darcs

Issue 1214 clearly identify patches as being conflict resolution patches

Title clearly identify patches as being conflict resolution patches
Priority wishlist Status deferred
Milestone Resolved in
Superseder Nosy List darcs-devel, ganesh, kowey
Assigned To
Topics Conflicts, UI

Created on 2008-11-10.00:18:39 by twb, last changed 2016-01-27.14:12:55 by gh.

Messages
msg6622 (view) Author: twb Date: 2008-11-10.00:18:35
Consider the transcript below.  If I pull only the first few patches,
I will get lots of conflicts and darcs will ask me to resolve them.
If I do, *my* conflict resolution will conflict with Eric's (in patch
5, I think)!  But if I pull *all* of the patches below, there are no
conflicts!

To me, this is a workflow problem.  I can easily see if a patch
introduces a conflict, but I can't easily see if a patch (and if so,
which one) resolves those conflicts.  The only hint is when the patch
name contains the word "conflict", which depends on the auspices of
the patch author.

This probably isn't a problem if I always "darcs pull --all".  But I
think that loses some nice benefits of using darcs.

$ darcs pull -s
Pulling from "http://darcs.net/unstable"...
[...]
Mon Oct 27 03:40:09 EST 2008  tux_rocker@reinier.de
  * add a get_unrecorded_in_files to check for unrecorded changes in a subset of working directory

    M! ./src/Darcs/Diff.hs -9 +11
    M! ./src/Darcs/Patch/FileName.lhs +2
    M! ./src/Darcs/Repository.lhs +2
    M! ./src/Darcs/Repository/Internal.hs -15 +27
Shall I pull this patch? (1/5)  [ynWvpxdaqjk], or ? for help: y
Mon Oct 27 06:06:12 EST 2008  tux_rocker@reinier.de
  * make get_unrecorded_private work with type witnesses again

    M ./src/Darcs/Repository/Internal.hs -5 +7
Shall I pull this patch? (2/5)  [ynWsfvpxdaqjk], or ? for help: y
Mon Oct 27 06:46:36 EST 2008  tux_rocker@reinier.de
  * make whatsnew use the lstat-saving functions to scan the working copy

    M ./src/Darcs/Commands/WhatsNew.lhs -5 +6
    M ./src/Darcs/Repository/Internal.hs -2 +2
Shall I pull this patch? (3/5)  [ynWvpxdaqjk], or ? for help: y
Sat Nov  1 08:59:44 EST 2008  Reinier Lamers <tux_rocker@reinier.de>
  * hopefully less buggy version of get_unrecorded_in_files

    M! ./src/Darcs/Commands/WhatsNew.lhs -15 +7
    M! ./src/Darcs/Diff.hs -15 +31
    M! ./src/Darcs/Repository/Internal.hs -11 +24
Shall I pull this patch? (4/5)  [ynWvpxdaqjk], or ? for help: y
Mon Nov 10 02:30:40 EST 2008  Eric Kow <E.Y.Kow@brighton.ac.uk>
  * Resolve conflicts (lstat vs de-literate)
  The conflits were between Reinier's lstat-saving patches on the one
  hand and the Diff haddockisation on the other.

    M ./src/Darcs/Diff.hs -1 +55
Shall I pull this patch? (5/5)  [ynWsfvpxdaqjk], or ? for help: y
Finished pulling and applying.
msg6847 (view) Author: kowey Date: 2008-12-21.17:16:22
This sounds like it might be tricky to implement unless we introduce some
information into new patches that marked them as being conflict resolution patches.

We could use the Ignore-this feature in the patch log, for example.
msg18982 (view) Author: gh Date: 2016-01-27.14:12:54
Relevant: http://hub.darcs.net/ganesh/darcs-conflict-marking
History
Date User Action Args
2008-11-10 00:18:39twbcreate
2008-12-21 17:16:30koweysetstatus: unread -> deferred
priority: wishlist
title: conflict resolution patches don't "stand out" -> clearly identify patches as being conflict resolution patches
nosy: kowey, dagit, simon, twb, thorkilnaur, dmitry.kurochkin
messages: + msg6847
topic: + UI
2009-08-10 23:51:05adminsetnosy: - dagit
2009-08-25 17:34:05adminsetnosy: + darcs-devel, - simon
2009-08-27 14:14:01adminsetnosy: kowey, darcs-devel, twb, thorkilnaur, dmitry.kurochkin
2009-09-04 17:13:05koweysettopic: + Conflicts
nosy: kowey, darcs-devel, twb, thorkilnaur, dmitry.kurochkin
2016-01-27 14:12:56ghsetnosy: + ganesh, - twb, thorkilnaur, dmitry.kurochkin
messages: + msg18982