5 bytes, #5B4Fh
@ RANSTO:        Long Real [unnormalized] ==>         S/SX/G/GX
D9D20D29512BF810DB46CCD20130008F060A2AF4808501B2CFA216F146136158
E048D34150B2130C093 @ 39.5 bytes, #390Ch
@ RANXXX:  Mystery bonus function (run it, and then try RAND)
@ No arguments, no results, 29.5 bytes, #9019h
CCD20630008FB97601B2CFA216F146136AF0808205814814158E8D341509109
The above programs are in ASC format; see Goodies Disk #1
at <www.hpcalc.org>, or see <http://www.engr.uvic.ca/~aschoorl/faq/>,
or TASC.EXE on Goodies Disk #8 to convert on PC, instead of in HP48.
Since the clueless Food and Drug Administration requires disclosure
of all ingredients (except, of course, for genetically engineered
organisms, which are not present herein),
here is the source code for the above objects
(you can skip this if you normally don't read labels :)
* RANPREV:   Get *previous* RAND value
* (identical to normal RAND except for the inverse multiplier)
CODE
        GOSBVL   =SAVPTR
        SETDEC
        D0=(5)   =%RAN ( find RANDOMSEED address from %RAN function )
        D0=D0+   16
        C=DAT0   A ( #80822 for G[X], #706A4 for S[X] )
        CD0EX
        A=DAT0   15
        A=0      S
        ?A#0     W
        GOYES    nz
        LCHEX    0999500333083533
        A=C      W
nz      B=0    W
        ABEX     W
        LCHEX    0953992389123803 ( originally 0002851130928467 )
        GOTO     lp2
lp1     A=A+B W
lp2     C=C-1 P
        GONC     lp1
        BSL      W
        P=P+1
        GONC     lp2
        DAT0=A   15
        ABEX     W
        P=       14
        A=A-1    A
lp3     ?B#0  P
        GOYES    end
        A=A-1    A
        BSL      WP
        GONC     lp3
end     A=B   M
        GOVLNG   =PUSH%LOOP
ENDCODE
* RANRCL:  Recall current random number as unnormalized Long Real
CODE
=PUSH%%LOOP     =     #2A235 ( unsupported, appears stable )
        GOSBVL   =SAVPTR
        D0=(5)   =%RAN
        D0=D0+   16
        C=DAT0   A
        CD0EX
        A=DAT0   15
        A=0      S
        B=A      W
        SETDEC
        A=0      A
        A=A-1    A
        ?B=0     M (detect the situation mentioned below)
        GOYES    nearzer
        GOVLNG   =PUSH%%LOOP
* PUSH%%LOOP in turn calls PUSH%% (2A24B), which tests for
* a non-zero *long* mantissa in B via ?B#0 M, rather than ?B#0 W;
* this misses testing the last 3 of the 15 digits; therefore
* unnormalized values less than 1E-12 will be considered zero!
* (could be fixed by normalizing/denormalizing, but I don't bother,
* since the probability of such a too low value is too low anyway :)
nearzer LC(5)     #304 ( Undefined Result if zero or *almost* zero)
        GOVLNG   =GPErrjmpC
ENDCODE
* RANSTO:  Restore current random number from unnormalized Long Real
* (presumably produced by the complementary function above)
::
CK1NoBlame
CK&DISPATCH1
SIXTYTHREE ( 3F, Long Real )
CODE
=POP1%%   =  #2A060 ( unsupported, appears stable )
        GOSBVL   =POP1%% ( includes SAVPTR, does SETDEC )
        A=B      W ( mantissa only, ignore exponent and sign )
        ABIT=1   0 ( reduce the effects of any inappropriate inputs )
        D0=(5)   =%RAN
        D0=D0+   16
        C=DAT0   A
        CD0EX
        DAT0=A   15
        SETHEX
        GOVLNG   =GETPTRLOOP
ENDCODE
;
* RANXXX:  Run this, and then observe the results of RAND
* ( after which you may lose faith in statistics, if you ever had any :)
CODE
        GOSBVL   =SAVPTR
        D0=(5)   =%RAN
        D0=D0+   16
        C=DAT0   A
        CD0EX
        A=0      W
        LAHEX    5
        ASRC
        ASRC
        DAT0=A   15
        GOVLNG   =GETPTRLOOP
ENDCODE
* The above function is solemnly dedicated to Dr. Albert Einstein,
* who asserted:  God does not play dice with the universe
Finally, quoting the celebrated philanthropist P. T. Barnum:
This way to the Egress:
-----------------------------------------------------------
With best wishes from:   John H Meyers   <jhmeyers@mum.edu>
----------------------------------------------------------------------
====
====================================================
HPShell 3.11    Integrated Developing Environment
 FREEWARE           for all HP48 on DOS PCs.
====================================================
* You program your HP48 and have never seen a real developing environment ?
* You look for an easy tool for file transfer ?
* You need someone managing your backups ?
* You would like somebody's doing all the stupid char conversion stuff ?
* And you want it all in one ?
Well, have a look on the HPShell !
After 5 years of continuous development it includes now a huge list of
features one would like to have while programming the HP48 or simply using 
it.
This is a (not complete) list of it's features:
-----------------------------------------------
* Multi functional Editor with Syntax highlighting, Word Wrapping,
  HP48 Displays, ...
* File Transfer (Kermit included, ANY other protocol is also useable!)
* Backup Management
* DOS and HP48 Directory Management
* Char Conversion (ASCII <> ISO8859-1)
* Macros, Hotkeys, Toolbar
* free defineable Toolprograms
* and much more !!!
The HPShell can be found online on the Web at
  <http://hpshell.home.ml.org>
Here you'll find a more complete list of features,
screen shots and a list of download-links spread all over world.
--
Tom Wellige
http://wellige.home.ml.org
http://www.dejanews.com/       Search, Read, Discuss, or Start Your Own    
----------------------------------------------------------------------
====
        After a quite long period of silence, here it comes again :
        The HProom has been updated yesterday (01/12/98)
        You will find a huge amount of games, utilities, links, pc 
softwares...
for your HP48 (mainly G and GX version).
        The URL is : http://www.chez.com/lekeno/
        In order to obtain the english version you have to click on the 
Union
Jack button...
                 Enjoy !
        
--
               __                    _
              +    Le Keno ___     /    Baheux Kenji
                   _______/ @_  8   /   Esial Info 1A
               +===/~ ~ ~ ~___------+
                   _|__|_/            Kenji.Baheux@esial.u-nancy.fr
                     L  L           http://www.chez.com/lekeno/
----------------------------------------------------------------------
====
My english isn't quite good... and i couldn't undesrtand the IR
upgrade....
it means that if i change IR resistor it will send signals to a larger
distance?
thanks.
P.D. sorry for my last message(manual for hp48) posted to all... 
                 _______     ________    ________     __    __     
                /______   /_______  /_______   /_  /_   

               / / ___  /  / / ____  / / / ______/  / / /_/ / /  
              / / /_/ /  / / / / / / / / /_____  / / /__/ /   
             / / _____/  / / /_/ / /  /_____   / / / ____  /     
            / / /       / / /__/ /  /____/  / / / / / / /   
            /_/        /______ /   /_______/  /_/  /_/
         
