On-Calc handwriting

[Mike K]

Ok, this may sound a little off, but It's a cool concept.

Some of you may know of the routine that was written a while back, that let you slide your finger around the calc buttons, to simulate a "mouse" moving around the screen.

This idea stems from that, but instead of a mouse, it would be like a paint program.

Things to include, well, the one thing to be added, depending on how long it takes to go from key A to key B would determine the length of the pixel line on screen. For example, I press the ON key, then immediatly shift my finger up 2 keys and press that. The stroke on screen would be long than if I had waited 1/2 second longer.

-Mike

[Madskillz]

wow...that would be an extremely cool idea! Any hope for this Mike?

[Patori]

Yeah, that's such a good idea!

[tr1p1ea]

Are you planning on coding this Mike K?

[soundbyte24]

So like, moving from 5 to 2 would make the mouse go up? Stuff like that? Because that's awesome!

[DarkAuron]

Even neater would be making the speed dependant on how long it takes you to go from one key to another, from up to 3 keys, and going in a weird diagonal direction could have more natural movement, or maybe how many keys you press in a certain direction within a certain amount of time, OR a combination of the two. Dunno if that made sense but it does to me

[Spengo]

That sounds cool, but it also sounds like a LOT of key checking every main loop.

[DJ_O]

yeah indeed, I hope it would be made in ASM so it runs fast enough

[Jim e]

Isn't that how that mouse program worked, I remember pressing [log] then pressing [*] would move a great deal. But making it like a mouse would mean taking samples, then based on where you finger is calculate the movement. Basicly sounds like putting a fancy key routine in the an interrupt.
What i don't understand is how time would affect this.

But the real question is application of such a mouse routine.

[Mike K]

Mirage 2.0!! Well the main idea was if I had the space between each key worth one pixel on the screen, all you would get is a jumbled mess or a clump of pixels. If the routine is time-dependant, meaning it will calculate in how long it took for your finger to go from key A to key B....

for example... Key A is one key to the left of Key B

Key A --(.1 sec)--> Key B = 1 pixel
Key A --(.2 sec)--> Key B = 2 pixels
Key A --(.3 sec)--> Key B = 4 pixels
Key A --(.4 sec)--> Key B = 8 pixels

etc. Now, .1 seconds is quite a lot so you could of couse use a much smaller unit of time.

-Mike

PS: applicational use? Maybe require a signature to sign-in? That would be fun to write...

[Jim e]

The problem with using time is that you are likely to hold your finger over one key for a good amount of time so if you try to move quickly you would be hindered by this calculation. A real mouse isn't based on time but relative movement.

Plus what if you hold you finger over many keys at once how would this be accounted for.

[Patori]

What we really need to ask, is "How does a pda account for multiple pressure points?"...

Anyways, a sig translator would be fun to code, but very difficult... some people have very different handwritings, and a pda, for example, requires a user to write in a certain way....

I don't know about the jumping, maybe it can do nothing or something.... I dunno....

But this is a cool idea for a project, and I would like to see it completed!

[lecks]

[quote="Mike K"]Ok, this may sound a little off, but It's a cool concept.

Some of you may know of the routine that was written a while back, that let you slide your finger around the calc buttons, to simulate a "mouse" moving around the screen.

[/quote]


where do i get the otiginal routine? my brother could probably code this. his whole life is computers and he's taking some programming classes in college. this could be his project.

[threefingeredguy]

he does ASM? send him to www.revsoft.org, we want to talk to him. And you.

[CoBB]

I made that routine, and it did account for the speed of action and could handle any key combination pressed too. However, you can't read handwriting on such limited hardware. It's not just a programming task, but also requires a lot of theoretical knowledge (neural and hybrid systems and the relevant mathematics).