B37


====
Would like some feedback on the above Survey ROM cards. Have had some
feedback on TDS and SMI but nothing on D'Zign survey card. Does anyone use
this card?
All replies appreciated.
Justin
====
You might want to try the message board at pobonline.com.
My coworker had thier downloadable software, not the card.  I just
remember it being really slow.  That may have changed by now.
Have a nice day.
====
it's been quite some time, since I last looked into the newsgroup. I
also checked if there are new programs for the HP49 and found some
nice ones. For the most part I'd now switch to the HP49, but I mostly
still use the 48. One reason is that some important programs are
missing on the 49. Namely PowerPlot with Table is just nice to have
and I miss it :-(
Isn't there a programmer with some ML knowledge who likes to try the
port of PowerPlot light (as this one is finished and absolutly
stable). I really would love that.
There have been quite some attempts to port the program (or better
said persons who said they wanted to give it a try), but it never got
reality until the Beta version by Yoann Desir. I was happy that then
PowerPlot would have got ported, but after the Beta the work stopped
:-( 
Are you waiting for me to learn SysRPL/ML? ;-)
This might be an option, but I'm sure that it would be _much_ faster
programmers would take the challenge :-)
On a side note. Too bad that now (I'm just working on my diploma
thesis) I don't need a calculator that often. My family is sort of
irritated that now there are days, when I don't take my HP48/49 with
me ;-)
Greetings from Cologne
Peter
-- 
Great HP48/49 links:
http://www.hpcalc.org
http://move.to/hpkb
to find *old* postings search:
http://groups.google.com/advanced_group_search
====
I have had a graph engine for a long while now, coded in SysRPL. It's
usually much faster than PowerPlot (compared on my HP48GX). At its worst it
mathces the speed of PP, at its best, its at least twenty times faster
(rarely). usually it's around 2-3 times faster. For normal functions, a
complete plot (131 points) take 0.8-1.2 seconds to plot. Such a normal plot
could be of 'SIN(X-1)/(X-1)', 'X^3+2*X-2' or 'LN(1/X-2)' for example.
I havent finished a front end for the porgram, but have tried a couple of
times - that's the reason nothing is released. I'm pretty sure I've written
this here a while back?
PowerPlot for the HP49G is definetely a big asset, but the lack of interest
just shows how few are actually using the HP49G.
The library I've coded that have drawn most attention, is definetely
SYMBTOOLZ (which is broken on newer ROM revisions). I have gotten maybe
15-20 different people contacting me about it - most had problems because
they didn't read the huge manual. That's too small a user group to
motivate.
...and now HP has shut down intelligent calculator development, and thrown
away their best guys doing exactly that. It doesn't really help progress
here, does it?
====
Will you fix it? Please.
X
for a new release...soon...
====
1.2 seconds for 131 calculations of 'SIN(X-1)/(X-1)'..?  On a HP48GX?  
That's
pretty incredible, considering that the fastest thing I could think of for
computing a single point in sysrpl would look like:
::
 % .6
 %1
 %-
 %SIN
 % .6
 %1
 %-
 %/
;
where .6 would be a hard-coded value of X into the expression.  On my 48, 
TIM
produces the result of the above program in 18.5ms.  If I multiply that by 
131,
that would be ~2.42 seconds just to make 131 such calculations, without
looping, substituion, error checking, line drawing, and other 
non-computational
stuff associated with graphing something...  So you're saying that you can
completely plot 131 points of the sample expression 'SIN(X-1)/(X-1)' in half 

of
the absolute minimal time that it takes the HP48 to come up with the 131
results??
Aaron
====
No, I'm doing it on the HP49G. My graph engine is for the '49. I just
checked - I use 2.16 seconds to calculate the points, and 0.2 seconds to
draw them on screen.
non-computational
My above timing include this.
...  So you're saying that you can
half of
No, sorry. Not in SysRPL. I'll provide the exact times (add 0.0-0.3 seconds
to draw the complete plot):
'SIN(X-1)/(X-1)': 2.16 s. (PP: 5.74 s.)
'X^3+2*X-2': 2.35 s. (PP: 5.90 s.)
'LN(1/X-2)': 0.93 s. (PP: 2.95 s.)
'2*X+1': 0.47 s. (PP: 3.42 s.)
'(9*X-EXP(X-1000)+8973)^2-X/2': 2.71 s. (PP: 6.95 s.)
'X': 0.00 s. (PP: 2.76 s.)
====
Really hope to see it in actions soon.
seconds
====
I'm afarid you won't. I'm not sure, but I think I have only one single
program left for release at some point in time. I haven't got any 
motivation
in this regard - the HP company killed the spark in me, doing what they 
have
done lately.
====
1)', 'X^3+2*X-2' or 'LN(1/X-2)' for example.
I am interested!!  Now I feel that my HP49G Graphical Calculator is not 
complete without your graph engine.
Santos Lucero
====
I for one am really excited about your program.  I understand that HP is
quitting development, but HP 49G is one of the best calculator.  I really
hope more programs from the 48 are ported since the 49G is lacking alittle
in 3rd party support.
I really hope to see you finish your program, the default plotter is just
too slow.
it
plot
written
interest
====
[Porting PowerPlot to the HP49]
Arggh, some of the best programs are from Raymond. Maybe he would like
to make that great port? :-)
Greetings from Cologne
Peter
-- 
Great HP48/49 links:
http://www.hpcalc.org
http://move.to/hpkb
to find *old* postings search:
http://groups.google.com/advanced_group_search
====
If someone is wasting BW here, it's definetely you. Stop complaining - it
just creates even more BW use.
====
You are quite correct.  0^0 is a singularity.  [inf]^[inf],
[inf]^0, and 0^[inf] are also.  There are a few more that I cannot
remember right now.    I could send you the calculus proof for this if
you want it.
Dylan
====
Yes, it is an essential singularity in real or complex arithmetic.  But is
useful to define 0^0 to be 1 for integer arithmetic.  This is covered in
the sci.math FAQ, I believe.
Scott
-- 
Scott Hemphill  hemphill@alumni.caltech.edu
This isn't flying.  This is falling, with style.  -- Buzz Lightyear
====
And in HP 49G
The Integer Math gives ?
but the real/complex gives 1.
is
====
If you will read  carefully, you will note that no one said that the 
HP49 evaluates 0^0 as 1 , although I think it should. 
In an evaluation of zero to the power zero, if either, or both, 
zeros are approximate form( with decimal points) or if the numerical 
mode is set to approximate, then the evaluation gives 1., with 
decimal point. Parisse says that this is for compatibility with the 
48, which always gives 1.
Only when both are exact zeros without decimal points AND the 
numerical mode is set to exact, does it evaluate to '?'.
About the proper meaning of 0^0 outside the HP world, see Concrete 
Mathematics, or do some Google research.
====
I just purchased a TI-89 calculator.  Now before you quit reading, I
am a very avid HP-48,49 user and have been for years.  I was curious
about the functionality of the TI-89 however now that I have two
calculators, I am curious on how to use th two.  I prefer the HP
primarily.  For those of you who have both the HP and the TI, which
calculator do you use primarily and how do you use the other
====
I use the TI89 to satisfy my curiosness regarding some mathematical 
aspects.
The TI is very fast for some symbolic calculations (Taylor series, although
somewhat limited, and summations, integrals, differentiations and so on).
The HP49G simply is way to slow to do certain tasks, for example this:
spends 19 seconds on the HP49G, but only 8.5 seconds on the TI89.
'9*X-EXP(X-1000)+8973' 'X' 1. TAYLR takes forever on the HP49G (must be a
bug), while it's calculated instantly on the TI89. In fact, the TI89 can
calculate a Taylor series to degree 25 in one second, of the above
expression. There a more than a few Taylor series expansions that take
forever on the '49.
The TI89 do have its limitations too - it can't develop a Taylor series for
'SIN(X)/X+1' for example.
====
I use the latest rom and TEVAL says 7.4142 seconds :-(
Right, very strange...
 
====
Now it finished with an Insufficient Memory error. It must be a bug
then...
====
Aaron, I had an 89(a gift from a friend),
I just can't get used to NON-RPN calculators.
I do alot of numerical computations, and I extensibly use the stack.
I could not make myself bring the 89 to work, and it was just sitting
there in my desk at home.
I finally gave the 89 to my cousin, he is in high school, he LOVES IT!
HM 
PS. My humble opinion is solely based on my experience. Others can have
a different view and/or taste.
-- 
====
Ok, I took it. You can download the new package at
http://youpibouh.thefreecat.org/info/prog/down/usinagaz-cur.zip (it is
still version 0.2), or on hpcalc.org when it is updated.
It now includes wget, which can be used as :
http://www.hpcalc.org/ wget
<HTML DOCTYPE.... (beware, this page is about 19Kb long ! it takes some
time to download it !)
-- 
 Moralit.8e : le modem et le cablerouteur font comme les filles, ils
 papotent toute la journ.8ee.
 -+- RB in NPC : Et en plus, ils ne parlent que de bits -+-
====
And what kind of record ?
You seem to forget the fact that ACO was profitable.
Even if I don't really care if it was closed or not.
====
Of developing products that fail in the market place. Of not gaining
nor even holding on to market share in their target market(s).
====
Are HP calculators still in production, by the way? How long will they 
still
be made?
Arnold
1a8f5fe5.0204300758.51f7143@posting.google.com, Helen
====
Oh, I'm sure helen really has a clue. (just wait...)
StarTrek devices ? ha ha ha. what's that? At least we know helen is an 
old lady.
.....and, is that your reason?   Come on !!!!! 
I have met many women.......and.......mmmmm.....   :-)
====
Just a few. Even my sis, who grew up on the ranch with me, who helped
me rebuild engines, put up hay, weld, shoe horses, drag calves to the
branding fire...she was the biggest tomboy I ever knew, yet even she
really didn't understand machinery. Where do you or someone you know
stand in comparison?
J.C. Randerson
Pueblo, Colorado
====
I know many very competent female engineers. One of my female
colleagues in engineering teaches courses like machine design at an
Ivy-League school. If you take a look outside your cave, you will find
many more examples of that sort.
====
I somehow agree. I personnaly think that the ACO was killed because we 
ended
up having 7 different bosses in 4 years. To make things worst, the managers
in carly's team, are, like her, strong willed directing types instead of
being leaders. No 1.5 years long R&D project can wistand having 2 or 3
different heavely directing peoples managing them. and Carly is solly
responsible for the choice of her reports and the type of peoples that she
choose. Personnaly, I like leaders, not 'directors'. but this is a personal
view.
Well, the only products we actually did finish were the ones that were
develop without non-technical peoples looking at. The reason being that
non-technical peoples (marketing, managers and the like) do live in an
everchanging word and only have short term vision (3 to 6 months). And thus
will change the any reasonably sized project's requierements 4 or 5 times
during it's life time, thus killing it or denaturing it at such a point 
that
it becomes a dog. I agree that this things should be looked at by non
techies, but in today's market, it is simpely unworkable. In the group 
where
I am now, we are working with 2-5 year time frames, and marketing peoples,
just don't want to join us, they simply have no training, nor will to see
that far in the future. We, poor litte tech are teherfore left to develop
products without any marketing input, which is as bad as what was done in
the ACO.
:-)
====
You know, this sounds a little bit like you have the choice of
developing a product that fails because there is no input from
non-technical folk, or developing one that fails because there was too
much meddling from non-techies. While that is of course an
exaggeration, it serves to illustrate a fundamental difficulty very
nicely: Ultimately, you need people who understand both worlds, and
those are extremely rare. Without them, however, it is very hard to
develop products that are really successful.
P.S.: I take it you understand that I am not blaming ACO's failure on
you or your co-workers. Developing successful products takes a team of
people with different backgrounds, working together in a fruitful
manner. Obviously, that did not happen at ACO, but the reason it did
not happen cannot simply be pinned on one side or the other. And, by
the way, that is one of the areas where the art and science of
management becomes so vitally important, the general disregard of
technicians towards the management notwithstanding...
====
I'm having trouble finding any good ones.  At hpcalc.org, there are tons of
board games for the 48, but not that many for the 49.  Anyone know if the
game Hex has ever been developed to any HP calculator?
====
I think I read about that somewhere. Dunno, something like ...
http://www.hpcalc.org/details.php?id=793
Another hex-shaped board game :-)
http://www.hpcalc.org/details.php?id=930
Bye.
HPCC member #1046
-
-- 
====
J.,
There's quite an interesting board game going on at the moment in the
courts between Hewlett and Fiorina.
Sorry, couldn't resist.
Mark.
of
====
I don't know how it got there, which pisses me off almost as much as the
fact that it's there!
I'm assuming it's impractical to replace the LCD display.  It would 
probably
cost more than buying a new one, right?  There's a portion where the screen
is visible, but it's at the top.  It's basically useless.
So what should I do with it, anyone have any suggestions?
====
You obviously do not know the inside story, or you know way more that you
would like us to believe.
The fact that the Xpander project was cancelled had little to do with the
product itself, rather the consequences of HP getting out of the education
business (which not surprisingly happened just after a change of president)
I do not believe that Xpander was an ill-conceived project. Maybe some
choices like using Windows CE raised its price over than what it should 
have
been if you are targeting students.
Xpander was the really first product designed by teachers for teachers.
Hundred of them were involved in the process. Of course this machine had
absolutely nothing to do in the calculator business. And it was not a
calculator as such, but I believe it was a great concept. The only good
thing about not releasing such appliance is that the existing HP community
can't complain about this product not being a traditional HP calculator :)
I am still in contact with various teachers all over the world, and they 
are
all waiting for a product similar to what Xpander could have been. And so
far there is nothing close to it.
Xpander was a pure learning tool aim at high-school student.
But did HP ever have a good marketing team ?
Many people here complain that HP is not what it used to be, that is a
company with products designed by engineers for engineers.
Look at the first Windows CE devices. That was a pure marketing product. 
But
totally useless with unusable feature. Obviously the people who first
designed it didn't really use it on a day to day basis.
The HP49G was made in response to the HP calculator community who have been
asking for a long time for an HP48 with more memory, faster and with more
CAS capabilities. That's exactly what the HP49 was.
But then you had marketing people who came over and said: it has to look
more like a TI, let's change the design so it looks more : cool. Then 
it
had to look even more like a TI with algebraic in it.
If the product had been designed by engineers only, it would have looked
much closer to an existing HP48 hardware...
So you claim that Xpander had no market whatsoever? On which ground.. I
would be interested to know how so many teachers could be wrong.
Not more surprised when I see how HP is doing when shutting down division 
or
reducing all the R&D budgets.
I was really happy when ACO was closed down, I would have left anyway and 
as
====
Yes, but so what? There are many wonderful things out there that lots
of people know nothing about, in all areas of human interest. Why
should everybody be interested in computer algebra?
No, it could not. The complexity of the interaction you need to do CA
is simply orders of magnitude above the one you need to handle a video
phone. That is so obvious that it should not require any further
explanation. No interface of any kind can remove that kind of
complexity, period.
Unless, of course, you have a device that is so intelligent that it
can do all the work for you. For now, and in the foreseeable future,
those devices are called mathematicians, scientists, etc. ;-)
====
Of course not computer algebra itself but computer algebra as the tool
for doing mathematics. As one door for entering the math universe.
*That* is the pity. That we have tools for making an easy entrance to
a very fascinating world, which could activate some more grey cells,
but instead of this we expect a black box to solve some integrals.
Now, of course you could ask me also, why should everybody be
interested in mathematics? Well, working the mathematical way, we can
learn to draw maximum of conclusions from a minimum of premises.
Exercising this helps abstracting from special to general, and helps
sharpening mind. Perhaps the answer to the question about interest for
mathematics could be a new question: Do we want it the easy way, that
is to believe anything without thinking for ourselves, or do we want
to wonder, think and ask about the world, life and anything else ;-)
If this interaction is by means of menus, dialogs, command lines and
other primitive user interfaces, then yes, you are right. But as
already said, a good recognition software, that can receive and
understand your writings on a touch screen could be the first step.
Instead of typing your input in the input lines of mathematica, you
could also write it on the screen, this doesn't change anything of the
inner workings of the CAS.
Ahhh, so may I say: I need new batteries, can you hear me? ;-)
Greetings,
Nick.
====
That is a very important point, and one you are getting completely
wrong, IMO. CA is not a tool for entering the universe of mathematics,
anymore than, say, computer-controlled offset printing machines are
the tools for entering the universe of literature. In order to use CAS
profitably, you already have to know a lot of mathematics.
There is no easy entrance into the world of mathematics. Quite
likely you have learnt so much mathematics in your life that you have
all but forgotten this fundamental truth. Are you seriously suggesting
that anybody would be able to use a CAS, without a deep knowledge of
the mathematics involved? And then even learn something in the
process? Have you ever taught mathematics at any level?
Maybe I am old-fashioned, but my experience with my own children tells
me that you cannot teach them to run before they can walk. As an
example, IMO it would be disastrous to give calculators to
six-year-olds in order to provide them with an easy entrance into
some of the basic mathematical concepts relating to whole numbers
(counting, adding, subtracting, etc.). I believe the same is true, at
a different level, for high-school math, college math, etc.
But it also does not change the fact that what you write on that
screen can potentially be very complex, and the (intermediate) output
usually is even more complex. What you describe above does not change
any of this at all.
====
Oh, but you will :-)
All young students in US will eventually become engineers of tommorow.
They don't have much choice beyond TI, they are tought in high schools
and most universities on TI, many corporations have basic TI training for 
it's
lower level technical personel but I've yet to see one designed for HP or
Casio. In that sense TI won (or Casio :-( ).
Precisely. I love TI automatic fast simplification of formulas, where many
consider that as fault.
In my case I have TI92. It's bulky, but it's keyobard happen to be 
superior,
thanks of course to QWERTY. I also hate TI Basic. It's not only
slow but seems like afterthought added to the system a the last hour.
Too litlle too late rarely is the outcome of some recent bad decisions.
Rather it is the product of years of neglecting the customer needs and
ignoring the competition.
Jack
====
And later:
Oh Helen, where have you been all my life?
I'll go burn some sack; 'tis too late to go to bed now.  ;-)
Shakespeare. Twelfth Night. Act II, Scene III.
Bye.
HPCC member #1046
-
-- 
====
Thanx everybody! I will try. ...and for sure I'll succeed ;-)
bye and good programming
ulrik
====
which is involved graphic.
Eduardos book does touch graphics. Try it out - you get nothing served. 
It's
one of the best starts for SysRPL out there.
====
maybe these could be links for you:
http://www.hpcalc.org/hp48/docs/books/graphics.html
http://www.hpcalc.org/details.php?id=1727
http://www.hpcalc.org/details.php?id=1763
http://www.hpcalc.org/details.php?id=1764
http://www.hpcalc.org/details.php?id=1765
Raymond
====
there are some PDF-Files on www.courbis.com available. (Click on
the files HP48 Machine Language is written in english. There are some
other files in french available as well. What a great chance to learn this
language ;-)
Some of them are dealing with Programming the HP48 (SX/GX). But I'm not 
much
of a (HP4X)-programmer so I'm not sure if it fits your needs.
have a lot of fun
Roman
====
Thanx verrry much!!
Your link seems very interesting (it is now in my bookmark), but only for 
ML.
I'm looking for sysrpl.
Mayby I will use it later when I'll be stronger in programming!
Bye, ulrik
====
i downloaded this file: fem48/49 (for instance..). 
a) how can i transfer the file *.lib to my calculator?
b) i transferred it through renaming the *.lib in *. without
extension, but there is only a string, and no library which i can
attach.
Pleas help!
thanx.
====
The convention has been:
[HARD] keys
|SOFT| keys
COMMANDS as from [CAT] or [alpha][alpha]COMMAND NAME[alpha]
Right Arrow = [RA]
Left Shift = [LS]
& means Hold Down the first key while pressing the second
[LS]&[LA]
====
Ahem. I assume you are using the HP-48, since there wasn't a different note 

