I agree with the author absolutely - TIs should have some real apps, not just games.
I think that TI apps should center on providing current PDA functionality, and use that as a jump-off point for expansion.
TI calcs are ready for these types of applications.
I have written a simple set of apps for TI-89 in TI-BASIC, which are simple yet useful, and I'm just a beginner - all those REAL programmers using assembly out there should be able to turn out great stuff.

I dont know as TI-92 programming is going to get much more popular at all until a solid C/C++ compiler/cross compiler is developed for it. Almost everyone has to take C or C++ at some point these days, just like we had to learn FORTRAN back in the early 80s. Its just a given. Once we can achieve this, I see the wanted/needed boom in programming apps happening.

Thanks for your time.

The chief reason for the lack of complex assembly programs (word processor, spreadsheet, database, etc.) is that the TI suffers from an extreme lack of output devices.

First, there is the screen. Ok, so I spend several minutes typing something, and then I get to look at it. Useful.

Second, you can use the linkport. So far, this has had several uses: to output to another calculator, a PC, or *other* things, such as an IR link or a pair of headphones.

Basically, the main use of any complex application is ultimately to produce a paper copy, or to distribute something over the Internet. Typing something on a TI may be short-term efficient, but only if you are someplace without desktops or laptops. The only way I can think of that this situation would occur is if you are in school and have vast amounts of spare time.

Essentially, I believe that with the current level of technology in schools, complex programs such as the ones you mentioned are not needed.

--Bill Barksdale

Thats not entirely true. I am on a first robotics team (Rush 27) and we all hate scouting. so we had our programmer write an app that asked us simple questions and recorded them. It stored it on our calc until we exported it to a program on the PC. The ability is out there and there is also a need.Its just people who spend the time to do it.

I need Help learning ASM for the Ti-83+

Does anyone, ANYONE, have a copy of chasm 2 beta? Its supposed to be really cool, development stopped a long time ago, and I can't find it anywhere. I'm learning asm, and I need a competent, stable oncalc compiler. Thank you!

Since the posting of this article, calculator programming has changed quite a bit. There is now a high demand for advanced programs, which one can now store a lot more of on his/her calculator (due to the fact that TI has expanded the memory). The Assembly language has also dramatically improved, which is shown by the abundance of large and unique programs using it. The author practically predicted the future when he talked about the programs that should be made. There are already a few programs that allow you to compose your own music, there isn't an end to the list of Word Processors and Spreadsheets that have been created, we've made shells with much better qualities than previously available, and we're currently in the process of making web browsers (we already have telnet browsers, but they're very confusing to the average user and are in most aspects obsolete). However, I don't believe we've even tapped what kind of things we could do with our calculators. We could make cell phones, radios, universal remote controls, and pretty much anything else you can think of. As the programming languages advance, so will the things we can do with them.

Having been passively following the events related to assembly programming on the graphing calculators for about a year now, I have used both Fargo, Fargo II, and various programs which run under them--mostly games.

However, as stated by Douglas Ward in the article The Future of Assembly Programs, games do not realize the full potential of the graphing calculator platform. That article seemed emphasize the fact that, although games are fun, much, much more could be achieved. I couldn't agree more.

Douglas Ward presents several theories for the lack powerful applications for the calculators.

First, it was hypothesized that there are not enough able programmers... I have seen many quite capable programmers (or at least their work) on the ticalc website.

Secondly, it was hypothesized that the shells (such as Fargo) may not be robust enough to harbor a power suite of applications... This may be partially true, but the shells are not in any way restrictive, and one can easily overstep them.

Third, it was hypothesized that the calculators themselves are not powerful enough for advanced applications. Nonsense! Microsoft wrote the first version of Excel way back in 1984 that was designed to run on an MC68000 chip running at a measly 7MHz. Don't tell me the calculators don't have enough power.

Fourth, it was hypothesized that there aren't any good development packages for the calculators. Not true. There are many packages, designed to compile programs for the 68k Macs that work well for MC68000 assembly, and even C programming. THINK C, Symantec, MPW, CodeWarrior all come to mind, a calculator assembly program does not differ much from a code resource like an 'XCMD'.

A hypothesis I've thought of is that there isn't a market, that is to say, people wouldn't be interested in a powerful program on a calculator... However, the platform of the TI-92 is very nice... how many PDAs the size of a TI-92 have a QWERTY keyboard?

I maintain that the reason we aren't seeing more powerful calculator applications is the lack of cooperation among programmers. Great programs are not written in a day, or by one person. With the release of the TI-92 plus module, I am very excited that there may be a common platform, with which we can build robust applications. It will be necessary for programmers to work together. I agree with Achton Nick Netherclift, in the article Documentation Lacking, that the documentation in programs is exactly that--lacking.

