[cvsnt] Multiple vendor branches
jo.kilian at gmx.de
Tue Mar 13 08:08:38 GMT 2007
Gerhard Fiedler schrieb:
> If I understand you correctly, this shouldn't be. What says the cvs status
> of one of these rogue files?
Yes you understood correct: this shouldn't be.
I'm not sure whether this is a "special behaviour" of vendorbranches or
of "cvs import" - or a combination of both.
The situation(CVS-behaviour is following - as I understand:
* When I have a new USERBRANCH and I ADD a new file to this branch, this
file will not be available on the trunk unless I do not merge from
userbranch to Trunk.
* When I do an INITIAL IMPORT of files to an VENDORBRANCH (nothing has
been there before in the repository) the files will be available on the
TRUNK (as revision 1.1) as well as on the vendorbranch (revision
18.104.22.168). Checking out the trunk will work as well as checking out the
vendorbranch - I will get BOTH times the same files ...
Furtheron unexpected behaviour occurs (and that's my problem):
* Importing on a second VENDORBRANCH the following problem occurs:
1. Modified files on the second vendorbranch (modified in comparison
to the initial vendorbranch) were imported correctly. They exist as
version 1.1 on trunk, 22.214.171.124 on the initial vendorbranch and 126.96.36.199 on
the second vendorbranch, where revisions 1.1 and 188.8.131.52 are identical
2. New files on the second vendorbranch (files that are new in
comparison to the initial vendorbranch) were imported in that way that
they exist as revision 1.1 on the trunk and 184.108.40.206 on the second
vendorbranch. They don't exist on the initial vendorbranch.
Checking out the trunk, I will get all 1.1 revisions - without any
differentiation whether they origin from initial vendorbranch or the
second vendorbranch ...
On one hand that's "correct" behaviour - that's how importing of
vendorbranches works: new files were imported as vendor-revision on the
branch (Revision 220.127.116.11 or 18.104.22.168 or 22.214.171.124 ...) and as 1.1 on the
trunk. Checking out the trunk I will get all 1.1 revisions ...
But on the other hand that's a "messy" behaviour, since it cannot be
distinguished on the trunk, which vendorbranch is the source of those
cvs status -v -- xxx.cgi (in directory G:\Projekte\Temp\Test)
File: xxx.cgi Status: Up-to-date
Working revision: 126.96.36.199
Repository revision: 188.8.131.52 /Repo/Test/xxx.cgi,v
Expansion option: kv
Commit Identifier: (none)
Sticky Tag: (none)
Sticky Date: (none)
Sticky Options: (none)
Merge From: (none)
second_Vendortag (revision: 184.108.40.206)
second_Vendorbranch (branch: 1.1.3)
***** CVS exited normally with code 0 *****
More information about the cvsnt