       _______________________________________________________   
      /______________________________________________________
     / /                                                      /
   / / * IO.SYS file not found. (A)bort,(R)etiry,(P)anic.   /
  / / * Pilot virtual:   EC-LCV                            /
  /______________________________________________________/
----------------------------------------------------------------------
====
On Wed, 2 Dec 1998 13:17:07 +0000, Posh - Miquel <is05140@salleurl.edu>
: My english isn't quite good... and i couldn't undesrtand the IR
: upgrade....
: it means that if i change IR resistor it will send signals to a larger
: distance?
No, it will receive IR signals from a larger distance.
If you take a look at the remote control programs, you'll find that the
sending range is more than 3 meters without any modification.
The reason for the short range of data transfer between two HPs are the
receivers. They are quite limited to reduce power consumption and
susceptibility to foreign signals (e.g. from a remote control).
The upgrade does make it a bit better, but for most users it's not worth
the effort. A) taking apart a HP is a major undertaking, B) the power
consumption is increased and C) it only really works if two upgraded HPs
talk to each other.
Pete
----------------------------------------------------------------------
Where could I find the assembly manual for my old  28s in internet?
(I've lost the string to create the {System Object} to accelerate my
calculator).
-- 
----------------------------------------------------------------------
pleese take a look at http://www.hpcalc.org
There are subpages related to the HP28.
Did you mean the 'Customize Your HP28' ?
Greetings
Raymond
----------------------------------------------------------------------
====
I'm searching for a program that compiles sources in c to HP48g machine
code language.
----------------------------------------------------------------------
I would like the HP48 to be able to generate matrices containing
numbers and variables eg {{0,1,n,m},{0,1,c,0}}.  I then need to be able
to multily this by a vector eg {0,0,1,1} for instance.
The calculator won't allow non-numerical matrices.
Which program will allow symbolic matrix maths?
Duncan
cochran@sjh.bi.umist.ac.uk 
----------------------------------------------------------------------
====
>
> I would like the HP48 to be able to generate matrices containing
> numbers and variables eg {{0,1,n,m},{0,1,c,0}}.  I then need to be able
> to multily this by a vector eg {0,0,1,1} for instance.
>
> The calculator won't allow non-numerical matrices.
>
> Which program will allow symbolic matrix maths?
>
>
> Duncan
> cochran@sjh.bi.umist.ac.uk
>
Duncan: Both the ALG48 and Erable programs have functions for symbolic
matrices. You will need the supporting programs UFL, AI048,and EQSTK to 
view
the matrices computed by ALG48 and Erable. You can download all of these
programs and their supporting documentation from the HP48 archive kept by
Eric Rechlin at: www.hpcalc.org/  - hope this helps, J.C. Randerson
http://www.dejanews.com/       Search, Read, Discuss, or Start Your Own    
----------------------------------------------------------------------
====
> I would like the HP48 to be able to generate matrices containing
> numbers and variables eg {{0,1,n,m},{0,1,c,0}}.  I then need to be able
> to multily this by a vector eg {0,0,1,1} for instance.
> 
> Which program will allow symbolic matrix maths?
ALG48 does... and it does a LOT more...
-- 
Erwann ABALEA
eabalea@certplus.com
----------------------------------------------------------------------
Does EMU48 could talk to a HPvia the serial port. I would like to use it to
transfer files and to play
----------------------------------------------------------------------
====
> Does EMU48 could talk to a HPvia the serial port. I would like to use it 
to
> transfer files and to play
>
>
Yes, use have to use one of the Unofficial Service Packs (included since 
SP3).
You'll find the actual version on my homepage at
http://privat.swol.de/ChristophGiesselink/.
Christoph
http://www.dejanews.com/       Search, Read, Discuss, or Start Your Own    
----------------------------------------------------------------------
Does EMU48 could talk to a HPvia the serial port. I would like to use it to
transfer files and to pay
----------------------------------------------------------------------
====
I'm writing a System RPL aplet for the HP38G which involves plotting
graphs. Does anyone know if its possible to use to ROM code behind the
'ZOOM' key which is available in the plot views of all the standard
aplets.
Any other info on re-using Rom code would be greatly appreciated because
it reduces program size and avoids re-inventing the wheel.
----------------------------------------------------------------------
====
hi
what does this error mean??
ERROR: unable to resolve symbol '_printf',
       referenced by object-file '/tmp/cca008321.o'
linkage failed due to unresolved references
I try to use the c-compiler for linux to create programms for my
hp48G
thanks
mathias
--
      Windows '95: how do you want to crash today ???
                 LINUX - join the revolution
 mathias kuhn                    phone:  +41-56-622-82-53
 CH-5612 villmergen              icq#:   22644419
----------------------------------------------------------------------
I've made a program for the HP48 but I can't find an alternative to the 
HALT
command which interrupts the execution of a program. I would like to do the
same but with other command like END or STOP from Basic.
----
Tito Nuno Alves Santos
----------------------------------------------------------------------
> I've made a program for the HP48 but I can't find an alternative
> to the HALT command which interrupts the execution of a program.
> I would like to do the same but with other command
> like END or STOP from Basic.
If you mean that you would like a command to exit from anywhere in
the middle of a UserRPL program, without having to follow structured
programming requirements (where programs only exit at the very end):
Either a KILL or CONT command will terminate a UserRPL program;
however, this also terminates any higher-level program which might
have contained (or called) the program which issued KILL or CONT.
A slightly more elegant (and useful) solution is to store the
following program in HOME, perhaps in a variable called 'EXIT'
You can then use the EXIT command in your UserRPL programs, and
only the specific user program which calls EXIT will be terminated.
This program is listed in SysRPL, then a downloadable binary
is given in ASC format (see the FAQ or Goodies Disk #1
or TASC.EXE on Goodies Disk #8 for the use of ASC format).
* EXIT from the current enclosing *UserRPL* program
* (without terminating any higher level program)
* E.g. << 123 EXIT 999 >> ==> 123
::
  CK0NOLASTWD
  BEGIN R> R> SWAP >R ( pop one stream )
  DUP ' x>>ABND EQUALPOSCOMP #0=?SKIP ABND ( local variables ? )
* << ... -> vars << ... EXIT ... >> ... skips this also ... >>
  DUPLENCOMP NTHELCOMP ?SKIP FALSE ( last token in composite )
  { x>> AtUserStack } ( User program, or popped all the way to SOL ? )
  SWAP EQUALPOSCOMP #0<> UNTIL
* Abort one more user level, in case DO, WHILE, START, FOR, etc.
* (which have copied their containing runstream via BEGIN)
  R@ DUPLENCOMP NTHELCOMP ?SKIP FALSE ' x>> EQUAL IT RDROP
* More examples:
* << 1 100 START 123 EXIT ... NEXT >> ==> 123
* << WHILE 1 REPEAT 123 EXIT ... END >> ==> 123
* << 123 P2 456 >>, P2: << 789 EXIT ... >> ==> 123 789 456
* << << << 123 EXIT ... >> EVAL ... >> EVAL 456 >> ==> 123 456
* << 0 -> x << << << 123 EXIT ... >> EVAL ... >> EVAL 456 >> 
>>
;
* ASC format: 90 Bytes, #045Fh
D9D2051A812A170F1070F107032230F9F608813079E60EF5323A446A33367947
0132366B650A21700CA3047A20936329CB04B2130322303A4467CC308C170210
70132366B650A21700CA3079E609363279B30CB9167BF60B2130F540
You asked for it -- you got it!
-----------------------------------------------------------
With best wishes from:   John H Meyers   <jhmeyers@mum.edu>
----------------------------------------------------------------------
[snipped]
There are still some neglected loopholes in my posted EXIT program,
such as failing to do ABND if a START or FOR loop was exited;
this has no severe consequences, but I will repost a new version
whenever I think I've caught up with all the loose ends, thanks.
Save five bytes by replacing R> R> SWAP >R with RSWAP R> ...
No Exit  - Jean Paul Sartre
You can't get there from here  - Farmer Brown
-----------------------------------------------------------
With best wishes from:   John H Meyers   <jhmeyers@mum.edu>
----------------------------------------------------------------------
====
----------------------------------------------------------------------
====
I know you said firm, but considering I can buy a brand new 48gx for
$170 (with manual, of course) and with yours I would have to spend an
extra $10-15 for the manual, your $130+15=$145 seems kinda steep. I'll
give you $100 shipped for it (pre-paid) but that's the only reasonable
amount I could pay for a used calc..
