Re: [Monotone-devel] slick pictures

[Nathaniel Smith]

On Thu, May 18, 2006 at 01:58:44PM -0400, Yury Polyanskiy wrote:
> On Thu, 2006-05-18 at 00:27 -0700, Nathaniel Smith wrote:
> > You need to add "-p0" to this last command.
> > 
> 
> Oh, thanks! My fault.
> 
> But even with it:
> ==============
> $ mtn diff -r 3895c20299d04 Makefile >~/tmp/p1
> $ cd ..
> $ mtn --db=/path/my.db co -r 3895c20299d04 Lm2/
> $ cd Lm2/
> $ patch -p0 <~/tmp/p1
> patching file Makefile
> Hunk #1 FAILED at 1.
> 1 out of 1 hunk FAILED -- saving rejects to file Makefile.rej

Hmm.  Could be a bug...

> ==============
> 
> Also it's impossible to generate GNU patch by
> $ mtn cat -r 3895c20299d04 Makefile | diff -u - Makefile
> 
> because ``mtn cat'' does not regard MT/monotonerc:get_linesep_conv().

Huh.  Separate issue -- could you file a bug report with enough
details to reproduce?  Thanks!

> Below are mtn (not working) patch ~/tmp/p1:
> ==============
> --- Makefile    de03d5fb694862f812dd93032eb27e8b76b3eae3
> +++ Makefile    0dbf55dcb12215a7cb72fb7a4dedbf37fc963280
> @@ -1,10 +1,10 @@
>  # Topdir makefile
>  # sets TOPDIR var
>  # $Id$
>  #
>  # NEW_PCB branch
>  
> -export VERSION := 1.4.2
> +export VERSION := 1.4.2e
>  export VERSION_IFACE := 1.3.3.2
> ==============
> 
> And this what regular diff says:
> $ diff -u Lm2/Makefile Lmega/Makefile
> --- Lm2/Makefile        2006-05-18 13:34:26.000000000 -0400
> +++ Lmega/Makefile      2006-04-12 01:02:03.000000000 -0400
> @@ -4,7 +4,7 @@
>  #
>  # NEW_PCB branch
>  
> -export VERSION := 1.4.2
> +export VERSION := 1.4.2e
>  export VERSION_IFACE := 1.3.3.2
> 
> ==============
> 
> 
> I think the problem might be that mtn adds to many context before? I
> think so because mtn's diffs work if I use patch -F100 usually.

I am suspicious about the $Id$ there.  What does the actual Makefile
that you are applying these patches to say?  If somehow the $Id$ is
expanded, then monotone's diff will fail simply because... there is a
conflict.  Possibly this means we _should_ dial down the amount of
context we include, to match 'diff', but it would at least mean that
our patches are fully valid :-).

-- Nathaniel

-- 
So let us espouse a less contested notion of truth and falsehood, even
if it is philosophically debatable (if we listen to philosophers, we
must debate everything, and there would be no end to the discussion).
  -- Serendipities, Umberto Eco