Re: [Erc-discuss] next ERC buffer

[Emanuel Berg]

"Aaron S. Hawley" <address@hidden> writes:

> 1) I like using `C-x <left>' and `C-x <right>' to
> quickly switch buffers but I believe I rebind them
> to `bury-buffer' and `unbury-buffer' since the
> defaults `next-buffer' and `previous-buffer' have
> some subtle behavior I don't like.

To each his own!

However...

Those shortcuts are not efficient as they will
require me to move my right hand to the arrow keys and
thus away from the "jkl;" position, where typing is
done and were the hand should be, constantly. It is
easy tho to have those shortcuts `C-x j' and `C-x
l' instead.

In Emacs, iterating the buffers in general isn't
a good idea of switching since there are so many
buffers for even the shortest session - for example,
this is what I have, and I started Emacs but a few
minutes ago!

      * *unsent followup...    1362 Message
     %* *Summary nntp+ne...      66 Summary          
     %* *Article*              1047 Article          
     %* *Group*                 207 Group            
        KILL                   5299 Kill
     %  *w3m*                   376 w3m              
        *scratch*                 0 Emacs-Lisp       
     %* *Messages*             7479 Messages         
      * .newsrc-dribble         601 Fundamental

Iteration is OK if it is confined to a small subset of
buffers, e.g. IRC #channels - at least in my case, as,
again, I typically have 2-4 open. If I were a more
passionate IRC user iteration would have to be
replaced by some other system, which would be faster
and more precise, perhaps a new "IRC" prefix keystroke
and then mnemonic shortcuts for each commonly used
channel, e.g.

    'e' - #emacs
    'z' - #zsh
    'd' - #debian

> 2) I run ERC in a separate Emacs instance to keep it
> separate from coding (and because of occasional ERC
> crashing problems). So in the Emacs frame running
> ERC, I only have to wade through ERC buffers and
> nothing else.

That sure is another way of getting the smaller set of
buffers to iterate but that brings along other
problems due to the lack of integration. You have to
switch between instances, and how do you kill/yank
between them? And how do you keep track what buffer is
where? Say you want to quote a man page in
Usenet/Gmane with Gnus, and ditto on IRC with ERC.
Where do you open the man page? If you open it in the
"non-communication" Emacs, you have to switch instance
and solve inter-instance kill/yank somehow. If you
open it in the communication Emacs, you will have to
switch instance, and then iterate the buffer when
iteration of #channels is desired. Later, when in some
other situation (which is likely to occur as you are
doing something that relates to the man page) you will
need the man page again, and then you will think
"good, that buffer is already open" - but where is it?
This situation will become out of hand very soon.

Several Emacs instances is almost never a good idea.

By the way - crashes?! That shouldn't happen. I can't
remember Emacs crashing on me ever in 5+ years of
every day (or night) usage. That problem should be
fixed rather than you relying on a second
Emacs instance.

-- 
underground experts united
http://user.it.uu.se/~embe8573