Help Triangle Solutions
James C. Beavers
I have been looking for 2 years to find a program for the HP48G that can do
Triangle Solutions. Anyone know of one that works similar to the one for the
HP11C?
If not, do you where I can get an HP11C repaired?
I have tried Kenneth Hobson's program from HP's ftp site but it doesn't
work.
Thanks,
James C. Beavers
Whitney Design Group, Inc.
1000 West Avenue Suite 319
Miami Beach, FL 33139
FAX (305)538-0114 TEL
WEB SITE http://home.att.net/~jcbwdg
E-MAIL jcb...@worldnet.att.net
WndaLouise
what do you want this program/directory to do...???
i just recently wrote a program to find the 12+ points for the 9 point
solution... ( find all the points where a circle intersects a triangle in a
bunch of different places...???...! )
???
and then there's the Sine ratios and so on to be used with the Solver...
what else...???
Norm Arnold
James C. Beavers wrote:
> I have been looking for 2 years to find a program for the HP48G that can do
> Triangle Solutions. Anyone know of one that works similar to the one for the
> HP11C?
<snip>
I am not aware of any built in triangle solutions for the HP11C. However I
have written two triangle solving programs for the HP48GX. If you or anyone
else is interested either send me an email or post to this group and I will send
them. Here are the characteristics of my programs:
The first one is called TRI and it deals exclusively with right triangles
(containing a 90 degree angle). The right triangle has four parts including
RISE (shortest side), BASE (side perpendicular to RISE), HYP (longest side) and
ANGLE (always the smallest angle and always opposite the RISE). The user
supplies any two of these parts and the GX will output all of them.
The second program is called OBLQ and it deals with all triangles. These
triangles consist of three sides (SA, SB, & SC) and three angles (AA, AB, &
AC). Side SA is opposite angle AA, etc. Given three of these items the GX will
output all of them. The user has 5 choices of input which are three sides given
(SSS), two sides and a non included angle given (SSA), two sides and an included
angle given (SAS), two angles and the included side given (ASA) or two angles
and a non included side given (AAS).
These programs have some common characteristics. For example they will work
in either imperial mode or metric mode. If user flag 1 is set then both input
and output will be imperial. For example 2'-4 3/4" would be represented as
2.0412. All angles are dealt with in degrees. The output is always sent to
variables as well as the screen. If user flag 2 is set then the output is also
sent to the printer via the IR port.
Hope this will help. Norm
---------------------------------------------------------
"It's not the size of the dog in the fight, but the size of the fight in the
dog."
- Archie Griffen
Norm_...@bc.sympatico.ca
Norm Arnold
Surrey, British Columbia
Canada
---------------------------------------------------------
Peter Karp
Hi James,
>I have been looking for 2 years to find a program for the HP48G that can do
>Triangle Solutions.
As WndaLouise writes, what do you want to calculate? There are many
Triangle Programms at the HP48 Software Archive (see link below).
My personal favorites are QTS (size 1KB), which solves the unknown
sides and angles of a triangle in a quick way, or SWS (size 12KB).
SWS also calculates *all* other unknowns (like height and so on (is
this the right english word?) and can graph the triangle and handles
also two possible solutions!) It's in german but it's selfexplaining
and easy to use.
>Anyone know of one that works similar to the one for the HP11C?
Sorry, don't know the HP11C.
Greetings from Cologne
Peter
________________________________
Newbies: Do you know the great Frequently Asked Questions?
http://www.engr.uvic.ca/~aschoorl/faq/
and the superb HP48 Software Archive?
http://hp48.ml.org
to look for *old* HP48 postings see
http://search.dejanews.com/filter.xp?groups=comp.sys.hp48
Ostcroix
You can take a look at QTSv2.1 at
http://hp48.ml.org/ in the math misc directory
QTS was written based on the old 11c triangle solver
Enjoy
James C Beaver wrote:
><HTML><PRE>I have been looking for 2 years to find a program for the HP48G
>that can do
>Triangle Solutions. Anyone know of one that works similar to the one for the
>HP11C?
>
>If not, do you where I can get an HP11C repaired?
>
>I have tried Kenneth Hobson's program from HP's ftp site but it doesn't
>work.
>
>Thanks,
>
>James C. Beavers
Othniel Rawlins
Ostc...@aol.com
==============
x@x.x
In article <355DF807...@bc.sympatico.ca>, Norm_...@bc.sympatico.ca (Norm Arnold) wrote:
>
>James C. Beavers wrote:
>
>> I have been looking for 2 years to find a program for the HP48G that can do
>> Triangle Solutions. Anyone know of one that works similar to the one for the
>> HP11C?
>
> I am not aware of any built in triangle solutions for the HP11C. However I
>have written two triangle solving programs for the HP48GX. If you or anyone
>else is interested either send me an email or post to this group and I will send
>them. Here are the characteristics of my programs:
> The first one is called TRI and it deals exclusively with right triangles
>(containing a 90 degree angle). The right triangle has four parts including
>RISE (shortest side), BASE (side perpendicular to RISE), HYP (longest side) and
>ANGLE (always the smallest angle and always opposite the RISE). The user
>supplies any two of these parts and the GX will output all of them.
> The second program is called OBLQ and it deals with all triangles. These
>triangles consist of three sides (SA, SB, & SC) and three angles (AA, AB, &
>AC). Side SA is opposite angle AA, etc. Given three of these items the GX will
>output all of them. The user has 5 choices of input which are three sides given
>(SSS), two sides and a non included angle given (SSA), two sides and an included
>angle given (SAS), two angles and the included side given (ASA) or two angles
>and a non included side given (AAS).
> These programs have some common characteristics. For example they will work
>in either imperial mode or metric mode. If user flag 1 is set then both input
>and output will be imperial. For example 2'-4 3/4" would be represented as
>2.0412. All angles are dealt with in degrees. The output is always sent to
>variables as well as the screen. If user flag 2 is set then the output is also
>sent to the printer via the IR port.
