4:09 p.m.
On Sunday 19 August 2007, Adrian Aichner wrote:
Ville Skyttä <scop(a)xemacs.org> writes:
> Hello,
Hi Ville!
> After three and half years of snail-pace porting work, I think the
> new vc from the unsupported area is now in a state where it makes
> sense to start pushing it towards the regular XEmacs packages tree.
Whow, has it really been that long?
Unfortunately, yes. Well, or at least that's when I started keeping a
ChangeLog for the sync, it might have actually been even a bit longer :P
For tramp I had to apply following quick patch but the issue needs
to
be investigated further.
Why is following patch not necessary with the latest official vc
package?
Can't say for certain offhand, but comments slightly above the line you
patched include:
;; The following defadvice is no longer necessary after changes in VC
;; on 2006-01-25, Andre.
[...]
;; With Emacs 22, the definition of `vc-user-login-name' has been
;; changed. It doesn't need to be adviced any longer.
...so I suppose the whole (defadvice vc-user-login-name) no longer applies to
the new vc. If it's not needed, perhaps there's a way to detect it and skip
the code altogether? There's already a (unless (fboundp 'process-file))
around it, but perhaps that's for some other purpose?
For easypg, following advice was suggested by Daiki and slightly
modified by me. It does not work first time accessing a .gpg file via
C-x v ~ (vc-version-other-window)
I have no experience with easypg, but let me know if something in vc needs to
change to accommodate its needs.
_______________________________________________
XEmacs-Beta mailing list
XEmacs-Beta(a)xemacs.org
http://calypso.tux.org/cgi-bin/mailman/listinfo/xemacs-beta
6:42 a.m.
New subject: Pushing new vc towards regular packages tree
Ville Skyttä <scop(a)xemacs.org> writes:
Hi,
> For tramp I had to apply following quick patch but the issue
needs to
> be investigated further.
>
> Why is following patch not necessary with the latest official vc
> package?
The patch is OK. `uid' is not known inside the advice, and it is an
optional parameter to `tramp-handle-vc-user-login-name'. I'll apply it
in Tramp CVS. Somebody else might sync it with the XEmacs Tramp
repository (I've stalled myself, you know).
Can't say for certain offhand, but comments slightly above the
line you
patched include:
;; The following defadvice is no longer necessary after changes in VC
;; on 2006-01-25, Andre.
[...]
;; With Emacs 22, the definition of `vc-user-login-name' has been
;; changed. It doesn't need to be adviced any longer.
...so I suppose the whole (defadvice vc-user-login-name) no longer applies to
the new vc.
Yes, indeed.
If it's not needed, perhaps there's a way to detect it and
skip the
code altogether? There's already a (unless (fboundp 'process-file))
around it, but perhaps that's for some other purpose?
(fboundp 'process-file) is a check for "GNU Emacs 22 existence", because
it is a new function introduced there. Implicitely, it means "new vc
existence". If there is a better check for the existence of "new vc", we
should apply that instead.
Btw, it would be nice if XEmacs could introduce the new functions
`process-file` and `start-file-process'. They are file name handler
based counterparts of `call-process' and `start-process', allowing to
run processes on remote machines. At least `process-file' is used in
vc.el of GNU Emacs 22.
Best regards, Michael.
_______________________________________________
XEmacs-Beta mailing list
XEmacs-Beta(a)xemacs.org
http://calypso.tux.org/cgi-bin/mailman/listinfo/xemacs-beta
6:17 p.m.
New subject: Pushing new vc towards regular packages tree
On Monday 20 August 2007, Michael Albinus wrote:
> Can't say for certain offhand, but comments slightly above the line you
> patched include:
>
> ;; The following defadvice is no longer necessary after changes in VC
> ;; on 2006-01-25, Andre.
> [...]
> ;; With Emacs 22, the definition of `vc-user-login-name' has been
> ;; changed. It doesn't need to be adviced any longer.
>
> ...so I suppose the whole (defadvice vc-user-login-name) no longer
> applies to the new vc.
Yes, indeed.
> If it's not needed, perhaps there's a way to detect it and skip the
> code altogether? There's already a (unless (fboundp 'process-file))
> around it, but perhaps that's for some other purpose?
(fboundp 'process-file) is a check for "GNU Emacs 22 existence", because
it is a new function introduced there. Implicitely, it means "new vc
existence". If there is a better check for the existence of "new vc", we
should apply that instead.
I think checking (fboundp 'vc-find-version) would be better. It exists in the
XEmacs "new vc" as well as in the one in GNU Emacs 22, but does not exist in
the XEmacs "old vc" nor in the GNU Emacs 21.3 one:
http://cvs.savannah.gnu.org/viewvc/emacs/lisp/vc.el?root=emacs&view=d...
Patch attached.
_______________________________________________
XEmacs-Beta mailing list
XEmacs-Beta(a)xemacs.org
http://calypso.tux.org/cgi-bin/mailman/listinfo/xemacs-beta
8:06 p.m.
New subject: Pushing new vc towards regular packages tree
Ville Skyttä <scop(a)xemacs.org> writes:
> (fboundp 'process-file) is a check for "GNU Emacs 22
existence", because
> it is a new function introduced there. Implicitely, it means "new vc
> existence". If there is a better check for the existence of "new vc",
we
> should apply that instead.
I think checking (fboundp 'vc-find-version) would be better. It exists in the
XEmacs "new vc" as well as in the one in GNU Emacs 22, but does not exist in
the XEmacs "old vc" nor in the GNU Emacs 21.3 one:
http://cvs.savannah.gnu.org/viewvc/emacs/lisp/vc.el?root=emacs&view=d...
Patch attached.
Looks OK to me. If you install it in Tramp's XEmacs package repository
(maybe a little bit more commentary), I'll be happy to apply it to
upstream Tramp as well.
Best regards, Michael.
_______________________________________________
XEmacs-Beta mailing list
XEmacs-Beta(a)xemacs.org
http://calypso.tux.org/cgi-bin/mailman/listinfo/xemacs-beta
6071
days inactive
6089
days old
3 comments
2 participants
participants (2)
-
Michael Albinus -
Ville Skyttä