[cvsnt] Problem branching, then error on commit in CVSNT
andrew at cushen.com
Thu May 8 22:37:13 BST 2008
So far so good - I created a new directory named "clean" in the root of the
C drive, checked everything out, modified the problem files, then started
Committing the problem files back in. Here is the output of one of the
C:\clean\RELEASE_5-01-2008>cvs ci -m "test commit" stock.aspx
Checking in stock.aspx;
/CLIENTwebsite/CLIENTsite/stock.aspx,v <-- stock.aspx
new revision: 220.127.116.11; previous revision: 18.104.22.168
["CLIENTwebsite" is actually the name of the repository, and I can't change
it at the moment...]
The Commits for the other problem files look similar.
So it looks good, right? My question then becomes: what next?
I'm guessing that you're going to recommend dumping my original sandbox, and
using the one I just created. However, as I mentioned, I have been using the
PushOK plug-in in Visual Studio 2005. I believe it is only the source files,
not the VS Project file, that are under source control; yet the Project file
seems tied to the original sandbox somehow, perhaps in a setting somewhere.
If I try to open Visual Studio then use the "Open from source control"
command on the File menu, and point VS to the Repository, I get an error
saying there is no project file in the directory. Which, of course, there
If I open the Project using the Open Project command, the project opens
automatically using the older, messed-up versions of the problem files from
the original sandbox.
Would it work to check in the files from the new sandbox at the command
prompt, perhaps then creating a new Branch, and *then* open the project in
VS? I'm afraid the messed up state of the original sandbox would just
corrupt things all over again, the way things are.
Any recommendations? I could switch to the March Hare plug-in if that would
make things easier...
More information about the cvsnt