> Hope this will help. Norm
Those sound like useful things to have. Have you sent them to the fellow
who makes the goodies disks?
Fabrizio Galia
if you need a easy program to solve all triangles go to my web site
Quack da Paperinik !
***********************************************
f.g...@studenti.to.it ; ICQ: 8627129
http://www.poli.studenti.to.it/~f.galia
joe...@usa.net
x@x.x wrote:
> Norm_...@bc.sympatico.ca wrote:
>> The user has 5 choices of input which are three sides given
>> (SSS), two sides and a non included angle given (SSA), two sides
>> and an included angle given (SAS), two angles and the included
>> side given (ASA) or two angles and a non included side given
>> (AAS).
> Those sound like useful things to have. Have you sent them to
> the fellow who makes the goodies disks?
I have received many programs & directories of this type, but
none have made it to the Goodies Disks because I'm waiting for a
program like the old HP-67/41 "automatic triangle solver" by John
Kennedy (I think). You input the three known values, and the
whole triangle was solved for *immediately*, without forcing the
user to figure out whether to press ASA, SSA, SAS, yadda, yadda,
yadda.
The closest we've yet gotten to this Golden Fleece (Holy Grail?)
of triangle solvers was one of Jim Donnelly's libraries that has
a triangle solver in it... but it uses an input form instead of
menu keys, so it's slow and a little cumbersome to use, whereas
the program of my dreams uses a menu to input three values and
as soon as the third one is entered, the triangle gets solved:
bing, bing, bing, POW! The first (or best) program like that has
a guaranteed place on Goodies Disk #12.
-jkh-
joe...@usa.net
-----== Posted via Deja News, The Leader in Internet Discussion ==-----
http://www.dejanews.com/ Now offering spam-free web-based newsreading
Dick Smith
In article <6jmvvf$t4j$1...@nnrp1.dejanews.com>,
joe...@usa.net wrote regarding triangle solving programs:
] I have received many programs & directories of this type, but
] none have made it to the Goodies Disks because I'm waiting for a
] program like the old HP-67/41 "automatic triangle solver" by John
] Kennedy (I think). You input the three known values, and the
] whole triangle was solved for *immediately*, without forcing the
] user to figure out whether to press ASA, SSA, SAS, yadda, yadda,
] yadda.
Ah, so it wasn't just me that found programs like that less than easy to use!
] The closest we've yet gotten to this Golden Fleece (Holy Grail?)
] of triangle solvers was one of Jim Donnelly's libraries that has
] a triangle solver in it... but it uses an input form instead of
] menu keys, so it's slow and a little cumbersome to use, whereas
] the program of my dreams uses a menu to input three values and
] as soon as the third one is entered, the triangle gets solved:
] bing, bing, bing, POW! The first (or best) program like that has
] a guaranteed place on Goodies Disk #12.
One day when I have more time to devote to enjoyable tasks like this I'd love
to learn enough about programming to be able to have a go at this one;
meanwhile let's hope that Joe's challenge is met.
Incidentally, there is a reasonable program for TI-85/6 calcs that gives you
a list of the three sides and three angles---you fill in the ones you know
and leave the others set to zero and away it goes (without Joe's sound
effects!), so it can't be too hard if it's doo-able on a TI, can it?
Hopefully,
Dick
--
=============================================================================
Dick Smith Acorn Risc PC di...@risctex.demon.co.uk
=============================================================================
Diego Berge
In article <6jmvvf$t4j$1...@nnrp1.dejanews.com>, joe...@usa.net says...
>
[ snippety snippety ]
>I have received many programs & directories of this type, but
>none have made it to the Goodies Disks because I'm waiting for a
>program like the old HP-67/41 "automatic triangle solver" by John
>Kennedy (I think). You input the three known values, and the
>whole triangle was solved for *immediately*, without forcing the
>user to figure out whether to press ASA, SSA, SAS, yadda, yadda,
>yadda.
You don't have the HP41 source code, do you? I'd sure like to have a look
at it.
btw, John Kennedy??
Regards,
Diego Berge.
Ostcroix
joe...@usa.net wrote:
>I have received many programs & directories of this type, but
>none have made it to the Goodies Disks because I'm waiting for >a program like
the old HP-67/41 "automatic triangle solver" by >John Kennedy (I think). You
input the three known values, and >the whole triangle was solved for
*immediately*, without
> forcing the user to figure out whether to press ASA, SSA, SAS, >yadda, yadda,
yadda.
......
>The closest we've yet gotten to this Golden Fleece (Holy Grail?)
>of triangle solvers was one of Jim Donnelly's libraries that has
>a triangle solver in it... but it uses an input form instead of
>menu keys, so it's slow and a little cumbersome to use, whereas
>the program of my dreams uses a menu to input three values and
>as soon as the third one is entered, the triangle gets solved:
>bing, bing, bing, POW! The first (or best) program like that has
>a guaranteed place on Goodies Disk #12.
Actually that's the way QTSv2.1 works. The Library has
SSA SAS SSS SAA ASA
All the user needs to do is enter the three known parameters
on the stack and it outputs all sides & angles. It automatically solves obtuse
angles or right angles. QTS even output all parameters in pretty display format
if the user installs JAVA or EQSTK.. I ported QTS from a 11C several years ago
when I was trying to learn SRPL.
Othniel Rawlins
Ostc...@aol.com
==============
x@x.x
>I have received many programs & directories of this type, but
>none have made it to the Goodies Disks because I'm waiting for a
>program like the old HP-67/41 "automatic triangle solver" by John
>Kennedy (I think). You input the three known values, and the
>whole triangle was solved for *immediately*, without forcing the
>user to figure out whether to press ASA, SSA, SAS, yadda, yadda,
>yadda.
>
>The closest we've yet gotten to this Golden Fleece (Holy Grail?)
