| 
Created on 2008-05-18.16:57:08 by btcoburn, last changed 2009-10-23.23:29:04 by admin. 
 
  
   | msg4750 (view) | Author: btcoburn | Date: 2008-05-18.16:57:04 |  |  
   | For example, take a patch with a traditional darcs hash of 
'20080513112149-05dcb-4c273a98eae16cc9d273f18af9907f9f2e567224' 
and a new hashed file name of 
'0000012256-2f91882e2149f43e2463fb9ee80edd67e379734ab0e395ae6cf36655a595bff0'.
The command "darcs2 changes --match 'hash ...'" should be smart enough to match
either hash type. Currently it only matches the old hash.
Tested on darcs 2.0.0 build Apr  9 2008, at 17:14:20, stable release binary for
MacOS X. |  
   | msg5458 (view) | Author: markstos | Date: 2008-08-13.01:58:07 |  |  
   | Do the new hashed file names correspond to files or patches? 
I'm personally OK with rejecting this request as 'wont-fix'. Exposing one hash
to match in the UI is enough for me. If a second hash really expresses the same
thing different, it could just be confusing document and interact with. 
I'm going to go ahead and mark this request 'wont-fix' now, but someone else may
re-open it if they disagree. |  
   | msg5483 (view) | Author: mornfall | Date: 2008-08-13.13:15:32 |  |  
   | The hashed filenames correspond to file contents. Same patch can (and will) carry
different "new-style" hashes depending on how it is commuted. You don't want to
refer to a patch through the hashed filename. The original "old-style" (but
that's really just misnomer, it's a "patch hash" vs "file hash") hash refers to
the patch uniquely and that's what you want to stick to. I agree with wont-fix. |  
   | msg5484 (view) | Author: kowey | Date: 2008-08-13.13:18:42 |  |  
   | Should this become an FAQ of some sort?  Like it or not, people do look in
_darcs/patches :-) (or they create tools that grab patches).  
As an aside, maybe it would be useful if darcs provided a utility to return the
hashed file name corresponding to a given patch. |  
   | msg5485 (view) | Author: mornfall | Date: 2008-08-13.13:23:07 |  |  
   | I believe adding the filename holding the patch to changes --xml output might not
be hard. *However*: this file is not guaranteed to exist in a local repository
(lazy repos won't have all of the patch files). Dunno about usefulness of such a
feature. |  |
 
| Date | User | Action | Args |  | 2008-05-18 16:57:08 | btcoburn | create |  |  | 2008-08-13 01:58:10 | markstos | set | status: unread -> wont-fix nosy:
  + markstos
 messages:
  + msg5458
 |  | 2008-08-13 13:15:35 | mornfall | set | nosy:
  + mornfall, simon, kowey messages:
  + msg5483
 |  | 2008-08-13 13:18:45 | kowey | set | nosy:
  tommy, beschmi, kowey, markstos, dagit, simon, btcoburn, Serware, mornfall messages:
  + msg5484
 |  | 2008-08-13 13:23:11 | mornfall | set | nosy:
  tommy, beschmi, kowey, markstos, dagit, simon, btcoburn, Serware, mornfall messages:
  + msg5485
 |  | 2009-08-06 21:05:08 | admin | set | nosy:
  + dmitry.kurochkin, thorkilnaur, - beschmi |  | 2009-08-11 00:14:39 | admin | set | nosy:
  - dagit |  | 2009-08-25 17:24:14 | admin | set | nosy:
  + darcs-devel, - simon |  | 2009-08-27 14:02:50 | admin | set | nosy:
  tommy, kowey, markstos, darcs-devel, thorkilnaur, btcoburn, dmitry.kurochkin, Serware, mornfall |  | 2009-10-23 22:43:06 | admin | set | nosy:
  + serware, - Serware |  | 2009-10-23 23:29:04 | admin | set | nosy:
  + Serware, - serware | 
 |