On Sun, Jan 30, 2011 at 3:06 PM, Peter Hutterer
<peter.hutte...@who-t.net> wrote:
> On Fri, Jan 28, 2011 at 10:49:04AM -0800, Ping Cheng wrote:
>> On Thu, Jan 27, 2011 at 8:57 PM, Peter Hutterer
>> <peter.hutte...@who-t.net> wrote:
>> > On Wed, Jan 26, 2011 at 10:20:39AM -0800, Ping Cheng wrote:
>> >> This change makes some of the lines in WacomModelDesc a bit too
>> >> long. Since we plan to move the table to the kernel driver soon,
>> >> I updated the table without updating the table format. It is only
>> >> to keep the code consistent between wcmUSB and wcmISDV4 for now.
>> >>
>> >> Once we move USB resolutions to the kernel, the defines can be
>> >> moved to wcmISDV4.c.
>> >>
>> >> Signed-off-by: Ping Cheng <pingli...@gmail.com>
>> >> ---
>> >>  src/wcmISDV4.c      |    6 +-
>> >>  src/wcmUSB.c        |  242 
>> >> +++++++++++++++++++++++++-------------------------
>> >>  src/xf86WacomDefs.h |   12 +++
>> >>  3 files changed, 136 insertions(+), 124 deletions(-)
>> >>
>> >
>> > [...]
>> >
>> >> diff --git a/src/xf86WacomDefs.h b/src/xf86WacomDefs.h
>> >> index 91adf72..7aa049d 100644
>> >> --- a/src/xf86WacomDefs.h
>> >> +++ b/src/xf86WacomDefs.h
>> >> @@ -45,6 +45,18 @@
>> >>  #define MIN_PAD_RING 0               /* I4 absolute scroll ring min 
>> >> value */
>> >>  #define MAX_PAD_RING 71              /* I4 absolute scroll ring max 
>> >> value */
>> >>
>> >> +/* Supported default resolutions in points/m */
>> >> +#define RES_100000 100000
>> >> +#define RES_200000 200000
>> >> +#define RES_80000   80000
>> >> +#define RES_50000   50000
>> >> +#define RES_44173   44173
>> >> +#define RES_39842   39842
>> >> +#define RES_39370   39370
>> >> +#define RES_36772   36772
>> >> +#define RES_20000   20000
>> >> +#define RES_10000   10000
>> >> +
>> >
>> > I don't think this is a good idea. defines are used for mainly two reasons:
>> > - improve readability of the code
>> >  x = MAX_X_VALUE is more expressive than x = 10000
>> > - ease refacturing, providing a single place to change the value without
>> >  changing the meaning.
>> >
>> > with the defines above, neither is realy met. RES_80000 is about as
>> > explanatory as 8000 but even worse: if we change this define we can never
>> > change the value without also changing every occurance of the macro.
>> > think of a how confusing the line below would be
>> >    #define RES_80000  40000
>> >
>> > so either we just leave the code as-is, or we find meaningful names for the
>> > defines. RES_INTUOS4 for example, or something like this if applicable.
>>
>> I thought about model or feature related names as well. But those
>> numbers do not have clean links to either the model or feature. Once
>> we made sure the actual numbers are right, I am sure someone will come
>> up with meaningful names for them.
>
> if you with access do the docs can't seem to find a decent grouping, I doubt
> we will anytime soon. The model table currently exists for that reason, we
> have a huge number of models that are slightly different and thus need
> different treatment. the defines above don't really help with that.
>
> besides, having a look at the patch again I noticed that this doesn't just
> change all 100000 into RES_100000, it also changes the values of a few
> tablets around (0x9F for example). this is in itself a showstopper for the
> patch, these changes must be in a separate patch.
>
> either way, I'm happy to merge a patch that changes those that are currently
> wrong, but I won't merge the define RES_100000 part.

Please totally ignore this patch. Those two entries will be retrieved
from the kernel since they are touch only devices. The default value
doesn't count anything.

Ping

------------------------------------------------------------------------------
Special Offer-- Download ArcSight Logger for FREE (a $49 USD value)!
Finally, a world-class log management solution at an even better price-free!
Download using promo code Free_Logger_4_Dev2Dev. Offer expires 
February 28th, so secure your free ArcSight Logger TODAY! 
http://p.sf.net/sfu/arcsight-sfd2d
_______________________________________________
Linuxwacom-devel mailing list
Linuxwacom-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/linuxwacom-devel

Reply via email to