Re: Info-cvs Digest, Vol 8, Issue 31

[Bogdan Serbanoiu]

On Wed, 30 Jul 2003 address@hidden wrote:

> Send Info-cvs mailing list submissions to
>       address@hidden
> 
> To subscribe or unsubscribe via the World Wide Web, visit
>       http://mail.gnu.org/mailman/listinfo/info-cvs
> or, via email, send a message with subject or body 'help' to
>       address@hidden
> 
> You can reach the person managing the list at
>       address@hidden
> 
> When replying, please edit your Subject line so it is more specific
> than "Re: Contents of Info-cvs digest..."
> 
> 
> Today's Topics:
> 
>    1. Re: setting up cvs and cvsweb.... (Robert Helmer)
>    2. Not waiting on #cvs.lock (gilmurra)
>    3. Re: Script for a CVS only shell available? (Wu Yongwei)
>    4. Re: Feature Request: admin files for "cvs import" and "cvs
>       add<dir>" (Wu Yongwei)
>    5. Re: Puzzled with sanity.sh (Wu Yongwei)
>    6. Re: Feature Request: admin files for "cvs import" and "cvs
>       add<dir>" (Greg A. Woods)
>    7. Re: RSE's cvs import patch against the current CVS source
>       (Julien Wajsberg)
>    8. Re: RSE's cvs import patch against the current CVS source 
>       (Mark D. Baushke)
>    9. Re: RSE's cvs import patch against the current CVS source
>       (Julien Wajsberg)
>   10. RE: Strange diff behavior on branch (Lemke, Michael  IZ/HZA-IC1)
> 
> 
> ----------------------------------------------------------------------
> 
> Message: 1
> Date: Tue, 29 Jul 2003 11:35:00 -0700
> From: Robert Helmer <address@hidden>
> Subject: Re: setting up cvs and cvsweb....
> To: bruce <address@hidden>
> Cc: address@hidden, address@hidden
> Message-ID: <address@hidden>
> Content-Type: text/plain; charset=us-ascii
> 
> On Tue, Jul 29, 2003 at 12:37:58AM -0700, bruce wrote:
> > I think I have CVS working... but I'm not really sure!!!
> 
> Have you tested it? Do this on the command line :
> 
> CVS_RSH=ssh
> CVSROOT=:ext:address@hidden:/CVS
> export CVS_RSH CVSROOT
> cvs co <module>
> 
> (where <module> is a name of a module you have in CVS).
> 
> > 
> > I have CVSROOT set to --> :ext:address@hidden:/CVS
> 
> Are you trying to use this CVSROOT in CVSWeb? Because CVSWeb cannot
> use a remote repository, it muts be ON lserver2.mesa.com, and use
> a CVSROOT like "CVSROOT=/CVS".
> 
> > When I try to access CVSWEB from the web server, I'm getting an error
> > indicating an error:
> 
> Check the Apache error.log file. What is the error there?
> 
> > within my /etc/profile, i have tried...
> >  CVSROOT="/CVS"
> 
> This is the only one that will work.
> 
> > within /etc/xinetd.d/cvspserver.. i have
> > service cvspserver
> 
> This does not matter for CVSWeb.
> 
> > My questions.... I understand that you really want SSH for the security...
> > But if I use CVSROOT="ext....", do I need to change the cvspserver
> > information, specifically the line that has pserver..???
> 
> ext and pserver are mutually exclusive. 
> 
> > Assuming I get this to work..what changes need to be made regarding CVSWEB.
> > I try to get CVSWEB running and I get a log error saying that the cvsweb.cgi
> > script is having an issue over an invalid character... I assume that I don't
> > have CVS actually running correctly....
> 
> No, that's a problem with the CGI. 
> You might want to try ViewCVS, it's easier to get running:
> http://viewcvs.sourceforge.net/
> 
> 
> > I'd like to basically insure that CVS is running.. and then get CVSWEB
> > running... and then be able to access CVS via a remote server as well as a
> > remote browser running CVSWEB....
> 
> 
> pserver doesn't have to be running to use CVS via SSH or locally (like
> CVSWeb uses). If you don't need pserver, don't run it. SSH is better.
