darcs

Issue 308 quitting from --ask-deps records the empty patch anyway

Title quitting from --ask-deps records the empty patch anyway
Priority bug Status resolved
Milestone Resolved in
Superseder Nosy List darcs-devel, dmitry.kurochkin, kowey, quick, thorkilnaur, tommy
Assigned To quick
Topics FromRequestTracker

Created on 2006-10-17.08:56:33 by kowey, last changed 2009-08-27.14:11:08 by admin.

Messages
msg1105 (view) Author: kowey Date: 2006-10-17.08:56:25
Import of RT #531

"Stephen J. Turnbull" <stephen@xemacs.org> 2005-10-01
-----------------------------------------------------
Subject: quitting from --ask-deps records the empty patch anyway

This is moderately annoying, although darcs unrecord is easy enough.
msg1106 (view) Author: kowey Date: 2006-10-17.09:02:33
I don't quite understand this - what do you mean by the empty patch?

- Do you mean say n to all the changes and then hit q during the ask-deps
interaction?
- Otherwise, accepting some changes and then hitting q during ask-deps seems ok
to me.
- Hitting q during the changes just quits the record dialogue

All three behave as I would expect them to, as a user.  Could you please clarify?

Thanks,

Eric
msg1859 (view) Author: kowey Date: 2007-07-18.06:47:07
Pasting in a mail exchange between me and Stephen (thanks!)

 > Stephen: Quick clarification question about a bug you filed in 2005         
                                  
 > (quitting from --ask-deps records the empty patch anyway)                   
                                  
 >   http://bugs.darcs.net/issue308                                            
                                  
 >                                                                             
                                  
 > > The only inconsistency is that if both recordings and dependencies are    
                                  
 > > empty darcs will not abort but create a noop patch; it's an awkward       
                                  
 > > feature.                                                                  
                                  
 >                                                                             
                                  
 > Is this what you meant by 'records the empty patch anyway'?                 
                                  

Yes, I think so.  Empty patches have their uses, but I would prefer
that "q" quit here.  It could either be hardcoded to prompt whether to
really create the empty patch (this is a rare use case either way, I
think), or there could be a --record-empty-patch flag with values
"never/no", "ask", "always/yes".  This would allow people to set
defaults in preferences.  That may be overkill though.
msg1998 (view) Author: kowey Date: 2007-08-05.19:59:33
Resolved in unstable with

Sat Jul 28 21:03:29 CEST 2007  Kevin Quick <quick@sparq.org>
  * Fix issue308: don't record empty patch with --ask-deps

Mon Jul 23 08:21:14 CEST 2007  Kevin Quick <quick@sparq.org>
  * Add test for issue308
History
Date User Action Args
2006-10-17 08:56:33koweycreate
2006-10-17 09:02:40koweysetstatus: unread -> unknown
nosy: droundy, tommy, kowey
messages: + msg1106
2007-07-18 06:47:12koweysetnosy: + beschmi
messages: + msg1859
2007-07-19 06:07:14koweysetassignedto: kowey -> quick
nosy: + quick
2007-08-05 19:59:34koweysetstatus: unknown -> resolved-in-unstable
messages: + msg1998
2007-08-09 12:03:01koweysetstatus: resolved-in-unstable -> resolved-in-stable
2008-09-16 21:30:28adminsetstatus: resolved-in-stable -> resolved
nosy: + dagit
2009-08-06 17:32:46adminsetnosy: + markstos, jast, Serware, dmitry.kurochkin, darcs-devel, zooko, mornfall, simon, thorkilnaur, - droundy, quick
2009-08-06 20:30:12adminsetnosy: - beschmi
2009-08-10 21:57:16adminsetnosy: + quick, - markstos, darcs-devel, zooko, jast, Serware, mornfall
2009-08-10 23:57:03adminsetnosy: - dagit
2009-08-25 17:47:32adminsetnosy: + darcs-devel, - simon
2009-08-27 14:11:08adminsetnosy: tommy, kowey, darcs-devel, quick, thorkilnaur, dmitry.kurochkin