Let me know,
Robert
----------------------------------------------------------------------
Ok, today i ordered the AUR and another 128k, but I cant wait for the book
to get here. This is driving me nuts, there must be something very 
intrinsic
to User RPL that I dont get. Pertaining to local variables, sometimes when 
I
write a program and use the -> name line I will recieve a syntax 
error. I
believe the -> command must be the last command in a program preceding
either an algebraic or another << program. Am I correct that you cannot
have any User RPL commands following a -> ? also, could you please pick
this one apart, i am writing this routine to take a rectangular coordinate
and return a labled HMS and distance. It is for my work, simple new
construction lot layouts.
So, the problem (1st of many im sure, as i havent fully developed anything
here yet), is that I cannot define the local variables of the truncated
angle. i want to truncate the angle so as to include them in a string with
decimals and the degree symbol. Please help with my -> usage...
%%HP: T(1)A(D)F(.);
<< Enter Coordinate
  {
  :Northing:
  :Easting :
  { 1 0 } } INPUT
  OBJ-> -> N E
 << RECT N E ->V2
   CYLIN V-> DUP ROT
   ROT ->HMS -> d a
  << IF a 0 <
    THEN 360 a HMS+
    ELSE a
    END
   << -> ams
    << ams 0 TRNC ams
      100 * 0 TRNC DUP
      100 * ams 10000 *
      SWAP - -> ddd mm ss
     << IF ddd 100 <
       THEN 0 ddd +
       END
      << d + -> hhh
       << hhh + mm + . +
         ss + :HHH.MM.SS
         + d :Feet +
       >>
      >>
     >>
    >>
   >>
  >>
 >>
>>
thanks for your help...
----------------------------------------------------------------------
====
> Ok, today i ordered the AUR and another 128k, but I cant wait for the 
book
> to get here. This is driving me nuts, there must be something very 
intrinsic
> to User RPL that I dont get. Pertaining to local variables, sometimes when 
I
> write a program and use the -> name line I will recieve a syntax 
error. I
> believe the -> command must be the last command in a program preceding
> either an algebraic or another << program. Am I correct that you cannot
> have any User RPL commands following a -> ? also, could you please pick
> this one apart, i am writing this routine to take a rectangular 
coordinate
> and return a labled HMS and distance. It is for my work, simple new
> construction lot layouts.
> 
> So, the problem (1st of many im sure, as i havent fully developed 
anything
> here yet), is that I cannot define the local variables of the truncated
> angle. i want to truncate the angle so as to include them in a string 
with
> decimals and the degree symbol. Please help with my -> usage...
> 
I send below the program you need.
You're doing at least 2 errors:
#1 the header you use (%%HP...) is misteriously wrong. See mine;
#2 you've to declare temp vars at the beginning of the program, with value 
0:
@program sample
%%HP: T(3)A(D)F(.);
<< 0 0 0 -> A B C  @ Declare 3 temp vars, with value 0. 
  << A?  INPUT
OBJ-> 'A' STO A 2 *  @ Use temps vars.
'B' STO B 3 / 'C'
STO
A A ->TAG  @ Show vars.
B B ->TAG
C C ->TAG
  >>
>>
This is a program like your's (I think. Your's is not running...)
@program coord:
%%HP: T(3)A(D)F(.);
<< DEG
Enter Coordinate
{
:Northing:
 :Easting:
{ 1 0 } } INPUT
OBJ-> RECT SWAP ->V2
CYLIN V-> RECT 360
HMS+ 360 MOD HMS->
DUP DUP IP DUP 10 <
0  IFTE SWAP
100 < 0  IFTE +
SWAP ->STR + SWAP FP
 . IFTE 00000
+ + DUP 1 6 SUB .
+ SWAP 7 12 SUB +
HHH.MM.S ->TAG
SWAP dist ->TAG
SWAP
>>
ciao.
enrico
TO REPLY, REMOVE .NOSPAM from my address!
----------------------------------------------------------------------------
-
: Enrico Carta                           : <                              > 
:
: University of Pisa - ITALY             : <            FOR               > 
:
----------------------------------------------------------------------------
-
----------------------------------------------------------------------
HI, I'm having a problem.......
I have an exam tomorrow, and I have some text on my computer wich I want to
put on my HP48gx.
I have xview installed on my hp but I can't get the text file copied to 
my
hp
Can anyone help me??
So;
-how do I get the textfile on my HP
-how can I view the text on my hp??
Please help me!!
Seb. Bevers
info@bevers.demon.nl
----------------------------------------------------------------------
====
I was playing with some real values, from -9 to +9, and found the HP
coded those using a 5 nibble address pointing to an encoded value in
ROM, rather than reproducing the value as a real on stack. I searched
for CK routines, and found one about reals:
CKREAL at #63B2D
It's a RPL stuff that does (inside) a DUP and a check; I went directly
there, and found:
DUP
.. -> something to a primitive code object
LC(5)   =DOREAL
GOTO    *6205B*
6205B --> RSTK=C
          C=DAT1    A
          CD1EX
          A=DAT1    A
          D1=C
          C=RSTK
          ?C=A
          <pushes a FALSE (with overwrite too)>
I thought checks would read adr on stack, point to it and check, and
then push a TRUE or FALSE accordingly. Why does the HP do a DUP, and
then an overwrite T/F? Is that better than checking and pushing a value,
leaving the enough RAM to push 5 nibbles test to the DUP function? I
never thought I'd find that routine coded.. that way. I expected a very
fast and simple routine, thus read - check - push_result, and I get a
dup - compare - overwrite_dupped_val_with_result, and several jumps
while a ML-only routine would have done fine, like:
C=DAT1   A
CD1EX
A=DAT1   A
LC(5)    =DOREAL
CD1EX
?C=A     A
GOYES    =PushTLoop  (1)
GONC     =PushFLoop  (2)
(1)  Assuming PushTLoop is close enough
     to allow a GOYES
(2)  I kept a GONC like they do in ROM,
     see last paragraph below for a question
     about that.
And when I check PushTLoop, I get something I didn't find yet:
LC(5)   =TRUE
A=C     A
PC=(A)
pushes a TRUE onto stack, but.. I doesn't change D or D1 at all!
I never thought of that because jumping to somewhere, like an object in
ROM, even if it's a valid one, wouldn't update D or D1, or does it?
And if I create an object, can I make PC jump to it at the end of my
program, rather than a {ĘD=D-1  A } and pushing the obj's addr 
using a
{ĘD1-5 DAT1=C A ..Loop.. } ?
What will happen to D, D1? Is a enough RAM check made?
(assuming C(A) contains object's address)
Also, sometimes I find a test, followed by a GOYES, and just after, a
GONC.. but according to context, the carry flag will *always* be
0/FALSE, so why is a GONC there, rather than a mere GOTO? Is GONC/GOC
faster than GOTO?
-Gilb-
00 E2 22 76 58 1F 72 BA  EA 94 78 F3 A2 63 E0 94
----------------------------------------------------------------------
[other stuff skipped]
> when I check PushTLoop, I get something I didn't find yet:
> LC(5)   =TRUE
> A=C     A
> PC=(A)
> pushes a TRUE onto stack, but.. I doesn't change D or D1 at all!
> I never thought of that because jumping to somewhere, like an object
> in ROM, even if it's a valid one, wouldn't update D or D1, or does it?
  TRUE is a PCO; the code above executes it.
Now disassemble TRUE:
LABEL   TRUE
        CON(5)   (*)+5
=PUSHA
        D=D-1    A
        GOC      ...
Mystery solved?
> And if I create an object, can I make PC jump to it at the end
> of my program, rather than a {ĘD=D-1  A } and pushing the 
obj's add...
  How would you know that you first had room to create the object?
  If your program calls GARBAGECOL itself, it could then be relocated
  during the GC, which would be a Bad Thing when the saved old PC
  was popped for RTN; OTOH, when your program has something like
  GOVLNG PUSH... [naming a function located in ROM] as its very last
  instruction, then the final code instructions are being executed
  in ROM, which is pretty safe against being moved during GC :)
  [but watch out for any new flash downloads updating your ROM
  while it's executing code :]
  Also, PCO's are not valid RPL objects in RAM, because their
  prologs would become instantly invalid if relocated.
  Now that we've said that, note how SysPtr@ jumps to DOBINT
  to push values onto the stack!  There are lots of tricks
  used in ROM, but a lot of them depend upon the fact that the code
  is frozen in ROM, and also has various related ROM functions
  defined right inside each other; trying to imitate some of the
  far-out-looking stunts in user programs, some of which might
  find their way into libraries and eventually get executed
  in temporary memory, could be unfulfilling, so remember
  what the fire-eaters and sword-swallowers always say:
  Don't try this at home, kids!    :)
In other words, understand the basis of what's being done
in ROM before trying to imitate it; standard coding practice
is much safer in user programming, especially when it's
to be executed in RAM (and perhaps even from temporary RAM)
-----------------------------------------------------------
With best wishes from:   John H Meyers   <jhmeyers@mum.edu>
----------------------------------------------------------------------
====
>I was playing with some real values, from -9 to +9, and found the HP
>coded those using a 5 nibble address pointing to an encoded value in
