Dapper status meeting followup

I wasn't able to attend the meeting due to the US Thanksgiving holiday, so
I'm going to post some comments inline here:

> Nov 24 06:05:21 <dholbach>	work: this week (today and tomorrow), c++ allocator change, BUG DAY, writing down a11y meeting notes, ubuntu-docs updating
> Nov 24 06:05:21 <dholbach>	work: next week: gnome-power spec'ing, more c++ allocator change, a11y spec'ing with hno73, getting the wiki up to scratch, bug triage (of course), motu mentoring

Currently there are no high-priority a11y specs and we don't have specific
a11y goals for Dapper; can you clarify what this is about?

> Nov 24 06:05:39 <Diziet>	Launchpad SSLCertificateAuthentication: some progress made yesterday; this will fail or succeed by tomorrow

What was the outcome here?  This was supposed to be a one-day project, so if
it has turned out to be more work than that, we need to discuss
(you/me/SteveA) before going any further.

> Nov 24 06:12:28 <infinity>	reducing-duplication: ongoing effort, as always, this last week saw good progress on kicking out multiple versions of libdb

While we will most likely continue to do work in this area on a regular
basis, please measure completion against the items listed in the spec so
that we know where we stand.

> Nov 24 06:12:28 <infinity>	usplash-*: tested vga16fb changes to use a lower resolution, muddled with assembly in syslinux to make the same change for CD booting, need more testing

How far off is this?  If we're going to do it, it needs to land ASAP in
order to be thoroughly tested.

> Nov 24 06:12:28 <infinity>	splash-down: Have a reasonable working implementation, will need to ping maintainers of display managers to hook things in correctly and coordinate uploads

Unless there is a specific reason to block on others, please go ahead and
make those modifications and uploads yourself.

Oliver: this doesn't affect ldm unless you've added a shutdown function to
it.

> Nov 24 06:12:30 <infinity>	general: helped with kernel security updates last week (l-r-m and linux-meta), need to get a new l-r-m and linux-meta in this week for the new kernel, spent plenty of time babysitting various components to make modular X buildable, currently polishing off most of that tonight, buildd tasks have been reasonably onerous this last week.

As discussed with BenC, please help him get up to speed on l-r-m so that he
can share this responsibility.

> Nov 24 06:14:42 <Kamion>	cd-bootloader: CD timeout restored. Upgraded to syslinux 3.11. Sent call for gfxboot testing to ubuntu-users.

Is it feasible to switch this on for Flight 2?

> Nov 24 06:14:45 <Kamion>	cd-build-process: Single-architecture rebuilds implemented.

Great, how does it work?

> Nov 24 06:15:41 <Kamion>	of my meetings with mdz, I still need to have a chat about merge progress, CD blockers has become obsolete, Kubuntu CD builds has been done by deciding to bring Riddell into the cdimage group, and we haven't had the syncs discussion yet

Regarding merges, Scott has switched off MOM so that we can catch up.  At
that time, I assigned all open merge bugs, so at this point, everyone just
needs to finish clearing out their assigned list of merge bugs.

I discussed syncs with James, and he expressed a preference that he continue
to be the only one to process them.  Please find some time to talk about
this with him directly; no need to block on me for it.

What's the RT ticket for the cdimage changes for Jonathan?

> Nov 24 06:16:52 <jbailey>	I looked through and noticed yesterday that I still have grub asssigned to me for merging, I'll try to grab that tomorrow.

If this turned out to be non-trivial, please hand it off.  It only ended up
on your list due to outdated Bugzilla configuration.

> Nov 24 06:18:05 <Keybuk>	udev-roadmap: mostly done.  rcS script needs polish and the whole thing needs a bit more testing.  grepmap is to be kept, and updated for input subsystem and libsane.  still got to sort out udeb.  waiting on 2.6.15-4

2.6.15-4 seems to have landed; do you have an estimate of when this can
land?

I saw your bootchart and am eager to try it out here.

> Nov 24 06:18:05 <Keybuk>	network-manager: largely no progress, discovered it doesn't seem to bring up wifi interfaces until a user logs in.  blocked on glibc dns patch.  might not be doable anyway.

If this really isn't the right solution, we need to be thinking right now
about an alternative to provide the bits of functionality we need: switching
between wireless networks, different network interfaces in the desktop,
automatic configuration attempts, etc.

> Nov 24 06:21:54 <Mithrandir>	livecd-performance: groundwork done, I can now get useful data out of the system.  Need to decide on further action to actually improve boot times.  Also, will get it into the default live cd  (togglable with a boot switch or similar) and call for testing.

Excellent; where can I see the output from your bootchart measurements?

> Nov 24 06:21:58 <Mithrandir>	embedded-ubuntu: I discovered that I got this spec last night, hence no status update on it.

I've fixed this.

> Nov 24 06:22:01 <Mithrandir>	network-authentication: Lots of interest around the spec, no work done on it so far.

How much of this is reasonably achievable for Dapper in your estimation?

> Nov 24 06:22:15 <Mithrandir>	Rant: It's very annoying when priorites are shifted around without any kind of notification.  The live-cd-performance spec was high priority, hence I spent time on it, now it's suddenly medium priority without anybody having told me.  I was also made approved for embedded-ubuntu, without any notification.
> Nov 24 06:22:20 <Mithrandir>	Until we use LP all the time and therefore see those kinds of things, notifications are needed.  They probably still will be needed then.

The spec tracker currently has some bugs.  Among these is the fact that the
access control model needs refinement; for example, anyone can create a new
spec and put your name on it.

