QA/Meetings/20080423

-!- wwoods changed the topic of #fedora-meeting to: Fedora QA Meeting | blocker list: <a href="https://bugzilla.redhat.com/showdependencytree.cgi?id=F9Blocker&hide_resolved=1">https://bugzilla.redhat.com/showdependencytree.cgi?id=F9Blocker&hide_resolved=1</a>		<a href="#t11:39" class="time">11:39</a>
wwoods	woo meeting	<a href="#t11:44" class="time">11:44</a>
wwoods	I think we're ~2 weeks from doing the serious final spin	<a href="#t11:45" class="time">11:45</a>
rdieter	woo indeed	<a href="#t11:47" class="time">11:47</a>
wwoods	I need to gather some better metrics but it looks like we have ~52 blockers	<a href="#t11:47" class="time">11:47</a>
wwoods	33 on the list, 24 children, 5 blockers	<a href="#t11:48" class="time">11:48</a>
wwoods	33+24-5=52	<a href="#t11:48" class="time">11:48</a>
wwoods	a *lot* of them are in MODIFIED	<a href="#t11:48" class="time">11:48</a>
wwoods	which is really good	<a href="#t11:48" class="time">11:48</a>
f13	yeah, and a lot have been closed too recently	<a href="#t11:48" class="time">11:48</a>
wwoods	yup	<a href="#t11:48" class="time">11:48</a>
f13	I've been watching the closures	<a href="#t11:48" class="time">11:48</a>
wwoods	yeah it's been steady	<a href="#t11:48" class="time">11:48</a>
* f13 stares at the rawhide compose process that is /still/ going.		<a href="#t11:49" class="time">11:49</a>
wwoods	I've been closing a couple a day	<a href="#t11:49" class="time">11:49</a>
wwoods	unfortunately we've also been adding a couple a day, so..	<a href="#t11:49" class="time">11:49</a>
f13	some of those are dubious "blockers"	<a href="#t11:50" class="time">11:50</a>
f13	maybe I need to be better at doing punt->target	<a href="#t11:50" class="time">11:50</a>
f13	problem is if it's not on the blocker, some developers don't feel the need to get it done for F9	<a href="#t11:50" class="time">11:50</a>
f13	but maybe that's not my problem to worry about	<a href="#t11:50" class="time">11:50</a>
wwoods	maybe we need to message that better	<a href="#t11:51" class="time">11:51</a>
cebbert	both of the kernel blockers that are left are kind of dubious	<a href="#t11:51" class="time">11:51</a>
f13	cebbert: feel free to toss them over at F9Target	<a href="#t11:51" class="time">11:51</a>
cebbert	one looks like it might be broken hardware	<a href="#t11:51" class="time">11:51</a>
wwoods	Target = "We really want this for F9 and will accept freeze breaks for it. But we won't delay the release any further if you don't fix it."	<a href="#t11:51" class="time">11:51</a>
wwoods	Blocker = "We can't release unless this is fixed."	<a href="#t11:51" class="time">11:51</a>
rdieter	I cheated, and added a pet bug or 2 to f9blocker, in the hopes of at least getting a maintainer comment.	<a href="#t11:52" class="time">11:52</a>
wwoods	heh	<a href="#t11:52" class="time">11:52</a>
wwoods	we'll play a little good QA guy / bad QA guy	<a href="#t11:52" class="time">11:52</a>
wwoods	WE CAN'T RELEASE UNTIL THIS IS FIXED. MAINTAINER, YOUR ASS IS ON THE LINE HERE	<a href="#t11:53" class="time">11:53</a>
wwoods	"oh now rdieter, don't be so hasty..."	<a href="#t11:53" class="time">11:53</a>
wwoods	etc.	<a href="#t11:53" class="time">11:53</a>
wwoods	heh	<a href="#t11:53" class="time">11:53</a>
wwoods	anyway, which 2 bugs?	<a href="#t11:53" class="time">11:53</a>
rdieter	<a href="https://bugzilla.redhat.com/show_bug.cgi?id=443320">https://bugzilla.redhat.com/show_bug.cgi?id=443320</a>	<a href="#t11:53" class="time">11:53</a>
buggbot	Bug 443320: low, low, ---, Kristian Høgsberg, ASSIGNED , Cannot enter keyboard input into kdm after logout	<a href="#t11:53" class="time">11:53</a>
wwoods	seeing as we have >50 bugs and 10 work days to fix 'em all, I'm anxious to be sure they're *seriously* blockers at this point	<a href="#t11:54" class="time">11:54</a>
rdieter	<a href="https://bugzilla.redhat.com/show_bug.cgi?id=427383">https://bugzilla.redhat.com/show_bug.cgi?id=427383</a>	<a href="#t11:54" class="time">11:54</a>
wwoods	ah, see, that's really nasty	<a href="#t11:54" class="time">11:54</a>
buggbot	Bug 427383: high, medium, ---, Adam Jackson, ASSIGNED , xorg-x11-drv-vesa causes KDE 4 to crash (on some hardware)	<a href="#t11:54" class="time">11:54</a>
wwoods	I'd consider that first one a blocker. -vesa... is.. different	<a href="#t11:54" class="time">11:54</a>
rdieter	the 1st one is my main concern.	<a href="#t11:54" class="time">11:54</a>
wwoods	I'd want ajax's opinion. which is usually: "If you're using -vesa you've already lost"	<a href="#t11:54" class="time">11:54</a>
