Preferred TI calculator???


Ok, so we are HP-calculator nuts! I have two questions regarding TI calculators?

1. Which vintage TI calcultor did you use/own besides your HP beauty and what did you use that TI beast for?

2. Which graphics TI calculator do you use now and what tasks do you use it for?

Happy Programming!




TI 58
TI 58C
TI 59

By the way, who knows how to repair the card reader of the TI 59
I know this is not the correct place here, but I take the opportunity....


I was actually a TI user until the HP41c was announced.

TI: More memories (TI59 - 100, HP67/97 - 26?) and more program steps (960 vs. 224), which really DID allow for more complicated problems to be programmed at times in the TI machines...and they were so much cheaper for a poor student!

But, programming in AOS was really painful.

Then the HP41c was announced and I was transfixed. Alpha on a calculator! Redefined keys! Forgot TI even made calculators! :-)

The TI's I used until then were: SR-16II (my first), SR-51A, SR-52 (bought second hand for $70 in 1978), TI-58C.

Graphing TI model? Forget it. Have a TI-84plus in case my son has to use that specific model in HS math classes, but otherwise...


Gene wrote:
> TI: More memories (TI59 - 100, HP67/97 - 26?) and more program steps (960 vs. 224)

not "and". Unlike the HP-67/97 where you had 224 steps AND 26 registers, the TI-58/59 had a movable partition (like the HP-41) so you could change the split between program and registers. The default config was 480 steps and 100 registers, so with 30 registers you had 720 steps which *is* more than the 224 steps on the HP67/97, but the comparison is deceptive as the HP67/97 had more merged instructions (e.g. GTO A would take one step on the HP-67/97 but two steps on the TI-58/59.

All in all I'd say that the TIs could support programs twice as long as the HPs, which was a real advantage.

> and they were so much cheaper for a poor student!

Indeed, but they were of significantly lower quality. The keyboard in particular was horrible.

> But, programming in AOS was really painful.


Though, you forgot to mention the printer (PC-100C). While the HP-97 beats the TI-59 + PC-100C hands down for quality, elegance, usability and looks, the TI approach allowed you to move up to a printing calculator when you could afford to do so while retaining the portability of a handheld calculator. With the HP-67/97 you need to have both of them (at a cost of more than $1000, serious money even nowadays).

For some strange reason, the PC-100C was a quality product. It was well built, solid, and was FAST! Boy was it fast! It had a row of thermal dots so it could print an entire line of dots at one time without moving the head (the paper moved against the print head).

Compared to the PC-100C, printing on an HP-97 gives you the impression that the 97 is doing you a favour :-)



good points...quite true!


My experience is that the PC-100A was a higher quality product than the PC-100C. The PC-100A was also quieter. To see the ultimate in what could be done with the TI-59/PC-100 combination go to Viktor Toth's site at, click on "The Library, click on "Texas Instruments", click on "PPC Notes" and click on "Volume 8 Number 2".


1. My first calculator: TI-57. Then I borrowed a TI-59 from a friend (with a PC-100C printer). I used the 57 to learn to program and the 59 to perfect it. I used the 59 to create self-modifying/program generating code.

2. None. I was grabbed by the HP-craze after my TI affair :)


The best TI calculators:

A) My wife's TI-92+, after it sells on eBay.

B) Any of them at 9.789 m/s^2. As you can see from the equation, the resulting improvement varies as the square of 's'.


Going back to school in my early twenties, I used a succession of TI-30s and TI-35s for ordinary scientific calculations. They would break, and I'd buy another one - completely disposable. I wanted a TI-59 because one or two other students I knew had them. But before I could get enough money together out of tutoring wages, I saw the 41C in the bookstore. The alpha capability, and the expansion just floored me. I saved a while longer than I might have otherwise and bought the 41, then a couple of memory modules, then a quad memory module, then a card reader. That machine was stolen from me as I slept on the lawn after class one afternoon. (Laziness was ever my downfall) So I saved and bought a 41CV. Any thought of a TI calculator was banished by the 41.

I recently bought a TI-59 + PC-100C printer. The 59 lacks a battery pack or battery cover, or I might play with it more. I also bought a TI-92+ to see what the last generation's top-of-the-line was like. I intend to play with them both some more. However I still have lots of projects lined up on HP machines, so it may be a while until I get around to it.




I received a TI-30 (original LED version) for Christmas 1978 (I think) and was quite delighted. I had been intrigued by some Casio models at the time, and had seen expensive HP models before, but the TI met my needs perfectly.

