[cvsnt] Cvs commit in batches of 1000

[Bo Berglund]

On Thu, 3 Nov 2005 11:02:36 -0700, "McCullough, Ryan"
<rmccullough at rightnow.com> wrote:

>From a *private* reply email (please do send responses to the list
instead of directly to me...):
.. snip ...
>What my script does is checkout the module into ./scripts. Under
>./scripts is a directory which matches the username, we will use john.
>Now, once the directory is checked out, I copy over the scripts, vp's
>and low-level recorded files into my cvs sandbox. I will then do about 6
>'cvs add' commands to pick up new files. Then I run one large commit
>command. Here is the command line:
>
>cvs -z3 commit -m "John - Automation daily checkin - 11/03/2005"
>
>This command returns:
>cvs commit: Examining .
>cvs commit: Examining lls
>cvs commit: Examining vp
>Protocol error: too many arguments
>
>I am going to experiment with using the -c option to the command line
>but from my understanding cvs should automatically only try to commit
>changed and new files right? So the -c option is kind of a moot point.
>Additionally, it is only occassionally that there are enough changed
>files that the import fails.

The only way I can figure out is that behind the scenes (without
feedback to the user) cvs is executing a shell command or similar with
a command line consisting of a list of files to commit and this list
gets way too long.

Now it is very important that you tell us what version you are
running. This is best done by executing the 'cvs ver' command from
within a sandbox and pasting the result back here.


/Bo
(Bo Berglund, developer in Sweden)