> 
> 
> 
> HTH,
> Rob Helmer
> 
> 
> 
> ------------------------------
> 
> Message: 2
> Date: Tue, 29 Jul 2003 11:55:10 -0700
> From: gilmurra <address@hidden>
> Subject: Not waiting on #cvs.lock
> To: address@hidden
> Message-ID: <address@hidden>
> Content-Type: text/plain; charset=us-ascii
> 
> I'm doing a "cvs update -P -d ..." every two hours from a cron job.
> Everything works fine 'til somehow a #cvs.lock doesn't get deleted.
> Every two hours another update job starts and eventually we run out of
> job space. I realize some locks are valid and I'm writing a script to
> inform me when a lock is more than one day old but that won't solve my
> immediate problem.
> 
> Is there a way to abort instead of waiting when a lock is set when doing
> an update? 
> 
> Thanks, Frank
> 
> 
> 
> ------------------------------
> 
> Message: 3
> Date: Wed, 30 Jul 2003 09:32:32 +0800
> From: "Wu Yongwei" <address@hidden>
> Subject: Re: Script for a CVS only shell available?
> To: <address@hidden>
> Message-ID: <address@hidden>
> Content-Type: text/plain;     charset="iso-8859-1"
> 
> Thank you very much.  The first script is *so* goddamnedly nice and slim.  I
> nearly thought myself an idiot at seeing it.
> 
> Best regards,
> 
> Wu Yongwei
> 
> --- Original Message from Mark Priest ---
> 
> Wu,
> 
> Here is a thread from the cygwin mailing list that has such a script:
> <http://cygwin.com/ml/cygwin/2003-04/msg00317.html>
> 
> Also, the scponly script at this URL http://www.sublimation.org/scponly/ has
> another example.
> 
> -Mark
> 
> 
> 
> 
> ------------------------------
> 
> Message: 4
> Date: Wed, 30 Jul 2003 09:39:35 +0800
> From: "Wu Yongwei" <address@hidden>
> Subject: Re: Feature Request: admin files for "cvs import" and "cvs
>       add<dir>"
> To: <address@hidden>
> Message-ID: <address@hidden>
> Content-Type: text/plain;     charset="iso-8859-1"
> 
> Greg A. Woods wrote:
> > If you're worried about that kind of threat then you have far bigger
> > problems than you think.
> >
> > CVS is not a security tool.
> 
> No, it is not.  Yet.  But I love CVS and love to have a secure CVS.  (BTW,
> my employer is a network security vendor :-).)
> 
> Don't you like CVS to be more secure?
> 
> Best regards,
> 
> Wu Yongwei
> 
> 
> 
> 
> ------------------------------
> 
> Message: 5
> Date: Wed, 30 Jul 2003 09:53:02 +0800
> From: "Wu Yongwei" <address@hidden>
> Subject: Re: Puzzled with sanity.sh
> To: <address@hidden>
> Message-ID: <address@hidden>
> Content-Type: text/plain; charset="iso-8859-1"
> 
> I knew I was having some extreme bad luck.  I had already checked the
> check.log file and had not found any problems.  To keep myself sane, I am
> sending the check.log to the list in this message.  (I suppose the sanity
> check is completely OK on your box?)
> 
> Or is my Red Hat 7.3 box going insane?
> 
> Best regards,
> 
> Wu Yongwei
> 
> --- Original Message from Mark D. Baushke ---
> 
> The immediate cause of the failure will be found by looking in the
> "/home/adah/src/ccvs/src/check.log" file.
> 
> The version-1 test just runs the command:
> 
>     `pwd`/cvs --version
> 
> and tries to compare the output with what it expects. This is a good
> sanity test for the expr command in your path among other things...
> 
>    Good luck,
>    -- Mark
> -------------- next part --------------
> A non-text attachment was scrubbed...
> Name: check.log
> Type: application/octet-stream
> Size: 779 bytes
> Desc: not available
> Url : 
> http://mail.gnu.org/pipermail/info-cvs/attachments/20030730/5e3fb8df/check.obj
> 
> ------------------------------
> 
> Message: 6
> Date: Wed, 30 Jul 2003 01:57:38 -0400 (EDT)
> From: "Greg A. Woods" <address@hidden>
> Subject: Re: Feature Request: admin files for "cvs import" and "cvs
>       add<dir>"
> To: "Wu Yongwei" <address@hidden>
> Cc: CVS-II Discussion Mailing List <address@hidden>
> Message-ID: <address@hidden>
> Content-Type: text/plain; charset=us-ascii
> 
> [ On Wednesday, July 30, 2003 at 09:39:35 (+0800), Wu Yongwei wrote: ]
> > Subject: Re: Feature Request: admin files for "cvs import" and "cvs 
> > add<dir>"
> >
> > Don't you like CVS to be more secure?
