can't tag module w/ PCL-CVS in Emacs 21.2

can't tag module w/ PCL-CVS in Emacs 21.2

[Richard V. Molen]

I am unable to tag a release using Emacs 21.2.1 PCL-CVS buffer using
the 't' command.  I've also tried to do this using a C-x v s (version
control snapshot) w/o success.

The PCL-CVS documentation suggests that it is a replacement for the VC
frontend (when using CVS).  This leads me to three questions:

1. What must I do to get the PCL-CVS *cvs* buffer's 't' command to tag
   a 'snapshot'?  Below are the details.

2. Does it make sense to use the VC front-end instead of PCL-CVS
   some/most of the time when working with a CVS backend?

3. Are there any plans to merge the two frontends?

-
|cvs-force-dir-tag's value is t
|
|Documentation:
|*If non-nil, tagging can only be applied to directories.
|Tagging should generally be applied a directory at a time,...
-

pressing 't' in  *cvs* after mouse-clicking on this spot...

...............v     
In directory .:

...here in the buffer presents this in the minibuffer...
Tag name:'
...adding a dummy tag name 'dumbd4' thus...

Tag name: dumbd4 <ret>

...yields an updated *cvs* buffer thus...


Repository : :server:user@hostname:/no/place/like/home/cvs
Module     : vendor/app
Working dir: c:/wrk/app/


In directory .:
              Up-To-Date  1.4         a.bat
.
.
.
              Up-To-Date  1.1.1.2     file.txt

--------------------- End ---------------------
-- last cmd: cvs -f tag dumbd4 . --

...showing that the cvs command was supposedly sent.
However, placing cursor on 'file.txt' then pressing '+'
to view tag tree shows only prior tags done outside Emacs.
In fact, doing it at a bash prompt thus...

$ cvs tag dumbcvs1

...then viewing the tag tree for file.txt shows that tagging
via the bash prompt works...

   Existing Tags:
   `---+-    1.1.1: IMPORT_TAG
       +---  1.1.1.1: OTHER_TAG
       `---  1.1.1.2: PRIOR_BASH_PROMPT_TAG
                      dumbcvs1


-- 
Richard V. Molen

Warning!!
Signature under construction, safety glasses required.

Re: can't tag module w/ PCL-CVS in Emacs 21.2

[Stefan Monnier <foo@acm.com>]

>>>>> "Richard" == Richard V Molen <rvmolen@bambecksystems.com> writes:
> The PCL-CVS documentation suggests that it is a replacement for the VC
> frontend (when using CVS).  This leads me to three questions:

Hmmm... I though I had been very careful to explain that it is a complement
to VC and a replacement only to VC-dired.

> 2. Does it make sense to use the VC front-end instead of PCL-CVS
>    some/most of the time when working with a CVS backend?

I use both all the time.

> 3. Are there any plans to merge the two frontends?

No.  There's a clear understanding that some kind of merging would be
great, but the two codes are very different and work very differently,
so I don't know of any actual plan to merge them.  They interact well
enough for my purposes right now.

> Repository : :server:user@hostname:/no/place/like/home/cvs

What is this :server: access method ?  I know of :ext:, :pserver:,
:fork:, and :gserver: but hadn't heard of that one yet.

> Module     : vendor/app
> Working dir: c:/wrk/app/


> In directory .:
>               Up-To-Date  1.4         a.bat
> .
> .
> .
>               Up-To-Date  1.1.1.2     file.txt

> --------------------- End ---------------------
> -- last cmd: cvs -f tag dumbd4 . --

So it seems everything is perfectly right, you've done the right thing
and Emacs seems to think we got the right result as well.

> ...showing that the cvs command was supposedly sent.
> However, placing cursor on 'file.txt' then pressing '+'
> to view tag tree shows only prior tags done outside Emacs.
> In fact, doing it at a bash prompt thus...

Hmm... That's strange!
Can you shows us the ` *cvs-tmp*' buffer right after the
command terminates (note the leading space in the buffer name) ?

