* bug#8054: 24.0.50; Who ever heard of a version number that doesn't change all year?
2011-02-16 12:54 ` bug#8054: 24.0.50; Who ever heard of a version number that doesn't change all year? jidanni
@ 2011-02-16 13:10 ` Lennart Borgman
2011-02-19 22:14 ` Chong Yidong
2011-02-26 3:04 ` debbugs-submit-bounces
` (2 subsequent siblings)
3 siblings, 1 reply; 18+ messages in thread
From: Lennart Borgman @ 2011-02-16 13:10 UTC (permalink / raw)
To: jidanni; +Cc: 8054
On Wed, Feb 16, 2011 at 1:54 PM, <jidanni@jidanni.org> wrote:
> Gentlemen, let's face it, your version number,
> emacs-version "24.0.50.1"
> is a total disaster for use in e.g., report-emacs-bug.
Yepp. I have suggested to use bzr ids instead. Which is what I am
using in my patched version of EmacsW32 (so I have some code for w32
for it).
I think this should be added to the makefiles.
^ permalink raw reply [flat|nested] 18+ messages in thread* bug#8054: 24.0.50; Who ever heard of a version number that doesn't change all year?
2011-02-16 13:10 ` Lennart Borgman
@ 2011-02-19 22:14 ` Chong Yidong
2011-02-19 22:22 ` Lennart Borgman
0 siblings, 1 reply; 18+ messages in thread
From: Chong Yidong @ 2011-02-19 22:14 UTC (permalink / raw)
To: Lennart Borgman; +Cc: 8054, jidanni
Lennart Borgman <lennart.borgman@gmail.com> writes:
> On Wed, Feb 16, 2011 at 1:54 PM, <jidanni@jidanni.org> wrote:
>> Gentlemen, let's face it, your version number,
>> emacs-version "24.0.50.1"
>> is a total disaster for use in e.g., report-emacs-bug.
>
> Yepp. I have suggested to use bzr ids instead. Which is what I am
> using in my patched version of EmacsW32 (so I have some code for w32
> for it).
>
> I think this should be added to the makefiles.
It would be rather annoying to change version.el or other working file
everything we make a bzr commit, so that's not a good idea.
What we could do is to check if Emacs is being run from inside a bzr
working directory bound to a savannah.gnu.org location (e.g. trunk), and
use vc-bzr-current-revision to find and report that revision number.
But such a feature would be pretty limited: it would not apply to anyone
working in a feature branch, nor would it apply to those running a "make
install"ed Emacs. So it doesn't seem worth the effort.
^ permalink raw reply [flat|nested] 18+ messages in thread
* bug#8054: 24.0.50; Who ever heard of a version number that doesn't change all year?
2011-02-19 22:14 ` Chong Yidong
@ 2011-02-19 22:22 ` Lennart Borgman
2011-02-19 22:23 ` Chong Yidong
0 siblings, 1 reply; 18+ messages in thread
From: Lennart Borgman @ 2011-02-19 22:22 UTC (permalink / raw)
To: Chong Yidong; +Cc: 8054, jidanni
On Sat, Feb 19, 2011 at 11:14 PM, Chong Yidong <cyd@stupidchicken.com> wrote:
> Lennart Borgman <lennart.borgman@gmail.com> writes:
>
>> On Wed, Feb 16, 2011 at 1:54 PM, <jidanni@jidanni.org> wrote:
>>> Gentlemen, let's face it, your version number,
>>> emacs-version "24.0.50.1"
>>> is a total disaster for use in e.g., report-emacs-bug.
>>
>> Yepp. I have suggested to use bzr ids instead. Which is what I am
>> using in my patched version of EmacsW32 (so I have some code for w32
>> for it).
>>
>> I think this should be added to the makefiles.
>
> It would be rather annoying to change version.el or other working file
> everything we make a bzr commit, so that's not a good idea.
>
> What we could do is to check if Emacs is being run from inside a bzr
> working directory bound to a savannah.gnu.org location (e.g. trunk), and
> use vc-bzr-current-revision to find and report that revision number.
Yes, something like that is what I have in mind.
> But such a feature would be pretty limited: it would not apply to anyone
> working in a feature branch,
Of course it should be extended to that.
> nor would it apply to those running a "make
> install"ed Emacs.
Why not?
^ permalink raw reply [flat|nested] 18+ messages in thread
* bug#8054: 24.0.50; Who ever heard of a version number that doesn't change all year?