Unfortunately, it had a 9V rechargeable NiCd battery that quickly lost its ability to hold a charge. One had to plug the calc into a nearby wall socket to use it. I started using disposable 9V batteries, until one of those went dead in electronics class in 1980-81. I hooked up the calc to a regulated DC power supply at 9.0 V, but might have done so backwards, as the display broke up and died.

The TI was replaced in Fall 1981 by a Casio fx-3600P (I hadn't the chutzpah to ask for a US$300+ HP-41C!), which I used until I bought an HP-15C in Fall 1983.

I bought a cheap TI-36X as a disposable for "old times' sake" in 2002 (or so), then got into HP collecting that year.

I'm intrigued by the TI-89, which seems to be an easier-to-use product with better firmware than the HP-49 series. Several of my younger colleages have TI-89's. However, I haven't yet committed to get one.

-- KS

Edited: 28 Jan 2006, 9:18 p.m.


I don't have any experience with any TI's newer than the 59, but I got an awful lot of good use from first the 58c and then the 59 before I had the HP-41cx. I also had the printer and used its alphanumeric capability a lot. I had four modules (the most valuable being the electrical engineering one), and a bunch of applications books, and a pile of users' group publications. My biggest program needed about 1500 bytes, so it would prompt for the appropriate cards if you'd call a part that wasn't in the memory at the time. That would of course mean overwriting another part; so if you needed that again before you were done, it would call for the other cards again. The major weaknesses of the 59 that immediately come to mind were the batteries' lasting less than four hours per charge, the memory's forgetting everything when you turn it off, only being able to hold one program at a time (unless multiple programs were planned around each other so they didn't require the same memory space and labels), only being able to hold one module at a time, the modules' only being able to hold user code, and the resulting slowness. There was no alpha capability without the printer, and the keyboard was cheap, and there were a lot of other less-major weaknesses; but the one thing that really drove me to the HP-41 was the fact that the TI had no I/O that could be interfaced to anything else. Still, the owner of the website had the gall to say the TI-59 is the best calculator in the world. Go figure.


The TI-59 was clearly the most powerful hand-held calculator during the three years from it's introduction in June 1977 until the introduction of the HP-41 in early 1980. It had more user memory than the HP-67, it had the Solid State Software modules which offered a wide range of subroutines which could be accessed by user programs, and it had a printer connection.

I didn't purchase an HP-41 when it came out because I couldn't afford it. When the time came that I could afford it the Radio Shack Model 100 had become available. I bought a Model 100 which offered a fourteen digit mantissa versus the ten digit mantissa in the 41, programming in a higher order language (BASIC) as opposed to the lower order language (RPN) in the 41, a built-in RS-232 connection, a built-in modem and a screen which supported graphics years before Casio offered the first graphing calculator.

Currently, I have a working HP-67, HP-28S, HP-41C, HP-41CV and HP-41CX all of which I purchased at thrift stores and garage sales. I got the CX and a fully functional TI-59 for two dollars each at a thrift store where they were being sold in the same bin with TV remote controls!


A company known as American Microproducts sold devices which allowed the user to connect multiple Solid State Sofware modules to the TI-59 calculator. I have one where the user manually switches the modules. I sort of remember that they also sold a unit where the user could switch modules under program control but I haven't found the reference.


1. Which vintage TI calcultor did you use/own besides your HP beauty and what did you use that TI beast for?

1A I presently own: SR50, SR52(3), TI58/58C/59(some) and PC-100, SR10, TI30, TI2550, 74S(3), TI68.

But the only one with some "character" is the SR22, a nice calculator!

1B Neither I ever used them in the past nor I do today.

2. Which graphics TI calculator do you use now and what tasks do you use it for?

2 I only own a Voyage 200: tried it for a little while (since I once used Derive for DOS) and came back to HP.



1. In 1983 as a Physics sophomore I used a TI-59, not that this was really needed but because I liked it. I remember I used it to evaluate a series for computing Pi I had 'discovered' using Fourier's series... a very very slowly convergent series first described at least two centuries before by a Fourier's countryman as I found out later... Oh, I used to use it at work too, as a military technician. My colleagues were amazed when I measured the diameter of a copper wire, enterered it on the calculator and it immediately displayed the AWG number :-).

The calculator seemed very powerful then, mostly when I remembered just four years earlier I did my calculations with the help of a logarithm table (calculators were just unaffordable then and slide rules didn't provide the accuracy I needed..., I didn't have one either). Too bad thieves broke into the house I lived with work colleagues then while we were at work and took it away... But then I would not have moved towards HP...

2. None. I just cannot get used to algebraic calculators any more. I always make mistakes when I try. I know there are people here who switch easily from one to another. Perhaps I should try harder... My inability to use the Windows calculator once made me (almost successfully) try to replace it:

My (brave) attempt to replace the Windows calculator




MY all-time favourite is still the Datamath!