in
the header;-)
====
On Sun, 28 Apr 2002 17:57:07 +0300, Veli-Pekka Nousiainen
But do you have an idea of how to EVAL/COLLECT in the MATCH command, so 
that
the result for example is ..x^v*x^(2*v)..x^2*x^1.5.. = ..x^(2*v)..x^3.5..
GruÛ
Alexandra
====
The next matrix gets TTRMs or memory looses in my calc and also in Emu48
with MK:
[[ 20.1 15]
 [20.3 38]
 [20.5 76]
 [20.7 57]
 [20.9 14]]
It happens using the matrix editor (I've heard about editor's bugs) but 
ALSO
writing it by hand on the editing line.
I'm ashamed...
====
Aaaarggh! I mean amazed. (My english!...)
====
I had a HP48 3 year ago, when I started university, everybody there
had one, but now I moved to the United States and everybody here uses
TI, so I sold my hp48 (to the only person here who have ever had one)
and bought a TI-89, I know this was a big mistake (dumb of me), I
didn&#8217;t like my TI-89 and I tried to use software to make it like
a HP (RPN, Equation Editor) but it didn&#8217;t work, so finally I
bought a HP49, and is so much better than the TI-89 that i feel alive
again.  I had used userRPL with my HP48 so I made some programs for my
HP49, all they are simple but useful for me, and now I feel I am ready
to take the next step to sysRPL, but now with the end of HP calculator
I have been wondering if it is worthy to learn it? What will happen if
in a couple years all that I have learned is useless because there are
not more HP calculators to program? Should I learn to program my TI-89
instead and keep using userRPL for my HP?
What do you think?
Marturo
Oops! you're right. It seems that I missed the first digit (5) when I 
copied
the factor.
will be implemented on the HP49G, and then we will be able to factor such 
big
numbers 
__
http://www.alpertron.com.ar/ecm.java
You will find a lot of information about ECM at the ECMNET project:
http://www.loria.fr/~zimmerma/records/ecmnet.html
http://www.alpertron.com.ar/ENGLISH.HTM
variables in the current directory.
If these commands were to work with wildcards, should they act on only
the first matching object that they find, or should they check for
additional matching objects?
I'll be darned, I never realized that PURGE would work with a wildcard
tagged name.  A feature to be used with a bit of caution, lest one
purge something unintentionally.
 
The 49G seems to have quite a few minor differences from the 48;
it's just full of surprises.
Did the experiment. I confirm that the HP49G PURGES the user variable.
If nothing with the name 'X' exists on any port, then nothing happens.
But yesterday I did something similar to answer the question of John
And the HP told me repeatedly Bad argument type. Strange, isn't it?
So J , excuse me for the wrong (?) answer in my previous
post. I really tried it out before answering, but something went
wrong, I guess...
Well, you can at least RCL something that resides on the ports.
The command itself is not the problem, it is the wild card that is a
wild card for RCL but not for PURGE. Consistency would IMHO require
that it is always a wild card. (Like * in UNIX. It doesn't care if you
use it with mv or rm or cp and so on) I would like such things to
behave consistently. As for accidentally PURGING something from port
memory, well I don't know what is the problem. If port memory contains
copies of objects that should stay unmodified then don't use the wild
card. If the copies in the ports should also be purged, use the wild
card. If a copy in one port should be PURGED then use :1:Obj .
 
Now, that is a good question. The commands with wild card arguments
were designed to act upon the first found object. If they should stay
that way, then the order of searching ports must be specified. Perhaps
two wild cards are necessary. One that indicates all ports and
current path and one that indicates first port with object found
That PURGE works on wildcarded names, but inconsistently with the way
that RCL works, and inconsistently with the documentation, I strongly
agree is a problem.
If a command doesn't work the same way on the 49G as it did on the 48,
then I could happily accept that, as long as there were some good reason
for the difference and it were well documented. In this case, I don't
see any good reason, and the new behaviour isn't well documented.
If PURGE didn't work with wildcards at all (and the documentation agreed
with this), then I wouldn't see that as a problem. These calculators are
not running UNIX, and I don't expect them to behave as if they were. In
UNIX you can set permissions (and in DOS, attributes) to help protect
against inadvertent removal or modification of a file, but these aren't
available on the calculator, so I wouldn't object to purging and
modifications being prohibited with wildcards any more than I object to
them being prohibited on variables not in the current directory. Must
wildcards work with every command?
The 48 documentation explicitly says that wildcards work with RCL, EVAL,
and PURGE and that the search is of all ports in reverse numerical order
and then main memory. It's not explicit that main memory is searched
starting in the current directory and working upwards through the path
to the home directory, but this is the normal search path, so one would
naturally assume so. So, for the 48, the search order is specified, and,
so far as I know, these commands work with wildcards as documented.
The 49G documentation leaves me wondering which commands wildcards are
supposed to work with, but says that it searches all of the ports (in
which order?) and then the HOME directory (rather misleading, as the
search actually starts in the current directory and searches the path
upwards; HOME directory apparently being meant as inclusive of its
subdirectories etc. in this context).
The problem is that the 49G's behaviour has been changed from the 48's
behaviour without any apparent reason, is poorly documented, and doesn't
match what scant documentation is provided.
Personally, I'd prefer that the current problems be fixed before any new
features were added.
As always, everything I write is IMHO, of course.
is it possible to do calculations with prefixes without having to put
units with it??
You can use the _? unit.
Well I tried some examples and I simply don't understand why the HP49G
behaves like it behaves.
Try the following: Enter 2_k and then enter 3_k again. Press + to add
them. The results is an error: Invalid Unit.
Well, perhaps because k is a built-in prefix for units?
Lets try 2_AAA , 2_AAA + 
It also errors. So we think, the HP49 needs complete and existent
units for its calculations.
But now enter 2_k and 3_k and press * to multiply them. Result:6_k^2
Also enter 2_AAA , 3_AAA and then press * to get 6_AAA^2.
Why is multiplication of non-existent/incomplete units possible while
addition isn't?
Because the multiplication of unit expressions is always defined to be
the product of the numbers and the product of the units, but the sum
can only be calculated with consistent units.  You cannot add 3 apples
plus 5 kids, but you can divide 3 apples between 5 kids.
So when you add two unit expressions, the HP *must* check if they are
consistent, and then it runs into an error.  When it multiplies, it
does not have to check, and so it does not.
Can you tell me example off usefull applications using alphaENTER or 
beteENTER ?
-0400)
Nonsence, the simplified answer is obtained not with EVAL but with 42
So you just couldn't resist, you did it! Now we also have to wait
until the quriosity of the intelligent HP49 also solves the bigest
mystery of this world. Women! 
Though I repeat my self here: *This* is the real advantage of the
HP49G. The ability to customize it, program it and turn it to your
personal problem solver. The question that one should ask is not,
what can I use it for? but is there anything that couldn't be done
with it, one way or the other?. Except time travels, of course.
 
I have now finished the documentation, and the program has been sent to all
that have requested it.
www.hpcalc.org has received it as well.
Smart Chembalance Studio is an integrated tool for chemical
equations.It has three major functions:
1.Quick balance for chemical equations. A complicated equation like
this:
119H2+18Ca(CN)2+6NaAlF4+12FeSO4+6MgSiO3+6KI+2H3PO4+12PbCrO4+24BrCl+6CF2Cl2+2

4SO2
=12PbBr2+12CrCl3+6MgCO3+6KAl(OH)4+12Fe(SCN)3+2PI3+6NaSiO3+18CaF2+110H2O
can be balanced in one second!So,balancing chemical equations will not
be your nightmare but a piece of cake from now on.
2.Quick calculation for chemical equations.It can calculate each
chemical's amount that takes part in a chemical reaction basing an
equation.
3.Quick browse for elements' propertypes.Just one click,you can know
any element's atomic number,atomic weight,melting
point,ElectroNagativity......
No matter you are a chemical researcher,teacher or student,It's a good
assistant for you.Try it!
 
Can you explain me the flag of LASTCMD= (~ -60) ?
-0400)
yeah mine does serial number beginning with CN.
are the 49s being produced in china now having the serial numbers programmed 
in?
because ive only had mine for about 2 months
So, Bernardo, perhaps you are now one of the lucky guys with a
 collection item . And you didn't even search for it.
I have HP49G and I want to run the TGV viewer within the stack
when I type TGV it appears the message
Undefined XLIB name
If I press the CAT key and search for TGV it appears with italic font
 
A wonderfull program was xfrc on hp48 erable, why isn't it on hp49 ?
I have got rom 19.5 , but it bugs : if I do EVAL,LIB it make an
infinite loop . So I reset but it stops on m216takernel loop , so that I
can't load another ROM .
e.g.: int( sin((n*PI*x)/L )dx
gives e complex solution with a lot of es,
but i need it expressed in sin, cos, etc.
i know that it works on a TI92,
but i'm not able to get a useable solution on my HP49.
        
Update your ROM. My ROM gives '-(COS(n*X*pi)/(n*pi)/L)'
Have you read the manuals? See the TRIG menu.
I've now updated to ROM 1.19-5, and now it works well.
        
ESC. BTW, the pictures or drawings are impressive. Let us know how it works
once your brother finish it.
--
Beto
PS: I'm using Acrobat Reader 4.0
when i opened that file with acrobat 5, it displays it in full screen
mode, and you can continue to the next page by pressing enter, you can
exit it by pressing esc and view it normaly as well
He he, that's why I like this group! You get generally educated, not
only calcs. And since music is one of my hobbies, I guess I know what
my next salary will be used for. 
 
-- 
http://users.auth.gr/iilias
As far as I know it is not possible.
Caspar
--
I finally got around writing the docs for Matrix49. I'm sorry for the long
delay :-(
Matrix49 contain the following commands. Most of them are self explanatory.
ADJOINT
MCOFACTOR
COFACTOR
MMINOR
MINOR
MHERMITE
KROENECKER
HAUGMENT
VEC
 
(z+1) / (z^2(z-1))
the result should be: 
(-1/(x+1)) + ((x+1)/(x^2+x+1))
 
simple use PARTFRAC
ftp://fourier.ujf-grenoble.fr/pub/hp48/
 
Yep!  That works just fine.  And it works as well in C.  But in C I
think my way is clearer.  All the error checking is done first and
the processing is done next.  The two are seperated visually.  It's
just a cleaner way to code it.  But I'm really only talking about
C.
I wouldn't suggest a goto for sysrpl.  It would probably be
difficult to implement and not worthwhile.
In Basic or ASM gotos are a natural part of the language and in ASM
they're essential.
I think a statement like Gotos are always bad is similar to
telling a child Don't ride your bike in the street.  It has it's
place but they outgrow the need for it.
 
 without any error checking,
::
  SYMBOL
  xi         @ Didn't try with just ID i
 ;
 '
 ID wb
 x*
 x=
 FPTR2 ^FLAGSYMBEXEC
 '
 ID wb
 FPTR2 ^SOLVE1EQ
;
 However in my opinion it's a BUG, a real big BUG.
Think of a list where a term is represented by a limit; if I want to change
any term accesing the list, when I'll go press ENTER the calc will say
SYNTAX ERROR. I have to split all the list, change the limit in the 
equation
writer or any other term and then rebuild up my list? I think it's almost
incredible!
Can I ask you, if you know, why the RPL interpreter works like this and 
only
for the limit AFAIK?
 
