libX11: Changes to 'master'

[Pander]

On 2012-03-19 23:03, Pander wrote:
> On 2012-03-19 18:45, Julien Cristau wrote:
>> On Mon, Mar 19, 2012 at 16:32:07 +0100, Pander wrote:
>>
>>> On 2012-03-14 22:25, James Cloos wrote:
>>>> Nuts.  I thought all of those were sorted already.
>>>>
>>>> The dstroke and Dstroke lines are arguably better than the U0110 and
>>>> U0111 lines.
>>>>
>>>> Since <Multi_key> <o> <apostrophe> <a> and <A> were already there,
>>>> the <Multi_key> <o> <apostrophe> entry for ó will have to go.
>>>
>>> Double nuts about the apostrophe. The <o> <apostrophe> for ó is used
>>> much much more then the sequences starting with <o> <apostrophe> for Ǻ
>>> and ǻ, for which alternative exist.
>>>
>>> Please remove:
>>>   <Multi_key> <o> <apostrophe> <A> 	: "Ǻ"   U01FA # LATIN CAPITAL LETTER
>>> A WITH RING ABOVE AND ACUTE
>>> and remove:
>>>   <Multi_key> <o> <apostrophe> <a> 	: "ǻ"   U01FB # LATIN SMALL LETTER A
>>> WITH RING ABOVE AND ACUTE
>>>
>>> because these are also available as:
>>>   <Multi_key> <asterisk> <apostrophe> <A> 	: "Ǻ"   U01FA # LATIN CAPITAL
>>> LETTER A WITH RING ABOVE AND ACUTE
>>> and as:
>>>   <Multi_key> <asterisk> <apostrophe> <a> 	: "ǻ"   U01FB # LATIN SMALL
>>> LETTER A WITH RING ABOVE AND ACUTE
>>>
>>> And at the same time, restore:
>>>   <Multi_key> <O> <apostrophe>           : "Ó"   Oacute # LATIN CAPITAL
>>> LETTER O WITH ACUTE
>>> and restore:
>>>   <Multi_key> <o> <apostrophe>           : "ó"   oacute # LATIN SMALL
>>> LETTER O WITH ACUTE
>>>
>>> There has been many requests about mirrored availability of compose key
>>> sequences for often used diacritics with vowels in the past. This is
>>
>> Where?
> 
> Hi Julien,
> 
> At least in the gtk+/GNOME realm, that is probably also why both orders
> are supported. If you want I can get you the URLs. Apart from that,
> please see my motivation below.
> 
>>> because, depending on their your native language, people use
>>> diacritic+vowel or the reverse, vowel+diacritic.
>>>
>>> Sorry for the late reply but this will prevent many irritations.
>>> Implementing the above mentioned change will make the overall patch from
>>> downstream gtk+ a success.
>>>
>> Sorry, I don't think adding more redundant compose sequences qualifies
>> as a success.  Changing existing compose sequences to add different
>> redundant ones even less so.
> 
> I agree complete with you by not adding the redunant codes for ¤¶§ etc.
> That were indeed too many for simply the same rarely used character and
> gtk-lookaside should drop these as well.
> 
> However for ó I beg to differ. The vowels aeiou and diacritics `´'~^¯¨
> are completely available in both orders. Because the are used so often,
> missing out on <o> <apostrophe> will break the consistency in this
> series and decrease user experience  and expectency of the system.
> 
> Ǻ has alternative codes and is so reaerly used, it is even hard to find
> examples of usage, except for side effect of wrong encoding. :S Can you
> find any? Also <o> <apostrophe> → ó was (until last week) an existing
> compose sequence, albeit conflicting with Ǻ, only, in my perspective,
> the wrong compose sequence was eliminated.

Via the link below you will find an illustrated overview to support my
point in the pink square. Additional requests would be in the red squares.

http://img40.imageshack.us/img40/2217/mirroredvowelseries.png

Other points of interest, but no that effect me personally, are the blue
squares on other missing sequences and orange squares on inconsistent
sequences.

Note the arrow in the corners of the overview which indicate the order
of the lookup of sequences.

> Perhaps some other members of the community would like to support this
> cause or comment on this?
> 
>> Cheers,
>> Julien
> 
> Best regards,
> 
> Pander
> _______________________________________________
> xorg-devel at lists.x.org: X.Org development
> Archives: http://lists.x.org/archives/xorg-devel
> Info: http://lists.x.org/mailman/listinfo/xorg-devel