line endings

[Andre Haupt]

On Thu, Jan 31, 2008 at 11:47:53AM +0200, Alexander Belchenko wrote:
> Stuart McGraw пишет:
> >Marcus Sundman wrote:
> >>Stuart McGraw <smcg4191 at frii.com> wrote:
> >>>So my question is, how does one use Bazaar so that
> >>>branches created on a Windows box will have the right
> >>>line endings?  Is there some process I overlooked?
> >>>A plugin?  A command option I missed?  Some third party
> >>>add-in somewhere?  Does Bazaar development have any
> >>>plans regarding this?
> >>
> >>Well, before changing line breaks one has to know the encoding of the
> >>file (or otherwise it might break completely), 
> >
> >How many encodings exist were this is actually a
> >problem?  And how often are they used?  How often
> >does the confluence of using such an encoding, and
> >requiring auto l.e. conversion actually occur in
> >the real world?  Its not a problem in the encodings
> >I have to deal with (the common 1-byte iso encodings,
> >utf8, iso-2022-jp, sjis).  I have not heard any
> >screams of outrage from the TortoiseCVS folks.
> >How about from the Mercurial folks who also have
> >this?
> 
> If you're using 1-byte encodings (including utf-8)
> the problem with line-endings pretty simple.
> It's always  \r\n or \n (CRLF or LF).
> 
> But for 2-bytes unicode encodings like UTF-16
> (and I think it's true for 4-bytes UTF-32 as well)
> line-endings becomes more complex, i.e. for UTF16-LE
> 
> \r\0\n\0 and \n\0 (CRLF or LF).
> 
> For UTF16-BE of course byte order will be different.
> 
> So for UTF-16 conversion CRLF <-> LF should be slightly
> different from 1-byte texts.
> And this conversion should be blazingly fast, of course.
> 
I really think a VCS should track the content of a file and must not
change its content in whatsoever way. IMHO its not the job of a VCS to
take care of such things like line endings.

My 0.02 Euros


Best regards,

Andre