>ROM, rather than reproducing the value as a real on stack.
It's a litte unsefull trick use by the compiler to squize the size of the
rpl programs.
>I searched for CK routines, and found one about reals:
>CKREAL at #63B2D
>It's a RPL stuff that does (inside) a DUP and a check; I went directly
>there, and found:
>I thought checks would read adr on stack, point to it and check, and
>then push a TRUE or FALSE accordingly. Why does the HP do a DUP, and
>then an overwrite T/F? Is that better than checking and pushing a value,
>leaving the enough RAM to push 5 nibbles test to the DUP function? I
>never thought I'd find that routine coded.. that way.
the reason is to win place, HP is quite short in memory in the ROM.
>Also, sometimes I find a test, followed by a GOYES, and just after, a
>GONC.. but according to context, the carry flag will *always* be
>0/FALSE, so why is a GONC there, rather than a mere GOTO? Is GONC/GOC
>faster than GOTO?
You might not have notice that a test is modifying the carry. For example,
you will find often in programs ?A=0.A RTNYES RTN
This basicaly set the carry if A=0 and clear it if not. A valid test set 
the
carry, a not valid test clear it.
A+ Cyrille de Brebisson
Le Meilleur moment pour planter un arbre etait il y a 20 ans. Le Deuxiemme
meilleur moment est maintenant
The Best Time to plant a tree was 20 years ago. The second best moment is
now.
http://www.capway.com/brebisso
----------------------------------------------------------------------
====
>
> CKREAL at #63B2D
>
> It's a RPL stuff that does (inside) a DUP and a check; I went directly
> there, and found:
>
> DUP
> .. -> something to a primitive code object
>
> LC(5)   =DOREAL
> GOTO    *6205B*
>
> 6205B --> RSTK=C
>           C=DAT1    A
>           CD1EX
>           A=DAT1    A
>           D1=C
>           C=RSTK
>           ?C=A
>           <pushes a FALSE (with overwrite too)>
>
> I thought checks would read adr on stack, point to it and check, and
> then push a TRUE or FALSE accordingly. Why does the HP do a DUP, and
> then an overwrite T/F? Is that better than checking and pushing a value,
> leaving the enough RAM to push 5 nibbles test to the DUP function?
It's mainly done to save space; This way, they have both the DUPTYPEREAL? 
and
TYPEREAL? objects, with minimal overhead.  The difference in execution time
of DupAndThen TYPEREAL? vs. your code further on is minimal, really.  But 
HP's
requires only 5 extra nibbles...
> I
> never thought I'd find that routine coded.. that way. I expected a very
> fast and simple routine, thus read - check - push_result, and I get a
> dup - compare - overwrite_dupped_val_with_result, and several jumps
> while a ML-only routine would have done fine, like:
>
> C=DAT1   A
> CD1EX
> A=DAT1   A
> LC(5)    =DOREAL
> CD1EX
> ?C=A     A
> GOYES    =PushTLoop  (1)
> GONC     =PushFLoop  (2)
>
> (1)  Assuming PushTLoop is close enough
>      to allow a GOYES
>
> (2)  I kept a GONC like they do in ROM,
>      see last paragraph below for a question
>      about that.
>
> And when I check PushTLoop, I get something I didn't find yet:
>
> LC(5)   =TRUE
> A=C     A
> PC=(A)
>
> pushes a TRUE onto stack, but.. I doesn't change D or D1 at all!
>
Of course it does.. in the execution of the object TRUE itself, which is
a Primary Code Object that pushes its own address on the stack.  PC=(A)
resumes execution at the pointee of =TRUE, which is =TRUE + 5 nibs
> I never thought of that because jumping to somewhere, like an object in
> ROM, even if it's a valid one, wouldn't update D or D1, or does it?
>
> And if I create an object, can I make PC jump to it at the end of my
> program, rather than a {ĘD=D-1  A } and pushing the obj's 
addr using a
> {ĘD1-5 DAT1=C A ..Loop.. } ?
>
> What will happen to D, D1? Is a enough RAM check made?
Yes you can.  And yes, the check is made.  It's implicit in every object's
prologue (well the ones that need a check anyway)
>
> (assuming C(A) contains object's address)
>
> Also, sometimes I find a test, followed by a GOYES, and just after, a
> GONC.. but according to context, the carry flag will *always* be
> 0/FALSE, so why is a GONC there, rather than a mere GOTO? Is GONC/GOC
> faster than GOTO?
Faster AND shorter. A GOC/GONC takes 3 nibs vs. 4 for a GOTO
--
Werner Huysegoms
remove the x before replying
http://www.dejanews.com/       Search, Read, Discuss, or Start Your Own    
----------------------------------------------------------------------
> GONC.. but according to context, the carry flag will *always* be
> 0/FALSE, so why is a GONC there, rather than a mere GOTO? Is GONC/GOC
> faster than GOTO?
it seems you should read the manuals (SASM.DOC in this case)
more carefully;-)
Since you were pointed to the docs (Mika...), why don't you use
them ?
Another pointer: See SASM.DOC page 109 ff...
There you will see that GONC/GOC consists of 3 nibs
instead of GOTO which consumes 4 nibs.
So there's one nib less to move around in memory:-)
Where applicable, one should use GONC/GOC to save space and time.
Greetings
Raymond
GONC http://www.hpcalc.org :-)
----------------------------------------------------------------------
====
I am about still to decide me for the suitable pocket calculator (HP48 G
with 768KB or TI89): -). Now still the size of the HP would interest me. I
unfortunately found no specification in addition. The TI is at least 1 x 
3.5
x 7.3 inch largely. But how it looks with the HP?
Thomas Vetter
PS. I think however that I will buy the HP. Symbolic algebra knows Mathcad,
Marple, Mupad anyway better than TI or HP. The application of the pocket
calculator is anyway forbidden to us at the university for maths. Why I 
will
use it particularly for electro-technology.
----------------------------------------------------------------------
====
concerning your p.s. :
you're right, mathcad & co. are better in symbolic algebra but you can't
really take your computer anywhere, you have to boot etc. The Hp and the
TI are intended to fill the gap between doing a quick calculation on a
piece of paper and an extended session using one of the math programs. 
Think about one special feature of the Hp :  it can convert almost all
units into another.
I' m german as well, doing an exchange in the US right now and it really
helps me a lot.
( how much bars are a psi, how much kJ is a BTU , how much 
m^2
is a ft^2 ...) 
Hope that helps in your decision ..
        ULLI
----------------------------------
remove the ..... you know it :-)
----------------------------------------------------------------------
====
Before you respond with RTFM, I did.  I just can't quite understand how 
to
go about doing what it says.
I have Frogger on my calculator that I would like to remove to make room 
for
some more important files.  I have no add-on RAM cards.  The manual says
this:
Enter the library identifier for the library in independent memory - it 
has
the form :port:number.
TIA,
Drew Boyles
----------------------------------------------------------------------
====
first of all you should know which number is has (lets take 778 for
example). Then, if haven«t got any memcard, it is in port 0. 
So you should
write:
:0:778 DUP DETACH PURGE    (press enter)
Note that 778 should be the correct number. After this, the library should
be removed.
-=UssU=-
Drew Boyles escribi.97 en mensaje <746ild$koo$1@nntp.msstate.edu>...
>Before you respond with RTFM, I did.  I just can't quite understand 
how
to
>go about doing what it says.
>
>I have Frogger on my calculator that I would like to remove to make room
for
>some more important files.  I have no add-on RAM cards.  The manual says
>this:
>
>Enter the library identifier for the library in independent memory - it
has
>the form :port:number.
>
>TIA,
>
>Drew Boyles
>
>
----------------------------------------------------------------------
Right.
If your library has (for example), the library number 789 (decimal),
you would have to key in ':0:789' (w/o the ').
This can be the following way:
1. Press right-shift then the plus_key (+)
2. Press the zero_key (0)
3. Press right_arrow (below the 'NXT' key)
4. Key in 7 8 9
Now you have the following expression on the stack
(In fact it's tagged object, but that doesn't matter here):
In level 1 & 2 of the stack display:    :0:789
6. Press left-shift then the two_key (2)
8. Press left-shift then the EEX_key (EEX)
Now the library should have been removed.
Hope this helps
Greetings
Raymond 
----------------------------------------------------------------------
====
Drew Boyles
abb6@ra.msstate.edu
>Right.
>
>If your library has (for example), the library number 789 (decimal),
>you would have to key in ':0:789' (w/o the ').
>This can be the following way:
>1. Press right-shift then the plus_key (+)
>2. Press the zero_key (0)
>3. Press right_arrow (below the 'NXT' key)
>4. Key in 7 8 9
>
>Now you have the following expression on the stack
>(In fact it's tagged object, but that doesn't matter here):
>In level 1 & 2 of the stack display:    :0:789
>
>6. Press left-shift then the two_key (2)
>8. Press left-shift then the EEX_key (EEX)
>
>Now the library should have been removed.
>
>Hope this helps
>
>Greetings
