Gnus v5.6.24 is released

Gnus v5.6.24 is released

[Lars Magne Ingebrigtsen]

The 498th release -- bug fixes.

Get it from <URL:http://www.gnus.org/gnus.tar.gz> or 
"/ftp@ftp.gnus.org:/pub/emacs/gnus/".  The patch is available as
<URL:http://www.gnus.org/patches/gnus-5.6.23-5.6.24.diff.gz>.

ChangeLog since last release:

Sat Jul 11 03:03:53 1998  Lars Magne Ingebrigtsen  <larsi@menja.ifi.uio.no>

	* gnus.el: Gnus v5.6.24 is released.

Fri Jul 10 04:23:24 1998  Hallvard B. Furuseth  <h.b.furuseth@usit.uio.no>

	* gnus-util.el (gnus-parse-netrc): Allow "default" values.

Fri Jul 10 04:15:35 1998  Lars Magne Ingebrigtsen  <larsi@ifi.uio.no>

	* nntp.el (nntp-server-opened-hook): Doc change.

Fri Jul 10 03:03:48 1998  François Pinard  <pinard@iro.umontreal.ca>

	* gnus-sum.el (gnus-summary-respool-trace): New command and
	keystroke. 

Fri Jul 10 02:18:01 1998  Lars Magne Ingebrigtsen  <larsi@ifi.uio.no>

	* gnus-util.el (gnus-prin1): Bind print-escape-multibyte to nil. 

Mon Jul  6 01:02:59 1998  Simon Josefsson  <jas@pdc.kth.se>

	* gnus-range.el (gnus-sorted-complement): Fix comments.

Thu Jul  2 11:16:14 1998  Lars Magne Ingebrigtsen  <larsi@ifi.uio.no>

	* gnus-sum.el (gnus-summary-iterate): New macro.

	* message.el (message-pop-to-buffer): Clone locals.

	* gnus-msg.el (gnus-posting-styles): Reinstated.
	(gnus-posting-style-alist): Ditto.

Wed Jul  1 18:02:31 1998  Lars Magne Ingebrigtsen  <larsi@ifi.uio.no>

	* gnus-int.el (gnus-get-function): Set funct to nil.

1998-07-01 16:57:38  Simon Josefsson  <jas@pdc.kth.se>

	* gnus-int.el (gnus-get-function): returned non-nil when
        function wasn't bound, if noerror=t

Wed Jul  1 17:30:41 1998  Lars Magne Ingebrigtsen  <larsi@ifi.uio.no>

	* gnus-topic.el (gnus-topic-mode-map): Bind TAB and M-TAB.

	* gnus-sum.el (gnus-build-sparse-threads): Make sure no dates are
	nil. 
	(gnus-summary-limit-mark-excluded-as-read): Use the intersection.

	* gnus-msg.el (gnus-setup-message): Clone all local variables from
	the summary buffer.

Wed Jul  1 14:03:52 1998  Richard Stallman  <rms@santafe.edu>

	* message.el (message-cite-original): Use mail-citation-hook.
	(message-cite-function): Ditto.

Wed Jul  1 14:00:53 1998  Rajappa Iyer  <rsi@lucent.com>

	* gnus-salt.el (gnus-pick-mode-map): Changed keymap.



-- 
(domestic pets only, the antidote for overdose, milk.)
  larsi@ifi.uio.no * Lars Magne Ingebrigtsen

Re: Gnus v5.6.24 is released

[Berry Kercheval]

OOPS.  This file URL:http://www.gnus.org/gnus.tar.gz unpacked into 
gnus-5.6.23, trashing my active installation (well. sorta) instead of
into gnus-5.6.24 as it should.

I can deal with it but you might want to fix it.

Re: Gnus v5.6.24 is released

[Lars Magne Ingebrigtsen]

Berry Kercheval <berry@kerch.com> writes:

> OOPS.  This file URL:http://www.gnus.org/gnus.tar.gz unpacked into 
> gnus-5.6.23, trashing my active installation (well. sorta) instead of
> into gnus-5.6.24 as it should.

Hm.  Looks like the disks ran over or something.  Things should be
better now.

