D-89
===
Subject: Re: EEPro HP49 latest version installation help pls
run on my 49g+. If you run the first command in the library, it
returns unauthorized. The 4th command will take {} as input. This
will start the interface. Look here for the original thread:
TW
===
Subject: Re: EEPro HP49 latest version installation help pls
could run on my 49g+. If you runthe first command in the library, it
returns unauthorized. The 4th command will take {} as input. This
will start the interface. Look here for the original thread:
TW
===
Subject: Hp 28S version Orlando 1.13?
I recently acquired an HP-28S calculator.
I know it's old but it's much smaller than the 48/49 series,
and I don't need the graphing capabilities.
But this calculator, entering #A SYSEVAL should return the version
number.
I was expecting 2BB or similar, and it displays Orlando 1.13.
Also, some of the menus aren't where they're supposed to be.
For example, pressing the plot key shows the statistics menu.
I tried a memory reset but it doesn't seem to help.
Could anyone tell me what's going on?
Ted
===
Subject: Re: Hp 28S version Orlando 1.13?
I've tried that, but nothing changes, unfortunately.
===
Subject: Re: HP-48GX/49G emulator for Linux?
http://s.cls.free.fr/wikini/wakka.php?wiki=Emu48
===
Subject: HP49G+ - strengths and weakneses?
Hi -
I'm trying to get a picture of the strengths and weaknesses of the
HP49G+ re the TI89 before I decide whether to order one.
The main strengths I care about seem to be better CAS (but in what
ways?), the equations editor (but could someone tell me briefly what
advantages this has over the TI89 equivalent in practice?). Matrix
crunching speed and SysRL aren't of much concern to me.
Weaknesses seem to be build quality, the fact that equations in the
stack are kept in user-friendly form, and a higher learning curve?
I'm not strongly for or against RPN.
Umm.. the keyboard IS fixed now, right?
Comments?
===
Subject: Re: HP49G+ - strengths and weakneses?
the equations editor (but could someone tell me briefly what
advantages this has over the TI89 equivalent in practice?).
I don't know what the TI89 equivalent is (if you're talking about the
$15 optional thing, I haven't used it) but I think the 49G+'s equation
editor is a great feature and a major advantage over the TI89.
For one thing, the equation editor is useful not only for entry, but
also simply for *looking* at results.
On the TI89, the pretty print feature will only display so much
information vertically. If you have compound fractions, or say a 3x3
matrix with symbolic fraction entries, the whole thing won't show, and
you'll have to copy and paste the result to the entry bar, and try to
decode it. It also takes forever to scroll through a long formula on
the TI89, by the way (you can jump to the beginning or the end, but
there's nothing that lets you jump an amount in between).
These are not problems on the 49G+.
Also, of course the equation writer is a godsend for entering complex
formulas. With the TI89, you can be guaranteed to get unpaired
paranthesis fairly frequently. The equation writer is also nice for
when you have a formula with similar parts repeated. For instance, if
you're entering the Fresnel equations for intensity change on
reflection, you can enter the numerator, copy it, and then paste that
in the denominator, and use the cursor to go right to where you need to
change the sign and change it. On the TI89 the problem is that you
can't edit the formula in pretty print; to find the necessary sign
change you'll have to scroll through the formula bar.
As for the CAS, as said, they're difficult to objectively compare. If
we're talking integrals, there are some the TI89 can do that the 49G+
can't, and some the 49G+ can do that the TI89 can't. And usually the
49G+ is faster, but there might be an odd problem that the TI89 can do
faster. Which of those sorts of problems will be the ones you'll
actually need to deal with? Who knows. But I'd say 3 out of 4 times,
the 49G+ is more powerful. While the TI89 will occasionally do
something faster than the 49G+, when the 49G+ is faster, it's often
MUCH faster.
Example, integrating squareroot of tangent of x. The 49G+ does this in
a couple of seconds. I tried this on the TI89, and it ran and ran, for
over 20 minutes, before I finally cancelled the operation to save the
batteries. I later tried it on a TI89 emulator that ran at a higher
speed, and it did manage to solve it.
(On the other hand, if you do the first substitution, z = tan(x), and
get squareroot(z)/(1+z^2), the TI89 will do that integral in seconds)
Another example, integrals of the form 1/(x^n-1).
If a human was solving this problem, the steps would be obvious:
Factor the denominator as x minus the various nth roots of 1, then use
partial fractions.
Well, in exact mode, the 49G+ can't do that integral if n = 5 or more.
The TI89 can. An advantage for the TI89?
