Discussion:
[fossil-users] no common ancestor
B Harder
2014-06-18 17:52:22 UTC
Permalink
If everything is (ultimately) spawned from the "initial empty
checkin", how can I be getting "WARNING - no common ancestor: <path>"
when merging one branch into another ?

-bch
Richard Hipp
2014-06-18 17:55:32 UTC
Permalink
Post by B Harder
If everything is (ultimately) spawned from the "initial empty
checkin", how can I be getting "WARNING - no common ancestor: <path>"
when merging one branch into another ?
Perhaps a "shun" broke the check-in chain somewhere along the way.

There is the "fossil test-ancestor-path HASH1 HASH2" command that you can
use to find common ancestors between two arbitrary check-ins. Perhaps you
can use that to figure out where the problem sits.
--
D. Richard Hipp
***@sqlite.org
Andy Bradford
2014-06-18 18:55:55 UTC
Permalink
There is the "fossil test-ancestor-path HASH1 HASH2" command that you
can use to find common ancestors between two arbitrary check-ins.
Perhaps you can use that to figure out where the problem sits.
Nifty command. Here's the output for the particular case I pointed out:

$ fossil test-ancestor-path d5de0c469b d07278c11c
1: 13 d5de0c469b04 2014-06-18 18:38:49 VERSION1
2: 11 61549ff3772e 2014-06-18 18:38:33
3: 9 b6e320bcdac2 2014-06-18 18:38:24
4: 3 a7c3818537c0 2014-06-18 18:35:59 PIVOT
5: 5 7fba82feb33a 2014-06-18 18:36:12
6: 7 82a44e997365 2014-06-18 18:35:00
7: 15 d07278c11c50 2014-06-18 18:39:25 VERSION2

Should I be expecting an error here given that if I try to merge
d07278c11c50 into d5de0c469b04 I get the error that there is no common
ancestor? Seems like this actually did find a common ancestor on line 4,
unless I misunderstand this output.

Thanks,

Andy
--
TAI64 timestamp: 4000000053a1e0dc
Matt Welland
2014-06-18 19:22:24 UTC
Permalink
IIRC this can happen if a file is created independently in both branches.
Post by Richard Hipp
Post by B Harder
If everything is (ultimately) spawned from the "initial empty
checkin", how can I be getting "WARNING - no common ancestor: <path>"
when merging one branch into another ?
Perhaps a "shun" broke the check-in chain somewhere along the way.
There is the "fossil test-ancestor-path HASH1 HASH2" command that you can
use to find common ancestors between two arbitrary check-ins. Perhaps you
can use that to figure out where the problem sits.
--
D. Richard Hipp
_______________________________________________
fossil-users mailing list
http://lists.fossil-scm.org:8080/cgi-bin/mailman/listinfo/fossil-users
--
Matt
-=-
90% of the nations wealth is held by 2% of the people. Bummer to be in the
majority...
Andy Bradford
2014-06-18 18:50:17 UTC
Permalink
If everything is (ultimately) spawned from the "initial empty
checkin", how can I be getting "WARNING - no common ancestor: <path>"
when merging one branch into another ?
How did you get to this error? In my case, I am able to cause it with
the following test Fossil:

http://fossil.bradfords.org:8080/

If you clone that and then try to merge the branch named ``four'' with
trunk you'll get this error:

$ fossil merge trunk
cannot find a common ancestor between the current checkout and trunk

The checkin in branch ``four'' is a merge with cherry picking of a part
of trunk.

What does the manifest for the checkin that gives the errorlook like?

$ fossil artifact `fossil info | grep checkout | awk '{ print $2 }'`

Andy
--
TAI64 timestamp: 4000000053a1df8b
Loading...