chuckr at telenix.org
Fri Sep 5 07:16:11 PDT 2008
-----BEGIN PGP SIGNED MESSAGE-----
Daniel Stone wrote:
> On Thu, Sep 04, 2008 at 03:35:17PM -0400, Chuck Robey wrote:
>> Soon as I figure out how other drivers do the function keys, so I can do my own
>> that way, I think it might be ready for release. My tablet comes with a thick
>> plastic overlay with the function areas delimited ... I can figure _A_ way to
>> handle this, I want it to be _the_ way to do it, if I possibly can. Been a lot
>> of learning to get this far, I surely am pleased to be here now.
> If they're actual keys, you can use xf86PostKeyEvent. If not ... just
> ignore it for now. We need to be able to map given areas of particular
> devices' co-ord space to given areas of CRTCs, AFAICT. Being able to
> grab particular (obviously non-overlapping) areas would probably be
> quite useful for shortcuts, etc, but this may be overkill: maybe we
> just want key events, rather than Button/Motion events on shortcut
I'm not completely certain I follow all of this, so let me try again, and
personally try to make NO assumptions this time.
I have a thick plastic overlay on my tablet (came with the device) which both
delineate 2 thin strips on on left/right, and two on the top/bottom of the
tablet area. The strips are subdivided (8 each on left/right, 13 each on
top/bottom), with each area having a (done in pretty small font) description of
a function, and an arrow showing a graphic sort of handwritten input you can
give. I don't know if you are supposed to just enter those squiggles anywhere
to get the keystroke action, or if maybe you are only supposed to give the input
over the designated area, or what. Or, maybe the squiggles are total overkill
to begin with?
I think it wouldn't be all that hard to implement active areas that, if you
clicked into, would dispatch either keys, or possibly batch files for commands.
Implementing the small handwritten squiggles, I've never seen any code like
that, so if you think this driver needs that, I'm really going to have to give
that more thought. Personally, doing the squiggle input (there's a better name
for it, I KNOW, but I forget it at the moment, help me out please??) is
overkill, and just implementing the function keys would be all that most users
would care to do. I mean, this isn't a PDA, right? Or is this maybe my
built-in laziness speaking? I need some senior guidance here, what would you do
in this case?
I may well be wrong here, so I would really value your opinions in this area. I
just don't have enough experience in this area to know what's right and what's
In the meantime, it seems obvious to me that I'm going to have to learn some xkb
stuff to correctly do the function-keys, allowing the correct management. heck,
I don't even like the way I'm detecting screen size, for scaling, I think I can
find better, if I really look harder.
It's not that my code does a bad visual job, it's that I'm not doing it "right",
and that offends my sense of right & wrong. I can't release code looking like
that. More time Just is increasingly hard to give, when I know that my current
driver actually works (I'm getting excited over this).
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v2.0.9 (FreeBSD)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org
-----END PGP SIGNATURE-----
More information about the xorg