I started with TI calculators at school when I was 16. The only caculator I had ever seen before was the HP 35 (see my Memories). HPs were simply unaffordable back then so I went for an SR-51A. At that time my teacher just had an SR-50 while my classmates were still using logarithm tables and slide rules in the math and physics classes.

Still at school, I switched to my very first programmable, an SR-56, giving my SR-51A to my sister. I added the PC-100A printer (the one that could connect to three different calculators, namely the SR 52, the SR 56 and the TI58/59 series.) This combo was seriously tweaked as my programming workbench but nothing is left from that era.

Later at University, when I had just seperated from my girl friend, I had left some spare money originally saved for a vacation for two. I spent it on a TI-59 (kind of revenge: my former girl friend's new lover just had an SR-52!).

The TI-59 was my main computing device in the first few years at university. I wrote some software for it and even made one listing appear in the German computer magazine "Chip". When the HP 41 appeared it was too late: I soon switched to a "Home Computer" (EACA Video Genie, TRS-80 clone) in 1980 or 81. This made my interest in calculators fade and it was the starting point for my current profession - I'm a professional programmer for over 20 years now.

From time to time I picked up some interesting machines just to play around a little: Casio PB-700, TI CC-40, TI-74 Basicalc, TI-95 Procalc, Sharp PC-1261. I still like them all. My first genuine HP was the HP 200lx, my "brain substitute" for several years, until replaced by a Psion netBook.

Some time ago, eBay provided a chance to get back to all my old calculators which had either failed or were sold or given away since long.

Now I own a bunch of HPs (11C, 12C, 15C, 16C, 25(C), 28S, 33S, 35, 41C/V/X, 42S, 45, 48G, 49G+, 97, 200lx, 720), Casios, Sharps, an Epson HX-20, a TRS-80 Model 100, and, of course, many TIs. I still don't really use any of these but I love to play around with them, solving more or less complicated problems mentioned in this forum. (Thanks to Valentin and other contributors who have helpd to awaken my mathematical brain which had slept for so long!)

I have the following working TIs in my collection (in numerical order):

TI-25 (early LCD)
TI-30 (original red LED)
TI-30 (original red LED) transplanted in TI-45 case
TI-36 Solar (LCD)
CC-40 with peripherals (Quick-Disk, RS-232, plotter, 80 column printer) and modules (Pascal, Math)
SR-50 (first scientific LED, no parentheses)
SR-51A (early LED, statistics, still no parentheses)
SR-51 II (2nd generation LED)
SR-52 (first programmable) with PC-100 (the original TI printer & security cradle!)
TI-56 (LCD)
TI-57 (3rd generation LED)
TI-57 II (LCD)
TI-59 with PC-100C and additional statistics module
TI-66 (LCD), made by Toshiba
TI-74 (S & Basicalc) with peripherals (PC-Interface, small printer) and some modules (math, statistics, finance, Deutsche Bank Leben)
TI-81 (first graphics)
TI-84 Plus
TI-92 Plus
TI-95 Procalc, sharing the peripherals with the TI-74, and some modules (8k RAM, Math, Statistics)
Voyage 200
CBL & CBR, Calculator Based Ranger and Calculator Based Laboratory with some probes for connection to the graphics calculators.

Many of the LED machines have such a bad keyboard that I barely use them. The older machines (50, 51, 52, 56) are better then the later models (30/57/58/59).

From the next generation, I like the TI-95, the last keystroke programmable from TI, and the Basic programmables. Playing around with the old peripherals is pure fun.

On the more modern side, I like the TI-92 Plus. This machine is solid like a brick, has a QUERTY keyboard and a huge display. I mainly use its connection to the CBL and/or CBR. The latter are educational devices, not serious lab equipment, but they provide a well designed link between the calculators and the "real world". You can have the CBL collect temperature and lighting data during 24 hours and have the calculator graph the results the next day. CBL can be operated on external power which saves on batteries.

CBL and CBR work well with most of the other graphics machines (except 81, rexstricted on 85). Software is available for free, mostly written in TI user code (sometimes referred to as "TI-BASIC" which it isn't.)

The Voyage 200 has the same operating system as the TI-92 Plus but more memory in a smaller footprint. I don't like the keys which are a little too tiny. While the TI-89 has the same CAS and is even more pocketable, its keyboard layout isn't very well suited to the vast functionality.

From the non-CAS graphics I prefer the 85/86 series for which production and development has sadly stopped, in favour of the less capable 81/82/83/84 line. The latter have only one-letter variable names and a less capable programming model while the former have long variable names and a more complete programming dialect. I like the 85/86 menu system better then both the systems in the simpler series and the CAS machines.



