help with vc.el and CVS

help with vc.el and CVS

Post by Derek Scherg » Sat, 13 Feb 1999 04:00:00



I've been trying to make use of the VC facilities provided in vc.el
with CVS and am having some problems.

Some things seem to work fine, for example while editing a file under
CVS control C-x v = (vc-diff) gives a diff of the current working file
with the latest revision and C-x v v (vc-next-action) does ask for a
log entry (comment) and commits the changes to the file fine. My
problems seem to be more centered around vc-dired-mode.

Doing C-x v d (vc-directory) does give me a listing of the directory
with cvs status info for some files, but only shows those files that
have been changed, and does not show files with no changes. Is this
right? Suppose I want to make changes to an unchanged file?!?

Once the vc-directory is up the mode line actually shows (Dired by
name) instead of (Dired under VC) which is what I expect it should say
if I was really in vc-dired-mode. Note that this mode is derived from
dired-mode -- have a look in vc.el. It seems that I am actually not in
vc-dired-mode since vc-dired-mode is nil, instead of t as it should
probably be.

I've also tried M-x vc-dired-mode instead of C-x v d (vc-directory) to
try and get in the mode properly but there are still problems. Again,
according to vc.el the vc-dired-mode-map should map = to be vc-diff
and C-x v v (vc-next-action) should (I think) ask for a log entry and
then commit all the changed files in the directory. Instead, for
vc-next-action after vc-directory, I get the message "Buffer foo is
not associated with a file" in the minibuffer, where foo is the
basename of the directory I'm in. If I do vc-dired-mode and then
vc-next-action I get the message "dired-subdir-alist seems to be
mangled" in the minibuffer which doesn't sound too good.

Note that this is with XEmacs 20.4 under both Linux (redhat 5.2 kernel
2.1.132) and AIX 4.3.1.

I've looked through vc.el and friends but I don't claim to understand
all of it -- yet!

I'm assuming that this must be reasonably popular stuff with all the
open source projects that are using CVS and very likely does work
properly. Which means, of course, that I'm probably just doing
something wrong. Any words of wisdom from someone who knows something
I don't would be greatly appreciated!
--
Cheers,
Derek
_____________________________________________________________________
Derek Scherger                         Echologic Software Corporation

[ Attached Message ]

From:
To:
Date: Fri, 05 Feb 1999 12:10:40 -0700
Local: Fri, Feb 5 1999 2:10 pm
Subject: help with vc.el and CVS

4.3.1 binary I now have XEmacs again and can get back to working
productively! (going back to vi is hard...)

I've been trying to make use of the VC facilities provided in vc.el
with CVS and am having some problems.

Some things seem to work fine, for example while editing a file under
CVS control C-x v = (vc-diff) gives a diff of the current working file
with the latest revision and C-x v v (vc-next-action) does ask for a
log entry (comment) and commits the changes to the file fine. My
problems seem to be more centered around vc-dired-mode.

Doing C-x v d (vc-directory) does give me a listing of the directory
with cvs status info for some files, but only shows those files that
have been changed, and does not show files with no changes. Is this
right? Suppose I want to make changes to an unchanged file?!?

Once the vc-directory is up the mode line actually shows (Dired by
name) instead of (Dired under VC) which is what I expect it should say
if I was really in vc-dired-mode. Note that this mode is derived from
dired-mode -- have a look in vc.el. It seems that I am actually not in
vc-dired-mode since vc-dired-mode is nil, instead of t as it should
probably be.

I've also tried M-x vc-dired-mode instead of C-x v d (vc-directory) to
try and get in the mode properly but there are still problems. Again,
according to vc.el the vc-dired-mode-map should map = to be vc-diff
and C-x v v (vc-next-action) should (I think) ask for a log entry and
then commit all the changed files in the directory. Instead, for
vc-next-action after vc-directory, I get the message "Buffer foo is
not associated with a file" in the minibuffer, where foo is the
basename of the directory I'm in. If I do vc-dired-mode and then
vc-next-action I get the message "dired-subdir-alist seems to be
mangled" in the minibuffer which doesn't sound too good.

Note that this is with XEmacs 20.4 under both Linux (redhat 5.2 kernel
2.1.132) and AIX 4.3.1.

I've looked through vc.el and friends but I don't claim to understand
all of it -- yet!

I'm assuming that this must be reasonably popular stuff with all the
open source projects that are using CVS and very likely does work
properly. Which means, of course, that I'm probably just doing
something wrong. Any words of wisdom from someone who knows something
I don't would be greatly appreciated!
--
Cheers,
Derek
_____________________________________________________________________
Derek Scherger                         Echologic Software Corporation

 
 
 

1. help with vc.el and CVS


4.3.1 binary I now have XEmacs again and can get back to working
productively! (going back to vi is hard...)

I've been trying to make use of the VC facilities provided in vc.el
with CVS and am having some problems.

Some things seem to work fine, for example while editing a file under
CVS control C-x v = (vc-diff) gives a diff of the current working file
with the latest revision and C-x v v (vc-next-action) does ask for a
log entry (comment) and commits the changes to the file fine. My
problems seem to be more centered around vc-dired-mode.

Doing C-x v d (vc-directory) does give me a listing of the directory
with cvs status info for some files, but only shows those files that
have been changed, and does not show files with no changes. Is this
right? Suppose I want to make changes to an unchanged file?!?

Once the vc-directory is up the mode line actually shows (Dired by
name) instead of (Dired under VC) which is what I expect it should say
if I was really in vc-dired-mode. Note that this mode is derived from
dired-mode -- have a look in vc.el. It seems that I am actually not in
vc-dired-mode since vc-dired-mode is nil, instead of t as it should
probably be.

I've also tried M-x vc-dired-mode instead of C-x v d (vc-directory) to
try and get in the mode properly but there are still problems. Again,
according to vc.el the vc-dired-mode-map should map = to be vc-diff
and C-x v v (vc-next-action) should (I think) ask for a log entry and
then commit all the changed files in the directory. Instead, for
vc-next-action after vc-directory, I get the message "Buffer foo is
not associated with a file" in the minibuffer, where foo is the
basename of the directory I'm in. If I do vc-dired-mode and then
vc-next-action I get the message "dired-subdir-alist seems to be
mangled" in the minibuffer which doesn't sound too good.

Note that this is with XEmacs 20.4 under both Linux (redhat 5.2 kernel
2.1.132) and AIX 4.3.1.

I've looked through vc.el and friends but I don't claim to understand
all of it -- yet!

I'm assuming that this must be reasonably popular stuff with all the
open source projects that are using CVS and very likely does work
properly. Which means, of course, that I'm probably just doing
something wrong. Any words of wisdom from someone who knows something
I don't would be greatly appreciated!
--
Cheers,
Derek
_____________________________________________________________________
Derek Scherger                         Echologic Software Corporation

2. rprinter on a Win95 workstation

3. cvs-1.6 + pcl-cvs/vc: cannot edit files

4. Tail-wags-dog approach to printer selection!

5. two bugs in XEmacs 19.12 .el files (vc.el & appt.el)

6. Where's the OS2KRNL fix for swapfile growth?

7. Y2K Problem: vc-hooks.el, `vc-parse-buffer'

8. samba-1.9.15p8 and AIX 4.1.3 (Netmask)

9. vc.el (v 1.19 from _pkg.el)

10. "shadowfile.el" versus "vc.el"

11. VC and CVS

12. Error on vc-checkin using cvs

13. vc-backend-checkin problem with CVS