Banner Orgy

Banner Orgy

[Robin S. Socha]

OK, so how do I get rid of /this/?

,----
| Maildrop mailing list - http://www.flounder.net/~mrsam/maildrop/
| 
| --------------------------- ListBot Sponsor --------------------------
| Get fast, easy info by phone: Call 800-555-TELL.
|    News, weather, restaurants... & much more!
| 
| http://on.linkexchange.com/?ATID=27&AID=2143
| ----------------------------------------------------------------------
| 
| [Article]
| 
| ______________________________________________________________________
| To unsubscribe, write to maildropl-unsubscribe@listbot.com
`----

Thanks!
Robin
-- 
Robin S. Socha <http://socha.net/>
"The new glue is, unfortunately, ignored by recent versions of the BIND
cache; the detailed technical explanation for this is that the BIND
company is a bunch of idiots." (DJB)

Re: Banner Orgy

[Michael Livshin]

"Robin S. Socha" <robin@socha.net> writes:

> OK, so how do I get rid of /this/?

you bug the list owner to move the list to, say, eGroups? ;)

-- 
(only legal replies to this address are accepted)

This machine is a piece of GAGH! I need quad processors if I am to do
battle with this code!
                                        -- Klingon Programmer

Re: Banner Orgy

[Raymond Scholz]

"Robin S. Socha" <robin@socha.net> schrieb:

> OK, so how do I get rid of /this/?

Something like

(banner . "Maildrop mailing list - http://www.flounder.net/~mrsam/maildrop/\n\\(.*\n\\)+To unsubscribe, write to maildropl-unsubscribe@listbot.com$")

in the Group Parameters should do that.

Cheers,
  Ray
-- 
Raymond Scholz - rscholz@zonix.de - PGP - http://www.zonix.de/

Re: Banner Orgy

[Toby Speight]

0> In article <deathsquad.m366j1ap84.fsf@socha.net>,
0> Robin S. Socha <URL:mailto:robin@socha.net> ("Robin") wrote:

Robin> OK, so how do I get rid of /this/?

 ,----
 | Maildrop mailing list - http://www.flounder.net/~mrsam/maildrop/
 |
 | --------------------------- ListBot Sponsor --------------------------
 | Get fast, easy info by phone: Call 800-555-TELL.
 |    News, weather, restaurants... & much more!
 |
 | http://on.linkexchange.com/?ATID=27&AID=2143
 | ----------------------------------------------------------------------
 |
 | [Article]
 |
 | ______________________________________________________________________
 | To unsubscribe, write to maildropl-unsubscribe@listbot.com
 `----

With a regexp something like

(concat "^Maildrop mailing list - .*\n\n"
        "----+ ListBot Sponsor ----+\n"
        "[^-].*\n"
        "----------------------+\n"
        "\\|"
        "^----------------------+\n"
        "To unsubscribe, .+\n*\\'")

Note that both banners will be matched, and both removed.  You may
want to allow for whitespace before each newline.

Here's a more complicated one I have for eGroups (which has changed
its format several times and which uses a different style in multipart
messages), which also strips quoted banners (bloody lookout users!):

> (concat "^\\(\\([ \t>]*>\\)?\\)[ \t]*\\(\n\\1[ \t]*\\)*"
>         "\\("
>         "To unsubscribe from .*\n\\1.*unsubscribe@egroups.com\n"
>         "\\|-------------------+\\( eGroups Sponsor -+\\)?....\n\\(\\1.+\n\\)+"
>         "\\|eGroups Sponsor[ \t]*\\(\n\\1[ \t]*\\)+<[^<>]+>\\([ \t]*\n\\1?\\)+<[^<>]+>\\([ \t]*\n\\1?\\)*"
>         "\\|eGroups Sponsor[ \t]*\\(\n\\1[ \t]*\\)+<[^<>]+> .*\\(\n\\1?[ \t]*\\)*"
>         "\\|eGroups Sponsor[ \t]*\\(\n\\1[ \t]*\\)+<http:\\(\n\\1\\|[^\n<>]\\)+> [Cc]lick here!?\\s-*$"
>         "\\)"
>         "\\(\\1?\n\\)*")