Ok, so we are HP-calculator nuts! I have two questions regarding TI calculators?

1. Which vintage TI calcultor did you use/own besides your HP beauty and what did you use that TI beast for?

2. Which graphics TI calculator do you use now and what tasks do you use it for?

Happy Programming!


1. None. I gave up on older TI's after purchasing an HP-34c in 1980.

2. Non-graphing: My 12 year old son uses a TI-34II periodically to check his pre-algebra homework. His teacher doesn't allow calculators in the classroom. The 34II is a nice machine and I found it on sale at Office Depot for $2.97.

Graphing: I have a TI-83+ Silver Edition that I use from time to time to graph functions, manipulate matrices and calculate statistics.




My favorite TI is whichever of the many TI36X solar-powered units that I have picked up at a local flea market.

I have written several times about buying them for $6-7 each when buying several of them at once. They are still selling around here for about $19. I go back to that well every now and then for more.

For that price, I can give them away, drop them, travel with them, or loan them out without worry. Just pull the drawer open, pull out another unit, and open the blister pack.

While the key feedback is not great, they are dependable, have plenty of computing power most of the time, and have a hard case. If I need more power, I pull out one of my 48GX's or turn around to the PC.


I used a Ti-55 before switching over to an Hp15c. Ti even exchange my old 55 for a newer Ti-55ii, but its keyboard still sucked! I had purchased an Hp15c when the keyboard of the Ti-55 was shot, then heard about the exchange program that Ti had. Until I got my new Ti-55ii, I was thinking about what a dope I was to buy a $100 calculator so quickly. That thought changed when I got my hands on the new Ti-55ii. Yep! Same Ti keyboard, glad I have an Hp, is what went through my head.

A friend of my bought the new Ti-66 and was really happy with it. Years later, I bought a Ti-66 and can see why (of course the Ti-66 was really made by Tosiba).

I have a good representation of the Ti graphics line, basically one of each family. I really have a dislike of the Ti-83/84 line as it is crippled in so many areas (but that is by design as Ti can sell another graphics, when you out grow this model). It does work as a general purpose graphics calculator. Shamefully, Hp followed suite with the Hp38/39G family as well.

I myself like the Ti-86, simple, yet powerful, with an easy to read screen. The Ti-89 may be far more powerful, but I find its units conversions awkward to use in comparison to the Ti-86, the screen very difficult to read, and its keyboard layout is awful (second shifting the trig functions).


One of the real problems with the TI-82 and T-83+ is that there are some instances where the devices fail to calculate determinants properly. I have seen two instances where my TI-82 and TI-83+ calculate the determinant of a matrix as zero but my TI-85 calculates the determinant properly.

The first case appeared last year when I was investigating the ability of various machines to find the determinants of sub-Hilbert matrices from order 7 to 10. The TI-83+ and TI-85 gave identical answers for orders 7 through 9. For order 10 the TI-85 gave the answer 2.207089...E-53 where the decimal equivalent of the exact answer obtained from the HP-49G is 2.164179...E-53 . The TI-83+ gave zero. I did not have a TI-82 at the time.

The second case appeared this year when working with the following linear equation problem proposed by Valentin Albillo:

1.3 x1 + 7.2 x2 + 5.7 x3 + 9.4 x4 + 9.0 x5 + 9.2 x6 + 3.5 x7 = 45.3

4.0 x1 + 9.3 x2 + 9.0 x3 + 9.9 x4 + 0.1 x5 + 9.5 x6 + 6.6 x7 = 48.4

4.8 x1 + 9.1 x2 + 7.1 x3 + 4.8 x4 + 9.3 x5 + 3.2 x6 + 6.7 x7 = 45.0

0.7 x1 + 9.3 x2 + 2.9 x3 + 0.2 x4 + 2.4 x5 + 2.4 x6 + 0.7 x7 = 18.6

4.1 x1 + 8.4 x2 + 4.4 x3 + 4.0 x4 + 8.2 x5 + 2.7 x6 + 4.9 x7 = 36.7

0.3 x1 + 7.2 x2 + 0.6 x3 + 3.3 x4 + 9.7 x5 + 3.4 x6 + 0.4 x7 = 24.9

4.3 x1 + 8.2 x2 + 6.6 x3 + 4.3 x4 + 8.3 x5 + 2.9 x6 + 6.1 x7 = 40.7

where the system has the unique solution:

x1 = x2 = x3 = x4 = x5 = x6 = x7 = 1.0

The exact determinant of the matrix is 1E-07. My TI-85 gave 0.990927530944E-07. Both my TI-82 and TI-83+ gave zero. This is not an easy problem. Determinants yielded by some other machines were

