Different key maps in different dired buffers

Different key maps in different dired buffers

[Whitfield Diffie]

    The function local-set-key seems to be local to major mode, not
local to an individual buffer.  Can anyone tell me how to set keys to
special functions in only some dired buffers?

                With anticipatory thanks,

                            Whit

Re: Different key maps in different dired buffers

[tomas]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

On Thu, May 26, 2016 at 01:14:45AM -0700, Whitfield Diffie wrote:
>     The function local-set-key seems to be local to major mode, not
> local to an individual buffer.  Can anyone tell me how to set keys to
> special functions in only some dired buffers?
> 
>                 With anticipatory thanks,
> 
>                             Whit

This happens because usually, the buffer's local keymap is set up as a
a *reference* to the major mode's keymap [1]. If you want to change
that, you'd have to dissociate this reference by copying. Something
like (warning: untested code ahead)

  (use-local-map (copy-keymap (current-local-map))

might do the trick [2], [3]

regards

[1] https://www.gnu.org/software/emacs/manual/html_node/elisp/Active-Keymaps.html#Active-Keymaps
[2] https://www.gnu.org/software/emacs/manual/html_node/elisp/Creating-Keymaps.html#Creating-Keymaps
[3] https://www.gnu.org/software/emacs/manual/html_node/elisp/Controlling-Active-Maps.html#Controlling-Active-Maps

- -- tomás
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.12 (GNU/Linux)

iEYEARECAAYFAldGvIUACgkQBcgs9XrR2kbYngCeOSPdXhhpy2DPoPyT9EG8NTqy
ehoAn1Mexm3CLo8YwgRCcj5z0M+JBx7B
=4a0D
-----END PGP SIGNATURE-----

Re: Different key maps in different dired buffers

[Stefan Monnier]

>   (use-local-map (copy-keymap (current-local-map))

Better would be

    (use-local-map (make-composed-keymap nil (current-local-map)))


-- Stefan

Re: Different key maps in different dired buffers

[tomas]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

On Thu, May 26, 2016 at 08:25:42AM -0400, Stefan Monnier wrote:
> >   (use-local-map (copy-keymap (current-local-map))
> 
> Better would be
> 
>     (use-local-map (make-composed-keymap nil (current-local-map)))

Hey, thanks. As the warning on the package said, I only half-knew what
I was doing ;-)

[trodding off to read on the difference between (copy-keymap foo) and
(make-composed-keymap nil foo)]

regards
- -- t
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.12 (GNU/Linux)

iEYEARECAAYFAldG7gUACgkQBcgs9XrR2kbSnQCdHpcTaVx/kpw3bFuoHXK4DqVW
p4UAnRDw+vMLDD+XPHj3olrg8d12Nv1+
=MLxm
-----END PGP SIGNATURE-----

RE: Different key maps in different dired buffers

[Drew Adams]

> The function local-set-key seems to be local to major mode, not
> local to an individual buffer.  Can anyone tell me how to set keys to
> special functions in only some dired buffers?

In addition to what others have said, which answers your question,
I'd ask what you really want to do.

In particular, are you trying to do something different for a given
class (subset, kind) of Dired buffers, or just arbitrary different
Dired buffers.  If the former, then consider defining a submode,
i.e., a mode that inherits from Dired mode but has its own keymap
and possibly does things a bit differently.  You can do this with
a major mode or a minor mode.

But if there is no way to recognize a Dired buffer that you might
want to have different key bindings for, then this approach would
not apply.  In that case, the Dired buffers you want to do this
in/for are, in effect, arbitrary.

Re: Different key maps in different dired buffers

[Whitfield Diffie]

> In addition to what others have said, which answers your question,
> I'd ask what you really want to do.

    I haven't seen anyone else say anything; did you see other
responses that I have somehow not received.


> In particular, are you trying to do something different for a given
> class (subset, kind) of Dired buffers, or just arbitrary different
> Dired buffers.

    Both, but in the case at hand, it was the former and you question
reveals an embarrassment of what I am doing in that the changes in
behavior of characters are just a means to doing something else.

    I have rewritten dired to have a dozen or more changes of
behavior, set by switches.  (If you are interested in this code, you
are welcome to it.  It is very variable in quality and is not
adequately documented.)  One of these is the switch
dired-directories-at-top, with the obvious meaning.  My calendar
consists of a directory of years, which are directories of days.  Each
day is a directory of events whose names begin with the times.  Some
of these are plain files; some are directories.  In the day
directories, I don't want to put the directories at the top.  I though
I could achieve this adequately (yes, it is a gross hack and not what
is really needed) by redefining ``f'' to be a function called
dired-find-file-in-future-dir.  I then found that ``f'' took on this
meaning in every dired buffer.

    I have other cases in which I have written a function that changes
the way a directory is displayed (for example comparing file names
numerically rather than as strings) as well as changing the functions
of some characters.  Probably the general form of my question is how
to put a dired buffer in a mode that does not become the mode of all
dired buffers.  I have not explored minor modes and I have don't think
I know the notion of a submode; that word does suggest the right
thing.


    Thank you very much for replying.  I have found the responses of
this list very variable.  I have asked questions that are more
interesting, more subtle, and less ignorant than this one, and been
met with silence.


                                   Whit

RE: Different key maps in different dired buffers

[Drew Adams]

> > In addition to what others have said, which answers your question,
> > I'd ask what you really want to do.
> 
> I haven't seen anyone else say anything; did you see other
> responses that I have somehow not received.

You will receive them.  Probably a delay in the mailing list.
You will get replies from Tomas and Stefan, at least.

> > In particular, are you trying to do something different for a given
> > class (subset, kind) of Dired buffers, or just arbitrary different
> > Dired buffers.
> 
> Both, but in the case at hand, it was the former and you question
> reveals an embarrassment of what I am doing in that the changes in
> behavior of characters are just a means to doing something else.

That's why I asked.  It is easy to think prematurely in terms of a
solution, without completely characterizing the problem.

Sorry, but I don't have the time to look into your fuller
description, but maybe someone else does.  It would probably
help people help you if you provided a more succinct description,
_in terms oF Dired only_.

You can ask also questions about specific bits of your code, of
course, but it is less likely that you will get help if you just
point people to all your code and ask for help.

IOW, be specific, narrow the scope, and describe in terms of
Dired.  That can help others help you.

Re: Different key maps in different dired buffers

[Dmitry Alexandrov]

On 26/05/16 18:05, Whitfield Diffie wrote:
>> In addition to what others have said, which answers your question,
>> I'd ask what you really want to do.
>
> I haven't seen anyone else say anything; did you see other
> responses that I have somehow not received.

Yes:

https://lists.gnu.org/archive/html/help-gnu-emacs/2016-05/msg00264.html
https://lists.gnu.org/archive/html/help-gnu-emacs/2016-05/msg00265.html
https://lists.gnu.org/archive/html/help-gnu-emacs/2016-05/msg00266.html

Are you subscribed to the list?

Re: Different key maps in different dired buffers

[Whitfield Diffie]

> Yes:
>
> https://lists.gnu.org/archive/html/help-gnu-emacs/2016-05/msg00264.html
> https://lists.gnu.org/archive/html/help-gnu-emacs/2016-05/msg00265.html
> https://lists.gnu.org/archive/html/help-gnu-emacs/2016-05/msg00266.html
>
> Are you subscribed to the list?

    No and that explains a lot.  It didn't occur to me that people
might just reply to the list.

        Thank you very much.

RE: Different key maps in different dired buffers

[Drew Adams]

> > Are you subscribed to the list?
> 
> No and that explains a lot.  It didn't occur to me that people
> might just reply to the list.

They shouldn't have (IMO).  But it happens.

Reply to list [was: Different key maps in different dired buffers]

[tomas]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

On Thu, May 26, 2016 at 11:05:11AM -0700, Drew Adams wrote:
> > > Are you subscribed to the list?
> > 
> > No and that explains a lot.  It didn't occur to me that people
> > might just reply to the list.
> 
> They shouldn't have (IMO).  But it happens.

Sorry for hijacking the thread: is it (roughly) consensus here
to "reply to all" instead of replying to list?

I happily handle both ways, but have been scolded elsewhere for not
(strictly) replying to list. That's why I didn't dare here.

Sorry, Whit.

regards
- -- tomás
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.12 (GNU/Linux)

iEYEARECAAYFAldH8scACgkQBcgs9XrR2kYi6ACfarGVWvkPTO8cmdLlheXA6vpP
N0wAn1LvigSL8hpIULRb1ICsXVoW76GY
=n9yC
-----END PGP SIGNATURE-----

Re: Reply to list [was: Different key maps in different dired buffers]

[Eli Zaretskii]

> Date: Fri, 27 May 2016 09:09:59 +0200
> From: <tomas@tuxteam.de>
> Cc: Dmitry Alexandrov <321942@gmail.com>,
> 	Whitfield Diffie <whitfield.diffie@gmail.com>, help-gnu-emacs@gnu.org
> 
> is it (roughly) consensus here to "reply to all" instead of replying
> to list?

Either one is okay.

> I happily handle both ways, but have been scolded elsewhere for not
> (strictly) replying to list.