-- 
(domestic pets only, the antidote for overdose, milk.)
  larsi@gnus.org * Lars Magne Ingebrigtsen

Re: Gnus v5.6.24 is released

[Bruce Stephens]

In gnus-topic.el, there's a [M-tab].  Should be [(meta tab)] or
something for XEmacs.

Re: Gnus v5.6.24 is released

[SL Baur]

Bruce Stephens <bruce@cenderis.demon.co.uk> writes:

> In gnus-topic.el, there's a [M-tab].  Should be [(meta tab)] or
> something for XEmacs.

Yup.  Here's a patch.

1998-07-11  SL Baur  <steve@altair.xemacs.org>

	* gnus-topic.el (gnus-topic-mode-map): Use modern key syntax.

===================================================================
RCS file: RCS/gnus-topic.el,v
retrieving revision 1.9
diff -u -r1.9 gnus-topic.el
--- lisp/gnus-topic.el	1998/07/11 23:04:22	1.9
+++ lisp/gnus-topic.el	1998/07/11 23:04:56
@@ -905,7 +905,7 @@
     "#" gnus-topic-mark-topic
     "\M-#" gnus-topic-unmark-topic
     [tab] gnus-topic-indent
-    [M-tab] gnus-topic-unindent
+    [(meta tab)] gnus-topic-unindent
     gnus-mouse-2 gnus-mouse-pick-topic)
 
   ;; Define a new submap.

Re: Gnus v5.6.24 is released

[Dave Love]

Can we have bindings which work everywhere, please (specifically Emacs
on tty, where `tab' isn't generated).

I'm not sure what XEmacs groks generally but this is compatible with
bindings used elsewhere.  There does, however, seem to be
inconsistency between other maps with \C-i and \t, for instance.

1998-07-12  Dave Love  <d.love@dl.ac.uk>

	* gnus-group.el (gnus-group-mode-map): Provide general bindings
	for gnus-group-select-group-ephemerally.

	* gnus-topic.el (gnus-topic-mode-map): Provide general bindings
	for gnus-topic-{un}indent.

--- gnus-topic.el	1998/07/12 16:36:51	1.1
+++ gnus-topic.el	1998/07/12 16:50:29
@@ -904,8 +904,10 @@
     "Gp" gnus-topic-edit-parameters
     "#" gnus-topic-mark-topic
     "\M-#" gnus-topic-unmark-topic
+    "\C-i" gnus-topic-indent
     [tab] gnus-topic-indent
-    [M-tab] gnus-topic-unindent
+    [(meta tab)] gnus-topic-unindent
+    "\M-\t" gnus-topic-unindent
     gnus-mouse-2 gnus-mouse-pick-topic)
 
   ;; Define a new submap.
--- gnus-group.el	1998/07/12 16:31:52	1.1
+++ gnus-group.el	1998/07/12 16:42:42
@@ -425,6 +425,7 @@
     "=" gnus-group-select-group
     "\r" gnus-group-select-group
     "\M-\r" gnus-group-quick-select-group
+    "\C-\M-\r" gnus-group-select-group-ephemerally
     [(meta control return)] gnus-group-select-group-ephemerally
     "j" gnus-group-jump-to-group
     "n" gnus-group-next-unread-group
.

Re: Gnus v5.6.24 is released

[Hrvoje Niksic]

Dave Love <d.love@dl.ac.uk> writes:

> Can we have bindings which work everywhere, please (specifically
> Emacs on tty, where `tab' isn't generated).

In XEmacs, [(meta tab)] binding will do the right thing on TTY's.

> +    "\C-i" gnus-topic-indent
>      [tab] gnus-topic-indent
> -    [M-tab] gnus-topic-unindent
> +    [(meta tab)] gnus-topic-unindent
> +    "\M-\t" gnus-topic-unindent

These will work correctly in XEmacs.

-- 
Hrvoje Niksic <hniksic@srce.hr> | Student at FER Zagreb, Croatia
--------------------------------+--------------------------------
Be nice to your kids.
They'll choose your nursing home.

