Please merge openipmi 2.0.16-1(main) from debian unstable(main)

[Bhavani Shankar]

Binary package hint: openipmi

Debian has a new version to be merged

openipmi (2.0.16-1) unstable; urgency=low

  * new upstream release from 2009-03-16
  * debian/compat debian/control raised to debhelper 5
  * debian/control removed URL in description new we
    have the Homepage: field
  * debian/control updated Standards-Version: (no changes needed)
  * debian/rules removed unused cdbd dependency

 -- Noèl Köthe <email address hidden> Tue, 28 Jul 2009 10:17:23 +0200

[Sebastien Bacher (seb128)]

thank you for your work there but how do decide what to update, you seem to work on a lot of updates but there is a reason the autosyncs are stopped that's not to play catching up manually now but to focus on stabilization, is there any change worth having there?

[Bhavani Shankar (bhavi)]

yes seb its a new upstream version with the following upstream changelog excerpt

2009-03-16 Corey Minyard <email address hidden>

 * configure.in: Move to version 2.0.16. It's been a while and
 a release is a good idea, I think.

2009-01-16 Corey Minyard <email address hidden>

 * lib/normal_fru.c: An Intel document (IPMI 0.9 to IPMI 1.0 Change
 Summary and Porting Considerations from October 1, 1998) mentioned
 the FRU version was updated from 1 to 2. Nothing else really
 changed this, but some systems used the version 2. So account for
 this change.

2009-01-09 Jan Safranek <email address hidden>

 * sample/rmcp_ping.c: I know, it's just a library sample, but
 people find rmcp_ping useful and it should report something when
 user provides wrong command line arguments.

2008-12-09 Corey Minyard <email address hidden>

 * lib/ipmi_lan.c, lib/oem_atca.c, sample/dump_sensors.c: Fix some
 compiler warnings.

2008-12-08 Dann Frazier

 * doc/IPMI.ltx: Fix some typos.

2008-11-19 Corey Minyard <email address hidden>

 * lib/domain.c: Allow scanning IPMB address 0x20 even if the the
 channel medium is not IPMB. This helps fake out things where
 there is no IPMB channel but an MC at 0x20 is still reported
 in the SDRs.

2008-11-07 Corey Minyard <email address hidden>

 * lib/ipmi_lan.c: Fix some bugs notices by Stephane Blain
 (stblain at users.sourceforge.net) dealing with the hash
 function for ipmi lan objects.

2008-09-07 Corey Minyard <email address hidden>

 * unix/selector.c: Fixed a problem with a missing NULL check
 on sending a signal.

 * sample/Makefile.am, sample/waiter_sample.c: Add a sample that
 does waiting using the OS handler waiter.

2008-09-05 Corey Minyard <email address hidden>

 * lib/ipmi_payload.c: Work around a bug in some systems that
 switch the request and response address in IPMI LAN messages.

2008-07-14 Corey Minyard <email address hidden>

 * unix/selector.c: Fixed handling of waking up threads.

2008-07-02 Corey Minyard <email address hidden>

 * utils/os_handler.c: Fixed a messed up refcount when a waiter
 times out.

 * lib/strings.c: Fixed another typo noticed by Nancy Nichols.

2008-06-29 Corey Minyard <email address hidden>

 * lib/hmac.c: Someone named "Sia" reported on Sourceforge that
 OpenIPMI was using the wrong key for HMAC-MD5-128. And they were
 right. Problem is now fixed.

2008-06-26 Corey Minyard <email address hidden>

 * lib/strings.c: Fix a typo noticed by Nancy Nichols.

 * configure.in: Fixed the dia test to work with 3-level version.

 * lib/hmac.c: A variable in the SHA1 init function was declared
 static instead of const. This means that in a multi-threaded
 program, it is possibly (though quite unlikely) that the data
 could get corrupted.

 * swig/python/openipmiguid/gui_domainDialog.py: Improve the error
 handling to make it easier to track down problems.

2008-05-21 Corey Minyard <email address hidden>

 * lib/domain.c: Don't remove everything from the IPMB scan if
 all the get channel commands fail. There may still be something
 on IPMB 0.

 * lib/domain.c: Fix an overflow problem dealing with IPMB
 addresses. If the IPMB ignores list covered all addresses, the
 scan code would go into an...

[Sebastien Bacher (seb128)]

you didn't get the point of my comment, the issue is that your are flooding the sponsor queue with new version updates, while you work is appreciate we do stop autosync on debian for a reason which is that we want to focus on stabilization now and not try to get every new version, do we really need those updates for karmic?

[Mathias Gug (mathiaz)]

I agree with Sebastien in the general sense. Only specific packages should now be picked from Debian and merged if relevant. We're focusing on stabilizing the packages for Karmic for now.

That being said for this specific package (openipmi) and this new version I think the merge is worth having in Ubuntu Karmic. However I'd like to get the Ubuntu diff (init script) pushed back to Debian first (by filling a bug in BTS and linking the Debian bug to this one) before sponsoring the merge in Ubuntu.

Thanks again for the work you're doing in Ubuntu - please be considerate in picking which packages to merge from Debian. If you have any doubt I'd suggest to stop by #ubuntu-devel and ask someone there if merging the new version from Debian is worthwhile.

[Bhavani Shankar (bhavi)]

okay matthias and seb

sorry for that.. I work and test only those packages which are mostly new upstream versions in debian containing fair amount of changes

Will curb myself down and forward that change to debian

Regards

[Bhavani Shankar (bhavi)]

change forwarded to debian

Regards

[Daniel Holbach (dholbach)]

If there's a very good reason to have the new upstream version we will consider it.

http://qa.ubuntuwire.com/bugs/rcbugs/ for example is a good indicator for things that we should probably have.

Is this merge important or not considering the release in a few weeks?

[Bhavani Shankar (bhavi)]

yes daniel but I opened this bug a long time back....

m unsure of its status now as karmic has reached FF

Regards

[Daniel Holbach (dholbach)]

Please review the changes and make up your mind.

Unsubscribing sponsors team for now.

[Mathias Gug (mathiaz)]

Too late for karmic. Marking Won't Fix.