bug#18725: Emacs 24.4. Say "no" to "erase customizations?". .emacs gets written.

bug#18725: Emacs 24.4. Say "no" to "erase customizations?". .emacs gets written.

[Alan Mackenzie]

Hi, Emacs.

In Emacs 24.4, started with a .emacs which includes a customisation for
face show-paren-match, do

M-x customize-group <RET> paren-showing-faces <RET>

.  One of the widgets is "Revert ...".  Hit <RET> on this widget, then 2
for "Erase customizations".  Emacs prompts with "Really erase
customizations? (yes or no)".  Type  "no".

.emacs is then written.  It shouldn't be.

-- 
Alan Mackenzie (Nuremberg, Germany).

bug#18725: [PATCH]: Emacs 24.4. Say "no" to "erase customizations?". .emacs gets written.

[Alan Mackenzie]

Alan Mackenzie <acm@muc.de> wrote:
> In Emacs 24.4, started with a .emacs which includes a customisation for
> face show-paren-match, do

> M-x customize-group <RET> paren-showing-faces <RET>

> .  One of the widgets is "Revert ...".  Hit <RET> on this widget, then 2
> for "Erase customizations".  Emacs prompts with "Really erase
> customizations? (yes or no)".  Type  "no".

> .emacs is then written.  It shouldn't be.

Here is a possible patch for the problem:


=== modified file 'lisp/cus-edit.el'
--- lisp/cus-edit.el    2014-06-29 02:33:50 +0000
+++ lisp/cus-edit.el    2014-10-15 10:11:00 +0000
@@ -779,7 +779,8 @@
   "Call function FUN on all widgets in `custom-options'.
 If there is more than one widget, ask user for confirmation using
 the query string QUERY, using `y-or-n-p' if STRONG-QUERY is nil,
-and `yes-or-no-p' otherwise."
+and `yes-or-no-p' otherwise.  Return t if the functionality has
+been executed, nil otherwise."
   (if (or (and (= 1 (length custom-options))
               (memq (widget-type (car custom-options))
                     '(custom-variable custom-face)))
@@ -892,16 +893,17 @@
   ;; Bind these temporarily.
   (let ((custom-reset-standard-variables-list '(t))
        (custom-reset-standard-faces-list '(t)))
-    (custom-command-apply
-     (lambda (widget)
-       (and (or (null (widget-get widget :custom-standard-value))
-               (widget-apply widget :custom-standard-value))
-           (memq (widget-get widget :custom-state)
-                 '(modified set changed saved rogue))
-           (widget-apply widget :custom-mark-to-reset-standard)))
-     "The settings will revert to their default values, in this
+    (if
+       (custom-command-apply
+        (lambda (widget)
+          (and (or (null (widget-get widget :custom-standard-value))
+                   (widget-apply widget :custom-standard-value))
+               (memq (widget-get widget :custom-state)
+                     '(modified set changed saved rogue))
+               (widget-apply widget :custom-mark-to-reset-standard)))
+        "The settings will revert to their default values, in this
 and future sessions.  Really erase customizations? " t)
-    (custom-reset-standard-save-and-update)))
+       (custom-reset-standard-save-and-update))))

 ;;; The Customize Commands


-- 
Alan Mackenzie (Nuremberg, Germany).

bug#18725: [PATCH]: Emacs 24.4. Say "no" to "erase customizations?". .emacs gets written

[Ivan Shmakov]

>>>>> Alan Mackenzie <acm@muc.de> writes:

[…]

 > +    (if
 > +       (custom-command-apply
 > +        (lambda (widget)

	Could this please be a ‘when’?

	TIA.

[…]

-- 
FSF associate member #7257  http://boycottsystemd.org/  … 3013 B6A0 230E 334A

bug#18725: [PATCH]: Emacs 24.4. Say "no" to "erase customizations?". .emacs gets written

[Alan Mackenzie]

Hello, Ivan.

Ivan Shmakov <ivan@siamics.net> wrote:
>>>>>> Alan Mackenzie <acm@muc.de> writes:

> [?]

