>>There are no comments on this article yet. Why don't you add one?

OK, I'll add one.

The website looks like a more technically-oriented 1996-vintage ticalc.org, like a mix of ticalc.org and TICT. I hope it's helpful to people learning to use their calculators without having to sift through comments by experienced posters on the same boards. The site looks well-organized, at least initially. Let's see how it functions over time with the community and staff.

TechniCalc is certainly an interesting idea with a promising future. I look forward to the site's development.

Yes, we definitally need a site for people to begin to learn how to use their new calculators....
Forums are supposed to be the place to goto last if you want to learn stuff, not first. Thats what FAQ's are for.

>>Forums are supposed to be the place to goto last if you want to learn stuff, not first. Thats what FAQ's are for.

Yes, I agree. However, for example, I am currently trying to make a program to display the MODE dialog, which is in the archives but will not run on VTI. The calculator manual isn't going to say anything, the SDK guide doesn't say anything, tiams.h only gives a _rom_call related to it in some way, and the tigcc help only goes as far as it can. This is a problem that cannot be solved by reading a manual. I'm not trying to get help here, obviously, because this isn't the correct forum, but I think you see what I mean. I define this as part of 'learning touse your calculator.'

There's surprisingly a lot that's undocumented. Worse off is that if you email TI they won't even give you a response. At least that's what happened to me..

<<Worse off is that if you email TI they won't even give you a response. At least that's what happened to me..>>

Odd. I've always heard good things about ti-cares response times (although possibly not the responses themselves ;)).

What was this about, specifically?

All I remember is that I was using OO_GetAttr and something with an undefined data type.

>I define this as part of 'learning touse your
>calculator.'

Realise though that the vast majority of calculator users are students who just want to get through high school math and play some assembly games. On this level, the manual really is everything they should need, in fact it's more than they should need. Of course most people won't or can't muddle through the whole thing, so it does fall to FAQs and forums.

As for assembly/C programming, since it's not TI's primary focus by far, it really doesn't surprise me that lots of stuff in that area is undocumented.

>>Realise though that the vast majority of calculator users are students who just want to get through high school math and play some assembly games. On this level, the manual really is everything they should need, in fact it's more than they should need. Of course most people won't or can't muddle through the whole thing, so it does fall to FAQs and forums.

This is why there needs to be a concerted effort to make the TI calculator interface a lot better, eliminating functions and making a set of promts in an application. Most people want to use their calculator, not to learn how to use it, and most people thus don't use all of it.

True enough, but there are two problems with that.

One, the teachers won't let TI change the interface, button layout, or menu layout, because they don't want to have to relearn anything. Most of the students only learn how do to things because the teachers and textbooks give button-by-button examples.

Two, there's no such thing as an OS or interface that's so simple that anybody can use it without instruction. The old saying is, after all, "Those who try to build idiot-proof systems always underestimate the persistence and ingenuity of idiots." I think part of the problem there is that lots of people are just skittish around computers and electronics and believe from the start that they won't be able to understand any of it, and therefore don't even try.

I guess there's a third thing I've observed too, which is that some people's problems in using the calculator stems from the fact that they don't understand the underlying mathematics of what they are trying to do. There's a general attitude out there with a lot of people that the calculator is magic and will do any homework problem, all you have to do is copy the problem from the page into the calculator and it will deliver an answer.

Oh. Is that why it still won't do my spanish homework? I type in the Spanish, and it just gives me this "Err:Syntax". What am I doing wrong???

The biggest UI task is to allow the user to use the device without worrying about whether you will break the device. It goes from milk jugs (skinny/thick handles) to typing (computers, where corrections are possible, vs. typewriters, where one mistake means starting over). The only barrier, if you know the math (or at least if you can read the problem) to using a basically idiot-proof program like Graphical Unit Circle [link] is confidence. If you accept that it is idiot-proof, reversible, and that whatever you do will not break, harm, or ruin anything, you can use it without instruction and with confidence to its full power. However, the UI has to make a user confident, and a command line doesn't do that. Seriously, try Graphical Unit Circle as an example. It accepts blank inputs, [ESC], letters (as symbolic for compatibility reasons), and, I think, almost anything you could throw at it. However, a good UI also needs to include everything and not allow the TIOS to interfere to build confidence on the part of the user.

I really think that everything else falls into place after you have confidence in a device.