Re: Gnus v5.6.24 is released

[Bud Rogers]

Stainless Steel Rat <ratinox@peorth.gweep.net> writes:

> -----BEGIN PGP SIGNED MESSAGE-----
> 
> Hrvoje Niksic writes:
> 
> > In XEmacs, [(meta tab)] binding will do the right thing on TTY's.
> 
> In FSF Emacs, M-Tab binding will do the right thing (maps to ESC TAB) in
> ttys.  However, certain window managers, such as my FVWM configuration and
> Microsoft Windows, use the ALT+TAB keys for other purposes.

KDE also captures ALT TAB for its own use to change windows.

-- 

Bud Rogers <budr@tanet.net>
I wore khaki when khaki wasn't cool.

Re: Gnus v5.6.24 is released

[Dave Love]

>>>>> "Rat" == Stainless Steel Rat <ratinox@peorth.gweep.net> writes:

 Rat> In FSF Emacs, M-Tab binding will do the right thing (maps to ESC
 Rat> TAB) in ttys.

What's the value of `FSF Emacs' for which that's true (for any
capitalization of `tab')?
.

Re: Gnus v5.6.24 is released

[Dave Love]

>>>>> "Hrvoje" == Hrvoje Niksic <hniksic@srce.hr> writes:

 Hrvoje> In XEmacs, [(meta tab)] binding will do the right thing on TTY's.

Out of interest, and not able to experiment, will "\M-\t" not DTRT in
XEamcs?  I.e. is that why there are two bindings in some cases?
.

Re: Gnus v5.6.24 is released

[Hrvoje Niksic]

Dave Love <d.love@dl.ac.uk> writes:

> >>>>> "Hrvoje" == Hrvoje Niksic <hniksic@srce.hr> writes:
> 
>  Hrvoje> In XEmacs, [(meta tab)] binding will do the right thing on TTY's.
> 
> Out of interest, and not able to experiment, will "\M-\t" not DTRT in
> XEamcs?

Here is what happens in XEmacs:

(define-key MAP "\M-\t" 'foo)
  ; binds M-C-i and M-tab to `foo'

(define-key MAP [(meta tab)] 'foo)
  ; binds M-tab (but not M-C-i!) to `foo'

(define-key MAP '(meta tab) 'foo)
  ; shorthand for previous

(define-key MAP [(meta control i)] 'foo)
  ; bind M-C-i (but not M-tab!) to `foo'

So, if you use the "old" string notation, you get the old semantics
(C-i and TAB being the same).  If you use the "new" vector notation,
you get new semantics (difference between C-i and TAB).  This is a
feature.

Under TTY's, however, C-i is treated exactly the same as TAB, so all
of the above examples bind ?\M-C-i or `ESC C-i' to `foo'.


Look at the XEmacs docs of `define-key' for more explanation of this
(I can mail the docs to you if you don't have XEmacs handy.)

-- 
Hrvoje Niksic <hniksic@srce.hr> | Student at FER Zagreb, Croatia
--------------------------------+--------------------------------
Lost in Static 18 / And the storm is closing in now

Re: Gnus v5.6.24 is released

[Dave Love]

Sorry, the part of the previous patch for gnus-group.el was clearly
bogus.
.

Re: Gnus v5.6.24 is released

[Mats Lofdahl]

Bruce Stephens <bruce@cenderis.demon.co.uk> writes:

> In gnus-topic.el, there's a [M-tab].  Should be [(meta tab)] or
> something for XEmacs.

M-tab (ALT-TAB, actually) lowers my emacs window and puts focus on
another window.  If that is the usual response from Motif/X-windows,
M-tab should be avoided in emacs.

If it is not, I'd like some advice on how to make my setup behave
better.  (I'm running the Common Desctop Environment.)

-- 
 Mats Löfdahl

 "Good luck.  If you turn out to be a genius after all, remember that I
 never once for a moment believed in you." -- Orson Scott Card (1996)

Re: Gnus v5.6.24 is released

[Vladimir Volovich]

"Mats" == Mats Lofdahl writes:

 Mats> M-tab (ALT-TAB, actually) lowers my emacs window and puts focus
 Mats> on another window.  If that is the usual response from
 Mats> Motif/X-windows, M-tab should be avoided in emacs.