rdieter	:) nod	<a href="#t11:54" class="time">11:54</a>
rdieter	I'd be happy with any comment, even that one.	<a href="#t11:55" class="time">11:55</a>
f13	ajax said he's making good progress on the f9-x-blockers.	<a href="#t11:55" class="time">11:55</a>
f13	and just going through them one at a time	<a href="#t11:56" class="time">11:56</a>
wwoods	yeah, he hasn't been complaining about massive impending doom	<a href="#t11:56" class="time">11:56</a>
cebbert	I reported bug 443729 last night; it seems some others are hitting nearly the same problem	<a href="#t11:56" class="time">11:56</a>
buggbot	Bug <a href="https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=443729">https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=443729</a> high, low, ---, Anaconda Maintenance Team, NEW , Cannot install preview release to encrypted partition	<a href="#t11:56" class="time">11:56</a>
wwoods	cebbert: sounds like bug 439633	<a href="#t11:57" class="time">11:57</a>
buggbot	Bug <a href="https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=439633">https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=439633</a> low, low, ---, Anaconda Maintenance Team, NEEDINFO , F9-Beta anaconda crash with PVCreateError	<a href="#t11:57" class="time">11:57</a>
wwoods	(which is on the blocker list)	<a href="#t11:57" class="time">11:57</a>
* jds2001 sits in the back		<a href="#t11:58" class="time">11:58</a>
wwoods	it's possible that udev hasn't created /dev/mapper/luks-sda2 yet	<a href="#t11:59" class="time">11:59</a>
wwoods	udev races are fuuuuun	<a href="#t11:59" class="time">11:59</a>
jeremy	not really	<a href="#t11:59" class="time">11:59</a>
jeremy	;)	<a href="#t11:59" class="time">11:59</a>
wwoods	you might try the debugging stuff as suggested in 439633 (check for udevd / device node)	<a href="#t11:59" class="time">11:59</a>
wwoods	fun with a capital F-U	<a href="#t12:00" class="time">12:00</a>
wwoods	oh, and I reproduced the notorious "Assertion c->xlib.lock' failed." bug	<a href="#t12:01" class="time">12:01</a>
wwoods	with vinagre, oddly	<a href="#t12:01" class="time">12:01</a>
wwoods	should be fixed, but the package hasn't hit rawhide yet	<a href="#t12:02" class="time">12:02</a>
wwoods	okay so. rdieter, I think you can ping ajax if you want a comment about keeping that -vesa bug as a blocker or not	<a href="#t12:02" class="time">12:02</a>
wwoods	anything that's in MODIFIED should be retested and closed ASAP	<a href="#t12:03" class="time">12:03</a>
wwoods	jds2001: if something's in MODIFIED, but the patch doesn't fix the bug	<a href="#t12:03" class="time">12:03</a>
rdieter	shrug, I think I'll move it to F9Target, ajax has bigger fish to fry.	<a href="#t12:03" class="time">12:03</a>
wwoods	does it go back to ASSIGNED or just stay in MODIFIED?	<a href="#t12:03" class="time">12:03</a>
wwoods	rdieter: well, the thing is, lots of reviewers use qemu+live image to test	<a href="#t12:04" class="time">12:04</a>
wwoods	so if that is busted, things look bad	<a href="#t12:04" class="time">12:04</a>
rdieter	qemu uses cirrus now, not vesa.  (apparently VirtualBox uses vesa still tho)	<a href="#t12:04" class="time">12:04</a>
wwoods	ohhhh	<a href="#t12:04" class="time">12:04</a>
rdieter	I'll ping ajax, hear what he has to say.	<a href="#t12:05" class="time">12:05</a>
poelcat	wwoods: 'not fixed' should be in ASSIGNED	<a href="#t12:05" class="time">12:05</a>
jds2001	wwoods: it should go back to MODIFIED	<a href="#t12:05" class="time">12:05</a>
poelcat	lol	<a href="#t12:06" class="time">12:06</a>
jds2001	oops	<a href="#t12:06" class="time">12:06</a>
jds2001	multitasking	<a href="#t12:06" class="time">12:06</a>
poelcat	jds2001: what?	<a href="#t12:06" class="time">12:06</a>
wwoods	ack. was using bugzilla wrong. we actually have 64 blockers.	<a href="#t12:06" class="time">12:06</a>
jds2001	try go back to ASSIGNED :)	<a href="#t12:06" class="time">12:06</a>
wwoods	k, gotcha	<a href="#t12:07" class="time">12:07</a>
wwoods	I don't have any special advice on working on the blockers, and I really don't want to run through all 64	<a href="#t12:07" class="time">12:07</a>
wwoods	if you've got stuff you find questionable, either punt to F9Target or bring it to #fedora-qa and we'll discuss	<a href="#t12:08" class="time">12:08</a>
wwoods	but really, the blocker list should only contain things we would delay the release if it was left unfixed	<a href="#t12:08" class="time">12:08</a>
wwoods	things for which ..	<a href="#t12:08" class="time">12:08</a>
jds2001	agreed.	<a href="#t12:08" class="time">12:08</a>