Don't be averted by such requirements: they are generally incorrect
(whoever subscribes to a list should be prepared to get two copies of
messages for a discussion in which they posted).  Unless the list has
a written policy that everyone is required to follow, you are fine
replying to all.

Re: Reply to list [was: Different key maps in different dired buffers]

[Whitfield Diffie]

> Sorry for hijacking the thread: is it (roughly) consensus here
> to "reply to all" instead of replying to list?

> I happily handle both ways, but have been scolded elsewhere for not
> (strictly) replying to list. That's why I didn't dare here.

    First: thank you very much.

> This happens because usually, the buffer's local keymap is set up as
> a *reference* to the major mode's keymap [1]. If you want to change
> that, you'd have to dissociate this reference by copying. Something
> like (warning: untested code ahead)

>   (use-local-map (copy-keymap (current-local-map)))

>might do the trick [2], [3]

    It did and the explanation makes clear what is not clear from the
documentation I could find.

    Drew, I also thank you for suggesting submodes that inherit; it is
a better solution and I will pursue it.

    Second: I was subscribed to this list for a long time but decided
that since I didn't have enough time to work on it, I was better just
reading it from the server when I was interested.  It didn't occur to
me that I might not see a reply if I asked a question.  Now I need to
go back and see if someone answered my previous question --- to change
echo-area editing, I had to recreate the whole line editor and it
didn't seem that should have been necessary.

     Again, this has been a wonderful experience with the list and I
thank you all very much.

                                        Whit

Re: Reply to list [was: Different key maps in different dired buffers]

[Glenn Morris]

Eli Zaretskii wrote:

>> is it (roughly) consensus here to "reply to all" instead of replying
>> to list?
>
> Either one is okay.

I'm so sure that reply-to-all is correct that I thought it was
official policy for GNU lists.

I don't understand why people think "recipient might get two copies"
is worse than "recipient might get no copies". Especially when the
former issue is trivially avoided by Mailman or MUA duplication suppression.

Re: Reply to list [was: Different key maps in different dired buffers]

[Eli Zaretskii]

> From: Glenn Morris <rgm@gnu.org>
> Cc: help-gnu-emacs@gnu.org
> Date: Fri, 27 May 2016 12:46:43 -0400
> 
> Eli Zaretskii wrote:
> 
> >> is it (roughly) consensus here to "reply to all" instead of replying
> >> to list?
> >
> > Either one is okay.
> 
> I'm so sure that reply-to-all is correct that I thought it was
> official policy for GNU lists.

You may be right.  I just don't think I've ever seen it written, but
it could be that I missed something.

> I don't understand why people think "recipient might get two copies"
> is worse than "recipient might get no copies". Especially when the
> former issue is trivially avoided by Mailman or MUA duplication suppression.

I fully agree.

Re: Reply to list [was: Different key maps in different dired buffers]

[Bob Proulx]

Eli Zaretskii wrote:
> Glenn Morris wrote:
> > Eli Zaretskii wrote:
> > > Either one is okay.
> > 
> > I'm so sure that reply-to-all is correct that I thought it was
> > official policy for GNU lists.
> 
> You may be right.  I just don't think I've ever seen it written, but
> it could be that I missed something.

I don't think anything is written down.  (But I have only been doing
this for ten years or so and still haven't learned all of the
secrets.)  Because for the GNU project we would need to get rms to
approve it and as far as I know no one has pushed an official policy
through for approval.  Until that happens it can only be an defacto
standard.  But by defacto standard reply-all is normal, unless you
know the poster is on the list already.

Pragmatically the bug reporting lists can't be reply only on the
mailing lists.  Because we don't require bug reporters to subscribe
before posting bug reports.  We don't require bug reporters to
subscribe first.  Doing so would be a burden and would alienate bug
reporters.  Therefore for any random bug report we should always
include the original reporter in the recipient list.  (Unless we know
them and know they are already subscribed.)

Because the bug reporting lists are open and need to include the
original sender this carries over to other lists too.  Like this
help-gnu-emacs list.  Should it require people to be subscribed?  That
is more of a gray area since it isn't a bug reporting list.  But
people writing for help are almost the same as bug reporters.  So by
default help lists get the same treatment as bug lists.

But what about discussion lists?  At some point maintainers of
individual projects may state a policy for their own project.  And so
there isn't a 100% uniform policy across all of the mailing lists.
Some lists.{non,}gnu.org lists are one way and some another way.

Other free libre projects have different policies.  For example on the
Debian lists the official policy is to reply only to the list unless
specifically requested by the poster that they be CC'd.  And many
other projects have that policy too.
 
Generally I reply-all unless I recognize the senders as already being
members of the mailing list.  For example here on this message it is
safe to reply only to the mailing list because all of the recipients
obviously read the message that was posted to the mailing list and so
must be subscribed.  I would hate to annoy them with a second copy.
The only unknown is the original poster who said he wasn't subscribed
and therefore I specifically added as a CC.

Additionally there is the problem of the News to Email gateway.
(Which coincidentally I see had a backlog of old messages push through
today.)  Many people read on the News side of things and post there.
I think for them it would be quite surprising to get an email copy of
something they posted by news.  However this is a mailing list and not
a news group.  People reading the news group have to expect that they
are still participating in a mailing list.

> > I don't understand why people think "recipient might get two copies"
> > is worse than "recipient might get no copies". Especially when the
> > former issue is trivially avoided by Mailman or MUA duplication
> > suppression.
> 
> I fully agree.

I completely disagree.

Mailman duplication suppression is mostly non-functional for this
issue.  And when it does come into play it can cause loss of messages
which is the exact opposite of what you want.  MUA duplication
suppression is extremely complicated in order to have it working
reasonably.  Since discarding the second by message-id is almost
always the wrong thing to do.

The problem with people getting two copies is that it is terribly
annoying to manage.  I will almost always get the direct copy first
before the mailing list copy.  And because of Mailman I never know if
I will actually get the Mailman copy.  I have to wait and see.  That
makes management of all of the messages much more complicated.  Which
is exactly why projects that have a policy of only replying to the
mailing list have that policy.

This is solved with Mail-Followup-To: but since that is only another
defacto standard it isn't implemented by many mailers.  Since most of
the popular mailers don't get basic things like threading right it
isn't surprising that they don't implement more subtle things.

And so we have an imperfect system that we simply have to learn to
live with regardless of the problems.  Even with the problems it is
much better than a web forum.

Bob

Re: Reply to list [was: Different key maps in different dired buffers]

[Emanuel Berg]

Bob Proulx <bob@proulx.com> writes:

> Pragmatically the bug reporting lists can't
> be reply only on the mailing lists.
> Because we don't require bug reporters to
> subscribe before posting bug reports.
> We don't require bug reporters to subscribe
> first. Doing so would be a burden and would
> alienate bug reporters.

That's right!

> Because the bug reporting lists are open and
> need to include the original sender this
> carries over to other lists too. Like this
> help-gnu-emacs list. Should it require people
> to be subscribed? That is more of a gray area
> since it isn't a bug reporting list.
> But people writing for help are almost the
> same as bug reporters. So by default help
> lists get the same treatment as bug lists.

There should not be any such requirements -
just the web form "input your data here, verify
that you are human, wait for a mail, input the
code and come back" all that makes my stomach
turn.

Unless I'm desperate for help (like with the
GMT - Generic Mapping Tools - they only have
a forum) I turn around when I'm asked to do
that, rather than to do it.

> But what about discussion lists? At some
> point maintainers of individual projects may
> state a policy for their own project. And so
> there isn't a 100% uniform policy across all
> of the mailing lists.
> Some lists.{non,}gnu.org lists are one way
> and some another way.

People shouldn't be so petty about their
projects! Projects are great fun and sometimes
useful but they shouldn't be a secluded
brotherhood "we know what it is about" than
shuns the darkness of the night...

> I would hate to annoy them with a second
> copy.

It is not that annoying. With Gnus and mail
splitting, you can get away with the extra mail
like this:

    (setq nnmail-split-methods
          '(
            ("mail.ml-ooa" "\\(To\\|Cc\\):.*\\(emacs-w3m@namazu.org\\|help-gnu-emacs@gnu.org\\|gnuplot-info@lists.sourceforge.net\\)")
            ("mail.misc"   "")) )

Even if you don't, it isn't terrible.
The important think is to send the
psychological message virtually immediately,
yes there are people, they may not agree, they
may be imperfect in many ways including solving
your issue, but keep working, keep working with
them, and the problem will be solved with time
and effort.

> Additionally there is the problem of the News
> to Email gateway. (Which coincidentally I see
> had a backlog of old messages push through
> today.) Many people read on the News side of
> things and post there. I think for them it
> would be quite surprising to get an email
> copy of something they posted by news.
> However this is a mailing list and not a news
> group. People reading the news group have to
> expect that they are still participating in
> a mailing list.

Some people may say, the best way to do mailing
lists is with Gnus and Gmane, which turns them
into newsgroups in all essence. And those
people might have good cause to say that...

>>> I don't understand why people think
>>> "recipient might get two copies" is worse
>>> than "recipient might get no copies".
>>> Especially when the former issue is
>>> trivially avoided by Mailman or MUA
>>> duplication suppression.
>>>
>> I fully agree.
>
> I completely disagree.
>
> Mailman duplication suppression is mostly
> non-functional for this issue.

