Re: some comments on the gutter
Andy Piper <andy(a)xemacs.org> writes in xemacs-beta(a)xemacs.org:
At 05:05 PM 8/12/99 +0900, SL Baur wrote:
...
> Restricting them to be buffers is pretty much useless in Gnus.
I
> don't know which rules you currently use to select which buffers get
> buttons, but they are almost always the wrong ones.
I don't understand this, please explain. Lead me :)
One example:
I am now in a message-mode buffer called `*wide reply to Andy Piper on
xemacs-beta(a)xemacs.org*'. There are currently two gutter buttons.
One labelled `*sent wide reply to Stephen J. Turnbull*<3>' and `*sent
mail*'. sent mail buffers should never have buttons, especially with
the tremendously long names most of them have. Good candidates for
buttons at this time would be the *Group* buffer, the
*Summary... buffer, the *Article... buffer and the SPC*Original
Article... buffer.
Continuing this example, I just clicked on the *sent mail* button. It
was replaced by *sent wide reply to Didier Verna*. When I C-x b'ed
back to this buffer (since it did not have a button) the *sent mail*
button disappeared.
Continuing this example, I just clicked on the remaining *sent wide
reply to Stephen J. Turnbull*<3> button. When I landed in that
buffer, the *sent wide reply to Didier Verna* button came back. When
I C-x b'ed back to this buffer (using C-x b because I was not offered
a button for it) the *sent wide reply to Didier Verna* button remained
and the other button became *sent wide reply to Stephen
J. Turnbull*<2>. ????
Another example:
Sometimes a button is generated for the *Article... buffer and
sometimes it isn't. I usually get one when I have just switched mail
groups and have started to read a new one. I have the Gnus option set
where an article is not automatically selected when I enter a group
and so in this case the *Article... buffer refers to a message from
the group I just stopped reading.
In this session, not once has there been a button in the gutter of a
buffer I have wanted to select. It's visually distracting on redisplay
and the apparently chaotic method in which buffers are chosen (if there
is a deterministic way in which they are chosen I have failed to figure
it out) is also distracting.
Andy, buttons appear and disappear with no apparent underlying reason.
I am competely confused and disoriented.