Maybe I can add some additional information.
The RPL compiler is able to handle the = sign in an equation in only two
cases:
1) expr=expr
Here the = symbol must be the most top operator
2)SIGMA(I=1,100,I)
Here it is used to separate 4 arguments, despite the fact that it seems you
entered only 3.
This will be compiled:
ID I , % 1 , % 100 , ID I
LIMIT takes only two arguments.
The way the RPL compiler is done, is that you have to specify a specific
rules for the syntax of LIMIT.
So, if we allow: LIMIT(expr,var=real) then LIMIT(expr,real) will not be
accepted anymore.
As LIMIT can take its argument on the stack, you can still do:
2: expr
1: var=real
then type LIMIT
We found that it is acceptable that you can enter the following expression
in Algebraic mode or in the Equation Writer only.
So unfortunately, that's the way the RPL compiler has been written more 
than
10 years ago. In its fundamental design, using the = anywhere or just in
LIMIT has some limitations...
Maybe if I have more time, I will investigate the subject a little bit
further, but I have already spent several hours on this.
 
Would it be possible to create a new CAS command? For example you could call 

it
LIMIT=. Its behavior would be identical to LIMIT, but would allow the 
algebraic
syntax here desired while rejecting the the LIMIT(real,real) syntax...
 
OK, I've understood.
So why not changing the LIMIT function to behave like
the SIGMA func?. I mean change the LIMIT to handle threee args, even if it
seems we enter only two, like the SIGMA handle 4 args but it seems we enter
only three.
The same was suggestes by Dennis Straley and I think a poll here on the NG
give as result that nobody dislike this change.
Is this to difficult to implement?
OK, here is the point. So maybe you can a function more, like we have for
integration. There we have INTVX which integrates respect to the current
default variable, and RISCH which need the integration variable
specification. You can distinguish LIMITVX (maybe LIMVX would be more
readable) and LIMIT; the first handle two variable (expression and point of
the limit func, and executes respect to the vurrent default variable in 
VX),
the second handle three vars: expr, var, point of limit.
than
Why? Could you suggest some docs about this that I can read?
I think the solutions proposed by me and Dannis would be great to 
implement.
 
terminal screen which explains why everything act in such a bad way...
Unfortunately, the bank0 was originally designed to replace the bootsector
in case we found a major bug in it which would make the calculator 
unusable.
So the bank0 can totally replace the bootsector, and if this one is
corrupted it can become tricky to re-update your calculator again...
I've uploaded the program to format the Port0 of the Flash on HPCALC:
P0ERASE.ZIP.
 
I hope that Cyrille doesn't misunderstand this statement...
Without it it's seemingly impossible to upgrade the 49
with a corrupted bank 0.
 
I was just wondering if there is a way to change all fonts displayed
in the screen of the 49.
For example I want to use homer font but it only changes the stack.
The rest of the screen (header) remains with the same system font.
 
    firstly sorry for my bad english; I recently started learning of sysRPL
programming for hp49 and I have two little question:
1 How I can know the time elapsed in execution of my program ?
2 I' am reading Programming in System RPL by Eduardo Mc and  I wish to
know how I can find the adress of the unsupported commands listed here, in
my hp49
 
sysRPL
Place your program in the stack and run TEVAL, it will return the time
elapsed in seconds.
You can't find them easily and would be bad programming habit using an
unsupported entry point because ROMs in the 49 changes from time to time.
There are some addresses that were told to be stable, but I wouldn't use an
unsupported entry point, I'd go the long way but with the chance of easy
portability.
Marco Tinarelli - Bologna - Italy
tittiXXYYZZ_tt@tin.it   (togliere XXYYZZ dall'indirizzo per
                         rispondere)
 
To find an entry you should disassemble it on the HP48, then put the
disassembly in the program below.
In this code I am looking for an entry that starts with :: DUP #*
The program returns all addresses that contain the above sequence of
commands. Then use a tool like nosy to check if the code at the address is
the same as in the HP48.
* Author: Daniel Lidstr232m
* Compile with jazz/hptools
* Searches for the 15 nibbles specified by the CON statements.
CODE
 GOSBVL =SAVPTR
 D1=(5) #80000
-- LC(N) 16            load C[W] with following
 CON(5) =DOCOL    ::
 CON(5) =DUP        DUP entry
 CON(5) =#*            #*
 CON(1) 0
 A=C W
- CD1EX
 C=C-1 A
 GOC exit
 D1=C
 C=DAT1 15
 ?C#A W
 GOYES -
 CD1EX
 RSTK=C
 A=C A
 P= 5-1
 GOSBVL =PUSHhxs
 GOSBVL =SAVPTR
 C=RSTK
 D1=C
 GOTO --
exit GOVLNG =GETPTRLOOP
ENDCODE
--
http://www.ite.mh.se/~danli97/
...
you are the most experienced expert in building this 
type of programs. Couldn't you deliver a program that
takes the first 15 nibbles or so (better a sufficiently
long start segment of the decompiled string with blanks as
word separators) as an argument, builds an asm-program 
similar to the above one which at once searches through the
49-rom and lists all pointers whose content start with input
segment? That would indeed be of great help for programmers.
For instance, there is an unsupported pointer on the HP48
which takes the next object from the runstream and executes
it via EvalNoCk. Probably there is a corresponding pointer
on the 49 for that and, I guess, even in a stable area. But
I didn't find it. On the other hand, I found a useful stable
pointer on the 49 which does :: TWO NTHCOMPDROP ; (used about
20 times in the 49-Unitman and hence saving more than 50 bytes
but I couldn't find it on the 48 :-) 
 
[...]
I'm not sure what you mean with the string part, why not just use the
program I sent? It's just a matter of changing the CON parts. Anyway, what
should this string look like? Do you have an example?
P.S.
='EvalNoCK: EQU #25F29 * U
At least in rom 19-5.
 
http://www.ite.mh.se/~danli97/
 
Because you gave just one example. Perhaps many users 
(including myself) are unshure whether after having
altered the CON parts, the result compiles with MASD
to a correct code-object (or do you suppose to have 
JAZZ in Port 0? I cannot affort this because I'd to 
reload it after each TTRM :-) 
A unnamed but stable pointer as I guessed, i.e., it
will be valid also in ROM 19-6. Could you shortly 
explain once again how you found it with your
procedure? 
It decompiles to :: 'R COLA EvalNoCK ;
So why not simply to feed a generalization of your 
program with the string :: 'R COLA, or, perhaps 
more easier with linebreakers instead of blanks
::
'R
COLA
which then should yield a list which looks
like { ... PTR 25F29 ... }. Then the user had
not to care on nibbles etc, except that its 
number must not exceed a certain limit.
- Wolfgang
I suppose that instead of TRUE I can put any obj I like. So maybe the
GETEXITMSG's stack diagram reported on Kalinowsky manual (my bible) in
pag 145 is to be changed from
to 
 
 
Yep! That's it. Now, if any TI (or HP) can give this or an equivalent
answer, I'm going to stop working with Mathematica and buy that device
without asking what else it can do. In other words: Will we ever get
the power of Mathematica or Maple or something similar on a handheld
device? (Nick has that dream again ;-) )
 
But the dream is nearer than it seems to be...
Probably Giac (B.Parisse's C++ CAS) or GP-Pari (Number theory 
great software) could be compiled or ported (they are GPL both, 
so the source code is available) to PocketPC.. just, my dream 
now, my Hp Jornada 525 doing advanced math...
Is there any PocketPC developer here? Could you give it a try? 
Avenard, have you ever thought about it? (You have some knowledge 
in WindowsCE programmnin, don't you?)
 
http://www.etsiig.uniovi.es/asociaciones/clubusu
 
http://www-fourier.ujf-grenoble.fr/~parisse/
http://www.parigp-home.de/
 
maple did a maple-like program for PDA's a while ago.they sold it to Casio.
(i don't know if it is any good). Maple said it would run in any device 
with
WinCE2.0 (i doubt it because of the CPU, but anyway it is there). 
http://www.casio.com/accessories/product.cfm?section=155&product=1871&displa

y=14&cid=4119
says...
 
I have a 49g and am having difficulties with local variables and
subroutines. I am trying to create something (in RPN of course) with the
The calc will not accept this as input. Can anyone tell me what I am doing
wrong?
 
What's up with the variable 'C'? Where is this supposed to exist?
 
If you want to STOre the result in a new global variable named C, do 'C' 
STO
two values, one on stack level one and other in stack level two.
 
doing
I am trying to define local variable C (in this case the sum of A and 2 x 
B)
for use further down in the program
 
Sorry for my bad English (and for the silly question :) )...
Is there a way to convert automatically a Real number to its Rational form
(if it exists?)
For example, could I convert the number 0.66666666666666666666667 (of 
course
with some approximation) to the fraction 2/3?
 
abc
 http://www.engr.uvic.ca/~aschoorl/faq/
http://www.finseth.com/~fin/hpdata.html
  
COPY by Rick Grevelle. Available on GD10 and here:
http://www.hpcalc.org/utils/memory/copy.zip
To make an archive backup directly to a computer, use the ARCHIVE
command
with an argument :IO:filename [the file name to store on the computer];
the ARCHIVE command ignores flag -35 and *always* makes a binary file.
If, however, you store your HOME directory in a port, and then treat
that port variable like any other and send it using Kermit,
if you send it in ascii mode (flag -35 clear) then any variables
which contain anything other than pure UserRPL objects will
not be restorable, sorry.  The meaning of ascii file transfer
is that objects are more or less printed out as text-only,
and if they can not be represented as plain text, then there is
nothing you can do to recover the original binary contents later.
Even if you sent the previously backed-up directory in binary mode,
you could not quite even restore it using the RESTORE command, unless
you actually transferred it first into a port (even Port 0 will do),
because the object you would end up with on the computer is
not a backup object, as RESTORE requires, but is instead
a directory object, which does not become a backup object again
until it is stored into a port!
You *can* cut and paste any UserRPL objects back into restorable
form, however, if you pare out any SysRPL/ML objects from the
file which you have saved on your PC; depending on available memory,
it might be a good idea to chop it up into parts if it is very large
and includes many variables.
 I no longer have the [backup] in port 9 ( I know.....big mistake
 to purge things you may need back later on, but now....... :~(
The moral is:  Use a brand new ARCHIVE command to transfer a
brand new backup to a computer; then no wrong transfer mode
can ever occur, and the RESTORE command can actually use it.
 I tried to load the prog ASC- (wich I never needed before) to Emu48
 then load the backup file also to Emu48 and convert it to binary and
 transfer this new file 'back home'.
  ASC- only converts *hex*dumps* of objects that are made via -ASC
  in the first place; printed text can not be understood by ASC-,
===
 
    Is there any way to merge 2 arrays into one?
    Eg: [ [ 1 2 3]  [ 4 5 6]  [ 7 8 9] ]       merge with   [ [.2 .4 .5] ]
                           to get   [ [ 1 2 3]  [ 4 5 6]  [ 7 8 9] [ .2 .4
.5] ]
mine (very heird) to do that.
Here it goes: (in UsrRPL):
                               << -STR 1   REPL SWAP - STR DUP SIZE  
REPL SWAP + OBJ- 
this program is about as lagubrious (?) as yours...
<< - a b <<
a OBJ- HEAD - c <<
b OBJ- EVAL SWAP c + SWAP 2 -LIST -ARRY 
To append [ [.2 .4 .5] ] to [ [ 1 2 3]  [ 4 5 6]  [ 7 8 9] ]
   (a) as a 4th row, do           -COL DROP 4 ROW+
   (b) as a 4th column, do        -COL DROP 4 COL+
To append a 1xn matrix to an mxn matrix as an (m+1)-th row, do
   -COL DROP OVER SIZE 1 GET 1 + SWAP ROW+
This assumes the two matrices have the same number of columns. 
An error will occur if this is not the case.
If you have a G series:
<< -ROW DUP 2 + ROLL -ROW DUP 2 + ROLL + ROW- 
===
What is the reason for the lack of flexibility of the UsrRpl notation when
it comes to declaration of local Vars. I mean, why do we always have to
start a secondary ( << ) every time we declare new local vars ( - a b ).
I know there must be an explanation based on the processor arquitecture, 
but
that's exactly what I'm asking and contesting... What«s the 
logic of that
This is a nearly philosophical question, yet to me there seems to 
be an obvious answer.
You are talking about _local_ variables.  They are local in the
sense that they are defined and usable in one part of the
program, and not in others.  So, in being local to one part of
the program, how does one know which part?  Well, it's the part
bounded by the << - . . .  construct.  Otherwise, how would
one know when that locality begins or ends?  What would a local
variable be without its locality?  (Global, I guess.)
That's right, but there's another reason: You are allowed to use almost any
names for your local variables, unlike globals. You need a mark (<<) to
tell the calc. when the line of local names after - stops (ie the number
of locals you wish to create), and when the program starts. Look at the
following (easy) example, and predict what is on the stack after it has
been run (I know that you know this, Steve):
<< 1 2 3 - a DUP << a DUP ^  + 
as to 'why do we always have to start a secondary, or thirdary, or
fourthary...program delimiter set every time we declare new local 
variables...'
is....
not entirely correct.
you can easily get around this, and treat all your local variables as 
global
variables, inside one big program block, by simply....!!!...defining all 
your
variables, of the local variety at the beginning of the main program...!
you can additionally use the <- ( left pointing arrow ) as a prefix to 
create
traveling local variables...!!!...that will jump from program to program 
while
the ( directory/program group ) is running...!!!
http://www.dawson-eng.demon.co.uk/nexus.htm
==
In an earlier post, you asserted:
UserRPL is a strict subset of SysRPL,
and in another post, someone re-stated the search
for the Holy Grail of the HP48:  Universal translators
for UserRPL == SysRPL == ML
Now, even though HP sometimes slips into such loose talk as this,
there is actually a very great distinction between UserRPL and
SysRPL; the similarities are based on considering how RPL is *executed*,
once it has been compiled, vs. the vastly different original
*source*languages* and *compilers* that each approach uses, each based
upon a very different view of the system and tasks to be performed.
As far as execution is concerned, all RPL is executed identically
(by ML code in ROM, pointed to by each object, rather than by code
within the object), so one can say that since the UserRPL compiler
(in the HP48 ROM, which compiles all the text you type into the
command line) only invokes a limited set of all ROM procedures,
whereas you can invoke all existing ROM procedures via SysRPL,
then what you can *execute* in UserRPL is indeed a subset of
what you can *execute* in SysRPL.  In this sense, there
is no conversion at all between one kind of RPL and another,
for it is all the same.  You can *substitute* some ROM functions
for others that were invoked from a user program, but every time
you make a substitution, you increase the number of situations
in which the converted program might act differently than
the original, or might fail entirely, as illustrated above,
because every substitution changes something, or doesn't
handle something which might in some cases occur, so you need
in general to have more knowledge of the whole problem and
environment than can be obtained by looking at
only one command at a time.
Even some of the most basic assumptions you may make may be violated
some of the time in some user programs, so the more liberties you
may take in any automatic translator, the less certain will be
the success rate of the outcome.
When we look at the *source*languages*, however, we see that
UserRPL isn't anything like a subset of SysRPL; SysRPL *looks*
rather dramatically different to begin with, has considerably
different syntax (if any), and in fact has no means at all
to code some of the things that we may be used to in UserRPL,
such as formulas that look like formulas, arrays, and directories
('formulas' can only be represented as their RPL equivalents,
arrays as hex strings, and directories, well... ?); in return,
you get access to functions which can enter or exit programs
at any old place, and violate many other user restrictions
with abandon (plus some risk), and you suddenly get some brand
new object types and a whole new stack to have to deal with
(e.g. long reals, bints, and the return stack).
Thus, when you read RPLMAN.DOC (the HP introduction to SysRPL),
or even any other book or guide, you probably find about
as much resemblance to the programming language of the
User's Guide and AUR as the landscape of the Moon
bears to where you live.
The whole environment structured by UserRPL is problem-oriented,
concealing much of the internals, even the fact of how programs are
structured, whereas SysRPL is the raw world of the ROM jungle --
more powerful, but more dangerous, and it's entirely up to you
to get through it without your memory being eaten alive;
in addition, the functions invoked by compiled UserRPL
perform for you certain things you take for granted,
such as the unrestricted scope of branching words,
the checking of arguments, *and* even the ability
to cancel your program with the CANCEL key.
Because of the great differences in the fundamentally presumed
environment between the two languages (and who would originally
use them), one might well take completely different approaches
to solving the same problem, depending on whether one would employ
UserRPL or SysRPL, just as the best route between two points might
take completely different courses when traversed by means of
an airplane vs. a truck, a ship, a railroad car, or skis,
and the most dramatic improvements between UserRPL and SysRPL
usually come about when the basic approach is restructured, rather
than when one plods along, just replacing equivalent commands.
This is why I originally said that the undertaking resembles
in some way the problem of translating spoken languages.
At first, the idea of just translating each word through
a dictionary sounds like a very simple solution; then as
more and more multiply-defined words, peculiar linguistic
structures, and idioms creep in, more and more of the
sense of the original gets distorted in any limited
attempts at translation.  It should of course eventually
be possible to create a translator of any desired level
of skill, and even of great ingenuity and creativity,
just as it is possible to construct a chess-playing program
which can defeat a world champion, but at a corresponding cost
in complexity and effort, so the issue may be: how much work
do you want to invest up-front, vs. how easy it may be
to just use your own brain and write your own SysRPL
when you find a candidate user program
worth re-casting into the mindset of SysRPL.
Some things that user commands do which many internal
system commands don't, from which most of your gain in
command-for-command replacement derives, and also
from which many of the benefits of UserRPL derive,
producing the inevitable trade-off:
o Save the command name (for automatic error messages)
o Save the command arguments (for errors and LASTARG command)
o Save the stack depth (for automatic stack cleanup after errors)
o Check the argument types, and in many cases choose different
  internal system functions, according to argument types,
  positively preventing crashes by checking on every command,
  rather than just checking original arguments.
o Abort if the CANCEL key has been pressed ( < 1_ms per test );
  generally tested only by program structure commands.
Sometimes there is a savings worth the investment, and sometimes
it may not be worth the time spent recovering from the first crash;
it's a matter of your own judgment for your own programs and purposes.
BTW, it may be that someone new to the HP48 asks for a translator
for UserRPL == SysRPL == ML, based upon the assumption that
UserRPL text gets re-interpreted every time it is executed,
rather than the fact that it gets *compiled* just once,
at the time it is first typed in, just as SysRPL gets compiled
once into the same kind of universal RPL object, and then both
sorts of RPL are executed in exactly the same manner thereafter;
this means that the potential execution time savings are much less
than might be expected, if the translation is just one-for-one,
replacing protected user commands by what they internally execute.
As to SysRPL == ML, everything that is executed in the HP48 is
always ML, but much of that ML is in ROM; if what you need to do
is not available at all, or is not efficiently performed by the
existing ML in ROM, or if it consists of many basic operations
which can be done in CPU registers, without using memory,
then there may be a lot to gain by writing your own new code,
but if there already exist ROM functions that efficiently do
exactly what you need, then you can write a small program to run
a huge amount of pre-written ROM code, and save yourself the trouble
(e.g. compare the almost-identical VARS and VARS2 commands in the
Hack library, one using a lot of ML instead of shorter SysRPL).
If you disassemble some HP48 ROM functions, you may even see that
HP may well invoke some *user* functions in its own SysRPL coding,
if it turns out that this can satisfactorily perform the desired
function without the need for any longer a program.  This may
have been motivated in part by the need to fit all the functions
into a finite-sized ROM, and also to get the product to market
in a finite time, but some of these same considerations may apply
even to your own everyday programming.
Well, with some sort of new product being always thought to be
around the corner, at least for the past several years, and with
the vast pile of software for everything in creation already
collected by Eric Rechlin into one monster site; investments
in large programming projects may happen to trail off a bit,
but so long as ungreased wheels need re-invention,
there may yet be more around the corner.
==
 