Re: Banner Orgy

[Raymond Scholz]

Raymond Scholz <ray-2001@zonix.de> schrieb:

> Something like
[really stupid regexp]

Sorry, my solution comletely sucks, if you are interested in reading
the article itself.  Toby's posting just opened by eyes, I overlooked
the [Article] placeholder :)

Cheers,
  Ray
-- 
Raymond Scholz - rscholz@zonix.de - PGP - http://www.zonix.de/

Am I stupid? (was: Banner Orgy)

[Robin S. Socha]

* Toby Speight <streapadair@gmx.net> writes:
>  Robin S. Socha <URL:mailto:robin@socha.net> ("Robin") wrote:

>>  OK, so how do I get rid of /this/?

[unspeakable horror]

> With a regexp something like

[potential solution]

OK. So I edited the Group paramaters (no keyboard shortcut for that
one?) - and as soon as I had said C-c C-c: *pooof* gone. What gives?
-- 
Robin S. Socha  <http://socha.net/Gnus/>

URGH! (was: Banner Orgy)

[Robin S. Socha]

[-- Attachment #1: Type: text/plain, Size: 123 bytes --]

* Toby Speight <streapadair@gmx.net> writes:

> With a regexp something like

Kinda following up to myself - what is this?

[-- Attachment #2: URGH --]
[-- Type: image/png, Size: 4046 bytes --]

[-- Attachment #3: Type: text/plain, Size: 45 bytes --]

-- 
Robin S. Socha  <http://socha.net/Gnus/>

Group parameters keybinding (was: Re: Am I stupid?, Banner Orgy)

[Charles Sebold]

On 6 Shevat 5761, Robin S. Socha wrote:

> OK. So I edited the Group paramaters (no keyboard shortcut for that
> one?)

G p works for me in the Groups buffer.  (Sorry no answers for the banner
questions.)
-- 
Charles Sebold
Random Answer to an Emacs Very Frequently Asked Question:
 M-x apropos is your friend, C-h a is not the same thing.
--
6th of Shevat, 5761
--
Well, folks, we can tell you what Ginger is. 

It's a manifestation of the sick modern world where style is more
important that substance, where perception is king, where people screw
their neighbour to buy an overpriced bit of clothing with a particular
name on and where the press report a story because other parts of the
press have reported it and so it must be a story.

 -- Kieran McCarthy, http://www.theregister.co.uk/content/28/16026.html

Re: Am I stupid? (was: Banner Orgy)

[Toby Speight]

0> In article <deathsquad.m31ytlsymc.fsf@socha.net>,
0> Robin S. Socha <URL:mailto:robin@socha.net> ("Robin") wrote:

Robin> OK. So I edited the Group paramaters (no keyboard shortcut for that
Robin> one?)

G p  in the *Group* buffer.

Robin> - and as soon as I had said C-c C-c: *pooof* gone. What gives?

I'm not sure whether I remembered to say that my solution was a Lisp
expression that must be evaluated and the *result* entered in the
Group parameters as

( ;; group parameters
 (banner . "<long result string>"))

I used `concat' to split it over several lines for clarity and
line-length, but unfortunately the group parameter is not evaluated
(Lars, ShengHuo, feature request?).

Re: Am I stupid? (was: Banner Orgy)

[Per Abrahamsen]

Toby Speight <streapadair@gmx.net> writes:

> I used `concat' to split it over several lines for clarity and
> line-length, but unfortunately the group parameter is not evaluated
> (Lars, ShengHuo, feature request?).

It would make the form interface, which I assume most people use, more
complex for very little gain.

Re: Am I stupid? (was: Banner Orgy)

[ShengHuo ZHU]

Toby Speight <streapadair@gmx.net> writes:

> I'm not sure whether I remembered to say that my solution was a Lisp
> expression that must be evaluated and the *result* entered in the
> Group parameters as
> 
> ( ;; group parameters
>  (banner . "<long result string>"))
> 
> I used `concat' to split it over several lines for clarity and
> line-length, but unfortunately the group parameter is not evaluated
> (Lars, ShengHuo, feature request?).

