Please attach all examples for this behaviour with a small explanation of the
desired output or the diff produced by other SCMs to this bug.

I'm closing this bug from 2006 due to lack of activity on it on the part of both
developers and users. 

Any re-opening of this issue should be specific about what's wrong with the
current "diff" output, and how it should be improved. 

Since we already have "--diff-command" and "--diff-opts", it seems like there is
a lot of possibility to improve the diffs using solutions external to darcs.

On a 2008-05 mailing list entry, David said that implementing the 'patience
diff' algorithm could be useful

http://lists.osuosl.org/pipermail/darcs-users/2008-May/012229.html

There's also Bram Cohen's algorithm to consider:
http://lists.osuosl.org/pipermail/darcs-devel/2007-May/005664.html

Is that different?

Small case where I think the generated diff is not the one I would expect:

marcot@quindinho:~/Modelos/darcs$ cat a 
a
b
c
marcot@quindinho:~/Modelos/darcs$ darcs init
marcot@quindinho:~/Modelos/darcs$ darcs add a
marcot@quindinho:~/Modelos/darcs$ darcs record
Darcs needs to know what name (conventionally an email address) to use as the
patch author, e.g. 'Fred Bloggs <fred@bloggs.invalid>'.  If you provide one
now it will be stored in the file '_darcs/prefs/author' and used as a default
in the future.  To change your preferred author address, simply delete or edit
this file.

What is your email address?
addfile ./a
Shall I record this change? (1/2)  [ynWsfvpxdaqjk], or ? for help: y
hunk ./a 1
+a
+b
+c
Shall I record this change? (2/2)  [ynWsfvpxdaqjk], or ? for help: y
What is the patch name?
What is the patch name? a
Do you want to add a long comment? [yn]n
Finished recording patch 'a'
marcot@quindinho:~/Modelos/darcs$ # Edit a
marcot@quindinho:~/Modelos/darcs$ cat a
c
b
a
marcot@quindinho:~/Modelos/darcs$ darcs whatsnew
hunk ./a 1
-a
-b
hunk ./a 2
+b
+a
marcot@quindinho:~/Modelos/darcs$ darcs --version
2.0.2 (release)

I would expect it to generate something as:
hunk ./a 1
-a
+c
hunk ./a 3
-c
+a

On Mon, Nov 17, 2008 at 06:35:27PM -0000, Marco Túlio Gontijo e Silva wrote:
> marcot@quindinho:~/Modelos/darcs$ darcs whatsnew
> hunk ./a 1
> -a
> -b
> hunk ./a 2
> +b
> +a
> marcot@quindinho:~/Modelos/darcs$ darcs --version
> 2.0.2 (release)
> 
> I would expect it to generate something as:
> hunk ./a 1
> -a
> +c
> hunk ./a 3
> -c
> +a

Although I agree that we could improve on our diff, I don't think this
is a good example.  I don't see any reason why darcs should decide
that the 'b' was preserved rather than the 'c' or the 'a'.  I guess
you're thinking that it should take into account the location in the
file, and somehow treat lines that haven't moved specially?

David

I think it would be good, but maybe it's just the way I'm looking at it.  I'm
convinced this is a subjective point of view, so it's ok not to consider this
example.

We should take note of http://hackage.haskell.org/package/patience-0.1.1

A GPL licensed Haskell implementation by David Roundy: 
https://github.com/droundy/iolaus-broken/blob/master/Iolaus/Lcs2.hs#L128

Patience diff has been implemented now. Can this ticket be resolved?

The following patch sent by Jose Luis Neder <jlneder@gmail.com> updated issue issue346 with
status=resolved;resolvedin=2.10.0 HEAD

* resolve issue346: implement "patience diff" from bzr 
Ignore-this: 67e8a1a8098c98a7ef7dcab816885b6d
Add patience diff algorithm to Darcs/Util/Diff/Patience.hs from David Roundy implementation with some tweaks.
Make some changes required to let the user choose between algorithms with a flag.