On 2007-07-04 23:02:55 +0100, Dave Shield wrote:
> On 04/07/07, Robert Story <[EMAIL PROTECTED]> wrote:
> > Have you tried Thomas' suggestion of using zip instead of tar?
>
> I hadn't had the chance, no. Hold on
>
> > I'm interested if that works.
>
> Yes - a zip archive seems to unpack cor
On 2007-07-04 15:25:01 -0400, Robert Story wrote:
> By sheer chance, I found a bug in read_config. I was checking to see what
> files an agent was reading, and saw lots of 'directory not present' messages
> logged for weird name (some including control characters). A short debugging
> session later
On 2007-07-04 15:50:46 -0400, Robert Story wrote:
> On Wed, 4 Jul 2007 20:15:27 +0100 Dave wrote:
> DS> On 04/07/07, Marcus Rueckert <[EMAIL PROTECTED]> wrote:
> DS> > ok after another discussion on irc with robert and thomas, we got v2 of
> DS> > the patch. instea
On 2007-07-04 17:42:57 +0200, Marcus Rueckert wrote:
> all the asserts removed in this patch use != or == for string
> comparison, which is undefined in C. after a discussion on irc with Wes
> and Robert, i followed Wes' suggestion and removed the offending
> asserts.
>
>
hi,
all the asserts removed in this patch use != or == for string
comparison, which is undefined in C. after a discussion on irc with Wes
and Robert, i followed Wes' suggestion and removed the offending
asserts.
Can we still apply it for 5.4.1?:)
darix
--
openSUSE - SUSE Linux is
On 2007-07-03 21:26:08 +0200, Thomas Anders wrote:
> I'm calling for votes to apply this to 5.4.1.
as it is all my "fault": +1 :)
darix
--
openSUSE - SUSE Linux is my linux
openSUSE is good for you
www.opensuse.org
-
On 2007-03-08 10:22:43 +, Dave Shield wrote:
> On 07/03/07, Marcus Rueckert <[EMAIL PROTECTED]> wrote:
> >On 2007-03-07 10:34:47 +, Dave Shield wrote:
> >> In particular, I propose the following:
>
> >> b) Move the following to a new directory
On 2007-03-07 10:34:47 +, Dave Shield wrote:
> The recent conversion of the Net-SNMP source code repository
> to be under the control of Subversion has done an outstanding
> job of retaining the previous CVS development history.
>However, I would suggest that the 'tags' and 'branches'
> tre
hi,
now that 5.4 is out of the door, we can return to this topic right?:)
iirc Wes promised to start a survey on -users about it.
any other planned steps i am missing?
darix
--
openSUSE - SUSE Linux is my linux
openSUSE is good for you
www.opensuse
On 2006-11-23 02:22:58 +0100, Thomas Anders wrote:
> The attached patch fixes dist/net-snmp.spec to allow for building RPMs on
> 64-bit
> platforms by consistently using proper %{_libdir} (translating to /usr/lib64
> on
> x86_64) instead of hard-coding /usr/lib. Tested on RHEL4/x86 and RHEL4/x86_
hi,
+1 for the patch
darix
--
openSUSE - SUSE Linux is my linux
openSUSE is good for you
www.opensuse.org
-
Take Surveys. Earn Cash. Influence the Future of IT
Join SourceForge
On 2006-11-16 01:23:36 +0100, Thomas Anders wrote:
I've committed a few additional manual pages that have been supplied by Debian:
>
> encode_keychange.1.def mib2c-update.1.def tkmib.1.def
> fixproc.1.def net-snmp-config.1.def traptoemail.1.def
>
> Attached please find the propos
hi,
the patch looks fine to me.
+1
darix
--
openSUSE - SUSE Linux is my linux
openSUSE is good for you
www.opensuse.org
-
Take Surveys. Earn Cash. Influence the Future of IT
J
On 2006-11-09 17:28:51 -0800, Fong Tsui wrote:
> /usr/lib/perl5/site_perl/5.8.0/i386-linux-thread-multi/auto/NetSNMP/Trap
> Receiver: Permission denied at /usr/lib/perl5/5.8.0/ExtUtils/Install.pm
> line 137
> make[1]: *** [pure_site_install] Error 255
> make[1]: Leaving directory
> `/home/tsui/p4/m
On 2006-11-12 15:18:10 +0100, Thomas Anders wrote:
> Magnus Fromreide wrote:
> > I am thus more inclined towards a patch that also changes the signature
> > of __scan_num_objid to
> >
> > static int __scan_num_objid _((char *, oid *, size_t *));
> >
> > to make the patch more correct.
>
> Agreed
On 2006-11-09 07:41:22 -0800, Wes Hardaker wrote:
> >>>>> "MR" == Marcus Rueckert <[EMAIL PROTECTED]> writes:
>
> MR> i just tested on multiple architectures. all 32bit apps are fine. all
> MR> 64bit architectures are broken.
>
> When you s
On 2006-11-08 13:00:08 -0800, Wes Hardaker wrote:
> > "TA" == Thomas Anders <[EMAIL PROTECTED]> writes:
>
> TA> Building CVS MAIN on Linux, darix and I noticed that "make
> TA> perltest" fails in the SNMP.pm MIB code tests:
>
> I'm actually not sure that the SNMP module has worked on 64 bit
>
On Wednesday 01 November 2006 18:49, Wes Hardaker wrote:
> Better alternative suggestions welcome (but be prepared to implement
> the changes).
svn log -v
darix
--
openSUSE - SUSE Linux is my linux
openSUSE is good for you
www.opensuse.org
-
taking into account the testing and how trivial the patch is:
+1
darix
--
openSUSE - SUSE Linux is my linux
openSUSE is good for you
www.opensuse.org
-
Using Tomcat but need to do m
On 2006-10-28 14:54:12 +0200, Thomas Anders wrote:
> I'm calling for votes to include this in 5.4.
+1
darix
--
openSUSE - SUSE Linux is my linux
openSUSE is good for you
www.opensuse.org
-
hi,
as the SF.net CVS is still/again unstable, i would like to suggest
moving to svn. I brought up this topic on irc a few times already.
Mediawiki used one of the last downtime for the migration aswell. [1]
Advantages I see for the move:
- similar UI as CVS
- changesets
- directory versioning
-
hi,
if you use tcpwrapper every connection gets logged.
without tcpwrapper this only happens if "-a" is specified at the
commandline. i dont think this is really intended.
The patch to fix that seems pretty easy:
[[[
Index: agent/snmp_agent.c
==
hi,
i was asked to provide a summary of all problems, we ran into so far:
first issue:
this is the first crash i encountered.
it only happens on bigendian 64bit arches.
http://pixel.global-banlist.de./bt.txt
i applied the following patch to fix it:
[[[
Index: agent/agent_registry.c
On 2006-02-01 17:25:04 +, Dave Shield wrote:
> it's objecting to, and how far through it gets.
> Could you perhaps try adding a debug print statement
> to that loop, that would indicate the value of the
> "oid_index" OID array at the point at which it fails.
while figuring out the best params
On 2006-02-01 17:25:04 +, Dave Shield wrote:
> H... that's a pity. The stack trace is only
> going to be of limited use. It would be helpful to
> know which bit of
>
> row->oid_index.oids[row->oid_index.len]++;
>
> it's objecting to, and how far through it gets.
> Could you perha
for debugging it directly.
i tested it with and without our patches/stackprotector/optimizations.
any ideas on this?
with kind regards
marcus rueckert
[1] we build it with -fstack-protector.
--
openSUSE - SUSE Linux is my linux
openSUSE is good for you
#0
26 matches
Mail list logo