With a group of programmers working on different aspects of the calculators, and each extensively documenting their work, more can be learned from each other, and bigger, better applications will certainly be a result.

The Linux operating system is a result of the colaboration of programmers, and an open source code environment. I am interested in working with the same idea in respect to the TI-92. Anyone else?

Hello Douglas Ward,

I just read your article The Future of Assembly Programs.

I disagree with many of the things that you said. The first thing is your want of "advanced" applications. These are graphing calculators, not laptops or palmtops. Their purpose is math. Of the examples you named, the only intriguing one you named was a spreadsheet. I think that a word processor would be a waste. Besides, in all actuality all of the calculators have one built in. In the 8x series you can use the program editor, and on the 92 you can use the text editor. I think that most of the advanced applications that you seek should remain on the platforms that we are all used to seeing them, namely the actual computers.

The other thing is that I don't think you realize how much of an actual PC's resources are wasted. Historicaly, as time goes programs have become bigger and sloppier then ever before imagined possible. As available memory increases, programmers don't feel so bad about using it all up for no good reason, and thats exactly what they do. Speed also becomes less and less of a problem. As to higher level languages and wizards and such, they have the same problem. Over all, the code that they output has become more bloated than ever. There are always exceptions, however. The biggest exception is GNU's C compiler, but that is probably an exception because it has been around and improved upon sinse a time when resources were scarce.

As to your theories, I think the 3rd and 4th are best. There are some points I'd like to hit upon though. In the 3rd you state that perhaps the calculators aren't ready for such advanced applications. Well, in my opinion, they are FREAKIN CALCULATORS for crying out loud! I think that what you are looking for is a palmtop. In the 4th theory, you suggest that if there were a compiler or something then perhaps these things would get done. Now you've entered into my domain. You see, I am currently working on a project to cure this need, one that I don't think you have heard of or you would have mentioned it. Its called HAL-IC, its a nice little compiler for the 86 that compiles TI-BASIC into assembly. Not just TI-BASIC, however. I have added a few extensions to make it more useful then that. It is very true that this project isn't near finished, but with the amount of work I've put into it there is no way I'm quitting any time soon. I am hoping on having a beta out by late August. Check it out at hal-ic.home.ml.org, you can find out a lot more about it there.

Perhaps if there was more of a calling for such advanced applications by the users of the calculators more programmers would be inclined to create them. Most people just want games though. Hehe, I can just imagine it now... Instead of asking to play tetris may friends will ask if they can type their reports... hahahaha.

Well, now that I know I can reply, I have something to say. Douglas Ward suggested in his article, The Future of Assembly Programs, that various programs be written for calculators. He gives some suggestions which I discuss in the next paragraph.

I must ask the question, "Why would you want a word processor?" I consider myself highly proficient at typing on a TI-85 or TI-86 keyboard, but I still only type at 20 words per minute. I can type 60 words per minute on my computer. Why would I want to type on my calculator when I can type three times as fast on my computer? Not only that, but how is it going to be printed? The application would only be useful for those calculator owners who also have graph links (which is surprisingly few). The spread sheet is a nice idea, but again, why bother? The numbers can be entered easily on the computer, and then they are already in a good spreadsheet program. If the numbers were entered on a calculator, they would have to be converted to Quattro Pro or Excel format. The database goes hand in hand with the spreadsheet, as many programmers know. Thus I have the same complaints about it as I do about the spreadsheet, combined with my complaints about a word processor. Music engraving? Why? Sure, it would be cool to be able to play a MIDI file on a calculator, but music engraving?!? What use is it unless you can print it? Again, a conversion program would have to be written: Calculator to Finale converter. Again, this limits the user base to those who own graph links.

I believe that his fifth (unnumbered) theory is the real reason behind a lack of so-called "powerful" programs. Programmers like to program games. They're fun to write, so they don't mind to do them for free. But I wouldn't write any of the above programs for free. They just aren't fun to write. All you get when you write a program is a few people who e-mail you and say, "Hey, your game is cool." No money, and really that's not much of a reward. This wouldn't be a major problem except calculator users seem to have a thing about paying for calculator programs. Since they don't want to pay, programmers don't make "powerful" applications.

I have more than once considered writing Windows NT for the TI-86. It would certainly work since I created a unique mouse routine. Bill Nagel said he would work together with me on it. But what would be the point? We would get no money, just a few e-mails saying, "It's cool". Not only that, but people on the Assembly-86 mailing list dismissed our idea as "pointless". I think it is quite clear that most calculator users use their calculator for two things: math and games. Sadly, an operating system like Windows NT, with preemptive multitasking and a message driven architecture would open the door for new applications, including word processors, spreadsheets, and databases, and even music engravers, to be developed (although I have already discussed the futility of such programs). But as long as the calculator users are not willing to pay for their programs, then I don't think any "powerful" applications will be developed.