2011-02-19 22:22 ` Lennart Borgman
@ 2011-02-19 22:23 ` Chong Yidong
2011-02-19 22:35 ` Lennart Borgman
2011-11-22 0:48 ` Glenn Morris
0 siblings, 2 replies; 18+ messages in thread
From: Chong Yidong @ 2011-02-19 22:23 UTC (permalink / raw)
To: Lennart Borgman; +Cc: 8054, jidanni
Lennart Borgman <lennart.borgman@gmail.com> writes:
>> But such a feature would be pretty limited: it would not apply to anyone
>> working in a feature branch,
>
> Of course it should be extended to that.
Revision numbers on feature branches are meaningless.
^ permalink raw reply [flat|nested] 18+ messages in thread
* bug#8054: 24.0.50; Who ever heard of a version number that doesn't change all year?
2011-02-19 22:23 ` Chong Yidong
@ 2011-02-19 22:35 ` Lennart Borgman
2011-11-22 0:48 ` Glenn Morris
1 sibling, 0 replies; 18+ messages in thread
From: Lennart Borgman @ 2011-02-19 22:35 UTC (permalink / raw)
To: Chong Yidong; +Cc: 8054, jidanni
On Sat, Feb 19, 2011 at 11:23 PM, Chong Yidong <cyd@stupidchicken.com> wrote:
> Lennart Borgman <lennart.borgman@gmail.com> writes:
>
>>> But such a feature would be pretty limited: it would not apply to anyone
>>> working in a feature branch,
>>
>> Of course it should be extended to that.
>
> Revision numbers on feature branches are meaningless.
I am not sure of that.
And can't you get the trunk's rev number?
^ permalink raw reply [flat|nested] 18+ messages in thread
* bug#8054: 24.0.50; Who ever heard of a version number that doesn't change all year?
2011-02-19 22:23 ` Chong Yidong
2011-02-19 22:35 ` Lennart Borgman
@ 2011-11-22 0:48 ` Glenn Morris
2011-11-23 13:03 ` Juanma Barranquero
2011-11-24 6:20 ` Chong Yidong
1 sibling, 2 replies; 18+ messages in thread
From: Glenn Morris @ 2011-11-22 0:48 UTC (permalink / raw)
To: 8054
Here's a patch to define an `emacs-bzr-version' during dumping and
include it in bug reports.
=== modified file 'lisp/loadup.el'
*** lisp/loadup.el 2011-07-06 22:43:48 +0000
--- lisp/loadup.el 2011-11-22 00:44:13 +0000
***************
*** 259,264 ****
--- 259,266 ----
(versions (mapcar (function (lambda (name)
(string-to-number (substring name (length base)))))
files)))
+ (setq emacs-bzr-version (condition-case nil (emacs-bzr-get-version)
+ (error nil)))
;; `emacs-version' is a constant, so we shouldn't change it with `setq'.
(defconst emacs-version
(format "%s.%d"
=== modified file 'lisp/mail/emacsbug.el'
*** lisp/mail/emacsbug.el 2011-09-23 23:47:27 +0000
--- lisp/mail/emacsbug.el 2011-11-22 00:25:12 +0000
***************
*** 224,229 ****
--- 224,231 ----
(add-text-properties (1+ user-point) (point) prompt-properties)
(insert "\n\nIn " (emacs-version) "\n")
+ (if (stringp emacs-bzr-version)
+ (insert "Bzr revision: " emacs-bzr-version "\n"))
(if (fboundp 'x-server-vendor)
(condition-case nil
;; This is used not only for X11 but also W32 and others.
=== modified file 'lisp/version.el'
*** lisp/version.el 2011-03-12 19:19:47 +0000
--- lisp/version.el 2011-11-22 00:37:43 +0000
***************
*** 78,89 ****
--- 78,110 ----
;; We hope that this alias is easier for people to find.
(defalias 'version 'emacs-version)
+ ;; Set during dumping, this is a defvar so that it can be setq'd.
+ (defvar emacs-bzr-version nil
+ "String giving the bzr revision number from which this Emacs was built.
+ This is nil if Emacs was not built from a bzr checkout, or if we could
+ not determine the revision.")
+
+ (defun emacs-bzr-get-version ()
+ "Try to return as a string the bzr revision number of the Emacs sources.
+ Returns nil if the sources do not seem to be under bzr, or if we could
+ not determine the revision. Note that this reports on the current state
+ of the sources, which may not correspond to the running Emacs."
