[Gnu-arch-users] how to recover from large conflicts?

[Denys Duchier]

Here is another vexing first for me:

I have been tracking:

        address@hidden/cscvs--experimental--1.1

in my own branch:

        address@hidden/cscvs--denys--1.1

Today, Charles needed to change the indentation of a large block of
code (to add exception handling).  Unfortunately, on my branch, I made
a small change in that block.  As a result, star-merge leaves behind a
fairly large .rej file.

What is a good strategy to recover from this?

I see the following possibilities:

1. bite the bullet and apply all the modifications by hand.
   this would certainly be fastest, but it is so intellectually
   unsatisfying that I'd like to investigate alternatives :-)

2. hack the .rej file to produce a diff that actually applies.  this
   file seems to be in unified diff diff format, but, when fed to
   patch, the latter complains about "missing header for context
   diff".  Is hacking this file a viable option?  I gave it a quick
   try and failed (and patch is woefully uninformative about possible
   causes)

3. using "replay" maybe.  I wasn't able to figure out how to do this.
   I was thinking: replay the cduffy stuff (the old and the new that I
   wanted to star-merge), then replay my patches and fix what needs
   fixing.  Experimenting with replay only made me utterly confused.

4. temporarily back out the patch of mine that is probably causing
   trouble, star-merge, then re-apply my patch slightly tweaked.  I
   don't know how to do this either except by explicitly applying it
   backwards and then committing - is that the only option here?

I think I need some practical tutoring ("remedial arch" anyone? :-)

Cheers,

-- 
Dr. Denys Duchier
Équipe Calligramme
LORIA, Nancy, FRANCE