I would like to say that Douglas Ward's article, The Future of Assembly Programs, had some good arguments, but I think the entire philosophy of assembly-on-calculators is being missed.

Having been following the development of ZShell and the assembly programs that came with it for several years now, I have seen many good programmers come and go. Who can forget the days of Henrik's Official ZShell Home Page which provided a nice list of the 20 or so programs available for ZShell 4.0 (including ZShell itself). Or the Unofficial TI Calculator Home Page? That was a long one to scroll through. :) Dan Eble started ZShell and then turned the project over to Magnus Hagander. There has been quite a bit of vaporware projects started and quite a bit of successful programs. Remember when Fargo was called MShell for a long time? It was thought to be vaporware until some "beta" tester finally uploaded it to some server. I have written some simple programs for the TI-85 and TI-92....ChemLib 85 and FChem. They are not great by any means, but they did do something for me. They allowed me to learn assembly, write a program, distribute it, and see people use it. It was fun. I've moved on now. I don't really do much with calculators anymore, and a lot of the original programmers in this community don't either.

I guess what I'm trying to say is this is not an industry, it's a hobby. People are learning things. Let it go at it's own pace and enjoy the ride. People find other interests and drop old projects. Live with it. I lost interest in assembly programming on the TI-85. It looks to me like Magnus Hagander has found some other interests besides ZShell. I'm not saying that any of these people did a bad job, but people move on. There's always a new generation of people interested in assembly on the calculators.

Use this as an opportunity to learn assembly, or see how a shell is written on the calculator, or whatever. But please, don't treat this like an industry. Who's in it to make money? Everyone is in it for fun....or am I wrong in thinking this is not like a Linux community?

Just my $0.02

Like a lot of users of TI calculators, I disagree with Douglas Ward in his article The Future Of Assembly Programs.

Let me expose my theories.

The first thing we must point out is that people who use these calculators are quite young people (more than 98% of users are below 24, see the Surveys section) and of course, are students. The aim of a calculator is to perform mathematical operations. Even when TI releases calculators that allow assembly programs, they don't want to make assembly-oriented calculators. A calculator must help you to pass exams, nothing more.

I've said users are young, and for the most interested in computers, and also in programming. Programming is not a knowledge you get from God, like mathematics, you must learn it. That's it, calculator users are students! They use and program their calculators all day. The best way to learn programming is to make little programs, often without any use, just as examples. The first theory of Douglas Ward is false. Programmers need practice.

The simpliest (and the funniest, don't forget we're working with young people, and school is boring, huh?) way to make examples of programming is to make games. It starts with Tic Tac Toe to end with Sorbo's Quest (which is quite powerful, don't you think?). Games, which can be played during a lecture, have few difficult functions to implement as a word processor would have and programmers can have inspiration into the world of gaming on computers.

It's the same reason why shells are not as powerful as Douglas Ward would want. If David Ellsworth wanted to make a very, very powerful shell, he wouldn't have made it for a calculator. He would have gone to Seattle to see Bill.

Programmers migrate to computers when they have become strong enough to make powerful programs for computers. They let their calculator do the math.

Calculators are designed to compute 2 + 2, or get a rounded value (you really need 25 decimals of Pi during an exam?) of results of functions, not for word processing or sound producing.

How do I see the future of assembly program? No changes. Games (more and more sophisticated with the arrival of new calculators), a few programs to view text files, and minimal shells.

That was what I wanted to say. Don't forget a calculator needs to know the result of 2 + 2, not to let the users write to his or her professor a nice letter with different fonts!

Bill at Seattle? I thought it was Redmond...

You are evil people. You really are. (All the replyers to Douglas Ward's article.)

I agree with Kevin Van Vechten in mosts respects (who, incidentily, is not evil.)

True, the user group for the TI calculators is young and students. True, they problably won't get a lot of use out of awesome OSs and big impressive data processors, buy who cares? True, the programmers don't get paid for anything.

So what!

The calculators, for all there limitations, are still computers, albiet old and restrictive. This represents a challenge, or should, to all programmers to program big, impressive, and amazing programs, regardless of purchasers being nonexistant and the probable users also being nonexistant. We should try to cram awesome programs on the calcs to see if we can. It will help us when we program for actual money by making us, forcing us, to write fast, optimized code. If we can do something difficult in assembly language on these calculators, we sure as hell can do it on big computers with high level languages on them.