HP-41 Math Pac  -86.31E-07 
HP-28S 0.0703238892937E-07
TI-59 1.024890772918E-07
TI-95 1.002914557528E-07
Rodger Rosenbaum modified Valentin's problem by changing A(3,7) from 6.7 to 6.8, changing A(7,7) from 6.1 to 6.0, and changing B(3) from 45.0 to 45.1. The exact determinant of the matrix becomes 798.9713675. The TI-82, TI-83+ and TI-85 calculate the value 798.97136749978 . Determinants from some other machines were
HP-41 Math Pac    798.9713752 
HP-28S 798.971367542
TI-59 798.9713674893
TI-95 798.971367499
Rodger's changes have essentially dumbed-down the problem so that almost any machine can solve for the determinant and get reasonable answers.

If one goes ahead and solves the linear equations using X = INV A * B then the solutions from the TI-82, the TI-83+ and TI-85 are exactly the same. If one uses the simultaneous equation function of the TI-95 one get a slightly more accurate solution.

The TI-82, TI-83+ and TI-85 all use a fourteen digit mantissa so it is not surprising that the three machines get the same results when they are asked to do the same thing. I can't explain the occasional improper evaluation of a determinant.


Hi, Palmer:

Palmer wrote:

"Rodger Rosenbaum modified Valentin's problem by changing A(3,7) from 6.7 to 6.8, changing A(7,7) from 6.1 to 6.0, and changing B(3) from 45.0 to 45.1. The exact determinant of the matrix becomes 798.9713675."

    I fail to see how changing the original problem so that you can then get 798.9713675 as the determinant's value with utmost ease has any relevance at all to finding the original determinant's value which is 0.0000001 instead.
"Rodger's changes have essentially dumbed-down the problem so that almost any machine can solve for the determinant and get reasonable answers."
    This sentence is almost surreal. This is not 'dumbing-down' the problem, this is changing it altogether. Now any machine can solve for the determinant and get a "reasonable answer", namely 798.9713675 when the true answer to the actual problem is 0.0000001 !! Must be a joke, isn't it ? You're not serious, right ?

Speaking of jokes, this whole situation strongly reminds me of this well-known joke:

    "A man who has lost his keys in the dark on one side of the street searches for them under a lamppost on the other side. When asked why he is looking over there, he says, "There's better light

This man would do well to search for his keys where he lost them, if he's to find them out.

The moral is clear: if your solving methods or machines can't deal with the original problem, you'd do well to change your methods and/or machines, as solving *another* problem simply won't do. You'll never find your keys that way.

Best regards from V.

P.D: By the way, never mind TI's, my SHARP PC-1475 has no problem at all solving the original problem, what with its 20-digit double-precision math.
You see, I changed the computing device, not the problem.



I admit that I don't really comprehend where we are going with all of this. I have been following Rodger's lead trusting that I will learn something in the process.

If you go back and review some of the threads covering Rodger's work with your original problem and with modifications to it you will find that his first two modifications only involved changes to single elements of B. Only fourteen digit machines delivered reasonable results. Other machines delivered X elements which were an order of magnitude in error and of the wrong sign. Rodger's third modification involved the changes mentioned to A(3,7), A(7,7) and B(3) previously mentioned. All machines then got excellent results, even my Model 100 operating in single precision mode which then has a mantissa of only six digits.


In answer to Namir's two questions:

1. Which vintage TI calculator did you use/own besides your HP
beauty and what did you use that TI beast for?

My first TI was an SR50. I had wanted the HP35 when they first
came out, but $395 for a calculator remained too high a price for
me for quite a long time. When TI came out with the SR50, I
perceived it as a cheaper, but equally capable, alternative to
the HP35. I knew before buying it that it was a lesser quality
machine than the HP35, but the SR50 was still a huge advancement
over my Casio "Minuteman III" 4-function calculator. I waited for
the SR50 to drop to $140 in 1975 (March?) before buying one. (I
was first made aware of the SR50 when they sold for $170.)

I was pretty pleased with the SR50 for most of Spring quarter.
(It was particularly exhilarating to have it for a Fluid
Mechanics course I took that term!) Then it developed charging
problems. I made it through final exams with it and then sent the
SR50 in for warranty repairs. The charging problem reappeared
during the summer. (Later I realized it wasn't a charging problem
as such but weak contacts between the calculator and the battery
pack---just bending the flat spring contacts out a little further
generally made the problem go away, at least for a while.)