What about my solution - does that work? (OK,
not everyone uses Gnus.)

Well, obviously we should try to make it as
pleasant as possible for our users. On the
other hand, to get two copies of help isn't
a problem! If people hit the ceiling in
frustration because of that perhaps they should
take a leave from computers and do Buddhism for
half a year, and then come back...

> And so we have an imperfect system that we
> simply have to learn to live with regardless
> of the problems. Even with the problems it is
> much better than a web forum.

Indeed. Tho there can be a gateway in between
that as well - the future of computing is
interface-agnostic!

Why not have the exact same material accessible
as a mailing list, as a Gmane newsgroup, as
a Usenet newsgroup, as a web forum, and as
a Facebook whatever-they-call-it! This isn't
hard to do, well, not impossible anyway. (With
Facebook politics may be a problem.)

Recall there is already gnu-emacs-help (the
listbot), there is gmane.emacs.help (Gmane),
and there is gnu.emacs.help (Usenet).

Is there a "GNU web forum" software? On Usenet,
there is rec.bicycles.tech - and here is the
same thing, as a web forum!

    http://www.cyclebanter.com/forumdisplay.php?f=8

-- 
underground experts united .... http://user.it.uu.se/~embe8573
Emacs Gnus Blogomatic ......... http://user.it.uu.se/~embe8573/blogomatic
                   - so far: 46 Blogomatic articles -                   

Re: Reply to list [was: Different key maps in different dired buffers]

[Bob Proulx]

Emanuel Berg wrote:
> Bob Proulx writes:
> > Because the bug reporting lists are open and need to include the
> > original sender this carries over to other lists too. Like this
> > help-gnu-emacs list. Should it require people to be subscribed?
> > That is more of a gray area since it isn't a bug reporting list.
> > But people writing for help are almost the same as bug
> > reporters. So by default help lists get the same treatment as bug
> > lists.
> 
> There should not be any such requirements - just the web form "input
> your data here, verify that you are human, wait for a mail, input
> the code and come back" all that makes my stomach turn.

What web form?  This is a mailing list.  There isn't any web form.
One sends email using your MUA mail user agent.  That could be any of
emacs, gnus, mutt, mailx, and so on and so forth.  Could be anything.
Which could include a web form is one if talking about Gmail or
Gmane.  But those are clearly web mail interfaces.

> Unless I'm desperate for help (like with the GMT - Generic Mapping
> Tools - they only have a forum) I turn around when I'm asked to do
> that, rather than to do it.

Me too.  At least for bug reporting systems such as Google Code which
required me to sign up for a Google account in order to post a bug
report.  And so I walked away from the code rather than do so.

> > But what about discussion lists? At some point maintainers of
> > individual projects may state a policy for their own project. And
> > so there isn't a 100% uniform policy across all of the mailing
> > lists.  Some lists.{non,}gnu.org lists are one way and some
> > another way.
> 
> People shouldn't be so petty about their projects! Projects are
> great fun and sometimes useful but they shouldn't be a secluded
> brotherhood "we know what it is about" than shuns the darkness of
> the night...

Mine was a statement of fact.  That is simply the way things currently
are.  It is rather a federalist system like the US government.  That
which is not required globally by GNU is left to the discretion of the
individual project maintainers.  Because this isn't a global GNU
Project policy it is allowed to be overridden by individual project
maintainers.  Which is why it isn't 100% uniform.

> > I would hate to annoy them with a second copy.
> 
> It is not that annoying.

It is annoying.

Is this a good time to note that I can tell you are using the
newsgroup interface.  Which comes through the gnu.emacs.help
newsgroup.  Is gatewayed by email to the mailing list.  And therefore
I assume is incapable of CC'ing the original poster who is not
subscribed?  This is a good example of one of the imperfections of the
loosely combined systems.  The original poster would need a reply-all
but coming through the newsgroup it is (IIRC) impossible to do this.
Which is one of the reasons I wish the mailing lists were NOT
gatewayed to the newsgroups.  The two are different things.  Trying to
tie them together cannot be 100% perfect.  But it has been this way
for many years.

> With Gnus and mail splitting, you can get away with the extra mail
> like this:

I am not a gnus user.

>     (setq nnmail-split-methods
>           '(
>             ("mail.ml-ooa" "\\(To\\|Cc\\):.*\\(emacs-w3m@namazu.org\\|help-gnu-emacs@gnu.org\\|gnuplot-info@lists.sourceforge.net\\)")
>             ("mail.misc"   "")) )

The canonical mailing list header to identify mailing lists is the
List-ID header.  Not the To header.  Mailman uses X-BeenThere which is
okay for Mailman but doesn't exist with other mailing list management
software and therefore isn't standard like List-Id.

> Some people may say, the best way to do mailing lists is with Gnus
> and Gmane, which turns them into newsgroups in all essence. And
> those people might have good cause to say that...

Many people do prefer Newsgroups or Gmane.  Gmane is not really a
newsgroup.  Although it has created a newsgroup-like interface.  Gmane
is still sending email.  It is more like a centralized webmail system.

> >>> I don't understand why people think "recipient might get two
> >>> copies" is worse than "recipient might get no copies".
> >>> Especially when the former issue is trivially avoided by Mailman
> >>> or MUA duplication suppression.
> >>>
> >> I fully agree.
> >
> > I completely disagree.
> >
> > Mailman duplication suppression is mostly non-functional for this
> > issue.
> 
> What about my solution - does that work? (OK, not everyone uses
> Gnus.)

I have no idea as I am not a Gnus user.  So that won't work for me.

I have spent a lot of time devising various procmail related
processes.  All of them have pros and cons.  I have tried many things
over the years.  In the end there isn't any perfect solution.

> Well, obviously we should try to make it as pleasant as possible for
> our users. On the other hand, to get two copies of help isn't a
> problem! If people hit the ceiling in frustration because of that
> perhaps they should take a leave from computers and do Buddhism for
> half a year, and then come back...

On that I agree.  Which is why you haven't seen complaints from me
about it.  But it does annoy me enough to discuss it. :-)

> > And so we have an imperfect system that we simply have to learn to
> > live with regardless of the problems. Even with the problems it is
> > much better than a web forum.
> 
> Indeed. Tho there can be a gateway in between that as well -

Indeed there are many interfaces already in existence.

> the future of computing is interface-agnostic!

Of that I am not so sure.  As time went by the typical interface all
looked like a web page.  But now I read blogs saying the web interface
is dead and that it all looks like an app with a touch screen.  We
shall see.

> Why not have the exact same material accessible as a mailing list,
> as a Gmane newsgroup, as a Usenet newsgroup, as a web forum, and as

We do already have all of those things.  Those are all different.  And
the seams between are not completely smooth between them.

> a Facebook whatever-they-call-it! This isn't hard to do, well, not
> impossible anyway. (With Facebook politics may be a problem.)

Facebook has a lot of non-free problems.  I would not recommend
expanding our free libre discussion onto that non-free platform.  It
would prevent many people from participating.

> Recall there is already gnu-emacs-help (the listbot), there is
> gmane.emacs.help (Gmane), and there is gnu.emacs.help (Usenet).

And unless I am mistaken I believe all of those already exist and are
all gatewayed to the same mailing list.

> Is there a "GNU web forum" software? On Usenet, there is
> rec.bicycles.tech - and here is the same thing, as a web forum!
>     http://www.cyclebanter.com/forumdisplay.php?f=8

Yes.  There are many.  Gmane is a good example.  And also all of the
web based newsgroup readers of which I am not familiar.

Bob

Re: Reply to list [was: Different key maps in different dired buffers]

[Emanuel Berg]

Bob Proulx <bob@proulx.com> writes:

> What web form? This is a mailing list.

There are many mailing lists which require you
to register using such forms.

> One sends email using your MUA mail user
> agent. That could be any of emacs, gnus,
> mutt, mailx, and so on and so forth. Could be
> anything. Which could include a web form is
> one if talking about Gmail or Gmane.
> But those are clearly web mail interfaces.

Gmane is a mail interface but it is also
a replication of the newsreader interface and
Usenet system of communication which is the
most advanced and powerful thus far.

> Is this a good time to note that I can tell
> you are using the newsgroup interface.
> Which comes through the gnu.emacs.help
> newsgroup. Is gatewayed by email to the
> mailing list. And therefore I assume is
> incapable of CC'ing the original poster who
> is not subscribed?

:O

You do me an injustice!

With Gnus I can send the same message to
several newsgroups and mail addresses with one
keystroke!

I'll double this up for you right away :)

> This is a good example of one of the
> imperfections of the loosely combined
> systems. ... Trying to tie them together
> cannot be 100% perfect. But it has been this
> way for many years.

Man, think BIG!

It is *never* going to be "100% perfect".
The binding together of mailing lists, Gmane,
blogs (Gwene), Usenet, RSS, and so on, all this
is wonderful as people can use whatever
interface they like!

It is just like using mpsyt instead of the
crappy, commercial web interface to YouTube.
I'm sure there are problems with that to but
the freedom of interface and access methods
outweighs them one hundred to one, just as it
does here!