>of triangle solvers was one of Jim Donnelly's libraries that has
>a triangle solver in it... but it uses an input form instead of
>menu keys, so it's slow and a little cumbersome to use, whereas
>the program of my dreams uses a menu to input three values and
>as soon as the third one is entered, the triangle gets solved:
>bing, bing, bing, POW! The first (or best) program like that has
>a guaranteed place on Goodies Disk #12.
Gentlemen and Ladies, we have a challenge! Will we let a puny HP41
enjoy a better triangle solver than anything available for the HP48?
Let it not be so! What fine programmer shall solve this challenge?
Fame, Fortune, and favors from desirable sexual partners await!
joe...@usa.net
Diego Berge wrote:
>
> You don't have the HP41 source code, do you? I'd sure like to have
> a look at it.
The automatic triangle solver program for the HP-41 was published in
the PPC Journal (Jake, do you know which issue?). I don't have it in
electronic form.
> btw, John Kennedy??
John was (and still is) a mathematics professor at Santa Monica
College. Very active in the HP calculator user community in the
PPC days.
-Joe-
Death to spammers.
joe...@usa.net
Norm Arnold
I am a structural steel detailer and I use these triangulation program
several times a day. That is why I wrote them. However I find that most
of the people in my profession do not have the expertize to write programs
(in fact a lot of them have very little understanding of their
calculators. However it is a widespread accepted fact that the HP48G
series is the best for our work. Consequently a lot of people I know and
work with have these mmachines but dont understand them. Therfore, my goal
when writing a program is not to produce the fastest or the most compact.
My goal is to produce the easiest to use and follow. Therefore choose
boxes and input forms are a staple for me. This is also why I write in
user RPL only (although I do use SYSEVAL & LIBEVAL when I can't find an
alternative). Although I believe I could write a program that does what is
suggested by Joe, it would not serve any purpose for me. I would
anticipate that the program would behave much like the solver: produce a
menu consisting of SA, SB, SC, AA, AB, and AC. The user fills in three
value then the program would search through the input to see which have
values and proceed to solve based on it's findings. Good luck to anyone
who takes it on. If anyone is interested in the code for my present
programs I have posted them seperately.
Happy programming... Norm
joe...@usa.net wrote:
> x@x.x wrote:
>
> > Norm_...@bc.sympatico.ca wrote:
>
> >> The user has 5 choices of input which are three sides given
> >> (SSS), two sides and a non included angle given (SSA), two sides
> >> and an included angle given (SAS), two angles and the included
> >> side given (ASA) or two angles and a non included side given
> >> (AAS).
>
> > Those sound like useful things to have. Have you sent them to
> > the fellow who makes the goodies disks?
>
> I have received many programs & directories of this type, but
> none have made it to the Goodies Disks because I'm waiting for a
> program like the old HP-67/41 "automatic triangle solver" by John
> Kennedy (I think). You input the three known values, and the
> whole triangle was solved for *immediately*, without forcing the
> user to figure out whether to press ASA, SSA, SAS, yadda, yadda,
> yadda.
>
> The closest we've yet gotten to this Golden Fleece (Holy Grail?)
> of triangle solvers was one of Jim Donnelly's libraries that has
> a triangle solver in it... but it uses an input form instead of
> menu keys, so it's slow and a little cumbersome to use, whereas
> the program of my dreams uses a menu to input three values and
> as soon as the third one is entered, the triangle gets solved:
> bing, bing, bing, POW! The first (or best) program like that has
> a guaranteed place on Goodies Disk #12.
>
> -jkh-
> joe...@usa.net
>
> -----== Posted via Deja News, The Leader in Internet Discussion ==-----
> http://www.dejanews.com/ Now offering spam-free web-based newsreading
--
---------------------------------------------------------
"To see what is in front of one's nose requires a constant struggle."
- George Orwell
joe...@usa.net
Othniel Rawlins wrote:
>
> Joe Horn wrote:
>
> > ... without forcing the user to figure out whether to press
> > ASA, SSA, SAS, yadda, yadda, yadda.
>
> SSA SAS SSS SAA ASA
Uh, that's precisely what I'm trying to avoid! Please don't
force the user to key in the angles and sides in some
predetermined order (clockwise/counterclockwise), and then
force him to decide which solver to activate (SAA? ASA? ??).
Give him six menu keys labeled with the sides and angles. He
keys in ANY value and presses ANY of those keys to store that
side or angle. Three values are entered thus, in any random
order. As soon as the third value is input, POW, the entire
triangle is solved. This is how the HP-41 triangle solver
worked. It was SO nice to use.
I am fully aware that the ASA/SAA/etc-style triangle solvers
are good and useful tools. But it haunts me that the older
HP calculators had BETTER triangle solvers written for them
than any that have been written for the HP48.
-Joe-
Death to spammers!
Roger Blake
joe...@usa.net wrote in message <6jmvvf$t4j$1...@nnrp1.dejanews.com>...
>x@x.x wrote:
>
>I have received many programs & directories of this type, but
>none have made it to the Goodies Disks because I'm waiting for a
>program like the old HP-67/41 "automatic triangle solver" by John
>Kennedy (I think). You input the three known values, and the
>whole triangle was solved for *immediately*, without forcing the
>user to figure out whether to press ASA, SSA, SAS, yadda, yadda,
>yadda.
>
>The closest we've yet gotten to this Golden Fleece (Holy Grail?)
>of triangle solvers was one of Jim Donnelly's libraries that has
>a triangle solver in it... but it uses an input form instead of
>menu keys, so it's slow and a little cumbersome to use, whereas
>the program of my dreams uses a menu to input three values and
>as soon as the third one is entered, the triangle gets solved:
>bing, bing, bing, POW! The first (or best) program like that has
>a guaranteed place on Goodies Disk #12.
However, I am unfamiliar with the programs you refer to. Could you be more
specific on how the input is. How does the calc know if your giving it an
angle or side length? Does the program use an input form?
As best I remember the program on HP41 math pack required the ASA or SSA
type entry.
