> Fix the word boundary handling of vi-eword...
> Provide shell variable $vimode to control VImode.
> If VImode=1, use "_" as the default for $wordchars, to match vi's default.
> Modify c_to_class() to use isword().
> rename e_capitolcase ->e_capitalcase (Luke Mewburn)
> Fix e_capitolcase to e_capitalcase
> Fix comparison of $EDITOR or $VISUAL against names of background
> processes so that it doesn't just prefix match.
> Rework run-fg-editor to use $editors exclusively if its set, and
> ignore the use of $EDITOR (default "ed") and $VISUAL (default "vi").
We're now starting the downstream patches at number 100, in order to
have a reserved numbers for upstream patches [0-99]. In case we reach
the number 99 for upstream patches, we should really consider doing a
rebase...
The patch has been updated to be more clear. This is a downstream
patch that we need to keep, it has been requested in BZ: #247637
Upstream did not accept the patch, because this is specific to Linux:
<http://mx.gw.com/pipermail/tcsh/2016-April/004972.html>
'tcsh_posix_status' is a downstream variable not accepted by upstream.
'anyerror' variable was introduced by upstream instead.
This patch keeps the 'tcsh_posix_status', but displays a warning to
user that this variable is deprecated and its support will be removed
in Fedora 25.
This patch was not working in latest tcsh packages in Fedora, and it
was not accepted by upstream. Generally, shells (bash, ksh, tcsh, ...)
does not want to have this kind of binary parsing inside of them. See:
<http://mx.gw.com/pipermail/tcsh-bugs/2013-April/000833.html>
This patch breaks things more common than the issue it solves. See:
http://mx.gw.com/pipermail/tcsh-bugs/2016-April/001030.html
Bottom-line is: tcsh can't process its keywords in HERE-DOC correctly,
and most likely it will never be possible to fix/solve this.
The issue solved with this patch is not reproducible with 6.19.00
upstream version of tcsh. It was most likely fixed/mitigated as a
side-effect of some commit in the upstream.
This patch was never accepted by upstream since upstream does not like
the solution: http://mx.gw.com/pipermail/tcsh-bugs/2016-April/001031.html
The issue solved by this patch has been tested on Fedora 23, but
reproduction was not successful. The package 'nss_ldap' is now obsolote
(has been retired) and is no longer part of Fedora. It was superseded by
'nss-pam-ldapd' and 'sssd'. The issue has been most likely mitigated
with the new identity management used in Fedora (FreeIPA/SSSD).
In case the issue will re-appear, we will need to find work with
upstream to find a solution which they would be willing to accept.