> $ cvs tag dumbcvs1
> ...then viewing the tag tree for file.txt shows that tagging
> via the bash prompt works...

Can you try `cvs -f tag dumbcvs2 .' to make it even more like
what Emacs runs ?
Are you sure bash and Emacs run the same `cvs' executable ?


        Stefan

Re: can't tag module w/ PCL-CVS in Emacs 21.2

[Richard V. Molen]

"Stefan Monnier <foo@acm.com>" <monnier+gnu.emacs.help/news/@flint.cs.yale.edu> writes:

> >>>>> "Richard" == Richard V Molen <rvmolen@bambecksystems.com> writes:
> > The PCL-CVS documentation suggests that it is a replacement for the VC
> > frontend (when using CVS).  This leads me to three questions:
> 
> Hmmm... I though I had been very careful to explain that it is a complement
> to VC and a replacement only to VC-dired.

Sorry, I misread.

> 
> > 2. Does it make sense to use the VC front-end instead of PCL-CVS
> >    some/most of the time when working with a CVS backend?
> 
> I use both all the time.
> 
> > 3. Are there any plans to merge the two frontends?
> 
> No.  There's a clear understanding that some kind of merging would be
> great, but the two codes are very different and work very differently,
> so I don't know of any actual plan to merge them.  They interact well
> enough for my purposes right now.
> 
> > Repository : :server:user@hostname:/no/place/like/home/cvs
> 
> What is this :server: access method ?  I know of :ext:, :pserver:,
> :fork:, and :gserver: but hadn't heard of that one yet.

From info pages on Cygwin install...

   This info manual describes how to use and administer CVS version
1.11.2.

... *Note Connecting via rsh:* ...

   If METHOD is not specified, and the repository name contains `:',
then the default is `ext' or `server', depending on your platform; both
are described in *Note Connecting via rsh::.

WinNT Environment variables
CVS_SERVER=/usr/bin/cvs
CVSROOT=:server:user@hostname:/no/place/like/home/cvs

> 
> > Module     : vendor/app
> > Working dir: c:/wrk/app/
> 
> 
> > In directory .:
> >               Up-To-Date  1.4         a.bat
> > .
> > .
> > .
> >               Up-To-Date  1.1.1.2     file.txt
> 
> > --------------------- End ---------------------
> > -- last cmd: cvs -f tag dumbd4 . --
> 
> So it seems everything is perfectly right, you've done the right thing
> and Emacs seems to think we got the right result as well.
> 
> > ...showing that the cvs command was supposedly sent.
> > However, placing cursor on 'file.txt' then pressing '+'
> > to view tag tree shows only prior tags done outside Emacs.
> > In fact, doing it at a bash prompt thus...
> 
> Hmm... That's strange!
> Can you shows us the ` *cvs-tmp*' buffer right after the
> command terminates (note the leading space in the buffer name) ?

I find no ' *cvs-tmp*' buffer in the buffer list.  Is there some
way to turn it on?  

I'm using Emacs on WinNT with Cygwin bash if that makes a difference.
 GNU Emacs 21.2.1 (i386-msvc-nt5.1.2600) of 2002-03-19 on buffy
 Concurrent Versions System (CVS) 1.11.1p1 (client)

Here's what I found in *Messages*
Loading pcvs...
Loading easy-mmode...done
Loading easymenu...done
Loading pcvs...done
(No files need saving)
Running cvs status ...
CVS process has completed
Loading cvs-status...done
(No files need saving)
Running cvs tag ...
CVS process has completed
(No files need saving)
Running cvs status ...
CVS process has completed

> 
> > $ cvs tag dumbcvs1
> > ...then viewing the tag tree for file.txt shows that tagging
> > via the bash prompt works...
> 
> Can you try `cvs -f tag dumbcvs2 .' to make it even more like
> what Emacs runs ?
> Are you sure bash and Emacs run the same `cvs' executable ?