> > +    (if
> > +       (custom-command-apply
> > +        (lambda (widget)

>        Could this please be a ?when??

It could.  But why?  Any special reason?

>        TIA.

> [?]

-- 
Alan Mackenzie (Nuremberg, Germany).

bug#18725: [PATCH]: Emacs 24.4. Say "no" to "erase customizations?". .emacs gets written.

[Stefan Monnier]

[ How did you end up with gnu-emacs-bug@moderators.isc.org in the "To:"
  field?  ]

> +and `yes-or-no-p' otherwise.  Return t if the functionality has
                                       ^^^
We prefer                            non-nil

Otherwise, looks good, please install,


        Stefan

bug#18725: [PATCH]: Emacs 24.4. Say "no" to "erase customizations?". .emacs gets written.

[Alan Mackenzie]

Hello, Stefan.

On Wed, Oct 15, 2014 at 10:39:48AM -0400, Stefan Monnier wrote:

> > +and `yes-or-no-p' otherwise.  Return t if the functionality has
>                                        ^^^
> We prefer                            non-nil

> Otherwise, looks good, please install,

Done, with requested amendment, revision #117588.

> [ How did you end up with gnu-emacs-bug@moderators.isc.org in the "To:"
>   field?  ]

This is going to make Glenn unhappy.  I posted it via Usenet, for lack
of getting a CC of my original bug report to which I could've replied.
It seems the bug-gnu-emacs mail server has been reconfigured not to send
CCs to people involved by default, and I'm not subscribed to
bug-gnu-emacs at the moment.

>         Stefan

-- 
Alan Mackenzie (Nuremberg, Germany).

bug#18725: [PATCH]: Emacs 24.4. Say "no" to "erase customizations?". .emacs gets written.

[Glenn Morris]

Stefan Monnier wrote:

> [ How did you end up with gnu-emacs-bug@moderators.isc.org in the "To:"
>   field?  ]

http://debbugs.gnu.org/cgi/bugreport.cgi?bug=16155#51

> Otherwise, looks good, please install,

Did you mean into emacs-24, because AFAIK this is not a new issue and is
only a minor thing. Are we sure this is safe?

replying to Debbugs

[Ivan Shmakov]

>>>>> Alan Mackenzie <acm@muc.de> writes:
>>>>> On Wed, Oct 15, 2014 at 10:39:48AM -0400, Stefan Monnier wrote:

[…]

 >> [ How did you end up with gnu-emacs-bug@moderators.isc.org in the
 >> "To:" field?  ]

 > This is going to make Glenn unhappy.  I posted it via Usenet, for
 > lack of getting a CC of my original bug report to which I could've
 > replied.  It seems the bug-gnu-emacs mail server has been
 > reconfigured not to send CCs to people involved by default, and I'm
 > not subscribed to bug-gnu-emacs at the moment.

	You can access the mailing list archive via NNTP at [1].  This
	particular version comes with ‘Message-Id:’s intact, so the
	follow-ups tend /not/ to break threads.

	My advice would, however, be to send replies to the messages
	there via email (/not/ over NNTP), possibly changing the header
	to use the relevant bug’s email address, like:

To: 18725@debbugs.gnu.org

	(The above is actually the approach I primarily use.)

	Alternatively, you can download a copy of any specific message –
	or of the whole archive (in mbox format) at once – from the
	respective bug report’s Web page (e. g., [2].)

[1] nntp://news.gmane.org/gmane.emacs.bugs
[2] http://debbugs.gnu.org/18725

-- 
FSF associate member #7257  http://boycottsystemd.org/  … 3013 B6A0 230E 334A

bug#18725: [PATCH]: Emacs 24.4. Say "no" to "erase customizations?". .emacs gets written.

[Glenn Morris]

Alan Mackenzie wrote:

> This is going to make Glenn unhappy.

There's no "going to" about it. I'm aware that you continue to do it.

> for lack of getting a CC of my original bug report to which I could've
> replied.

I checked just now and the automated bug acknowledgment emails are still
going out, just as they always have done, and they still contain
instructions on how to reply. If you did not get one, maybe check into
your email settings.