> The canonical mailing list header to identify
> mailing lists is the List-ID header. Not the To
> header.

Do all lists adhere to that?

The To header is on the contrary quite
indispensible...

> Many people do prefer Newsgroups or Gmane.
> Gmane is not really a newsgroup.

It is thousands.

> We do already have all of those things.
> Those are all different. And the seams
> between are not completely smooth
> between them.

They are smooth enough. With your efforts they
will be even smoother. But the objective is not
industrial smoothing but freedom of interface
and data access!

> Facebook has a lot of non-free problems.
> I would not recommend expanding our free
> libre discussion onto that non-free platform.
> It would prevent many people
> from participating.

We can't all but exclude ourself from a whole
generation of people either, can we?

People use it. They are much more comfortable
using it than sending mails. It is a sad state.
But it is the truth.

> And unless I am mistaken I believe all of
> those already exist and are all gatewayed to
> the same mailing list.

Even so, on Usenet gnu.emacs.help is
a newsgroup.

>> Is there a "GNU web forum" software?
>> On Usenet, there is rec.bicycles.tech - and
>> here is the same thing, as a web forum!
>> http://www.cyclebanter.com/forumdisplay.php?f=8
>
> Yes. There are many. Gmane is a good example.
> And also all of the web based newsgroup readers
> of which I am not familiar.

Gmane is *archived* online.

Here is a forum:

    https://www.flashback.org

If we had software for such forums - I don't
know if we do? - it would be a good thing if
that software included an interface to the
traffic here - i.e., yet another gateway!

-- 
underground experts united .... http://user.it.uu.se/~embe8573
Emacs Gnus Blogomatic ......... http://user.it.uu.se/~embe8573/blogomatic
                   - so far: 46 Blogomatic articles -                   

Re: Reply to list [was: Different key maps in different dired buffers]

[Dmitry Alexandrov]

Emanuel Berg <embe8573@student.uu.se> writes:

>>> Is there a "GNU web forum" software?
>>> On Usenet, there is rec.bicycles.tech - and
>>> here is the same thing, as a web forum!
>>> http://www.cyclebanter.com/forumdisplay.php?f=8
>>
>> Yes. There are many. Gmane is a good example.
>> And also all of the web based newsgroup readers
>> of which I am not familiar.
>
> Gmane is *archived* online.
>
> Here is a forum:
>
>     https://www.flashback.org
>
> If we had software for such forums - I don't
> know if we do? - it would be a good thing if
> that software included an interface to the
> traffic here - i.e., yet another gateway!

Gmane actually *does* allow posting trough a web interface.  However, I
doubt that anyone uses it — it’s pretty ugly, imho, and requires solving
a captcha.

By the way, so far as we are discussing techniques of using mailing
lists, let me turn your attention, that your messages are systematically
appear in ‘gmane.emacs.help’ as replies to non-existent messages, thus
breaking the tree.  That might be some bug of Gmane, of course, however
the archive of GNU Mailman [0] shares this views.

[0] https://lists.gnu.org/archive/html/help-gnu-emacs/2016-06/threads.html

Re: Reply to list [was: Different key maps in different dired buffers]

[Bob Proulx]

Emanuel Berg wrote:
> Bob Proulx writes:
> > What web form? This is a mailing list.
> 
> There are many mailing lists which require you
> to register using such forms.

Sites like Gmane and Nabble have forms for you to register with their
site.  If you like you can put up a web form somewhere too.  But those
sites are not this mailing list.  The mailing list has no web forms.
The mailing list only deals with mail.

> > One sends email using your MUA mail user agent. That could be any
> > of emacs, gnus, mutt, mailx, and so on and so forth. Could be
> > anything. Which could include a web form is one if talking about
> > Gmail or Gmane.  But those are clearly web mail interfaces.
> 
> Gmane is a mail interface but it is also a replication of the
> newsreader interface and Usenet system of communication which is the
> most advanced and powerful thus far.

And Gmane does a very good job of it.  Even though I don't use them
myself they are usually my recommendation for people who do want a web
interface.  Gmane has created a very nice user interface to it.

> > Is this a good time to note that I can tell you are using the
> > newsgroup interface.  Which comes through the gnu.emacs.help
> > newsgroup. Is gatewayed by email to the mailing list. And
> > therefore I assume is incapable of CC'ing the original poster who
> > is not subscribed?
> 
> :O
> 
> You do me an injustice!
> 
> With Gnus I can send the same message to
> several newsgroups and mail addresses with one
> keystroke!
> 
> I'll double this up for you right away :)

I did get your courtesy copy.  And for the rest of the list what I
received was this in a direct email

>> The following message is a courtesy copy of an article that has been
>> posted to gnu.emacs.help as well.
>>...

But for *this* message here it was posted to the newsgroup.  That came
through the news to mail gateway.  And as far as anyone who follows up
here can see it only went to the mailing list.  Which means that any
CC was lost.  Which means my CC to Whitfield was lost.

While you were able to manually send me a copy it still wasn't perfect
because anyone following up to your message doesn't know that
Whitfield wasn't subscribed and is now cut out of the conversation.
That is exactly what happened with Dmitry's follow-up message.

This is the situation that Mail-Followup-To is designed to handle.
But the news to mail gateway doesn't deal with it.  I am not sure how
it would since news doesn't have the concept.

An imperfect situation.

And this doesn't even mention that your follow-ups to the mailing list
are always in-reply-to the message before the message you are replying
to and not the one it should be in reply to.  Some breakage in
postings coming through the newsgroup gateway.

> > This is a good example of one of the imperfections of the loosely
> > combined systems. ... Trying to tie them together cannot be 100%
> > perfect. But it has been this way for many years.
> 
> Man, think BIG!
> 
> It is *never* going to be "100% perfect".  The binding together of
> mailing lists, Gmane, blogs (Gwene), Usenet, RSS, and so on, all
> this is wonderful as people can use whatever interface they like!

Wonderful?  Yes!  Imperfect?  Yes. :-(

> It is just like using mpsyt instead of the crappy, commercial web
> interface to YouTube.  I'm sure there are problems with that to but
> the freedom of interface and access methods outweighs them one
> hundred to one, just as it does here!
> 
> > The canonical mailing list header to identify mailing lists is the
> > List-ID header. Not the To header.
> 
> Do all lists adhere to that?

As far as I can see.  Yes.  But I am sure there are some bad lists
somewhere that don't.

> The To header is on the contrary quite indispensible...
> 
> > Many people do prefer Newsgroups or Gmane.
> > Gmane is not really a newsgroup.
> 
> It is thousands.

It is a web mail interface to an archive of thousands of mailing
lists.  It has an user interface that looks a lot like the newsgroup
interface.  But looking like something and being something are two
different things.

> > We do already have all of those things.  Those are all
> > different. And the seams between are not completely smooth between
> > them.
> 
> They are smooth enough. With your efforts they will be even
> smoother. But the objective is not industrial smoothing but freedom
> of interface and data access!

A good time for me to say that I do wish people to be able to read and
respond using their preferred interfaces.  For me that is reading mail
with mutt and responding to mail using emacs.  I hate captured web
form interfaces where I can't use my favorite editor to craft my reply.

> > Facebook has a lot of non-free problems.  I would not recommend
> > expanding our free libre discussion onto that non-free platform.
> > It would prevent many people from participating.
> 
> We can't all but exclude ourself from a whole generation of people
> either, can we?
> 
> People use it. They are much more comfortable using it than sending
> mails. It is a sad state.  But it is the truth.

You are proposing that we make a compromise of values in order to
bring in the young generation of Facebook users?  To paraphrase the
often quoted phrase, "Those that would give up some essential liberty
for some security deserve neither."

Also the Facebook interface is designed for very short comments.  What
do you think it would do with our exchange here?  We have already lost
most people on this list who read email and are used to long
diatribes.  FB skimmers spinning the touchscreen with their finger
would have passed this by a page in from the start.  I would oppose an
FB to mailing list gateway as being generally incompatible.

> > And unless I am mistaken I believe all of those already exist and
> > are all gatewayed to the same mailing list.
> 
> Even so, on Usenet gnu.emacs.help is a newsgroup.

Yes.  A newsgroup that is gateway'd into this mailing list.  From my
perspective it is gateway'd into the mailing list because I read the
mailing list.  From your perspective the mailing list is gateway'd
into the newsgroup because you read the newsgroup.  I realize the
directionality is a matter of perspective.  I used to read the
newsgroups of USENET before the noise there became too much for me to
wade through.  Personally due to the interface differences I wish they
would not gateway between each other.  Gmane handles the mailing lists
fine because it is a web mail interface.  But newsgroup readers cannot
because they are using the newsgroup which is gateway'd to the mailing
list and there are concepts that do not transfer across them.  Such as
keeping a non-subscribed person who has requested CC's in the
discussion.  That is a mailing list concept and not a newsgroup concept.

> >> Is there a "GNU web forum" software?  On Usenet, there is
> >> rec.bicycles.tech - and here is the same thing, as a web forum!
> >> http://www.cyclebanter.com/forumdisplay.php?f=8
> >
> > Yes. There are many. Gmane is a good example.  And also all of the
> > web based newsgroup readers of which I am not familiar.
> 
> Gmane is *archived* online.

