Mercurial confusion - commit / push, backouts
- by Madmanguruman
I'm trying to set up a repository on a shared filesystem. I'm using Mercurial 2.1.2 on a Windows-based architecture.
I start with an empty folder on the shared filesystem and create a repository in it. After this, I dump in the baseline files, and add them to versioning, then commit the changes.
I then clone the repository to my local hard drive.
I then make a change in my local repository, commit it, then push back to the shared filesystem repository. The shared repo graph I get in TortoiseHG looks strange (to me).
This is the shared repo:
This is the local repo:
On the shared repo, the working directory always shows up on the top, then the graph goes 'down' to rev. 0 then back 'up' again through various revisions.
It looks to me like I have two different branches, even though everything is on the default branch. Also, that 'top' revision always says "* Working Directory * Not a head revision!"
I noticed that in my local repository, I don't get that dangling working directory at the top of the list - everything is in one branch.
I also noticed that on my local repository, I can back out the tip revision with no problem. On the shared filesystem repository, I cannot, since I get an error ("Cannot backout change on a different branch"). How can this be? Aren't they supposed to be identical to each other?
Am I fundamentally doing something wrong?