Yes, Cygwin mounts 
d:\tools\cygwin\bin on /usr/bin so /usr/bin = /bin
a symbolic link in /bin/cvs points to d:\bin\cvs.exe

so bash finds the link in /usr/bin/cvs and Emacs has this
exec-path...

("d:/tools/cygwin/bin"   ;find link first?
      ; ... paths
 "d:/bin"                ;or find real program first?
      ; ... paths
 "d:/tools/emacs-21.2/bin")

Also this is my WinNT path, if it matters...
Path=%SystemRoot%\system32;%SystemRoot%;%SystemRoot%\System32\Wbem;d:\bin


-- 
Richard V. Molen

Warning!!
Signature under construction, safety glasses required.

Re: can't tag module w/ PCL-CVS in Emacs 21.2

[Stefan Monnier <foo@acm.com>]

>> Hmmm... I thought I had been very careful to explain that it is a complement
>> to VC and a replacement only to VC-dired.
> Sorry, I misread.

That probably indicates that it's miswritten.  If you could provide us
with a rewritten version of the text which wouldn't confuse you, it would
be most helpful.

> I find no ' *cvs-tmp*' buffer in the buffer list.  Is there some
> way to turn it on?  

Never use the buffer list for such debugging things.  Use C-x b instead.
After all, you don't debug all the time, so you don't want those things to
clutter the buffer-list.
More specifically, the leading space acts very much like a leading .
in file names.

>> Are you sure bash and Emacs run the same `cvs' executable ?
> Yes, Cygwin mounts 
> d:\tools\cygwin\bin on /usr/bin so /usr/bin = /bin
> a symbolic link in /bin/cvs points to d:\bin\cvs.exe

Instead of reasoning I generally prefer to replace the executable
with another one and see if it is changed in the same way in both
cases.  I don't trust my ability to reason about such things.


        Stefan

Re: can't tag module w/ PCL-CVS in Emacs 21.2

[Richard V. Molen]

Sorry for being slow to respond, I've been having some hardware
problems.


"Stefan Monnier <foo@acm.com>"
<monnier+gnu.emacs.help/news/@flint.cs.yale.edu> writes:

> >> Hmmm... I thought I had been very careful to explain that it is a
> >> complement to VC and a replacement only to VC-dired.  > > Sorry,
> >> I misread.
> 
> That probably indicates that it's miswritten.  If you could provide
> us with a rewritten version of the text which wouldn't confuse you,
> it would be most helpful.
>

You honor me, and normally I would accept such an offer, but the
section 'About PCL-CVS' seems pretty clear as I re-read it.  It is
concise in that PCL-CVS is a replacement for VC-dired (note the
'dired', not the entire VC).  However, I do wonder what the 'PCL'
means.

> 
> >> Are you sure bash and Emacs run the same `cvs' executable ?
> > Yes, Cygwin mounts 
> > d:\tools\cygwin\bin on /usr/bin so /usr/bin = /bin
> > a symbolic link in /bin/cvs points to d:\bin\cvs.exe
> 
> Instead of reasoning I generally prefer to replace the executable
> with another one and see if it is changed in the same way in both
> cases.  I don't trust my ability to reason about such things.

I agree, so I've simplified by using only d:\tools\cygwin\bin\cvs.exe
(no links or shortcuts).  I cut 'exec-path' to two paths in my .emacs.
I also tried running 'emacs --no-init-file' & made a sample repository
on the local hard drive to avoid the :server: method.

Summarizing, tagging with PCL-CVS works great on RH7.3 Emacs 21.2 by
placing the cursor on the spot just past '.:' then pressing 't' etc.
So I know I'm doing that part right.

And, on a WinNT box, I can checkout a directory, do directory status,
examine, use *cvs* buffer to commit, view log, view status of files,
use C-x C-q to commit files -- PCL-CVS & that part of VC work great.