 Mats> If it is not, I'd like some advice on how to make my setup
 Mats> behave better.  (I'm running the Common Desctop Environment.)

Under X, it is not necessary to map the ALT key to Meta. If your
keyboard has some extra keys (like e.g. M$ window keys), you can map
those keys to Meta. So, ALT and Meta will be different keys (and Emacs
understands both of them as different keys). This is very convenient.

	Best regards, -- Vladimir.

Re: Gnus v5.6.24 is released

[Aaron M. Ucko]

Stainless Steel Rat <ratinox@peorth.gweep.net> writes:

> Which is fine if you happen to have a relatively new Intel or Alpha box.
> But this ignores Suns, HPs, IBMs (RS/6000 and PC/RT), and numerous others,
> of which there are many more in the world than Linux or *BSD.

Um, pretty much all the Sun keyboards I've seen have been Type 4 or 5
keyboards hence had both meta (diamond) and alt keys.  OTOH, IBM and
SGI keyboards all seem to use the PC101 layout, and HP and "classic
Digital" keyboards have their own layouts with only one shifting
modifier besides shift and control, so perhaps you have a point.  On
the gripping hand, CDE is ugly and to be avoided anyway. ;-)

-- 
Aaron M. Ucko, KB1CJC <amu@mit.edu> (finger amu@monk.mit.edu)

Re: Gnus v5.6.24 is released

[Aaron M. Ucko]

Stainless Steel Rat <ratinox@peorth.gweep.net> writes:

> Aaron M Ucko writes:
> 
> > Um, pretty much all the Sun keyboards I've seen have been Type 4 or 5
> > keyboards hence had both meta (diamond) and alt keys.
> 
> Take a good look at a Sun 3 type keyboard sometime.

Easier said than done these days.

> > OTOH, IBM and SGI keyboards all seem to use the PC101 layout,
> 
> None of the dozen or so SGIs or RS/6000s I maintain have "Windows" keys.

Hence PC101 rather than PC104.

> CDE is not a window manager.  Motif (mwm) is a window manager that uses
> ALT.  So do many FVWM configurations.  So do many of the variety of other
> "twm" window managers.  Enlightenment does it; so does AfterStep.

Yes, but the original question was whether M-tab is safe, which is a
function of WM configuration.  CDE-like configurations typically bind
M-tab; many others don't.  At any rate, no WM I've seen messes with
keycode -> keysym translations, which are the X server's
responsibility; I think you're just confused because many WMs pretend
Mod1 is always Meta, even though it's sometimes Alt when they are
distinct.

-- 
Aaron M. Ucko, KB1CJC <amu@mit.edu> (finger amu@monk.mit.edu)

Re: Gnus v5.6.24 is released

[Alan Shutko]

>>>>> "R" == Stainless Steel Rat <ratinox@peorth.gweep.net> writes:

R> So yes, on a 101-key keyboard, Alt and Meta are the same thing.

No.  On a 101 key keyboard, there is no keycap with Meta on it.
However, one can tell X that the Meta key is anything you like.  I
could tell X that my Meta key is the key labeled on my keyboard
"Scroll Lock".

You can similarly tell X that Alt and Meta are separate keys, by
giving the keysyms to two different keycodes and assigning different
mods to them.  I'm not sure if dtwm and mwm bind Alt or Meta... I'll
have to check at work.

(Another nit: CDE and Motif are two different UI standards, although
I think CDE is a superset.)

In any case, M-tab may not be safe, but Emacs has enough other things
bound to M-tab in various modes that it should be perfectly fine
here.  An Emacs user should either use ESC-Tab or convince their wm
not to snag the M-tab sequence, by turning off that mapping (my
preferred route) or by separating Alt and Meta.

-- 
Alan Shutko <ats@acm.org> - By consent of the corrupted
My other computer is a Connection Machine.

Re: Gnus v5.6.24 is released

[Alan Shutko]

>>>>> "R" == Stainless Steel Rat <ratinox@peorth.gweep.net> writes:



R> So yes, on a 101-key keyboard, Alt and Meta are the same thing.
>> No.  On a 101 key keyboard, there is no keycap with Meta on it.