>
>Raymond
>
----------------------------------------------------------------------
: Right.
: If your library has (for example), the library number 789 (decimal),
: you would have to key in ':0:789' (w/o the ').
Sure, but maybe you don't know that number: here's how to find out:
Press left-shift, then 2 (libraries), press A (ports).
Now select the port where you've put you're lib. (most likely port 0 or
:0: ). It then lists all libraries numbers in that port (on the menu).
Selecting a number will put in on the stack (together with the first
characters of the lib's name).
Bart Puype
--
----------------------------------------------------------------------
====
I'd really like to have one of these things. I'm willing to buy or trade. I
have several HP calculator items.
----------------------------------------------------------------------
====
Does anyone know if there is a nifty tool for calculating the shortest 
route
in a network on a 48G
----------------------------------------------------------------------
Folks,
I just purchased a 48GX and would like to know if anyone's got one to talk
to a HP 200LX via the *infrared port*.  I can get them talking using wire
(two connectivity kit serial adapter cables and one serial male gender
changer) and Kermit.  However attempts using the HP 200LX Data Comm
terminal set to Kermit server mode and configured for the infrared port
does not seem to work (the 48GX was set to infrared as well).
Rob
----------------------------------------------------------------------
====
Yes! I have transferred files to and from my 200LX.  The procedure is
available at http://www.hp.com/calculators/service/ir_transfer.html
Dave Zelkowski
> 
> Folks,
> 
> I just purchased a 48GX and would like to know if anyone's got one to 
talk
> to a HP 200LX via the *infrared port*.  I can get them talking using wire
> (two connectivity kit serial adapter cables and one serial male gender
> changer) and Kermit.  However attempts using the HP 200LX Data Comm
> terminal set to Kermit server mode and configured for the infrared 
port
> does not seem to work (the 48GX was set to infrared as well).
> 
> Rob
----------------------------------------------------------------------
====
I use the hp48gx a lot for my math classes and have started writing short
programs for fun and for things in my work. I hate cluttering my menu bar
with all those variables that the programs use. I've noticed that the
programs that I download from the net don't have that problem. In the 
manual
it says that a library handles varibles in a different way, something like
globe and local variables. So I was wondering if there's a way for a non
programer like me to turn my very small programs into librarys. I already
know how to attach and detach a library, I've done this a lot. And I also
have meta kernal and a memory card installed. I also know all the 
procedures
for attaching to different ports. The meta kernal is the latest version to.
Now with this limited imformation about my abilties as a programer can 
anyone
tell me how to do this.
Gary
gwiese@wwnet.net
http://www.dejanews.com/       Search, Read, Discuss, or Start Your Own    
----------------------------------------------------------------------
====
Can't Alg48 or Erable solve.......
My examples are relatively simple. I am using them because they cover a
lot. I would like to know if I need to enter them in there
characteristic form
( r^2+r*a+b = 0 ) on the stack. I tried LDEC in Erable. 1st on stack 0,
2ed on stack x ^ 2 + 4*x +3 and got something close to a general
solution for my 2ed example.
Non homogenous differential  equations ?
i.e.
(y' +1) / 3 = 2*x -y
    The general answer is of course y = C * e^-3*x  + 2*x-1
Or how about a homogeneous 2ed order differential equation.
i.e.
y'' + 4*y' +3*y = 0
      y(0) = 2 and y'(0) = -1
      The general answer is of course y = C * e ^ - 3*x + D*e ^ -x
      The perticuler answer is  y = - 0.5*e ^ -3*x + 2.5*e ^ -x
BTW I would like to know from an American what level
mathematics this is in school. I am a Swedish 17 years old
and might start to study in the USA.When do you start Liner algebra ?
Is it true that calculators are allowed to be used on exams ?!
All the best
Eva
----------------------------------------------------
To John M.
Now I am become Death, the destroyer of worlds.
 J. Robert Oppenheimer. I believe Robert was speaking to Seth Nedermyer,
the man who realized that plutonium could be brought to critical mass by
a symetrical chock wave.
Sanskrit text, 'Bhagavad-gita'
 (the song of God).
After witnessing the first atomic detonation at Trinity Site, just west
of Socorro, New Mexico on July 16, 1945.
Vishnu is trying to persuade the Prince that he should do his duty and
to impress him he takes on his multi-armed form (Shiva) and says:
If the radiance of a thousand suns
Were to burst at once into the sky
That would be like the splendor of the Mighty one...
 I am become Death,
The shatterer of Worlds.
Eva
----------------------------------------------------------------------
====
    SORRY THE IDENTITY HAD MY SISTERS NAME TIRIL ON IT
Can't Alg48 or Erable solve.......
My examples are relatively simple. I am using them because they cover a
lot. I would like to know if I need to enter them in there
characteristic form
( r^2+r*a+b = 0 ) on the stack. I tried LDEC in Erable. 1st on stack 0,
2ed on stack x ^ 2 + 4*x +3 and got something close to a general
solution for my 2ed example.
Non homogenous differential  equations ?
i.e.
(y' +1) / 3 = 2*x -y
    The general answer is of course y = C * e^-3*x  + 2*x-1
Or how about a homogeneous 2ed order differential equation.
i.e.
y'' + 4*y' +3*y = 0
      y(0) = 2 and y'(0) = -1
      The general answer is of course y = C * e ^ - 3*x + D*e ^ -x
      The perticuler answer is  y = - 0.5*e ^ -3*x + 2.5*e ^ -x
BTW I would like to know from an American what level
mathematics this is in school. I am a Swedish 17 years old
and might start to study in the USA.When do you start Liner algebra ?
Is it true that calculators are allowed to be used on exams ?!
All the best
Eva
----------------------------------------------------
To John M.
Now I am become Death, the destroyer of worlds.
 J. Robert Oppenheimer. I believe Robert was speaking to Seth Nedermyer,
the man who realized that plutonium could be brought to critical mass by
a symetrical chock wave.
Sanskrit text, 'Bhagavad-gita'
 (the song of God).
After witnessing the first atomic detonation at Trinity Site, just west
of Socorro, New Mexico on July 16, 1945.
Vishnu is trying to persuade the Prince that he should do his duty and
to impress him he takes on his multi-armed form (Shiva) and says:
If the radiance of a thousand suns
Were to burst at once into the sky
That would be like the splendor of the Mighty one...
 I am become Death,
The shatterer of Worlds.
Eva
----------------------------------------------------------------------
====
> Non homogenous differential  equations ?
> i.e.
> 
> (y' +1) / 3 = 2*x -y
>     The general answer is of course y = C * e^-3*x  + 2*x-1
> 
> Or how about a homogeneous 2ed order differential equation.
> 
> i.e.
> 
> y'' + 4*y' +3*y = 0
>       y(0) = 2 and y'(0) = -1
>       The general answer is of course y = C * e ^ - 3*x + D*e ^ -x
>       The perticuler answer is  y = - 0.5*e ^ -3*x + 2.5*e ^ -x
> 
> BTW I would like to know from an American what level
> mathematics this is in school. I am a Swedish 17 years old
> and might start to study in the USA.When do you start Liner algebra ?
> Is it true that calculators are allowed to be used on exams ?!
Your examples are basically the level of an AP (Advanced Placement) 
Calculus course offered in some (not all) American high schools, so 
basically kids your age would be taking the course as seniors before 
they go off to university for their first year. Some instructors let 
students use calculators, some don't, and those who do allow 
calculator use often have different thresholds of tolerance for 
features like graphing, programability, solvers, etc. It all really 
depends on the school and the teacher.
Good luck with your calculus! Viva la Wronskian!
katz
-- 
-----------+
----------------------------------------------------------------------
====
You are sure ? That work on your 49 ??
On my 49 both #0 or # 0 give the same result, the entry point is
generated, not the object.
!RPL
!NO CODE
::
  #0  or # 0  doesn't change anything 
;
@
====
====
Sorry.
I've worked with the beta of the beta 1.19-6 that I forgot that this
====
? both generate BINT0 for me.
Moreover, in response to the originator of this thread,
either you use BINT0 or you use BINT0 TOTEMPOB.  I fail
to see the use of expressly including the BINT object
in the secondary i.o. a pointer to BINT0.
If you execute the program, it will push onto the stack
a pointer to the BINT object in your secondary - you should not
====
At least on my HP49 in MASD SysRPL mode  
           # 0   ( with space ) 
           @
is compiled to BINT0, and *not* to # 0
====
More strange:
!RPL
!NO CODE
::
#0
;
@
D9D2070133B2130
so prolog + bint0 + semi
now try:
!RPL
!NO CODE
#0
@
119200000
the object!!
- Denis.
====
 