The calculators prepare us for the future, and what better way to prepare then by stretching the limits of the "possible"?

This reply may not be very big, but I hope it gets my opinion across.

Thank you for your time.

First sorry for gramatical errors because I'm french and I don't speak very well in english... ;)

When you say that there is no reaon to put a spreadsheet or a GOOD text editor, I think you're wrong because :
1) YOU don't need that, but someone else can find that very usefull (like me). I haven't a notebook and if a want to write a mail, for exemple, I write it on my TI92 and after I send it by the computer... It will be better if someone (why not me ?) write a program which send directy from the TI when it is connected to the computer...
2) On the TI8x, it's sure, it's difficult to write, but on the TI92, I can write faster than with a pen ! (But not faster as a computer)
3) A functionnality is always a good thing ! Why do we use the last version of a software ? Not only but the bug corrections !

If you think the calc is not enough powerful, you're wrong ! If you want Win x.x, it's sure you can't (beurk !). But for exemple, look at Geoworks, this program is a REAL multitasking and it run very fastly on an old 286 !!! It is programed in Assembly, so it's fast and small ! If you are a GOOD programer, you use assembler, if you are a programer that use only high level language, it's sure your program will need many power and memory ! :(

Actually, I don't know very well the 68k, but I program on the 80x86 since 2 years (pascal since 3 years, basic since 6 years, and C++ since 1 year)

To have an idea of the power of the assembly, compare the scrolling of the TI's text editor and Xetal's one !

We can do very GOOD programs on a calculator (I'm thinking about the very usefull PCTools 98 and the debugger), and also a Multitasking kernel, why not ?

Ivan Fiodorovich Karamazov (Yefka)
Web : http://www.geocities.com/SiliconValley/4617
Sayena's e-mail : sayena@innocent.com

One of the ideas left out of that article is probably both the most feasible and most needed applications on the calculator: an assembly programming tool. Programs could be stored in normal program files or a nice pretty GUI could be made that would have online help and a list of the instructions. Maybe even a menu driven system of finding just the right instruction with bit, math, ld, port, and other major groups of functions could be implemented. After you pressed the button for one of these major
groups, a list of the instructions in the group would appear, then all the legal combinations for each instruction. Programming would be made much easier, and more possible when most people would use it most -- during classes.

I said this would be the most feasible of the large assembly programs because of two things. 1) speed does not matter. Even the BASIC compiler on the calculator takes a little while to tokenize a program. Why not spend the time compiling a much better faster program for assembly? 2) size won't matter. With the release of the expander and a driver for the TI-85, 86, and 92, any size program can be created. Since most compilers run through the source 3 (?) times anyway, if the program is to large it can be broken down into smaller parts. The editor could be deleted and replace with code containing the first run-through. That code could be deleted and replaced..... This would leave room for the largest of programs. If the program is too large, it too can be broken down into smaller parts, i.e. a function at a time can be coded and assembled.

I also said that this would be the most needed of any other program. If any other large programs are to be built, coders will need time to code
them. Since most of us do BASIC (at least in all my math classes), this would be the perfect time to do assembly. Since coders would have more
time to write in assembly, more assembly programs will be made, and they will probably be higher quality.

These are just a few of my thoughts regarding the future of assembly programming.

In my opinion, the article posted completely misses the whole art of TI programming.

First of all, why is assembly so important? Go to http://titan2.gunn.palo-alto.ca.us/GSNSTDIO.83G and download a beta copy of GSN Studio G3 which my friend and I have written. It's a TI-Basic math program for the TI-83 that spans from algrebra, geometry, trig, to command line, user-config, shutdown/booting, and even games. In the 14.5 K program files, less than 500 bytes are in assembly, yet it achieves the main purpose of the TI - math and games. Powerful math programs are important to further strengthen the power of the TI, but not word processors...

Second, spreadsheets. You can deal with matrixes, so why bother to have a separate spreadsheet? Word processor!!! Is it comfortable to type without a QWERTY keyboard?

Third of all, TI-asm is risky. A bad programmer can easily hang his/her TI, thus forcing him/her to take out the battery and clear the RAM. BAD. Therefore, it's not good for beginners. TI-Basic virtually never hangs the TI, because faulty syntax results in an error message, then quit, without crashing.

Fourth, why so many games?
Games are fun when you're bored in school, but since a TI-82/83/85 has only 27 K of RAM, and since games are so huge, what's the point of wasting those space when the RAM could be better used to store useful math programs.