I have written a couple of triangle programs myself. I was thinking it
could be done using an input form where you have six blanks S1, S2, S3, A1,
A2, A3. You would fill in any three (except AAA) and the program would
solve and place the unknown values in the form. Is this what you had in
mind?
Milo
There is a program that is in the grapevine book "algebra and pre-calculus
on the hp 48g/gx written by Dan Coffin on page 32 tha is called TRIGX. it
will take any info you have (sin, cos,tan,angle in DMS or RAD,area, radius,
etc.) and solve the remaining fields. I have typed it in and it works fairly
quick as it is written in user RPL. The book is excelent! if you need more
info feel free to write me.
James C. Beavers wrote in message <6jke68$s...@bgtnsc02.worldnet.att.net>...
>I have been looking for 2 years to find a program for the HP48G that can do
>Triangle Solutions. Anyone know of one that works similar to the one for
the
>HP11C?
>
>
>I have tried Kenneth Hobson's program from HP's ftp site but it doesn't
>work.
>
>Thanks,
>
>James C. Beavers
Guy Macon
As far as how to input the information goes, may I strongly suggest
that it be done exactly like the way the HP48 solves voltage dividers,
ohm's law, etc.? I know that joe wants it to instantly solve when the
third parameter is entered, but I prefer HP's way.
Ken Alverson
>I am fully aware that the ASA/SAA/etc-style triangle solvers
>are good and useful tools. But it haunts me that the older
>HP calculators had BETTER triangle solvers written for them
>than any that have been written for the HP48.
How does the equation editor combine several equations into a single solver
(with the muse thing). If someone knows how to initialize muse with a set
of equations, it would be as simple as using any other equation.
Ken
Ken Alverson
>Uh, that's precisely what I'm trying to avoid! Please don't
>force the user to key in the angles and sides in some
>predetermined order (clockwise/counterclockwise), and then
>force him to decide which solver to activate (SAA? ASA? ??).
>Give him six menu keys labeled with the sides and angles. He
>keys in ANY value and presses ANY of those keys to store that
>side or angle. Three values are entered thus, in any random
>order. As soon as the third value is input, POW, the entire
>triangle is solved. This is how the HP-41 triangle solver
>worked. It was SO nice to use.
Here we go. What about this (I forget the cosine rule, so you'll have to
put that in, but here goes)
<<
{ 'sin(aa)/sa=sin(ab)/sb' 'sin(ab)/sb=sin(ac)/sc' cosinerules }
'EQ' STO
MINIT MSOLVR
>>
It should look and act like the regular solver (it is) and as long as sine
and cosine rules are all you need to solve the triangles, you're set. If
those aren't enough, add in the required equations and let muse do the
solving.
Ken
joe...@usa.net
Ken Alverson writes:
> Joe Horn writes:
>
>> As soon as the third value is input, POW, the entire
>> triangle is solved. This is how the HP-41 triangle solver
>> worked. It was SO nice to use.
>
> Here we go. What about this (I forget the cosine rule, so you'll have to
> put that in, but here goes)
>
> <<
> { 'sin(aa)/sa=sin(ab)/sb' 'sin(ab)/sb=sin(ac)/sc' cosinerules }
> 'EQ' STO
> MINIT MSOLVR
> >>
Uh... no! Good idea, but it misses my point. THE ONLY TRIANGLE
SOLVER PROGRAM THAT WILL BE WORTHY OF GOODIES DISK INCLUSION IS
ONE THAT SOLVES *INSTANTLY* FOR THE WHOLE TRIANGLE THE *MOMENT*
THAT SUFFICIENT DATA HAS BEEN ENTERED.
No, I don't want to press shift-solve, as somebody has suggested.
Why should I have to? When the data reaches critical mass, I want
the damn explosion to occur! I don't want to have to push some
damn button! This is a TOOL we're desigining here, and as such
it should do as much of the work as possible, freeing the mind of
the user to be able to focus entirely on the PROBLEM and not have
to worry about the personality and psychology of the damn
programmer!
No, I don't want to wait for any damn input forms. They're slow
and annoying. Forget it! Ghah! Yuck! Bleh!
No, I don't want to wait for the damn solver to cogitate; this is
very simple math, people! How would you like it if the solver were
invoked every time you pressed the 1/x key? They could have done
it that way, but it'd be painfully slow. So they programmed it to
perform 1 SWAP /. The same idea MUST be used here; program the
literal math for the various combinations of inputs, and just DO
IT. NOW. NO WAITING.
Possible menu layout:
[SIDE.A] [Angle A] [SIDE.B] [Angle B] [SIDE.C] [Angle C]
Type a number. Press a menu key. *bing*
Type a number. Press a menu key. *bing*
Type a number. Press a menu key. *POW* The solution appears.
Sound effects optional. Ease of use and speed MANDATORY.
-Joe-
Norm Arnold
joe...@usa.net wrote:
<snip>
> THE ONLY TRIANGLE
> SOLVER PROGRAM THAT WILL BE WORTHY OF GOODIES DISK INCLUSION IS
> ONE THAT SOLVES *INSTANTLY* FOR THE WHOLE TRIANGLE THE *MOMENT*
> THAT SUFFICIENT DATA HAS BEEN ENTERED.
<snip> By the way, Why are you shouting?
> I don't want to wait for any damn input forms.
<snip>
It sounds to me like the only type of triangle program to get on the goodie
disk will be an ML routine. Once the three inputs are entered the program will
have to determine how many sides (or angles) are given and if the relationship
is two to one, whether the one is included or excluded. There are twenty
scenarios that have to be checked (including three angles given which is
unsolvable), then in eighteen of those you have to check if the odd ball data
is included or excluded. Then finally you can do the calculations. All that
code to find out something that the user already knows???? You can't possibly
think that someone will try to solve a triangle without knowing what input he
is going to give. Even with the menu format you suggest, the user has to know
what he is supplying in order to put it in the correct variables. It seems to
me that all this checking and comparing would be a hell of a lot slower than
just getting the info from the user (at least in user rpl). Perhaps from a
purist's point of view this would be a "better" program, but I live in the real
world. I don't want to sit around waiting for my HP to figure out something I
already know! My criterion for my programs are: Does it work? Does it handle
every case I can throw at it? Then it is GOOD!