 
 There seems to be a memory leak in this program. If someone could point
 it out to me i would be grateful. ( I have a feeling im making an
 obvious mistake. Oh well... )
 
 --------------------------------------------------------------------------
 
 ASSEMBLE
 
         NIBASC /HPHP48-Z/
 
 unscroll        EQU     #13695
 abuffdfltdim    EQU     #130CA
 *wait            EQU     #1A7ED
 
 RPL
 
 ::
 
  ClrDA1IsStat
  ZEROZERO  DOLCD XYGROBDISP
  TURNMENUOFF
 
  ABUFF SIXTYFOUR HEIGHTENGROB
 
  TOADISP
 
  ::
  65 ZERO_DO (DO)
 
         ATTN? IT :: ZEROISTOPSTO [LOOP] RDROP ;
 
         INDEX@ ZERO WINDOWXY
         150 ZERO_DO (DO) LOOP
 *        HXS 10 7C00000000000000 PTR wait
 
  LOOP
  ;
  DEFINE ay  4GETLAM
  DEFINE vy  3GETLAM
  DEFINE  y  2GETLAM
 
  DEFINE  aySTO 4PUTLAM
  DEFINE  vySTO 3PUTLAM
  DEFINE   ySTO 2PUTLAM
 
  % -0.5  %0  % 65   THREE ' NULLLAM CACHE
 
  BEGIN
 
         ATTN? NOT
 
  WHILE
 
         vy ay %+ vySTO
         y vy %+ ySTO
 
 
         y %0 %<= IT
         ::
                 vy %CHS % 0.925 %* vySTO
                 %0 ySTO
                 vy ay %+ %0< IT :: 3RDROP ;
         ;
 
         y COERCE ZERO WINDOWXY
 
         FIFTY ZERO_DO (DO) LOOP
 *        HXS 10 1500000000000000 PTR wait
 
  REPEAT
 
  ABND
  FLUSHKEYS
  ATTNFLGCLR
  TURNMENUON
  PTR unscroll
  PTR abuffdfltdim
 ;
==
i've had this idea for a long time of creating a program that would somehow
emulate the process of evolution...???
The idea in it's various conceptual forms was to allow a computer to create 

a
picture while humanimal(s), Acting as the force of nature...or the 
statistical
chance of survival, Would determine which elements of the picture would 
survive
and which would die off.
Then--A psychological facet would become active...As the humanimal 
controller
picks surviving elements over and over again, Working with psuedo-random
forces, The picture would begin to resemble graphical iconography that the
controller subliminally projects onto the evolutionary image. (?)
Well...that's what i was thinking...
Regretfully, The resolution on the 48 may be the restrictive factor that 
makes
the subliminal image all the harder to pick out...??? (!)
Anyways--
This is how the program works:
The Directory consists of 4 Objects
Start   The Evolutionary Program
Save    Saves PICT
Dot     Returns a graphical morsel
PPAR    Plot Parameters
'Start' may take either the GROB from Save, which was automatically saved 
when
you quit your last session with Evolv, or nothing on the stack, in which 
case,
a new PICT template is created.
---------------:: <
As a technical note:
This bit of programming requires a small amount of cleverness, As the 
program
is asking 3 questions;
1) Is there anything on the Stack?
2) If there is something on the Stack, what is it?
3) If there isn't anything on the Stack, what should i do?
Included in question (2) is the additional question, Is it the kind of 
thing
that i'm looking for & is that it...?
The solution i used was this:
Since i was looking for a GROB ( TYPE 11 ), 
The routine first asks if there is anything on the Stack, 
If there is, Then it TYPEs it, 
If there isn't anything on the Stack, 
Then it returns a 99--as a generic solution...
Since there are no Type 99's...
Thus, No matter what i'm checking for, It would be equivalent with it not 
being
there...!
Then it continues with the program solution for a GROB or NoGROB.
---------------:: <
Once the program starts, It delineates two pictorial regions, And adds a
graphical element to each, You then pick one or the other, Or neither.
If you pick neither, Another graphical element will be added to both.
If you pick one over the other, Then the selected picture will be copied 
over
to the other side, And the process begins over again.
Once it gets going you hit either the first 3 White Keys; A, B or C, to 
select
the image on the left side, 
Or the latter 3 White Keys; D, E or F, to select the image on the right 
side.
How you choose which image is entirely up to you. 
Choose it for it's aesthetic balance, or because it looks like 
something...?
If you hit any key other than the White Keys...
Except for the DEL, BACK or ON Keys, Which will terminate the program,
...All the other Keys will add graphical elements to both sides of the
Evolutionary palettes without killing one side off.
After allow several hundred generations to be born, struggle and die...
You might want to allow your therapist to see the results.
Directory Name:    Evolv
Checksum:        # F818h
Size:               1178
 
%%HP: T(3)A(D)F(.);
DIR
  Start
    << DEPTH
      << DUP TYPE
==
      << PICT STO
      << ERASE {
# 0d # 0d } { # 64d
# 63d } BOX { # 66d
# 0d } { # 130d
# 63d } BOX
# 0d } PVIEW
      DO { # 0d
# 0d } { # 66d # 0d
} 0 1
183 pi
          << RAND
{ GXOR GOR } RAND 2
* CEIL GET EVAL
        NEXT
        DO
        UNTIL KEY
        << { 0 54 55
} 162 POS
          << PICT
RCL 'Save' STO 1
          << { 1 2 3
4 5 6 } 162 10 - POS
            << PICT
{ { { # 0d # 0d } {
# 64d # 63d } } { {
# 66d # 0d } {
# 130d # 63d } } }
162 10 - 3 / CEIL GET
EVAL SUB PICT {
# 0d # 0d } 3 PICK
REPL PICT { # 66d
# 0d } ROT REPL 0
IFTE
      UNTIL
      END
  Save 0
  Dot
    << {
GROB 7 7 C1E3F7F7F7E3C1
GROB 7 7 F7F7F7F7F7F7F7
GROB 23 23
00F7000CFF100EFF308FFFF08FFFF0CF1CF1E700F3E300E3F300E7F100C7F100C7F100C7F1
00C7F100C7F300E7E300E3E700F3CF1CF18FFFF08FFFF00EFF300CFF1000F700
GROB 15 15 0E308FF0CFF1EFF3EFF3FFF7FFF7FFF7FFF7FFF7EFF3EFF3CFF18FF00E30
} RAND OVER SIZE *
CEIL GET DUP SIZE 2
  PPAR {
(-6.5,-3.1)
(6.5,3.2) X 0 (0,0)
FUNCTION Y }
END
===
 
For what I know, there is only one that allow you to continue to work with
line with a new one much more powerful.
http://www-miaif.lip6.fr/geraldenglish/mk.html
==
I can confirm that the component is capacitor. There is actually a
100K resistor connected between the HC174 reset line and VCC, and this
capacitor connected between the reset line and ground.
This timing circuit resets the HC174 when the calculator is powered
on.
I opened several HP48GX's, and all had a capacitor. And I have
upgraded more than 20 HP48G calculators to date, always using 100K and
100nF, with no problems at all.
I talked to the author of the document that described upgrading the
HP48G to 384K and above, stating that his document was wrong at this
point. The mistake was confirmed, but he said he never had a problem
in any of the many calculators he had upgraded. This is probably
because two 100K resistors will set 2.5V fixed at the reset input of
the HC174, which will be accepted as a HIGH or inactive state. As the
calculator probably resets the latch during the boot-up sequence, the
reset produced by the timing circuit is not really required. Note that
this 2.5V level will cause excessive current consumption by the
74HC174 when it's on, according to 74HC series electrical
characteristics.
Of course, anyway the right way is to solder a capacitor. The exact
value is not critical, and 22nF-100nF should do.
And now that we are talking about hardware, I'll post some information
I learned from upgrading so many calculators, and which I think is
very IMPORTANT:
In most of the upgrade documents around there, the card detect pins
(pin 37, counting VCC as pin 1) are connected, through a switch to
ground or the card slot VCC in order to disable/enable memory
protection. When switches are not used, these pins are connected
permanently to the card slot VCC.
Note that the card slot VCC is not powered when the calculator is off.
Also note that the card detect pins are inputs which recognice three
states: low, high, and open. In order to detect the open state, the
HP48 chip contains a voltage divider using two resistors which set
these pin's voltage to 2.,5V when they're open.
If you connect these pins to the card slot VCC, then when the
calculator is OFF, you're powering ALL the circuits connected to this
VCC from the internal calculator VCC (which is always on to retain RAM
contents) through the voltage divider.
The internal resistors that form the voltage divider have very high
values, and so the current that can flow from permanent VCC to the
card slot VCC is very low, but still enough to allow about 1.5V in the
card slot VCC WHEN THE CALCULATOR IS OFF.
The consequence is excess of current consumtion when the calculator is
OFF. The current is still very low, and will result in still very long
battery life, but it is high compared with current consumption if
things are done properly.
The SOLUTION is obvious: connect these pins to permanent VCC instead
of card slot VCC (no mather if you do it directly or through
switches). 
As I mentioned before, I've upgraded more thatn 20 HP48G to 384K and
768K. When the card detect pins are connected to card slot VCC, and
the calculator is off, measured current consumption varies from 10uA
to 100uA. When the card detect pins are connected to permanent VCC,
current consumption was not measurable with my instruments, which
means it was below 1uA.
==
Whenever you create a UserRPL program which calls a library function,
the library has to be present at the time you edit (or download)
the program, so that the built-in compiler can find the function
by its name, embedded within the library, and then translate that
name into numerical references (library number, function number).
If you display the original program at a later time, when the
library is not present, the original library function name
then appears as XLIB xxx yyy, because of the missing library,
or an Undefined XLIB Name error occurs when running the program.
If you suspect that a library might not be present when you
create or download a program which uses it, or that different
versions of the library may be re-arranged to have different
internal function numbers, you can use a strategy which
defers looking for the library function until the program
is actually executing, e.g. to execute a function
If the library is not present when you execute this, then
the result will probably instead be the name 'XXXX'
left on the stack; if you want to go further,
testing whether or not the program was found,
you might instead say  XXXX EXEC
Where EXEC is a program like this:
The above program also handles the case where the function XXXX
sometimes might even be stored in an ordinary variable,
or might at other times be present in a library;
it trades a small amount of extra execution time
for a greater amount of flexibility
in the configuration of your programs and libraries.
==
Let A(n,n) a square matrix. We want to calculate eigenvalues and 
eigenvectors.
Eigenvalues are the roots of the characteristic poly of the matrix.
If you rename eigenvalues as: a[0]..a[n-1], eigenvectors have to solve:
(A-a[i]*I)*x[i]=0, for i=0..(n-1). x[i] is the eigenvector of a[i] 
Where I is identity matrix.
Example:
finding eigenvalues and eigenvectors of
A=[[2 3]
   [4 6]]
det(A-aI)=det[[2-a 3]
              [4 6-a]]
=(2-a)*(6-a)-12=a^2-8*a (characteristic poly).
roots are: a0=0 and a1=8
eigenvector for a0:
[[2-a0 3]  *[[x0]  =[[0]
 [4 6-a0]]   [x1]]  [0]]
=
[[2 3]   *[[x0]  =[[0]
 [4 6]]    [x1]]   [0]]
=
row1: 2*x0+3*x1=0
row2: 4*x0+6*x1=0
solution is in the form= x0=(-3/2)*x1
for example, an eigenvector for a0 is [[-1.5 1]], or in general,
alpha*[[-1.5 1]], where alpha is any nonzero value. 
Soving in the same way for a1, you get
-2*x0+x1=0
so, an eigenvector for a1 is [[1 2]]
if you enter in your 48G(X):
[[2 3][4 6]] EGV
you get:
1: [8 4E-14] i.e. [8 0] : eigenvalues
2: [[.5 1][1 -2/3]] :eigenvectors for 8 and 0, respectively.
==
 
==
read the complete story on this web page:
http://science.nas.nasa.gov/Pubs/NASnews/97/05/hexpi.html
from the address
        http://www.cecm.sfu.ca/~pborwein/PAPERS/P130.ps
For those wanting to program some PI algorithms in assembly language,
here's a nice candicate by Dik T. Winter:
     int a=10000,b,c=2800,d,e,f[2801],g;main(){for(;b-c;)f[b++]=a/5;
     for(;d=0,g=c*2;c-=14,printf(%.4d,e+d/a),e=d%a)for(b=c;d+=f[b]*a,
     f[b]=d%--g,d/=g--,--b;d*=b);}
==
Also, does anybody have a small program (sys/user RPL) that returns the
coefficients of a polynomial written out in symbolic form, ie for 'x^2 +
2*x - 1' it would return [1 2 -1]?  Preferably it would handle symbolics
- then you would also pass a variable, I suppose.
 