> 
> No, actually I don't, at least not in the way you're suggesting.
> 
> I want CVS to be a good source-code change tracking tool, and nothing
> more.
> 
> -- 
>                                               Greg A. Woods
> 
> +1 416 218-0098                  VE3TCP            RoboHack <address@hidden>
> Planix, Inc. <address@hidden>          Secrets of the Weird <address@hidden>
> 
> 
> 
> ------------------------------
> 
> Message: 7
> Date: Wed, 30 Jul 2003 12:07:38 +0200
> From: "Julien Wajsberg" <address@hidden>
> Subject: Re: RSE's cvs import patch against the current CVS source
> To: address@hidden
> Message-ID:
>       <address@hidden>
> Content-Type: text/plain; charset=us-ascii
> 
> 
> I just saw that 'verifymsg' is processed before 'importinfo' with your
> patch.
> I don't think it is the correct behaviour :)
> 
> How about modifying this ?
> Somewhere before the call to do_verify ?
> 
> --
> Julien Wajsberg
> 
> ----------
> Answer to Ralf S. Engelschall <address@hidden> :
> ----------
> 
> 
> On Tue, Jul 29, 2003, Mark D. Baushke wrote:
> 
> > > Any possibility to make it a feature of the mainstream CVS?
> > [...]
> > Also, the patch should not be conditional on the macros RSE_PATCHES or
> > RSE_PATCH_IMPORTINFO, so those parts of the patch need to be removed.
> > [...]
> 
> Sure, the #ifdef/#endif wrappers are there just for two reason: to allow
> me to exactly identify what code changes belong to which parts of my
> larger patch set and to allow me to selectively activate/deactivate
> some parts of it. For a possible inclusion into the mainstream CVS
> source these would be gone, of course. As already mentioned to Mark
> in a private mail, I'll try to find some time over the next weeks and
> work-off my patch set by adding sanity.sh and documentation changes and
> moving it up to the latest CVS HEAD version. I also will separate the
> large patch set into individual smaller ones (for easier review by the
> CVS developer team).
>                                        Ralf S. Engelschall
>                                        address@hidden
>                                        www.engelschall.com
> 
> 
> 
> _______________________________________________
> Info-cvs mailing list
> address@hidden
> http://mail.gnu.org/mailman/listinfo/info-cvs
> 
> 
> 
> 
> 
> 
> 
> ------------------------------
> 
> Message: 8
> Date: Wed, 30 Jul 2003 03:59:41 -0700
> From: "Mark D. Baushke" <address@hidden>
> Subject: Re: RSE's cvs import patch against the current CVS source 
> To: "Julien Wajsberg" <address@hidden>
> Cc: address@hidden
> Message-ID: <address@hidden>
> 
> Julien Wajsberg <address@hidden> writes:
> 
> > I just saw that 'verifymsg' is processed before 'importinfo' with your
> > patch. I don't think it is the correct behaviour :)
> 
> I suggest that you should not depend on the ordering of the verifymsg as
> compared to the importinfo or commitinfo.
> 
> In client/server mode, verifymsg is processed first. In local mode,
> verifymsg is usually called after the commitinfo has been called.
> 
> It is just as reasonable to reject a commit due to a bad log message as
> it is to reject it because a file is being committed does not pass some
> commitinfo check.
>  
> > How about modifying this ?
> 
> I believe it would be a waste of time. There is still an 'enhancement'
> request http://ccvs.cvshome.org/issues/show_bug.cgi?id=27 to have the
> verifymsg processed only once rather than for each directory...
> 
> > Somewhere before the call to do_verify ?
> 
> I'd suggest that this is not a useful distraction for RSE right now.
> 
>       Thanks,
>       -- Mark
> 
> 
> 
> ------------------------------
> 
> Message: 9
> Date: Wed, 30 Jul 2003 13:28:08 +0200
> From: "Julien Wajsberg" <address@hidden>
> Subject: Re: RSE's cvs import patch against the current CVS source
> To: address@hidden
> Message-ID:
>       <address@hidden>
> Content-Type: text/plain; charset=us-ascii
> 
> 
> Mark D. Baushke wrote:
> 
> 
> > Julien Wajsberg <address@hidden> writes:
> >
> > > I just saw that 'verifymsg' is processed before 'importinfo' with your
> > > patch. I don't think it is the correct behaviour :)
> >
> > I suggest that you should not depend on the ordering of the verifymsg as
> > compared to the importinfo or commitinfo.