+ (let ((file (expand-file-name ".bzr/branch/last-revision" source-directory)))
+ (if (file-readable-p file)
+ (with-temp-buffer
+ (insert-file-contents file)
+ (goto-char (point-max))
+ (if (looking-back "\n")
+ (delete-char -1))
+ (buffer-string)))))
+
;; We put version info into the executable in the form that `ident' uses.
(or (eq system-type 'windows-nt)
(purecopy (concat "\n$Id: " (subst-char-in-string ?\n ?\s (emacs-version))
^ permalink raw reply [flat|nested] 18+ messages in thread* bug#8054: 24.0.50; Who ever heard of a version number that doesn't change all year?
2011-11-22 0:48 ` Glenn Morris
@ 2011-11-23 13:03 ` Juanma Barranquero
2011-11-23 13:24 ` Christoph Scholtes
2011-11-23 13:37 ` Lennart Borgman
2011-11-24 6:20 ` Chong Yidong
1 sibling, 2 replies; 18+ messages in thread
From: Juanma Barranquero @ 2011-11-23 13:03 UTC (permalink / raw)
To: Glenn Morris; +Cc: 8054
> Here's a patch to define an `emacs-bzr-version' during dumping and
> include it in bug reports.
If this were a democracy, I'd vote to commit it to the trunk now. It's
not going to break anything, and it can be useful in bug reports.
Juanma
^ permalink raw reply [flat|nested] 18+ messages in thread
* bug#8054: 24.0.50; Who ever heard of a version number that doesn't change all year?
2011-11-22 0:48 ` Glenn Morris
2011-11-23 13:03 ` Juanma Barranquero
@ 2011-11-24 6:20 ` Chong Yidong
2011-11-24 7:06 ` Glenn Morris
1 sibling, 1 reply; 18+ messages in thread
From: Chong Yidong @ 2011-11-24 6:20 UTC (permalink / raw)
To: Glenn Morris; +Cc: 8054
Glenn Morris <rgm@gnu.org> writes:
> Here's a patch to define an `emacs-bzr-version' during dumping and
> include it in bug reports.
I think this would need the branch name too, in order to be useful.
^ permalink raw reply [flat|nested] 18+ messages in thread
* bug#8054: 24.0.50; Who ever heard of a version number that doesn't change all year?
2011-11-24 6:20 ` Chong Yidong
@ 2011-11-24 7:06 ` Glenn Morris
0 siblings, 0 replies; 18+ messages in thread
From: Glenn Morris @ 2011-11-24 7:06 UTC (permalink / raw)
To: Chong Yidong; +Cc: 8054
Chong Yidong wrote:
>> Here's a patch to define an `emacs-bzr-version' during dumping and
>> include it in bug reports.
>
> I think this would need the branch name too, in order to be useful.
Really? At present, emacs-bzr-version will contain something like:
106501 rgm@gnu.org-20111124035909-rx8lttx9f1j29t69
So it's not just a revision number. The second bit is "a globally unique
identifier for the revision so that it can be referred to without
ambiguity".
But it's trivial to add eg parent_location from branch.conf, if you
really think this is necessary.
^ permalink raw reply [flat|nested] 18+ messages in thread
* bug#8054: 24.0.50; Who ever heard of a version number that doesn't change all year?
2011-02-16 12:54 ` bug#8054: 24.0.50; Who ever heard of a version number that doesn't change all year? jidanni
2011-02-16 13:10 ` Lennart Borgman
@ 2011-02-26 3:04 ` debbugs-submit-bounces
2011-02-26 3:51 ` Glenn Morris
2012-04-13 0:26 ` bug#8054: closed (Re: bug#8054: 24.0.50; Who ever heard of a version number that doesn't change all year?) jidanni
2012-04-13 18:48 ` bug#8054: 24.0.50; Who ever heard of a version number that doesn't change all year? jidanni
3 siblings, 1 reply; 18+ messages in thread
From: debbugs-submit-bounces @ 2011-02-26 3:04 UTC (permalink / raw)
To: 8054
Would it not be possible to add a script that could be used by configure to
check for the .bzr directory and if it exists, extract the value from the
branch/last-revision file and use that value to either append to the version or
set some other variable. If the .bzr directory does not exist, set it to nil or
'unknown' or something?
This would have the advantage of linking the revison number to the final build,
eliminating differences that depend on whether the code is run from the install
directory or installed somewhere else.
Tim
P.S. or alternatively, update the text presented in report-emacs-bug to
encourage people to include the bzr revno if they are running from bzr sources
and explain how to get this value
--
Tim Cross
tcross@rapttech.com.au
There are two types of people in IT - those who do not manage what they
understand and those who do not understand what they manage.