I did this a while ago and had posted it to this forum.  At the time I
was using GROBs for the labels, but text is easier to do, I think.  I
did it in User RPL by making a self-modifying entry in the CST list.
Shown below is a self modifying 4th entry of a CST list with commments.  
{
 OFF                         @ Current label
 << 
  PATH                                  @ Save path 
  HOME CST                              @ recall contents of CST
  DUP                                   @ Save a copy of CST 
  4 GET                                 @ Get 4th item of CST (this
item!)
  IF 5 DUP FC? THEN                     @ Check if user flag 5 clear (1)
   SF                                   @   If so, set it and leave
   ON                                 @   Label #1
  ELSE 
   CF                                   @   If not, clear it and leave
   OFF                                @   Label #2 on the stack
  END 
  1 SWAP PUT                            @ Replace label with new one 
  4 SWAP PUT                            @ Replace 4th item in CST (4)
  'CST' STO                             @ STO CST to save change 
  1 MENU                                @ Goto CST menu (new label
wont
                                          show up until you press CST
                                          without this step) 
  EVAL                                  @ Return to original path
}
This actually isn't horribly slow.  All you'd need to do to is change
labels, the flag being checked (1) and the position in the CST list (2)
to suit your needs.
Enjoy,
Pat
 In sysrpl you can do this pretty easily. Here is a program that puts a
box beside EQ if the EQ variable exists. Try inserting the following
list into your CST list:
{ 
 :: TakeOver $ EQ ' ID EQ SAFE@_HERE DUP IT SWAPDROP Box/StdLabel ;
 ID EQ
}
That is to say, you want your CST list to look like this:
{ A B C D ETC
  { 
 :: TakeOver $ EQ ' ID EQ SAFE@_HERE DUP IT SWAPDROP Box/StdLabel ;
 ID EQ 
  }
 D E F G H ETC
}
I.e., don't do THIS:
{ A B C D ETC
 :: TakeOver $ EQ ' ID EQ SAFE@_HERE DUP IT SWAPDROP Box/StdLabel ;
 ID EQ 
} 
==
   Is there a wildcard in sysrpl I've looked through James Donnellys
 book
   and Eduardo Kalinowskis book and cannot find anything about it.
   Any help would be greatly appreciated. Thnx in advance.
 
  Wildcard for what?
 
  For port access, the ampersand (&) is the wildcard character.
  For example, whereas :1:FOO RCL only searches for FOO in port
  1, :&:FOO RCL searches all your ports for FOO.
 
  For algebraic expression match-and-replace operations using
  the upMATCH and downMATCH commands, any name beginning with
  an ampersand (e.g. '&A') acts as a wildcard.  For example,
  whereas { 'SIN(X)' 'COS(X)' } upMATCH searches the algebraic
  object on level 1 and replaces all SIN(X)'s with COS(X)'s,
  { 'SIN(&X)' 'COS(&X)' } upMATCH replaces all SIN's of
  *anything* with their COS.
 
  There is no wildcard feature for general object names.  For
  example, there is no single built-in way to PURGE all
  objects whose name begins with 'A' (and not purge anything
  else).  Such operations would require a program.
 
  If you have any other meaning of wildcard in mind, please
  advise.
 
  Meanwhile, please note that wildcards is in your manual's
  index.  Please look it up and read the pages shown there.
 I know how to use it in user rpl with MATCHup MATCHdown. I was thinking
 of
 trying to do some experimenting with SYSRPL with (I hope I've got this
 correct matchob?) which is in SYSRPL. I want to be able use it much the
 same
 as in user RPL, that is matching an object to get back a certain object.
 Ex:
 { &a 'a*x'} or even
 { 'x^&n' '(x^n+1)/(n+1)' }
 I was thinking of making my own integral tables using pattern matching.
 I
 believe it would be too slow and cumbersome in user RPL but what about
 SYSRPL. I am just curious if the ampersand is the same in both. Some of
 the
 other symbols are not for Ex: Binary nos. in user RPL use #14bc5h
 in SYSRPL <14bc5 is used.
As in my previous post I was refering to SysRPL not UserRPL and the user's
guide does not cover anything remotely resembling SysRPL.
 
Does it have a built in Determinant
function?
Yes.  On an unaltered G series, it is under MATH MATR NORM next, DET.  Or 
just
type DET to get the determinant.  
What I'm trying to do is calculate the inverse of a 3x3 matrix
For the inverse, just put the matrix on the first level of the stack and 
then
press 1/X.  Note that this doesn;t work with matrices whose determinate is 
0.
For symbolic matrices you will have to use either ALG48 or Erable ( there 
may
be other, smaller programs that do only sumbolic matrices, but these two 
are
the best and have many more useful features.
Was there ever a mini-challenge in UserRPL to find the prime factorization 
of a
number?  One of the first things I did when I learned how to program
(well...thought i learned how to program) my HP was write two programs in
UserRPL that found the prime factorization of a number on stack level one 
and
returned its prime factorization in a list on stack level one.  Has anyone 
else
done this?
Yes, in 1990.  It was one of the first mini-challenges.
My entry can be found on Goodies Disk #1, as 'PRIME'.
A much faster version by Mark Adler can be found on GD2
as 'FACNUM'.
Both of the above use the simple sieving algorithm, so they
are very slow if the smallest prime factor is large.  Much
faster algorithms exist, but to my knowledge nobody has
coded them *in User RPL*.
Is the fastest ML factorizer the FACTOR command in ALG48?
It factors (10^17-1)/9 into 2071723 * 5363222357 in 5.3
seconds, and determines that (10^19-1)/9 is prime in 4
seconds.  I don't know of any faster *HP48* factorizer.
In this case ALG48 uses Pollard-Rho, which is indeed quite fast for
such smallish factors. A method which often is equally fast and
occasionally a lot faster is Pollard p-1. However, in this case
        2071723-1 = 2*3*17*19*1069
        536322257 = 2*2*17*389*202753
so that Pollard p-1 would be fairly slow due to the large last factor
(1069). Since Pollard p-1 requires calculating powers by just about
all numbers until 1069 (ignoring the possible optimizations..), it
seems obvious that the algorithm would be slow. However, I'm not
really sure whether it would take more than 5.3 seconds or not.
As for other Pollard-Rho implementations, the one in ALG48 is
heavilily optimized, so taking a large enough test set I'm confident
no implementation is significantly faster. However, for any single
test there can be significant variations based on a lucky choice of
the initial values in the algorithm.
Also, ALG48 only performs the Miller-Rabin primality test, which is
probabilistic. If one uses all prime bases in succession, then some
limits are known. For example, if all prime bases 2,3,...,19 are used,
then it is known that if the number passes the test in all those
bases, it must be a prime if it is less than 341550071728321 [1].  If
one continues up to base 31, the first known failure is the number
56897193526942024370326972321, which factors into 137716125329053 and
a 15-digit number. However, it is *not* known whether any smaller
ones exist.
is a vanishingly small probability that the number isn't a prime after
all. Actually, in this case the limit is incorrect, since when the
number does not fit into 64 bits ALG48 uses a SRPL version of the
Miller-Rabin test. For speed only bases up to 11 are tested, in which
case the first known failure comes at 2152302898747.
That is, PRIM? returns 1 for the number in 4 seconds only because
it doesn't want to waste more time to make a vanishlingly small
probability of failure even smaller.
==
==
I'm not sure if factoring is the proper terminology,  but here's what I 
want
to do...
If I have a rational polynomial in the form of (x^3+3x^2)/x^4,  how can 
I
get the calc to present it as 1/x+3/x^2?
danny
====
X
X
<<
{ # 0d # 0d } PVIEW
# 0d # 63d FOR j
{ # 0d } j +
{ # 131d } j +
LINE
NEXT
0. WAIT
 
====
After looking into EFnet(#hp48) and looking that everything was quiet, I've
made my #hp48 channel on BrasNet.org(Brazilian network). To connect,
us.brasnet.org or eu.brasnet.org(Europe). Then join #hp48 and speak with
coerce(me)!
====
 
Take oo as
        a
lim     -
where a is any (real, haven't had the complex ones yet =) number. So for
oo-oo you have
      / a            / b 
lim   | - |  -  lim   | - |
The limits can be put together:
       a - b 
lim    -----
Now you can see that
(1) if a-b=0 then it is a singularity (undefined)
(3) if a-b<0 then you get -oo
I hope everything's correct. You can use limits to prove all these
undefineds =)
Greetings
Thomas
-- 
If you cannot convince them, confuse them.
    -- Harry S. Truman
====
Absolutely! And so the problem is not that oo is itself undefined, but
rather how quickly the terms a/x and b/x fly to infinity :-)
please? )
Greetings,
Nick.
====
I'm studying Industrial Engineering (Mechanics and Structures 
specialization)
In the first and second course (Basic calculus and algebra, ODE, and PDE(I'm 

still stuck with this one!) calcs are banned in exams, even non-progs or 
Mickey 
Mouse calcs.
But in advanced math classes in the 3rd, 4th (and 5th when I reach)courses 
they
can be used, I took my 49 to the Numerical Methods in Engineering final 
exam.
And I can't take it to de PDE exam in june... How could I pass :(
====
Greetings,
Nick.
====
Well, what software will you use to write your euation on the PC?
Since each math software has its own way to code such data, you must
have some translator on the HP, that converts, say, equations from
Mathematica format to HP49 equations. Of course if you just write your
equations on the PC using an ASCII editor and following the HP49G
conventions, then the equations should be also readable on the HP49G.
Greetings,
Nick.
====
You can use the command TAYLOR to get partial sums of the desired
order. Example:
SIN(X)
X
4
TAYLOR
to get a TAYLOR series of degree 4. The resulting polynomial can be of
course plotted. You can also plot all taylor polynomials up to a
certain degree, if you do something like
<< 1 4 FOR I
   'SIN(X)' X I TAYLOR
   NEXT
   ERASE DRAX DRAW
The variable EQ will then contain a list with all taylor polynomials
which will be plotted with the command DRAW.
Greetings,
Nick. 
====
outdating is reported in rfc-index.txt
-- 
(By Dr. Greg Wettstein, Roger Maris Cancer Center)
====
I would bet any CEO who buys a company with a 1.5 billion dollar
liability, would have to dump anything not rally profitable to bail
her whims out.  Most of wall street is betting against her as are
the H.P.  families.  They seem to think it is better to do what you
are good at rather than buy competition because you can't beat
them.
Don
====
If those golden years are over, let's make our own golden years.
Greetings,
Nick.
====
Nick, I like your optimism.  Let's learn the lessons of this fiasco and 
move
on to a new and brighter future... how can we support it?
Greg
single pixel width lines of about 100 pixels in length at the rate of 
two
193
====
check the help text. One of the multimedia controls is NOT
distributable. I believe you have to get the media player as a
seperate download and install it first.
====
Have you any problem about Visual basic?
You want ask somebody?
Go to www.tutorialbox.com
P.S. All this is for free so please click on our banner!
====
I am since a long time a visual basic programmer
I find this product very interesting. I can execute interpreted code
directly on my PC microprocessor
http://www.hyperpanel.com/boot_direct.php3?ID=13240502
====
I'm sure Keystone would sell them to you.  Check out
http://www.keystonelearning.com/
====
text format.
The programme displays them to view only, not for printing out. The trouble
is anyone could go to the directory where they reside and open them, then
print them out.
This would not be desirable at all, has anybody got any suggestions.
====
When the files are saved in Rich text format, they are not Word files 
anymore.
If you want to restrict access to these files why don't you not save them 
in
Word format and use a password. You distribute the password to valide 
users,
who would be able to access the files.
Chris & Kathie a .8ecrit :
====
Need Senior N tier VB developer with Citrix metaframe
REQ: Senior N tier VB developer with Citrix metaframe
Skills: Visual Basic v.6 Ntier development, Citrix metaframe
platform-must , SQL 2000 Database, POS (Point of Sale) Apps experience
a must
Duration: 3 mos + 
Location: New Orleans, LA 
Positions: 2
Others: Citrix AND POS (Point of Sale) are MUSTS
Please send rate / resume / availability to
pradeep@idsysinc.com
Pradeep
Vice President - Business Development
Intelligent Decisions Systems, Inc
Fort Lee, NJ
www.idsysinc.com
T: 1 206 280 3636
F: 1 425 228 8519
====
Works on my machine but when I distribute it, the user gets programe has
caused an error in MSDXM.OCX
PLEASE ANYBODY GOT ANYS IDEAS?????
3837
====
how i can send files from the PC to the calc using the xmodem?  if the
only i can see is the contents of the home directory?  or it only
works to receive?
====
...and if you have a 49 or 48 with MK, you can write a program called
STARTSEND for transfering directly from the Filer using XMODEM instead
KERMIT.
====
Use Emu48, put the object in stack and press XSEND. In your calc, put a 
name
(as 'A') in stack and press XRECV.
====
HPhreacker.
====
oops, the last link was suposed to be
http://www.icircuits.com/osd-232.htm
====
you could make one cheap
http://www.airborn.com.au/serial/sertopar.html
or buy one ready made (needs 25 to 9 pin adapter)
www.jameco.com part-number 182983 $55 
you could get one cheaper looking on the net. i think they should work
but i haven't tried myself
also, you could connect your HP-calculator to the television and
record a video :)(you need to send characters), maybe you would do
that better with a PC and the EMU.
http://www.airborn.com.au/serial/sertopar.html
====
Everybody can check the following page:
http://securityresponse.symantec.com/avcenter/venc/data/w32.klez.removal.too

l.html
====
I had installed Norton Antivirus 2002, I'm subscribed to Live-Update
and that wasn't enough. A W32.Klez.XXXXX virus mutation ripped up the
auto-protect system.
I'm angry.
Max.
====
antivirus program since it's the only one integrating
====
My NAV2002 has cought all the W32.Klez viruses I've received, so it works
fine.
====
this is a little extrange question, but can anybody try this:
use a keyboard (foldable) like targus to enter characters into the
emu48/49 of a PDA. use enter to execute command and obtain a result.
does it work?
i don't have one of those keyboards myself but i am thinking on making
me a home-made one and using software like winwedge for CE.
thank you much
====
never
I agree. I remember it as something infuriating...
====
Yesterday while I was perusing 'User's guide' on page H-28 I found
SYSEVAL command. It says: Evaluates unnamed operating system objects
specified by their memory addresses. I was interested how it works
and I typed #54h, put it in stack and then typed SYSEVAL. After that
the calc began to do something for long time. I understood that I had
made mistake and I was in panic. I didn't know what to do, I wanted to
stop it ( system halt ) but I couldn't remember how to do it. I was
searching the book all in rush, but I couldn't find it. Then I
remembered some combination of keystrokes: ON-A-F. I did it, and it
asked me :Try to recover memory?. I said YES. And then it was doing
something for very long time and I had to do something. You people,
who use Windows OS, will understand my logic. I removed the
battery-compartment door and quickly pulled out one battery and turned
it back again. I looked at grafic display and it was completely black.
Then I did it again ( You know that it sometimes works with Windows ),
but this time there was nothing on display except translucent
sandwatch. It didn't react at all. That was 10 hours before, and now
it behaves in same way. It doesn't work ( only sandwatch ).  
 
