For new forum users:
MU= Multi User
MT= Multi-Touch

The thread is meant for idea generation/mock up generation. For those who have the time, working prototypes or a fully working OS (hopefully OSS) are welcome.

On http://nuigroup.com/forums/viewthread/3893 (and its reply: http://nuigroup.com/forums/viewreply/25480/) there are ideas on how to activate or call a system menu.

For a general MT OS, what would/should the default launch/landing screen be?

Any ideas how to manage multiple user accounts where only one user is logged in but many share the table? Should user accounts use a virtual keyboard to enter user name and password, online gestures to enter username and password or another system (or combination of systems) such as click an icon to represent a user (name is entered on a virtual keyboard) and click another set of icons to represent their password so that when prompted for their password, the screen fills up with 20 icons out of which 3 (or more) are related to the user and the other 17 are randomly generated?

We had a similar discussion to this a while ago.  You might want to look it over.

http://nuigroup.com/forums/viewthread/2538/

Just for fun!

Why not use special IR cards for identification (don’t know if there are any)? It’s just my thought that you can make a credit card size “something” that emits IR light in form of bar code, so that a camera can detect it. You place your card on a table and OS loads your profile.

Or just use a USB or Bluetooth key. Then there could be some simple graphical pass key, like choose the color sequence or something.

Everything that makes you input something with keyboard is wrong, I think.

@Xela, that link really brings me up to speed with the ideas. The ZUI approach seems interesting for file management. When I think of it, I can’t help but think of drawers underneath the first ‘desktop’ and shelves above the first ‘desktop.’

Bumptop (link from the thread Xela referred to) have similar ideas in their paper.

@silenz, the usb sticks works when no one should ever see the passwords being entered or even user name being entered. As for IR cards, it could be possible for the cards to have an image printed on it that appears only under IR light but perhaps RFID cared would be better (and probably what you had in mind) as that technology is already in use. I’d say a keyboard at least in virtual form needs to be present as a fallback device in case gestures don’t pick up characters for users entering their information into the system for the first time.

Here are 2 screens that can appear when starting or waking up a system from stand by:

After tapping the corner hotspots the following screen can appear

While not in the mock up, a count down timer can be used to tell users how long to hold the hotspot until it is registered as a full hold.

Some ideas on what can happen next are discussed on: http://nuigroup.com/forums/viewthread/3893/

@ SteveJB - That looks understandable. Only the amount of text in the mock up is a little confusing.

About the count down timer - Could there be a simple circular progress bar going around the tap area when holding? That would indicate that something is happening and the rough amount of time needed to complete the action. Just a quick animation, since the hold time would be what, about 2 seconds?

@nolodude The count down timer makes sense.

A circular progress bar could be difficult for the general user to understand although it has its use cases. Perhaps a more generally used horizontal progress bar would work as the default.

How does a person read ‘tap to start’ when he’s standing on the other (top of your image) side?
ie: the ‘tap to start’ text has to be readable from every side.

@SilentSimon Good Catch there. For illustrative purposes, I probably left it in to be not confusing for first time viewers of the mock up.

Estimating the orientation of the viewer is another issue. How does would the system know the orientation of the user based on corner hot spots alone?

I’ve been thinking about having a hot spot in the center of the screen with a displayed ‘coin’ that can be dragged towards the user. When the user holds the coin (a count down time or progress bar can appear) a circle appears around the coin and the user needs to drag the coin towards the edge of the circle closes to the user. Another alternative is to keep the corner hot spots and have the user drag an image in the hot spot display to the edge (of the table) closest to them.

SteveJB - 25 March 2009 08:46 AM

@SilentSimon Good Catch there. For illustrative purposes, I probably left it in to be not confusing for first time viewers of the mock up.

Estimating the orientation of the viewer is another issue. How does would the system know the orientation of the user based on corner hot spots alone?

I’ve been thinking about having a hot spot in the center of the screen with a displayed ‘coin’ that can be dragged towards the user. When the user holds the coin (a count down time or progress bar can appear) a circle appears around the coin and the user needs to drag the coin towards the edge of the circle closes to the user. Another alternative is to keep the corner hot spots and have the user drag an image in the hot spot display to the edge (of the table) closest to them.

First of all, wow! What a group, the NUIgroup! Give all of yourselves a round of applause!

This was by far the most interesting thread I´ve read in a long time. I mean, I´m constantly digging through the hardware sections to learn as much as I can about building my own MT device, but this is where it really happens. Interacting with a MT setup is a fascinating subject I´ve also been thinking about, and more precisely about something that I feel haven´t been discussed:

Almost every MT setup I´ve seen built is based on either a webcam or a videorecorder registering touches, and in both cases they have microphones in them. Especially like the PS3eye which has three microphones. Is there beeing work done to make the MT setup “hear” you, and solve the issue with user orientation in table setups by, for example, knocking on the table and everything rotates towards you automatically?

/Sebulon

I’m just getting back into the group again, so please forgive me if my ideas have been stated elsewhere.

Why not simply put the hot spots on the edges instead of the corners? Wouldn’t users most likely tap the one nearest to them?

I think one of the main difficulties with multi-user interaction is the orientation of objects. There is usually one “main” orientation.

One scenario I have been thinking of is when there is initially 1 user. The “main” orientation is suited for them. When another person approaches the table, it now becomes a collaborative experience. So for example, if the initial user was browsing the web. They should be able to create a duplicate window for the other person to see what they are looking at and collaboratively browse. User #2 can now also interact with the browser. Let’s say user #2 clicks on a link, both windows should change.  I think this is a good solution to user orientations in the case of multiple users, but it does take up more screen real estate. I don’t know if I’ve explained myself clearly or not, but I think it’s an interesting problem to tackle.

Phyllers

did you think of vein pattern recognition? i tried it through special glass from “schott” ( a german specialist in glass-manufacturing), and it lets through about 85% of IR light and only very little of visible light… you could try to do something like this with some special projection material.
i can´t help you much with this since i´m not a programme, but just google for vein pattern recognition and stuff and you´ll see that it works with 850nm IR LEDs…