[Openicc] colord 0.1.7 released!

[Richard Hughes]

On 6 May 2011 18:41, Alastair M. Robinson <blackfive at fakenhamweb.co.uk> wrote:
> Hmmmm - could you explain that to me please?  I trust the sensor won't be
> locked except when specifically required, and thus won't prevent other
> software from accessing the instrument?

Sorry, that was a pretty crappy commit message.

Basically, the model we've got at the moment is that udev matches the
sensor, and colord gets notified and a DBus object gets created. If
the session component wants to use the sensor, it calls Lock() on this
object and then is allowed to called GetSample() as many times as the
session wants. Once Lock() is called, then the device is connected-to
using libusb (for native drivers) and this of course locks out other
applications. The client can either then call Unlock() or just exit
(where colord cleans up automatically to remove the lock), which of
course disconnects the device using libusb. At the moment only the
Huey has this kind of native driver, but hopefully if the argyllcms
drivers are released as GPLv2+ then I can dl_load those and handle all
the hardware in this way.

This means in normal usage the device is freely available to be used
by Argyllcms.

The commit message was basically saying that there's now a "colormgr
sensor-lock" command that just calls Lock() on all the sensors and
waits for any events from the object. It's mainly for testing the
auto-cleanup and the ButtonPressed signals, and isn't really useful
for end users at the moment.

Richard.