R> Yes, there is... it is called "Alt" rather than "Meta" but it
R> serves an identical purpose.

So Alt is really Meta?  It doesn't serve the same purpose.  It
generates Alt.  It doesn't generate Meta.  X may be set up to treat
the two as the same, but that's not something the keyboard has any
control of.  There are two levels here.  At the keyboard level,
there's no Meta.

>> However, one can tell X that the Meta key is anything you like.  I
>> could tell X that my Meta key is the key labeled on my keyboard
>> "Scroll Lock".

R> Of course, you would lose that key's normal function, in this case
R> XON/XOFF toggle. 

What nromal function?  It doesn't toggle XON and XOFF for me.  Not
surprisingly, because XON and XOFF don't mean anything for X.  It
toggles a nice light on my keyboard and does nothing, even in xterm.

Besides, the point was that as above, there are two different layers,
and in neither layer are Alt and Meta synonymous.  They aren't to the
keyboard, which doesn't _have_ Meta, and they aren't to X.  I just
checked and I can put Meta on keycode 57 (the right key labeled Alt on
my keyboard), remove Meta_R from mod1 and add it to mod2, and suddenly
everything treats the two differently.  Namely, Alt-Tab will do the
fun window switch, and Meta-Tab is passed to Emacs.

Hell, if I wanted to I could go adsd Super and Hyper into the mix and
map Hyper-space.

>> (Another nit: CDE and Motif are two different UI standards,
>> although I think CDE is a superset.)

R> Inasmuch as CDE uses Motif as its default window manager, this is
R> true.

Wrong.  CDE uses dtwm.  

NAME

  dtwm - The CDE Window Manager

SYNOPSIS

  [options]

DESCRIPTION

  The dtwm window manager is an X Window System window manager based
  upon the OSF/Motif window manager, mwm (version 1.2.4).  It provides
  mwm compatible window management functionality.  This includes
  functions that facilitate control (by the user and the programmer)
  of elements of window state such as placement, size, icon/normal
  display, and input-focus ownership.


-- 
Alan Shutko <ats@acm.org> - By consent of the corrupted
Rainy days and automatic weapons always get me down.

Re: Gnus v5.6.24 is released

[Alan Shutko]

>>>>> "R" == Stainless Steel Rat <ratinox@peorth.gweep.net> writes:

R> Alt does not "generate" anything; neither does Meta.  They are
R> shifting keys, similar to Ctrl.  The only difference at that level
R> is that Symbolics called it "Meta" while IBM called it "Alt".

Sure Alt generates something.  It generates a keycode.  How else do
you think software can tell when you tap the alt key (to display
menus, perhaps)?

You'd might as well say that the only difference between Hyper and
Ctrl is that they have different names.  

>> What nromal function?  It doesn't toggle XON and XOFF for me.  Not
>> surprisingly, because XON and XOFF don't mean anything for X.

R> X is not the be-all and end-all.  Scrol Lock on a tty equates to
R> toggling XON and XOFF.  That is its originally intended purpose in
R> life.

Right, but we were talking about X.  If you'd like to discuss the
equivelence of Meta and Alt on terminals, say so.  But there's no
equivalence under X, and there's no documentation that I know of where
IBM engineers said "We need another modifier key, but instead of
calling it Meta, we'll call it Alt and really _mean_ Meta."

Sure, it is a commonly made equivalence, but there's nothing that
means it has to be so.

-- 
Alan Shutko <ats@acm.org> - By consent of the corrupted
You humans are all alike.

Re: Gnus v5.6.24 is released

[Robert Bihlmeyer]

Hi,

>>>>> On 15 Jul 1998 21:42:12 -0400
>>>>> Stainless Steel Rat <ratinox@peorth.gweep.net> said:

 Rat> Alan Shutko writes:

 >> Sure Alt generates something. It generates a keycode.

 Rat> It does not, in and of itself, generate anything that Emacs,
 Rat> thus Gnus, recognizes.

That's only the default setup. Of course Alt generates a keycode, that 
gets from the keyboard to the cpu to the os to X. If you tell X to map 
this keycode to something other than a modifier, it produces this
keysym with no problem. I just used xmodmap to make Alt generate an "r".

