Re: Prompt length calculation fails when UTF-8 is used within \[ \]

Lasse Kärkkäinen wrote: >> ftp://ftp.cwru.edu/pub/bash/bash-4.0-patches > > So, what format are these and how do I patch (on Linux)? Apparently the > patch program doesn't recognize them automatically. The patch program recognizes them okay on my machine. Works for me. > P.S. is there some spe

Re: Prompt length calculation fails when UTF-8 is used within \[ \]

ftp://ftp.cwru.edu/pub/bash/bash-4.0-patches So, what format are these and how do I patch (on Linux)? Apparently the patch program doesn't recognize them automatically. P.S. is there some specific reason why you do things in so non-standard way? I would expect a project of this size use a ve

Re: 3-lines long segfault

Marc Herbert wrote: > - removing the "-t" option from "read" dodges the immediate crash, but >not the memory corruption? So bash might still crash later? > > - in other words, triggering a trap from a command expansion will >always corrupt memory? (I mean: without your recent fix) It i

time hangs or seg faults

On one system, I have installed bash from one version to the next, from 3.2.17 to 4.0.10. With each version, the built-in "time" command hangs. On the same system, I have installed tcsh, and that one's built in time function works. Previously, I thought it was a versioning issue, but, now I a

Re: Errors when patching bash4.0

On Wed, 22 Jul 2009, Alex Reed wrote: Hello, I've downloaded and decompressed the bash-4.0 tarball, and downloaded bash40-001 through bash40-028 into a subdirectory bash-4.0/patches, using: bash $ tar xzvf bash-4.0.tar.gz bash $ cd bash-4.0 bash-4.0 $ mkdir patches bash-4.0 $ # line breaks adde

Negative COMP_CWORD values?

Hello, Is it intentional that COMP_CWORD may have a negative value within a completion function? I did not find any indication of that in the bash man page. For example: foo() { echo "COMP_CWORD:$COMP_CWORD" } complete -F foo bar Then, type "bar " (sans quotes, t

Re: Errors when patching bash4.0

On Jul 22, 4:09 pm, Bob Proulx wrote: > Alex Reed wrote: > > At least one "hunk" fails on every patch file.  What am I doing wrong? > > Hmm...  Works for me.  Here is a trace of the important bits. > >   $ wgetftp://ftp.gnu.org/pub/gnu/bash/bash-4.0.tar.gz >   $ wgetftp://ftp.gnu.org/pub/gnu/bash/

Re: read -t 0 not supported? (poll)

On Wed, 22 Jul 2009, Marc Herbert wrote: It seems polling using "read -t 0" is not supported. It is supported in bash 4.0. -- Chris F.A. Johnson === Author: Shell

Re: Errors when patching bash4.0

Alex Reed wrote: > At least one "hunk" fails on every patch file. What am I doing wrong? Hmm... Works for me. Here is a trace of the important bits. $ wget ftp://ftp.gnu.org/pub/gnu/bash/bash-4.0.tar.gz $ wget ftp://ftp.gnu.org/pub/gnu/bash/bash-4.0.tar.gz.sig $ wget ftp://ftp.gnu.org/p

Errors when patching bash4.0

Hello, I've downloaded and decompressed the bash-4.0 tarball, and downloaded bash40-001 through bash40-028 into a subdirectory bash-4.0/patches, using: bash $ tar xzvf bash-4.0.tar.gz bash $ cd bash-4.0 bash-4.0 $ mkdir patches bash-4.0 $ # line breaks added for clarity to you, the reader for i in

Re: Prompt length calculation fails when UTF-8 is used within \[ \]

Lasse Kärkkäinen wrote: >> ftp://ftp.cwru.edu/pub/bash/bash-4.0-patches > > So, what format are these and how do I patch (on Linux)? Apparently the > patch program doesn't recognize them automatically. They can be applied with `patch -p0'. > P.S. is there some specific reason why you do things i

Re: read -t 0 not supported? (poll)

On Wed, Jul 22, 2009 at 12:03 PM, Marc Herbert > wrote: > Hi, > > It seems polling using "read -t 0" is not supported. Apparently because > the implementation is based on alarm() (as opposed to be based on > poll()/select()). > > - is this true? > - if yes, doesn't this deserve more explicit doc

read -t 0 not supported? (poll)

Hi, It seems polling using "read -t 0" is not supported. Apparently because the implementation is based on alarm() (as opposed to be based on poll()/select()). - is this true? - if yes, doesn't this deserve more explicit documentation? Would anyone know about an alternative, convenient way to

Re: builtin test command file existence fails with negation

On Tue, Jul 21, 2009 at 8:42 PM, Lynn Kerby wrote: > > Thanks, I hadn't read the Open Group test command doc. However, I find the > doc inconsistent on this matter when read in its entirety. > > For starters, POSIX doesn't allow for the use of '-a' as a unary operator > (according to the RATIONA