Hi,
I just ran into the following problem: I have a branch A in which I made a
modification to a file X, and now I'm trying to merge that delta into a
separate branch B which doesn't come from a common ancestor of A (i.e. I
imported both A and B distinctly). File X has never existed in B. Once I try
the merge, it aborts at that file X saying that it is not under revision
control. While I know that there is nothing that it can do for my file, it
shouldn't abort the merge halfway through because of that IMHO.
To give a more precise description of the problem: I have imported linux
kernels (as you all might be aware by now) into svn. What I did at the time
was to create distinct import trees for the 2.4 kernels and 2.5 kernels. Now
I'm trying to move some mods I did on a 2.4 kernel into a 2.5 kernel and it
is failing halfway through the merge - this happened with 3 files and I could
only get past it once I commited dummy versions of these files into the 2.5
tree. I know that the problem stems from the fact that I don't have a common
ancestor to the 2.4 and 2.5 trees but I did that because they don't derive
from each other. I guess I could have picked a 2.3 version and imported that
as a common ancestor and then imported both 2.4 and 2.5 trees as branches
from it but I'm not sure that would solve the problem completely - if I have
a mod in a file that had been created only during 2.4 (i.e. didn't exist at
2.3 and didn't exist in 2.5 either) I think I would have run into the same
problem.
I tried the --force option but that didn't help me. At least not when I was
also running with --dry-run to see what was going to happen.
Cheers,
csd
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
Received on Mon Mar 10 20:40:14 2003