I'm wondering if it's possible to do a recursive diff operation to show
differences for a whole directory at once.
cvs diff
will do this for you.
Use the command line or WinCvs, since Tortoise does not provide this option
for a directory (at least not in the version I use)
However, I can't
Failing that, there are various patches that try to implement strict
locking on top of cvs floating around out there, a search for cvs edit
- -c might turn up Noel Yap's patches... I have not looked at them in a
long time... once upon a time there was an rcvs sourceforge.net project,
but I
You situation is rather uncommon, because the web interface of
http://membled.com/cvsroot/ give you access to the CVS repository tree.
Tools like Viewcvs Cvsweb give you access to individual revisions, but
this view gives you access to the actual RCS files.
On simple solution would be to use a
Saad Malik writes:
Basically what I
want to do, is for the cvs update command to somehow get rid of all
the (?) mark files it finds when updating the module.
Larry Jones writes:
Sorry, there's no way to do that in CVS.
But CVSgui (aka Wincvs) can help you here: use a CVSgui version
[EMAIL PROTECTED] wrote:
My question is: How can I set up CVS server to deny adding of a file to a
folder in repository when there is a file with the same name from windows
point of view ?
I could not find a script doing this; but I had some perl scripts for other
commitinfo tasks lying
I wish it were possible to just let the fooweb module check out into
c:/cvs/foonetbeanswebapp/fooweb... but unfortunately, there's just no good
Have you tried doing the whole thing, but in a toplevel directory not called
CVS or cvs ?
(this has a special meaning to CVS and it will defenitely get
Write a commitinfo script.
This is a serverside script that is run during the commit process before the
actual commit is done.
It allows you to deny the commit based on your rules.
In the script , you can access the repository and check if rcs files with
similar names (i.e. only differing in
We have a problem whereby some developers checkin files (via commandline
on
Unix/CVS host) that may have been edited on Windows PCs.
That is the root of the problem.
Developers working on PCs should use a PC cvs client to checkout their text
files in a sandbox on a PC- and they will work in CRLF
Build them right where they are now.
The sources you have now got locally, are _not_ the repository.
They are your private copy -often called a sandbox by CVS users.
Like a real sandbox, they are your spot for doing anything you like.
You can change them, without any risk for the real Wine
Paul Gelderblom (ptok) wrote:
edit -c (regrettably in my opinion) is not a part of the official
version of
CVS which is the subject of this mailing list. There is a patch available
Derek Rober Price wrote:
Other than that, I have and still do consider that patch to be a great
idea
edit -c (regrettably in my opinion) is not a part of the official version of
CVS which is the subject of this mailing list. There is a patch available
(see http://mail.gnu.org/archive/html/info-cvs/2003-12/msg00135.html)
which gives you cvs edit -c in the standard cvs, but it is more likely you
William Mallouk writes:
We need to version large files (400MB+) such as cd-rom images and
large binaries. We realized that CVS gets very slow when versioning
these files. Did anybody come across the same problem? Any suggestions?
Larry Jones writes:
Don't use CVS.
Jep, right, but some of
You are probably working in a directory which was checked out by mark.
CVS remembers the user who has checked out a sandbox in the CVS/Root file in
the sandbox.
- Original Message -
From: Anthony Ettinger [EMAIL PROTECTED]
To: [EMAIL PROTECTED]
Sent: Thursday, January 22, 2004 20:40
I checked it out in my own directory as 'ettinger'.
CVS/Root has /usr/local/shared/cvs in it. The
subdirectory kp2 is owned by mark, but this is the
repository, as far as i know.
Oh, you are working in local mode (in contrast with client/server) .
It's kind of odd that it asks you to login
I normaly don't want to use locking on the cvs archive.
On Windows projects I think it's simplier to use strickt locking on
the resource and the corrosponding header file. Merging changes in
that files does not make happy anyone :-)
If not how can I enshure that these files can only be
If we're already using commitinfo, commit_prep and log_accum to log and
send email notifications, can this script co-exist?
See http://www.cvshome.org/docs/manual/cvs-1.11.10/cvs_18.html#SEC165
All occurrences of the name `ALL' appearing as a regular expression are
used in addition to the
[EMAIL PROTECTED] wrote:
Hi everyone,
Is there a way of enforcing exclusive checkout in CVS?
Todd Denniston [EMAIL PROTECTED] wrote:
read the following message if you want more insight as to why, and if
you are not afraid of patching your tools, there patches for you.
You could have trouble if Windows and Linux have a different idea of
file line endings - you'd probably need to use the Cygwin CVS client
or another that won't insert carriage returns when checking out text
files on Windows.
Most windows clients have an option for Linux line endings - but
Hi all,
We often checkout using a sticky tag, so that
developers can work with
baselines of code.
Normally, checkins on files or directories with a
sticky non-branch tag
are disabled, which is correct.
However, we have found out that sometimes commit on
these sticky tags succeed, because
19 matches
Mail list logo
