Re: update corrupts wc

Philip Martin <philip.martin_at_wandisco.com> writes:

> The first commit leaves A/B,not-present,op_depth=0 and that's correct.
> The second delete converts that to op_depth=2. That's wrong (I
> think). The A/B,not-present,op_depth=0 node should continue to exist,
> but perhaps we need an op_depth=2 node as well?

I've added an XFAIL regression test for this.

We have a related problem with excluded nodes:

svn up --set-depth exclude wc/A/B # A/B op_depth=0
svn rm wc/A # A/B op_depth=2
svn revert wc/A # A/B op_depth=2

I think delete is supposed to add a working node for every child with
a base_node, not convert base_nodes into working nodes.

-- 
Philip