Re: [Monotone-devel] Re: line endings with 0.31

[hendrik]

On Tue, Nov 21, 2006 at 03:58:58PM +0100, Richard Levitte - VMS Whacker wrote:
> In message <address@hidden> on Tue, 21 Nov 2006 09:48:54 -0500, 
> address@hidden said:
> 
> hendrik> In monotone, I suggest that a file that has been
> hendrik> character-converted on checkout have its line-end coding
> hendrik> reverted on checkin, on a line-by-line basis.  Thus only when
> hendrik> the user explicitly edits line ends will the end-of-line
> hendrik> coding be changed in the repository.  This would have the
> hendrik> effect that if massive damage is done to a true binary file
> hendrik> if it is mistakenly line-end-converted, the damage would be
> hendrik> mostly undone on subsequent checkin.
> 
> What I hear you say is that eol conversion should happen upon file
> extraction.  Then, unless the user explicitely does eol conversion,
> the file is converted back to the original eol format before being
> stored (or the diff being performed).

I had on mind something like this:

eol conversion happens on file checkout.

Upon checkin:

diff ignores the detailed end-of-line coding, except of course to 
recognise all end-of-lines as being end-of-lines.  Then after the 
regular diff, has been performed,
  for each line
  do
    if it exists only in the new version,
    then we use the new encoding.
    elsif it exists in both the original and changed versions
      if the user has modified the coding
      then we use the user's coding
      else we use the iriginal coding
      endif
    else if the line exists only in the original
    then no problem, it's gone.
    endif
  endfor

> 
> I'm sure I misunderstood what you wrote, because my interpretation
> doesn't make sense at all.

By the way, *does* monotone use a line-by-line diff?
I thought it used a character-by-character diff based on compression 
technology.

-- hendrik