mm-inline-media-tests change causes more changes than expected

mm-inline-media-tests change causes more changes than expected

[Dale Hagglund]

I'm running gnu emacs 20.3 with pgnus 0.98.  I added the following
lines to my .gnus file:

      (setq mm-inline-media-tests 
        (cons '("text/html" nil (lambda (h) nil)) mm-inline-media-tests))

The goal here was to prevent gnus from attempting to inline html parts
using w3.  This was failing trying to find x-mouse.el, which probably
means that I have an old version of w3.

I could, of course, have upgraded to a newer version of w3, but my
actual goal was to view the occasional bit of html-only mail with
lynx, and avoid w3 entirely.  Somewhat heretical, I realize, but there
you have it.

After modifying mm-inline-media-tests as shown above, text/html
messages show up with a button, and the raw html shown inline.
Hitting return on the button twice runs lynx, as specified in my
.mailcap file.

However, now, whenever I get multipart/mixed messages, any text/plain
parts are presented with buttons as well.  Messages bodies marked as
text/plain in the headers, either explicitly or implicitly, display
normally.

So, my questions are:

1) Is changing mm-inline-media-tests as above the correct way to
   achieve my goal?

2) Why does the modification affect the display of text/plain parts?

3) What do I need to do to prevent gnus from using w3 to inline
   text/html parts, without disturbing the display of other text/*
   parts.

4) With the above change, text/html parts are displayed with buttons
   in the article buffer, and the raw html displayed inline.  This
   means that I have to click (or the equivalent) on the button twice
   to start lynx.  How can I tell gnus to display text/html parts as
   initially hidden.

Dale.

Re: mm-inline-media-tests change causes more changes than expected

[Lars Magne Ingebrigtsen]

Dale Hagglund <rdh@best.com> writes:

> However, now, whenever I get multipart/mixed messages, any text/plain
> parts are presented with buttons as well.

Do the text/plain parts have disposition "attachment"?

> 1) Is changing mm-inline-media-tests as above the correct way to
>    achieve my goal?

Yes.

> 2) Why does the modification affect the display of text/plain parts?

It shouldn't.

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

Re: mm-inline-media-tests change causes more changes than expected

[Dale Hagglund]

[-- Attachment #1: Type: text/plain, Size: 397 bytes --]

Lars Magne Ingebrigtsen <larsi@gnus.org> writes:

> Dale Hagglund <rdh@best.com> writes:

> > However, now, whenever I get multipart/mixed messages, any text/plain
> > parts are presented with buttons as well.

> Do the text/plain parts have disposition "attachment"?

Nope.  I've attached a short message that (with my current settings
for mm-inline-media-tests) illustrates the problem.

Dale.


[-- Attachment #2: simple multipart message --]
[-- Type: message/rfc822, Size: 2864 bytes --]

[-- Attachment #2.1.1: Type: text/plain, Size: 346 bytes --]

A New Internet-Draft is available from the on-line Internet-Drafts directories.
This draft is a work item of the IETF Steering Group Working Group of the IETF.

	Title		: Use of HTTP State Management
	Author(s)	: K. Moore, N. Freed
	Filename	: draft-iesg-http-cookies-01.txt
	Pages		: 10
	Date		: 02-Dec-99
	
[ a bunch of stuff deleted... --rdh]

[-- Attachment #2.1.2.1: Type: Message/External-body, Size: 135 bytes --]

[-- Attachment #2.1.2.2: draft-iesg-http-cookies-01.txt --]
[-- Type: Message/External-body, Size: 67 bytes --]

Re: mm-inline-media-tests change causes more changes than expected

[Dale Hagglund]

Well, I figured it out.  Somewhere in the recent past of using gnus
(which I rarely restart), I'd pressed ``M-t'' in the article buffer,
probably while intending to type ``t'' to toggle header display.  This
toggles the stuff that inhibits buttonizing.  ``M-t'' again makes all
well.

Sigh...  Sometimes, just sometimes mind you, I begin to think there
are a few too many legal key bindings.

Dale.

Re: mm-inline-media-tests change causes more changes than expected

[Lars Magne Ingebrigtsen]

Dale Hagglund <rdh@best.com> writes:

> Sigh...  Sometimes, just sometimes mind you, I begin to think there
> are a few too many legal key bindings.

There can never, ever be too many valid key bindings.

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