Well, maybe. The 49G+ can still do those integrals, but it needs to be
in approximate mode, because it factors the denominator numerically.
I've only tried it up to n = 11, but it solves that just about
instantly.
On the other hand, the TI89 can do it exactly for n = 7, but it takes
FOREVER. I don't know if a real TI89 could do it before the batteries
would run out. I tried it again on an emulator running at faster
speed, and after leaving it all night it had the answer in the morning.
If you'd hope you could speed the TI89 up by doing it approximately,
you're out of luck. It won't do anything with those integrals in
approx. mode.
Well, it's easy to go on and on with comparisons. I've been working
with some matrices lately, and it seems the TI89 can take
transcendental functions of matrices, but can only do that -- or find
eigenvalues or eigenvectors -- approximately with numerical data.
The 49G+ can't take transcendental functions of matrices, but it can
find eigenvalues and -vectors of symbolic matrices. (Using those
functions, you should be able to write a little program to use
functions on the matrices)
Despite that this post probably seems excessively biased towards the
49G+, I still think the TI89 is a great calculator.
My general guideline for a recommendation would be that it depends on
what you want. If you want the calculator that's easier to use, that'd
probably be the TI89. It automatically simplifies expressions, which
can be annoying if don't want that, but a great time saver if you do.
If you want sheer power, go with the 49G+. I think that eventually the
49G+'s sheer power will eventually allow it to make up for (with
programming or user knowledge) any relative weaknesses it may have to
the TI89 (as in the matrix example above).
I think the 49G+ is for those who are a little more serious about
science and math, while the TI89 is more for people who don't want to
need to invest as much time into learning a calculator.
===
Subject: Re: HP49G+ - strengths and weakneses?
X
MAP
===
Subject: Re: HP49G+ - strengths and weakneses?
MAP won't allow you to get the equivalent to 'EXP(M)' where M is a
square matrix, which is NOT the same as taking EXP of every entry in the
matrix.
One definition of EXP(M), for an arbitrary square matrix M is
Sum(j=0..+oo,M^j/(j!)),
where M^0 is the appropriate identity matrix, and M^j is otherwise the
appropriate power of M.
If the n by n square matrix, M, factors into form INV(U)*D*U, where D is
diagonal and U obviously must be invertible, then EXP(M) =  INV(U)*F*U,
applied to diagonal D.
This can be done in the HP49 series approximate mode when M has a basis
of eigenvectors.
===
Subject: Re: HP49G+ - strengths and weakneses?
DIAGMAP
===
Subject: Re: HP49G+ - strengths and weakneses?
No, that's not it.
MAP takes the function of each individual term, not of the matrix
itself.
Example, using MAP to apply the exponential function to the 2x2
identity matrix gives
[e,1]
[1,e]
While really the exponential function of the 2x2 identity should be
[e,0]
[0,e]
===
Subject: Re: HP49G+ - strengths and weakneses?
Now I faintly remember that this has discussed before
Google anyone not as lazy?
VPN
===
Subject: Re: HP49G+ - strengths and weakneses?
By the way, here's a little program for the 49G+ that will really apply
the exponential function to a 2x2 matrix:
===
Subject: Re: HP49G+ - strengths and weakneses?
THNX
===
Subject: Re: HP49G+ - strengths and weakneses?
Sheesh. I was looking through the 49G+'s catalogue, and apparently it
can do it on its own.
You just need to use DIAGMAP rather than simply MAP.
I'll probably be learning more about this calculator 10 years from now.
===
Subject: Re: HP49G+ - strengths and weakneses?
Then you have to buy them both
as they are strong in different areas of CAS
so there is no clear winner in this category
No!
The TI 89 Ti has still the old TI 89 style
mushy feeling
bad layout
poorly tailorable keyboard
Buy the 49g+
You can speed it up using software only (keeping warranty)
VPN
===
Subject: Re: Which calculator should I get?
It is quite capable. True the project has not been around as long as
TIGCC so of course it is not as mature. I am not trying to persuade
you either way. I just think calling someone's work crappy when
comparing it to something that is YEARS older is kind of not very nice.
. .
TW
PS -- Even your overclocked 89 won't be able to spit close to the
49g+ running at even 12 mHz slow speed. Not to mention you wouldn't
have an SD card.
===
Subject: Re: Which calculator should I get?
If you love programming, go for 49G+. Learn ARM instead of MC68000.
HPGCC is usable.
cslim
===
Subject: Re: Which calculator should I get?