 >> Right, but we were talking about X.

 Rat> No, we were discussing a key binding that needs to work in X and
 Rat> in ttys, and whether or not that binding is "safe". And the
 Rat> point I think I (and others) have made is that it is not "safe".

We're discussing whether to support stupid wms that steal bindings
from programs. Remember that ICCCM requests wms to offer users a way
to send ANY key-combo to applications.

	Robbe

-- 
Robert Bihlmeyer	reads: Deutsch, English, MIME, Latin-1, NO SPAM!
<robbe@orcus.priv.at>	<http://stud2.tuwien.ac.at/~e9426626/sig.html>

Re: Gnus v5.6.24 is released

[Mats Lofdahl]

Vladimir Volovich <vvv@vvv.vsu.ru> writes:

> Under X, it is not necessary to map the ALT key to Meta. If your
> keyboard has some extra keys (like e.g. M$ window keys), you can map
> those keys to Meta. So, ALT and Meta will be different keys (and
> Emacs understands both of them as different keys). This is very
> convenient.

Hmmm...  I have the Alt key and an Alt-Gr key.  I use the Alt key for
Meta and the Alt-Gr key is needed to access some extra characters on
my keyboard.  I don't have the extra M$ window keys because I use an
older Swedish keyboard, but I would be interested in how to map those
keys anyway.  Can it be done in .emacs or is it an X thing?

I know I can (as Hrvoje points out) use ESC.  In fact, that is what I
do use when I need to do M-tab.  It works but it is a bit
inconvenient.

-- 
 Mats Löfdahl

 "Good luck.  If you turn out to be a genius after all, remember that I
 never once for a moment believed in you." -- Orson Scott Card (1996)

Re: Gnus v5.6.24 is released

[Aaron M. Ucko]

Mats Lofdahl <mats@usa.net> writes:

> Hmmm...  I have the Alt key and an Alt-Gr key.  I use the Alt key for
> Meta and the Alt-Gr key is needed to access some extra characters on
> my keyboard.  I don't have the extra M$ window keys because I use an
> older Swedish keyboard, but I would be interested in how to map those
> keys anyway.  Can it be done in .emacs or is it an X thing?

The latter.  man xmodmap.

-- 
Aaron M. Ucko, KB1CJC <amu@mit.edu> (finger amu@monk.mit.edu)

Re: Gnus v5.6.24 is released

[SL Baur]

Mats Lofdahl <mats@usa.net> writes:
 ...
> Hmmm...  I have the Alt key and an Alt-Gr key.  I use the Alt key for
> Meta and the Alt-Gr key is needed to access some extra characters on
> my keyboard.  I don't have the extra M$ window keys because I use an
> older Swedish keyboard, but I would be interested in how to map those
> keys anyway.  Can it be done in .emacs or is it an X thing?

It's an X thing.  I map one Alt key to Meta, the other Alt to Alt,
two of the three ugly picture keys to Super and Hyper, and the third
ugly picture key to Kanji.  It took some fiddling (I am not
particularly X literate) to get it right.  I do it with the following
in .Xmodmap:

keycode  64 = Meta_L
keycode 113 = Alt_R
keycode 115 = Super_L
add mod2 = Super_L
keycode 116 = Hyper_R
add mod3 = Hyper_R
keycode 117 = Kanji

clear mod1
add mod1 = Meta_L
add mod4 = Alt_R

Re: Gnus v5.6.24 is released

[Yair Friedman]

SL Baur <steve@xemacs.org> writes:

> It's an X thing.  I map one Alt key to Meta, the other Alt to Alt,
> two of the three ugly picture keys to Super and Hyper, and the third
> ugly picture key to Kanji.  It took some fiddling (I am not
> particularly X literate) to get it right.  I do it with the following
> in .Xmodmap:
> 


Note that keycode and keysyms values depend on the X server.  If you
want easly manipulate your keyboard there is a utility xkeycaps by jwz.
--
Yair.