====
Have you tried answering NO when TTRM? or pulling out the batteries for a
couple of hours?
Don't use SYSEVAL in this way (I hope you've learnt it yet) This commnd
evaluates system objects (runs ROM programs) and if you enter in a random
adress...
====
you  can pull the batts out and press the ON bottom to drain the
capacitor that holds the cahrge.
off-topic: i am test driving the sharp zaurus and i have to day that
things have gone a long way since my philips nino (winCE). this sucker
can do C/C++ java. maybe i'll be able to put Parisse code too an
compile on machine. nice keyboard too!
back to the shark
 
====
Perhaps you could try the reset hole under one of the two upper rubber feet 

(the right one is marked with a 'R'). For me it worked fine to get out of 
the recover memory routine (removing batteries did not work). See manual 
page 5-18 (according to the german version; I don't know about the other 
languages) for details.
Manuel P.
 
====
Well two out of three then (and one the right way)
====
???
I don't think this is a good idea.
He'd use the pinhole for reset instead.
This works in most cases, and there's no risk of damging the calc.
I think this issue is in the HP-48 FAQ...
Raymond
====
Well I have done this on all my HP calculators
and never ever has the working power come out
(blondes say that they work on blue smoke
and went that leaks out, they won't work anymore
that=electric equipments)
PS: Others don't have to test they protection
    diodes this way....
====
NO, don't do that! You risk frying the power loop diode in the battery
regulation.
====
some time ago i used powerplot on my HP49, and it is a fine program. But i
think there is a bug in it.
it sems to have some influence on my key assignments. they did not work
after i used powerplot. Have anybody elles noticed that?
====
The 48 version of PowerPlot utilizes user flag 1, but it is designed to
restore that flag when you exit the program.  If for some reason the 
program
crashes in the middle, it is likely that this flag will be set.  Other than
that, the 48 version didn't make any key assignments.
Aaron
====
I tried it this week.  I got an hour glass on all the time after a plot.  I
found some operations very slow after a graph.     I then got a lockup
trying to use zoom and had to take the batteries out.
Danny
====
I'm not surprised to hear that there are some bugs.  I'll send your reports 

to
the guys who did the port.  As I said, I'm not sure if he's still working on 

it
or not.
Danny, which version are you using?  HP49 version or one of the 48 
versions?
 
 
====
Is there a posibility in UserRPL to collect/eval the whole or only some
terms in the converted-expression within the MATCH command?
 
====
You have to operate on parts, I think.
So the answer is: no - not in ONE step.
Remember also that there is IF clause available
for both MATCH|^ and MATCH |v
====
I'm new in ASM programmation, and in this newsgroup too.
I'm looking for a while about a little asm program that return a number
stored in a register on the stack.
Like :
2:
1:
number EVAL number RCL
2: 25
1: code
So please, excuse-me for my ignorance, I'm very new.
I use HP ASM v1.0 (by J.Y Avenard). Jazz is too big for my little HP48GX
(I've a lot of other program and no card... :-((
I've know how to put a new element in the stack, and I've seen how to
reserve memory for a string, but I don't know how to reserve memory for
other type.
thks.
====
I understand perfectly that YOU do like the behaviour because that's
the way you designed it - for you. But since I'm not the only one to
complain about this there should be a choise to the user.
The save flags should be automatic and in ML and those flags
that might change during CAS processing.
Maybe the Forbit Mode Change should not error
but instead give the behaviour I have described earlier.
Change myself? Are you kidding? You - Professor - as the Master Mind
behind the most advanced original Calculator CAS
are much more cabable to do the job. Trivial 4 U, hard 4 me.
with best 
 
====
I get a lot of inquiries from 39/40G users who would like to know how to 
write games and so forth using sRPL.  If you're looking for something to 
add to the book, how about an appendix on how to program on the 39/40G?
====
Great source of info. For me it will be rated as Get It NOW!
====
Is there ANY way to rcobo the new model 12C so
that it's faster? The IRR calculation iterations are really slow
AND if you have to use it in a LOOP to back-solve
a sweet-spot (a break-even point + marginal)
the program will take forever. 10 x speed is needed.
Should I use the 19B II instead?
Yes, but it doesn't calculate fractional periods...
I need both devices.
Provided that the 12C can be speeded up, I'll buy one.
====
well, we could make it vibrate while thinking and be able to feel it
(biofeedback). let's see, we do an infinite loop and we get a massager
at the same time for those stressfull ocasions (exams). the LX model
would include infrared heat option.
wait a minute!
HP agilent has a micro that is configurable. they call it ST210 (
stmicroelectronics http://us.st.com made it) it defines its circuitry
from a source program. it runs at 250 Mhz yoouup. could they use it
for a supercalculator-HP49G ?
====
How in the world to you turn off the italics in the CAT menu when
using mini font to read the programs you have stored in the
calculator???  Is it a flag setting, if so I must have missed it.  Or
is it just that way so you can recognize what you have stored, and is
unchangable?  Any help in this would be appreciated.
====
Italics = external library command
normal = internal OS library command
To turn the italics off you have to hack the external library
into the ROM image and then re-flash your calc.
Happy hacking
====
I have a big probelm connecting my hp49 to the computer and transfer 
data...
i think i have wrong sttings for the connection.. on my pc i use the normal
software, provided form hp and this should find the calculator on the 
serial
board.. on the calculator i don't know what to choose for all the options 
to
get this data transfer working..
Port: Wire (surely)
Type: Kermit
Name : haven't specified one
Fmt: ASC
Xlat: New1
Chk: 3
Baud: 9600
Parity: None
OvrW = Checked
so can anybody help me with my problem..?
thx in advance
christoph
====
CAS = Computer Algebra System. It refers to the symbolic manipulation of 
expressions.  
If you would like to learn more about HP's implementation of a CAS in the 
http://www-fourier.ujf-grenoble.fr/~degraeve/cas49_1.pdf
If you're interested in more robust CAS you can always look into 
applications like Mathematica or Maple. A good place to start with a wealth  

of information is http://www.wolfram.com (the makers of the ubiquitous 
Mathematica). Please don't make this thread into a full blown discussion 
about the long standing dilemma of which one is better, Mathematica or 
Maple. IMHO, Ma... is the best. :)
====
functions
for each procedure I need.
Ok, is there an ability to bind them on some keys so to access them fast?
--
====
Yes. Probably the best way is to download and install Keyman 2.2002, 
which you can find at URL www.hpcalc.org.
A convenient and omnipresent set of keys to use might be 
right-shift-hold F1 through F6,
====
What is the equivalent command of NIBB (syntaxe JAZZ, RPL mod) in the
HP-Tools ?
====
Maybe NIBHEX...
====
Beginner:
- Learning to use Matrix Writer and Filer.
- Uses algebraic logic and 'quotes'.
- Knows that ON aborts.
- Learning to use CAS.
- Learning how to store variables.
A bit advanced:
- Uses CAS normally.
- Uses User-RPL to program.
- Uses plot features.
Medium:
- Uses SYSEVAL to program.
- Uses LIBEVAL to program.
- Ask on newsgroups what is an External.
- Makes advanced symbolic manip.
- Uses I/O, HPComm..
- Uses Libraries.
Expert:
- Uses the Lib 256.
- Uses Xtable and Emacs to code in SysRPL.
- Uses Kermit to send files.
- Uses Flash Memory.
Guru:
- Uses ML and direct memory access.
- Uses FLASHEVALS and direct Flash access.
Wizard:
- Uses HPTools to compile, Emu48 to test, HyperTerminal to send.
- Uses direct memory edition.
- Write RAW files(direct HPHP49-B...)
modeled after an UNIX sample.
====
It seems you're posting binaries. Please, Check your newsclient and your 
system.
 Michael
====
without
filer.
I could reproduce it. But in a different way, that when I pressed A it
showed TTRM.
====
without
filer.
Didn't work, but the JHM bug is bad!
====
Why none of the ML gurus made an PDF/PS viewer(would be cool to have pocket
Erable/Alg48/Int48Pro manual at my calculator) or an ZIP/RAR extractor(to
mantain the PDF/PS).
====
I think that the pdf format carries way to much overhead in formatting, it 
would be a waste to have such a file in the calc to display only a fraction  

of the file content. These formats (PDF/ps/LaTeX) are great for precision 
formatting across many platforms, especially for printing, but in the HP 
you don't need a cross-platform, bulky file swapping standard. Furthermore, 

the decoding will be painfully slow. You would be better off (in terms of 
content output to file size ratio) converting the pdf to ASCII and the 
images to grobs and then storing it into the calc. 
What it would be cool though is to have PC parser for pdf/ps-2-tgv files. 
Diego. 
====
So, why buy an TI89?
Buy an HP25 and shut up.
====
fault if
My
tests
it's
Even MuPAD is better than Derive.
====
LAP(1/SQRT(x)),LAP(EXP(-A*X}/SQRT(pi*X)),LAP((1-cos(x))/t),LAP((1+2*a*X)/(SQ

RT(pi*X)),LAP(2/x*(cos(b*X-cos(a*X)),
Or HP48 with Erable+Int48Pro+Alg48?
Like me? I have 188 libraries.
====
LAP(1/SQRT(x)),LAP(EXP(-A*X}/SQRT(pi*X)),LAP((1-cos(x))/t),LAP((1+2*a*X)/(SQ

RT(pi*X)),LAP(2/x*(cos(b*X-cos(a*X)),
I don't buy calculators because of a program. It's just like buying Windows
because of Freecell.
It's better to have more commands and have more control.
Stat48Pro, AutoSimp49, OT49, Emacs, just to say a few.
What I want with editing matrices on EQW? Calculate cos(det(x))?
No.
Just assume that you have an big directory structure like:
|-HOME
|
|
|
|- GAMES
 | MINEHUNT
 | TETRIS
 | TEST
|- MATH
 | GEOMETRY
 |- CALCULUS
   | ANALYZE
|- UTILS
|- HACKER
on 49 it has subdirectory.
I have an TI-89, an TI-92+ and an 86(very messed).
That is, you don't know how to use it.
Debug2, from hpcalc.org.
Yes but the PC-like menuing is shitty :-PPPPPPPPPPPPPPPPPPPPPPPP
And there's no SYSEVAL or LIBEVAL.
Sys-RPL rulez!!!!!!!!!!!!!!!
And on it, can you store files into Flash ROM?
No compressor for it.
And...
Can you evaluate gamma(x) on TI?
Can you evaluate 0,5! on TI?
Can you graph a conic on it?
Can you get roots of 'a*x^4+b*x^3+c*x^2+d*x+e' on it?
I can: [a b c d e] PROOT.
Can you make libraries direct from it?
Can you connect it to any serial port?
Can you use it connected to an 286 - MS-DOS 3.3 computer?
and...
====
I'm in complete agreement with you both. HP Corvallis Service center
sent me a complete piece of junk HP49 as a replacement unit recently.
HP calculators, especially the 49, which I really liked, are dead.
Just last week I started using a TI89 again. This is really sad, I
never thought all this would never happen.
You can lay the blame squarely on Carly Fiorina's shoulders. Her
obsession with the Compaq merger meant killing off any little part of
HP she could to pay for it. Never let a nontechnical person run a
technical company - especially a nontechnical woman! I've never met a
woman yet who really understood machinery or other technical stuff
like a man.
J.C. Randerson
Pueblo, Colorado
====
This is just plain silly. She was not even around when HPs bad
decisions in the calculator business were made. And as far as ACO is
concerned, dismantling them was a rational decision, see my other
post. There is no sane company on this planet that would have kept a
division with that kind of a track record.
I would note that you clearly have no understanding at all of what
caused the final demise of HP's calculator business. One of the main
reasons it failed so spectacularly was precisely that there seemed to
have been no non-technical competence with sufficient influence to
remind the technicians that it is not enough for a commercial product
to be cool (while dreaming of StarTrek devices). There has to be a
market for it, in which it can sell in sufficient numbers. It's that
simple.
You have not met many women then, I take it.
====
You can have a look at usinagaz at http://www.hpcalc.org/hp48/utils/comms/
for internet connectivity. I'm sorry not to have had time to write a Wget
it doesn't let you really download files.
-- 
 * [...] Note that 120 sec is defined in the protocol as the maximum
 * possible RTT.  I guess we'll have to use something other than TCP
 * to talk to the University of Mars.
 * PAWS allows us longer timeouts and large windows, so once implemented
 * ftp to mars will work nicely.
 */
(from /usr/src/linux/net/inet/tcp.c, concerning RTT [retransmission 
timeout])
 
====
 
Use KML. YorkeM generated nice BSOD's while using emacs+extable+OT49+nosy.
 
====
Remember to announce the next upgrade in Arabic, not Roman numerals.
It runs the chance of being automatically filtered out 
  