> 
> I don't depend of the order. It just happens in our setup that verifymsg
> takes longer than other hooks.
> 
> > In client/server mode, verifymsg is processed first. In local mode,
> > verifymsg is usually called after the commitinfo has been called.
> 
> My understanding was that commitinfo/taginfo always occurs before
> verifymsg. It seems I was wrong...
> 
> > It is just as reasonable to reject a commit due to a bad log message as
> > it is to reject it because a file is being committed does not pass some
> > commitinfo check.
> 
> Yes I agree with that.
> 
> > > How about modifying this ?
> >
> > I believe it would be a waste of time. There is still an 'enhancement'
> > request http://ccvs.cvshome.org/issues/show_bug.cgi?id=27 to have the
> > verifymsg processed only once rather than for each directory...
> >
> > Somewhere before the call to do_verify ?
> >
> > I'd suggest that this is not a useful distraction for RSE right now.
> 
> If there is no advantage at doing that, surely it is a waste of time.
> But if it adds coherence, why don't you want to add this feature in the
> good way at the first time ?
> 
> Moreover, it could help (later) if you want to modify the code in order not
> to show an editor if
> commitinfo is failing, for example (just a thought).
> 
> --
> Julien
> PS: sorry Mark, I hit the wrong reply button at first...
> 
> 
> 
> 
> ------------------------------
> 
> Message: 10
> Date: Wed, 30 Jul 2003 14:51:02 +0200
> From: "Lemke, Michael  IZ/HZA-IC1" <address@hidden>
> Subject: RE: Strange diff behavior on branch
> To: "'address@hidden'" <address@hidden>
> Message-ID:
>       <address@hidden>
> Content-Type: text/plain
> 
> Not having seen a single reply I try again and ask:
> 
> > I have a file that was initially added on a branch.  If I
> > do a diff based on a date I get this:
> > 
> > $ cvs diff -D 16-jul-2003  koucopy.sh
> > cvs diff: koucopy.sh is a new entry, no comparison available
> > $ cvs stat koucopy.sh
> > ===================================================================
> > File: koucopy.sh        Status: Up-to-date
> > 
> >    Working revision:    1.1.2.12        Mon Jul 21 05:48:12 2003
> >    Repository revision: 1.1.2.12        
> > /mnt/bflow/cvs/repos/inacom/Attic/koucopy.sh,v
> >    Sticky Tag:          B_PROD (branch: 1.1.2)
> >    Sticky Date:         (none)
> >    Sticky Options:      (none)
> > $ cvs log koucopy.sh
> > 
> > RCS file: /mnt/bflow/cvs/repos/inacom/Attic/koucopy.sh,v
> > Working file: koucopy.sh
> > head: 1.1
> > branch:
> > locks: strict
> > access list:
> > symbolic names:
> >         B_PROD: 1.1.0.2
> > keyword substitution: kv
> > total revisions: 13;    selected revisions: 13
> > description:
> > ----------------------------
> > revision 1.1
> > date: 2002/12/13 14:08:29;  author: bflowadm;  state: dead;
> > branches:  1.1.2;
> > file koucopy.sh was initially added on branch B_PROD.
> > ----------------------------
> > revision 1.1.2.12
> > date: 2003/07/21 05:48:12;  author: bflowadm;  state: Exp;  
> > lines: +5 -2
> > ...
> > $ cvs -v
> > 
> > Concurrent Versions System (CVS) 1.12.1 (client/server)
> > 
> > 
> > 
> > Is this expected behavior?  Diffs on revisions (-r 1.1.2.10) work.
> > 
> > Thanks,
> > Michael

Hi M,

Try to update everythin' first, co module and you should get the
retrieving revisions if the diffs work. 
 
Cheers,
B
> 
> 
> 
> ------------------------------
> 
> _______________________________________________
> Info-cvs mailing list
> address@hidden
> http://mail.gnu.org/mailman/listinfo/info-cvs
> 
> 
> End of Info-cvs Digest, Vol 8, Issue 31
> ***************************************
>