Attached patch fixes the optstring parameters for the
getopt_long call in aide.c file.
[...]
Could you please publish enhanced CVS snapshots which
contain a prebuilt configure script and generated
autoconf/automake files? That would make it easier to
test the snaphots because hardly everyone has
compatible versions of autoconf/automake and so on to
generate those files.
[...]
Additional feedback: On Red Hat Linux 9, I'm seeing
several issues with aide 0.9 and 20030902 CVS snapshot.
In particular, but not limited to:
* Option --check-config or -D gives a segfault.
* Running "aide" without any options does not perform a
check as explained in the manual (only explicit --check
does it).
* Any entry like "=/ p+u+g" (without the quotes) in
aide.conf recurses into sub-directories upon "aide
--check". I shouldn't when the '=' equal sign is in front.
* Upon aide --check I get lots of questionable
warnings, such as:
open_dir():Not a directory: /boot/grub/stage1
open_dir():Not a directory: /boot/grub/stage2
open_dir():Not a directory: /boot/grub/device.map
* The verbose debug output of the form "Matches rule
from line #155: ^/" is off by one. In my aide.conf, the
line is 154, not 155.
* Line 119 of compare_db.c contains a typo:
- error(255,"Debug, compare_md_entreis %p %p\n",e1,e2);
+ error(255,"Debug, compare_md_entries %p %p\n",e1,e2);
patch against aide.c
Logged In: YES
user_id=246445
I've committed your patch to CVS. Could you possibly
consolidate your other comments into a separate request (or
several separate requests) for the sake of organization?
All of your concerns can probably go into bug reports.
Logged In: YES
user_id=62241
> Could you possibly consolidate your other comments into a
separate
> request (or several separate requests) for the sake of
organization?
Will do so once I find the time to see what changes have
been applied to CVS since I have had encountered these issues.
Logged In: YES
user_id=246445
The getopt patch has been applied and separate tracker
requests for each of the other issues has been opened.
Therefore, I am closing this bug.