TDiff with unicode files without BOM

Hi all,
I have a file in a SVN repository that is unicode encoded, but misses
the BOM, the initial two bytes (FF FE) to make editors detect the
endianess of the file.
Using TDiff between versions resuts in strange behavior.
Characters are displayed with an interleaving blank like:
h e l l o
But the difference coloring is normally at the correct line but at half
the horizontal position, as if the position is somewhat counted
correctly in unicode-16 chars.
I can confirm that if the file do have the BOM the TDiff behavior is
correct.

Since I have other diff tools that looks like they correctly interprets
the difference among versions, is this to be considered a possible bug?

Do you have had similar experience?
Is it normal to have unicode-16 files without BOM?

Thanks for any help.
Mau.

------------------------------------------------------
http://tortoisesvn.tigris.org/ds/viewMessage.do?dsForumId=4061&dsMessageId=3060833

To unsubscribe from this discussion, e-mail: [users-unsubscribe_at_tortoisesvn.tigris.org].
Received on 2013-07-19 14:59:37 CEST