difference between iGesture and iGesture NumPad?

[SputNik]

from what I can tell, using the wayback machine (www.archive.org), there is no functional difference between the regular iGesture and the iGesture NumPad. Is this correct? Is the only difference that the NumPad has the number keys printed on it?

I'd be fine with a regular iGesture if this was true, as I don't really need to look at the numpad to hit the right keys. Can anyone confirm this for me?

Thanks,

Nik

[Rqyteqto]

Hey SputNik -

Did you ever get any answer on this question? I am trying to find out if an iGesture pad can be programmed for characters instead if numbers or whatevdr. Basically, my interest is if two iGestures can add up to one TouchStream, sort of.

[SputNik]

never got a reply...from all I can tell, they are identical, but that's simply speculation -- I wouldn't rely on it unless someone else can back it up.

I gave up, tho, because I found a full TS on craigslist for incredibly cheap, which is what I really wanted.

[Rqyteqto]

Hey SputNik -
Good to hear from you. Congrats on finding yourself a TS.

I am looking for a second myself, sort of a backup. just in case mine fails. That's why I am looking at the iGesture pads. The Human Solution has a pack of them for sale at $200 a crack. I've been trying to pick one up on EBay, but keep getting outbid.

I figure if I could get two iGestures working like one TS, it would be almost as good.

From what I can make out from the documentation, it appears it should be possible. It seems the boards can be programmed virtually any way you can think of and if someone was to get into the software, a whole lot more ways. As best I can tell, the boards are all functionally similar, except for the size and number.

I'll have to check out Craig's List, thanks for the tip.

If I get a couple of the iGestures, I'll experiment with them and publish what I find out.

[rolfwind]

I figure if I could get two iGestures working like one TS, it would be almost as good.

Actually that's a very good idea - I checked out the documention and this could even be better than a keyboard - I don't see why it couldn't be programmed that way.

It could be even better than a keyboard if you stick in the time programming it because there's no key layout printed on two igestures (I'm thinking of switching to something even more exotic than Dvorak) - you'd simply need to make one - I think I will try this - I have transparent paper for making transparencies with inkjets and a color laser printer - SWEET!

Anybody know if thin transparencies on top of the pad still let this thing work?

Anybody know if the igesture can be this programmed? I'd like to use the programmer's notepad by changing modes and don't know if the regular igesture is more limited than the datastream keyboard.

Plus it'd act like a split keyboard. Or maybe have three units (don't know if connecting to a USB hub for this may have issues.)

[TorbenGB]

get two iGestures working like one TS, it would be almost as good.
...
As best I can tell, the boards are all functionally similar, except for the size and number.

A great idea!
I guess the only thing that makes the TS Different is that the two halves can work together -- when you use a modifier chord on one side to get a shift/punct/programmers pad on the other hand. It would be tricky to get this benefit from two iGestures...

Anybody know if thin transparencies on top of the pad still let this thing work?

You can put any thin sheet on top and it still works. I sometimes experiment with plain inkjet paper (to make my own Dvorak or other overlays) and this doesn't bother the TS at all. It bothers me in so far as the dimples get hidden so you have even less tactile feedback, and the surface texture of the original TS is really unique -- paper or foils give a poorer feel for the surface.

[Rqyteqto]

Well, I intend to find out what the iGestures will do. They are supposedly the same as the separate halves of the TS but as noted, without the interconnect. I doubt it would be possible to get them to be interactive, though I wouldn't rule that out and I imagine somebody with some very good programming skills might pull it off.

There is no question one can reprogram the various keys laid out anyway you want. And its clear those keys can be repositioned to some extent. What's not clear is if its possible to completely redefine the key layouts. I would think so, as in mouse mode, there's no pattern whatsoever.

[TorbenGB]

There is no question one can reprogram the various keys laid out anyway you want.

