RandR future proposals

[Graeme Gill]

Keith Packard wrote:
> Sure, we need monitor-specific implementations of much of this; is there
> some reason to prefer a place outside of the X server?

Not at all, I was just concerned that aiming for that level might make it
a task with too many imponderables, and therefore prevent even the basic
centralized DDC access being implemented any time soon.

 > We're already
> starting to capture monitor-specific quirks related to EDID data which
> the X server must be able to process; surely we should work to
> centralize such data in a single project. 

Sounds good. Some degree of extensibility to accommodate new
monitors or model specific features without having to
upgrade X would be a nice quality.

 > If ddccontrol has created a
> database format, it's quite possible we can just adopt that, and perhaps
> even add our existing quirk data to it.

It would be sensible to leverage their work, if they are willing.

> Given the potential for abuse, I'd really like to avoid opening up an
> uninterpreted I2C channel from clients to the monitor; is there some
> piece of ddccontrol which is not appropriately exposed as a RandR output
> property?

I don't know enough about the detail to comment on that. The fact that
information is all locked up behind VESA standards makes it rather
inaccessible to those without corporate backing.

Graeme Gill.