You can use concat in gnus-article-banner-alist (since 5.8.8), then
put a simple symbol instead of long result string in group parameters.

ShengHuo

-- 
(setq gnus-posting-styles '((".*" (signature-file "~/.signature"))))

Re: Am I stupid? (was: Banner Orgy)

[Kai Großjohann]

On 31 Jan 2001, Per Abrahamsen wrote:

> It would make the form interface, which I assume most people use,
> more complex for very little gain.

I wish people were using `G c' rather than `G p', but it appears they
don't.  At least there are a lot of questions due to hosed group
parameters, which could have been avoided using `G c'.

Why are people using `G p' and not `G c'?  Do they not know about `G
c'?  Or was the explanation not promising enough?

kai
-- 
Be indiscrete.  Do it continuously.

Re: Am I stupid? (was: Banner Orgy)

[Jesper Harder]

Kai.Grossjohann@CS.Uni-Dortmund.DE (Kai Großjohann) writes:

> Why are people using `G p' and not `G c'?  Do they not know about `G
> c'?

Probably not -- only 'G p' is mentioned in the "Group Parameters" info
node.

Re: Am I stupid? (was: Banner Orgy)

[Paul Jarc]

Kai.Grossjohann@CS.Uni-Dortmund.DE (Kai Großjohann) writes:
> Why are people using `G p' and not `G c'?  Do they not know about `G
> c'?  Or was the explanation not promising enough?

I didn't know about 'G c'.  (But I do now, thanks.)  I expect I'll
still use 'G p' when I want to find out what's really in there, but if
I remember, I'll use 'G c' to make changes.

Side note: is it possible for backends to receive notification when
the user changes group parameters, so that they could be stored in the
backend, like marks?  Or is there some reason why this would be a bad
idea?


paul

Re: Am I stupid? (was: Banner Orgy)

[Steven E. Harris]

Kai.Grossjohann@CS.Uni-Dortmund.DE (Kai Großjohann) writes:

> Why are people using `G p' and not `G c'?  Do they not know about `G
> c'?  Or was the explanation not promising enough?

I usually use `G p' because I'm almost always *copying* a parameter
from one group to another. Once I find a setting I like on one group,
I'll copy it around and modify it as necessary, possibly promoting it
to a Topic parameter. `G c' just seems a little too heavyweight for
this purpose, as my copy-and-paste keystrokes aren't as fast.

-- 
Steven E. Harris        :: steven.harris@tenzing.com
Tenzing                 :: http://www.tenzing.com

Group parameters Customize (was Re: Am I stupid? (was: Banner Orgy))

[Raja R Harinath]

Hi,

Kai.Grossjohann@CS.Uni-Dortmund.DE (Kai Großjohann) writes:
> On 31 Jan 2001, Per Abrahamsen wrote:
> > It would make the form interface, which I assume most people use,
> > more complex for very little gain.
> 
> I wish people were using `G c' rather than `G p', but it appears they
> don't.  At least there are a lot of questions due to hosed group
> parameters, which could have been avoided using `G c'.
> 
> Why are people using `G p' and not `G c'?  Do they not know about `G
> c'?  Or was the explanation not promising enough?

'G c' doesn't seem to handle all of supported syntax of the group
parameters.  For example, 

  (total-expire
   (expiry-wait . 7))

is a valid group parameter, but 'G c' on a group with that setting
will complain or misbehave (like not showing the "Total Expire" box as
being checked, IIRC).  'G c' appears to support only

  ((total-expire . t)
   (expiry-wait . 7))

(I may be wrong about a couple of parantheses, but I guess you get the
idea). 

This makes it painful, at least to me, to migrate to using 'G c',
since I'll have to remember to 'G p' beforehand to fix up.

