Karl Fogel <kfogel@newton.ch.collab.net> writes:
> Blair Zajac <blair@orcaware.com> writes:
> > Are there going to be any interoperability issues between an svn
> > client and server where one has ut8 enabled and the other doesn't?
>
> Nope. The whole reason we're defining UTF-8 as the standard for
> certain kinds of data in Subversion is to avoid such interop problems.
>
> It might be possible to make such problems happen at the moment, while
> there is such a thing as the --enable-utf8 compile-time option, but
> that is a temporary artifact of development, not a long-term problem.
Actually, there will be no munging even if one of client/server is
compiled with --enable-utf8 and the other with --disable-utf8 as long
as all data subject to translation is in US-ASCII. The one compiled
with --disable-utf8 will enforce that no eigthbit characters or
ISO2022 escapes are used, so as long as you're not trying to use
EBCDIC or ISO646-SE (or, heaven forbid, UTF-7) or suchlike which
assigns a different meaning to any of the US-ASCII codepoints, any
offending character should be caught before it can do any damage.
// Marcus
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
Received on Thu Jul 11 22:10:56 2002