Dear all I have a nasty problem with my "old-and-faithful" CVS version 1.12.13.
I have two files "FileA" and "FileB", where I use locking "cvs watch on". The files are in different parts of a directory structure. Assume that I have two sandboxes, and the files have been cvs commit'ed many times. In Sandbox A I do $ cvs edit FileA $ date > FileA # Just to get whatever new content in the file With "cvs editors FileA" and "cvs status FileA" I see my self as editor and that the file is changed. In Sandbox B I do $ cvs update -r 1.1 FileA $ cvs update -A FileA In Sandbox A I then notice the CVS error happens - as described in https://savannah.nongnu.org/bugs/index.php?23370 I am no longer the editor of FileA. I then do $ cvs edit FileB $ date > FileB # Just to get whatever new content in the file $ cvs edit FileA # I now try to restore my edit state on FileA cvs editors now show that I have editor status on both FileA and FileB .... Here is STATE_1 (I will refer to this spot below) I can also verify that in $CVSROOT/myproject/..../CVS/fileattr that I am indeed the editor of each of the files - good! BUT, if I do $ cvs unedit FileA then I get nothing on the command line - I simply loose the file lock (not mentioned in cvs editors) - and here comes the problem: The file is STILL locally changed - it is not reverted to the last check-in state or the truck... whatever. However if I do $ cvs unedit FileB then CVS kindly asks if I want to revert my changes - which I then accept, and get back to the last version - i.e. drop my local changes. I will kindly ask you to explain to me, why CVS sees FileA and FileB differently after I am at STATE_1 (see above). Best -- Peter Toft <[email protected]> _______________________________________________ Bug-cvs mailing list [email protected] http://lists.nongnu.org/mailman/listinfo/bug-cvs