====
Nope, sorry but 0^0 is NOT 1. It's a singular form, IIRC, and as such is
undefined and hence is displayed as '?'.
WigglePig
====
Alban.
====
HI
In which menue can i change from hex base 16 to Dezimal base 10??
thanx
Hot PAPA
====
I assume you are using the HP49G.
Use right-shift 3 (BASE) 2 to change from hex to dec.
have a lot of fun
Roman
====
Raymond
Legend: MTH is a hard key, keys in square brackets are soft keys
====
Oh yes, Maple is also terrific but Danny has been rather specific
asking which one to choose between Matlab and Mathematica.
====
price.
product
they
I prefer MuPAD. It's very programmable and updatable. It's very easy to 
use.
It's only 25 MB when installed. Uses only 24 MB ram.
====
apple!
Nice. Can you stop spending bandwidth?
====
I am programming my first game in sysrpl.
I draw my grobs with my 48 but I don't know how to manage them in my code.
Could someone help me, please?       
====
Have you tried RPLMAN.DOC from www.hpcalc.org ?
There are some other documents related to graphics.
Raymond
====
This document seems similar to the one written by Eduardo Kaliniowski.
I am sorry but it is not very useful. I would like to read some code in 
which is involved graphic.
My problem at the moment is: how to use grobs or graphics in general in 
sysrpl?
This is my first experiment and the list of the commands is not enough.
and, as always, sorry for my english!
====
But one thing I don't understand. Can I use the cerf for mathematics
right out of the box? Or would I have to develop first?
Yes, and when you want to go for a job, everybody asks you about
experience with tools you never had the opportunity to work with. With
a small afordable package, something like a math specialized tool, you
could at least experience working with a CAS, be it different from the
standard tools.
So perhaps some of the companies out there, have mercy and stop
thinking that we want a non math companion. Mathematizing you can do
anything, even management, but show me a palm that would solve a deq
====
And if we think really consequently, this way (adaptation time) is the
best possible. The system evolves with time, and fits better and
better to your handwriting. This is a *real* individual solution and a
very intelligent one.
Quality has its price. A Mercedes is more expensive than a an Opel,
but close their doors and the sound alone will tell you which one was
the door of a Mercedes.
Well, neither does the HP49G.
Yes, Jean-Denis, tell them the difference 
Yes, *that* would have been great. Imagine a Mathematica-like
software, that accepts your *written* input instead of 100s of key
presses for a single matrix.
====
Yes captain! It is a little (big?) wonder that the 49G is as good as
it is, with all its strengths and flows. If HP had gave the ACO
adequate resources, then the HP49G would be unbelievable (which it
sometimes already is...). But then HP kicked the ACO out and that was
it. I can't get rid of the feeling that this was a minutiously planed
and carried out death operation right from the start.
Exactly! It looks like technology sells not only because of necessity
nowadays, but also because of, well,... I wanna have it-reasons. Or
else, why did Tamagochi sell that good? And if it did sell good, then
why not calcs with a fantastic CAS?
====
The screen of the HP49G is indeed often too small. But remember that
people before a couple of years named it big. As the time passes by
always bigger screens for computers come out, which let the older ones
appear smaller. The first question is: Is it big enough for the work
that must be done? If the answer is yes then it can stay that way.
If the answer is no, then the second question arises? Are ther any
available technologies for making it bigger? I think that there are.
The screen of the HP95LX was much bigger and also with a better
resolution, and that was already in the eighties. Of course the screen
(as we know it) of pocket calculators can't be as big as a computer
monitor, as long as the shell of the calculator das so small physical
dimensions. But there are many other ways to construct calculators,
like for example with shells that fold, and the like. Another possible
way would be to have a flexible screen that you can connect to the
calc, and disconnect/fold it when you don#t need it any more.
Amen! Regarding the huge amount of available commands, some of which
sound or work similarly, it is a real huge question, why so little
attention has been given to decent documentation. If I remember well,
one of the first (and also remaing) complaints about the HP49G was
(and still is) how does the darn thing work?
Yes but starting from a certain degree of enshrinkment you'll get
problems with resolution 
Sufficient number of keys? Well for math work, I would prefer no keys
at all, but rather the possibility to write my stuff directly on the
screen. Best user interface would be a living paper sheet with math
knowledge for me.
Oh yes, it did work and how it did work! Much better than anything
that has been brought to the market until today. Comparing a Pilot to
the Newton is much like comparing a cro-magnon to a homo sapiens.
Especially the last update of the software was a huge step ahead. But
so it goes. Apple does something today, everybody is yelling that it
is not good, and after some years they present us the novities which
we knew and forgot about.
Wish I could say something against this last sentence, but I can't.
Though I am a zero in economics (can't even manage my own economy
), I think that it is true that markets are not made of a few
hundreds of people. (Or even a few thousands.) But according to
nowadays marketing books, which I tried to read and understand (with
little success) , the first imperative should be *Create* demand.
Doesn't that mean that the company XYZ which is manufacturing the
calculator ABC (or steen-o-cobo GX ) should try first to convince
Helen and Nick that they can't live without it? Or did I completely
misunderstood that? 
 
====
By the way, if that is what was going on, then I am not surprised they
closed down ACO. First of all, it is enlightening to see you openly
admit that the HP 49 was just some sort of a scam, trying to milk some
money out of a cheap knock-off of a successful product (the HP 48).
had suspected all along.
Second, so you were putting all your resources into something as
ill-conceived as the Xpander? That stillborn child that never even
made it out the door (and for good reasons I might add)? Who the hell
do you think would have wanted to buy this thing, and at what price,
other than for a curiosity cabinet?
One thing is entirely clear to me: There seemed to have been no
marketing competence at all at ACO. There were people with a good
understanding of computer algebra and calculator design and related
issues, but precious few seem to have understood that a commercial
product, once it has been designed, ultimately must be sold, and sold
in sufficient numbers at that. Thus they designed products (like the
HP 49G and, even worse, this silly contraption named the Xpander) that
were completely out of touch with reality, i.e. any existing markets.
And you people were surprised that HP closed you down???
====
     While your work looks great, I'd like to point out that maxima
also works on a a PDA (the new Sharp Zaurus).  I suspect the Zaurus
binaries will also run as-is on the iPaq.
     If anyone's interested, see my other posting in this thread.
  
====
     Going off on a slight tangent, if you want to see Maxima in action
on a PDA, do this:
[ Note: the following is a bit tedious and ugly.  However, it's just
  software, and the issues can be resolved.    I'm only mentioning it
  now just to give a possible glimpse of the future.  ]
* Find someone with a Sharp Zaurus 5000D or 5500 PDA.  This is a
  recently-released PDA with a 206MHz StrongArm processor & color
  very nice, but the battery life sucks.
  [ The following may also work on a Compaq Ipaq that's been modified to
* If they haven't already done it, have them install the terminal
  application (on the CDROM that comes with the Zaurus).
* Go ye here and download the maxima package for the Zaurus (it's at the
  very bottom of the page):
        http://web.njit.edu/~rxt1077/clisp-maxima-zaurus.html
  For convenience, here's a direct download link:
        http://web.njit.edu/~rxt1077/clisp-maxima-zaurus.tar.bz2
* Under Unix or cygwin (Windows), extract the files:
        bunzip2 < clisp-maxima-zaurus.tar.bz2 | tar xvf -
  This will produce two files:
        lisp.run
        maxima-clisp.mem
* Transfer the files to the Zaurus, and put them both into some
  convenient directory.  Go to the directory in the terminal window.
        set +m 
        stty susp '^-'
        ./lisp.run -M maxima-clisp.mem
  The set +m is important, as it works around a bash issue when the
  Zaurus is suspended and resumed.
  This will start up maxima in the terminal window.  Unfortunately,
  maxima is configured, by default, to work with an 80-column screen,
  and the the terminal window is only 44 columns wide.  To fix this, run
  (at the maxima prompt):
        linel: 43
  (Yes, it's 43 and not 44.  You've got to use a value one less 
than
  the screen width.  Unfortunately, you've got to do this each time you
  start maxima.  I haven't found out a way to make this automatic.
  AFAICT, there is none.)
* At this point, you can try out maxima commands (don't forget the
  trailing semicolon in the following!):
  Indefinite integrals:
        integrate(x/(1+x^3),x);
  Definite integrals:
        integrate(x/(1+x^2),x,0,1);
  Limits:
        limit((2*x+1)/(3*x+2),x,inf);
        limit(sin(3*x)/x,x,0);
  Equation solving:
        solve([x+y+z=5,3*x-5*y=10,y+2*z=3],[x,y,z]);
* For an introduction to maxima, see:
        http://members.tripod.com/~Probability/intromax.htm
  (Note that this intro talks about macsyma and not maxima.  Maxima
  was derived from Macsyma, and so the tutorial is still good.  However,
  one Zaurus difference is that Ctrl-Z doesn't work -- don't even try to
  hit it.)
  The main web site for maxima is:
        http://maxima.sourceforge.net/
  And online documentation can be found as:
        http://maxima.sourceforge.net/referencemanual/maxima_toc.html
  (Note that the chapter on Symmetries is in French.  
* To quit, enter:
        quit();
Note that maxima isn't perfect.  Although it's derived from the old DOE
Macsyma, some bugs have crept into it, which people are working on
fixing.
====
No, the BIOS was in the last 40K of the 1M space.  (Nowdays it's often
the last 128K or more.)
The original PC didn't even officially support 640K.  It used DIP switch
settings for the RAM size, instead of autosizing like a modern BIOS.  The
largest memory size supported by the DIP switch settings was somewhere
between 500 and 600K, IIRC.  To use memory beyond the DIP switch settings,
people used programs run from their autoexec.bat file to tell the BIOS and
DOS to use more memory.
The original CGA and MDA used address space staring at 704K, so that was
considered to be the practical limit on early PCs.  Later video cards
took over the space from 640K up, and BIOS autosizing was limited to 640K.
The claim that noone would ever need more than 640K has commonly been
attributed to Bill Gates.  But no one has ever found an authoritative
citation, so this is really an urban legend.
====
You are right about HP but TI actually managed to do so in the past. When
you consider HP28, HP48S and HP48G line, they were far superior to the
existing TI calulators at the time. TI manged to make TI89/92 line, that 
was
not that much better than HP48G but was better enough to win the market
with a combination of good hardware, good marketing and sufficient 
software.
When HP finally got w wake up call they blunder by sticking to already
outdated  hardware that could not run efficiently improved software. Note, 
that
despite opinions on this forum, it is still an open debate if HP49 software 
is
 in
fact much better than TI89/92 and very often depends on personal 
preferences
of the users. I like HP over TI strictly because of RPN, but many hate HP
exactly for that. Its a perfect case of too little too late which I
spoted over a year ago and predicted, that it will just bankrupt HP 
calculator
operations.
HP48 line and earlier HP calculators were carried by every significant 
retailer in US.
They had even better exposure than TI at the time. It was sudden flow of 
very
 good
calculators TI8x and TI92 line which finally made most of retailerd to not 
even
 sell
HP49 model. Only home depot did ti but with marginal sales. That is why 
they
dumped the product.
 
====
It is somewhat hard to argue this without any hard data, but I doubt
that TI ever won the market that HP was in. I don't see many engineers
and scientists use TI89s either.
Yes, that was unfortunate, or rather the effect of HP not being
committed to developing new calculator hardware anymore; which might
have been caused by their (correct) perception that there was no
market _for them_ in that area anymore.
Yes, I think that is debateable. Particularly so if you restrict the
debate to the (small) subset of CAS features that most people will
realistically be interested in using.
Same here... Also, the keyboard layout of the TI89 is a disaster,
plain and simple; one gets the impression that the various functions
were strewn over the keyboard at random.
I don't know if much of the situation really is HP's fault, and if it
is, then it definitely is not a recent fault. You would then have to
argue that HP should have seen the need of a substantial presence in
the educational market decades ago. You are right, it was a case of
too late, but much too late, not just a few years too late.
579
1974
====
Sircam strikes again.
I've just received a message from tamu.edu. That's a university
in Texas, isn't it? If you're reading this, send it again please -
I can't read it.
Bye.
HPCC member #1046
-
-- 
====
Would like some feedback on the above Survey ROM cards. Have had some
feedback on TDS and SMI but nothing on D'Zign survey card. Does anyone use
this card?
All replies appreciated.
Justin
====
You might want to try the message board at pobonline.com.
My coworker had thier downloadable software, not the card.  I just
remember it being really slow.  That may have changed by now.
Have a nice day.
====
it's been quite some time, since I last looked into the newsgroup. I
also checked if there are new programs for the HP49 and found some
nice ones. For the most part I'd now switch to the HP49, but I mostly
still use the 48. One reason is that some important programs are
missing on the 49. Namely PowerPlot with Table is just nice to have
and I miss it :-(
Isn't there a programmer with some ML knowledge who likes to try the
port of PowerPlot light (as this one is finished and absolutly
stable). I really would love that.
There have been quite some attempts to port the program (or better
said persons who said they wanted to give it a try), but it never got
reality until the Beta version by Yoann Desir. I was happy that then
PowerPlot would have got ported, but after the Beta the work stopped
:-( 
Are you waiting for me to learn SysRPL/ML? ;-)
This might be an option, but I'm sure that it would be _much_ faster
programmers would take the challenge :-)
On a side note. Too bad that now (I'm just working on my diploma
thesis) I don't need a calculator that often. My family is sort of
irritated that now there are days, when I don't take my HP48/49 with
me ;-)
Greetings from Cologne
Peter
-- 
Great HP48/49 links:
http://www.hpcalc.org
http://move.to/hpkb
to find *old* postings search:
http://groups.google.com/advanced_group_search
====
I have had a graph engine for a long while now, coded in SysRPL. It's
usually much faster than PowerPlot (compared on my HP48GX). At its worst it
mathces the speed of PP, at its best, its at least twenty times faster
(rarely). usually it's around 2-3 times faster. For normal functions, a
complete plot (131 points) take 0.8-1.2 seconds to plot. Such a normal plot
could be of 'SIN(X-1)/(X-1)', 'X^3+2*X-2' or 'LN(1/X-2)' for example.
I havent finished a front end for the porgram, but have tried a couple of
times - that's the reason nothing is released. I'm pretty sure I've written
this here a while back?
PowerPlot for the HP49G is definetely a big asset, but the lack of interest
just shows how few are actually using the HP49G.
The library I've coded that have drawn most attention, is definetely
SYMBTOOLZ (which is broken on newer ROM revisions). I have gotten maybe
15-20 different people contacting me about it - most had problems because
they didn't read the huge manual. That's too small a user group to
motivate.
...and now HP has shut down intelligent calculator development, and thrown
away their best guys doing exactly that. It doesn't really help progress
here, does it?
====
Will you fix it? Please.
X
for a new release...soon...
====
1.2 seconds for 131 calculations of 'SIN(X-1)/(X-1)'..?  On a HP48GX?  
That's
pretty incredible, considering that the fastest thing I could think of for
computing a single point in sysrpl would look like:
::
 % .6
 %1
 %-
 %SIN
 % .6
 %1
 %-
 %/
;
where .6 would be a hard-coded value of X into the expression.  On my 48, 
TIM
produces the result of the above program in 18.5ms.  If I multiply that by 
131,
that would be ~2.42 seconds just to make 131 such calculations, without
looping, substituion, error checking, line drawing, and other 
non-computational
stuff associated with graphing something...  So you're saying that you can
completely plot 131 points of the sample expression 'SIN(X-1)/(X-1)' in half 

of
the absolute minimal time that it takes the HP48 to come up with the 131
results??
Aaron
====
No, I'm doing it on the HP49G. My graph engine is for the '49. I just
checked - I use 2.16 seconds to calculate the points, and 0.2 seconds to
draw them on screen.
non-computational
My above timing include this.
...  So you're saying that you can
half of
No, sorry. Not in SysRPL. I'll provide the exact times (add 0.0-0.3 seconds
to draw the complete plot):
'SIN(X-1)/(X-1)': 2.16 s. (PP: 5.74 s.)
'X^3+2*X-2': 2.35 s. (PP: 5.90 s.)
'LN(1/X-2)': 0.93 s. (PP: 2.95 s.)
'2*X+1': 0.47 s. (PP: 3.42 s.)
'(9*X-EXP(X-1000)+8973)^2-X/2': 2.71 s. (PP: 6.95 s.)
'X': 0.00 s. (PP: 2.76 s.)
====
Really hope to see it in actions soon.
seconds
====
I'm afarid you won't. I'm not sure, but I think I have only one single
program left for release at some point in time. I haven't got any 
motivation
in this regard - the HP company killed the spark in me, doing what they 
have
done lately.
====
1)', 'X^3+2*X-2' or 'LN(1/X-2)' for example.
I am interested!!  Now I feel that my HP49G Graphical Calculator is not 
complete without your graph engine.
Santos Lucero
====
I for one am really excited about your program.  I understand that HP is
quitting development, but HP 49G is one of the best calculator.  I really
hope more programs from the 48 are ported since the 49G is lacking alittle
in 3rd party support.
I really hope to see you finish your program, the default plotter is just
too slow.
it
plot
written
interest
====
[Porting PowerPlot to the HP49]
Arggh, some of the best programs are from Raymond. Maybe he would like
to make that great port? :-)
Greetings from Cologne
Peter
-- 
Great HP48/49 links:
http://www.hpcalc.org
http://move.to/hpkb
to find *old* postings search:
http://groups.google.com/advanced_group_search
====
If someone is wasting BW here, it's definetely you. Stop complaining - it
just creates even more BW use.
====
You are quite correct.  0^0 is a singularity.  [inf]^[inf],
[inf]^0, and 0^[inf] are also.  There are a few more that I cannot
remember right now.    I could send you the calculus proof for this if
you want it.
Dylan
====
Yes, it is an essential singularity in real or complex arithmetic.  But is
useful to define 0^0 to be 1 for integer arithmetic.  This is covered in
the sci.math FAQ, I believe.
Scott
-- 
Scott Hemphill  hemphill@alumni.caltech.edu
This isn't flying.  This is falling, with style.  -- Buzz Lightyear
====
And in HP 49G
The Integer Math gives ?
but the real/complex gives 1.
is
====
If you will read  carefully, you will note that no one said that the 
HP49 evaluates 0^0 as 1 , although I think it should. 
In an evaluation of zero to the power zero, if either, or both, 
zeros are approximate form( with decimal points) or if the numerical 
mode is set to approximate, then the evaluation gives 1., with 
decimal point. Parisse says that this is for compatibility with the 
48, which always gives 1.
Only when both are exact zeros without decimal points AND the 
numerical mode is set to exact, does it evaluate to '?'.
About the proper meaning of 0^0 outside the HP world, see Concrete 
Mathematics, or do some Google research.
====
I just purchased a TI-89 calculator.  Now before you quit reading, I
am a very avid HP-48,49 user and have been for years.  I was curious
about the functionality of the TI-89 however now that I have two
calculators, I am curious on how to use th two.  I prefer the HP
primarily.  For those of you who have both the HP and the TI, which
calculator do you use primarily and how do you use the other
====
I use the TI89 to satisfy my curiosness regarding some mathematical 
aspects.
The TI is very fast for some symbolic calculations (Taylor series, although
somewhat limited, and summations, integrals, differentiations and so on).
The HP49G simply is way to slow to do certain tasks, for example this:
spends 19 seconds on the HP49G, but only 8.5 seconds on the TI89.
'9*X-EXP(X-1000)+8973' 'X' 1. TAYLR takes forever on the HP49G (must be a
bug), while it's calculated instantly on the TI89. In fact, the TI89 can
calculate a Taylor series to degree 25 in one second, of the above
expression. There a more than a few Taylor series expansions that take
forever on the '49.
The TI89 do have its limitations too - it can't develop a Taylor series for
'SIN(X)/X+1' for example.
====
I use the latest rom and TEVAL says 7.4142 seconds :-(
Right, very strange...
 
