In short... a life with only 1 monitor is not a life worth living. Once you go to two screens it is very hard to go back.

@DizzyOfCRN

Doing a HIDD doesn't have to be too hard, because you can subclass and inherit a lot of methods.

Also, you could copy or subclass the VESA HIDD, and keep a standard frame buffer in host memory, and then convert it to JPEG and send it to the device in a refresh function. If you look at rom/hidds/vesa/hardware.c, vesaDoRefreshArea() could perhaps be replaced to do this.

I have an interest in getting dual display support from which I can use castAR stereoscopic glasses.

Basically, one can think of castAR as two projectors. Each projector would display an 1280x720 pixel display. Since it is intended for resolving 3d stereoscopic output of a 3d virtual environment.

However, castAR is also a headtracking/position device that will feed data down USB to the computer from which a software can then be process the data.

In this case, it would probably make logical sense to consider each display to be separate content because left eye projector and right eye projector will not have the exact same content. Even though, in theory, we can have some degree of common data when we think of a 2d desktop but when running a 3d app, it would be two different viewing perspective of the same 3d environment. Just like your left eye's view is not exactly the same as the right eye due to displacement. In a 3d game, you have a virtual camera for your view. In our case, we would have actually two virtual cameras.... each representing an eye's viewing perspective.

AROS desktop would be 2d plane.... so in our case, we would more or less display the desktop which would be a moderately larger bitmap desktop that is say.... 1920x720 but left eye projector would display the left 1280x720 while the right eye would display the right 1280x720 but there would be overlap.

The combined pixel resolution of the left and right projectors is 2560x720.

Some of you may already know about castAR. I'm just trying to make it understandable as possible to those not familiar of castAR that is under development.

My goal is to work in getting this display device supported by AROS/IcarosDesktop.

This would be more or less new to me.

castAR uses HDMI video output from the computer to display the whole output.

As more details becomes available, I would like to continue the discussion as I become more familiar with how the device works.

I got the lowlevel stuff to work:http://youtu.be/X04M7VPC_1o

It only shows one statically linked JPEG of our Aros kitty for now.

What if I open a public screen which is hidden and just grap the screen and convert it to JPEG and send it to the frame? It would sound a lot simpler than doing gfx hidd? But would allow content for the frame to be created...

I also found out that the brightness of the frame can also be controlled and I assume that the key presses are diverted to int endpoint, don't know waht use there could be for them thought.

Thank you, it sounds more complicated than I anticipated...

I vaguely remember that perhaps Sonic added multiple monitor support, meaning that the main thing for you to do would be to implement a graphics HIDD for this device (i.e. a subclass of the top level graphics HIDD).

Resurrecting this very old topic...

I bought for 10? a Samsung SPF-87H digital photo frame and it has a nice feature that one can send a jpeg picture to it via usb (mini monitor mode) and the frame displays the picture. Protocol for it is very simple but what sort of higher level API should the usb class have if implemented?

Update rate is quite adequatehttp://youtu.be/mN_FDZX3iyc

USB driver could make it's own "graphics" library for simple gfx output but I'd really like to just openscreen() on it...

well, i had allready put some thoughts in this topics...but i would like to show ggi again...it has, afaik, the feature of split one gfx target to several

I've got a stack of old video cards. I plan to stick them all in one machine so I can work on multiple monitors at some point. There was a Mac flight simulator that had the ability to use three displays - it was really cool. Being able to do something like than in AROS would be a real plus.

Quote:

Thats what I find difficult to picture (ehem) - what about software that directly draws to the bitmap?

Its this that lead me to think we would need to do it through a rastport or something =/

(most operations work on a rastport so it would be trivial to check when it occurs and redirect to the appropriate driver)

Quote:

No - I'm aware of this, and its exactly "why" that I am trying to identify =/

.. and more importantly how to start changing AROS so it IS possible..