New manufacturing data flags for keyboards (2nd draft).

[Bernardo Innocenti]

(adding devel@ to cc)

Mitch Bradley wrote:
> Jim Gettys wrote:
>> What are the olpc vs. olpc2 variants?
> 
> The way that I understand it based on Bernie's IRC explanation, is that 
> these are different usage models based on how you interpret one of the keys.
> 
> It is not a property of the keyboard per se, so I question whether this 
> is something that should be in manufacturing data.

As you say, it doesn't belong to manifacturing data.

BUT... do we want keyboard autoconfiguration?  If so, I don't know where
else we could hardcode this bit of knowledge.

Unless we admit shipping forked versions of hal and xkeyboard-config
with a table for mapping manifacturing data to the actual XKB parameters.


>> On Sat, 2007-10-06 at 15:25 -0400, Bernardo Innocenti wrote:
>>   
>>> Jim Gettys wrote:
>>>
>>>     
>>>> Bernie, please fill in the table below for the keyboard layout for each
>>>> keyboard type we have ASAP, so we can get this to Quanta and I can get
>>>> this all into the Wiki.
>>>>       
>>> Based on the olpc-configure script, plus our recent additions:
>>>
>>>  Keyboard Layout          KM       KL        KV
>>>                           olpc     us        olpc
>>>  OLPC_Argentina_Keyboard  olpc     es        olpc
>>>  OLPC_Brasil_Keyboard     olpc     br        olpc
>>>  OLPC_Ethiopia_Keyboard   olpc     us,et     olpc2,olpc
>>>  OLPC_Libya_Keyboard      ????     ????      ????
>>>  OLPC_Nigeria_Keyboard    olpc     ng        olpc
>>>  OLPC_Rwanda_Keyboard     ????     ????      ????
>>>  ????                     olpc     us,ara    olpc2,olpc
>>>  OLPC_Thailand_Keyboard   olpc     us,th     olpc2,olpc
>>>  Urdu_Keyboard            olpc     us,ur     olpc2,olpc
>>>  Cyrillic_Keyboard        olpc     us,ru     olpc2,olpc
>>>  OLPC_Turkey_Keyboard     olpc     us,tr     olpc2,olpc
>>>  OLPC_Nepal_Keyboard      olpc     us,np     olpc2,olpc  (UNSURE)
>>>
>>>
>>> For Libya and Rwanda I have no idea.  For Nepal, please double check.
>>> I also don't know who gets to use the arabic keymap.
>>>
>>>     
>>>> The KV flag is included for completeness: it should normally be a null
>>>> string as Xkb has a notion of default variant.
>>>>       
>>> Yes, it does.  But the default in xkeyboard-config is rarely ours.
>>> And it would be a pity if we couldn't use the pristine upstream package.
>>>
>>> I have xkb rules to deduce the variant from KM+KL, but olpc2 need
>>> to be specified manually.

-- 
 \___/
 |___|  Bernardo Innocenti - http://www.codewiz.org/
  \___\ One Laptop Per Child - http://www.laptop.org/