I agree. If you're clever enough to feel at home with the raw XML configuration file behind the FW software, you could probably do pretty good magic right there. The software lets us configure the surfaces easily, but if one thoroughly understands the XML then all you need is Notepad and you only need the FW software to upload the XML to the device.

[Rqyteqto]

"If you're clever enough ..."

" ... if one thoroughly understands the XML ... "

Two mighty big "ifs", at least for me.

Could be a great project for a vacation, if I ever get around to taking one.
Perhaps a good reason to take a vacation, I figure at least a month with no interuptions, no distractions.

[ivanw]

No need to be so smart... XML is not some kind of mysterious thing, you can even learn it, mind you! (peace of cake)

This is the customization xml you need to move your Q* around every now and then:

Code: Select all

<xml>
<DOCTYPE>
<MTS_config>
 <TapArea><MacroRefEvent> </TapArea>
</MTS_config>

• 

The different positions correspond to these values for line 4:


And, believe it or not, you can do that with the other letters too.
* ...in French, when you spell the letter Q, it makes the same sound as when you say "ass"

[Rqyteqto]

That's amazing. Certainly looks simple enough, but then I've always found most programing and indeed most everything to do with computers is relatively simple, or at least relatively logical and rational, which for me is one and the same. If I understand the rationale and the structure of the language, I can learn it relatively easily.

So I take it if the 'xcenter' and 'ycenter' or the 'xradius' is not defined, its value defaults to zero (or '1' in the case of 'xradius'). I also assume the units are not related to any measurement system, but are essentially "fingers" or perhaps "touches", a measurement unique to MultiTouch boards and relate to the board itself. Are they fixed units, ie - 1 touch is the same for all products or are they relative, with the same number of units for each board.

What's the resolution available? Can one specify .01 touches? Or .001?

I also take it 0,0 is at the bottom left corner for right hand pads and the bottom right hand corner for all else.

[ivanw]

So I take it if the 'xcenter' and 'ycenter' or the 'xradius' is not defined, its value defaults to zero (or '1' in the case of 'xradius')

It looks this way. All this reverse-engineering makes us equals for the guesswork.

I also assume the units are not related to any measurement system, but are essentially "fingers" or perhaps "touches", a measurement unique to MultiTouch boards and relate to the board itself

Same thought here too. Centimeters seem to apply for positioning.

Are they fixed units, ie - 1 touch is the same for all products or are they relative, with the same number of units for each board.

I think that units are to be loosely considered here, common sense should give the answers each time.

What's the resolution available? Can one specify .01 touches? Or .001?

I played with +/- 0.01 to adjust at the pixel level in the editor. But this is absolutely not an issue when you know that the selection process will still choose the closest key, should you have touched one centimeter astray in an empty area.

I also take it 0,0 is at the bottom left corner for right hand pads and the bottom right hand corner for all else.

Same conclusion here too.

[Rqyteqto]

Well, this is very interesting. I have to believe these boards are the epitome of good design, ie- the designers didn't take the easy way out and simple figure, "Hey, nobody will ever need more than 64 KB of RAM" and so build a 64 KB limit into the OS.

It always pisses me off when I come across something that's really powerful, or at least could be if the designer had just allowed for such and such to happen rather than assume what the functions will be and limit the usage to that. The printing press was originally intended to print bibles, because the demand for bibles was growing exponentially and doing them by hand was simply too slow. You could only train new monks to copy the texts so quickly and that took up a lot of monk-power that could be used for better purposes, like making cheese, wine, etc. Think of where we would how be if Guttenburg had somehow made the press so it could only make bibles. As it was, it took nearly two centuries before the presses were used much for common purposes.

Really have to thank you ivanw, this will all be of immense help. It looks like its possible to reprogram these buggers just about anyway one could imagine. I just love it.

Eventually though, we will need to figure out how to build the boards themselves, or something equivalent. Even with no moving parts, they will wear out sooner or, hopefully, later. I'm committed to getting myself another TouchStream and at least a pair of iGestures. If I had the resources, I get a third TouchStream and a second pair of IGestures just for good measure, but I fear the better half will balk at that.