>The biggest UI task is to allow the user to use the
>device without worrying about whether you will break
>the device.

That's more the user's problem than the UI's. There's not really a way to break the TI calculators from within the TIOS, but the user has no way of knowing that. What does happen is people get a syntax error and throw up their hands. The issue is that the user must speak the machine's language, because there's no way to make the machine so it can read the minds of each different user.

>I really think that everything else falls into place
>after you have confidence in a device.

Call me old fashioned, but I see it as confidence in myself that I know how to use the device, and that comes from reading the manual. Confidence is no replacement for knowledge. That and 25 cents, as they say, will get you a cup of coffee.

It's the device's job to be user friendly, and part of being user-friendly is that the device doesn't scare off the user. What is appropriate for industrial applications (UNIX & Co. command line OSs) is usually counterproductive in the consumer market. For example, grandmothers don't usually respond well to

ERROR:562 SYNTAX LINE:_1504 ILLEGAL _ROM_CALL DEC
VOID: OK PRESS ENTER TO CONT OR HOLD TO STOP

You can't expect them to. Only a programmer understands that (and many programmers won't). As you say, to understand it, you need to speak the device's language. However, once that error is thrown, our grandma-a nonspeaker-thinks (these are all examples I've come across):
1) Somehow, it's my fault, so I shouldn't put my hands on it because I don't understand it.
2) I don't know this. I'm inadequate. Woe is me.
3) Computers are bad.
4) Whoever told the computer to do this right now is young enough to be my grandkid. They're doing this to rebel against my authority.
5) I give up.

The point is to eliminate the language that only the computer only speaks. Our function should be to turn unintelligible gibberish into something usable, and something like the error above, while sounding user-friendly to a person who is used to seeing things like F38FNV9AER, actually drives people from computers and promotes the most dangerous thing in society today, antiscience. It doesn't need to read the user's mind. It simply needs to stop using jargon, commands, and code.

All true, though I have observed certain things:

For some people, they are scared of the technology before even using it. No amount of user friendliness can change that.

Even error messages that aren't so cryptic cause some people to run to a forum and ask "What does this mean?"

>>It simply needs to stop using jargon, commands, and code.

An odd statement. Without commands and code, what is left? Clearly a complete paradigm shift is needed for what you are describing to occur.

Yes. A complete paradigm shift. The user does not want to know what is going on inside a computer. Telling them what is happening is counterproductive. The programmer must simply solve the problem in the best way they see fit because the user sure isn't going to.

This obviously excludes system utilities which programmers use. I was simply talking about application programs.

I know what you mean about the obvious error messages...a message box saying "page not found!" will come up in Netscape sometimes when my mom types a nonexistent URL, and she'll ask what it means. This has happened before in many different ways, like in installer programs which clearly say "Press the Next button to install [application name]", she'll again ask me what to do. It gets REALLY annoying...

>>This has happened before in many different ways, like in installer programs which clearly say "Press the Next button to install [application name]", she'll again ask me what to do. It gets REALLY annoying...

So why didn't it just start while displaying the string?

Do you see what I mean about the computer not showing its guts? It should only tell people what they need to know to get done what they're doing.

I agree with you in most ways about the interface issues, but it is nice not having the install program immediately install when you run it. Many times, people (including me) download install programs and run them later. It's usually hard to tell from the file name just what you're installing, so it's helpful to see the name of the program and make sure it's the one you want. Also, what if you accidentally double-clicked on the installer?

What it should do is tell you what it is going to do, ask whether to do it, then do it. It shouldn't have multiple redundant steps.

However, ordinary users messing around with installers shouldn't be necessary. The OS should be user-oriented, not object-oriented, code-oriented, procedural, or whatever.

That's an impossible expectation though. Not every ordinary user has a geek in the house they can go to for getting a new program installed. You must allow for some interaction between installers and ordinary users, because the programs won't install themselves without being told to do it.

If it's an intuitive UI, you don't need someone to explain it to you (most of the time). My point was that the installer shouldn't be there. The computer should not operate like that. Remember, in 1994, when the internet was first starting up as a commercial medium, and when rumblings about Windows 95 were starting, that Microsoft could have taken the UI wherever it wanted to. It chose one route. It turns out that that route wasn't nearly as intuitive as most people thought. Strike One. From what I gather, Longhorn is going to be Strike Two (if Linux hasn't taken that place of honor, and it probably has). We simply cannot afford another five or six years of mess-ups before we start listening to the general public (Strike Three).