 
Why so agressive today? All I sad below turned
meanwhile out to be true. Well, my choice of words in
MASD may also does not make such a clever distinction
was a bit too sarcastic. But as you know my aim is to 
to stimulate CdB and you to make MASD superior to JAZZ
which it is already in many respects, but not in yet 
====
I may have over reacted a bit today.
But you keep asking the same thing over and over and it's frustrating
else.
every
an
====
You're welcome =)
It's possible to make a BINT with e.g. the address of BINT0 in it, like
this:
!RPL
!NOCODE
::
  DOBINT
  BINT0
;
@
However #00000h doesn't have a name, so you can't get # 0 this way.
====
Well, you could do (Jazz syntax) :
:: DOBINT PTR 0 ;
and all while in rpl mode.
You'll have to translate this into MASD syntax though.
                                     
====
A question for the ML geeks out there:
PC=(A) jumps to the address located at A[A]. BUT: If I have, say, a list
containing BINT0, this list actually only contains its pointer. How can
the Saturn know whether it's the address at A[A] or the address at the
address at A[A]?? Does this have something to do with the entry =PRLG?

====
No idea what =PRLG? does.  But the Saturn doesn't know anything about
how you've set up your objects.  It just does what you tell it.  The
semantics of the instructions are well defined; for details see sasm.doc.
PC=A:    Copy the contents of the A field of the A register into the PC.
         It does not access memory.
         Before:  A register = xxxxxxxxxxx12345,
         After:   PC = 12345
PC=(A):  Copy the five nibbles from the memory location pointed to by
         the A field of the A register into the PC.
         Before:  A register = xxxxxxxxxxx12345
                  memory loc 12345: a
                             12346: 9
====
You made a mistake about how many keystrokes you need:
if you do 2*(2+3)=
you're right it's 8 keystrokes.
However
2 3 + 2 *
is 6 keystrokes as you need to either press ENTER or SPACE to separate two
numbers (like between 2 and 3)
It is still better than algebraic.
However, if you do:
2+3=
or 2 3 +
it's the same number of keystrokes
Who just wanted to post something and took the first opprtunity to do so
head
'5
====
What is the value of the expression '5 + 3 * 2' ?
The correct value is 11, but you will only get to it with a rule that
specifies that multiplication and division happen before addition and
subtraction. That rule is an arbitrary one the mathematical community
has decided upon and something which the notation does not imply. In
fact, the rule requires operating in a manner that is at odds with the
way infix expressions are normally handled. Normally it is understood
that an operator evaluates over its immeadiately surrounding
arguments, hence '5 + 3' works out to be 8. Reading from left to right
we would encounter '5 + 3' and have all the necessary components for
resolving the '+'. That we look ahead to see that there is a '*' and
therefore put our dealing with the '+' on hold is not immeadiately
clear. The order of evaluation is determined by the placement of the
operators.
In contrast the expression '5 3 2 * +' doesn't leave a lot of room for
arriving at a result other than 11.  The principle of evaluating an
extension that says that the most recently encountered operator when
read from right to left is evaluated before any previously encountered
operators. Hence evaulating the '+' waits for the evaluation of the
'*', which having two numbers to work with evaluates to 6 and gives
the '+' two numberical arguments to work with, yielding 11. The order
of evaluation is fixed.
I don't see how it can be argued that infix notation is more compact
as the number of symbols used to form expressions is the same, and
certainly postfix seems to be more clear.
You will note that when you do addition in the traditional method
you are operating ala postifx.
627
174 +
-----
801
====
(snip) 
It's not fixed, so much as explicit.  '2 3 + 4 5 + 2 /' can be evaluated
in several equally valid ways =)
 
