Re: *shell* buffer in GNU Emacs 23.0.0 has faulty terminal control

[Kenichi Handa]

In article <address@hidden>, Peter Dyballa <address@hidden> writes:

> Am 27.09.2006 um 02:52 schrieb Kenichi Handa:

> > In article <address@hidden>, Peter  
> > Dyballa <address@hidden> writes:
> >
>>> Launching GNU Emacs 23.0.0 with -Q I create a *shell* buffer (tcsh is
>>> my login shell) and run git to clone some repository (X11R7.1). While
>>> git runs it produces progress output. Different from the behaviour of
>>> GNU Emacs 22.0.50's *shell* buffer there is /not/ one spot
>>> overwritten with the new text, but the new text is added on a new
>>> spot. In the end the 23.0.0 *shell* buffer has a contents like:
> >
> > Please show me the output of C-h C RET in your *shell*
> > buffer.

[...]
> Defaults for subprocess I/O:
>    decoding: U -- utf-8-unix (alias: mule-utf-8-unix)

>    encoding: U -- utf-8-unix (alias: mule-utf-8-unix)

With the same coding system, I can't reproduce the bug at
least by using "wget".  It prints the progress report as
something like below:

41% [==============>                      ] 1,590,232    308.01K/s    ETA 00:0

And, in *shell* buffer, the same line is updated.  But, your
test case of "git" may be different.

Peter Dyballa <address@hidden> writes:
>       pete 115 /\  ./git_xorg.sh
>       creating font
>       cloning font/adobe-100dpi
[...]

Please show me the contents of "git_xorg.sh".  As I've never
used git, I have no idea what you did in that shell script.

---
Kenichi Handa
address@hidden