Gmane is an email archive and web mail interface.  A good one.  You
appear to disagree.  In what way does Gmane not meet your definition
of a web interface to the mailing list?

> Here is a forum:
> 
>     https://www.flashback.org
> 
> If we had software for such forums - I don't know if we do? - it
> would be a good thing if that software included an interface to the
> traffic here - i.e., yet another gateway!

That appears to me like a typical web forum.  (I can't really tell.  I
didn't browse more than a few pages.)  Personally I don't like web
forums.  I realize many others prefer them.  That's fine.  Each to
their own.  But let's not join incompatible technologies together into
a Frankenstein's monster.

I think for those preferring a web interface to the mailing lists that
Gmane is the best user experience available.

Bob

Re: Reply to list [was: Different key maps in different dired buffers]

[Emanuel Berg]

Dmitry Alexandrov <321942@gmail.com> writes:

> Gmane actually *does* allow posting trough
> a web interface. However, I doubt that anyone
> uses it — it’s pretty ugly, imho, and
> requires solving a captcha.

Plus a forum is often also a small community
with users - moderators, post counts, guest
books, private messages, and so on. At least it
was like that some ten years ago when I got
banned from such forums on a regular basis...
You want the truth? You can't *handle*
the truth!

> By the way, so far as we are discussing
> techniques of using mailing lists, let me
> turn your attention, that your messages are
> systematically appear in ‘gmane.emacs.help’
> as replies to non-existent messages, thus
> breaking the tree. That might be some bug of
> Gmane, of course, however the archive of GNU
> Mailman [0] shares this views.

No, that was the case with the "Gnus IMAP
groups appear twice" issue, right?

Or does it happen all the time? For example,
does it happen now?

If it is only the IMAP posts that was because
the OP sent me questions not as CC, but To my
mail.

As he started it out as a post to the list, and
as the communication was non-sensitive,
I thought I'd put it on the list with full
quotes to bring it to the public and perhaps
show the OP where the stuff belongs.

-- 
underground experts united .... http://user.it.uu.se/~embe8573
Emacs Gnus Blogomatic ......... http://user.it.uu.se/~embe8573/blogomatic
                   - so far: 47 Blogomatic articles -                   

Re: Reply to list [was: Different key maps in different dired buffers]

[Emanuel Berg]

Bob Proulx <bob@proulx.com> writes:

> Sites like Gmane and Nabble have forms for
> you to register with their site.

Perhaps, but Gmane isn't a web site lite any
other, it is a unique piece of technology.
It is not a thing of the web except for
archiving and distribution, which is two things
the web is good for. There is no web interface
to it that you are forced to use. The best
thing is to use Gmane with Gnus and then you
don't have to use the web at all. And you don't
have to register with forms. The first time you
send a post to a group, you get a mail to
confirm. You reply to that mail and thereafter
you just post to the group. There is no
entering of data, no codes back and forth, no
web browser. Simply put, it is brilliant!

Here - mails, RSS, Usenet, Gmane, mail, my own
project blogomatic:

    http://user.it.uu.se/~embe8573/figures/gnus/gnus-group.png

> If you like you can put up a web form
> somewhere too. But those sites are not this
> mailing list. The mailing list has no
> web forms.

But many do otherwise you don't get access.

> And Gmane does a very good job of it.
> Even though I don't use them myself they are
> usually my recommendation for people who do
> want a web interface.

Take a look at my screenshot above! Does that
look like a web interface? (And whose questions
are you calling rhetorical?)

> But for *this* message here it was posted to
> the newsgroup. That came through the news to
> mail gateway. And as far as anyone who
> follows up here can see it only went to the
> mailing list. Which means that any CC
> was lost.

OK, I can start putting in a small note: also
"CC'd to Whitfield" if that makes you feel
better about it. (It can even be automatized.)

> And this doesn't even mention that your
> follow-ups to the mailing list are always
> in-reply-to the message before the message
> you are replying to and not the one it should
> be in reply to. Some breakage in postings
> coming through the newsgroup gateway.

OK? To me the threads look good so I never
thought about it.

> But looking like something and being
> something are two different things.

The challenge is to make it look and behave
like the user wants and still be accessible to
everyone else the way *they* want it!

> I hate captured web form interfaces where
> I can't use my favorite editor to craft
> my reply.

Don't hate the interface. Become the interface!
That's what I did, with a little help from my
friends...

> You are proposing that we make a compromise
> of values in order to bring in the young
> generation of Facebook users? To paraphrase
> the often quoted phrase, "Those that would
> give up some essential liberty for some
> security deserve neither."

We can't have watertight between us and the
rest of humanity, i.e. millions of educated,
creative, and active people. It is
a bad situation. Much worse than listbot vs.
Gmane vs. Usenet. However I don't have any
suggestions what to do about it as of now.

> Also the Facebook interface is designed for
> very short comments. What do you think it
> would do with our exchange here?

"Our" exchanges? I want *them* to be *us*!

> I would oppose an FB to mailing list gateway
> as being generally incompatible.

There is nothing incompatible in computer
technology. It is all Os and 1s. Because of
politics, sadly it might never happen with FB.
In terms of technology it is very possible.

> Gmane handles the mailing lists fine because
> it is a web mail interface. But newsgroup
> readers cannot because they are using the
> newsgroup which is gateway'd to the mailing
> list and there are concepts that do not
> transfer across them. ... Gmane is an email
> archive and web mail interface. A good one.
> You appear to disagree. In what way does
> Gmane not meet your definition of a web
> interface to the mailing list?

Perhaps it is easier if you think of Gmane in
combination with Gnus, which is indeed
a newsreader - and not just any but the best.
Gmane/Gnus is a way to organize and take part
in communication. It is Usenet Reloaded!

-- 
underground experts united .... http://user.it.uu.se/~embe8573
Emacs Gnus Blogomatic ......... http://user.it.uu.se/~embe8573/blogomatic
                   - so far: 47 Blogomatic articles -                   

Re: Reply to list [was: Different key maps in different dired buffers]

[Bob Proulx]

Dmitry Alexandrov wrote:
> Emanuel Berg writes:
> > If we had software for such forums - I don't know if we do? - it
> > would be a good thing if that software included an interface to
> > the traffic here - i.e., yet another gateway!
> 
> Gmane actually *does* allow posting trough a web interface.  However, I
> doubt that anyone uses it — it’s pretty ugly, imho, and requires solving
> a captcha.

Really?  I thought most people that used Gmane used it for posting
through their interface.  No?  I guess it is really hard to tell by
anyone outside of Gmane.  But I thought that was one of the advantages
of using it.  That one need not have any other connection.

Plus I know that during the recent network connection problems to the
Boston datacenter site that much mail from Gmane was backlogged unable
to transmit messages.  All of those came through Gmane so there are at
least some users using it.

> By the way, so far as we are discussing techniques of using mailing
> lists, let me turn your attention, that your messages are systematically
> appear in ‘gmane.emacs.help’ as replies to non-existent messages, thus
> breaking the tree.  That might be some bug of Gmane, of course, however
> the archive of GNU Mailman [0] shares this views.
> 
> [0] https://lists.gnu.org/archive/html/help-gnu-emacs/2016-06/threads.html

Yes.  I have tried to determine why this is before.  I don't think it
is on Emanuel's end.  It appears with other posters too.  Emanual is
just one of the more prolific posters and so we see it more often with
his posts.  :-)

Plus when the messages do thread they thread to the parent of the
message and not the message being responded to.  And of course the
unusual software in the path is Gnus, News, the news2mail gateway, any
of the three of which could be causing problems.  I think it is the
Stanford newsgroup gateway causing the problem.

Details follow for those who want to skip now.

Here was my message:

  From: Bob Proulx <bob@proulx.com>
  References: <CAF+O-CXAWEKtmRpcw2j1ewXbZNMZXhsst9BXE3tGpVRr0RnLaQ@mail.gmail.com>
          <57473688.4020503@gmail.com>
          <CAF+O-CVmonxXDuvXOVtO-eqo9eBTA2ogurG_nUhB6vaVMuas+Q@mail.gmail.com>
          <e0650b7a-110d-43b4-9825-5e2a8ced304e@default>
          <20160527070959.GB27615@tuxteam.de> <83mvnc0vze.fsf@gnu.org>
          <wsk2ifsb2k.fsf@fencepost.gnu.org> <83pos7z6gv.fsf@gnu.org>
          <mailman.787.1464997128.1216.help-gnu-emacs@gnu.org>
          <86zir1n347.fsf@student.uu.se>
  In-Reply-To: <86zir1n347.fsf@student.uu.se>
  Message-ID: <20160603195221435910074@bob.proulx.com>

And here was the follow-up message:

  From: Emanuel Berg <embe8573@student.uu.se>
  Path: usenet.stanford.edu!news.tele.dk!news.tele.dk!small.news.tele.dk!newsfeed.xs4all.nl!newsfeed9.news.xs4all.nl!feeder1.xsusenet.com!nntpspool.opticnetworks.net!nntpspool01.opticnetworks.net!aioe.org!.POSTED!not-for-mail
  Newsgroups: gnu.emacs.help
  Organization: Aioe.org NNTP Server
  Xref: usenet.stanford.edu gnu.emacs.help:217984
  User-Agent: Gnus/5.13 (Gnus v5.13) Emacs/24.4 (gnu/linux)
  References: <CAF+O-CXAWEKtmRpcw2j1ewXbZNMZXhsst9BXE3tGpVRr0RnLaQ@mail.gmail.com>
          <57473688.4020503@gmail.com>
          <CAF+O-CVmonxXDuvXOVtO-eqo9eBTA2ogurG_nUhB6vaVMuas+Q@mail.gmail.com>
          <e0650b7a-110d-43b4-9825-5e2a8ced304e@default>
          <20160527070959.GB27615@tuxteam.de> <83mvnc0vze.fsf@gnu.org>
          <wsk2ifsb2k.fsf@fencepost.gnu.org> <83pos7z6gv.fsf@gnu.org>
          <mailman.787.1464997128.1216.help-gnu-emacs@gnu.org>
          <86zir1n347.fsf@student.uu.se>
          <mailman.789.1465006376.1216.help-gnu-emacs@gnu.org>
  Message-ID: <8660tpmztw.fsf@student.uu.se>

There is no In-Reply-To on the follow-up.  Only the References header
can be used for threading.  Checking through the references in the
place of my message 20160603195221435910074@bob.proulx.com I see
mailman.789.1465006376.1216.help-gnu-emacs@gnu.org instead.  That
appears to be the point where threading gets broken.  Something
appears to have converted one message id into another one.

But this isn't only Emanuel.  Here is another example of the same
which has neither Emanuel nor myself in the conversation.

  https://lists.gnu.org/archive/html/help-gnu-emacs/2016-06/threads.html#00084

The original posting has:

  From: "Tory S. Anderson" <torys.anderson@gmail.com>
  Message-ID: <874manawag.fsf@torysa-worldsendless.byu.edu>

The follow-up with the broken threading has:

  From: Marco Wahl <marcowahlsoft@gmail.com>
  Path: usenet.stanford.edu!news.tele.dk!news.tele.dk!small.news.tele.dk!newsfeed.xs4all.nl!newsfeed8.news.xs4all.nl!feeder.erje.net!1.eu.feeder.erje.net!weretis.net!feeder4.news.weretis.net!news.mixmin.net!aioe.org!.POSTED!not-for-mail
  Newsgroups: gnu.emacs.help
  Message-ID: <84vb33j93h.fsf@gmail.com>
  References: <mailman.1096.1461772070.7477.help-gnu-emacs@gnu.org>
  NNTP-Posting-Host: ghePZ09roGjaAbQfu6+I9w.user.gioia.aioe.org
  User-Agent: Gnus/5.13 (Gnus v5.13) Emacs/25.1.50 (gnu/linux)
  Xref: usenet.stanford.edu gnu.emacs.help:217524
  References: <mailman.1096.1461772070.7477.help-gnu-emacs@gnu.org>

Same thing.  The missing message-id has been replaced with a
non-matching one.

In these messages I see two things in common.  One is that both came
through the news gateway.  Another is that both users posted using
Gnus.  Can we find an example that is broken that only uses one of
them to eliminate the other as a problem?  At this moment I suspect
the gateway.  But that is always my suspicion bias since it is outside
of our control.

Lars Magne Ingebrigtsen has several in this thread.  But my mail was
in there too and I am looking for something without.

  https://lists.gnu.org/archive/html/help-gnu-emacs/2016-05/threads.html#00092

Here is one.

  https://lists.gnu.org/archive/html/help-gnu-emacs/2016-04/msg00019.html

Original:

  From: Shiyao Ma <I@introo.me>
  Message-ID: <CAJQX3DzrbHsHOHGjdqQAEtu3jypnpZy+sgLT8NrTayvCweOnOQ@mail.gmail.com>

Follow-up:

  From: "Pascal J. Bourguignon" <pjb@informatimago.com>
  Path: usenet.stanford.edu!fu-berlin.de!uni-berlin.de!individual.net!not-for-mail
  Newsgroups: gnu.emacs.help
  User-Agent: Gnus/5.13 (Gnus v5.13) Emacs/24.5 (gnu/linux)
  Xref: usenet.stanford.edu gnu.emacs.help:217422

Again Gnus and News.  Can't tell.

Ah!  Here is one.  Another example of a broken mail link.

  https://lists.gnu.org/archive/html/help-gnu-emacs/2016-02/threads.html#00390

Original:

  From: Eli Zaretskii <eliz@gnu.org>
  Message-Id: <83mvqmd1mo.fsf@gnu.org>
  References: <2b5da096-ab4d-4803-9a62-dbed5c036514@googlegroups.com>

Follow-up:

  From: Vladimir Panteleev <thecybershadow@gmail.com>
  Message-ID: <9f89c7fe-19e2-4d24-89f9-630d830f8b67@googlegroups.com>
  Xref: usenet.stanford.edu gnu.emacs.help:217048
  References: <2b5da096-ab4d-4803-9a62-dbed5c036514@googlegroups.com>
          <mailman.6021.1456560027.843.help-gnu-emacs@gnu.org>
  In-Reply-To: <mailman.6021.1456560027.843.help-gnu-emacs@gnu.org>
  User-Agent: G2/1.0

What User-Agent is represented by "G2/1.0"?  Since that isn't Gnus it
would seem to let Gnus off the hook.  The commonality is the Stanford
news gateway.  The converted message id is of the same form.  Looks to
be the same problem.  Does not include Gnus so seems to implicate the
Stanford newsgroup gateway.

Bob

Re: Reply to list [was: Different key maps in different dired buffers]

[Emanuel Berg]

Bob Proulx <bob@proulx.com> writes:

> Yes. I have tried to determine why this is
> before. I don't think it is on Emanuel's end.
> It appears with other posters too. Emanual is
> just one of the more prolific posters and so
> we see it more often with his posts. :-)

You better believe it!

Anyway this message is not by way of
gnu.emacs.help but instead gmane.emacs.help -
did it help in this respect?

-- 
underground experts united .... http://user.it.uu.se/~embe8573
Emacs Gnus Blogomatic ......... http://user.it.uu.se/~embe8573/blogomatic
                   - so far: 48 Blogomatic articles -                   

Re: Reply to list [was: Different key maps in different dired buffers]

[Stefan Monnier]

>> Gmane actually *does* allow posting trough a web interface.  However, I
>> doubt that anyone uses it — it’s pretty ugly, imho, and requires solving
>> a captcha.
> Really?  I thought most people that used Gmane used it for posting
> through their interface.

I'm not sure what you mean.  But I expect that most users of Gmane use
it as an NNTP server.  They probably do use it for reading *and*
posting, but the web-interface for posting via Gmane is probably *much*
less frequently used than the NNTP-interface.


        Stefan

Re: Reply to list [was: Different key maps in different dired buffers]

[Dmitry Alexandrov]

Stefan Monnier <monnier@iro.umontreal.ca> writes:

>>> Gmane actually *does* allow posting trough a web interface.  However, I
>>> doubt that anyone uses it — it’s pretty ugly, imho, and requires solving
>>> a captcha.
>>
>> Really?  I thought most people that used Gmane used it for posting
>> through their interface.
>
> I'm not sure what you mean.  But I expect that most users of Gmane use
> it as an NNTP server.  They probably do use it for reading *and*
> posting, but the web-interface for posting via Gmane is probably *much*
> less frequently used than the NNTP-interface.

Well, in terms of most *people* Bob is probably right.  Consider a
random passerby who came to http://comments.gmane.org from Google search
and decided to post a comment — he clicked ‘reply’, entered something,
solved captcha, clicked ‘post’, and never came back.  He might not even
realized that it was a mailing list, but +1 person to users of Gmane.

Re: Reply to list [was: Different key maps in different dired buffers]

[Dmitry Alexandrov]

Emanuel Berg <embe8573@student.uu.se> writes:

>> By the way, so far as we are discussing
>> techniques of using mailing lists, let me
>> turn your attention, that your messages are
>> systematically appear in ‘gmane.emacs.help’
>> as replies to non-existent messages, thus
>> breaking the tree. That might be some bug of
>> Gmane, of course, however the archive of GNU
>> Mailman [0] shares this views.
>
> No, that was the case with the "Gnus IMAP
> groups appear twice" issue, right?
> Or does it happen all the time?

From time to time, but quite often.

> For example, does it happen now?

Yes, it actually does.  You could check it by yourself with a help of
Gmane’s web-interface [0], where this message is shown on the top level
(not belonging to any thread); or GNU Mailman archive [1], which seems
to be smarter as it can associate this message with the right thread,
but shows ‘Message not available’ as its parent; or (I believe) with a
fresh Gnus instance [3].

[0] http://news.gmane.org/gmane.emacs.help
[1] https://lists.gnu.org/archive/html/help-gnu-emacs/2016-06/threads.html
[2] $ HOME='/tmp' NNTPSERVER='news.gmane.org' emacs --eval '(gnus)'

Re: Reply to list [was: Different key maps in different dired buffers]

[Dmitry Alexandrov]

On 04/06/16 22:41, Bob Proulx wrote:
> Yes.  I have tried to determine why this is before.  I don't think it
> is on Emanuel's end.