- Hari
-- 
Raja R Harinath ------------------------------ harinath@cs.umn.edu
"When all else fails, read the instructions."      -- Cahn's Axiom
"Our policy is, when in doubt, do the right thing."   -- Roy L Ash

Re: Am I stupid? (was: Banner Orgy)

[Simon Josefsson]

prj@po.cwru.edu (Paul Jarc) writes:

> Side note: is it possible for backends to receive notification when
> the user changes group parameters, so that they could be stored in the
> backend, like marks?  Or is there some reason why this would be a bad
> idea?

Nnimap used to send flags to server when they were set (in a ad-hoc
way), but it was changed to mimic all other backends -- flags are only
saved when you `q' a group.  `Q' or `C-k' simply forgets about what
you've been doing in a mailbox (modulo `B DEL's and similar).

Changing this now would probably be confusing..

Re: Am I stupid? (was: Banner Orgy)

[Harry Putnam]

Kai.Grossjohann@CS.Uni-Dortmund.DE (Kai Großjohann) writes:

> On 31 Jan 2001, Per Abrahamsen wrote:
> 
> > It would make the form interface, which I assume most people use,
> > more complex for very little gain.
> 
> I wish people were using `G c' rather than `G p', but it appears they
> don't.  At least there are a lot of questions due to hosed group
> parameters, which could have been avoided using `G c'.
> 
> Why are people using `G p' and not `G c'?  Do they not know about `G
> c'?  Or was the explanation not promising enough?

Butting in here a bit... I think `G c' is a good way to learn how to
edit Group params since you can do `G c' then look with `G p' at what was
written.  Soon you have an idea how it is done.  

Once you know how its done, then `G c' is really pretty cumbersome
compared to a quick hand edit.  However, this will no doubt lead to
more errors.

I even find myself looking at groups with `G p' for  something
similar to what I'm currently after, rather than use `G c'.  

`G c' is very nice but in fact is slow and cumbersome.

Re: Am I stupid? (was: Banner Orgy)

[Kai Großjohann]

On Wed, 07 Feb 2001, Jesper Harder wrote:
> Kai.Grossjohann@CS.Uni-Dortmund.DE (Kai Großjohann) writes:
> 
>> Why are people using `G p' and not `G c'?  Do they not know about
>> `G c'?
> 
> Probably not -- only 'G p' is mentioned in the "Group Parameters"
> info node.

Eeek.  I'll add it right away.  Thanks!

kai
-- 
Be indiscrete.  Do it continuously.

Re: Am I stupid? (was: Banner Orgy)

[Kai Großjohann]

On 07 Feb 2001, Paul Jarc wrote:

> Side note: is it possible for backends to receive notification when
> the user changes group parameters, so that they could be stored in
> the backend, like marks?  Or is there some reason why this would be
> a bad idea?

Hm.  I guess it's a nice idea.  Never thought about this.  Hm.  Might
be nifty to allow this.

kai
-- 
Be indiscrete.  Do it continuously.

Re: Am I stupid? (was: Banner Orgy)

[Kai Großjohann]

On 07 Feb 2001, Simon Josefsson wrote:
> prj@po.cwru.edu (Paul Jarc) writes:
> 
>> Side note: is it possible for backends to receive notification when
>> the user changes group parameters, so that they could be stored in
>> the backend, like marks?  Or is there some reason why this would be
>> a bad idea?
> 
> Nnimap used to send flags to server when they were set

flags != group parameters

Did nnimap ever store group parameters on the server?

kai
-- 
Be indiscrete.  Do it continuously.

Re: Group parameters Customize (was Re: Am I stupid? (was: Banner Orgy))

[Kai Großjohann]

On 07 Feb 2001, Raja R. Harinath wrote:

> 'G c' doesn't seem to handle all of supported syntax of the group
> parameters.  For example, 
> 
>   (total-expire
>    (expiry-wait . 7))

Ah.  Maybe Gnus should contain code to auto-migrate this to

((total-expire . t)
 (expiry-wait . 7))

I guess that would be a useful thing to do.  Thoughts?

kai
-- 
Be indiscrete.  Do it continuously.