>> My point was that the installer shouldn't be there. <<

How would software get on the PC then? This isn't a Playstation where you put the disc in, play the game, then take the disc out when you're done.

It would be installed by the OS, with 'features' (the programs) selected by the user without the user knowing the technical or semi-technical details. 'Installer' should be 'wizard.'

That's exactly how most instllers already work. Most of the time all you have to do is keep hitting OK or Next, because with most of the choices you want the default selection anyway. Wizard is just a buzzword for installer, and they are still plenty confusing to people who go in expecting to be confused.

But now you're contradicting yourself, because above you complained that users shouldn't have to be bothered with hitting OK or Next.

( Do you see what I mean about the computer not showing its guts? It should only tell people what they need to know to get done what they're doing. )

You can't have it both ways, you know. Leaving in the options confuses the people who are reluctant with computers, and taking them out screws over anybody who needs to vary from the default install.

My point is that if the person needs to install something, they should be able to install it without having to worry about options that have nothing to do with their objectives. As astounding as it may seem, "Click OK to install" makes no sense to a lot of people.

But these devices are not meant for grandmothers. They are meant for people who know at least some basic math. If they can learn math, surely it should be easy to learn how to operate a calculator, once they skim through the manual?

I've never seen the error you mention, or anything similar to it. When I get a syntax error, it's usually a missing parenthesis or something, in which case I get an error message like "Missing )".

I know what he's getting at though. It's like getting a DIM error when trying to graph something. It is descriptive of the problem, but only if you already know what the problem might be. Thus most people have to run to a form and go "What the heck is THIS?"

Of course, the people we were talking about before are the type of people whose VCRs flash 12:00 because they don't know how to set the clock. The Grandmother is the cliché example, but it spans all ages. It reaches the point where they can't be helped unless they can watch somebody do it for them, because our appliances can't work on plain English voice commands like the ship's computer on Star Trek.

There are two questions that I see often enough that I hold up as evidence that it's the people who are the problem moreso than the calculator. First, the people who can't type a fraction because the TI-83 doesn't have a "a b/c" button like some of the scientific calculators. Second, the people who don't know how to type the comma. Sometimes it seems like a cliché "What's the number for 911?" type joke, but these things really do get asked.

My great-aunt was a prolific comptometer operator. She knew a lot of math, as well as how to operate a comptometer (a fairly arcane device). However, she was never exposed to computers, and would have been completely lost in a DOS prompt.

Sure. You know the saying, we were all newbies once. I don't think people should expect to be able to dive right in and do all sorts of stuff with no instruction, and I don't think computers should be expected to allow people to dive right in and do all sorts of stuff with no instruction.

What should be expected is that the people are willing to learn, willing to read that manual or that FAQ or that "... For Dummies" book. Certainly some of us will pick it up more quickly, or with a manual that isn't quite "plain English", but that's life. Everything can't be as simple to use as a toaster oven.

>>Everything can't be as simple to use as a toaster oven.

What the toaster oven could easily do and does not do:
-Ask you to set the optimum temperature for specific items
-Ask you to look up the temperature/voltage/current relationship on a conversion chart and dial that in on a variable transformer
-Ask you to type in the 'food code' or something-like 753579297 for rye bread, etc.
-Require the adjustment of the trays to the optimum height above the elements
-Ask you to remove elements
-Ask you to specifically open vents or circulators
-[nightmare] Be 'incompatible' with certain foods
-Have numerical error codes

The toaster oven could be a lot more complicated, but it isn't. The fact that it isn't is a testament to the people who made it a three-step process when it could easily have been an eight- or nine-step process.

It's much easier to make a simple thing complicated than vice-versa.

Actually, you help make my point. A simple toaster oven only does one thing. A complicated toaster oven would be more versatile. A computer is expected to do more than just one thing. A four function calculator is simple. It does four things. A graphing calculator is expected to do much more than four things.

There's a definate trade off between ease of use and the ability to do more things with just one gadget.

A toaster oven is one of the most versatile things there is. My point was that it absorbs a lot of differences in foods (liquid vs. solid, frozen vs. fresh, specific heats, combined foods, etc.) and still performs its task. It's almost the perfect UI: does a lot, doesn't expect user programming, universal interface for vastly different items, simple, and quickly understandable. You could take one out of the box and start using it immediately.