Personally I find input forms, choose boxes, etc. to be very useful. I
have seen programs written so that each input must first be stored into a
specific variable and then run the program just so they don't have to use an
input form! So the user has to memorize where each input goes and do it by
hand just because the programmer doesn't like input forms. Why not lighten up
a bit. If an input form fills the bill then use it. Writing kilobytes of code
just to avoid an input form or a choose box makes absolutely no sense.
Well that's my 2 cents. Norm
Guy Macon
In article <6jsdfr$4n$1...@nnrp1.dejanews.com>, joe...@usa.net (joe...@usa.net) wrote:
>
>Ken Alverson writes:
>
>> Joe Horn writes:
>>
>>> As soon as the third value is input, POW, the entire
>>> triangle is solved. This is how the HP-41 triangle solver
>>> worked. It was SO nice to use.
>>
>> Here we go. What about this (I forget the cosine rule, so you'll have to
>> put that in, but here goes)
>>
>> <<
>> { 'sin(aa)/sa=sin(ab)/sb' 'sin(ab)/sb=sin(ac)/sc' cosinerules }
>> 'EQ' STO
>> MINIT MSOLVR
>> >>
>
>SOLVER PROGRAM THAT WILL BE WORTHY OF GOODIES DISK INCLUSION IS
>ONE THAT SOLVES *INSTANTLY* FOR THE WHOLE TRIANGLE THE *MOMENT*
>THAT SUFFICIENT DATA HAS BEEN ENTERED.
>
>Why should I have to? When the data reaches critical mass, I want
>the damn explosion to occur! I don't want to have to push some
>damn button! This is a TOOL we're desigining here, and as such
>it should do as much of the work as possible, freeing the mind of
>the user to be able to focus entirely on the PROBLEM and not have
>to worry about the personality and psychology of the damn
>programmer!
>
>No, I don't want to wait for any damn input forms. They're slow
>and annoying. Forget it! Ghah! Yuck! Bleh!
>
>No, I don't want to wait for the damn solver to cogitate; this is
>very simple math, people! How would you like it if the solver were
>invoked every time you pressed the 1/x key? They could have done
>it that way, but it'd be painfully slow. So they programmed it to
>perform 1 SWAP /. The same idea MUST be used here; program the
>literal math for the various combinations of inputs, and just DO
>IT. NOW. NO WAITING.
>
>Possible menu layout:
>
>[SIDE.A] [Angle A] [SIDE.B] [Angle B] [SIDE.C] [Angle C]
>
>Type a number. Press a menu key. *bing*
>Type a number. Press a menu key. *bing*
>Type a number. Press a menu key. *POW* The solution appears.
>
>Sound effects optional. Ease of use and speed MANDATORY.
Ah. I see the light. Yes, that's the Right Way To Do It.
I can see a nice place for a series of solver routines for triangles,
squares, circles, etc, etc. as well. They would be a nice compact
and memory miserly thing to have available, but they wouldn't be the
holy grail that you seek.
joe...@usa.net
A younger and more foolish Joe Horn wrote:
>
> I'm waiting for a
> program like the old HP-67/41 "automatic triangle solver" by John
> Kennedy (I think). You input the three known values, and the
> whole triangle was solved for *immediately*, without forcing the
> user to figure out whether to press ASA, SSA, SAS, yadda, yadda,
> yadda.
Although John Kennedy (PPC member 918) did publish a triangle solver
somwhat like this in PPC Journal V7N3P10 (April 1980), and so did
Keith Jarrett (PPC member 4360) three pages later in the same issue,
the one I was thinking of was by Geroge Donaldson (PPC member 3825)
which was on the page after Kennedy's program! It solves the whole
triangle as soon as the third value is input, and does not require
any special order for the inputs. It kicks ass. And there's nothing
even close to it for the HP48... even though that was 18 years ago!
Please do not ask me for a copy of this HP-41 program, since I do
not have it in electronic form. Don't worry; somebody will have
the Perfect HP48 Triangle Solver done soon. I hope.
-Joe-
(1537) [13] <53> {NaN}
WndaLouise
Concerning Norm's complaint that there is a -balance- point to be found between
the ease of use for the User and the ease of use for the 48...
i personally love programs that take the view that the User should have the
overwhelming advantage in ease of use...
i occasionally like to program an input structure that allows me to input
various objects in any order and then let the program figure out what to do
with them...!
such as in one case that takes a grob and a complex number, the input can be
either
2: grob or 2: (x,y)
1: (x,y) 1: grob
i wrote a triangle solver like the one being sought along time ago, and i'll
look for it...i seem to recall however that it doesn't work as quickly as some
people might like it to work...i've always been satisfied with a reasonable
solution in a reasonable amount of time...???
Ping-Tao Fok
My concern about a triangle solver is that there are many ways to
determine a triangle uniquely. For example, 2 sides and angle or 2 angles
and one side can determine a triangle uniquely. You as the user has to
tell the calculator what you are entering, ie, side or angle. Also, how
much information do you want? The following is just a small list of
things that may be useful:
length of all three sides
angles(interior and exterior)
area
bisection points
trisection points
centroid
center of mass
bisection intersection point
trisection intersection points
perpendicular lines the drop from one angle to its opposite side
inscribing circle radius (circle that is inscribed by the triangle)
conscribing circle radius (circle that inscribe the triangle)
and so on.
If you just want to know the sides and the angles, anyone can write a
program for that pretty easily. But if you want to know more about the
triangle, you need to provide more selection for the user.