Jane, please collect a list of problems like this and see that they are
communicated to Mark, Steve and Kiko.

If you have any questions about your assignments or their relative
priorities, contact me directly.

> Nov 24 06:25:53 <mvo>	* pyhton-apt improvments (needed for unattended pkg upgrades, g-a-i, cmd-not-found magic)
> Nov 24 06:25:53 <mvo>	* unattended upgrades script ready (in python-apt, need to be integrated into the APT::Periodic system  [Unattended package upgrades]
> Nov 24 06:25:53 <mvo>	* gdebi (direct deb install tool) almost ready for upload into archive (needs new python-apt first) [Support for packages by
> Nov 24 06:25:53 <mvo>	  third parties in external repos]
> Nov 24 06:25:58 <mvo>	* cmd-not-found magic work with zyga (prototype in   pyhton ready, misses still some stuff) ["Sugest packages to install to proivde missing programs"]
> Nov 24 06:25:59 <mvo>	* tried to make apt more smart about failed network    sources [Default apt sources.list configuration]-> might be used for [Improved handling of APT   cdrom: sources as well]
> Nov 24 06:26:00 <mvo>	* other goals: not started
> Nov 24 06:26:16 <mvo>	Will do:
> Nov 24 06:26:16 <mvo>	* upload new apt/python-apt/gdebi + all depenencies (abi break)
> Nov 24 06:26:16 <mvo>	* integrate unattended upgrades script into the APT::Periodic system
> Nov 24 06:26:16 <mvo>	* 3rd party packages channel work
> Nov 24 06:26:16 <mvo>	* use hide-admin-tools for update-notifier

release-upgrades should be your next priority; that's a big chunk of work
that needs to get started right away.

Is there a spec for gdebi?  I don't remember putting it on the Dapper list.

> Nov 24 06:28:28 <ogra>	* thin-client-sound: updated the spec (pending approval), started adopting the proof of concept in real code in my local bzr branch, discussed changes to gstreamer with pitti, updated the spec accordingly

Please make sure you do this in a separate branch from your ldm changes.

> Nov 24 06:28:28 <ogra>	* thin-client-local-devices: started some initial tests, ltspfs/ltspfsd cvs version uploaded and build, binary is in the archive
> Nov 24 06:28:28 <ogra>	* thin-client-memory-usage: updated the spec (pending approval), did some further testing
> Nov 24 06:28:28 <ogra>	* thin-client-faster-startup: updated the spec (pending approval)

If you are blocked on spec approvals for any of this, please mail me
explicitly, as I am not frequently polling Launchpad.

> Nov 24 06:34:35 <Diziet>	I think it might have been me that changed the Target to Release because it looked like it had just been forgotten.  Was that wrong of me ?
> Nov 24 06:35:08 <ogra>	i think thats a mdz job normally
> Nov 24 06:35:21 <Diziet>	Oh.

Again, spec tracker bugs.  DapperGoals in the wiki is authoritative until
the spec tracker is fixed.

> Nov 24 06:36:21 <pitti>	hide-admin-tools-to-users: no actions since last week; BLOCK: sudo logging decision, see u-d mail

Will take a look at your post.

> Nov 24 06:36:26 <pitti>	automated-problem-reports: BLOCK: approval from Keybuk, would appreciate opinions from other people, too (elmo, Kamion, mdz) since this is very intrusive

Low priority; if it's intrusive and/or a lot of work, it will be deferred.
I explicitly left it off of DapperGoals.

> Nov 24 06:46:06 <seb128>	video-playback: waiting for gstreamer0.10 which is due first week of december

Is someone working on the xine side of this?  Daniel?

> Nov 24 06:48:01 <Kamion>	sivang: ok, anything on home-user-backup? for the moment it's targetted to dapper; we'll need to talk to mdz and find out if it deliberately wasn't that way earlier
> Nov 24 06:48:05 <Kamion>	(and if so why)

I haven't read the spec yet.

> Nov 24 06:52:04 <doko>	toolchain: gcc/gcj packages split, libstdc++ allocator change in the works (pestering several people with NEW processing, promotions and package uploads), ia32-libs removal stalled (blocked by glibc biarch on amd64)

ia32-libs removal?

> Nov 24 06:56:40 <Kamion>	I make it 127 open merge bugs currently
> Nov 24 06:56:42 <Keybuk>	there are 120 merges outstanding, of which 10% date before dapper opened
> Nov 24 06:56:42 <pitti>	doko: #u-d, let's discuss that later
> Nov 24 06:56:50 <Keybuk>	127?  I have 122
> Nov 24 06:56:51 <doko>	pitti: ok
> Nov 24 06:56:51 <seb128>	pitti, doko: xulrunner?
> Nov 24 06:56:59 <Keybuk>	this is only a slight decrease on this time last week
> Nov 24 06:57:03 <Keybuk>	ergo. PEOPLE ARE NOT DOING ENOUGH MERGES
> Nov 24 06:57:14 <pitti>	all merge bugs are assigned now
> Nov 24 06:57:25 <doko>	Keybuk: I DONT KNOW THAT THERE ARE OPEN BUGS, you can't see them, if you are not the assigne

As above, all open merge bugs are currently assigned to a developer.  You
only need to worry about what's already on your list.

Remaining merges should be completed next week.

Thanks to everyone for your status updates (you're looking good overall) and
to Colin and Jane for running the meeting in my absence.


-- 
 - mdz