[darcs-users] annotate output (Was: machine-readable formats)

Jason Dagit dagitj at gmail.com
Thu Sep 9 03:42:31 UTC 2010


On Sat, Sep 4, 2010 at 3:38 AM, Eric Kow <kowey at darcs.net> wrote:

> $ git blame sha1_file.c
> ...
> 0fcfd160 (Linus Torvalds  2005-04-18 13:04:43 -0700    8)  */
> 0fcfd160 (Linus Torvalds  2005-04-18 13:04:43 -0700    9) #include "cache.h"
> 1f688557 (Junio C Hamano  2005-06-27 03:35:33 -0700   10) #include "delta.h"
> a733cb60 (Linus Torvalds  2005-06-28 14:21:02 -0700   11) #include "pack.h"
> 8e440259 (Peter Eriksen   2006-04-02 14:44:09 +0200   12) #include "blob.h"
> 8e440259 (Peter Eriksen   2006-04-02 14:44:09 +0200   13) #include "commit.h"
> 8e440259 (Peter Eriksen   2006-04-02 14:44:09 +0200   14) #include "tag.h"
> 8e440259 (Peter Eriksen   2006-04-02 14:44:09 +0200   15) #include "tree.h"
> f35a6d3b (Linus Torvalds  2007-04-09 21:20:29 -0700   16) #include "refs.h"
> 70f5d5d3 (Nicolas Pitre   2008-02-28 00:25:19 -0500   17) #include "pack-revindex.h"628522ec (Junio C Hamano

I had to double check that.  I think you copy&pasted correctly and
that snippet is actually broken.  Looks like the last two lines were
accidentally joined.

The real point of this reply is just to put a +1 vote on
git-blame-compatible output if that's feasible.  By feasible I mean,
doesn't force us to break the way darcs works.  If it's not feasible
then, oh well.

My reason is simple:  We can leverage the tools that already support
git's blame output just a little bit more easily that way.

Jason


More information about the darcs-users mailing list