[ksh93-integration-discuss] Problems with stty erase <backspace>
Bernd Finger - Sun Germany
Bernd.Finger at Sun.COM
Thu Sep 14 23:01:42 PDT 2006
Glenn,
here'e a more detailed description of what I originally posted.
Problem:
From host1, I am doing an "ssh host2". Then, I am typing "test<backspace_key>",
as you can see in the first attachment. After that, I am typing
"test<bs><bs><bs><bs>" so that the funny symbols can be seen better. As the 3rd
step, I am typing "stty erase <bs>" (which I, by the way, I wish I would never
have to do again in the final version of ksh93. But maybe that's a topic for
another thread on this list. Seems to be a much more complex issue than I
originally thought. For example, even with bash, after starting vi, the
incorrectly mapped backspace key occurred although it worked fine in the shell).
Settings:
--->
host1:
$ head -1 /etc/release
Solaris Nevada snv_45 X86
Version M 1993-12-28 r+
(http://www.opensolaris.org/os/project/ksh93-integration/downloads/ksh93_integration_20060827_snapshot_i386.tar.bz2)
$ set -o | /bin/fgrep on
Current option settings
bgnice on
braceexpand on
clobber on
exec on
glob on
gmacs on
interactive on
log on
monitor on
multiline on
rc on
unset on
viraw on
$ cat /etc/ksh.kshrc
# enable gmacs editor mode
set -o gmacs
# enable multiline input mode
set -o multiline
$ echo ${LANG}
en_US.ISO8859-1
$ echo ${TERM}
xterm
host2:
$ head -1 /etc/release
Solaris Nevada snv_46 X86
Version M 1993-12-28 r+
(http://www.opensolaris.org/os/project/ksh93-integration/downloads/ksh93_integration_20060827_snapshot_i386.tar.bz2)
$ set -o | /bin/fgrep on
Current option settings
bgnice on
braceexpand on
clobber on
exec on
glob on
interactive on
log on
monitor on
multiline on
rc on
unset on
vi on
viraw on
$ echo ${LANG}
en_US.ISO8859-1
$ echo ${TERM}
xterm
$ cat /etc/ksh.kshrc
# enable gmacs editor mode
set -o gmacs
# enable multiline input mode
set -o multiline
<---
Regards,
Bernd
Glenn Fowler wrote:
> On Thu, 14 Sep 2006 14:57:40 -0400 James Carlson wrote:
>> Dan Price writes:
>>>> we're still trying to reproduce this
>>>> did you get it to fail without attaching dtrace?
>>> I would hope so: dtrace is instrumenting the kernel in Roland's example
>>> (in other words we're seeing all calls on the system to pgsignal());
>>> it's just a passive observer here.
>
>> It failed in a different way with dtrace running, but only because ksh
>> seemed quite confused about having a job started with '&' and ended up
>> waiting for it to exit. Something's definitely wrong with job
>> control.
>
> ampersand you say
>
> I was just running
> /usr/bin/stty intr '^?'
> I didn't notice & in the original post
>
> so it may be >1 problems
>
> can someone post the exact command sequences and misbehaving output
> (or <HANGS HERE> etc.) for the problem(s)
>
> thanks
>
> -- Glenn Fowler -- AT&T Research, Florham Park NJ --
>
> _______________________________________________
> ksh93-integration-discuss mailing list
> ksh93-integration-discuss at opensolaris.org
> http://mail.opensolaris.org/mailman/listinfo/ksh93-integration-discuss
>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: stty-erase-backspace.jpg
Type: image/jpeg
Size: 161000 bytes
Desc: not available
Url : http://oss-beta1.opensolaris.org/pipermail/ksh93-integration-discuss/attachments/20060915/a54a6f4f/attachment.jpg
More information about the ksh93-integration-discuss
mailing list