There's also M-x gnus-read-ephemeral-emacs-bug-group, which is great.

> It seems the bug-gnu-emacs mail server has been reconfigured not to
> send CCs to people involved by default, and I'm not subscribed to
> bug-gnu-emacs at the moment.

There's been no change of configuration on bug-gnu-emacs.
People who report bugs have never been expected to be subscribed to the
list, and never will be.

I've said everything I have to say on this before.

bug#18725: [PATCH]: Emacs 24.4. Say "no" to "erase customizations?". .emacs gets written.

[Michael Albinus]

Alan Mackenzie <acm@muc.de> writes:

>> [ How did you end up with gnu-emacs-bug@moderators.isc.org in the "To:"
>>   field?  ]
>
> This is going to make Glenn unhappy.  I posted it via Usenet, for lack
> of getting a CC of my original bug report to which I could've replied.
> It seems the bug-gnu-emacs mail server has been reconfigured not to send
> CCs to people involved by default, and I'm not subscribed to
> bug-gnu-emacs at the moment.

I do read all gnu-emacs MLs via gmane, which allows me to reply w/o such
problems. Maybe you give it a try.

(setq gnus-secondary-select-methods
        '((nnml "")
	  ...
	  (nntp "news" (nntp-address "news.gmane.org")))
      gnus-select-method '(nnnil))

Best regards, Michael.

replying to Debbugs

[Ivan Shmakov]

>>>>> Michael Albinus <michael.albinus@gmx.de> writes:
>>>>> Alan Mackenzie <acm@muc.de> writes:

 > Cc: 18725@debbugs.gnu.org

	There’s no reason to abuse this particular bug report for this
	discussion; follow-up To: emacs-devel@ instead.

[…]

 >> This is going to make Glenn unhappy.  I posted it via Usenet, for
 >> lack of getting a CC of my original bug report to which I could've
 >> replied.  It seems the bug-gnu-emacs mail server has been
 >> reconfigured not to send CCs to people involved by default, and I'm
 >> not subscribed to bug-gnu-emacs at the moment.

 > I do read all gnu-emacs MLs via gmane, which allows me to reply w/o
 > such problems.  Maybe you give it a try.

	Incidentally, this is what I’ve just suggested.  Still, I’d
	rather appreciate if someone could take care of actually fixing
	the email to Usenet gateway for these mailing lists.

 > (setq gnus-secondary-select-methods
 >         '((nnml "")
 > 	  ...
 > 	  (nntp "news" (nntp-address "news.gmane.org")))
 >       gnus-select-method '(nnnil))

	Per User-Agent: values, it seems that Alan is not a Gnus user.

-- 
FSF associate member #7257  http://boycottsystemd.org/  … 3013 B6A0 230E 334A

bug#18725: [PATCH]: Emacs 24.4. Say "no" to "erase customizations?". .emacs gets written.

[Stefan Monnier]

> This is going to make Glenn unhappy.  I posted it via Usenet, for lack
> of getting a CC of my original bug report to which I could've replied.

IIUC by default you don't get a Cc, indeed.
But you do receive a acknowledgment, instead, which should work fine
for replying.  Maybe the acknowledgment should contain the original
report as an attachment (or does it already?  I can't remember and
don't have one on hand).
I'd expect that many users would be confused if they received an
automatic Cc: for their bug-reports.


        Stefan

bug#18725: [PATCH]: Emacs 24.4. Say "no" to "erase customizations?". .emacs gets written.

[Stefan Monnier]

> Are we sure this is safe?

It looks pretty safe to me, yes.
It looks more like an omission, really (it makes it work more like Custom-save).


        Stefan

Re: replying to Debbugs

[Glenn Morris]

Ivan Shmakov wrote:

> 	rather appreciate if someone could take care of actually fixing
> 	the email to Usenet gateway for these mailing lists.

It's impossible. The reasons why have been explained in the past.
Please just let gnu.emacs.bug the newsgroup die. Pretty please.

(I know your concern is message-ids, I have other ones.)