Do you use hpgcc? I've heard the library is extremely small compared to
tigcc. How is it? When you say usable, do you mean usable without
issues?
Also, I believe the MC68000 is much more used in smaller devices, but I
could be wrong.
How long does the battery take to run out at that speed and at stock?
applications you write in C?
There are actually a few apps for the TI-89 that let you do that...
however, I've heard they're pretty buggy (to be expected).
I also heard that a lot of the ARM CPU is used to emulate the Saturn.
Is this true?
===
Subject: Re: Which calculator should I get?
Ok, I didn't literally mean it was crappy; I just meant it didn't work
so well. I know it's still quite an immature project, and I stated that.
:) No offense to the developers.
I don't really care about an SD card. And remember, for now, I'm mostly
going to be coding the sucker. I guess I'll be doing some more research
yet. :)
--
Patrick M.
/* EOF */
===
Subject: Re: Which calculator should I get?
The 49g+ can be easily  SOFTWARE overclocked
which does not void the warranty
It will go fully working up to 120MHz
and 203MHz is the max CPU limit
(MEM dragging and LCD usually blanked)
I wouldn't buy a TI-89 anymore
VPN
===
Subject: Re: Which calculator should I get?
Erm... if you overclock something, and it gets destroyed... even if it's
software overclocked, it's still overclocked, and hence shouldn't the
warranty be void? Just like on a PC, you overclock via the BIOS (which,
as you know, is software :)), and if your CPU is destroyed (even if not)
your warranty is void. Though it's much harder to know if the chip has
been OCd or not if no hardware modifications have taken place.
I know that the 49g+ is much faster in all aspects... but my true focus
right now is C programming support (I'm not in the higher grades yet, so
don't need a ton of power, though I may for my apps/games and it'll be
nice for the future) so I can code the sucker in C. :)
--
Patrick M.
/* EOF */
===
Subject: Re: Which calculator should I get?
How many days you think should prove that an overclocked calc
will not be destroyed?
I run flat a fresh set of batteries at 119MHz on a 49g+
hpgcc is great and getting better (you need a PC)
I hope I C your C contributions in the future
You're welcome
VPN
PS: Does anyone know a pocket device
with a C development environment witheen??
===
Subject: Re: HP-12C Platinum is slower than vintage HP-12?
No, no , you must press them to the wall. Tell them that his individual
unit
is defective and that it runs much slower than the ones in the store.
Forget
about the fact that a large lot of them are junk.
Scott
===
Subject: Re: HP-12C Platinum is slower than vintage HP-12?
The algorithm is more accurate, thus slower
VPN
===
Subject: Re: A new trigonometry?
That's a truism, but irrelevant,
because that's not *quite* the proper criterion
for determining whether a Euclidean trisecting construction
exists or not! (keep reading to find out why)
To get the meaning, first let's investigate something simpler --
do you all remember the construction you learned in school
for *bisecting* an angle?
Good, now I want you to get out your compass,
straight-edge and pencil, and show me how to apply
that famous (and seemingly infallible) angle *bisecting*
procedure to an angle of *zero* degrees,
then to an angle of 360 degrees, and so forth.
Have you actually held the tools in your hands
and tried to carry out the instructions you once learned?
If you have, then you know what I'm getting at --
the construction which you learned requires you
first to draw an arc intersecting each side of the
angle to be bisected, then from those two intersection points
to draw two more arcs having equal (and suitable) radii,
and finally to join the original angle's vertex
to an intersection of the latter pair of arcs.
But when you are busy bisecting an angle of
360 degrees, say, you encounter an insurmountable
problem -- each of the second pair of arcs
is forced to have the same center, hence
there is no *unique* intersection point
of those final two arcs, hence said
construction recipe can not be properly completed!
I therefore put the case to you that there exist angles
(360 degrees, for one), which can NOT be bisected
by this construction -- does that prove
that angle bisection is *in*general* impossible?
No, of course not, I'm sure you'll agree,
so how then can you say that exhibiting *one*
angle for which trisection is impossible
proves that angle trisection is *in*general* impossible,
any more than exhibiting *one* angle whose bisection
is impossible would prove *in*general*
that angle bisection is impossible?
Oh, you may say why, I happen to *know* what the
answer is for 360 degrees, but your knowing the answer
doesn't mean that your general construction procedure
can actually construct it!