Ben
--
**************************************************************************
Nature is as complex as it needs to be, and no more. -Einstein
Check out my web page at http://www.umn.edu/nlhome/g062/fokx0002/ben.html
**************************************************************************
joe...@usa.net
Norm Arnold wrote:
> joe...@usa.net wrote:
> > THE ONLY TRIANGLE
> > SOLVER PROGRAM THAT WILL BE WORTHY OF GOODIES DISK INCLUSION IS
> > ONE THAT SOLVES *INSTANTLY* FOR THE WHOLE TRIANGLE THE *MOMENT*
> > THAT SUFFICIENT DATA HAS BEEN ENTERED.
> Why are you shouting?
To get your attention. I see it worked!
> It sounds to me like the only type of triangle program to get
> on the goodie disk will be an ML routine.
No. It was done in less than 400 bytes in the HP-41, in purely
user code. It can be done in HP48 User RPL. Nobody's done it
yet, that's all.
> Once the three inputs
> are entered the program will have to determine how many sides
> (or angles) are given and if the relationship is two to one,
> whether the one is included or excluded. There are twenty
> scenarios that have to be checked (including three angles given
> which is unsolvable), then in eighteen of those you have to
> check if the odd ball data is included or excluded.
> Then finally you can do the calculations.
Correct.
> All that code to find out something that the user already
> knows????
The user NEED NOT know which equation needs to be used with each
combination of inputs. That's the benefit of the program; it
figures out which equations to use, and then uses them. And it
is easy to do this. You don't need to "check" all 19 cases.
As the inputs are accumulated, a running total of binary values
can be kept, so that when the 3rd input is made, a unique value
has been accumulated and can be used for a single calculated
branch. No IF's, no CASE, no data shuffling. Fast. Easy.
> You can't possibly think that someone will try to solve a
> triangle without knowing what input he is going to give.
Huh? I know three sides, or two sides and an angle, or one
side and two angles. I do not know (nor care) which of the
19 cases I have (nor whether it's the ambiguous SSA case
that has two solutions): the program should figure that out
transparently to the user.
> Even with the menu format you suggest, the user has to know
> what he is supplying in order to put it in the correct
> variables.
If I implied that the program should be able to solve a triangle
about which nothing is known, I'm sorry, I did not intend for it
to perform miracles.
> It seems to me that all this checking and comparing would be a
> hell of a lot slower than just getting the info from the user
> (at least in user rpl).
No. Since it requires the smallest possible number of keystrokes
from start to finish, and since it gives the answer *instantly*
(remember, it uses the case-appropriate equations, not the
solver; and finding the case is instantaneous), it will not be
slow. PROOF: Try the HP-41 program; even *that* was faster than
the existing HP48 programs, despite the fact that the HP-41 was
a much slower machine than the HP48.
> Perhaps from a purist's point of view this would be a "better"
> program, but I live in the real world.
So you're saying that you focus on the way things really are.
And I'm saying that I focus on the way things ought to be.
Yes, this means that you live in the real world and I am an
idealist. As a programmer (read: artist engaged in a CREATIVE
process) I can think of no higher compliment. Thank you.
Current HP48 triangle solvers are good. I have an idea of how
to do it better (read: faster and easier to use). That's all.
Actually, don't real-world people appreciate "faster" and
"easier to use"? Or is human inertia (the instinct to do things
the same way that we've always done them) really that strong?
> I don't want to sit around waiting for my HP to figure out
> something I already know!
Then don't. Run your own real-world program. I won't mind.
But just don't challenge me to a triangle solving race, or
I'll smoke ya! ;-)
> My criterion for my programs are: Does it work? Does it
> handle every case I can throw at it? Then it is GOOD!
Correct. I never said they were bad. Why resist suggestions
for improvement? Nothing and nobody is being threatened here,
and if my idea plays out, a lot of people might be helped.
> Personally I find input forms, choose boxes, etc. to be very useful.
So do I. I use 'em all the time in programs when I cannot
think of a user interface that's easier to use and faster.
> Why not lighten up a bit.
Uh, ok.
> If an input form fills the bill then use it. Writing
> kilobytes of code just to avoid an input form or a choose
> box makes absolutely no sense.
That it makes no sense to you, I can understand.
Please understand that it does make sense to me.
Joseph K. Horn (1537) [13] <53> {NaN}
http://users.kcyb.com/joehorn/HP48.htm
Guy Macon
For those working on the Joe Horn Triangle Challenge, be sure to test
your code with the following inputs:
Side A + Side B < Side C
Zero or negative side length
Zero or negative angle.
Angle A + Angle B >= 180 degrees
...etc, etc.
It's fine to exit with an error message on these inputs, but it
should tell the user what he did wrong. Better would be to
reject the bad input with an error message saying why. Crashing
with an overflow error is bad, but not the worst. Running forever
in search of an answer that will never come is worse but still not
the worst. Giving a wrong answer is the worst possible error
(unless you count giving a wrong answer then clearing memory).
A good program handles errors with style and grace.
Roger Blake
Guy Macon wrote in message <6jtmtq$2d0$2...@news01.deltanet.com>...
To me this input is ambiguous. [SIDE.A] the side opposite from [Angle A]
and so on. If so this seems very much like the ASA style. I
Diego Berge
In article <35622BC1...@bc.sympatico.ca>, Norm_...@bc.sympatico.ca
says...
>
[ big snips everywhere ]
> Personally I find input forms, choose boxes, etc. to be very useful.
In my programs, whenever it's convenient I implement both types of input.
Generally novice users will like the input forms and once they get to know the
program they'll use the stack. I copied the concept from the 48GX itself (
[RightShift] + [Solver] vs [LeftShift] + [Solver]).
Regards,
Diego Berge.
Michael Hoppe
On <19.05.98> joe...@usa.net (joehorn) wrote:
J>When the data reaches critical mass, I want the damn explosion to occur!
You're so right.
J>Type a number. Press a menu key. *bing*
J>Type a number. Press a menu key. *bing*
J>Type a number. Press a menu key. *POW* The solution appears.
Bing-bing-pow sounds good. But what output is expected in the case
the user types in three angles? The ratio of the sides?
