Re: Ukrainian support in GNU Emacs 21.1

[Serhii Hlodin]

>>>>> "DL" == Dave Love writes:

 DL> I have already done a koi8-u environment, based on a proper
 DL> (Unicode-based, and thus complete) definition of koi8-u, to be
 DL> invoked in a UA locale.  I also have proper versions of
 DL> windows-1251 and cp866.  These shouldn't need specific changes
 DL> elsewhere, like in Gnus, with Gnus coding conversion fixed to do
 DL> the right thing.

 When I get release of GNU Emacs 21.1, I can't find KOI8-U in MULE. This
 is very strange, because I send patches to FSF for version 20.7 and to
 give one's word for me include this in release 21.x.

 Gnus use MULE configuration for MIME. I add koi8-u as default charset
 for Ukrainian users (in mm-util) and add koi8-u in
 rfc2047-charset-encoding-alist (convert to Base64).

 KOI8-U, CP1125 and windows-1251 -- this is various codepages who has
 use in Ukraine:

        koi8-u       -- Unix-like, Linux OS;
        cp1125       -- OS/2, MS Windows, MS-DOS (in place of CP866);
        windows-1251 -- MS Windows.

 KOI8-RU -- this is first name for KOI8-U from Microsoft Corp. A great
 many old mail and news programs for MS Windows use this as KOI8-U.

 CP866U -- this is first name for CP1125 in Ukraine, when this codepage
 is maked.

 Accordingly next aliases (as aliases only) must be defined as:

        (define-coding-system-alias 'koi8-ru 'koi8-u)
        (define-coding-system-alias 'cp866u 'cp1125)

 CP1125 must be defined as:

        (codepage-setup cp1125)

 DL> Are `koi8-ru' and `cp866u' common and correct aliases for koi8-u
 DL> and windows-1251, as the code in README.UK implies?  (By correct, I
 DL> mean the definitions are identical, not just overlapping.)  In GNU
 DL> recode koi8-u and koi8-ru are different, and I can't find a
 DL> reference to cp866u.

 DL> If you can tell me where to find the definition of cp1125 in terms
 DL> of Unicode, I can make that trivially.

 On 
        http://www.unicode.org

 find link to

        http://oss.software.ibm.com/icu/charset/

 On this page find link to 

 http://oss.software.ibm.com/cvs/icu/charset/data/ucm/

 and see ibm-1125_P100-2000.ucm

 DL> Why have you redefined the cyrillic-ukrainian input method rather
 DL> than keeping it or renaming it?

 No. Don't rename. Redefine only.

 For russian users I make cyrillic-ruskrainian (base -- russian chars,
 additional -- ukrainian). This is West Ukrainian specific.

 For Windows-users I make cyrillic-ukrainian-ms.

 Default input method -- cyrillic-ukrainian.

 DL> Is the new version what Ukrainian users will always want as
 DL> default, and does it conform to some standard?  [I assume someone
 DL> has asked for a copyright assignment for the input methods.]

 This pathes already includes in all Cyrillic distribution's Linux:

    ALT Linux (http:///www.altlinux.ru)
    ASPLInux (must be new release) (http://www.asplinux.ru)

 DL> Sorry if README.UK explains some of this -- the local Ukrainian
 DL> speaker isn't around at present.

 This file -- little HOW-TO for Ukrainian users.

 I'm sorry for bad English.

-- 
With best regards,
                Serhii Hlodin