[TorbenGB]

Wow Ivan, that's a great introduction to "FWXML"!! We should distill this forum post into a sort of manual FWXML, so that we have a useful reference point.

I've started this page:
http://fingerfans.dreamhosters.com/bin/view/Main/XmlManual
and I would like everybody's help to complete it.

[ivanw]

I've started this page: http://fingerfans.dreamhosters.com/bin/view/Main/XmlManual
and I would like everybody's help to complete it.

I like the contribution idea and I will insert my own Question for fingerfans FITB.

BTW, attachment does not work for me, I never could upload anything.
My last attempt of this morning in Sandbox/WebHome failed too.
...no error, no timeout, just nothing happens when I submit with the Upload File button.


• It looks like the Upload up to 0 KB limit applies :
  
  

[TorbenGB]

It looks like the Upload up to 0 KB limit applies

"0 kb" just means no limit. I've set it to 100MB now, should be enough... retry, abort, fail?

[ivanw]

I see the new limit now: Upload up to 1000000 KB but it did not help.
To make sure I was correctly identified, I edited my personal page Main/IvanW. (added a link to stumbleupon for those not yet addicted to). Editing worked fine.
When I tried the Attach procedure, page freeze, same story. It may be on my part though. It would be good to know if some others have the same issue.
Sorry for the burden.
- ivan

[Rqyteqto]

The TWiki manual is fantastic, you both have been formally knighted and will beatified into sainthood soon.

[TorbenGB]

I need your help ... Could somebody please try to attach a file to that page?

Ivan just added a lot more details, and I've moved those around to match the existing format. All, please keep the contributions coming! And let's use this thread for coordinating this effort.
It seems to me that the Fingerfans forum here has gained a lot of momentum. Now let's see if we can do the same with the wiki knowledge base?

[ivanw]

I intend to dig a bit further in those places where I've just dropped by while playing with the editor. It looks like Wayne Westerman did everything to invite anyone interested to grasp the idea and get involved...

Still being convinced that practice is the best way to get the most out of the TouchStream keyboard, I think that two other commitments of effort are necessary: practice and practice!

But it is so much fun customizing this keyboard when you get to discover how much it is far from being as stupid as what we dealt with before.

It's like playing god in a corner of your attic

BTW: I've been unable to edit the XmlManual page for pretty some time yesterday. I found that we should check the Release edit lock when we are done with our last modification:

• editing
• re-editing
• correcting
• ...done

I am not sure if locks apply starting with the first preview -- but I guess it could be so
I read about edit locks and came up with adding the following two settings in my personal page to avoid that:

Code: Select all

 Set RELEASEEDITLOCKCHECKBOX = checked="checked"
 Set DONTNOTIFYCHECKBOX = checked="checked"

This has the result of automatically checking these two options when I a the page:

• Release edit lock
• Minor changes, don't notify

The price to pay is that concurrent editing is not prevented! ...for those who understand what it means and are ready deal with (by looking at the original page right before submitting to see if anyone did change it while you were working), it has the advantage of not holding an exclusive editing permission of a Wiki page for the next coming 60mn (default value).

I woul like to read from anyone who has some practical advice on the subject.
...even if it is something like "Hey, just go read ThisWikiPageOnTWikiPageEditing"
____________________________________________________________
edit:
Locks do apply starting with the first preview according to this TWiki FAQ link: SimultaneousEdits

[TorbenGB]

I am not sure if locks apply starting with the first preview

Ivan, you're on the right track. When people VIEW a page, nothing is locked. When somebody clicks the EDIT link, that page is locked from others for 60 minutes. Then it automatically unlocks again. One can choose to open the lock earlier by ticking the "unlock" checkbox before clicking the SAVE button. That's all there is to it.

The price to pay is that concurrent editing is not prevented!