Just wondering,
Michael
--
-= Michael Hoppe <michae...@k2.maus.de>, <mho...@hightek.com> =--
-= Key fingerprint = 74 FD 0A E3 8B 2A 79 82 25 D0 AD 2B 75 6A AE 63
-= PGP public key available on request. =---------------------------
joe...@usa.net
Michael Hoppe writes:
>
> But what output is expected in the case
> the user types in three angles? The ratio of the sides?
Hey, good idea! I like it! (I blush to confess that the
program I'm working on merely beeps if you attempt to enter
a third angle, and won't accept the input). If you can think
of a clean way of representing the sides as ratios (perhaps
as algebraic objects?) please post it. Yours is the first
really new idea (regarding HP-calculator triangle solvers)
in 18 years! It'd be cool if it could be implemented.
-Joe-
--
Ostcroix
......
>Possible menu layout:
>[SIDE.A] [Angle A] [SIDE.B] [Angle B] [SIDE.C] [Angle C]
>Type a number. Press a menu key. *bing*
>Type a number. Press a menu key. *bing*
>Type a number. Press a menu key. *POW* The solution >appears.
>Sound effects optional. Ease of use and speed MANDATORY.
Hey Joe, What about this simple menu layout:
SIDE ANGLE
all you do is enter in your case and it ouputs 6 paramters
Side1 Angle1
Side2 Angle2
Side3 Angle3
You can't get more simpler than that :)
Othniel Rawlins
Ostc...@aol.com
==============
Guy Macon
In article <6k37po$lad$1...@nnrp1.dejanews.com>, joe...@usa.net (joe...@usa.net) wrote:
>
>Michael Hoppe writes:
>>
>> But what output is expected in the case
>> the user types in three angles? The ratio of the sides?
>
>Hey, good idea! I like it! (I blush to confess that the
>program I'm working on merely beeps if you attempt to enter
>a third angle, and won't accept the input). If you can think
>of a clean way of representing the sides as ratios (perhaps
>as algebraic objects?) please post it. Yours is the first
>really new idea (regarding HP-calculator triangle solvers)
>in 18 years! It'd be cool if it could be implemented.
In my opinion, when you enter the second angle, the third
angle should be instantly calculated and filled in for you.
Then, when you enter a side length, the other two sides
should be instantly calculated and filled in.
It would also be nice if the main program entry screen just
gave you the sides and angles, but a menu selection led to a
long list of additional calculated information, such as area,
height, largest circle that fits inside, and as many other
parameters as you can think of. Would a menu selection to
draw the triangle on the screen be useful?
joe...@usa.net
Guy Macon writes:
>
> In my opinion, when you enter the second angle, the third
> angle should be instantly calculated and filled in for you.
Excellent idea! Why didn't I think of that? (smacks forehead)
> It would also be nice if the main program entry screen just
> gave you the sides and angles, but a menu selection led to a
> long list of additional calculated information, such as area,
> height, largest circle that fits inside, and as many other
> parameters as you can think of.
Another good idea!
> Would a menu selection to
> draw the triangle on the screen be useful?
Already have the code done for that. If the largest side is on
the bottom, it's easy to fit any possible triangle on the screen
with sides and angles labeled. Useful in two ways: (a) when
used by students, it helps the learning process because most
students struggle with visualizing data; and (b) when used by
engineers it can serve as a quick verification that there were
no gross errors made in the inputs.
WndaLouise
ack! i hate to agree with Othniel about anything...!
SIDE ANGLE
all you do is enter in your case and it ouputs 6 paramters
Side1 Angle1
Side2 Angle2
Side3 Angle3
You can't get more simpler than that :)
------------------------------------------:
but this would also solve some of the other problems inherient in determining
which sides and which angles are being provided...
it would however require the use to understand that they are supposed to enter
in connecting sides and angles...
AND
it would allow the user to enter more information from one menu
i.e., SIDE ANGLE X Y
x & Y refering to the hieght and base of a non-right angle triangle...
you might also want to display a picture of the triangle with the menu...???
Roger Blake
WndaLouise wrote in message
<199805221831...@ladder01.news.aol.com>...
>ack! i hate to agree with Othniel about anything...!
>
>SIDE ANGLE
>
>all you do is enter in your case and it ouputs 6 paramters
>
>Side1 Angle1
>Side2 Angle2
>Side3 Angle3
>
>You can't get more simpler than that :)
>------------------------------------------:
>but this would also solve some of the other problems inherient in
determining
>which sides and which angles are being provided...
>it would however require the use to understand that they are supposed to
enter
>in connecting sides and angles...
This is all well and good and I would agree that this is the way to go but
what is the difference between this kind of entry and the ASA SSA etc. type
? I have already written a program using the latter type entry and it would
be easy enough to change the front end but to what advantage?
WndaLouise
Roger asks:
This is all well and good and I would agree that this is the way to go but
what is the difference between this kind of entry and the ASA SSA etc. type
? I have already written a program using the latter type entry and it would
be easy enough to change the front end but to what advantage?
i think that this was thoroughly covered earlier on in this thread, but
essentially...
1) just because
2) so an idiot, that has no idea what they're doing, can use it
3) it's the kind of thing recreational programmers like to do
Roger Blake
WndaLouise wrote in message
<199805231849...@ladder01.news.aol.com>...
Thank for your thoughtful reply.
I guess I did not make my question clear. I was trying to determine if the
program is to say take any two angles and a side and compute all the
possible triangles (3 of them) or if the input is to ordered like [Side1]
[Angle1] etc where there is usually just one solution. I think the posts,
including Joe's have been ambiguous on this point.
I guess I really don't care too much as to what motivates people to make the
program. Just because, works for me. It had seemed too me that if the
ordered input where [Side1] and [Angle1] are opposite from each other is
what is be asked for then that is really not much different from existing
programs. It would only require a rewrite of the front end of the existing
ASA AAS type program. I would not be adverse to doing this though I don't
really see how it makes things any easier.
