RandR version 1.2 revisited

[Graeme Gill]

Adam Jackson wrote:

> It appears to.  I don't have the base EDID docs, just the implementors guide, 
> which mutters something about requiring sRGB displays to encode exactly 
> certain values.  The EDID code already in the server includes a function 
> called "get_whitepoint_section" so I imagine that's what it does.

 From my perspective, this doesn't address DDC/CI concerns at all -
getting the EDID passively doesn't help much, because what I want
to be able to do is to communicate with the monitor using DDC/CI, so
that I can set the monitor controls (brightness, contrast, etc.),
to set a calibration condition.

What this translates into is that ideally I would like to
be able to send and receive the DDC/CI packets connected to
an Output. For an example of what I'm talking about, see
how Apple does it:
<http://developer.apple.com/documentation/Darwin/Reference/IOKit/IOI2CInterface/CompositePage.html#top>

Alternatively, if DDC/CI communication is traditionally provided by some
other sub-system, and if there is no desire to tie this into X11,
I would need a way of locating the DDC/CI channel that is associated
with a given Output.

Now I can quite understand this probably falls outside the anticipated
scope of RandR, and maybe it is better left to a latter revision,
or some other extension, but I thought I would at least flag it
in peoples thoughts.

Graeme Gill.