2 patches for repository http://darcs.net/screened:

patch 58901b1ae2881f1b11d385610d959be53a095bdb
Author: Ben Franksen <ben.franksen@online.de>
Date:   Sat Dec 24 07:29:45 CET 2022
  * RepoPatchV3: improved display of conflictors

  It now displays the effect, followed by the (in-effective) conflicting
  alternatives in a way that is similar to conflict markup (though minus
  display of the base). This is easier to parse for humans and more
  informative than a dump of the internal structure of a conflictor.

  Unfortunately it is not so easy to do this for RepoPatchV1 or V2.

patch a467d1bd53846201882a37a2e4c0d7042c2cd3f7
Author: Ben Franksen <ben.franksen@online.de>
Date:   Sun Dec 25 07:29:29 CET 2022
  * improve display of "conflicted" RebaseChange

  The previous, simpler method displayed them as a "large" conflict of the
  whole patch with the inverse of the fixups. It is now more fine-grained,
  showing conflicts inline per prim. This works by choosing RepoPatchV3 to
  wrap prims and then display the result of a forceCommute with the residual
  fixups. Special care needs to be taken to filter out leading inverse pairs
  that forceCommute temporarily injects into the patch (a hack that was needed
  to make it work with RepoPatchV3).

Attachments

• patch-preview.txt
• repopatchv3_-improved-display-of-conflictors.dpatch

Both fine, though I would like to disentangle Rebase from RepoPatchV3.

> disentangle Rebase from RepoPatchV3

I agree tha would be nice. The problem is that we have to deal with 
a series of fixups stuck against another series of prims (from the 
Named patch). So far I haven't found a way to calculate conflicting 
alternatives from that in some elementary way i.e. w/o re-inventing 
conflictors.