I did not imply it either, of course.  However, if the problematic
software or service might be avoided, it’d better be avoided, I believe.
But it have to be found first, of course.

> Details follow

Good investigation. :-)

> There is no In-Reply-To on the follow-up.  Only the References header
> can be used for threading.  Checking through the references in the
> place of my message 20160603195221435910074@bob.proulx.com I see
> mailman.789.1465006376.1216.help-gnu-emacs@gnu.org instead.  That
> appears to be the point where threading gets broken.  Something
> appears to have converted one message id into another one.
>
> <...>
>
> Again Gnus and News.  Can't tell.
>
> Ah!  Here is one.  Another example of a broken mail link.
>
> https://lists.gnu.org/archive/html/help-gnu-emacs/2016-02/threads.html#00390
>
> Original:
>
>   From: Eli Zaretskii <eliz@gnu.org>
>   Message-Id: <83mvqmd1mo.fsf@gnu.org>
>   References: <2b5da096-ab4d-4803-9a62-dbed5c036514@googlegroups.com>
>
> Follow-up:
>
>   From: Vladimir Panteleev <thecybershadow@gmail.com>
>   Message-ID: <9f89c7fe-19e2-4d24-89f9-630d830f8b67@googlegroups.com>
>   Xref: usenet.stanford.edu gnu.emacs.help:217048
>   References: <2b5da096-ab4d-4803-9a62-dbed5c036514@googlegroups.com>
>           <mailman.6021.1456560027.843.help-gnu-emacs@gnu.org>
>   In-Reply-To: <mailman.6021.1456560027.843.help-gnu-emacs@gnu.org>
>   User-Agent: G2/1.0
>
> What User-Agent is represented by "G2/1.0"?

 From the ‘Path’ header I can guess that this is Google Groups web
interface [0].

[0] https://groups.google.com/forum/#!forum/gnu.emacs.help

> Since that isn't Gnus it would seem to let Gnus off the hook.  The
> commonality is the Stanford news gateway.  The converted message id is
> of the same form.  Looks to be the same problem.  Does not include
> Gnus so seems to implicate the Stanford newsgroup gateway.

I’ll try to post this message to ‘gnu.emacs.help‘ via a proper news
server (news.gnus.org) and not with Gnus, let see what will happen.

Re: Reply to list [was: Different key maps in different dired buffers]

[Dmitry Alexandrov]

>> Since that isn't Gnus it would seem to let Gnus off the hook.  The
>> commonality is the Stanford news gateway.  The converted message id is
>> of the same form.  Looks to be the same problem.  Does not include
>> Gnus so seems to implicate the Stanford newsgroup gateway.
>
> I’ll try to post this message to ‘gnu.emacs.help‘ via a proper news
> server (news.gnus.org) and not with Gnus, let see what will happen.

That is it!  References are broken.

So what we could do with it, besides avoiding posting here via
‘gnu.emacs.help’?

It would be logical to report this issue to the gateway maintainers (or
how else should be they called?) at Stanford, I have no idea though,
where to report to them to.

Or is it <listmaster@gnu.org>’s (is there any?) business to deal with
broken gateways?

Re: Reply to list [was: Different key maps in different dired buffers]

[Emanuel Berg]

Dmitry Alexandrov <321942@gmail.com> writes:

> $ HOME='/tmp' NNTPSERVER='news.gmane.org' emacs --eval '(gnus)'

I get this error:

    Cannot open load file: no such file or directory, image

-- 
underground experts united .... http://user.it.uu.se/~embe8573
Emacs Gnus Blogomatic ......... http://user.it.uu.se/~embe8573/blogomatic
                   - so far: 48 Blogomatic articles -                   

Re: Reply to list [was: Different key maps in different dired buffers]

[Bob Proulx]

Emanuel Berg wrote:
> Anyway this message is not by way of gnu.emacs.help but instead
> gmane.emacs.help - did it help in this respect?

Yes.  This message threaded okay.  In my mail reader it shows as a
reply to my message correctly.  And here it is in the archive.

  http://lists.gnu.org/archive/html/help-gnu-emacs/2016-06/threads.html#00148

Doing it that way worked okay.

Bob

Re: Reply to list [was: Different key maps in different dired buffers]

[Emanuel Berg]

Bob Proulx <bob@proulx.com> writes:

>> Anyway this message is not by way of
>> gnu.emacs.help but instead gmane.emacs.help
>> - did it help in this respect?
>
> Yes. This message threaded okay. In my mail
> reader it shows as a reply to my message
> correctly. And here it is in the archive.

Anyone else feels like getting on nntp.aioe.org
to send a reply on gnu.emacs.help?

You know the definition of insanity proposed by
General Lee - trying the same thing over and
over, each time getting a different result :)

-- 
underground experts united .... http://user.it.uu.se/~embe8573
Emacs Gnus Blogomatic ......... http://user.it.uu.se/~embe8573/blogomatic
                   - so far: 48 Blogomatic articles -                   

Re: Reply to list [was: Different key maps in different dired buffers]

[Bob Proulx]

Dmitry Alexandrov wrote:
> >> ... seems to implicate the Stanford newsgroup gateway.
> >
> > I’ll try to post this message to ‘gnu.emacs.help‘ via a proper news
> > server (news.gnus.org) and not with Gnus, let see what will happen.
> 
> That is it!  References are broken.

I think it is confirmed.

> So what we could do with it, besides avoiding posting here via
> ‘gnu.emacs.help’?

That is the problem.  It is available.  People will use it.  We can't
stop people from posting to the news group.

> It would be logical to report this issue to the gateway maintainers (or
> how else should be they called?) at Stanford, I have no idea though,
> where to report to them to.

Now that we have determined it to be the problem I guess we should
contact them and see if they can address the problem somehow.  I don't
know who maintains the gateway.  I have never had any interaction with
them.

> Or is it <listmaster@gnu.org>’s (is there any?) business to deal with
> broken gateways?

Well... I think that is me!

AFAIK the listmaster address is now untended.  AFAIK it was replaced
with mailman@gnu.org for use with Mailman.  AFAIK (a lot of softening
here) that happened around 2003 according to notes left behind in
files so it has been a while.  Regardless of the actual history it
predates my involvement by many years.  Which is why I have AFAIK so
many times.

These days the mailing list admins on the non-FSF side of things are
myself and Karl.  Karl and I handle almost all of the mailing list day
to day activity.  That includes all of the anti-spam and helping
people with subscription and with large file problems and so forth.
The FSF admins maintain the Mailman installation and Mailman upgrades.
But they don't get involved in the day to day maintenance.

And so you could write to mailman@gnu.org but that would just fall
through to me or Karl at some point. :-) You could write to
help-gnu-emacs-owner@gnu.org and that would reach the owners of this
mailing list.  Which includes several other people in addition to
myself and the -owner address is always a good thing to do when
dealing with a specific mailing list.  Anyone of the group there could
help.

Meanwhile, while investigating the listmaster address, I ran across
another related untended mailbox that is 1.2G in size holding 71345
messages piling up from 2006 that all appear to be spam.  Oh joy.
Not! :-)

In any case, I will try to chase down who is the maintainer of the
Stanford news to mail gateway and contact them about the problem.  It
has been this way for years.  It will probably take a while to get
fixed.

Bob

Re: Reply to list [was: Different key maps in different dired buffers]

[Emanuel Berg]

Bob Proulx <bob@proulx.com> writes:

> That is the problem. It is available.
> People will use it. We can't stop people from
> posting to the news group.

...!?

Why would anyone want to stop people from
using it?

And why it is a problem that it is available?

-- 
underground experts united .... http://user.it.uu.se/~embe8573
Emacs Gnus Blogomatic ......... http://user.it.uu.se/~embe8573/blogomatic
                   - so far: 48 Blogomatic articles -                   

Re: Reply to list [was: Different key maps in different dired buffers]

[Bob Proulx]

Emanuel Berg wrote:
> Bob Proulx writes:
> > That is the problem. It is available.
> > People will use it. We can't stop people from
> > posting to the news group.
> 
> ...!?
> 
> Why would anyone want to stop people from
> using it?
> 
> And why it is a problem that it is available?

The message you responded to told it all.  I will repeat it for you.

> Dmitry Alexandrov wrote:
> > That is it!  References are broken.
> >
> > So what we could do with it, besides avoiding posting here via
> > ‘gnu.emacs.help’?
> 
> That is the problem.  It is available.  People will use it.  We can't
> stop people from posting to the news group.

At this moment the only way to avoid the problem of broken References
is to avoid using it.  But it is available.  Therefore people will
continue to use it.  Therefore we will continue to see broken
References whenever anyone posts through the news gateway.

Bob

Re: Reply to list [was: Different key maps in different dired buffers]

[Emanuel Berg]

Bob Proulx <bob@proulx.com> writes:

> At this moment the only way to avoid the
> problem of broken References is to avoid
> using it.

And to do that would be to burn down the house
in order to kill the rats.

