What is your default environment locale?
Submitted by matej on Tue 13 Jun 2006
| I don't use locales at all |
   9% | 54 votes |
| en_* |
13% | 76 votes |
| en_* with UTF-8 |
29% | 172 votes |
| other than en_* |
19% | 112 votes |
| other than en_* with UTF-8 |
27% | 160 votes |
| Total 574 votes |
Posted by Anonymous (85.99.xx.xx) on Wed 14 Jun 2006 at 09:29
tr_TR.UTF-8 though sometimes when Turkish case conversion becomes a problem* I set LC_CTYPE to en_US.UTF-8
* I think everybody knows that... In Turkish "I" when converted to lowercase becomes ı (dotless i) and when converted to uppercase "i" becomes "İ" (I with dot above). So when you use locale-dependent case conversion inyour code to lower/upper case some keywords, all of the app blows. (imap becomes İMAP for example) Even most successful applications like Samba or PostgreSQL is effected by that bug. For more information:
http://www-306.ibm.com/software/globalization/facts/archive.jsp
http://en.wikipedia.org/wiki/Turkish_dotted_and_dotless_I
http://www.i18nguy.com/unicode/turkish-i18n.html
PS: Why did I explained it here? Because the problem is mostly about developers not knowing that issue about Turkish locales.
* I think everybody knows that... In Turkish "I" when converted to lowercase becomes ı (dotless i) and when converted to uppercase "i" becomes "İ" (I with dot above). So when you use locale-dependent case conversion inyour code to lower/upper case some keywords, all of the app blows. (imap becomes İMAP for example) Even most successful applications like Samba or PostgreSQL is effected by that bug. For more information:
http://www-306.ibm.com/software/globalization/facts/archive.jsp
http://en.wikipedia.org/wiki/Turkish_dotted_and_dotless_I
http://www.i18nguy.com/unicode/turkish-i18n.html
PS: Why did I explained it here? Because the problem is mostly about developers not knowing that issue about Turkish locales.
[ Parent ]
yes, I do understand :)
but I was of course, unaware of this. as I'm unaware of many other problems in other locales - east asia etc, far, far away from me.
does applications you use and fail to convert case use libc correctly? using toupper vs towupper or tolower vs towlower is often overseen. many apps are still "thinking-english"-only. I guess you'd have to complain to authors...
but I was of course, unaware of this. as I'm unaware of many other problems in other locales - east asia etc, far, far away from me.
does applications you use and fail to convert case use libc correctly? using toupper vs towupper or tolower vs towlower is often overseen. many apps are still "thinking-english"-only. I guess you'd have to complain to authors...
[ Parent ]
Posted by Anonymous (82.82.xx.xx) on Thu 15 Jun 2006 at 23:16
I fixed that bug in Samba some time ago (3.0.9 or so...), recent versions whould work without a problem, less had this bug, recent versions should be fixed, too now. What other programs do you know which are affected? Please send mail to bjoern[at]j3e[dot]de. Thanks ...
[ Parent ]
en_* with UTF-8.
[ Parent ]
root .. default (C)
normal user account .. cs_CZ iso-8859-2
normal user account .. cs_CZ iso-8859-2
[ Parent ]
Posted by Anonymous (213.224.xx.xx) on Thu 15 Jun 2006 at 06:33
Hmmm, it looks like Debian uses "en_*"by default, I better start searching how to make it use the UTF_8 version :(.
[ Parent ]
Posted by Anonymous (24.182.xx.xx) on Sun 18 Jun 2006 at 18:18
Is easy to do, but a lot of stuff doesn't work as expected yet, at least in some X terminals. Sarge is mostly usable with UTF_8 although in many cases it doesn't do the right thing, but Etch seems to vary in usability sort of at-random.
Is frustrating, because I regularly work in OSX as well as Debian, and I keep forgetting whether I'm in Debian(Gotta watch out for UTF problems) or OSX (More limited terminal environment, need to use slightly different arguments to rsync, ls, etc)
Of course matters aren't helped by the fact that I use the same bashrc in both places, and define functions or aliases to emulate missing functionality in either environment...
Is frustrating, because I regularly work in OSX as well as Debian, and I keep forgetting whether I'm in Debian(Gotta watch out for UTF problems) or OSX (More limited terminal environment, need to use slightly different arguments to rsync, ls, etc)
Of course matters aren't helped by the fact that I use the same bashrc in both places, and define functions or aliases to emulate missing functionality in either environment...
[ Parent ]
Where's the "I don't know" option?
[ Parent ]
Posted by Steve (62.30.xx.xx) on Thu 15 Jun 2006 at 16:10
[ Send Message | View Steve's Scratchpad | View Weblogs ]
[ Send Message | View Steve's Scratchpad | View Weblogs ]
[ Parent ]
LANG=lt_LT.UTF-8
LC_MESSAGES=en_US.UTF-8
I hate seeing messages, menus etc. in my native language :)
LC_MESSAGES=en_US.UTF-8
I hate seeing messages, menus etc. in my native language :)
[ Parent ]
Posted by Anonymous (222.225.xx.xx) on Sat 17 Jun 2006 at 16:20
* en_US.UTF-8 for most of programs under X
* some tereminals running under all sorts of traditional locales: ja_JP.eucJP, pt_BR, ... (to edit old files under old encoding system.) -- this is m17n-env package trick provided through menu system.
I wish Debian kept Linux console to be C locale. (The current my system shows en_US.UTF-8 which must be Debian system default these days.)
osamu at debian.org
* some tereminals running under all sorts of traditional locales: ja_JP.eucJP, pt_BR, ... (to edit old files under old encoding system.) -- this is m17n-env package trick provided through menu system.
I wish Debian kept Linux console to be C locale. (The current my system shows en_US.UTF-8 which must be Debian system default these days.)
osamu at debian.org
[ Parent ]
$ locale
LANG=en_GB
LANGUAGE=en_GB:en_US:en
LC_CTYPE="en_GB"
LC_NUMERIC="en_GB"
LC_TIME="en_GB"
LC_COLLATE=C
LC_MONETARY=nl_NL@euro
LC_MESSAGES="en_GB"
LC_PAPER=nl_NL
LC_NAME=nl_NL
LC_ADDRESS=nl_NL
LC_TELEPHONE=nl_NL
LC_MEASUREMENT=nl_NL
LC_IDENTIFICATION="en_GB"
LC_ALL=
- Felix
LANG=en_GB
LANGUAGE=en_GB:en_US:en
LC_CTYPE="en_GB"
LC_NUMERIC="en_GB"
LC_TIME="en_GB"
LC_COLLATE=C
LC_MONETARY=nl_NL@euro
LC_MESSAGES="en_GB"
LC_PAPER=nl_NL
LC_NAME=nl_NL
LC_ADDRESS=nl_NL
LC_TELEPHONE=nl_NL
LC_MEASUREMENT=nl_NL
LC_IDENTIFICATION="en_GB"
LC_ALL=
- Felix
[ Parent ]
Posted by Anonymous (213.216.xx.xx) on Mon 19 Jun 2006 at 11:56
Locale is not a single tag! It contains both actual locale and language information. Besides, C/POSIX would have been a neccessary option in this poll.
[ Parent ]