Problem with CVS-1.11, W2K client, Linux server.

[Kerry Kurian]

Hello all,

Apologies in advance for what may turn out to be a stupid question. I'm
only familiar with the discussions on the list over the past few months.

I've done everything I can think of to try to get a W2K CVS client to
work with a Linux CVS server. It almost works, but I have a specific
problem: I can't checkout a repository after I've created it. I can
checkout the CVSROOT directory, if I specify it. But, "checkout .",
which works on my Linux machine, doesn't work on the W2K client. As a
result, I can't add files or directories to the repository I am able to
create.

I'm guessing that I'm doing something completely boneheaded (some of you
might say: "Yeah, you're using Windows"), because I can't find any
reference to this problem anywhere.

Thanks in advance for your help.

Examples follow:

I can init a repository:

Win2K>  cvs -d :server:address@hidden:/home/user/repository init

But, when I try to check out this new repository, here's what happens:

Win2K>  cvs -t -d :server:address@hidden:/home/user/repository checkout .

cvs checkout: notice: main loop with
CVSROOT=:server:address@hidden:/home/user/repository
S-> so_module (., Updating, , )
cvs checkout: cannot open CVS/Entries for reading: No such file or
directory
cvs [checkout aborted]: no repository

I've tried with the '.' in quotes, in apostrophes, escaped, etc.

As a check of my own sanity, I logged into linuxbox and tried to
checkout there:

linuxbox>   cvs -d /home/user/repository checkout .
cvs checkout: Updating .
cvs checkout: Updating CVSROOT
U CVSROOT/commitinfo
[snip]
U CVSROOT/verifymsg

It turns out that cvs checkout on the W2K client does work if I specify
a named directory. For example:

W2K> cvs -d :server:address@hidden:/home/user/repository checkout CVSROOT
cvs server: Updating CVSROOT
U CVSROOT/checkoutlist
etc. etc. etc.

What's the deal? Thanks again.

-K