Re: replying to Debbugs

[Richard Stallman]

[[[ To any NSA and FBI agents reading my email: please consider    ]]]
[[[ whether defending the US Constitution against all enemies,     ]]]
[[[ foreign or domestic, requires you to follow Snowden's example. ]]]

    It's impossible. The reasons why have been explained in the past.
    Please just let gnu.emacs.bug the newsgroup die. Pretty please.

If we are going to discontinue the newsgroup, we should inform users
who use that newsgroup.  For instance, by posting something there
automatically once a month, telling users to use the mailing list
instead.

-- 
Dr Richard Stallman
President, Free Software Foundation
51 Franklin St
Boston MA 02110
USA
www.fsf.org  www.gnu.org
Skype: No way! That's nonfree (freedom-denying) software.
  Use Ekiga or an ordinary phone call.

Re: replying to Debbugs

[Glenn Morris]

Richard Stallman wrote:

> If we are going to discontinue the newsgroup, we should inform users
> who use that newsgroup.

Nobody uses it. A handful of times a year, someone tries to post
something to it. It gets trapped in the bug-gnu-emacs mailman
moderation, and I reject it with a request to resend it by mail.
(Alan is also a list moderator and approves his own messages.)

If it was up to me, we'd simply make that rejection automatic and get on
with life.

Re: replying to Debbugs

[Eli Zaretskii]

> From: Glenn Morris <rgm@gnu.org>
> Date: Thu, 16 Oct 2014 03:16:05 -0400
> Cc: michael.albinus@gmx.de, ivan@siamics.net, emacs-devel@gnu.org
> 
> If it was up to me, we'd simply make that rejection automatic and get on
> with life.

Why isn't it up to you?

replying to Debbugs [ was: bug#18725: ]

[Alan Mackenzie]

[ Cc: changed to emacs-devel@gnu.org ]

Hello, Stefan.

On Wed, Oct 15, 2014 at 02:04:37PM -0400, Stefan Monnier wrote:
> > This is going to make Glenn unhappy.  I posted it via Usenet, for lack
> > of getting a CC of my original bug report to which I could've replied.

> IIUC by default you don't get a Cc, indeed.
> But you do receive a acknowledgment, instead, which should work fine
> for replying.

It's fine for the address, but doesn't contain the text one wants to
reply to.

> Maybe the acknowledgment should contain the original
> report as an attachment (or does it already?  I can't remember and
> don't have one on hand).

> I'd expect that many users would be confused if they received an
> automatic Cc: for their bug-reports.

We get an automatic Cc: for posts to emacs-devel, and that doesn't seem
to confuse people.  Well, it confused me, because I thought we did
usually get automatic Cc:'s for bug-gnu-emacs too.  It seems we don't,
and haven't done for a long time.

It would seem the only practicable way to participate in bug-gnu-emacs
threads is to be subscribed to the mailing list.  There seems to be no
other way (aside from NNTP posting) that one gets the addresses and the
text together.

I don't understand the bit in .../admin/notes/bugtracker that says "We
generally don't assume anyone who posts to a list is subscribed to it,
so we cc everyone on replies.", which seems to contradict the bit which
says "NB this [replying to 123@debbugs.gnu.org] only sends mail to the
bug-list, it does NOT send a CC to the original bug submitter.".

>         Stefan

-- 
Alan Mackenzie (Nuremberg, Germany).

Re: replying to Debbugs

[Alan Mackenzie]

On Thu, Oct 16, 2014 at 03:16:05AM -0400, Glenn Morris wrote:
> Richard Stallman wrote:

> > If we are going to discontinue the newsgroup, we should inform users
> > who use that newsgroup.

> Nobody uses it.

It is unknown how many people use it.

> A handful of times a year, someone tries to post something to it. It
> gets trapped in the bug-gnu-emacs mailman moderation, and I reject it
> with a request to resend it by mail.

How often does such a poster actually resubmit via email?

> (Alan is also a list moderator and approves his own messages.)

> If it was up to me, we'd simply make that rejection automatic and get on
> with life.