^ permalink raw reply [flat|nested] 18+ messages in thread
* bug#8054: closed (Re: bug#8054: 24.0.50; Who ever heard of a version number that doesn't change all year?)
2011-02-16 12:54 ` bug#8054: 24.0.50; Who ever heard of a version number that doesn't change all year? jidanni
2011-02-16 13:10 ` Lennart Borgman
2011-02-26 3:04 ` debbugs-submit-bounces
@ 2012-04-13 0:26 ` jidanni
2012-04-13 3:00 ` Glenn Morris
2012-04-13 18:48 ` bug#8054: 24.0.50; Who ever heard of a version number that doesn't change all year? jidanni
3 siblings, 1 reply; 18+ messages in thread
From: jidanni @ 2012-04-13 0:26 UTC (permalink / raw)
To: 8054; +Cc: control
reopen 8054
thanks
Thank you but emacs-bzr-version is nil and thus will not help here in
$ apt-cache policy emacs-snapshot
emacs-snapshot:
Installed: 2:20120410-1
Candidate: 2:20120410-1
Version table:
*** 2:20120410-1 0
500 http://emacs.naquadah.org/ unstable/ Packages
100 /var/lib/dpkg/status
^ permalink raw reply [flat|nested] 18+ messages in thread* bug#8054: 24.0.50; Who ever heard of a version number that doesn't change all year?
2011-02-16 12:54 ` bug#8054: 24.0.50; Who ever heard of a version number that doesn't change all year? jidanni
` (2 preceding siblings ...)
2012-04-13 0:26 ` bug#8054: closed (Re: bug#8054: 24.0.50; Who ever heard of a version number that doesn't change all year?) jidanni
@ 2012-04-13 18:48 ` jidanni
3 siblings, 0 replies; 18+ messages in thread
From: jidanni @ 2012-04-13 18:48 UTC (permalink / raw)
To: debian-emacsen; +Cc: 8054
Dear debian-emacsen: shouldn't emacs be able to .,. OK never mind.
>>>>> "GbTS" == GNU bug Tracking System <help-debbugs@gnu.org> writes:
GbTS> Your bug report
GbTS> #8054: 24.0.50; Who ever heard of a version number that doesn't change all year?
GbTS> which was filed against the emacs package, has been closed.
GbTS> The explanation is attached below, along with your original report.
GbTS> If you require more details, please reply to 8054@debbugs.gnu.org.
GbTS> --
GbTS> 8054: http://debbugs.gnu.org/cgi/bugreport.cgi?bug=8054
GbTS> GNU Bug Tracking System
GbTS> Contact help-debbugs@gnu.org with problems
GbTS> From: Glenn Morris <rgm@gnu.org>
GbTS> Subject: Re: bug#8054: closed (Re: bug#8054: 24.0.50; Who ever heard of a version number that doesn't change all year?)
GbTS> To: 8054-done@debbugs.gnu.org
GbTS> Date: Thu, 12 Apr 2012 23:00:55 -0400 (15 hours, 44 minutes, 28 seconds ago)
GbTS> jidanni@jidanni.org wrote:
>> Thank you but emacs-bzr-version is nil and thus will not help here in
>> $ apt-cache policy emacs-snapshot
GbTS> I suggest you take that up with the person who is kind enough to build
GbTS> Emacs for you. Or learn to build it yourself.
GbTS> ----------
GbTS> From: jidanni@jidanni.org
GbTS> Subject: 24.0.50; Who ever heard of a version number that doesn't change all year?
GbTS> To: bug-gnu-emacs@gnu.org
GbTS> Date: Wed, 16 Feb 2011 20:54:37 +0800 (1 year, 8 weeks, 23 hours ago)
GbTS> Gentlemen, let's face it, your version number,
GbTS> emacs-version "24.0.50.1"
GbTS> is a total disaster for use in e.g., report-emacs-bug.
GbTS> Who ever heard of a version number that doesn't change all year?
GbTS> Even Microsoft with its 6.5235345532543 numbers are better than yours.
GbTS> I especially like e.g.,
GbTS> http://en.wikipedia.org/wiki/Special:Version MediaWiki|1.17wmf1 (r82223)
GbTS> So do contemplate providing _version information down to the commit
GbTS> granularity_ like the MediaWiki people do ... available right there to
GbTS> the report-emacs-bug command, etc.
GbTS> Don't expect the user to have to explain what time of what day he got his
GbTS> copy from where, when asked for more details than "24.0.50.1".
GbTS> ----------
^ permalink raw reply [flat|nested] 18+ messages in thread