OK, I think by this that you didn't try my suggestion, to sit down
with a pencil and paper and actually try it.  If you had, you'd probably
quickly note that '2+3' is easier to write and read than '2 3 +'.  Why?
Because it looks an awful lot like '23 +', if you do it wrong, and this'll
get worse with more complex expressions.  If you add enough space to make
everything clear, you're not writing as compactly; if you don't, you're
not writing as clearly.  This is especially evident when you get into
talking about the fact that you're dealing in two-dimensional space when
you write on paper.  You've superscripts, and multileveled rationals, and
radicals and logarithms (but standard notation for these is /awful/) and
all of these are three dimensional.  If you want to represent a ratio
nicely, you can use a:b or a/b or a (over) b, and only the first two
really translate to postfix well.  With complex expressions you want the
third.  Not that the postfix alternatives to these aren't as clear, just
that they aren't as compact.  With skill, maybe this difference will
disappear.  I'll start doing my calculus with a postfix style, and see
a bit more of what it's like.
(snip)
 
Yes, I know =/  It was this argument combined with the arguments from
both camps that the other was innately nonsensical that inspired me to
start this thread in the first place.  It's true that you learn to do
math in this style.  This doesn't mean anything, though =)  What you're
learning is the process, and the only reason the numbers are arrayed in
that form is so you can see the place structures clearly.  They could
also teach it like this:
627
+++
174
===
801
Which could be described as vertical infix.  Your argument seems to
be constructed against people who declare that postfix is harder to
understand than infix.  I'm not saying that.  What I'm saying is that
they are /equally/ hard to understand in general, and unequally so
depending on where you use them.  Infix can be really hard to understand
when all its notation is crushed into a one-dimensional space.  You're
====
I'm gonna put my two cents in, then shut up...
Both algebraic and rpn are easy...I used algebraic forever, but then just 

to
be different (and to make my calculator unusable to anyone else in my
class...lol) I starting using rpn...I must say that I would rather put
2+3*(5+4) then 2 3 + 5 4 + *
But either way, I don't think one is faster than the other...it will always
====
For doing complex calculations, RPN is way more natural.  Got something 
comples
====
But you forget to mention what you have to do in RPN to accomplish the
affect of the 4-7 levels of parenthesis...it's a lot more easy in algebraic
mode...just enter it like you see it
e.com...
to
always
whichever
comples
calculate
4-7
====
First of all usually you don't see it with all the parenthesis, but with
some paper style shortcuts like horizontal lines for fractions.
And in RPN you can (I can, at least!) enter it the way you *think*.
====
well, I don't think three spc two space plus  I think three plus
two...but that's beside the point...back to what I said, doing complex
numbers in BOTH modes is NOT EASY...for things like that I just assume use
the equation writer!!! =P
algebraic
 I personally prefer RPN but which is right or best seems to be 
largely a
 religous issue; people have generally made up their minds about it.
 However, most calculators follow RPN when keyboard functions are involved.
====
true very true...but when entering several different calculations along 
that
line, it can be a bit confusing (Like I said, I use RPN, I'm just 
saying...)
since you can't just enter SIN(30) + COS(58) / TAN(34)...that would be 30
SIN 58 COS 34 TAN / + in RPN mode....Please don't get me wrong, I use RPN
and love it, but truly it's not as convenient maybe? when doing things like
that....like in one of my other posts, I don't think like that...I think
SIN(30) not 30 space SIN....
<slrn9qi4p7.pus.The.Central.Scrutinizer.nospam@C1459607-A.arvada1.co.home.co

just to
always
whichever
====
====
Just to be precise:
If it's a vector the list is then { #n }
Also keep in mind that it works with real and complex arrays but not with
====
====
Anybody know hoy I can convert a number hexadecimal to decimal or binary
format, with the HP49G.
PD: In the manual don't especify anythink.
------------------------------------------
Hola
Alguien sabe como puedo pasar numeros de hexadecimal a decimal o binario 
con
la HP49G.
PD: El manual no dice nada al respecto.
Gracias.
====
1. Input the number in the format # xxxxxh, e.g. # 12345h
2. Use one of the commands HEX DEC OCT BIN, depending on which format
you want.
Binary integers are not actually converted, they're only *displayed*
====
(snip spanish)
  DEC # F001F001h 
is a simple way, but you might want something with IFERR.
do simple string manipulation in both to hide the #...x stuff from
yourself.
You can also do the conversions manually, which is something I would
====
Just remember to be in RPN mode.
1. Input the number in the format # xxxxxh, e.g. # 12345h
2. Use one of the commands HEX DEC OCT BIN, depending on which format
you want.
Binary integers are not actually converted, they're only *displayed*
====
For sale HP 41CV & Stats R.O.M
In it's day, a top-of-the-range programmable calculators. 
Used, but not abused? Comes with non-HP case.
and also:-
Hewlett Packard HP-41 stats module, part number 
000041-14001
Used, but not abused?
Best offer secures !!!
Buyer collects from LONDON NW2 or ads postage/SAE Jiffy 
bag.
Stephen 07956-544202
====
The point is, after the HP has stopped working on an object and doesn't
need it any more, it (the object) is not instantly purged from memory.
As soon as there's no more room for another new object, the old, now
unneeded ones must be thrown away. This time consuming process is known
as garbage collection. Depending on how many objects of what size are
waiting to be collected, it can take between almost no time and 4-5
====
====
====
No ! It says specifically 48 KB !
  The DETLEF subdirectory on this disk contain's Detlef Mueller's 
complete
System RPL development system and library creator.  It almost obsoletes
Goodies Disk #4!  Check it out.
Here's a sample of useful System RPL which I just finished today, developed
entirely on the HP 48 keyboard.  It sorts a list of real numbers into
ascending order, and it's *fast*, over 400 times faster (!) than HP's SORT
program in the Owner's Manual (in User RPL, of course) on a list of 100
numbers:
:: CK1&Dispatch # 5h
  :: INNERCOMP 1LAMBIND 1GETLAM
    #1+_ONE_DO 1GETLAM ROLL # 0h INDEX@  don't work syntax err at 1+
      WHILE
          ITE ROTDROPSWAP SWAPDROP
        ;
      REPEAT ROT2DROP UNROLL
    LOOP 1GETABND {}N
  ;
;
========================
<ccslitha@yahoo.com
 Perhaps though, you connected your keyboard to your hp48?
====
Developed entirely on the HP 48 keyboard doesn't imply it's UserRPL.
====
Just another something to watch out for.  Maybe it's what you've run
into?:
If you use JAZZ, you do not need to worry about that. Note that there
is no space after the parenthesis.
For example:
::
  ...
  ZERO_DO
  ... *** WRONG ***
  DROPLOOP
  ...
;
::
  ...
  ZERO_DO (DO)
  ... *** RIGHT ***
  DROPLOOP (LOOP)
  ...
;
* #1+_ONE_DO    ...
cc
On Wed, 19 Sep 2001 08:22:35 -0700, _______TVD________
====
www.greatunpublished.com seems to be down.  Anybody know if they're
in New York?  That might be why.
Barry
====
====
====
I am trying to connect my 49g to my PC using a homemade cable (cause
the real one hasn't arrived yet) and it doesn't work. I thought i read
somewhere about some of the early 49g's having a problem with the comm
port, so I don't know if the problem is my cable or the calc :) The
serial number is ID9460.... and the ROM is version #1.10
oh yeah, i also stupidly tried to use one of the old pc connector
cables (the one's that you are not supposed to use with the 49g...i
didn't know until after i had tried). anyway, the unit still appears
====
only the version numbers ID93.... (like mine) have the bug
I did the same thing when I used a 38 cable (came with my 48G)
My calc went berserk and dsplayed a totally black screen, beeping and
stuff
====
I suppose that you use last version of connectivity kit software
====
In my linear algebra class we've been doing elimination, matrix row
reduction to solve systems of linear equations.  We've also been using
Derive 5.0.
My question is that sometimes when the by hand method/book/derive
lists an answer that there is no solution for a system, my hp49 will
give solutions. Usually they are really small decimals ex. .382774399
or something on that line.  Do I have a setting wrong or is the HP
linear solve not able to tell that there are no solutions.
My secondary question is that when there are infinate solutions the HP
again returns a single solution where derive will set one variable
equal to itself then give the other two in terms of the first.
My problem is that if I'm going to use linear solv on the HP that it's
pretty well useless if I can't tell if there are infinate solutions or
====
====
Okay, yeah I should have included that the first time, doh.
2 examples one with infinate solutions the other with no solutions
no solution:  (accoridng to derive, by hand and book)
x + y + 3z = 12
2x +2y + 6z = 6
HP 49G using NumSlv and Solve linear system returns....
.436363....
.436363....
1.30909....
Infinte solutions (derive returns x = z+4 y = z-4  z=z
x + 2y - 3z = -4
2x + y - 3Z = 4
====
====
Can you give my the syntax for the entries for linsolv?  I looked in the 
cas manual but evidently I don't understand their description of the input 
cuase I get a bad argument type.  
====
Please explain this, Julian. I don't understand what you mean by much 
more
dynamic at runtime.
I have a HP-48sx, and have programmed in RPL, but the speed penalty was 
just
too painful to continue.
own defining words doesn't seem to be possible. (A big loss, IMHO.)
--
- Bill Zimmerly, http://www.zimmerly.com
Reasonable people adapt themselves to the world. Unreasonable people
====
must be a great (and old) printer to spend so much money in a USB/RS232
converter.
I've seen printers cheaper than the price of a USB adapter
Nick.
This would seem to indicate a space was used after #1+ rather than the
#1+_ONE_DO mnemonic required. How are you compiling the program, Jazz or
HPTools? (think think) guessing Jazz from the syntax error description.
Are you using quotation marks around the indicated string? Also note that 

if
you attempt to edit a string by pressing the down arrow, it will begin with
something like C$ 56 (then the rest of the string) and the HP will only 
accept
56 characters in the string. So if you add three characters they will be
chopped off. You can change the 56 to a very large number (say 999) and 
then
add as many characters as you need (up to 999!) and the press enter to 
accept
all of your changes.
 
