As the perl folks are about to release another version with yet more
major unicode corrections (apparently they have had to redo alot of
namespace and unicode stuff in every major version since 5.6 due to the
changing of the guard (amongst other reasons).
But perl has the ability in the new version to specify how chars shall
be handled in a search, for example --
locale specific, unicode specific, ascii rules(C I think), and a
'd'epends mode that handles most things like Unicode, except for bytes
between 128-255, where they will default to staying with Latin1 (as I
understand it -- they aren't especially forth coming with explanations).
I tried to ask, under what circumstances would I get the collating order
that bash now defaults to in the en_US locale -- and was told NONE. I
asked them why -- how did they justify that behavior -- what documents
were they following and was told that to ask for technical
justifications was rude, with no one volunteering an answer. Sounds a
bit sketchy to me... like they have no reasons....but that could be a
projection.
I told them I thought that them taking such technical questions as
'rudeness' was their personal trip, and not related to the engineering
task(s) at hand. A few understood, a few other loud ones joined in
with those who had no answer, and said asking for any sort of reason or
justification as to why perl is the way it is, argumentative and
rude...etc....blah blah blah... (bunch of self-absorbed prima donna's -
worse than me on a bad day!)...
Anyway... so WHY does bash collate this way? Under what rules is bash
operating? I.e. justification?
If other programs claim to have locale specific sorting & character
collation, should they be sorting the same way?
So far, i've seen Microsoft do the a<A<z<Z collation, in win7 and above,
BUT.. I hardly think Bash did it to follow windows.... So.. where did
this come from and how should it apply to other "locale honoring" programs?
Thanks!
Linda
- locale specific ordering in EN_US -- why is a<A<b... Linda Walsh
-