A toaster oven is not versatile. It only does one thing. It heats up. Any food you may put inside is irrelevant. The toaster oven doesn't know or care what food is inside or even if there is any.

(Here reveals the vast differences in how the two of us see things.)

I agree. Even though the toaster doesn't ask you to do things like adjust the tray height or set the optimal temperature, IT DOESN'T DO THOSE THINGS BY ITSELF, EITHER.

The toaster oven meets my definition of 'versatile' by doing many subsets of one task.

There is a huge difference between reheating a sandwich and heating a science project. Yet it does both.

It's a lot more complicated to do once you start trying to do it.

>> There is a huge difference between reheating a sandwich and heating a science project. <<

I fail to see why. It's dependant on the user to decide what he want to heat in the toaster. The fellow who designed it does not plan for the specific uses of sandwiches or science projects.

Or let's put it this way. You couldn't make a gizmo that will heat a sandwich but won't heat a science project if you wanted to. Therefore the versatiliy you see is not something inherent in the toaster, but in the imagination of the user.

...and oddly the point I was originally trying to make with the "toaster" argument is long gone.

It's very easy to make the toaster oven not heat a science project or a sandwich. It could not have vents, or have the vents be the wrong size, or the heat flow analysis be based on the erroneous assumption that cp=1.

If you think that the point we were discussing earlier was not resolved, please bring it up again. I thought that I made the point that the toaster oven UI was designed properly-user oriented and simple-and the computer's-expecting paradigm shifts-wasn't.

You did make that point, and I agree with it. My point, however, is that your point is irrelevant because the toaster oven does one or a small number of tasks, and is thus inherently simple, while the computer does a large number of dissimilar tasks, and is thus inherently complex. Any complex task, even those predating electricity, is a paradigm shift for those who are completely unfamiliar with it. Yet only now, with computers, is the task expected to adapt to the person who is unwilling to learn new things. I find that to be an unreasonable expectation made by lazy people.

Users aren't lazy. They have jobs to do, and simply don't have time to learn to use a complex piece of equipment, especially if the person selling them the complex piece of equipment says that it will interface smoothly with their jobs. They then turn it on, expecting it to smoothly interface with their jobs, and it displays a string of numbers, introducing them to new vocabulary, and expecting them to know how to use the device without telling them how, step-by-step in each task, when and if the user gets lost.

Let's say that this person is a teacher. They want their internet, their word processor, and their grading program. They don't want a taskbar, a desktop, a few icons, a bunch of switches, boards, sockets, cards, software, installers, or any of the other buzzwords and jargon floating around. Computers have to come to them. They have enough on their hands as it is and will not change what they do unless the new way is better.

My comparison of the VTI V200 with the card catalog fits here, as well. They told me that they needed to digitize their card catalog. So, I copied the interface of the card catalog into a program, asked them whether they liked it. Eventually they approved it, after which I began writing the program.

This does apply to mass-market software. If this was mass-market, the result would have been more copies of the same user-oriented and familiar interface.

Never before in the past has someone wanting to do something new been able to do so without first learning how to do it. I don't expect that to change.

Just because you say there can be a computer that people don't have to learn doesn't make it so, anymore than if it were 1930 and you were talking about a car you didn't have to learn how to drive, or if it were 1800 and you talking about a horse you didn't have to learn how to ride.

Your card catalog comparison only fits half the argument. Yes, you can make a computerized card catalog that works like the paper version. What if, five years from now, a grade school student using the library for the first time accuses you of making the computerized card catalog system confusing and impossible to use? Now, the blame would be misplaced. It has nothing to do with you, it's really the card catalog system, the one you were emulating when you designed the system, that is confusing the student. But he doesn't know that. To him, you are the one who has made his life difficult. Why doesn't the computer just look up the book I want, he will ask. I shouldn't have to deal with search term boxes or Dewy Decimals, he'll say. All this even though the computerized system is actually far simpler and more convenient to use than the paper version it replaced.

It isn't for a school. You're assuming I'm in high school, aren't you? :)

Anyway, my point is that the current system that they have in place should be compatible with the computer. If the user, after a time, believes the interface is deficient, they then should specify what they want changed. I then would do it. Like I said, a car only looks completely different from a horse because it has evolved by customer demand. That evolution is caused by manufacturers either (a) being perceptive to popular demand or (b) getting extremely lucky. Our job is to do (a).

