Native character set support

[Jeremy Allison]

Andrej Borsenkow wrote:

> What are plans (_are_ there any plans) to add support for native charsets to
> other SAMBA parts as well? Off the head, at least following can be non-ASCII
> 
>   - anything related to user management (names, comments and oh God!
> passwords)
>   - server comments
>   - PC names (yes, it has all sorts of implications, but Windows runs with
> it nicely)
>   - share names? Well, as long as share is just a directory, why not?
> 

I have been thinking a lot about I18N issues for a while.

My current plan is to add UNICODE throughout the internals of Samba
by using UTF8, translating to microsoft unicode (little endian
16 bit unicode) and microsoft code page support (multi-byte character
support) on entry and exit to the network.

The only issue with doing this is that many UNIXs have
native multi-byte character support (EUC for Japanese
etc.) and UTF8 is not compatible with these.

However, adding a "UTF8 <-> EUC" conversion layer (or a similar
one for any other multi-byte character sets) will probably
touch many areas of the system unfortunately.

> Of all these, user management is absolute must if domain support in SAMBA is
> going to be actively used outside of ASCII speaking world. I have some
> ideas, but I would like to ask about current state of art first ...

Yes - this is the problem with our current domain support,
it will not work with asian character sets or non US-ASCI
usernames.

Jeremy.

-- 
--------------------------------------------------------
Buying an operating system without source is like buying
a self-assembly Space Shuttle with no instructions.
--------------------------------------------------------