Re: [PATCH v2 1/1] xinput-calibrator: move it from meta-oe to oe-core

["Burton, Ross" <ross.burton@intel.com>]

On 2 July 2013 17:21, Laurentiu Palcu <laurentiu.palcu@intel.com> wrote:
>> The file always has content because the file in the main recipe has a
>> comment in.  Personally I think not installing a useless file is the
>> right thing to do, so we should change the default pointercal.xinput
>> to be empty.
> Then why install it at all if it's empty? I believe a zero sized file in
> /etc it's even worse.

Exactly, this test will spot the empty file and not install it.

>> Relatedly, interestingly Weston has a different approach and attaches
>> the calibration to the device using udev properties, which means
>> systems where touch devices can be hotplugged actually work.  I'm
>> incredibly tempted to extend xinput-calibrator to write out udev
>> fragments and make Xi respect them if they exist...
>
> Ross, you requested (according to the description in bugzilla:
> https://bugzilla.yoctoproject.org/show_bug.cgi?id=4416#c0 ) to just move
> xinput-calibrator to oe-core from meta-oe. Technically, the first
> patchset did just that. It would have been easier for me, from the
> planning point of view at least, if I had known in advance the
> additional changes you needed for this package. I admit I don't know how
> all packages in oe-core or meta-oe should or shouldn't work... I'm
> working on improving that though. ;)

My talk of udev properties is just talk, I should have made that
clear: it's nothing I expect you to do.  Once the systemd bits are
purged entirely, I'm happy to see this merged.

Cheers,
Ross