Suppose, for example, that I conjecture the existence
of an angle trisecting procedure, and that steps #17 and #18
of this marvelous procedure are very similar to those of the
famous angle bisecting procedure, requiring that the intersection
of two arcs be found. But lo, when it comes to attempting this
procedure upon an angle of exactly 60 degrees (the
non-trisectable angle mentioned in the book I was critiquing),
why, it turns out that the centers of those two arcs
must coincide, hence *just*for*such*a*special*angle*,
the construction fails, much as angle bisection
likewise fails for special angles such as 360 degrees!
I now have a real shocker to report -- let's think of
measuring all angles as multiples of a right angle,
e.g. what we call 135 degrees is equivalent to
exactly 1.5 right angles (when expressed
in decimal notation), and so forth.
Further, let us express the measurement of all angles
in the binary system (135 degrees in terms of right
angles, expressed in binary, is thus exactly 1.1
right angles).
The shocker is, that any angle having a finite number
of 1 digits in said expression is certainly trisectable
in a finite number of steps (proof left to you :)
Since the set of such angles is obviously everywhere dense,
there is an *everywhere*dense* set of angles
which *can* be trisected in a finite number of steps!
Before you choke on your doughnut, I suppose it should be
admitted that this isn't the proper criterion either,
as to whether a Euclidean trisecting construction
actually exists.
What *is* the proper way of putting it, then?
By the way, has this exercise stretched anyone's mind a bit,
to consider whether we all really know how to know
whether we know at all what we think we know?
===
Subject: Re: Link-cable to my HP48Gll
Well, if you're talking about being a l33t hax0r, the interface *I*
miss is the old Centronics parallel port.
guy who still boots an IBM PS/1 each year to run the cub scout's pine
car derby because the printer port is cabled straight to some simple
trip wire sensors in the race track.
===
Subject: Re: Link-cable to my HP48Gll
Me thinks one can easily fit both RS and USB (along with IR)
to the top of the case
===
Subject: Re: So whats next
: just what I said & thought
: I want the old HP back
It's gone forever. It was once a company of engineers run by engineers. It
is
now run by bean counters.
--
-------------------
Keep working millions on welfare depend on you
===
Subject: Re: So whats next
X
wait-a-minute -
I'm going to hang myself on a chain of calculators...
:-(
===
Subject: Erable 3.x bug?
GCD3 is supposed to return u, v, d given arguments a and b. u, v, and d
are such that d = GCD(a,b) and
ua + vb = d
But if you try any pair a, b:
2. a
1. b
GCD3
3. b
2. 0
1. 1
This is also the case in other versions. What am I doing wrong?
===
Subject: Re: Erable 3.x bug?
If my recollection is correct, you have to configure
INTEGER instead of POLYNOMIAL somewhere.
===
Subject: Re: Erable 3.x bug?
===
Subject: Assign other key to Right-Shift
Hi!
Among other HP calculators I have a faulty HP48. I opened it and fixed
the bad screen but I have no luck with Right shift button. I love my
calculator and I lik eto use it if I only can assign the RS button to
another key. I installed keyman but had no luck there, I cannot get to
the library...
I think there are a way to make another key to work as RS, either with
double click or by pressing it longer. Can anyone advice me how to do
this, the key # is 81.x and it can be moved to any button. Just so I
can install and properly run keyman or other program and then do better
configuration there.
/Chris
===
Subject: hptalx not working for regular user
I've compiled and installed hptalx 1.3.0 for linux. It works nicely when
running as root, but for a regular user it immediately comes up with a
The
calculator is not know message, when trying to connect.
The regular user has access to /dev/ttyS0 and /dev/ttyS1. Does anyone know
what else is needed ?
TIA, Jaap
===
Subject: Re: hptalx not working for regular user
Already found it. kermit requires write access to the /var/lock directory,
which it doesn't have in a standard debian install...
so either the user access to /var/lock should be granted or kermit should
be
run setuid root.
Jaap
===
Subject: MINEHUNT ROM 2.0
Hello together,
Hello Cyrille, Hello Yean-Yves,
I have been reported that MINEHUNT on ROM 1.23,
on ROM 1.19-6 and older, has made spontanously
garbage collection, which was possibly interupting the game.
Has this been corrected in ROM 2.0?
Heiko
===
Subject: Re: MINEHUNT ROM 2.0
Playing with it with ROM 2, I believe I can still have garbage collections
but they are MUCH faster.
Arnaud
===
Subject: remote34
I am using the remote34 program,
but my Kenwood Receiver & DVD use more than one GROUP code for the
commands.
But the TRANSFER function that is called by the REM function with a
4-byte GROUP and a 2-byte FUNCTION code is in machine language so I
cannot modify it to store GROUP & FUNCTION as an 8-byte code for each
command.
Any help out there?