Re: [Monotone-devel] Re: Bug in CRLF conversions

[Richard Levitte - VMS Whacker]

In message <address@hidden> on Mon, 30 Jan 2006 12:31:42 +0100, Zbynek Winkler 
<address@hidden> said:

zwin> Richard Levitte - VMS Whacker wrote:
zwin> 
zwin> >monotone> However, the basic idea seems sound: distinguish
zwin> >monotone> binary and text, and for text, assume (by default)
zwin> >monotone> native line-endings.
zwin> >
zwin> >Or any line ending.  It doesn't really matter because for text
zwin> >files, the line ending isn't really interesting in itself, it's
zwin> >just a layout indicator, saying "break the line here".  (I
zwin> >realised this yesterday, while discussing this issue)
zwin> >
zwin> Let me disagree. An example where this is not true is python. It
zwin> is not layout indicator, it says "this is the end of statement".

That's a really bad example.  Python uses the common property of a
text file that there is an end to each line.  Python makes no other
requirement or just wouldn't work on Windows, where at least half if
not all the python scripts have lines that end with CRLF rather than
LF.

zwin> I think I like more the way mercurial does this - it does not
zwin> convert anything by itself. But it allows you to specify
zwin> "filters" (in/out) that are called to process the files on
zwin> commit or checkout. I think it follows the principle of least
zwin> surprise (by default *do not play with user data*).

I think it's time I take a look at hg to see how it does things.  I
hope it won't make me scream and hide...  :-)

Cheers,
Richard

-----
Please consider sponsoring my work on free software.
See http://www.free.lp.se/sponsoring.html for details.

-- 
Richard Levitte                         address@hidden
                                        http://richard.levitte.org/

"When I became a man I put away childish things, including
 the fear of childishness and the desire to be very grown up."
                                                -- C.S. Lewis