Wait! These checkboxes are there for a reason: to make sure nothing is lost.
EITHER keep that checkbox checked and DON'T GO BACK in your browser after you've saved --
OR leave it unchecked (so that you don't unlock the page) and then it's okay to go back in your browser to edit some more.
DO NOT mix these practices because you risk killing somebody else's contributions.

I woul like to read from anyone who has some practical advice on the subject

It's simple, really. Think of the lightswitch to your bathroom: if it's on, it's occupied. So remember to switch it off when you're done = remember to unlock the page when you're done editing BUT NO SOONER.

[ivanw]

I'm quite comfortable with the bathroom switch metaphor.

But something still puzzles me: I'm pretty sure that yesterday, I have manage to lock the XmlManual page in such a way that I couldn't even open it myself. It was like if I got outside of a toilet room without opening the door. Then I could not do the trick the other way around to get back inside.

At the time, I did not bother to fully understand about editing locks and that must be it! Now I would like to really stumble upon the color-sign indicating someone is currently editing the page I was about to modify.
I've not seen any yet. Yesterday I've just been unable to preview or save my changes, that was the only sign I got.

Still learning Wiki behavior...

[TorbenGB]

Yes, wiki can be tricky sometimes but compare it to coding HTML and FTP'ing your changes: wiki takes a little learning, but gives so much ease.

sign indicating someone is currently editing the page I was about to modify

There is no lock indication until you try to edit. When you click EDIT, you either get to the edit mode or you will be told that "user X has locked the page for editing, and it will be unlocked in X minutes". Most of the time you'll only want to view-not-edit anyway, so this is less clutter on the screen. Also, consider if 2 people view a page and then one decides to edit: the other wouldn't see a lock anyway.

[rolfwind]

Anyone get typing with iGesture working?

I've been playing about with the software and xml....... after inserting the first tap area into the XML file - the gesture editor reads it and the surface keys icon pop up to my joy.......

But still can't get anything to type...... the gesture editor automatically adds lines like these:

<EventMacro macro_name=" B " lib="user" user_selectable="false">

<KeyEvent keyCode="B" driverContext=" qwerty qwack" />

<KeyEvent keyCode=";" driverContext=" qwerak" />

<KeyEvent keyCode="X" driverContext=" dvorak" />

<KeyEvent keyCode="B" duration="50" />

</EventMacro>

to the configuration file. I wonder if it's because under Global Feature Editor all the typing configuration is grayed out.

I tried swapping device drivers (to the numpad model) but the software doesn't allow that.

The other thing I noticed is that even though I move the keys around graphically in the gesture editor - it doesn't change the x/y center in the XML.

The only way I'v been able to pass a key code is turning back on the 1 finger clicking and overiding that event with a keystroke. But I can't figure out how if it's possible to have different areas by that method (otherwise I'm typing a grand total of 1 letter everywhere I type).

Anybody?

Edit: Update - I made the radius of one of the keys 10 and still nothing so it's definitely not for lack of hitting it:-)

[ivanw]

Hey rolfwind, your post got us back to this very topic after a bunch of 15 relating to the configuration file format.
_______________________________________________________

It happens to be difficult to get on with this thread without shouting loud!

...Should our admin -- thank you Torben -- move things around in some new Forum Index titled something like About knowledge base Wiki ?
- It looks like this alien posts here belong to some topics named after their subject like XmlManual or Wiki page editing.

Torben, this is just a suggestion on my part. Please, consider doing this only if it does not require more time that you can afford to the task and that it may not get you into trouble.
_______________________________________________________

Back to what rolfwind is doing, only those who have an iGesture at hand can experiment further and I only have a TS so I can't.
But I have some inputs on this:

1. MyGestureEditor is driven by hardware dependent conditional attributes - they could just be deleted to loosen up relevance controls.
2. A fairly precise understanding of all these hardware dependent attributes must be achieved to elaborate on how to make sure that:
   2 iGesture + cable + = TS ...
3. Wiki XmlManual will help.