Pipe symbol not being recognised

[GoogleCodeExporter]

What steps will reproduce the problem?
1. ssh to host
2. use a command that pipes into another, e.g. ls | wc -l

What is the expected output? 
Results of executing first command piped into the second command, i.e:
$ ls | wc -l
    216

What do you see instead?
$ ls | wc -l
ls: |: No such file or directory
ls: wc: No such file or directory

What version of the product are you using (you can see this by using Menu
-> About in the Host List)?

What type of system are you trying to connect to?
1.5.5 r409 2009.09.06

If you are able to connect, what is the output of "echo $TERM", "uname -a",
and any other relevant information on the host?
TERM=screen
"uname -a" = Linux <hostname> 2.6.24.5-85.fc8 #1 SMP Sat Apr 19 12:39:34
EDT 2008 i686 i686 i386 GNU/Linux

Please provide any additional information below.
It's as if it is interpreting "|" as a string

Original issue reported on code.google.com by hit...@gmail.com on 12 Nov 2009 at 1:57

• Merged into: #166

[GoogleCodeExporter]

It might be helpful if you do the following:

echo | > /tmp/pipe
xxd /tmp/pipe

so we can see what character your system thinks that is, if not a pipe. (I 
could be 
off base, this is just a conjecture.) Of course, this will fail if you are 
having the 
same problem with '>', but give it a shot.

Also, what shell are you running?

Original comment by gwillen@gmail.com on 12 Nov 2009 at 11:12

[GoogleCodeExporter]

Also, which phone are you using?

Original comment by gwillen@gmail.com on 12 Nov 2009 at 11:14

[GoogleCodeExporter]

I knew I forgot some crucial information! The phone is an HTC Hero (UK Orange 
unmodified)

$ echo | >/tmm/pipe
$ xxd /tmp/pipe
0000000: c2a6 0a                                         ...
$ echo $SHELL
/bin/bash

Original comment by hit...@gmail.com on 13 Nov 2009 at 8:09

[GoogleCodeExporter]

That's the UTF-8 code for pipe. Try: echo $LANG

I have a feeling you're using a non-UTF-8 locale but a UTF-8 setting on 
ConnectBot.

Also screen requires a -U upon startup to recognize UTF-8 or "defutf8 on" in 
your
.screenrc

Original comment by kenny@the-b.org on 13 Nov 2009 at 5:56

[GoogleCodeExporter]

$ echo $LANG
en_GB.UTF-8

Looks like my locale is UTF8 I guess?

I couldn't find where to change the language settings in ConnectBot, but I tried
different terminal emulations such as xterm and vt100 with the same results.

I also tried setting defutf8 in .screenrc but still no change.

Original comment by hit...@gmail.com on 14 Nov 2009 at 8:45

[GoogleCodeExporter]

The HTC keyboard *should* be outputting | U+7C VERTICAL LINE, but it's 
outputting ¦
U+A6 BROKEN BAR. Strange mistake.

Original comment by kenny@the-b.org on 14 Nov 2009 at 3:00

[GoogleCodeExporter]

This is also occuring on an HTC Eris running Android 1.5.  ConnectBot 
v1.5.5(r409
2009.09.06)

$ echo | >/tmp/pipe
$ xxd /tmp/pipe
0000000: c2a6 0a                                         ...
$ echo $SHELL
/bin/bash
$ echo $LANG
en_US.UTF-8
$ echo $TERM
xterm
$ uname -a
Linux host-name 2.6.28-16-generic #55-Ubuntu SMP Tue Oct 20 19:48:24 UTC 2009 
i686
GNU/Linux

Original comment by scott.st...@gmail.com on 25 Nov 2009 at 3:58

[GoogleCodeExporter]

This is basically Issue 166 again. Thanks, HTC!

Original comment by kenny@the-b.org on 28 Nov 2009 at 4:32

• Changed state: Duplicate