Re: Gnus v5.6.24 is released

[Hrvoje Niksic]

Mats Lofdahl <mats@astro.su.se> writes:

> Bruce Stephens <bruce@cenderis.demon.co.uk> writes:
> 
> > In gnus-topic.el, there's a [M-tab].  Should be [(meta tab)] or
> > something for XEmacs.
> 
> M-tab (ALT-TAB, actually) lowers my emacs window and puts focus on
> another window.  If that is the usual response from Motif/X-windows,
> M-tab should be avoided in emacs.

You can press `ESC TAB' if your WM steals Alt+TAB from you.  I would
rather rebind the WM keys, but that's me.

-- 
Hrvoje Niksic <hniksic@srce.hr> | Student at FER Zagreb, Croatia
--------------------------------+--------------------------------
"A Real Programmer's code can awe with its fiendish brilliance, even
as its crockishness appalls."

Re: Gnus v5.6.24 is released

[Kai Grossjohann]

>>>>> Mats Lofdahl <mats@astro.su.se> writes:

  > M-tab (ALT-TAB, actually) lowers my emacs window and puts focus on
  > another window.  If that is the usual response from Motif/X-windows,
  > M-tab should be avoided in emacs.

So use ESC TAB in Emacs.

kai
-- 
You ate somebody? -- Just a leg. -- That's terrible! -- Not with mustard.
(Terry Pratchett: Interesting Times)

Re: Gnus v5.6.24 is released

[François Pinard]

[-- Warning: decoded text below may be mangled, UTF-8 assumed --]
[-- Attachment #1: Type: text/plain, Size: 669 bytes --]

Kai Grossjohann <grossjohann@charly.cs.uni-dortmund.de> writes:

> >>>>> Mats Lofdahl <mats@astro.su.se> writes:
> 
>   > M-tab (ALT-TAB, actually) lowers my emacs window and puts focus on
>   > another window.  If that is the usual response from Motif/X-windows,
>   > M-tab should be avoided in emacs.
> 
> So use ESC TAB in Emacs.

FWIW, Alt-TAB never gave any kind of trouble to me under X, using `fvwm2'
as a manager, and plain Emacs 20.2.97.  I use Linux SuSE 5.0, in case it
matters (yet I do not think so).

-- 
François Pinard                            mailto:pinard@iro.umontreal.ca
Join the free Translation Project!    http://www.iro.umontreal.ca/~pinard

Re: Gnus v5.6.24 is released

[Kai Grossjohann]

>>>>> François Pinard <pinard@iro.umontreal.ca> writes:

  > FWIW, Alt-TAB never gave any kind of trouble to me under X, using
  > `fvwm2' as a manager, and plain Emacs 20.2.97.  I use Linux SuSE
  > 5.0, in case it matters (yet I do not think so).

What Alt-tab does, depends on the configuration of fvwm.  I'm not sure
about the *default* configuration of fvwm, whether that has this
problem.  I know that the default configuration of KDE has the problem
of using Alt-tab for its own purposes.

kai
-- 
You ate somebody? -- Just a leg. -- That's terrible! -- Not with mustard.
(Terry Pratchett: Interesting Times)

Re: Gnus v5.6.24 is released

[Jochen_Hayek]

>>>>> "LMI" == Lars Magne Ingebrigtsen <larsi@ifi.uio.no> writes:

    LMI> The 498th release -- bug fixes.
    LMI> Get it from <URL:http://www.gnus.org/gnus.tar.gz> or 
    LMI> "/ftp@ftp.gnus.org:/pub/emacs/gnus/".  The patch is available as
    LMI> <URL:http://www.gnus.org/patches/gnus-5.6.23-5.6.24.diff.gz>.

    LMI> ChangeLog since last release:

    LMI> Sat Jul 11 03:03:53 1998  Lars Magne Ingebrigtsen  <larsi@menja.ifi.uio.no>

    LMI> 	* gnus.el: Gnus v5.6.24 is released.

    LMI> [...]