On another note, it's not simpler or more convenient if the user doesn't like it. Our opinion on whether it's simple or convenient (and I agree that it is) doesn't count in the free market. I would like to remind you of the literary term 'writer-based prose' in this context.

I am not trying to say that people shouldn't learn how to use a computer. The computer should do what they need in the way they have done it before. In the extreme and with the best software, the person would have learned to use a computer when they learned to do what they currently do (which after all is what they want to learn).

>> It isn't for a school. You're assuming I'm in high school, aren't you? <<
Or a public library, or a university library. Most libraries are used by students at one time or another. I just said student instead of "generic computer-newbie user".

>> The computer should do what they need in the way they have done it before. <<

But still, many people use computers or graphing calculators do to things they have never done before. Your excuse doesn't hold up then. I'm sure many students and other people have been confused by card catalog systems in the centuries since its development. Did anybody, in all that time, fix it so that it is easier for the novice to use? No, because given what a card catalog is meant to do, it was already using the best system possible. The burden is on the user to gain some minimum set of knowledge about the system in order to use it. Saying "the system should be easier to use" didn't get them anywhere 50 or 100 years ago, and I don't see why that should suddenly be different today.

>>But still, many people use computers or graphing calculators do to things they have never done before.

Like what?

>>I'm sure many students and other people have been confused by card catalog systems in the centuries since its development.

They should learn how to perform the task, not to use a computer. The computer program should be tailored to the task instead of reinventing the wheel. That's all I've tried to say.

>> Like what? <<

Like math. Ask any high school student.

>> They should learn how to perform the task, not to use a computer. The computer program should be tailored to the task instead of reinventing the wheel. That's all I've tried to say. <<

I agree that they need to learn the task, but now the task is linked to the computer. You have to learn both simultaneously.

That's not what I meant. I meant that hasn't been done before so that there is no prior knowledge by anyone as to how to do it.

You seem to not be getting my point that the computer should not exist as separate from the task. The user should be able to learn the task and then do it on the computer.

I think I am indeed missing your point. Or we're missing each other's. Learn the task and then do it? How is that different from how people do things right now?

>>How is that different from how people do things right now?

It basically isn't. What we need to do as a profession is to not go down the tempting slippery slope of making people learn how to use a computer before they can do their job. They learned to do their job. As a profession, we need to tailor the software to the job, not the other way around.

I agree with that in principle. But these days, using a computer IS the job, or at least an integral part of it. And if it's a job that doesn't actually require the use of a computer, then fine. They can do it without the computer. It's their choice to not learn the computer. They just don't get the benefits of the computer either. There's no free lunch.

And on the high school/graphing calculator side, the argument doesn't hold. We don't impose graphing calculators on students who have already learned the math. They are there TO learn the math, and that much is actually made easier if they simultaneously learn to use the calculator effectively.

For people whose job is IT, then their job is IT. However, most people have a job outside of that and they shouldn't be forced to learn a whole new profession to do what they already do. It is also, in a professional context, usually not possible to simply not use a computer. For example, the Chicago Public Schools has almost entirely signed on to the use of a real-time grade reporting system called MTG/EDLINE. Unfortunately, the system is unusable. Teachers-and also lower-level administration-have no option to not use it, so they have to modify their reporting systems to meet this program. It is ridiculous, obviously, but everyone has to use it. My problem is with the structure of the program. As a programmer, I know that it could have been written much better (I've seen it, BTW-they use custom Open and Save dialogs to show how 'good' they are at programming but they don't program the system with the user in mind).

As for your second statement, I just keep saying that the calculator should be designed around the user so that they don't have to actively learn to use it. It's not easy-which is why it's called software engineering-but it can be done.

People are not in love with technology for technology's sake. I know, as a person interested in computers, it is hard to come to terms with. However, people need computers to get something done, and are completely uninterested in how it gets done. They want as little interference by the computer as possible, and, as surprising as it seems, are unwilling to read manuals. They have enough to worry about (i.e., their jobs) without having to figure out how to use industrial process equipment, which is the general perception, even today, of computers.

I agree that this is the attitude of people, but it mystifies me as to why. Why do people balk at reading a graphing calculator manual but readily accept the need to take driver's ed to learn how to operate an automobile?

Why should I care what the proper command syntax is when I just want to do my calculus homework? And why should I care about the difference between the gas pedal and the brake pedal, or which road goes in which direction, when I just want my car to take me to the mall?

I wouldn't go to France and expect the people will understand my English, I must learn their language if I am to communicate. Nor do I expect the computer to know my language, I must learn how it expects to be talked to.

The kids today need to just stop whining. Oh, is the graphing calculator too hard for you? Maybe the scientific calculator is too? How would you like to use a slide rule and tables like your grandparents? Hey, guess what? You can't use a slide rule without being taught how either. I myself couldn't pick up a slide rule right now and actually do any math with it.

Doing something you haven't done before involves LEARNING, whether it involves electronics or not. There are no shortcuts.

>>The kids today need to just stop whining.

Unfortunately, they're the market. Successful enterprises (at anything-sales, info, etc.) don't say "Hey, Customer! You're dumb!"

The reason people learn something is motivation. Motivation cannot be drilled into people. It is an internal value, and nobody on the outside can change that. You can only attempt to make your product something that a person wants to learn to use. You don't do that by saying "Hey, Customer! You're dumb!" or worse, "Learn this or else!" but by relating it to what they're doing. To use your example,

>>Why should I care what the proper command syntax is when I just want to do my calculus homework?

You shouldn't. It should be a prompt. What does nInt(expression,var,low,up) have to do with calculus? Nothing. It is a computer command. The computer should provide a framework in which the user can get a result they can understand. It doesn't need to pander. It just needs to "speak math" to peoople using a cacluating program, to "speak publishing" to a person using a DTP app, etc.

>The reason people learn something is motivation.

True enough. But people's expectations are contradictory. Perhaps I want to enjoy swimming without all the bother of learning how to swim.

Really it's a product of having the calculators. Back when calculus had to be done by hand, any student who couldn't do it had no one to blame but themselves. But now we have these graphing calculators, and people think "I don't understand how this works, so it must be able to do anything I want it to do." and think it's the calculator's fault when they don't know how to do their homework.

>You shouldn't. It should be a prompt.

Sure. And then all the internet forums would be filled with "What do I type at this prompt?" and "How do I get to the prompt?" That's hardly any different than the existing nInt() command.

Answering a question is different from asking one.

>>And why should I care about the difference between the gas pedal and the brake pedal, or which road goes in which direction, when I just want my car to take me to the mall?

What do the brake and gas pedals have to do with getting to the mall? Nothing, by themselves. However, they speak your language. Press on one, it goes. Press on the other, it stops. Simple. Intuitive. Example #1 of a car that doesn't work like this: a BMW 740i (I think). It has a knob where the transmission usually is that allows you to navigate the stereo system, GPS, pedal height, etc. Who expects you to learn that? 99% of people who drive it start playing music while trying to shift into gear. Why? It's counterintuitive. It's a non-ubiquitous computer interfering with how we normally do things. In conclusion, anything that any innovator plans to do cannot reinvent anything that anybody currently uses. It needs to build on it. Computers don't do that. They try to force new paradigms on people-pointing, clicking, menus, toolbars, windows, login, logoff, prompts, functions, programs, code.

In an ironic twist, you inadvertently give another example. Why did everyone use a slide rule? They had comptometers, after all. Comptometers were complicated! They tried to force a paradigm shift on people who had no time to shift paradigms. If you look back on the test for certified comptometer operators, it is very long and deals with very arcane subjects. The average person had no time for this, and it fell out of use. Many used upwards of 50 buttons to do what a simple four-function calculator that you can place in your smallest pocket does today.

>> What do the brake and gas pedals have to do with getting to the mall? Nothing, by themselves. <<

And what does nInt() have to do with calculus? Nothing, by itself.

>> Press on one, it goes. Press on the other, it stops. <<

Learning which pedal performs which action is no different than learning the syntax for using nInt().

>> It's counterintuitive. <<

Intuitiveness is in the eye of the beholder. Even monkies need to be taught how to use a stick to get ants out of an anthill or how to use a rock to crack open a nut, something that one might expect to be intuitive to any random fellow pulled off the street.

A person who is already comfortable with computers will naturally apply such knowledge to unfamilar software and probably figure out how to make it work. A person familiar with mathematics but unfamiliar with computers may even be able to use your theoretical intuitive calculus software to get the correct answers. But the students we are dealing with aren't familiar with the calculator OR the math that they are using it to perform. That's a double whammy and the ONLY WAY they can expect to get anything done is if they are actually WILLING to learn these things. The problem with that is of course that most of them couldn't care less about math because they don't plan on being in a career that requires it.

>> They try to force new paradigms on people-pointing, clicking, menus, toolbars, windows, login, logoff, prompts, functions, programs, code. <<

Sure. And I'll learn the new paradigm and use word processors and calculators while they stick with their typewriters and slide rules. It's a trade off. You learn new things, you can perform your tasks faster and easier. You don't learn new things, you do it with the same old familiar inferior methods.

Let's take these disagreements one at a time

I think you would agree that a computer instead of the old mechanical systems or electronic replacements thereof is a paradigm shift. I see people coming into unfamiliar institutions and forcing new paradigms on people all the time. The result is never peole using the new system. The result, 100% of the time, is that at least 75-80% of the people involved hate the person and permanently associate technology with intruders. Non-coincidentally, that is the same proportion that is composed of non-computer-nerds in the general public. If you're doing something, the point is not to preach to the choir and leave the other 75-80% who don't know what to do immediately and innately hung out to dry. The point is to provide a smooth transition from the old to the new, no growing pains, as few paradigm shifts as possible, with as manysimilarities, even if only superficial, between the old and the new.

What you say is true, however, that attitude serves only to eliminate innovation. Following your logic, we should be operating our automobiles with reigns because the steering wheel, gear shift, and pedals were too complicated for people used to the horse and carriage. Guns would never have caught on, loading and firing a musket it more involved than shooting an arrow from a bow. Etc.

We're drifting from the central issue of graphing calculators anyway. It's one thing to go to somebody who is used to method #1 and making them switch to method #2. But these students, they've never done this math at all before. It would be a new paradigm for them whether it's the 1990s with graphing calculators or the 1950s with slide rules.

The central issue is that they are in math class to learn something. At one time it was to learn how math worked. Now it's more about learning how to operate the calculator. Whether that change is a good or bad thing is another great debate. But regardless they must learn something that is completely unfamiliar in order to do the prescribed math problems. And of course, they don't want to learn anything if they can avoid it. They expect the graphing calculator to be exceedingly simple, which would allow them to avoid learning anything. Naturally they are disappointed. A theoretical calculator that you desire that would indeed allow them to do all this math without learning anything would be quite the new paradigm, it would eliminate the need for math class altogether.

Early cars were, indeed, operated with reins until sticks, and eventually steering wheels, caught on because of customer demand. Guns were based on crossbows, not long bows. This is exactly my point. Programmers simply need to make the calculator as inobtrusive as possible, with the first step not being to force the existing technology on people but to seamlessly integrate calculators with existing systems. Example: a current consulting job that I am assisting with involves the exact duplication of a card catalog system into a VTI Voyage 200. The goal is to have a seamless reproduction of their card catalog. I don't see why setting this objective is so difficult. My point is that most people would try to 'improve the interface.' When the customer wants the interface improved, do it for them. Just don't force it on them.

Okay, let's talk about existing systems. Before graphing calculators, people did math using scientific calculators, pencil, paper, and their heads. The calculator didn't even have the ability to do a definate integral in one fell swoop, the students had to know how to set up the problem if they were to use the calculator to crunch the numbers.

Today, in fact, they could do the exact same thing. That would be a direct flow from the preexisting system as you describe.

However, it is also easier now, because there is that integration command there. Instead of setting up the entire calculation beforehand on paper, you only need to know how to lay out the command with the starting bits of the problem. You cannot convince me this results in the process being more complicated or more difficult. You don't even have to know how to do it the old way. You just have to copy it from the textbook into the calculator in a certain order.

Of course, what I have just described is irrelevant anyway. The thing about students twenty years ago, they KNEW that they were going to have to use scientific calculators, pencil, paper, and their heads. The students today, they think the graphing calculator is the only thing they need. They think pencil, paper, and their heads are not required. This is not an instance of the technolgy paradigm shifting away from the user. In fact, it is the users who expect a very large paradigm shift when one has not occured. They underestimate the ability of the calculator, thinking it will do all the work for them.

You revere the old methods of doing things, well, the students could easily use any old method of doing their homework instead of the new methods. But wait, no they can't, because they refused to learn those methods thinking that the calculator would take care of everything for them with no intervention from them required.

The calculator needs to make clear that if you don't know the math, you aren't going to be able to tell the calculator what to do (uh, what's a mantissa? whaddaya mean a mantissa? why da calc askin' me fer da mantissa?). However, it also needs to bemore user-oriented than a command line. For you, it is much easier, as you know how to use the calculator as well as the math. However, if a person knows only the math, it is much harder for them. I have seen people take a TI-30Xa into a calculus test because they believed that they couldn't do any better with a TI-89. The "certain order" you mention is the hangup for a lot of people.
Somebody could make a calculator that is just like a pencil and paper, or that uses a prompt system, or a command line made intelligently (hard). However, whatever the interface is, it needs to be in response to extensive market research as to what people want in an interface, can and can't do with a certain interface, and whether they're comfortable with an interface. Current systems simply do not do that.

Ever use an equation writer program, like the HP 49 has, or one of the third-party versions for TI's 68k calculators? I haven't either, but they do away with command line entry in favor of a more "pretty print" style entry, to borrow TI's term for the 68k output method. However, I'm certain that even with an equation writer removing the impediment of the entry line, there's still plenty to screw up a user who hasn't made at least a small effort to familiarize himself with the system before trying to use it to pass a test or something.

>>The calculator needs to make clear that if you don't know the math, you aren't going to be able to tell the calculator what to do<<

That's laughable, and statements like that just make you lose credibility in my eyes. The calculator must be responsible for telling the user common sense things? My problem with users is that many of them think that given an 89 they can do anything math oriented with little effort or training. Apply that standard to anything else and nobody would even consider it. Does buying a tool set from the hardware store make me a master carpenter?

Or on a similar note, some of them are just plain lazy. They ask how to do something like linear interpolation. I'm certain that in most cases they are staring at a textbook page that gives the equation. They would only need to use the standard old four functions and the number keypad, and they'd still get the answer quickly. But no, they rush to a forum to ask how to make the calculator do it automagically, and have to wait half a day for an answer.

>>Does buying a tool set from the hardware store make me a master carpenter?

No. That's my point. You learn carpentry first and then buy tools. Most often you don't know what you're looking at until you learn carpentry, at which point you can use the tools without cutting off important body parts. The computer/calculator should have a code block like this:

:Dialog
:Title "Logarithms"
:Request "Log",qwer,0
:Request "Base",tyui,0
:EndDlog
:(C) tyui defaults to e

instead of allowing log(qwer) or ln(qwer) at a command line. This more accurately reflects knowledge of math instead of computer languages.

The TI-89 allows users (and indeed is subtly marketed as such) to think that they can do anything with it without knowing any math.
If you know TI-BASIC in depth, I can guarantee you that you can ace most trig finals without knowing any of the material. I don't want a debate about whether this is good or not (as it's off topic) but it is true, unfortunately. Some people, like you say, are just plain lazy, and the calculator shouldn't help them. It should be an aid, not a crutch. I think that we agree on this, and that our main disagreement is on whether calculators can be transformed as such. I think they can; in fact, I work for a developer working on such a system called Morvlon 3.00. We'll prove it to you.

>> No. That's my point. You learn carpentry first and then buy tools. <<
>> The TI-89 allows users (and indeed is subtly marketed as such) to think that they can do anything with it without knowing any math. <<

So you do agree with me on that. Good.

>> The computer/calculator should have a code block like this: <<

That strikes me as somewhat cumbersome. More importantly, the command line version is closer to being a direct descendant of the way it was done on one-line scientific calculators. I guess that's the dilemma, keep it close to what it's replacing for the sake of existing users or redesign it completely for the sake of new users now that the technology is somewhat better able to do that? Obviously the former is what the industry has gone with. From what I've read here, you seem to want it both ways, even though the choices are mutually exclusive.

And once you go down that path, where does it end? It can go to any level of absurdity:

:Dialog
:Title "Addition"
:Request "First number",a,0
:Request "Second number",b,0
:EndDlog
:Disp expr(a)+expr(b)

This is something I worry about when writing FAQs and such. At what point in simplifying things for the dumbest people do you start insulting the smarter ones? I think I do err on the side of assuming a certain level of knowledge in the reader, and that's my failing, but I'm certainly willing to go into more depth for the people who can at least point at the parts they don't understand. The problem is, most of them do a throw-up-the-hands sort of thing and just say they don't understand any of it. That doesn't help me or them.