Re: Ukrainian locale

["Volodymyr M. Lisivka" <lvm at mystery dot lviv dot net>]

В нд, 15-чер-03 14:31 Petter Reinholdtsen писали:
> [Volodymyr M. Lisivka]
> > Fine, but I am change my employer and _very_ busy now, sorry.
>
> OK.  I'm in no hurry to get the locale included in glibc, so we can
> continue when you have more time.
Ok, but current locale has bugs in collation order for Ukrainian letters.
Can You update at least LC_COLLATION section to fix bugs?


> There is also a licensing issue.  Your locale contains the following
> copyright comment:
>
>   % Copyright (C) 2003 Volodymyr M. Lisivka <v_lisivka@sourceforge.net>
>   %
>   % Distribution and use is free, also for commercial purposes
>   % if this copyright notice are preserved.
>
> I am not sure if this is compatible with the GNU GPL.  Did you intend
> it to be?

What I want:
  - locale definition must be free for all (even for non-GPL OS'es and
systems);
  - information about author must be preserved in source file (I want
receive any bugreports and questions about current Ukrainian locale).

AFAIK these requirements is compatible with GNU GPL v2.
But if not, I can remove all incompatibles from copyright comment. just say,
what I must change.


> It might also be required for you to sigh some legal
> paperwork to assign the copyright over to the Free Software Fundation.
> Is this OK with you?
Yes, but my previous letter was found in 1.5 year after arriving to FSF. ;)

> Then some comments on the locale itself.
>
>  - I've been told by the glibc maintainers that the <U#> notation
>    should be used for ASCII values as well as non-ASCII values.  The
>    locale uses ASCII texts several places.  This should probably be
>    changed.
Is there any tool to do that?


>  - The glibc locales seem to contain some information in comments at
>    the top of the file.  Your file is missing these.  The included
>    patch adds these headers.
Where I can find documentation about format and purpose of this comment?


> +% Charset: KOI8-U
What this string mean? (Default encoding for locale?)


>  - I believe the full name of the language and region should be used
>    in LC_IDENTIFICATION.  Your locale uses the codes.  This patch
>    changes this.
>
>  - The category in LC_IDENTIFICATION should probably use uk_UA instead
>    of il18n to match the other locales in glibc.  This patch fixes it.
>
>  - Why do you use posix:1993 for LC_NUMERIC?
Please read ISO TR14652 about LC_IDENTIFICATION section.

>  - The locale uses your sourceforge address.  I do not think it is
>    working.
Yes, You right. Correct address - "v_lisivka@users.sourceforge.net", sorry.


See newer version in atachment.


--
Best regards,         mailto:lvm@mystery.lviv.net
Volodymyr M. Lisivka  ICQ#14549856

Attachment: uk_UA-2.0
Description: Text document