I tried to build it:

    johayek@Hayek5 $ ./configure --cache-file=/usr/local/config.cache
    loading cache /usr/local/config.cache
    checking whether make sets ${MAKE}... (cached) yes
    checking for a BSD compatible install... (cached) /usr/bin/install -c
    checking for emacs... (cached) /usr/local/bin/emacs
    checking where .elc files should go... $(prefix)/share/emacs/site-lisp

that's where the problem started, I assume -> `$(prefix)'

    checking for makeinfo... (cached) /usr/bin/makeinfo
    creating ./config.status
    creating Makefile
    creating lisp/Makefile
    creating texi/Makefile
    johayek@Hayek5 $ make
    cd lisp && make EMACS=/usr/local/bin/emacs lispdir=/share/emacs/site-lisp all
    [...]

the lispdir is not quite appropriate,
and that's how the Makefile looks:

    johayek@Hayek5 $ head Makefile
    # Generated automatically from Makefile.in by configure.
    lispdir = $(prefix)/share/emacs/site-lisp
    srcdir = .

I personally made it then to look like, but ...:

    johayek@Hayek5 $ head Makefile
    # Generated automatically from Makefile.in by configure.
    prefix  = /usr/local
    lispdir = $(prefix)/share/emacs/site-lisp
    srcdir = .
.

Re: Gnus v5.6.24 is released

[Jochen_Hayek]

As always, when I install a new *gnus,
I apply a single personal patch to message.el .

Yes, I want to let sendmail use the `From:',
that I insert into the mail header, as its proper `From:'.

Yes, if life were easier and people and their tools smarter,
I would only use a proper `Reply-to:',
but you know the way they are.

And message-user-mail-address is quite a resonable function to use
instead of user-login-name, isn't it.

And at least in my case sendmail never ever produces a valid
e-mail address (to deliver e-mail to) from user-login-name,
because either I'm behind a firewall,
or my ISP assigns me a dynamical IP address and a deliberate
temporary hostname, or ...

What do you think about this suggestion?

(Courtesy copies to me appreciated.)

================================================================================

*** message-5.6.24.el	1998/07/23 13:26:26	1.1
--- message-5.6.24.el	1998/07/23 13:30:48	1.2
***************
*** 2026,2032 ****
  		     ;; But some systems are more broken with -f, so
  		     ;; we'll let users override this.
  		     (if (null message-sendmail-f-is-evil)
! 			 (list "-f" (user-login-name)))
  		     ;; These mean "report errors by mail"
  		     ;; and "deliver in background".
  		     (if (null message-interactive) '("-oem" "-odb"))
--- 2026,2032 ----
  		     ;; But some systems are more broken with -f, so
  		     ;; we'll let users override this.
  		     (if (null message-sendmail-f-is-evil)
! 			 (list "-f" (message-user-mail-address)))
  		     ;; These mean "report errors by mail"
  		     ;; and "deliver in background".
  		     (if (null message-interactive) '("-oem" "-odb"))
.

Re: Gnus v5.6.24 is released

[Kai Grossjohann]

>>>>> Jochen_Hayek@ACM.org writes:

  > And at least in my case sendmail never ever produces a valid
  > e-mail address (to deliver e-mail to) from user-login-name,
  > because either I'm behind a firewall,
  > or my ISP assigns me a dynamical IP address and a deliberate
  > temporary hostname, or ...

What happens when you set message-sendmail-f-is-evil such that no `-f'
option is generated at all?

kai
-- 
Abort this operation?   [Abort]  [Cancel]

Re: Gnus v5.6.24 is released

[Jochen_Hayek]

>>>>> "KG" == Kai Grossjohann <grossjoh@amaunet.informatik.uni-dortmund.de> writes:

>>>>> Jochen_Hayek@ACM.org writes:
    >> And at least in my case sendmail never ever produces a valid
    >> e-mail address (to deliver e-mail to) from user-login-name,
    >> because either I'm behind a firewall,
    >> or my ISP assigns me a dynamical IP address and a deliberate
    >> temporary hostname, or ...

    KG> What happens when you set message-sendmail-f-is-evil such that no `-f'
    KG> option is generated at all?

Good point, mate!

My patch is not well integrated at all, you are perfectly right.

I just happened to find a place, where I could plug in
the mail address I want to use.

But this message.el seems to be a quite worrying and historical
piece of software to me,
and I wouldn't know where a better place would be.

----------

Off-topic to KG: isn't it strange, we `huns' have to talk in english,
if we actually could use our funny native language.
.