[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
bug#22785: 24.5; comint/shell modes should be merged with term mode
From: |
Ingo Lohmar |
Subject: |
bug#22785: 24.5; comint/shell modes should be merged with term mode |
Date: |
Tue, 23 Feb 2016 22:05:18 +0100 |
User-agent: |
Notmuch/0.20.2+113~g6332e6e (http://notmuchmail.org) Emacs/25.0.90.1 (x86_64-pc-linux-gnu) |
On Tue, Feb 23 2016 12:33 (-0800), Per Bothner wrote:
> The "line" submode of term mode duplicates much of the functionality
> of comint mode. When I wrote term mode, my hope was that it would
> subsume comint mode, or that comint mode would be built on top of
> term mode. That never happened - so far.
>
> 'shell mode' should basically be the same as 'term' mode, except
> that shell mode starts of in line-editing mode, while term mode
> starts up in character mode.
>
> There are obvious advantages to merging these modes. One is
> we have ansi escape sequence handling in a single place.
> Another is that you can run console applications from shell mode.
To me, that seems to be a bad idea. They are two different modes
precisely *because* they are quite different. If you want to run
console applications, why would you use shell-mode at all?
I dumped any use of term.el because I could not control its
complexities, no matter what overrides, patches and advices I used. I
have not looked back after using shell-mode where it's useful instead,
and not running console applications anymore.
Term is *significantly* more complex than shell (175k vs 54k), and does
a lot more things (for better or worse). Maybe factoring out the ANSI
sequence parsing is what you really want?