====
Now it finished with an Insufficient Memory error. It must be a bug
then...
====
Aaron, I had an 89(a gift from a friend),
I just can't get used to NON-RPN calculators.
I do alot of numerical computations, and I extensibly use the stack.
I could not make myself bring the 89 to work, and it was just sitting
there in my desk at home.
I finally gave the 89 to my cousin, he is in high school, he LOVES IT!
HM 
PS. My humble opinion is solely based on my experience. Others can have
a different view and/or taste.
-- 
====
Ok, I took it. You can download the new package at
http://youpibouh.thefreecat.org/info/prog/down/usinagaz-cur.zip (it is
still version 0.2), or on hpcalc.org when it is updated.
It now includes wget, which can be used as :
http://www.hpcalc.org/ wget
<HTML DOCTYPE.... (beware, this page is about 19Kb long ! it takes some
time to download it !)
-- 
 Moralit.8e : le modem et le cablerouteur font comme les filles, ils
 papotent toute la journ.8ee.
 -+- RB in NPC : Et en plus, ils ne parlent que de bits -+-
====
And what kind of record ?
You seem to forget the fact that ACO was profitable.
Even if I don't really care if it was closed or not.
====
Of developing products that fail in the market place. Of not gaining
nor even holding on to market share in their target market(s).
====
Are HP calculators still in production, by the way? How long will they 
still
be made?
Arnold
1a8f5fe5.0204300758.51f7143@posting.google.com, Helen
====
Oh, I'm sure helen really has a clue. (just wait...)
StarTrek devices ? ha ha ha. what's that? At least we know helen is an 
old lady.
.....and, is that your reason?   Come on !!!!! 
I have met many women.......and.......mmmmm.....   :-)
====
Just a few. Even my sis, who grew up on the ranch with me, who helped
me rebuild engines, put up hay, weld, shoe horses, drag calves to the
branding fire...she was the biggest tomboy I ever knew, yet even she
really didn't understand machinery. Where do you or someone you know
stand in comparison?
J.C. Randerson
Pueblo, Colorado
====
I know many very competent female engineers. One of my female
colleagues in engineering teaches courses like machine design at an
Ivy-League school. If you take a look outside your cave, you will find
many more examples of that sort.
====
I somehow agree. I personnaly think that the ACO was killed because we 
ended
up having 7 different bosses in 4 years. To make things worst, the managers
in carly's team, are, like her, strong willed directing types instead of
being leaders. No 1.5 years long R&D project can wistand having 2 or 3
different heavely directing peoples managing them. and Carly is solly
responsible for the choice of her reports and the type of peoples that she
choose. Personnaly, I like leaders, not 'directors'. but this is a personal
view.
Well, the only products we actually did finish were the ones that were
develop without non-technical peoples looking at. The reason being that
non-technical peoples (marketing, managers and the like) do live in an
everchanging word and only have short term vision (3 to 6 months). And thus
will change the any reasonably sized project's requierements 4 or 5 times
during it's life time, thus killing it or denaturing it at such a point 
that
it becomes a dog. I agree that this things should be looked at by non
techies, but in today's market, it is simpely unworkable. In the group 
where
I am now, we are working with 2-5 year time frames, and marketing peoples,
just don't want to join us, they simply have no training, nor will to see
that far in the future. We, poor litte tech are teherfore left to develop
products without any marketing input, which is as bad as what was done in
the ACO.
:-)
====
You know, this sounds a little bit like you have the choice of
developing a product that fails because there is no input from
non-technical folk, or developing one that fails because there was too
much meddling from non-techies. While that is of course an
exaggeration, it serves to illustrate a fundamental difficulty very
nicely: Ultimately, you need people who understand both worlds, and
those are extremely rare. Without them, however, it is very hard to
develop products that are really successful.
P.S.: I take it you understand that I am not blaming ACO's failure on
you or your co-workers. Developing successful products takes a team of
people with different backgrounds, working together in a fruitful
manner. Obviously, that did not happen at ACO, but the reason it did
not happen cannot simply be pinned on one side or the other. And, by
the way, that is one of the areas where the art and science of
management becomes so vitally important, the general disregard of
technicians towards the management notwithstanding...
====
I'm having trouble finding any good ones.  At hpcalc.org, there are tons of
board games for the 48, but not that many for the 49.  Anyone know if the
game Hex has ever been developed to any HP calculator?
====
I think I read about that somewhere. Dunno, something like ...
http://www.hpcalc.org/details.php?id=793
Another hex-shaped board game :-)
http://www.hpcalc.org/details.php?id=930
Bye.
HPCC member #1046
-
-- 
====
J.,
There's quite an interesting board game going on at the moment in the
courts between Hewlett and Fiorina.
Sorry, couldn't resist.
Mark.
of
====
I don't know how it got there, which pisses me off almost as much as the
fact that it's there!
I'm assuming it's impractical to replace the LCD display.  It would 
probably
cost more than buying a new one, right?  There's a portion where the screen
is visible, but it's at the top.  It's basically useless.
So what should I do with it, anyone have any suggestions?
====
You obviously do not know the inside story, or you know way more that you
would like us to believe.
The fact that the Xpander project was cancelled had little to do with the
product itself, rather the consequences of HP getting out of the education
business (which not surprisingly happened just after a change of president)
I do not believe that Xpander was an ill-conceived project. Maybe some
choices like using Windows CE raised its price over than what it should 
have
been if you are targeting students.
Xpander was the really first product designed by teachers for teachers.
Hundred of them were involved in the process. Of course this machine had
absolutely nothing to do in the calculator business. And it was not a
calculator as such, but I believe it was a great concept. The only good
thing about not releasing such appliance is that the existing HP community
can't complain about this product not being a traditional HP calculator :)
I am still in contact with various teachers all over the world, and they 
are
all waiting for a product similar to what Xpander could have been. And so
far there is nothing close to it.
Xpander was a pure learning tool aim at high-school student.
But did HP ever have a good marketing team ?
Many people here complain that HP is not what it used to be, that is a
company with products designed by engineers for engineers.
Look at the first Windows CE devices. That was a pure marketing product. 
But
totally useless with unusable feature. Obviously the people who first
designed it didn't really use it on a day to day basis.
The HP49G was made in response to the HP calculator community who have been
asking for a long time for an HP48 with more memory, faster and with more
CAS capabilities. That's exactly what the HP49 was.
But then you had marketing people who came over and said: it has to look
more like a TI, let's change the design so it looks more : cool. Then 
it
had to look even more like a TI with algebraic in it.
If the product had been designed by engineers only, it would have looked
much closer to an existing HP48 hardware...
So you claim that Xpander had no market whatsoever? On which ground.. I
would be interested to know how so many teachers could be wrong.
Not more surprised when I see how HP is doing when shutting down division 
or
reducing all the R&D budgets.
I was really happy when ACO was closed down, I would have left anyway and 
as
====
Yes, but so what? There are many wonderful things out there that lots
of people know nothing about, in all areas of human interest. Why
should everybody be interested in computer algebra?
No, it could not. The complexity of the interaction you need to do CA
is simply orders of magnitude above the one you need to handle a video
phone. That is so obvious that it should not require any further
explanation. No interface of any kind can remove that kind of
complexity, period.
Unless, of course, you have a device that is so intelligent that it
can do all the work for you. For now, and in the foreseeable future,
those devices are called mathematicians, scientists, etc. ;-)
====
Of course not computer algebra itself but computer algebra as the tool
for doing mathematics. As one door for entering the math universe.
*That* is the pity. That we have tools for making an easy entrance to
a very fascinating world, which could activate some more grey cells,
but instead of this we expect a black box to solve some integrals.
Now, of course you could ask me also, why should everybody be
interested in mathematics? Well, working the mathematical way, we can
learn to draw maximum of conclusions from a minimum of premises.
Exercising this helps abstracting from special to general, and helps
sharpening mind. Perhaps the answer to the question about interest for
mathematics could be a new question: Do we want it the easy way, that
is to believe anything without thinking for ourselves, or do we want
to wonder, think and ask about the world, life and anything else ;-)
If this interaction is by means of menus, dialogs, command lines and
other primitive user interfaces, then yes, you are right. But as
already said, a good recognition software, that can receive and
understand your writings on a touch screen could be the first step.
Instead of typing your input in the input lines of mathematica, you
could also write it on the screen, this doesn't change anything of the
inner workings of the CAS.
Ahhh, so may I say: I need new batteries, can you hear me? ;-)
Greetings,
Nick.
====
That is a very important point, and one you are getting completely
wrong, IMO. CA is not a tool for entering the universe of mathematics,
anymore than, say, computer-controlled offset printing machines are
the tools for entering the universe of literature. In order to use CAS
profitably, you already have to know a lot of mathematics.
There is no easy entrance into the world of mathematics. Quite
likely you have learnt so much mathematics in your life that you have
all but forgotten this fundamental truth. Are you seriously suggesting
that anybody would be able to use a CAS, without a deep knowledge of
the mathematics involved? And then even learn something in the
process? Have you ever taught mathematics at any level?
Maybe I am old-fashioned, but my experience with my own children tells
me that you cannot teach them to run before they can walk. As an
example, IMO it would be disastrous to give calculators to
six-year-olds in order to provide them with an easy entrance into
some of the basic mathematical concepts relating to whole numbers
(counting, adding, subtracting, etc.). I believe the same is true, at
a different level, for high-school math, college math, etc.
But it also does not change the fact that what you write on that
screen can potentially be very complex, and the (intermediate) output
usually is even more complex. What you describe above does not change
any of this at all.
====
Oh, but you will :-)
All young students in US will eventually become engineers of tommorow.
They don't have much choice beyond TI, they are tought in high schools
and most universities on TI, many corporations have basic TI training for 
it's
lower level technical personel but I've yet to see one designed for HP or
Casio. In that sense TI won (or Casio :-( ).
Precisely. I love TI automatic fast simplification of formulas, where many
consider that as fault.
In my case I have TI92. It's bulky, but it's keyobard happen to be 
superior,
thanks of course to QWERTY. I also hate TI Basic. It's not only
slow but seems like afterthought added to the system a the last hour.
Too litlle too late rarely is the outcome of some recent bad decisions.
Rather it is the product of years of neglecting the customer needs and
ignoring the competition.
Jack
====
And later:
Oh Helen, where have you been all my life?
I'll go burn some sack; 'tis too late to go to bed now.  ;-)
Shakespeare. Twelfth Night. Act II, Scene III.
Bye.
HPCC member #1046
-
-- 
====
Thanx everybody! I will try. ...and for sure I'll succeed ;-)
bye and good programming
ulrik
====
which is involved graphic.
Eduardos book does touch graphics. Try it out - you get nothing served. 
It's
one of the best starts for SysRPL out there.
====
maybe these could be links for you:
http://www.hpcalc.org/hp48/docs/books/graphics.html
http://www.hpcalc.org/details.php?id=1727
http://www.hpcalc.org/details.php?id=1763
http://www.hpcalc.org/details.php?id=1764
http://www.hpcalc.org/details.php?id=1765
Raymond
====
there are some PDF-Files on www.courbis.com available. (Click on
the files HP48 Machine Language is written in english. There are some
other files in french available as well. What a great chance to learn this
language ;-)
Some of them are dealing with Programming the HP48 (SX/GX). But I'm not 
much
of a (HP4X)-programmer so I'm not sure if it fits your needs.
have a lot of fun
Roman
====
Thanx verrry much!!
Your link seems very interesting (it is now in my bookmark), but only for 
ML.
I'm looking for sysrpl.
Mayby I will use it later when I'll be stronger in programming!
Bye, ulrik
====
i downloaded this file: fem48/49 (for instance..). 
a) how can i transfer the file *.lib to my calculator?
b) i transferred it through renaming the *.lib in *. without
extension, but there is only a string, and no library which i can
attach.
Pleas help!
thanx.
====
The convention has been:
[HARD] keys
|SOFT| keys
COMMANDS as from [CAT] or [alpha][alpha]COMMAND NAME[alpha]
Right Arrow = [RA]
Left Shift = [LS]
& means Hold Down the first key while pressing the second
[LS]&[LA]
====
Ahem. I assume you are using the HP-48, since there wasn't a different note 

in
the header;-)
====
On Sun, 28 Apr 2002 17:57:07 +0300, Veli-Pekka Nousiainen
But do you have an idea of how to EVAL/COLLECT in the MATCH command, so 
that
the result for example is ..x^v*x^(2*v)..x^2*x^1.5.. = ..x^(2*v)..x^3.5..
GruÛ
Alexandra
====
The next matrix gets TTRMs or memory looses in my calc and also in Emu48
with MK:
[[ 20.1 15]
 [20.3 38]
 [20.5 76]
 [20.7 57]
 [20.9 14]]
It happens using the matrix editor (I've heard about editor's bugs) but 
ALSO
writing it by hand on the editing line.
I'm ashamed...
====
Aaaarggh! I mean amazed. (My english!...)
====
I had a HP48 3 year ago, when I started university, everybody there
had one, but now I moved to the United States and everybody here uses
TI, so I sold my hp48 (to the only person here who have ever had one)
and bought a TI-89, I know this was a big mistake (dumb of me), I
didn&#8217;t like my TI-89 and I tried to use software to make it like
a HP (RPN, Equation Editor) but it didn&#8217;t work, so finally I
bought a HP49, and is so much better than the TI-89 that i feel alive
again.  I had used userRPL with my HP48 so I made some programs for my
HP49, all they are simple but useful for me, and now I feel I am ready
to take the next step to sysRPL, but now with the end of HP calculator
I have been wondering if it is worthy to learn it? What will happen if
in a couple years all that I have learned is useless because there are
not more HP calculators to program? Should I learn to program my TI-89
instead and keep using userRPL for my HP?
What do you think?
Marturo