In August of 1975, I saw an advertisement for an HP25C in Time
magazine. I decided right then to buy one. The idea of
programmability in a calculator had always fascinated me, but
having to type the program in again every time you turned the
machine off made no sense to me! (A Physics Professor of mine had
one, so I knew about the HP65 and its little magnetic cards, but
$800 for a calculator way beyond anything I could imagine buying
at the time.)

My HP25C arrived in the middle October of 1975 and I was very
content with it for quite a while.

NOTE: currently, the HP Museum site incorrectly lists the
introduction of the HP25C as 1976. (I've emailed David Hicks
about this issue.)

I will admit I was tempted by the HP29C when it came out, but I
managed to resist the urge to upgrade. My HP25C was my "good"
calculator and my TI SR50 was my "loaner" calculator. I knew my
two calculators were on borrowed time when I encountered the
HP41C in graduate school. I held out until the 41CV came out!
After becoming addicted to my 41CV, I sold the 25C to a friend. I
threw the SR50 in the garbage!

I later sold the HP41CV to help pay for an HP75C. In the spring
of 1984, I traded my "HP-IL to Centronics Parallel" adapter for a
new HP41CX. (The adapter was of little interest to me after I
purchased the HP-IL version of the HP "Thinkjet" printer.)

For a brief period, I had the habit of buying whatever the newest
and most expensive HP scientific calculator was and selling my
previous one. That practice started with the HP28C and lasted
through the HP28S and HP48SX. (I kept the 41CX all along as my
"loaner" calculator.) Later I bought the HP48G, HP49G and the

2. Which graphics TI calculator do you use now and what tasks do
you use it for?

I use the TI Voyage-200 for just about everything. Of the
graphing calculators I still own (including the TI-85, TI-84 Plus
SE, HP48SX, HP48G, HP49G and HP49G+), it is the only one I
actually use for graphing. I use it routinely for working out
solutions to homework problems for my Physics classes.

I bought a TI-84 Plus SE mainly for working with TI's CBL
("Calculator Based Laboratory") data acquisition device, but I
haven't had time yet to do anything with the CBL.

I realize that students need to worry about the QWERTY keyboard
of the Voyage-200 (and its predecessors the TI-92/92+) because
some standardized tests ban anything with a QWERTY keyboard. So
the TI-89 is sometimes allowed while the same calculator with a
QWERTY keyboard is not. Others just prefer the TI-89 form factor.

For old eyes, I find the TI-89 and TI-89 Titanium hard to use.
Calculations aren't so bad, but just try setting the preferences!
The text on the preferences screen is so microscopic that I need
to take off my glasses to read it! (I'm very nearsighted so
taking off my glasses is almost like using a magnifying glass!)

One further thing to note about the TI-89/92/Voyage-200 models is
that they finally allow the user to write their own "functions"
(as opposed to "programs"). On these TI models a "function" is a
user-written program that can be called by other user written
programs or used in a calculation. Being able to call my own
program from another of my programs is a capability I just took
for granted from my HP experience. Finding that I couldn't do
that on the TI-85 was a real disappointment at the time I got
that calculator.




Mark wrote:

"My HP25C arrived in the middle October of 1975 and I was very content with it for quite a while. NOTE: currently, the HP Museum site incorrectly lists the introduction of the HP25C as 1976. (I've emailed David Hicks about this issue.)"

Gene: Actually, in this case, you are incorrect. :-) The HP25C was introduced on July 1, 1976.

This was covered in the 1976 issues of the PPC Journal. I guarantee that the 1975 issues of the PPC Journal do not mention the HP25c introduction in that year. It was introduced in 1976.

Link to July 1976 PPC front page in PDF format

(Note: I took this page from TOS)

And, Craig Finseth's excellent dump of HP calculator information also lists the date as 1976.

Craig Finseth's HP25c page

So, very much more likely that somehow you've got the year's mixed up.


I can't help you on your publications Gene. The dates I cited are correct! That Fluid Dynamics course I mentioned was Spring quarter 1975; I saw the advertisement for the 25C in August of 1975; and I received my 25C in October of 1975.

I can't explain the discrepancy with your sources.



I know you believe you are correct, but multiple sources contradict your memory.

I believe probability suggests that your memory is the one that is not correct. That explains the discrepancy. Sorry!

The HP25C was introduced July 1976, not in 1975.

Any other HP users believe it was 1975?


The HP Journal indicates, along with other information on this site, that the HP 25 was introduced in 1975. The HP-25C was, in fact, introduced in 1976. Perhaps that is where the confusion or foggy memories come to play.



Interesting thread.

A small remark to begin with : the TI85 and all others can call programs from within programs : just type the target program's name and voilà ! it's called.