Re: Am I stupid? (was: Banner Orgy)

[Simon Josefsson]

Kai.Grossjohann@CS.Uni-Dortmund.DE (Kai Großjohann) writes:

> >> Side note: is it possible for backends to receive notification when
> >> the user changes group parameters, so that they could be stored in
> >> the backend, like marks?  Or is there some reason why this would be
> >> a bad idea?
> > 
> > Nnimap used to send flags to server when they were set
> 
> flags != group parameters

Oops!   Vocabulary overflow.

> Did nnimap ever store group parameters on the server?

No.  I'm not sure I want that in all cases, I'm using different group
parameters on all three of my Gnus configurations (laptop, work,
school).

But it would be nice to somehow put things on the server, group
parameters and other stuff.  It's really boring to synchronize Gnus
configs.  Pine puts config in a IMAP mailbox.  I think it's ugly, but
ACAP and similar doesn't seem to get much deployment, so I guess it's
the only solution right now...

Re: Group parameters Customize (was Re: Am I stupid? (was: Banner Orgy))

[Paul Jarc]

Kai.Grossjohann@CS.Uni-Dortmund.DE (Kai Großjohann) writes:
> On 07 Feb 2001, Raja R. Harinath wrote:
> >   (total-expire
> >    (expiry-wait . 7))
> 
> Ah.  Maybe Gnus should contain code to auto-migrate this to
> 
> ((total-expire . t)
>  (expiry-wait . 7))
> 
> I guess that would be a useful thing to do.  Thoughts?

My only worry would be that Gnus might accidentally change the meaning
of what was there.  I think I'd prefer if the customization stuff
could be made to understand a wider range of expressions, always
preserving values that the user didn't change, but probably always
writing expressions in some canonical form.  So if you change
something, and then change it back, it might end up as a different
expression with the same meaning, but if you don't change it, you're
guaranteed that customization won't screw it up due to not
understanding it.


paul

backends storing group info (was: Am I stupid?)

[Paul Jarc]

Kai.Grossjohann@CS.Uni-Dortmund.DE (Kai Großjohann) writes:
> On 07 Feb 2001, Paul Jarc wrote:
> > Side note: is it possible for backends to receive notification when
> > the user changes group parameters, so that they could be stored in
> > the backend, like marks?  Or is there some reason why this would be
> > a bad idea?
> 
> Hm.  I guess it's a nice idea.  Never thought about this.  Hm.  Might
> be nifty to allow this.

While we're at it, I'd suggest that we might do this for *all* parts
of the group info.  Gnus gives the backend the opportunity to change
the whole info, but the backend can't take advantage of that if it
can't be notified when the user changes the values.  I suppose the
name is already taken care of with -request-rename-group, and it may
not make sense for a backend to change the select method for a group.
Article marks are handled by -request-set-mark, so that leaves rank
and parameters.

If a backend stores parameters, should it also receive the parameters
of any ancestor topics?  That would reduce the usefulness of
inheriting parameters from topics, since they'd always be overridden
by the full set returned by the backend, so that's probably a bad
idea.  Of course, if we store only the parameters that belong to the
group, then we can't just copy the backend's storage from one machine
to another and get quite the same configuration.  Maybe if topics were
more group-like, and had a virtual servers, and were stored in
backends... that's probably overkill, though.


paul

Re: Group parameters Customize (was Re: Am I stupid? (was: Banner Orgy))

[Per Abrahamsen]

Raja R Harinath <harinath@cs.umn.edu> writes:

> 'G c' doesn't seem to handle all of supported syntax of the group
> parameters.  For example, 
> 
>   (total-expire
>    (expiry-wait . 7))

Is that really valid syntax, or just something that works by accident?
Or does it work at all differently from (total-expire . t)?  Often
"non-nil" is taken for a synonym of "t".

The documentation says:

`total-expire'
     If the group parameter has an element that looks like
     `(total-expire . t)', all read articles will be put through the
     expiry process, even if they are not marked as expirable.  Use with
     caution.  Unread, ticked and dormant articles are not eligible for
     expiry.