However, after simplifying my setup by ensuring use of only one
executable in one location with a minimal path and using 'emacs
--no-init-file', making a dummy repository on the local hard drive (d:)
and avoiding the :server: method, " *cvs-tmp*" reports...

cvs.exe tag: nothing known about pads
cvs.exe [tag aborted]: correct the above errors first!

...tagging works fine with bash or even the native OS 'cmd.exe'
thus...

/cygdrive/d/tools/cygwin/bin/cvs -f tag -c bashtag2  <- ala bash
d:\tools\cygwin\bin\cvs tag -c cmd_exetag1           <- ala cmd.exe

the CVS subdirectory in the 'sandbox' looks like this.
=============================================
$ cd /home/ric/pads
$ cd CVS
$ cat Repository
pads
$ cat Root
/home/ric/wrk/cvs
$ cat Entries
/HTR8B.txt/1.1.1.1/Mon Oct 21 20:13:14 2002//
/blr8c.txt/1.1.1.1/Tue Oct 22 18:51:44 2002//
/cocal.txt/1.1.1.1/Fri Sep  6 19:12:18 2002//
/emacs.txt/1.1.1.1/Fri Sep  6 19:10:00 2002//
/htr8f.txt/1.1.1.1/Tue Oct  1 15:14:20 2002//
/legacy.txt/1.1.1.1/Mon Sep 16 19:34:26 2002//
/scsi.txt/1.1.1.1/Mon Oct 14 16:42:44 2002//
/misc.txt/1.2/Tue Oct 22 21:11:50 2002//
D
===============================================


Thanks for your interest; any other ideas I can try?


-- 
Richard V. Molen

Warning!!
Signature under construction, safety glasses required.

Re: can't tag module w/ PCL-CVS in Emacs 21.2

[Stefan Monnier <foo@acm.com>]

> You honor me, and normally I would accept such an offer, but the
> section 'About PCL-CVS' seems pretty clear as I re-read it.  It is
> concise in that PCL-CVS is a replacement for VC-dired (note the
> 'dired', not the entire VC).

In that case...

> However, I do wonder what the 'PCL' means.

I can't give you an authoritative answer, but the original author
is Per Cederqvist and I have always assumed that it explained the
P and the C, although the L is still a mystery.

> However, after simplifying my setup by ensuring use of only one
> executable in one location with a minimal path and using 'emacs
> --no-init-file', making a dummy repository on the local hard drive (d:)
> and avoiding the :server: method, " *cvs-tmp*" reports...

> cvs.exe tag: nothing known about pads
> cvs.exe [tag aborted]: correct the above errors first!

Is that the complete content of ` *cvs-tmp*' ?
What was the command executed (and more generally the content of *cvs*) ?

> ...tagging works fine with bash or even the native OS 'cmd.exe' thus...
> /cygdrive/d/tools/cygwin/bin/cvs -f tag -c bashtag2  <- ala bash
> d:\tools\cygwin\bin\cvs tag -c cmd_exetag1           <- ala cmd.exe

IIRC, PCL-CVS puts a dot over there, so could you try out
`cvs -f tag bashtag2 .' ?


        Stefan

Re: can't tag module w/ PCL-CVS in Emacs 21.2

[Richard V. Molen]

"Stefan Monnier <foo@acm.com>"
<monnier+gnu.emacs.help/news/@flint.cs.yale.edu> writes:
 
> Is that the complete content of ` *cvs-tmp*' ?

Yes.

> What was the command executed (and more generally the content of
> *cvs*) ?

...Selecting menubar Tools.PCL-CVS.Directory Status...

Repository : /home/ric/wrk/cvs
Module     : pads
Working dir: ~/wrk/tmp/pads/



In directory .:
              Up-To-Date  1.1.1.1     HTR8B.txt
              Up-To-Date  1.1.1.1     blr8c.txt
              Up-To-Date  1.1.1.1     cocal.txt
              Up-To-Date  1.1.1.1     emacs.txt
              Up-To-Date  1.1.1.1     htr8f.txt
              Up-To-Date  1.1.1.1     legacy.txt
              Up-To-Date  1.2         misc.txt
              Up-To-Date  1.1.1.1     scsi.txt

--------------------- End ---------------------
-- last cmd: cvs -f status --

...using mouse to position cursor just after the '.:' 
and pressing 't' results in the following...