You'd shut down the entire newgroup, I think you said somewhere else on
the thread.  But at the moment it's consuming only machine resources
(apart from the efforts of the moderators).  It would take effort to
shut down and is currently useful (no matter how little).

-- 
Alan Mackenzie (Nuremberg, Germany).

Re: replying to Debbugs [ was: bug#18725: ]

[Stefan Monnier]

> We get an automatic Cc: for posts to emacs-devel,

No, there's no such thing on emacs-devel.  Members of emacs-devel
receive the message, of course, but if you post and you're not a member,
you won't get a copy of your message.

> and that doesn't seem to confuse people.  Well, it confused me,
> because I thought we did usually get automatic Cc:'s for bug-gnu-emacs
> too.  It seems we don't, and haven't done for a long time.

You'll get a copy of your message if you're registered to bug-gnu-emacs,
just as is the case for emacs-devel.

> I don't understand the bit in .../admin/notes/bugtracker that says "We
> generally don't assume anyone who posts to a list is subscribed to it,
> so we cc everyone on replies.",

I think this "we" is meant to say "we, mailing-list members, should do
that when replying", not "we, mailing-list administrators, have
configured the list in such a way".
I guess the wording is ambiguous.


        Stefan

Re: replying to Debbugs

[Glenn Morris]

Eli Zaretskii wrote:

> Why isn't it up to you?

I don't make policy for Emacs and GNU at that level.

Re: replying to Debbugs

[Glenn Morris]

Alan Mackenzie wrote:

> It is unknown how many people use it.

It's totally known.
I could go back and look at the bug-gnu-emacs moderation mails over the
last year, but I did not save them. I stand by my statement that is is
very few (if we exclude you).

> How often does such a poster actually resubmit via email?

The short answer is: sometimes.
The longer answer is: I don't care.

There are already 1000s more bugs than we can ever fix. Literally 1000s.
And the number grows every day.

It's documented all over the place (has been for years) how to report
Emacs bugs, and it not "post to a newsgroup". Eg:

https://lists.gnu.org/mailman/listinfo/bug-gnu-emacs

   This list is connected (note that the connection can be unreliable) to
   the gnu.emacs.bug newsgroup, but please report bugs by mail, not news.
   Treat the newsgroup as a read-only mirror.

https://www.gnu.org/software/emacs/manual/html_node/emacs/Checklist.html

   we ask you not to post bug reports (or replies) via the newsgroup. It
   can make it much harder to contact you if we need to ask for more
   information, and it does not integrate well with the bug tracker. 

The cold facts are: if someone reports a bug via the newsgroup, then I
would feel quite happy just ignoring them (not that I do that, as I
said). Most of the time it is not a very good bug report (because they
are incapable of following basic instructions). If it was truly
important, somebody else will report it properly at some point.

> shut down and is currently useful (no matter how little).

It is not usefuul, it is harmful.
The problem is http://debbugs.gnu.org/cgi/bugreport.cgi?bug=936
For everybody other than you, with whatever weird ISP you found,
newsgroup messages do not make it into the tracker.

I won't spend any more energy discussing this. Do what you like.

Re: replying to Debbugs

[James Cloos]

AM>> It is unknown how many people use it.

GM> It's totally known.

You may know how many posts have been made to it, but not how many read it.

(I presume read vs write is the disparity.)

-JimC
-- 
James Cloos <cloos@jhcloos.com>         OpenPGP: 0x997A9F17ED7DAEA6

Re: replying to Debbugs

[Richard Stallman]

[[[ To any NSA and FBI agents reading my email: please consider    ]]]
[[[ whether defending the US Constitution against all enemies,     ]]]
[[[ foreign or domestic, requires you to follow Snowden's example. ]]]

    Nobody uses it. A handful of times a year, someone tries to post
    something to it. It gets trapped in the bug-gnu-emacs mailman
    moderation, and I reject it with a request to resend it by mail.
    (Alan is also a list moderator and approves his own messages.)

I guess that is good enough.

