Re: [PATCH/RFC v5 7/7] Careful with CRLF when using e.g. UTF-16 for working-tree-encoding

From: Torsten Bögershausen <tboegi@web.de>
Date: Jan 31, 2018, 17:28 UTC
Message-ID: <20180131172837.GA32723@tor.lan>
In-Reply-To: <10091BA4-1069-4A65-9057-CAAD87F9B55F@gmail.com>

[]

Show 8 quoted lines

> > That is a good one.
> > If you ever plan a re-roll (I don't at the moment) the *.proj extemsion
> > make much more sense in Documentation/gitattributes that *.tx
> > There no text files encoded in UTF-16 wich are called xxx.txt, but those
> > are non-ideal examples. *.proj makes good sense as an example.
> 
> OK, I'll do that. Would that fix the problem which this patch tries to address for you?
> (I would also explicitly add a paragraph to discuss line endings)

Please let me see the patch first, before I can have a comment.

But back to the more general question:

How should Git handle the line endings of UTF-16 files in the woring-tree, which are UTF-8 in the index?

There are 2 opposite opionions/user expectations here:

a) They are binary in the working tree, so git should leave the line endings
   as is. (Unless specified otherwise in the .attributes file)
b) They are text files in the index. Git will convert line endings
   if core.autocrlf is true (or the .gitattributes file specifies "-text")

My feeling is that both arguments are valid, so let's ask for opinions and thoughts of others. Erik, Junio, Johannes, Johannes, Jeff, Ramsay, everybody: What do yo think ?