Actually, there is a way to "avoid the problem
of broken References", and that is to use Gnus
and gnu.emacs.help.

Because to me threads always looked good.

> But it is available. Therefore people will
> continue to use it.

People use it because it is *useful*.

-- 
underground experts united .... http://user.it.uu.se/~embe8573
Emacs Gnus Blogomatic ......... http://user.it.uu.se/~embe8573/blogomatic
                   - so far: 48 Blogomatic articles -                   

Re: Reply to list [was: Different key maps in different dired buffers]

[Dmitry Alexandrov]

Bob Proulx <bob@proulx.com> writes:

> Dmitry Alexandrov wrote:
>> >> ... seems to implicate the Stanford newsgroup gateway.
>> >
>> > I’ll try to post this message to ‘gnu.emacs.help‘ via a proper news
>> > server (news.gnus.org) and not with Gnus, let see what will happen.
>>
>> That is it!  References are broken.
>
> I think it is confirmed.
>
>> So what we could do with it, besides avoiding posting here via
>> ‘gnu.emacs.help’?
>
> That is the problem.  It is available.  People will use it.  We can't
> stop people from posting to the news group.

Well, being a listmaster, actually, you could, I guess.  I am not
familiar with GNU Mailman, but there must be sophisticated filtering
features.  But that would be too radical solution, of course.

Instead, I imagined, sending to anyone who posts via Usenet an
auto-reply mail that explains the problem and advises switching to Gmane
might be feasible as a temporary measure.

I’ve started to dig through GNU Mailman’s documentation (which seems to
be not ready yet) to find something about auto-responder and ran across
a note that Mailman has NNTP support on its own.

Then I’ve searched the Web a bit and found a page [0] which hints, that
usenet.stanford.org was blamed unfairly — it is a mere Usenet server,
while mail-to-news gating is done here, on gnu.org.

[0] https://www.gnu.org/usenet/usenet.html

Is that page still actual?

>> It would be logical to report this issue to the gateway maintainers (or
>> how else should be they called?) at Stanford, I have no idea though,
>> where to report to them to.
>
> Now that we have determined it to be the problem I guess we should
> contact them and see if they can address the problem somehow.  I don't
> know who maintains the gateway.  I have never had any interaction with
> them.
>
>> Or is it <listmaster@gnu.org>’s (is there any?) business to deal with
>> broken gateways?
>
> Well... I think that is me!

Nice to meet you, by the way.  :-)

> AFAIK the listmaster address is now untended.  AFAIK it was replaced
> with mailman@gnu.org for use with Mailman.  AFAIK (a lot of softening
> here) that happened around 2003 according to notes left behind in
> files so it has been a while.  Regardless of the actual history it
> predates my involvement by many years.  Which is why I have AFAIK so
> many times.
>
> These days the mailing list admins on the non-FSF side of things are
> myself and Karl.  Karl and I handle almost all of the mailing list day
> to day activity.  That includes all of the anti-spam and helping
> people with subscription and with large file problems and so forth.
> The FSF admins maintain the Mailman installation and Mailman upgrades.
> But they don't get involved in the day to day maintenance.
>
> And so you could write to mailman@gnu.org but that would just fall
> through to me or Karl at some point. :-) You could write to
> help-gnu-emacs-owner@gnu.org and that would reach the owners of this
> mailing list.  Which includes several other people in addition to
> myself and the -owner address is always a good thing to do when
> dealing with a specific mailing list.  Anyone of the group there could
> help.

Thank you for the details.

> Meanwhile, while investigating the listmaster address, I ran across
> another related untended mailbox that is 1.2G in size holding 71345
> messages piling up from 2006 that all appear to be spam.  Oh joy.
> Not! :-)
>
> In any case, I will try to chase down who is the maintainer of the
> Stanford news to mail gateway and contact them about the problem.  It
> has been this way for years.  It will probably take a while to get
> fixed.

Re: Reply to list [was: Different key maps in different dired buffers]

[Emanuel Berg]

Dmitry Alexandrov <321942@gmail.com> writes:

> Instead, I imagined, sending to anyone who
> posts via Usenet an auto-reply mail that
> explains the problem and advises switching to
> Gmane might be feasible as
> a temporary measure.

There are problems associated with auto-replies
that are much worse than some malfunction that
amounts to inconsistent references...

Did you locate the software that does the
translation between Usenet posts and mails?

Because this sounds like a mere BUG which, God
willing, can be identified and fixed with but
a small effort...

-- 
underground experts united .... http://user.it.uu.se/~embe8573
Emacs Gnus Blogomatic ......... http://user.it.uu.se/~embe8573/blogomatic
                   - so far: 48 Blogomatic articles -                   

Re: Reply to list [was: Different key maps in different dired buffers]

[Bob Proulx]

Dmitry Alexandrov wrote:
> Bob Proulx writes:
> > That is the problem.  It is available.  People will use it.  We can't
> > stop people from posting to the news group.
> 
> Well, being a listmaster, actually, you could, I guess.  I am not
> familiar with GNU Mailman, but there must be sophisticated filtering
> features.  But that would be too radical solution, of course.

I am sure you have heard the wisdom, "With great power comes great
responsibility."  I take that very seriously and try very hard not to
abuse positions of responsibility.  I would see unilaterally breaking
off the newsgroup from the mailing list where it has been working for
many years to be an abuse of that trust.

As to your comment about Mailman having sophisticated filtering,
actually no it doesn't.  Mailman is actually pretty simple in terms of
available features.  But it is sufficient.  Just barely.

> Instead, I imagined, sending to anyone who posts via Usenet an
> auto-reply mail that explains the problem and advises switching to Gmane
> might be feasible as a temporary measure.

That would potentially create a lot of backscatter spam.  Not good.
Let's not do it.

And for the people who solely use usenet the newsgroups appear to them
to be the primary interface.  It is completely symetrical and depends
upon your point of view.

> I’ve started to dig through GNU Mailman’s documentation (which seems to
> be not ready yet) to find something about auto-responder and ran across
> a note that Mailman has NNTP support on its own.

AFAIK Mailman 2 does not have NNTP support.  Are you reading about
Mailman 3 perhaps?  I doubt there will be a migration to 3 anytime
soon.  The FSF would need to do it and everything about it is
different.  It would be large undertaking and time resources are not
available for it.

> Then I’ve searched the Web a bit and found a page [0] which hints, that
> usenet.stanford.org was blamed unfairly — it is a mere Usenet server,
> while mail-to-news gating is done here, on gnu.org.
> 
> [0] https://www.gnu.org/usenet/usenet.html
> 
> Is that page still actual?

Hmm...  Very interesting.  That is very possible.  And since it is
documented there I think very likely!  But I think anyone who ever had
anything to do with it has retired from the project.  We will need to
re-learn it.  I will kick that over to the FSF admins and see what
they say about it.

Bob

Re: Reply to list [was: Different key maps in different dired buffers]

[Emanuel Berg]

Bob Proulx <bob@proulx.com> writes:

>> https://www.gnu.org/usenet/usenet.html
>> 
>> Is that page still actual?
>
> Hmm... Very interesting. That is very
> possible. And since it is documented there
> I think very likely! But I think anyone who
> ever had anything to do with it has retired
> from the project. We will need to re-learn
> it. I will kick that over to the FSF admins
> and see what they say about it.

Ha ha, "gnUSENET" :)

Find out if it is real on gnu.gnusenet.test...

Here are the groups what I can find on
nntp.aieo.org -

    K     58: gnu.announce
    K    438: gnu.bash.bug
    K      0: gnu.cfengine.bug
    K      1: gnu.cfengine.help
    K      0: gnu.chess
    K      2: gnu.chess.bug
    K      1: gnu.cvs.bug
    K      0: gnu.cvs.help
    K      3: gnu.emacs.announce
    K   5088: gnu.emacs.bug
    K      1: gnu.emacs.gnews
    K    108: gnu.emacs.gnus
    U    978: gnu.emacs.help
    K     77: gnu.emacs.sources
    K      0: gnu.emacs.vm.bug
    K      8: gnu.emacs.vm.info
    K      0: gnu.emacs.vms
    K      0: gnu.g++.announce
    K      1: gnu.g++.bug
    K      2: gnu.g++.help
    K      0: gnu.g++.lib.bug
    K      0: gnu.gcc.announce
    K      0: gnu.gcc.bug
    K      4: gnu.gcc.help
    K      0: gnu.gdb.bug
    K      0: gnu.ghostscript.bug
    K      0: gnu.glibc.bug
    K      0: gnu.gnats.bug
    K      0: gnu.gnusenet.config
    K      1: gnu.gnusenet.test
    K      1: gnu.gnustep.announce
    K     40: gnu.gnustep.bug
    K    142: gnu.gnustep.discuss
    K     23: gnu.gnustep.help
    K      3: gnu.groff.bug
    K    299: gnu.hurd.bug
    K      1: gnu.hurd.help
    K      2: gnu.misc.discuss
    K      0: gnu.smalltalk.bug

-- 
underground experts united .... http://user.it.uu.se/~embe8573
Emacs Gnus Blogomatic ......... http://user.it.uu.se/~embe8573/blogomatic
                   - so far: 48 Blogomatic articles -