 
Well I think I'm going to take some time to write a document specifically 

aimed
at new users of the 49.  I've been mulling over some of the things to talk 

about
dealing with common issues and problems.  I'd like to get it to a list of 

about
20 things to discuss.  I don't really want this to be a huge document or
completely re-written help file, or a manual for programs, but rather a 
guide of
things every user should learn soon to help them get going.  Here are some
things I've thought of.
1) fixing programs when transfer incorrectly (ObFx, etc)
2) Library management (various installation methods, access, maybe Libman)
3) RPN (why? benefits, etc.)
4) Key shortcuts (r-shift hold ENTER, etc)
5) Plotting (why do so many programs appear after I plot named Y1...N? 
Changing
speed, etc.)
6) MATR & EQW tutorial (the important things to know, condensed)
7) KEYTIME, STARTED, EXITED, etc.
8) ROMs (is it safe to use a beta? ;-)
9) usefull doc locations (manuals, 48 manuals, command reference, 
programming,
etc)
10) CHOOSE vs. SOFT
11) list of required programs to make operating easier
12) transferring files (what program to use, modes, etc.)
13) basic CAS operations
Basically I'm wanting to make a FAQ/tutorial that isn't as intimidating as 

other
documents, specificallly designed for new users.  I need some more things 

people
should learn quickly (like those things that we learned at one point and 
said
how did i live without knowing this before)  This isn't a solid lkist 

right
suggestions.
that is an excellent idea. maybe it can be put in a mini brochure with all 

the
essential info (that can be printed an folded in half). maybe you could 
write a
book called HP49G for dummies. we need something like that. thanks for 

your
effort and i'll wait for your release. congratulations in advance.
 i think it also could be interesting key assignments (keyman) they make
my life easier! when i bought my hp (2 years ago) i was about 1 year 
wondering what the user key was!! a lot of new people (i'm not an
expert at all! :-(  ) don't know about it! 
see you!!!
vicent
 i don't know if anybody has already come up with something. if not, i'll 

look it
up in the near future.
http://www.ibhdoran.com/usb_link.html#Silicon : Interface
 Where can I buy the Advanced user manual for the HP49G calculator.  I 
tried
to find it on the HP site but couldn't find it.
 If you use FOR command, the index variable (local) is created 
automatically
so these lines:
My version:
<< [[0,0] [0,0]] sto A ;
      FOR(I,1,2)
          FOR(J,1,2)
               PUT(A;{I,J},9) sto A ;
          NEXT ;
      NEXT
 
 
In my opinion you should do it in UserRPL, becuase it is faster and file is
smaller.
<<
[[ 0 0 ]
 [ 0 0 ]]
'A'  STO
A
1  2  FOR  I
    1  2  FOR  J
    {  I  J  } 9  PUT
    NEXT
  NEXT
  'A'  STO
111.5 bytes
The previous version (my previous post) in HP-Basic takes 251.5 bytes and 

is
much slower.
second sample in UserRPL:
<<
 [[ 0 0 ]
  [ 0 0 ]]
  'A' STO
  A 9 CON
  'A' STO
Works too but takes only 68.5 bytes and is very fast.
You mean that one can use only five variables in programming? Surely
For doing symbolic calculations in a program or on the home menu, those have 

to
be used.  Otherwise there would be problems with the A-Z variable names.  

You
can use those to get numerical results (such as with integer commands and 

such).
In the EQW where you do almost all the symbolic calculations you can use 
any
variable you'd like.
For regular programming a-z and theta are there 10 lists, matrices, stat 
lists,
equations, and a bunch of other variables availible to use.
 Many English words can be looked up nicely from dict.org, port 2628.  Just
telnet to it and enter 'h' or 'help' to get the commands.  'd fogie' would
give you definitions for fogie, for instance.  What Matt is really saying
here, is that the people who find his wording offensive are old, implying
that they are behind the times and oversensitive and not adjusting properly
to the modern world.  ...or maybe I'm reading too much into his attitude.
 I donwloaded the extable files from Eric«s site but it seem 

that they are 
for 
Win plattform
I dropped the nine files unzipped to the 49g but it didnt accepted them.
Can you send the right file to marxpio@ig.com.br?
You will be in my heroes gallery ;)
  
have to agree with Tim there...he didn't do anything wrong...plus there is
nothing wrong with saying that some people aren't just as bright as
others...has nothing to do with being a insensitive person...or a bad
customer support specialist at that...
But it should'nt be too hard to implement in asm? Little program which
Then makes the required changes to system variables and restores the 
port configuration... Just a tought :)
 
I don't have any suggestions for programs, but I've some thoughts and I 
hope
you can make some use out of them:
  1) With a disassembly of the ROM, one thing you might want to look at 
first
     would be the code at the supported entry points.  Then you could at 
points
     jump to a location after the supported entry point, and only execute 

part
     of the code there.
  2) The HP49G has an diassembler in it.  If you could emulate the HP49G on
     your computer, you could possibly utiltize this to disassemble pieces 

of
     the ROM at a time, which would let you explore in a odd way =)  I'm 
not
     sure about how you'd get ROM code to the stack, except that it would 

be
     trivial to build a CODE object from arbitrary points at memory in ML.
     Maybe there is a neater way.
  3) www.hpcalc.org has a whole section on PC utilities.
  4) Saturn machine code is simple enough that you could write a 
disassembler
     yourself in any high-level language, and then the only problem would 

be
     getting ROM code to your program.  The experience might make you a 
better
     programmer =)
  5) Maybe this is obvious, but once you have the code diassembled with
     addresses the thing to do is search for RTN-type instructions.  I can
     imagine an ML program that'd search through the ROM, producing CODE
     objects to match ROM code that's jumped to at some point and the next
     RTN.  Of course when you use the disassembler you can notice that 
there's
     more to the entry than given, and move forward or backward.
  6) OK, now I'm getting interested for myself.  How about an on-the-fly
     disassembler that could follow jumps, go to arbitrary points of 
memory,
     etc?  These thoughts are becoming less useful to your computer-based
     interests, so I'll stop now =)
Do you really need to use a computer?  Maybe I'm biased.  On weekdays I
spend so much time away from my computers that it's not economical for me
to rely on them, when my 49 will serve.
Have you tried Derek Nickel's Voyager program?  It's an old DOS
program, but it does exactly what you want to do, and it's FAST!  Use
its SEARCH command to look through the entire ROM or any specified
part of it for any desired hex sequence, and use the FIND command to
search for substrings in the comment field of all entry points in the
easily-edited entry point table.  I had months of fun hunting for
unsupported entry points and then using Voyager to disassemble them.
Voyager takes up half of Goodies Disk #2.  It's also available here:
http://www.hpcalc.org/details.php?id=3824
 
Is there a way of using it on a rom image without dumping.  I ask because 

I've
been working with Alain Robillard who just recently ported Chemlab to the 49 

on
getting a version to the 39/40.  I've managed to find most of the 
unsupported
entries for the 39 using emu48, but the few others aren't in the same 
location
so I need something to search the ROM.  I couldn't figure out how to dump 

it.
Any other programs that will work on a regular ROM image that you (or 
anyone
else) knows of?
 
Funny reading :-)
But how come that F = System RAM card pulled?