Okay, here goes... My reply never got posted, so I'll have to start all over again. The reason that ASM is so important is because your 14.5K program is too large. It can be squashed down to far less than that, and I'm sure that even the most inefficient ASM code would beat out your equivalent Basic program in size and speed.
As for spreadsheets, a program would be useful, since the Matrices can't store and constantly recalculate formulae inserted into them. We need to use a similar data type setup, but we still hit a memory problem, not to mention that it would be useless if we wanted to use it under a real application on the PC, unless someone wrote up a PC-side program or a converter to a more useable format, such as Excel or Lotus 123 (yes, people still use Lotus).
Finally, I have to go with the concept that most utilities are overkill, but decent math, chemistry, and physics programs should be done in assembly, and maybe we should write up a few TSRs to allow for a better interface to the calculator itself, including things like symbolic I/O and other things that TI forgot to put in.
Also, I LIKE games. I've gotten bored in my AP calculus class so many times, and the only thing that kept me awake was the most recent game that I had slapped onto my 85. And yes, I did ace the course, so it's not like games can pose a significant distraction. Let the coders do what they want, and we'll see larger programs. Right now, we're seeing more collaborative efforts, such as the ZTetris project, and things like Super Mario 86 is definite proof of complex work. If we need utilities, let someone learn the assembly and work on it. I doubt all the well known coders here knew z80 ASM from the day that they were born, but instead, a few of them learned it as they went along. Stop complaining and follow that example.

IS there a current future to ANY assembly on the CALC's below the 85? I have seen many of the same programs running aorund, and not very many new ones, at least from where I'm sitting. If anyone out there is able to, I would love for you to prove me wrong. I want my self to learn more ASM coding skills, but it just seems that there aren't enough people willing to put forth their knowledge into a book, ebook, or difinitive reference on a site like TIcalc.org. It would be nice to have a pure reference on how to program, what calls do what, and find it all in one place. so far I have found many mixed items, and non helpful information, unless you are already experienced in ASM. Back to topic though, we are programming on CALCULATORS. read that again. CALCULATORS. it is amazing to me that people actually got THIS far on a calc, and maybe our standards are very high. It is not the most profitable, and glory filled job being able to program on a calc, and you will really not fill some large gap in society, or cure some major problem by doing this. The only reason people really do this, in my opinion, is because they enjoy it, and it only becomes a well organized process if we ALL join in and share info and keep updated. I am seeing, unfortunately, a lack of everything having to do with the calcs. updates are less and less, and the more experienced programmers move on to bigger and better things, and often do not pass down their pearls of wisdom to many. CAN we keep ASM and basic on the TIcalcs ALIVE is more the question in my mind. It seems to be wilting slowly, yet surely. Keep the faith is what we need, and more concentration on making good commenting on the work that actually makes it to the public. comment your code. ok i'm done ranting and raving. Please email me if you have ideas. Thanx!
Christopher

What a waste...

I bought my 92 to help with calculus and physics. I understand ward's disappointment in the state of things, if not accepting all his theories as to why it is so.

I enjoy programming, but there comes a point when you ought to try programming something more functional than the another version of blackjack. Looks like a lot of people think the TI is a Gameboy with math functions, not a calculator that can be exploited for games.

I'm not attacking people for writing games, just for not seeing the potential of the calculator. Look at how functional the thing is! If a word processor is not needed or practical, try telling that to all the PDA users who type with that stupid little pen! Nobody said your spreadsheet has to compete with Excel 98. If visicalc ran on my Commodore 64, (64k memory & 6502 processor was quite sufficient thank you) then I'm pretty sure it'll work on my 92plus. I see other college students with palmpilots; I wonder if they carry around a graphing calculator too.

Assembler is necessary to realise the future of the TI calculators, which is way bigger than Super Mario Brothers.

I have been looking for a good telnet program/ppp emulator for the ti92. It would be very useful to be able to hook up to an RJ-45 hub via a custom made cable, use ppp to connect to the net and be able to run programs off a linux machine. That would solve most of the problems of not having a decent spreadsheet or word processor; sure it may all be forced into text mode but it will serve the purpose with virtually no memory restrictions. combine the telnet with a multitasking kernel and you can run almost anything text based and make it multitasking just like a real linux machine or even dos if you have a dos telnet server. A simple telnet program and ppp emulator would make the most sense to solve most of the useful app problems.

You seem to know alot about computers and programing in many perspectives that I'm just learning about and I was wondering if anyone or you might know if it's possable to trade something from the computer to a gameboy by using the old gameshark cable that hooks up to the printer port of the computer and links up to the link port of the gameboy. I know this about gameboys and not calculators, but most people here are good at programing and might hold the key to beating games!
ex. pokemon games where trading is a big part of realy beating the game and not "beating" the game.