> USEUNICODEFILENAMES NO (The Default)
> TSM won't backup Japanese filenames unless you
> run the client on a Japanese NT server. Adding
> the Japanese codepage to a UK build won't work.

> USEUNICODEFILENAMES YES
> TSM will backup Japanese filenames most of the
> time, but occassionally dies a horrible death
> on certain filenames. With lots of people with
> lots of Japanese filenames, the client will fail
> more than it suceeeds.

> USEUNICODEFILENAMES is suppose to be used with
> Macintosh files, any other use is convenently
> unsupported by Tivoli.

This is not a matter of "convenience". USEUNICODEFILENAMES
was *never* intended to provide the support you are seeking.
It's purpose is strictly for support of Macintosh volumes on
NTFS file systems (granted, though, the option would have
been better named ENABLEMACFILESUPPORT, or something along
those lines).

> Tivoli support (especially, the Japanese end) are
> no help whatsoever. As a company, Tivoli seem to
> have no concept that people do business in multiple
> languages, and as a backup product Tivoli need to
> support it. Recents failures, where the NT server
> product failed as soon as you run it under Japanese,
> German etc.. bears testiment to the fact, Tivoli
> aren't even testing their products under different
> languages.

IBM/Tivoli fully understands the need for a global perspective,
as we are a global company doing business all over the world.
Please do not equate language support issues with "have no
concept that people do business in multiple languages". This is
just not true.

Agreed, the recent problems we have had with non-English
character sets do not instill the greatest confidence in
our NLS support. However, we do in fact have a large number
of resources dedicated to NLS support, translation, and
testing, and we are continuously working to improve our
processes to (among other things) eliminate the kinds of
problems you mention. Yes, we have stumbled in this arena,
especially recently, but we have also made every effort to
respond to the problems in as timely a fashion as possible,
because we *do* understand the need for this support.

Regarding support for file names comprised of characters from
different character sets (i.e. Japanese file names on English
systems), this is a long-standing requirement. It isn't here
yet because we are ignoring it; rather, the implementation is
not trivial. But it is something that we are actively working
on and hope to deliver this year. (Standard caveat: this does
not constitute a formal announcement or commitment.)

Regards,

Andy

Andy Raibeck
IBM Tivoli Systems
Tivoli Storage Manager Client Development
e-mail: araibeck AT tivoli DOT com
"The only dumb question is the one that goes unasked."
"The command line is your friend"