Now back to the questions :

1. Which vintage TI calculator did you use/own besides your HP beauty and what did you use that TI beast for?

After a wonderful Radofin 2660 that my dad offered to his amazed son "a few" years ago, I got a TI30, the legendary LED one. It served me faithfully through high school, but those machines were not really mandatory at the time. Math teaching rarely involved long strings of calculations, it was more theoretical. During later studies, my parents offered me (again) a TI58C, because of its powerful function set mixed with the constant memory, a must. At that time funding was extremely low and I knew HP only by name. Honestly I was not impressed : machines 4 times more expensive had to offer 4 times more 'things', and obviously it was not the case. I never had quality problems with my TI, maybe because I knew that the price of my beauty was very high from my point of view, so I was taking *real* care (I wonder if some quality problems on TI57/58/59s are not due to bad care - granted some later TI models were low low quality). I was using this machine for school related subjects, and for programmming the usual fan stuff : multiprecision, tricks, prime factors, special functions...

2. Which graphics TI calculator do you use now and what tasks do you use it for?

The TI I use now is a TI83-family member, because when commuting I can't always take good care of my machine, so these are considered expandable. Also, it seems they are quite easy to program, speed is reasonable and the function set is enough for my stuff.

My interets now are number theory things, like finding how to factor (3+7*sqrt(5)), (quickly) calculating the Mascheroni constant, what is the value of the infinite sum of i^p/i! for a given p, what is the curve limiting the domain of (n,p) for which the Comb(n,p) < 10^500, why the continuous fraction representation of the square root of any integer is periodic, ...

For this kind of problems CAS machines are completely unusable, so I need higher end models only for pure calculating power, and actually the game is to use better algorithms instead of brute force. So low-end machines are OK, provided they are easy to program.

As a mater of fact, no current machine fits that bill :

- HP49G+ : uses RPL... yuck

- TI89 : functions can't access global variables !!!

- ClassPad 300 : functions can only be a single mathematical expression (this one is the most wounded of all)

- other graphical machines : no local variables, can't pass parameters

- vintage Basic machines : some could be good, but we got accustomed to big screens. Granted, I like the HP71B and the TI74 which have real functions / procedures. If the 71B had a 4-lines screen, it would be in my pocket all day.



- TI89 : functions can't access global variables !!!

From a theoretical (information technology) point of view this is a desirable feature! Why? These functions cannot have side effects, the output is completely defined by the input. If you really need side effects, just write a procedure that acts on global variables.

If the 71B had a 4-lines screen, it would be in my pocket all day.

That would be equally true for the TI-74. As a student I was looking for such a beast and came across the Casio PB-700. Nice screen, reasonable keyboard but way to sloooooooow.

Another very interesting machine is the Casio PB-2000C which is programmable in a C dialect (pre ANSI, no switch/case).

(All links go to Viktor Thot's excellent RSKEY site:



There are many cases when those side effects are desirable.

For example a function :

1. based on a table of coefficients

2. which returns many results

3. which output trace information

4. which is defined based on itself and can be accelerated by storing the result of previous calls

5. very deeply recursively defined, overflowing the built-in subroutine stack

6. which needs some input at some point (yes, this one is weird)

7. not numeric, but handling the user interface (example CreateWindow()...)

By the way, procedures cannot access the caller's local variables either, and there is NO EXCUSE for this.

TI-ish discussion, so OT.



It's no at all OT, because this is a general programming question.

Just to keep other readers informed what our discussion is all about:

    Functions (as implemented on certain machines) must not have side effects by modifying global variables or states while programs are allowed to do so but cannot return values easily.

    I think it is a good idea to restrict functions in this way but GE complains.

GE, let me answer some of your points:

>> There are many cases when those side effects are desirable.

>> For example a function :

>> 1. based on a table of coefficients

    Put them in a { list } and pass them as an argument.

>> 2. which returns many results

    Return a { list }.

>> 3. which output trace information

    I agree, that's a real problem if you cannot write to the environment outside the function.

>> 4. which is defined based on itself and can be accelerated

>> by storing the result of previous calls

    You still need a way to initialize such a function. If you just call it with arbitrary parameters, how do you know the current state?

    If you really need state inside a function it is better to write a program instead and have it return a value (you can pass a variablename as a string and store the result into it or you can create some global variables in a special directory an modify these.)

>> 5. very deeply recursively defined, overflowing the built-in subroutine stack.

    That can probably be better handled by unrolling the recursion into a loop.

>> 6. which needs some input at some point (yes, this one is weird)

    You'll need to rewrite it as a program. It's better to collect al input beforehand and pass the values to the function.

>> 7. not numeric, but handling the user interface (example CreateWindow()...)

    Again better suited to a program instead of a function.

>> By the way, procedures cannot access the caller's local variables either,

>> and there is NO EXCUSE for this.