Repository : /home/ric/wrk/cvs
Module     : pads
Working dir: ~/wrk/tmp/pads/



In directory .:
              Up-To-Date  1.1.1.1     HTR8B.txt
              Up-To-Date  1.1.1.1     blr8c.txt
              Up-To-Date  1.1.1.1     cocal.txt
              Up-To-Date  1.1.1.1     emacs.txt
              Up-To-Date  1.1.1.1     htr8f.txt
              Up-To-Date  1.1.1.1     legacy.txt
              Up-To-Date  1.2         misc.txt
              Up-To-Date  1.1.1.1     scsi.txt

--------------------- End ---------------------
-- last cmd: cvs -f tag emacstag3 . --

...with a ' *cvs-tmp*' file containing this...

cvs.exe tag: nothing known about pads
cvs.exe [tag aborted]: correct the above errors first!

...that's all...

> IIRC, PCL-CVS puts a dot over there, so could you try out
> `cvs -f tag bashtag2 .' ?

Yes, that reproduces the problem.  Here's what I did at the bash
prompt...

$ cvs -f tag bashtag3 .

...here's what I got on the console...

cvs.exe tag: nothing known about pads
cvs.exe [tag aborted]: correct the above errors first!

...It's strange that this seems to work on the RH 7.3 box, but not on
the WinNT box.  My RH box is down right now or I'd try it there.

-- 
Richard V. Molen

Warning!!
Signature under construction, safety glasses required.

Re: can't tag module w/ PCL-CVS in Emacs 21.2

[Richard V. Molen]

rvmolen@bambecksystems.com (Richard V. Molen) writes:

> "Stefan Monnier <foo@acm.com>"
> <monnier+gnu.emacs.help/news/@flint.cs.yale.edu> writes:
>  
> > Is that the complete content of ` *cvs-tmp*' ?
> 
> Yes.
> 
> > What was the command executed (and more generally the content of
> > *cvs*) ?
> 
> ...Selecting menubar Tools.PCL-CVS.Directory Status...
> 
> Repository : /home/ric/wrk/cvs
> Module     : pads
> Working dir: ~/wrk/tmp/pads/
> 
> 
> 
> In directory .:
>               Up-To-Date  1.1.1.1     HTR8B.txt
>               Up-To-Date  1.1.1.1     blr8c.txt
>               Up-To-Date  1.1.1.1     cocal.txt
>               Up-To-Date  1.1.1.1     emacs.txt
>               Up-To-Date  1.1.1.1     htr8f.txt
>               Up-To-Date  1.1.1.1     legacy.txt
>               Up-To-Date  1.2         misc.txt
>               Up-To-Date  1.1.1.1     scsi.txt
> 
> --------------------- End ---------------------
> -- last cmd: cvs -f status --
> 
> ...using mouse to position cursor just after the '.:' 
> and pressing 't' results in the following...
> 
> 
> Repository : /home/ric/wrk/cvs
> Module     : pads
> Working dir: ~/wrk/tmp/pads/
> 
> 
> 
> In directory .:
>               Up-To-Date  1.1.1.1     HTR8B.txt
>               Up-To-Date  1.1.1.1     blr8c.txt
>               Up-To-Date  1.1.1.1     cocal.txt
>               Up-To-Date  1.1.1.1     emacs.txt
>               Up-To-Date  1.1.1.1     htr8f.txt
>               Up-To-Date  1.1.1.1     legacy.txt
>               Up-To-Date  1.2         misc.txt
>               Up-To-Date  1.1.1.1     scsi.txt
> 
> --------------------- End ---------------------
> -- last cmd: cvs -f tag emacstag3 . --
> 
> ...with a ' *cvs-tmp*' file containing this...
> 
> cvs.exe tag: nothing known about pads
> cvs.exe [tag aborted]: correct the above errors first!
> 
> ...that's all...
> 
> > IIRC, PCL-CVS puts a dot over there, so could you try out
> > `cvs -f tag bashtag2 .' ?
> 
> Yes, that reproduces the problem.  Here's what I did at the bash
> prompt...
> 
> $ cvs -f tag bashtag3 .
> 
> ...here's what I got on the console...
> 
> cvs.exe tag: nothing known about pads
> cvs.exe [tag aborted]: correct the above errors first!
> 
> ...It's strange that this seems to work on the RH 7.3 box, but not on
> the WinNT box.  My RH box is down right now or I'd try it there.