On the hand if the program is to decide where the two angles are then we are
being asked to come up with a program that solves for the three solutions.
More challenging and more academic. And thats well and good. All I am
saying that most users of a triangle program are really after a solution of
the one triangle that is of interest to them.
So which is it? The former or the latter?
joe...@usa.net
Roger Blake writes:
> I guess I did not make my question clear. I was trying to determine if the
> program is to say take any two angles and a side and compute all the
> possible triangles (3 of them) or if the input is to ordered like [Side1]
> [Angle1] etc where there is usually just one solution.
Neither, but closer to the latter. Although SideA and AngleA are opposite
each other (this is the way it's done in all the textbooks that I've ever
taught with), PLEASE NOTE: the inputs are NOT "ordered". The user can put
the three known values in ANY desired order. This is radically different
*from the user-interface point of view* than the old ASA/SAA/etc user
interface which requires the input to be in a specific ORDER. One of my
previous posts in this thread had an example of specific keystrokes; another
example is given below.
> It had seemed to me that if the
> ordered input where [Side1] and [Angle1] are opposite from each other is
> what is be asked for then that is really not much different from existing
> programs. It would only require a rewrite of the front end of the existing
> ASA AAS type program. I would not be adverse to doing this though I don't
> really see how it makes things any easier.
Here's an example of how it makes it easier:
Suppose Side A = 15, Angle A = 60 degrees, and Angle C = 70 degrees.
== CURRENT METHOD ==
To solve this using existing programs, I have to rearrange these values
to some *specific order* that the programmer demands, type them into the
HP48 in that order, and then *find the menu key* that solves for *this
specific case* (SAA? ASA? AAS?). What order does the programmer require?
Which triangle case applies? Think carefully, because you'll get the
wrong solution otherwise!
== SUGGESTED METHOD ==
With the Kennedy/Jarrett/Donaldson interface that I've been describing,
you type any one of the known data (ANY one of them) and then press the
menu key with *its name* on it. No thinking required, no rearranging of
inputs ever, no thinking about ASA/AAS/etc. When the third input is made
like this, the solution is displayed, and each side and angle is then
available on its own named key.
== EXAMPLE KEYSTROKES ==
15 [Side A] \
60 [Angle A] > in any order --> solution, on the 6 menu keys.
70 [Angle C] /
No worries about input order. No worries about triangle case.
> All I am saying that most users of a triangle program are really after a
> solution of the one triangle that is of interest to them.
Agreed. A program that solves for all the permutations of the inputs
would be a fun programming exercise but is not my goal. I really believe
that the requested program would be useful both in the real world and in
school.
> I think the posts, including Joe's have been ambiguous on this point.
I hope I was more clear this time.
Guy Macon
>On the hand if the program is to decide where the two angles are then we are
>being asked to come up with a program that solves for the three solutions.
>More challenging and more academic. And thats well and good. All I am
>saying that most users of a triangle program are really after a solution of
>the one triangle that is of interest to them.
>
>So which is it? The former or the latter?
When I run a triangle solver, in my mind rotations don't make
different solutions, but reflections do. I would just make the
longest side the bottom and be done with it.
Another feature for the Joe Horn Perfect Triangle Solver Contest:
If there are two mirror image solutions, have a key that reflects
the triangle. Some might also like a rotate key, but I wouldn't
have much use for it, and it would complicate the part of the
program that displays the triangle.
I would be willing to write documentation for whatever program
meets Joe's criteria. I am a pretty good technical writer, even
if I don't always bother to spellcheck my usenet posts...
Roger Blake
joe...@usa.net wrote in message <6k7s6g$n97$1...@nnrp1.dejanews.com>...
>Roger Blake writes:
>
>I hope I was more clear this time.
>
>-Joe-
>
Thanks, I got it now.
WndaLouise
What is What...!!! ???
why hasn't anyone posted the definitive solution yet...!!!
i've been suffering from wiggly brain the last the couple of days/weeks and so
i haven't been making much progress on the assignment myself...!!!
and/or/but...i was also wondering if the program is supposed to work on Right
Triangles only or all kinds of Triangles...???
the approach i've been thinking of, is instead of waiting until a critical
number of values are input...
( also, if it's a right triangle only approach, then angle C would always be
90...so why is that key on the menu...??? )
...the program would have an equation for each variable, then ask if the
variables that each unknown variable -needs- (???) is available, and if so,
then it solves for that, then cycles around again, if any unknowns were
found...and then checks to see if all the variables are known yet, and if so,
then it quits...!
this sounds like it would need a CASE structure though...
is such a thing absolutely forboddin...???
also: a little graphic at the beginning might be useful of what sides and
angles are called what...???
joe...@usa.net
WndaLouise wrote:
>
> i was also wondering if the program is supposed to work on Right
> Triangles only or all kinds of Triangles...???
All plane triangles.
> this sounds like it would need a CASE structure though...
> is such a thing absolutely forboddin...???
Foreboding, perhaps, but neither forbidden nor verboten, as long as
it's fast. The built-in multiple-equation solver works sorta like
that, but it's *way* too slow.
> also: a little graphic at the beginning might be useful of what sides and
> angles are called what...???
I agree. An optional graphic of the solution might be useful too.
The same code could produce both graphics, of course.
Gerald Squelart
joe...@usa.net wrote in message <6kb58d$vte$1...@nnrp1.dejanews.com>...
>WndaLouise wrote:
>> i was also wondering if the program is supposed to work on Right
>> Triangles only or all kinds of Triangles...???
>All plane triangles.
Discrimination!
It must work on 3D triangles.
Jean-Yves also wants spheric triangles, but I think it's a little hard in
this universe.
Gerald.
I don't speak in a curved space when I post here.
Guy Macon
Solving triangles that are on the surface of spheres is also a
legitimate endeavor. It is outside of the scope of what Joe
is looking for, but there is a use for such a program. You
would have to tell the program how big the sphere is. of course.