New draft submitted of 3066bis...

[Chris Lilley]

On Tuesday, November 2, 2004, 7:20:43 AM, Martin wrote:

MD> At 03:23 04/09/11, Addison Phillips [wM] wrote:
 >>http://www.inter-locale.com/ID/draft-phillips-langtags-07.html

MD> This is a short comment on the above draft, based on a thread
MD> that recently came up on www-tag at w3.org. (see
MD> http://www.w3.org/mid/417F7D2B.6000109@metalab.unc.edu and
MD> http://www.w3.org/mid/p06110407bda45535ab9b@%5B192.168.254.47%5D).

MD> Elliotte Harold
MD> mentioned that it may not be totally clear what kind of
MD> case equivalence is meant to be used for language tags.
MD> It would be good if the draft could explicitly say that
MD> this is case equivalence on US-ASCII in the straightforward
MD> way such that such as match a-z with A-Z.


MD> There are some other side tracks in that draft, such as
MD> the question of how non-US-ASCII language tags are case
MD> equivalent or not, which I think can safely be ignored.

Even better if it could distinguish between three cases and say which
applies:

a) all the data is un US-ASCII, so case folding is on US-ASCII only
b) the data allows characters beyond US-ASCII, Unicode case folding
rules are used
c) the data allows characters beyond US-ASCII, ASCII parts are case
folded and non-ASCII parts are not.

If option c) applies, and i hope t does not, please indicate any impact
of locale settings (eg case folding of I in a turkish locale, etc)


Actually it would be good to clarify which of these applies for 3066 as
well; it seems abundantly clear to me that a) applies but this has been
disputed.


-- 
 Chris Lilley                    mailto:chris at w3.org
 Chair, W3C SVG Working Group
 Member, W3C Technical Architecture Group