Oops, I installed RH 8.0 on that box and now I have the same problem.
Here the directory status...

Repository : /vc/user
Module     : linux/user
Working dir: ~/

In directory .:
              Up-To-Date  1.1.1.1     .bash_logout
.
.
.
--------------------- End ---------------------
-- last cmd: cvs -f -z6 status --


...After using mouse to put cursor '.: '<-here and pressing 't'...

.
.
.
--------------------- End ---------------------
-- last cmd: cvs -f -z6 tag emacsTag1 . --


...peaking at ' *cvs-tmp*' buffer shows...

cvs tag: nothing known about config
cvs [tag aborted]: correct the above errors first!

...'M-x version' shows...

GNU Emacs 21.2.1 (i686-pc-linux-gnu, X toolkit, Xaw3d scroll bars) of 2002-08-28 on astest

...'cvs status -v .bash_logout shows...

ric@tsunami ric]$ cvs status -v .bash_logout
===================================================================
File: .bash_logout      Status: Up-to-date

   Working revision:    1.1.1.1 Fri Apr 12 16:09:53 2002
   Repository revision: 1.1.1.1 /vc/user/linux/user/.bash_logout,v
   Sticky Tag:          (none)
   Sticky Date:         (none)
   Sticky Options:      (none)

   Existing Tags:
        start                           (revision: 1.1.1.1)
        linux_user                      (branch: 1.1.1)

...Any ideas?

It appears to that cvs does not like the '.' in the tag command.
Can Emacs be configured to remove it or am I using the command
incorrectly?

Thanks for your patience.

-- 
Richard V. Molen

Warning!!
Signature under construction, safety glasses required.

Re: can't tag module w/ PCL-CVS in Emacs 21.2

[Stefan Monnier <foo@acm.com>]

>>>>> "Richard" == Richard V Molen <rvmolen@bambecksystems.com> writes:
> Repository : /vc/user
> Module     : linux/user
> Working dir: ~/

> In directory .:
>               Up-To-Date  1.1.1.1     .bash_logout
> .
> .
> .
> --------------------- End ---------------------
> -- last cmd: cvs -f -z6 status --
[...]
> -- last cmd: cvs -f -z6 tag emacsTag1 . --
[...]
> ...peaking at ' *cvs-tmp*' buffer shows...
> cvs tag: nothing known about config
> cvs [tag aborted]: correct the above errors first!

What is this `config' thingy ?
Have you tried to run `cvs -f -z6 tag emacsTag1 .' and
`cvs -f -z6 tag emacsTag1' from the shell ?

This thing has been working fine under unix for a while, so I can
believe that the . is posing problems under w32, but under unix it's
highly suspect.