    There is an excuse! At least for functions which are called from somewhere but are defined elsewhere. How should these functions ever know about local variables defined in the calling function or program? I agree partly in that it is very restrictive that even local programs cannot access the local variables of their defining program.

There is a paradigm called "Functional Programming" (Wikipedia) which relies totally on side effect free coding. Everything is treated as a function!


Edited: 2 Feb 2006, 9:06 a.m.


Marcus wrote:

" hat would be equally true for the TI-74. As a student I was looking for such a beast and came across the Casio PB-700. Nice screen, reasonable keyboard but way to sloooooooow."

    The SHARP models PC-1350, PC-1360, PC-1600, and PC-E500/E500S, among others, all have 4-line, dot-matrix quality graphical displays. And none of them is 'slow', quite on the contrary the SHARP PC-E500 is among the fastest vintage handhelds you can find. It's really fast, some 4 times faster than an HP-71B.

    Nevertheless, as I have as many as four or five SHARP PC-1350 models, all acquired en-masse in eBay for peanuts from a seller that was clearing some obsoleted surplus military equipment (which used preprogrammed PC-1350 for artillery computations), I use one of them all the time, as an all-use, scratch machine, and it's as comfortable, easy to use, and powerful programming and calculator environment as it gets.

    I did enter several very short programs I specifically wrote to compute integrals (Gauss), solve for roots (enhanced Newton) and do matrix inversion, and I can do with it most anything I want or need, very fast, and still got 12 Kb or so for one-shot programs or variables.

    And the extreme usefulness and ease of programming afforded by the 4-line screen can't be stressed enough, a real pleasure. Getting back to the 22-char, 1-line display of the HP-71B feels unbearably cramped in comparison.

"Another very interesting machine is the Casio PB-2000C which is programmable in a C dialect (pre ANSI, no switch/case)"
    I do own a Casio PB-2000C. It features a pretty good, standard C language, with filesystem and all, but unfortunately it always runs interpreted instead of compiled, so if you're used to fast, compiled C programs you'll find it slow in comparison. It also accepts plug-in modules, and one of them does provide BASIC language programming so you get the best of both worlds in a single machine.

    Matter of fact, I'd rather prefer HP had issued a C/Assembler ROM for the HP-71B instead of the actual FORTH/Assembler ROM. After all, there's also a full FORTH-language system in the 41C Translation ROM, so who needs two incompatible, distinct FORTH implementations ?

Best regards from V.

Edited: 2 Feb 2006, 9:52 a.m.


Preferred TI calculator??? "Does not compute", said the mechanized female voice (known only as "computer") from the original 60's Star Trek episodes. j/k :-)


ps. OK, I'll admit I had a TI30 in middle school in the seventies. Currently have a TI Programmer (also from the 70's), a TI ProCalc95 w/ printer, a TI92, and a TI Voyage200. Anyone interested in any of these (trades for HP's preferred)?


I have a like new made in the USA HP-12C in a slip case with a spiral bound manual. Is your equipment in working order? I don't need the TI-30.

Possibly Related Threads...
Thread Author Replies Views Last Post
  OT: Simulating a TI calculator with crazy 11-bit opcodes Egan Ford 8 1,418 08-13-2013, 12:06 AM
Last Post: Paul Dale
  OT: anyone use the TI BAII Plus calculator a lot ? Gene Wright 1 614 06-15-2013, 01:08 PM
Last Post: John B. Smitherman
  New solar TI calculator Mic 8 1,316 12-18-2012, 04:49 PM
Last Post: Chris Smith
  Yet a new TI calculator with color screen ?! Mic 20 2,435 11-24-2012, 12:30 PM
Last Post: Al
  OT: TI 36X Pro and Wishes for a solar RPN calculator Eddie W. Shore 15 1,830 05-04-2011, 07:16 PM
Last Post: Kiyoshi Akima
  TI introduces new calculator! Takes world by storm! HP quits business and asks forgiveness! Gene Wright 4 797 03-23-2011, 09:56 PM
Last Post: gene wright
  Calculator for CFA; HP 12C Platinum or TI BA II Plus Professional Sujith Abraham 23 2,645 03-22-2011, 11:46 PM
Last Post: Martin Pinckney
  TI-NSPIRE calculator plays DOOM Gene Wright 12 1,413 02-22-2011, 05:58 PM
Last Post: gene wright
  TI Selling new TI NSpire Touchpad Namir 9 1,370 04-20-2010, 06:20 PM
Last Post: Bart (UK)
  TI calculator collection Michael Meyer 30 3,000 11-19-2009, 11:14 AM
Last Post: Michael Meyer

Forum Jump: