Digital Joysticks provide better control than Analog Joysticks

[atariksi]

Back to the 5 points:

 

(1) eliminating regions of uncertainty in DK and other games for analog joysticks

(2) reducing throw to less than or equal to digital joysticks for analog joysticks

(3) feedback reliant is inferior to knowing states a priori

(4) analog interface itself produces inexact states even after the joysticks are read compared to digital states

(5) analog joysticks have inferior sampling rates (speaking of joysticks in post #1)

 

For #1, it's obvious in games like DK when you jump over barrels, jump from one ledge to another especially in the pie screen, etc. in games where you wiggle between tight spaces (no trigger required here), etc.

[ApolloBoy]

You should be banned for life from every forum in existence according to what you have written and how you behave. I gave truthful remakrs about you like this one.

Jesus Christ, talk about overreacting. So what, is everyone who disagrees with you a troll now?

[kool kitty89]

If you side with a troll, then you can also be labeled as a troll. I am only comparing analog joysticks with digital joysticks. Yes, if there's no digital version of flight controls the current set up would provide the best control available.

 

Thumb sticks was refuted as well earlier. If you expect the user to come to a complete stop without relying on the PCB (going extreme) to try to shorten the throw then that's still inferior to a digital joystick. In a digital joystick, not only do you have the easier control of hitting the extreme but you come to a stop immediately whereas in a thumb stick you have to come to a stop. A car hitting a brick wall traveling distance d compared to the same car coming to a stop before hitting a brick wall. Which one travels distance d faster? Also, the bigger level also helps the digital joystick compared to a thumb stick. A thumb has less force as well compared to using the full hand/wrist.

 

Point #3 you made is clearly false if you think all games require feedback for their controllers in order to figure out what the controller did. The rotary phone example shows how a priori is superior to waiting for feedback.

Hardly ad admission - "if there's no digital version" then the analog flight stick is the best. Find any flight sim on a modern console and play it with a digital 8-way stick then any analog controller and tell me which gives less control. Hitting any extreme on the controller will not make for a steady flight. And fast tapping the controller doesn't count. By your own standards that would require looking at the screen to get feedback. No a priori knowledge there.

 

With regards to the rotary phone. How can you have a priori knowledge about what number will be dialed before looking at the number you want? Unless you memorize every position on the dial and hit it with your eyes closed. And if you are capable of that, it is the same thing as memorizing the position of an analog stick.

 

Try a bit harder next time.

 

In a rotary phone vs. an analog joystick, you don't know the state of the analog joystick in selecting the number before you hit trigger or something to dial that number. In a rotary phone, you spin the dial for the number so no need to wait for some feedback. You are also looking at the analog joystick, but there's no way to tell what state it's in. That's not the feedback, the feedback is after you dial do you know what you dialed a priori? You don't. It's uncertain.

 

If you rely on feedback, you can have flight simulation controls with a digital joystick. They already exist. I think what he was talking about was the REAL flight controls in an airplane which don't have any digital equivalent.

 

@appolloboy: if you side with a criminal in a crime, you are an accomplice. Same logic. Oops, I forget, you don't know logic.

 

On another note, the mechanism used for a rotary phone is much like a 1-bit digital/serial/pulse mechanism... as such you can quite literally "dial" by tapping out the correct number of pulses manually or using a digital device to output 1-bit clicks to the analog audio line. (rotary phones use 1=1, 2=2, etc up to 10=0)

[+Stephen]

You look at both the joystick and the phone. The feedback referred to is to find out what number you dialed. Imageine the analog joystick is hooked up to a phone device, you hold the joystick in some state on an axis and maybe press the trigger to register the number. This is inferior control than using the rotary system since you need feedback for the analog joystick to decide which number you pressed. Don't mix it with other feedbacks-- just for the dialing is what the device was implemented to do.

If that's the case (dialing only), then we will restrict this conversation to paddle controllers, because they operate in the same manner as a phone dial. In this case, you could have the numbers painted on the dial. Same as phone, same control. You said WAY back in this thread that paddle controllers offer better control for games like breakout.

 

Therefore, your comparison of an analog joystick to a rotary dial phone is both null&void, and refuted.

[JamesD]

You should be banned for life from every forum in existence according to what you have written and how you behave. I gave truthful remakrs about you like this one.

Jesus Christ, talk about overreacting. So what, is everyone who disagrees with you a troll now?

You are just figuring that out?

[sloopy]

/me looks for last post before lock...

 

 

sloopy.

[ApolloBoy]

You should be banned for life from every forum in existence according to what you have written and how you behave. I gave truthful remakrs about you like this one.

Jesus Christ, talk about overreacting. So what, is everyone who disagrees with you a troll now?

You are just figuring that out?

Just figured I'd ask him. He'll probably deny it though.

 

@appolloboy: if you side with a criminal in a crime, you are an accomplice. Same logic. Oops, I forget, you don't know logic.

Ooh, that burned. Also, we're not talking about criminal acts so your analogy is flawed.

Edited December 18, 2010 by ApolloBoy

[atariksi]

You look at both the joystick and the phone. The feedback referred to is to find out what number you dialed. Imageine the analog joystick is hooked up to a phone device, you hold the joystick in some state on an axis and maybe press the trigger to register the number. This is inferior control than using the rotary system since you need feedback for the analog joystick to decide which number you pressed. Don't mix it with other feedbacks-- just for the dialing is what the device was implemented to do.

If that's the case (dialing only), then we will restrict this conversation to paddle controllers, because they operate in the same manner as a phone dial. In this case, you could have the numbers painted on the dial. Same as phone, same control. You said WAY back in this thread that paddle controllers offer better control for games like breakout.

 

Therefore, your comparison of an analog joystick to a rotary dial phone is both null&void, and refuted.

 

Think about it some more. I am giving example of item # (3) -- a priori control vs. waiting for feedback.

[potatohead]

A call to be banned for life?

 

From the guy with a spare female persona?

 

From the guy WHO ASKED TO BE REFUTED, WITHOUT QUALIFICATION?

 

Let's talk about that for a moment, shall we? Now, before I go any farther, I'm not pissed at all. Truth is, I'm laughing at some clown calling me a ass. That's what you do when that happens, because it's just not serious in any way.

 

Where was I? Oh yeah, qualifiers! You asked "Let's see who can refute it" Well, here I am callingforbanski! Refuted. Refuting any way I can actually, because YOU ASKED FOR IT!

 

Don't go crying about how tough it gets, WHEN YOU ASKED FOR IT. Life is funny that way. When YOU ASK FOR IT, posting up self-refuted things, panties in a bunch, pissed at other people actually dishing it out to you, PER YOUR REQUEST, going off like that looks just a bit silly, like you've a ego to hide or something!

 

Do you have a big ego to hide, by the way? I can't tell just yet. Help me out there, would you? It would make things easier.

 

Now, the last time we went down this road, you decided to rant to me on PM's, about how bad of a person I was for making you look wrong. This time YOU ASKED FOR IT!

 

The basic reason you really, really want me off this thread is I've brought the kind of info in that makes it difficult to maintain the illusion of authority.

 

THAT IS BY DESIGN, BECAUSE YOU ASKED FOR IT. Remember that. The record is clear.

 

Now the stuff I wrote is all about that end. Last time I cracked you on one of these hold-outs, you got pissed as hell, and just flailed. Refuted, BTW, just like right now, REFUTED.

 

Truth is, I took the data you provided to us, on this thread that YOU ASKED FOR, and simply posted up my conclusions. It's not really my problem that data sort of looks bad. After all, I didn't generate it, did I?

 

Now, you've got the FIVE POINTS, all based on a SELF REFUTED initial point of discussion that lacks the qualification necessary for rational consensus. Again, I didn't do that, nobody else here did, and none of us ASKED FOR IT either. You did.

 

Given that mess, is there any wonder you are pissed off right now? Hell, I would be, fucking it up that badly by my own hand. I feel for you. Really, that's just gotta suck.

 

So the FIVE POINTS DON'T MATTER, because they are in support of a primary one that is self-refuted, because it's form is wrong, not allowing for solid discussion, or proof via scientific method. Subjective. Until you bring NEW DATA to the table, that can qualify that initial point, you are on losing ground.

 

Now, as for locking the thread! I'll be up, front and completely honest, like I was earlier.

 

You have no good end game right now. The stuff you posted up, THAT YOU ASKED FOR, is refuted, solid. You don't have the hearts and minds, and that's everything on a subjective matter of opinion, and so we play a game.

 

The game is, either we get pissed and leave, or break protocol, leaving you able to claim the win, just by attrition, or foul, or we stay. If we stay, the game is ongoing, tiresome, really only worth it for the entertainment value, and that's only true when it's healthy.

 

We have kind of left healthy here as of late, and that tends to lower the entertainment value, and puts the thing at risk of giving you the out of claiming some mess, being able to establish some authority on the matter throughout.

 

Not gonna happen. And again, to be really, really clear, YOU ASKED FOR IT, and your buddy gonnabringatariskitojesushead is gonna deliver proper, largely because I find it worth it to answer this kind of call, because that's how I roll.

 

That leaves the lock. If the lock happens, like now, it won't go well. You, Atariski, do not have the higher ground right now, and so it makes the very best sense to lock it up solid, REFUTED.

 

Finally, why didn't I leave?

 

Well, YOU DID ASK FOR IT, and frankly, I just can't stand that kind of ego running unchecked. That too is how I roll, and I find it entertaining enough to continue, no personal harm, no foul, BECAUSE YOU ASKED FOR IT.

 

Now, had you not asked for it, I would have been outta here long ago, but sadly for you, YOU DID ASK FOR IT.

 

REFUTED

 

I suggest you reconsider offering up challenges of this kind, if you are not well prepared to deal with the product of them.

 

***It's worth pointing out that suddenly, we've got single +1 marks on a lot of the dynamic duo posts! Hmmm, could that be circling the wagons, triggered by some comments about not having the hearts and minds? Maybe so, and that's by no means scientific, just curious, that's all, given the whole illusion of authority bit going on right now.

Edited December 18, 2010 by potatohead

[potatohead]

Yes, absolutely.

 

Nicely done. I wonder if that wouldn't work faster in the Tandy BASIC, just because it's got proper bit-ops? The Atari BASIC doesn't, though it does do the nice comparisons, as I wrote above.

 

Just thought about this a bit, and wouldn't you need to shift one or the other, so that the values are clear?? Seems to me, just adding them would leave some confusion, or am I missing something?

 

 

 

 

 

 

Nice visual program way to illustrate analog! I see a controller map! Cool!

 

(prolly, comparisons are faster, but that's nicely arranged)

The advantage of this approach is that you can easily configure the throw length and center to suit the player/joystick.

Something that is rarely done on any 8 bit let alone the 5200.

It could make all the difference when it comes to response time on a twitch game.

 

If I were to use this approach I'd actually use a table for each axis and use lower bits for one axis and higher bits for the other. You bitwise or them or add them.

30 J=A(JOYSTICK(0)) + B(JOYSTICK(1))

Then you have one value that represents both axises.

Edited December 18, 2010 by potatohead

[potatohead]

On another note, the mechanism used for a rotary phone is much like a 1-bit digital/serial/pulse mechanism... as such you can quite literally "dial" by tapping out the correct number of pulses manually or using a digital device to output 1-bit clicks to the analog audio line. (rotary phones use 1=1, 2=2, etc up to 10=0)

 

Yep, I had made a phone in High School, that packed into a old APF Paddle controller canister. Mic on one side, speaker on the other, with two alligator clips and a single button. Strip a bit of wire, then dial out with the button. Did it a coupla times when I was grounded from the phone. Good times

 

One can get pretty good at the taps. It was kind of fun actually!

Edited December 18, 2010 by potatohead

[JamesD]

Yes, absolutely.

 

Nicely done. I wonder if that wouldn't work faster in the Tandy BASIC, just because it's got proper bit-ops? The Atari BASIC doesn't, though it does do the nice comparisons, as I wrote above.

 

Just thought about this a bit, and wouldn't you need to shift one or the other, so that the values are clear?? Seems to me, just adding them would leave some confusion, or am I missing something?

Well, there is no need for bit opps in this case, just proper support for arrays. You see, if no bits collide, a bitwise OR can be done with addition.

 

If array A uses 1, 2, and 4 as values, and array B uses 8, 16, 32... then no bits collide during the addition.

Add a 3rd array to give the bit to binary joystick conversion, and on the TRS-80 you could do this:

40 ON (C(A(JOYSTICK(0)) + B(JOYSTICK(1)))) GOTO 100,200,300,400,500,600,700,800,900

 

Each combination of values from the joystick test results in a unique value representing a digital joystick position.

Those unique positions translate to a unique array index which contains the digital joystick position.

Each line number represents a different binary joystick position.

 

The bit value to position conversion values would be:

Upper Left = 1+8 = 9

Left = 1+16 = 17

Up = 2+8 = 10

Center = 2+16 = 18

Upper Right = 4+8 = 12

Right = 4+16 = 20

Lower Left = 1+32 = 33

Down = 2+32 = 34

Lower Right = 4+32 = 36

 

So you need an array size of 36+1 (+1 because arrays start at zero) and any invalid number is zero.

The data for array C is as follows... if I counted right:

1000 DATA 0,0,0,0,0,0,0,0,0,1,2,0,3,0,0,0,0,4
1010 DATA 5,0,6,0,0,0,0,0,0,0,0,0,0,0,0,7,8,0,9

 

The speed of the code depends on how fast your BASIC handles arrays.

<edit>

The third array isn't needed if you have 36 line numbers in the ON GOTO. I'm not sure which method would be faster.

 

With the Atari's ability to goto line numbers in a variable you could put those in the C array.

Edited December 18, 2010 by JamesD

[potatohead]

Ahh! That's what I was missing. Didn't see how you ran it through the array. Thought, for a moment, the joystick outputs were being used directly.

 

Seems to me, for simple 8 direction, comparisons would be quicker, but if zones are needed, the "map" method with the arrays makes more sense.

 

Atari BASIC is cool for comparisons and goto variables, and it's string handling. Not so cool, because it's got no bit ops...

 

Gonna have to go do some twiddling with my CoCo 3. Finally got a joystick for it. It's the older, non centering one though. The deluxe is much better, a lot like the Apple ][ style. Loved those.

 

Anyway, just kind of wondering which is quicker. A little bit of retro fun tomorrow.

[JamesD]

Ahh! That's what I was missing. Didn't see how you ran it through the array. Thought, for a moment, the joystick outputs were being used directly.

 

Seems to me, for simple 8 direction, comparisons would be quicker, but if zones are needed, the "map" method with the arrays makes more sense.

 

Atari BASIC is cool for comparisons and goto variables, and it's string handling. Not so cool, because it's got no bit ops...

 

Gonna have to go do some twiddling with my CoCo 3. Finally got a joystick for it. It's the older, non centering one though. The deluxe is much better, a lot like the Apple ][ style. Loved those.

 

Anyway, just kind of wondering which is quicker. A little bit of retro fun tomorrow.

BTW, I doubt you'll find any old CoCo software that actually uses that. It's something I came up with during the discussion.

 

<edit>

 

I think this is the complete code minus the routines for handling the directions.

ON GOSUB could be used in place of ON GOTO

10 DIM A(64):DIM B(64):DIM C(3):FOR I=0 TO 63:READ A(I):NEXTI:FOR I=0 TO 63:READ B(I):NEXTI:FOR I=0 TO 36:READ C(I):NEXT I
20 ON (C(A(JOYSTICK(0)) + B(JOYSTICK(1)))) GOTO 100,200,300,400,500,600,700,800,900
30 GOTO 20


1000 DATA 1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1
1010 DATA 1,1,1,1,1,1,1,1,1,1,1,1,2,2,2,2
1020 DATA 2,2,2,2,4,4,4,4,4,4,4,4,4,4,4,4
1030 DATA 4,4,4,4,4,4,4,4,4,4,4,4,4,4,4,4

1040 DATA 8,8,8,8,8,8,8,8,8,8,8,8,8,8,8,8
1050 DATA 8,8,8,8,8,8,8,8,8,8,8,8,16,16,16,16
1060 DATA 16,16,16,16,32,32,32,32,32,32,32,32,32,32,32,32
1070 DATA 32,32,32,32,32,32,32,32,32,32,32,32,32,32,32,32

1080 DATA 0,0,0,0,0,0,0,0,0,1,2,0,3,0,0,0,0,4
1090 DATA 5,0,6,0,0,0,0,0,0,0,0,0,0,0,0,7,8,0,9

Edited December 18, 2010 by JamesD

[Divya16]

You should be banned for life from every forum in existence according to what you have written and how you behave. I gave truthful remakrs about you like this one.

Jesus Christ, talk about overreacting. So what, is everyone who disagrees with you a troll now?

You are just figuring that out?

Just figured I'd ask him. He'll probably deny it though.

 

@appolloboy: if you side with a criminal in a crime, you are an accomplice. Same logic. Oops, I forget, you don't know logic.

Ooh, that burned. Also, we're not talking about criminal acts so your analogy is flawed.

 

An analogy doesn't have to be exactly what you are comparing it to. Look up "analogy" and then figure out the logic. The logic applies. If someone says "Appolloboy is mean rotten apple" and someone else agrees or sides with him then that latter person can also be considered to be in the same boat as the former.

 

As for the over-reacting, I don't think so. If he behaves like that, any forum would consider him a negative asset. And it's good for him that he stop since if people look back at the forum and see how the undeniable truth about the digital joysticks' control being superior to analog joysticks and find this wretched person was so opposed and stopped the forum, they would mock him for life.

[Divya16]

I have yet to see one refutation to

By your standards, I could point to posts that do just that and say you are blind and it wouldn't be a personal attack.

 

(1) eliminating regions of uncertainty in DK and other games for analog joysticks

(2) reducing throw to less than or equal to digital joysticks for analog joysticks

(3) feedback reliant is inferior to knowing states a priori

(4) analog interface itself produces inexact states even after the joysticks are read compared to digital states

(5) analog joysticks have inferior sampling rates (speaking of joysticks in post #1)

Those have been addressed but it is a moot point anyway as example games have been provided where using an analog controller is superior to using a digital one and your original hypothesis is incorrect even if any or all of those are true.

Btw, you never proved any are true, you only stated they are true. Since you are the one making those claims the burden of proof is on you. Feel free to post some code demonstrating where inexact states exist.

Just move an analog joystick all the way right and let go, you will see center is inexact (different from what it was). Now move back to the right and you'll see it's not the same value at the extreme. Even if you hold at the extreme, you won't get the same reading. Your points were refuted many times but I'm going to breifly address them one more time since some people forget.

 

1) DK is a game designed for digital joysticks, and what regions of uncertainty? There are no in between or uncertain states. The joystick always returns the position of the joystick and the software does range checking. If the code in DK is flawed that is not the fault of the controller, a fact which you have never attempted to rule out in your "experiment". That entire argument has been covered, you just ignored it.

You just don't understand it. When you jump or move around back and forth, you have thresholds rather than exact values at the extreme so there's uncertainty of the motion.

 

2) As I pointed out with sample code, the throw required by an analog stick is configurable. You don't need to use maximum range of movement when simulating digital control on an analog stick. Throw is also dependent on the joysticks used which we pointed out was a small sample size and since tests only seem to have been conducted on the 5200, I highly doubt you even used all the joysticks pictured. Rather than providing custom code that tests reaction times and control in a controlled experiment where variables can be ruled out, your experiment depends on specific games, specific systems, one test subject, and your entire results are limited to them. If you don't like the maximum throw of an analog stick, you could design a piece of plastic to shrink the area of throw and the software could be modified for a shorter throw. But you made no attempts to rule out controller design rather than analog being the factor.

That would be really dumb thing to do. Now how much do you move in the joystick if you shorten the range from center-- it's even worse. Now you added more uncertainty which is worse than the throw. And you haven't beaten the digital joystick anyways. For any distinguishable state from center, you can have a digital joystick with the same distance that's better-- no uncertainty and help of the PCB for the extremes. Again already refuted many many times.

 

3) But when you are simulating a digital joystick, you slam the joystick just like a digital joystick. You aren't searching for an exact position. And as I pointed out, once in a while when you try to move at an angle with a digital joystick, you will trigger only one axis slightly before the other axis is detected. You are relying on feedback to identify where the corner is with the digital joystick. As such, no priori can be assumed for digital and your argument would apply to both type of sticks.

You already screwed up in understanding the rotary phone analogy so this point is moot. Digital joysticks allow a priori knowledge of the state not so with analog joysticks.

 

4) Inexact... "Not strictly accurate or precise; not exact" The controller returns the position of the stick as resistance values indicating the location of the stick. Where are the inexact states? At the very least you have used the wrong word. The state is not as simple as on or off but certainly isn't inexact. Any inexactness would be in the player's control of the stick, or how the software translates the position to a state. The first is dependent on experience and skill which is learned, and the last is dependent on the software implementation. As such, it is not the interface and your statement is incorrect either way.

That's total rubbish. Analog signals don't provide consistent values even if you had a digital joystick hooked as in the Atari 5200. The values change over time, heat, etc. Again already refuted many times. This is about the interface now not about the joystick.

 

5) The analog joystick itself returns a resistance for each axis that changes as fast as you change the position of the stick. Sampling rate would depend on the joystick interface implementation, not the joystick, and as such it is dependent on what machines you do the tests on. Some systems interface digital joysticks through the ports on an AY sound chip and commands must be sent to the chip before you can read the ports. That is no better than configuring the PIA on the CoCo for reading the joystick. If it takes multiple clock cycles for the ADC to read the stick position from the time you trigger it, you just trigger it before you do something else and read the results later. As long as you can read the stick once per frame anything faster is pointless anyway. Your argument is flawed in multiple ways.

Your understanding is flawed. If you move around the joystick, then the longer you take to sample the worse your sample. Because joystick is dynamic and your sample is better if it's instantly latched like in the digital case. You need to think about it before you reply again.

[JamesD]

I have yet to see one refutation to

By your standards, I could point to posts that do just that and say you are blind and it wouldn't be a personal attack.

 

(1) eliminating regions of uncertainty in DK and other games for analog joysticks

(2) reducing throw to less than or equal to digital joysticks for analog joysticks

(3) feedback reliant is inferior to knowing states a priori

(4) analog interface itself produces inexact states even after the joysticks are read compared to digital states

(5) analog joysticks have inferior sampling rates (speaking of joysticks in post #1)

Those have been addressed but it is a moot point anyway as example games have been provided where using an analog controller is superior to using a digital one and your original hypothesis is incorrect even if any or all of those are true.

Btw, you never proved any are true, you only stated they are true. Since you are the one making those claims the burden of proof is on you. Feel free to post some code demonstrating where inexact states exist.

Just move an analog joystick all the way right and let go, you will see center is inexact (different from what it was). Now move back to the right and you'll see it's not the same value at the extreme. Even if you hold at the extreme, you won't get the same reading. Your points were refuted many times but I'm going to breifly address them one more time since some people forget.

But it always returns to the the range that is considered center for games like DK which is what you keep arguing about so there is no uncertainty. Either you are arguing about games designed for digital joysticks or you aren't. Your arguments try to have it both ways.

 

FWIW, the Tandy/Kraft deluxe joysticks returned to a very small region. I'm sure the 5200 wasn't quite so reliable due to it's centering mechanism but that is a problem with the design.

 

In games that require an analog stick, a non-self centering stick is usually required and it's a moot point.

 

1) DK is a game designed for digital joysticks, and what regions of uncertainty? There are no in between or uncertain states. The joystick always returns the position of the joystick and the software does range checking. If the code in DK is flawed that is not the fault of the controller, a fact which you have never attempted to rule out in your "experiment". That entire argument has been covered, you just ignored it.

You just don't understand it. When you jump or move around back and forth, you have thresholds rather than exact values at the extreme so there's uncertainty of the motion.

Yes I do understand but we see it differently. You certainly do have thresholds. But you aren't trying to just get to the threshold when you move the stick. Anywhere past the threshold will do.

And as for the extreme, on a game designed for digital joysticks it doesn't matter since the threshold isn't at the extreme.

The digital states are consistent.

 

And if you want to argue that in an analog game you can't consistently find those positions, if you attach a digital joystick in place of the analog stick, you can never find the intermediate positions so you have no control in those areas.

 

2) As I pointed out with sample code, the throw required by an analog stick is configurable. You don't need to use maximum range of movement when simulating digital control on an analog stick. Throw is also dependent on the joysticks used which we pointed out was a small sample size and since tests only seem to have been conducted on the 5200, I highly doubt you even used all the joysticks pictured. Rather than providing custom code that tests reaction times and control in a controlled experiment where variables can be ruled out, your experiment depends on specific games, specific systems, one test subject, and your entire results are limited to them. If you don't like the maximum throw of an analog stick, you could design a piece of plastic to shrink the area of throw and the software could be modified for a shorter throw. But you made no attempts to rule out controller design rather than analog being the factor.

That would be really dumb thing to do. Now how much do you move in the joystick if you shorten the range from center-- it's even worse. Now you added more uncertainty which is worse than the throw. And you haven't beaten the digital joystick anyways. For any distinguishable state from center, you can have a digital joystick with the same distance that's better-- no uncertainty and help of the PCB for the extremes. Again already refuted many many times.

First of all, you have made several statements that have no proof to back them up but that isn't unusual so we'll go with it.

 

If you shrink the area and configure the software to match, you reduce the reaction time since you have shortened the throw.

The maximum range is reduced, you are still doing range checking and there is no freakin uncertainty.

 

Now, on an Atari 5200, the self centering mechanism may not allow such a mod but on the Tandy Deluxe joysticks that center with springs it would work quite well.

 

And you have never addressed attaching a digital joystick to an analog game one single time because you know you'll loose that argument. Which disproves your hypothesis and all these points are moot... as is clearly stated and you did not even attempt to refute that.

 

3) But when you are simulating a digital joystick, you slam the joystick just like a digital joystick. You aren't searching for an exact position. And as I pointed out, once in a while when you try to move at an angle with a digital joystick, you will trigger only one axis slightly before the other axis is detected. You are relying on feedback to identify where the corner is with the digital joystick. As such, no priori can be assumed for digital and your argument would apply to both type of sticks.

You already screwed up in understanding the rotary phone analogy so this point is moot. Digital joysticks allow a priori knowledge of the state not so with analog joysticks.

Argue it all you want, feedback is involved.

 

4) Inexact... "Not strictly accurate or precise; not exact" The controller returns the position of the stick as resistance values indicating the location of the stick. Where are the inexact states? At the very least you have used the wrong word. The state is not as simple as on or off but certainly isn't inexact. Any inexactness would be in the player's control of the stick, or how the software translates the position to a state. The first is dependent on experience and skill which is learned, and the last is dependent on the software implementation. As such, it is not the interface and your statement is incorrect either way.

That's total rubbish. Analog signals don't provide consistent values even if you had a digital joystick hooked as in the Atari 5200. The values change over time, heat, etc. Again already refuted many times. This is about the interface now not about the joystick.

I could sit down and do the math similar to what I did at the bottom but I would need to know what metal is used in the joystick's internal wiring to do the exact calculations. I can safely say that if you had the joystick at room temp (72) and you held it until it was just below skin temp (90, hey, it has some air flow and would cool) the error would be unnoticeable since temperature coefficients for most metals are expressed as x10-3 to x10-5. The resolution of the joystick interface is too small to detect the change. You certainly wont impact the range checking.

 

Now, I can't say what would happen as the temp of the interface changes since the accuracy of the ADC may be temperature sensitive. However, that depends on the implementation and some interfaces may not be sensitive so it would be system specific.

 

Besides you are talking about games like DK which have regions.

 

Again you are trying to have things both ways without ever addressing games intended for analog controllers.

 

5) The analog joystick itself returns a resistance for each axis that changes as fast as you change the position of the stick. Sampling rate would depend on the joystick interface implementation, not the joystick, and as such it is dependent on what machines you do the tests on. Some systems interface digital joysticks through the ports on an AY sound chip and commands must be sent to the chip before you can read the ports. That is no better than configuring the PIA on the CoCo for reading the joystick. If it takes multiple clock cycles for the ADC to read the stick position from the time you trigger it, you just trigger it before you do something else and read the results later. As long as you can read the stick once per frame anything faster is pointless anyway. Your argument is flawed in multiple ways.

Your understanding is flawed. If you move around the joystick, then the longer you take to sample the worse your sample. Because joystick is dynamic and your sample is better if it's instantly latched like in the digital case. You need to think about it before you reply again.

My understanding was based on your statement which could be taken multiple ways.

I like this way better anyway.

 

So you are going to tell me that a computer running l MHz or more that reads a joystick position within a few clock cycles will not get an accurate reading if you are moving the joystick?

 

If we were talking about something in the KHz range you might have a point. But we are talking about millions of cycles per second. In the case of the Atari 5200, 1.7 million. It takes one tenth to one and one half tenths of a second just to blink. Think how long it takes to move that joystick.

 

Suppose the throw from center to a side is 1.5 inches at the top of the joystick.

Suppose there are 32 different values from center to the side. (CoCo's 64/2)

1.5 / 32 = .046875 inches from one value to the next.

(for the sake of ease of typing the seconds below were truncated but I used the un-truncated value in actual calculations)

If it takes 4 clock cycles (just guessing) to read the joystick at 1.7MHz, that is 4/1,700,000 = 2.35294-6 seconds to read the joystick.

.046875/2.35294-6 = ~19,921.875 inches / second, or 1,131.92472 miles per hour, or .314423533 miles / second.

That is how fast you would need to move the joystick to get from one joystick position to the next in the amount of time it takes to read the joystick for this example.

Since it probably takes close to the minimum amount of time it takes to blink to actually move the joystick that far, by comparison you would move the joystick more than .0314423533 miles or 1,992.18751 inches in the same amount of time you moved the joystick 1.5 inches.

 

In the event you just happen to read the joystick on a boundary the result will either be the previous or the new value. It's a tiny difference and in the next screen refresh you will have the new value either way.

[JamesD]

10 DIM A(64):DIM B(64):DIM C(3):FOR I=0 TO 63:READ A(I):NEXTI:FOR I=0 TO 63:READ B(I):NEXTI:FOR I=0 TO 36:READ C(I):NEXT I
20 ON (C(A(JOYSTICK(0)) + B(JOYSTICK(1)))) GOTO 100,200,300,400,500,600,700,800,900
30 GOTO 20


1000 DATA 1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1
1010 DATA 1,1,1,1,1,1,1,1,1,1,1,1,2,2,2,2
1020 DATA 2,2,2,2,4,4,4,4,4,4,4,4,4,4,4,4
1030 DATA 4,4,4,4,4,4,4,4,4,4,4,4,4,4,4,4

1040 DATA 8,8,8,8,8,8,8,8,8,8,8,8,8,8,8,8
1050 DATA 8,8,8,8,8,8,8,8,8,8,8,8,16,16,16,16
1060 DATA 16,16,16,16,32,32,32,32,32,32,32,32,32,32,32,32
1070 DATA 32,32,32,32,32,32,32,32,32,32,32,32,32,32,32,32

1080 DATA 0,0,0,0,0,0,0,0,0,1,2,0,3,0,0,0,0,4
1090 DATA 5,0,6,0,0,0,0,0,0,0,0,0,0,0,0,7,8,0,9

The dimension of C should be 37, not 3

[+poobah]

Yes, rotary phones don't rely on feedback. You know their states a priori.

All of my 5 points made earlier today still stand.

Explain to me how you know what number you are dialing without FIRST LOOKING TO SEE WHERE TO PUT YOUR FINGER??? That my dense friend is VISUAL FEEDBACK.

Do you stop moving your finger because you know:

 

a) EXACTLY where to stop (as "aprioriski" might suggest)

or

b) you finger hits the stop... oh wait, that's feedback =P

 

Shooting fish in a barrel... with all the water drained out

[Divya16]

Yes, rotary phones don't rely on feedback. You know their states a priori.

All of my 5 points made earlier today still stand.

Explain to me how you know what number you are dialing without FIRST LOOKING TO SEE WHERE TO PUT YOUR FINGER??? That my dense friend is VISUAL FEEDBACK.

Do you stop moving your finger because you know:

 

a) EXACTLY where to stop (as "aprioriski" might suggest)

or

b) you finger hits the stop... oh wait, that's feedback =P

 

Shooting fish in a barrel... with all the water drained out

 

All joysticks have extremes that rely on hitting something. So the analog joystick and the rotary phone have extremes. But in the analog joystick case, you can't tell what number you dialed unless you rely on the phone device to tell you or help you out.

 

@JamesD: games were given earlier that were analog and also work with digital joysticks (and better I may add) like Missile Command, Star Wars, Pole Position,etc. So they figured out a way to incorporate in-between values.

[JamesD]

@JamesD: games were given earlier that were analog and also work with digital joysticks (and better I may add) like Missile Command, Star Wars, Pole Position,etc. So they figured out a way to incorporate in-between values.

Even if that is true, and you have not provided any data that can be verified it is, how does that eliminate the fact that there are games where an analog controller is superior?

 

 

BTW, that wasn't what I meant by in between values.

<edit>

In the case of a game like Doubleback it would allow you to play the game but it is tougher to control.

For a flight sim, an implementation similar to what is done in the games you list would not respond like a real airplane. It would be similar enough to pass for fun but there is no position of the stick that would instantly give you a gentle or steep climb immediately. The software would have to do some sort of timing. But I think I mentioned that about page 3.

Edited December 18, 2010 by JamesD

[JamesD]

Yes, rotary phones don't rely on feedback. You know their states a priori.

All of my 5 points made earlier today still stand.

Explain to me how you know what number you are dialing without FIRST LOOKING TO SEE WHERE TO PUT YOUR FINGER??? That my dense friend is VISUAL FEEDBACK.

Do you stop moving your finger because you know:

 

a) EXACTLY where to stop (as "aprioriski" might suggest)

or

b) you finger hits the stop... oh wait, that's feedback =P

 

Shooting fish in a barrel... with all the water drained out

Exactly.

[JamesD]

10 DIM A(64):DIM B(64):DIM C(3):FOR I=0 TO 63:READ A(I):NEXTI:FOR I=0 TO 63:READ B(I):NEXTI:FOR I=0 TO 36:READ C(I):NEXT I
20 ON (C(A(JOYSTICK(0)) + B(JOYSTICK(1)))) GOTO 100,200,300,400,500,600,700,800,900
30 GOTO 20


1000 DATA 1,1,1,1,1,1,1,1,1,1,1,1,1,1,1,1
1010 DATA 1,1,1,1,1,1,1,1,1,1,1,1,2,2,2,2
1020 DATA 2,2,2,2,4,4,4,4,4,4,4,4,4,4,4,4
1030 DATA 4,4,4,4,4,4,4,4,4,4,4,4,4,4,4,4

1040 DATA 8,8,8,8,8,8,8,8,8,8,8,8,8,8,8,8
1050 DATA 8,8,8,8,8,8,8,8,8,8,8,8,16,16,16,16
1060 DATA 16,16,16,16,32,32,32,32,32,32,32,32,32,32,32,32
1070 DATA 32,32,32,32,32,32,32,32,32,32,32,32,32,32,32,32

1080 DATA 0,0,0,0,0,0,0,0,0,1,2,0,3,0,0,0,0,4
1090 DATA 5,0,6,0,0,0,0,0,0,0,0,0,0,0,0,7,8,0,9

The dimension of C should be 37, not 3

And the function is JOYSTK(), not JOYSTICK()

[atarian63]

Yes, rotary phones don't rely on feedback. You know their states a priori.

All of my 5 points made earlier today still stand.

Explain to me how you know what number you are dialing without FIRST LOOKING TO SEE WHERE TO PUT YOUR FINGER??? That my dense friend is VISUAL FEEDBACK.

Do you stop moving your finger because you know:

 

a) EXACTLY where to stop (as "aprioriski" might suggest)

or

b) you finger hits the stop... oh wait, that's feedback =P

 

Shooting fish in a barrel... with all the water drained out

uh.. no... :roll

 

Gaming de evolution/ Joystick (Atari) >Gamepad (Nintendo)> analog joystick> Kinect (MS)

bottom reached.

[JamesD]

uh.. no... :roll

 

Gaming de evolution/ Joystick (Atari) >Gamepad (Nintendo)> analog joystick> Kinect (MS)

bottom reached.

Which is your opinion, not a fact.