        Stefan

Re: can't tag module w/ PCL-CVS in Emacs 21.2

[Richard V. Molen]

"Stefan Monnier <foo@acm.com>" <monnier+gnu.emacs.help/news/@flint.cs.yale.edu> writes:

> >>>>> "Richard" == Richard V Molen <rvmolen@bambecksystems.com> writes:
> > Repository : /vc/user
> > Module     : linux/user
> > Working dir: ~/
> 
> > In directory .:
> >               Up-To-Date  1.1.1.1     .bash_logout
> > .
> > .
> > .
> > --------------------- End ---------------------
> > -- last cmd: cvs -f -z6 status --
> [...]
> > -- last cmd: cvs -f -z6 tag emacsTag1 . --
> [...]
> > ...peaking at ' *cvs-tmp*' buffer shows...
> > cvs tag: nothing known about config
> > cvs [tag aborted]: correct the above errors first!
> 
> What is this `config' thingy ?
> Have you tried to run `cvs -f -z6 tag emacsTag1 .' and
> `cvs -f -z6 tag emacsTag1' from the shell ?

Oops!? Please accept my apologies, the Linux example must be from a broken
repository. When I re'checkout'ed repository the new checkout works fine.
Also, of course, both the above lines work in bash & Emacs tags fine.

However, in WinNT I get this with a newly created dummy repository.
(Pardon the Halloween humor, the sandbox directory is 'coffin')...

$ cvs -f -z6 tag bashtag000 .
cvs.exe tag: nothing known about coffin
cvs.exe [tag aborted]: correct the above errors first!

$ cvs -f -z6 tag bashtag001
cvs.exe tag: Tagging .
T corpse
T treasure

$ cvs -f status -v corpse
===================================================================
File: corpse            Status: Up-to-date

   Working revision:    1.1.1.1 Tue Oct 29 21:20:02 2002
   Repository revision: 1.1.1.1 /home/ric/vc/tomb/coffin/corpse,v
   Sticky Tag:          (none)
   Sticky Date:         (none)
   Sticky Options:      (none)

   Existing Tags:
        bashtag001                      (revision: 1.1.1.1)
        HEARSE_01                       (revision: 1.1.1.1)
        ACME                            (branch: 1.1.1)

$ cat CVS/Root
/home/user/vc/tomb

...I wonder if PCL-CVS dropped the '.' in the tag command if that 
would work?  Is there an easy way to try?...

-- 
Richard V. Molen

Warning!!
Signature under construction, safety glasses required.

Re: can't tag module w/ PCL-CVS in Emacs 21.2

[Stefan Monnier <foo@acm.com>]

>>>>> "Richard" == Richard V Molen <rvmolen@bambecksystems.com> writes:
> $ cvs -f -z6 tag bashtag000 .
> cvs.exe tag: nothing known about coffin
> cvs.exe [tag aborted]: correct the above errors first!
> $ cvs -f -z6 tag bashtag001
> cvs.exe tag: Tagging .
[...]

So that's what it is, huh!
I'll see how I can get PCL-CVS to avoid passing this `.' thing.
Of course, I'd argue it's a bug in CVS (after all, W32 also has the
. and .. convention, just like unix).
Thanks for your help.  I'll send/post a patch as soon as I have it.


        Stefan

Re: can't tag module w/ PCL-CVS in Emacs 21.2

[Richard V. Molen]

"Stefan Monnier <foo@acm.com>" <monnier+gnu.emacs.help/news/@flint.cs.yale.edu> writes:

> >>>>> "Richard" == Richard V Molen <rvmolen@bambecksystems.com> writes:
> > $ cvs -f -z6 tag bashtag000 .
> > cvs.exe tag: nothing known about coffin
> > cvs.exe [tag aborted]: correct the above errors first!
> > $ cvs -f -z6 tag bashtag001
> > cvs.exe tag: Tagging .
> [...]
> 
> So that's what it is, huh!
> I'll see how I can get PCL-CVS to avoid passing this `.' thing.
> Of course, I'd argue it's a bug in CVS (after all, W32 also has the
> . and .. convention, just like unix).
> Thanks for your help.  I'll send/post a patch as soon as I have it.

Thank you, Steve, for maintaining this nice package.  I'll be glad to
help test out the patch on my box.

-- 
Richard V. Molen

Warning!!
Signature under construction, safety glasses required.