Testiness.

[Ihno Krumreich]

On Wed, Aug 31, 2005 at 04:16:44PM +0200, Ihno Krumreich wrote:
> On Wed, Aug 31, 2005 at 02:29:08AM +0200, Wolfgang Denk wrote:
> > In message <200508301909.03225.rob at landley.net> you wrote:
> > >
> > > Is this ubuntu brokenness, gnu brokenness, some standards body that 
> > > collectively dropped acid, or a simple mistake nobody's bothered to correct 
> > > because they think it's intentional?
> > 
> > I see the same in several current Linux distros (Ubuntu, Fedora  Core
> > 2/3/4, SuSE 9.2/3...) when using such locale settings.
> > 
> > For example:
> > 
> > Fedora Core 2:
> > 
> > 	$ cat /etc/issue
> > 	Fedora Core release 2 (Tettnang)
> > 	Kernel \r on an \m
> > 
> > 	$ echo [a-z]*
> > 	config.mk CVS flash.c load_sernum_ethaddr.c Makefile tqm8xx.c u-boot.lds u-boot.lds.debug
> > 
> > [mind the "CVS' and "Makefile" entries which definitely do NOT  match
> > "[a-z]*' - at least in my understanding.]
> > 
> > Fedora Core 4:
> > 
> > 	$ cat /etc/issue
> > 	Fedora Core release 4 (Stentz)
> > 	Kernel \r on an \m
> > 
> > 	$ echo [a-z]*
> > 	config.mk CVS flash.c load_sernum_ethaddr.c Makefile tqm8xx.c u-boot.lds u-boot.lds.debug
> > 	$ echo $LANG
> > 	en_US.UTF-8
> > 	$ LANG=C
> > 	$ echo [a-z]*
> > 	config.mk flash.c load_sernum_ethaddr.c tqm8xx.c u-boot.lds u-boot.lds.debug
> > 
> > SuSE 9.2:
> > 
> > 	$ cat /etc/issue
> > 
> > 	Welcome to SuSE Linux 9.2 (i586) - Kernel \r (\l).
> > 
> > 
> > 	$ echo [a-z]*
> > 	config.mk CVS flash.c load_sernum_ethaddr.c Makefile tqm8xx.c u-boot.lds u-boot.lds.debug
> > 	$ echo $LANG
> > 	en_US.UTF-8
> > 	$ LANG=C
> > 	$ echo [a-z]*
> > 	config.mk flash.c load_sernum_ethaddr.c tqm8xx.c u-boot.lds u-boot.lds.debug
> > 
> > > This is clearly broken behavior, I'm wondering where the brokenness 
> > > originated.  (I'm guessing Microsoft might have been involved if it's a 
> > > standards body.  After all, MS filesystems are case insensitive...)
> > 
> > Looks broken to me, too. But it seems very few  people  even  notice,
> > and even fewer care.
> 
> Depends on what you want to have. If the collating sequence for en_US.UTF-8
> is according to the dictionary (like it is for the german UTF-8). Then
> everything works perfect.
> 
> What all the above example have in mind is the American ASCII code
> and the collating sequence according to the numeric value.
> If you want to have this you have to set LANG=C.
> 
> So its not a bug its a feature.
> 

Just an example of the diffirences in collating sequences (I personally hate it
to see the dot-files mixed into the normal file names):

ihno at s390z08:~/tmp/collate> export LC_ALL=en_US.UTF-8
ihno at s390z08:~/tmp/collate> ls -la
total 3
drwxr-xr-x  2 ihno suse  192 2005-08-31 16:27 .
drwxr-xr-x  4 ihno suse 3528 2005-08-31 16:24 ..
-rw-r--r--  1 ihno suse    0 2005-08-31 16:24 .kde
-rw-r--r--  1 ihno suse    0 2005-08-31 16:25 mail
-rw-r--r--  1 ihno suse    0 2005-08-31 16:24 Mail
-rw-r--r--  1 ihno suse    0 2005-08-31 16:27 Ma-il
-rw-r--r--  1 ihno suse    0 2005-08-31 16:27 Ma.il
-rw-r--r--  1 ihno suse    0 2005-08-31 16:24 .mcop
ihno at s390z08:~/tmp/collate> export LC_ALL=C
ihno at s390z08:~/tmp/collate> ls -la
total 3
drwxr-xr-x  2 ihno suse  192 Aug 31 16:27 .
drwxr-xr-x  4 ihno suse 3528 Aug 31 16:24 ..
-rw-r--r--  1 ihno suse    0 Aug 31 16:24 .kde
-rw-r--r--  1 ihno suse    0 Aug 31 16:24 .mcop
-rw-r--r--  1 ihno suse    0 Aug 31 16:27 Ma-il
-rw-r--r--  1 ihno suse    0 Aug 31 16:27 Ma.il
-rw-r--r--  1 ihno suse    0 Aug 31 16:24 Mail
-rw-r--r--  1 ihno suse    0 Aug 31 16:25 mail
ihno at s390z08:~/tmp/collate> cat /etc/issue

Welcome to SUSE LINUX Enterprise Server 9 (s390x) - Kernel \r (\l).


ihno at s390z08:~/tmp/collate> 

Ihno


"Never trust a computer you can lift."
--
Ihno Krumreich            ihno at suse.de
SUSE LINUX Products GmbH  Projectmanager S390 & zSeries
Maxfeldstr. 5             +49-911-74053-439
D-90409 Nürnberg          http://www.suse.de