> This makes it painful, at least to me, to migrate to using 'G c',
> since I'll have to remember to 'G p' beforehand to fix up.

I don't think it is important to allow "migration", as 'G c' should be
what new users use.  'G p' should be reserved for experts, and maybe
for cut&paste.

Re: Group parameters Customize (was Re: Am I stupid? (was: Banner Orgy))

[ShengHuo ZHU]

Per Abrahamsen <abraham@dina.kvl.dk> writes:

> Raja R Harinath <harinath@cs.umn.edu> writes:
> 
> > 'G c' doesn't seem to handle all of supported syntax of the group
> > parameters.  For example, 
> > 
> >   (total-expire
> >    (expiry-wait . 7))
> 
> Is that really valid syntax, or just something that works by accident?
> Or does it work at all differently from (total-expire . t)?  Often
> "non-nil" is taken for a synonym of "t".

It is valid and works in the same manner (see
gnus-group-parameter-value).

ShengHuo

-- 
(setq gnus-posting-styles 
      '((".*" (signature (format "(setq gnus-posting-styles 
      '%S)" gnus-posting-styles)))))

Re: Group parameters Customize (was Re: Am I stupid? (was: Banner Orgy))

[Per Abrahamsen]

ShengHuo ZHU <zsh@cs.rochester.edu> writes:

> Per Abrahamsen <abraham@dina.kvl.dk> writes:
> 
> > Raja R Harinath <harinath@cs.umn.edu> writes:
> > 
> > > 'G c' doesn't seem to handle all of supported syntax of the group
> > > parameters.  For example, 
> > > 
> > >   (total-expire
> > >    (expiry-wait . 7))
> > 
> > Is that really valid syntax, or just something that works by accident?
> > Or does it work at all differently from (total-expire . t)?  Often
> > "non-nil" is taken for a synonym of "t".
> 
> It is valid and works in the same manner (see
> gnus-group-parameter-value).

Valid according to the code, or valid according to the docmuentation?
I cannot find any support in the documentation for that syntax.

Re: Group parameters Customize (was Re: Am I stupid? (was: Banner Orgy))

[Raja R Harinath]

Per Abrahamsen <abraham@dina.kvl.dk> writes:
> ShengHuo ZHU <zsh@cs.rochester.edu> writes:
> > Per Abrahamsen <abraham@dina.kvl.dk> writes:
> > > Raja R Harinath <harinath@cs.umn.edu> writes:
> > > > 'G c' doesn't seem to handle all of supported syntax of the group
> > > > parameters.  For example, 
> > > > 
> > > >   (total-expire
> > > >    (expiry-wait . 7))
> > > 
> > > Is that really valid syntax, or just something that works by accident?
> > > Or does it work at all differently from (total-expire . t)?  Often
> > > "non-nil" is taken for a synonym of "t".
> > 
> > It is valid and works in the same manner (see
> > gnus-group-parameter-value).
> 
> Valid according to the code, or valid according to the docmuentation?
> I cannot find any support in the documentation for that syntax.

I seem to remember early (pre-customize) versions of the documentation
that talked about this, but I may be mis-remembering.  The closest I
see is from the "Expiring Mail" page:

  If you put `total-expire' in the group parameters, articles will not
  be marked as expirable, but all read articles will be put through
  the expiry process.

(I know that's somewhat of a stretch.)

- Hari
-- 
Raja R Harinath ------------------------------ harinath@cs.umn.edu
"When all else fails, read the instructions."      -- Cahn's Axiom
"Our policy is, when in doubt, do the right thing."   -- Roy L Ash

Re: Group parameters Customize (was Re: Am I stupid? (was: Banner Orgy))

[Per Abrahamsen]

Raja R Harinath <harinath@cs.umn.edu> writes:

> Per Abrahamsen <abraham@dina.kvl.dk> writes:
>
> > Valid according to the code, or valid according to the docmuentation?
> > I cannot find any support in the documentation for that syntax.
> 
> I seem to remember early (pre-customize) versions of the documentation
> that talked about this, but I may be mis-remembering.

The form interface was written straight from the documentation...