-- 
Dr Richard Stallman
President, Free Software Foundation
51 Franklin St
Boston MA 02110
USA
www.fsf.org  www.gnu.org
Skype: No way! That's nonfree (freedom-denying) software.
  Use Ekiga or an ordinary phone call.

news:gnu.emacs.*

[Ivan Shmakov]

>>>>> Glenn Morris <rgm@gnu.org> writes:
>>>>> Ivan Shmakov wrote:
>>>>> Michael Albinus <michael.albinus@gmx.de> writes:

 >>> I do read all gnu-emacs MLs via Gmane, which allows me to reply w/o
 >>> such problems.  Maybe you give it a try.

 >> Incidentally, this is what I’ve just suggested.  Still, I’d rather
 >> appreciate if someone could take care of actually fixing the email
 >> to Usenet gateway for these mailing lists.

 > It's impossible.  The reasons why have been explained in the past.
 > Please just let gnu.emacs.bug the newsgroup die.  Pretty please.

 > (I know your concern is message-ids, I have other ones.)

	FTR, – the Message-Id:s are broken not just for
	news:gnu.emacs.bug alone, – news:gnu.emacs.help shows exactly
	the same issue (see, e. g., the example I’ve posted earlier
	[1]), and presumably the same applies all across the gnu.emacs.*
	groups (or perhaps even all the gnu.* ones.)

	Naturally, the gmane.emacs.* groups do /not/ exhibit this issue,
	so it’s certainly not something inherently impossible to fix.

	If the issue with news:gnu.emacs.bug is specifically its
	(lack of) integration with the BTS, – it should still be
	possible to deduce the bug report to add the incoming articles
	to (or the address to forward to) based on either References:,
	In-Reply-To:, or (perhaps preferably) the Subject: field.

	(The latter is, incidentally, how some other bug trackers –
	like, say, RT – behave.)

[1] news:874mv2wzpp.fsf@violet.siamics.net

-- 
FSF associate member #7257  http://boycottsystemd.org/  … 3013 B6A0 230E 334A

Re: news:gnu.emacs.*

[Glenn Morris]

Ivan Shmakov wrote:

> 	FTR, – the Message-Id:s are broken not just for
> 	news:gnu.emacs.bug alone, – news:gnu.emacs.help shows exactly
> 	the same issue (see, e. g., the example I've posted earlier
> 	[1]), and presumably the same applies all across the gnu.emacs.*
> 	groups (or perhaps even all the gnu.* ones.)

This is a Mailman FAQ.

http://wiki.list.org/pages/viewpage.action?pageId=4030712

Nobody here at emacs-devel can do anything about the GNU newsgroups or
mailing lists, so please don't email us about them.

Re: news:gnu.emacs.*

[Ivan Shmakov]

>>>>> Glenn Morris <rgm@gnu.org> writes:
>>>>> Ivan Shmakov wrote:

 >> FTR, – the Message-Id:s are broken not just for news:gnu.emacs.bug
 >> alone, – news:gnu.emacs.help shows exactly the same issue (see,
 >> e. g., the example I've posted earlier [1]), and presumably the same
 >> applies all across the gnu.emacs.* groups (or perhaps even all the
 >> gnu.* ones.)

 > This is a Mailman FAQ.

 > http://wiki.list.org/pages/viewpage.action?pageId=4030712

	ACK, thanks for the pointer.

 > Nobody here at emacs-devel can do anything about the GNU newsgroups
 > or mailing lists, so please don't email us about them.

	The page above reads:

 > If your mail program or newsreader implements message threading
 > correctly (see http://www.jwz.org/doc/threading.html), then it should
 > be able to deal with these differences.  If not, then you should
 > complain to the people responsible for implementing and supporting
 > your mail program or newsreader.

	I’ll be just as happy if Gnus could be made to properly handle
	threading in those same newsgroups.  Could we discuss possible
	fixes here, please?

	(I assume that Gnus is within the emacs-devel@ scope, given that
	Gnus is part of GNU Emacs.  Cross-posting to ding@gnus.org,
	though, so feel free to drop the former on follow-up.)

-- 
FSF associate member #7257  http://boycottsystemd.org/  … 3013 B6A0 230E 334A