wwoods	so.. yeah. keep an eye on 'em? everything seems to be going OK, though	<a href="#t12:08" class="time">12:08</a>
* jds2001 has been somewhat agressive removing things		<a href="#t12:08" class="time">12:08</a>
wwoods	and now, a quick word about RCs	<a href="#t12:09" class="time">12:09</a>
wwoods	so, the RC images will not typically be made public	<a href="#t12:09" class="time">12:09</a>
wwoods	but that's OK 'cuz anyone who actually wants to test stuff with images can build their own	<a href="#t12:09" class="time">12:09</a>
wwoods	if there's anything *really* important we might put some up on the torrent server	<a href="#t12:09" class="time">12:09</a>
jds2001	yep, i do  that anyways	<a href="#t12:09" class="time">12:09</a>
wwoods	but, for the most part, I think RHT-internal people will take care of the media-based testcases	<a href="#t12:10" class="time">12:10</a>
* f13 points out FreeBSD where the binaries themselves aren't relaly made public for release testing, releases consisit fo building everything from source on a workstation the day of release		<a href="#t12:10" class="time">12:10</a>
f13	well, if my reading of their release engineering guide is correct	<a href="#t12:10" class="time">12:10</a>
* rdieter shudders		<a href="#t12:10" class="time">12:10</a>
wwoods	unless you're testing media-based stuff, 99% of the time testing rawhide is *exactly* the same as testing whatever RC we have floating around	<a href="#t12:10" class="time">12:10</a>
wwoods	when we make a first actual RC I'll have a test results page available	<a href="#t12:11" class="time">12:11</a>
wwoods	and the community folks are welcomed and encouraged to test and fill in squares on the matrix, as always	<a href="#t12:11" class="time">12:11</a>
wwoods	I'll make a post to -devel-list and -test-list when that happens	<a href="#t12:12" class="time">12:12</a>
wwoods	but I just want to make sure everything's clear	<a href="#t12:12" class="time">12:12</a>
wwoods	we aren't hiding anything and we aren't preventing participation	<a href="#t12:12" class="time">12:12</a>
wwoods	we're dividing up the work in a way that makes sense	<a href="#t12:12" class="time">12:12</a>
jds2001	wwoods: does it matter if we use pungi to build images and fill in image-based results?	<a href="#t12:12" class="time">12:12</a>
wwoods	jds2001: as long as you're using the same ks + tree, that should be perfectly acceptable	<a href="#t12:13" class="time">12:13</a>
jds2001	k	<a href="#t12:13" class="time">12:13</a>
wwoods	f13: /usr/share/fedora-release/rawhide-fedora.ks is still the canonical ks for images, right?	<a href="#t12:13" class="time">12:13</a>
wwoods	will that get an update when we update fedora-release for the RC, or will there be a new file?	<a href="#t12:13" class="time">12:13</a>
f13	wwoods: it'll be named appropriately in fedora-release	<a href="#t12:14" class="time">12:14</a>
f13	and the F-9 package of pungi will also gain an appropriately named copy	<a href="#t12:14" class="time">12:14</a>
wwoods	f13: gotcha	<a href="#t12:14" class="time">12:14</a>
wwoods	how likely is it that we'll have rc-specific changes that we'll need to put on a web page or something?	<a href="#t12:15" class="time">12:15</a>
wwoods	or is the ks basically stable at this point	<a href="#t12:15" class="time">12:15</a>
wwoods	and using the packaged one will always be preferred?	<a href="#t12:15" class="time">12:15</a>
f13	using the packaged one is preferred, and if we do have to make last minute changes, it'll be commited to the upstream source control so we can point people there	<a href="#t12:17" class="time">12:17</a>
wwoods	gotcha	<a href="#t12:17" class="time">12:17</a>
wwoods	I'll make a note of that on the matrix page, for testers who want to produce their own images	<a href="#t12:18" class="time">12:18</a>
wwoods	okay then - anything else we want to discuss? is the plan for the next couple of weeks clear?	<a href="#t12:19" class="time">12:19</a>
wwoods	f13: is the schedule accurate for when we're going to start producing RCs?	<a href="#t12:19" class="time">12:19</a>
f13	wwoods: I do believe so.  however that Friday is going to be.. interesting.  We're supposed to be moving offices	<a href="#t12:20" class="time">12:20</a>
f13	depending on when the move happens is when the first RC starts getting composed	<a href="#t12:20" class="time">12:20</a>
wwoods	oh crap, right	<a href="#t12:20" class="time">12:20</a>
wwoods	well have fun with that	<a href="#t12:20" class="time">12:20</a>
wwoods	heh	<a href="#t12:20" class="time">12:20</a>
f13	I may be able to spin up something thursday after I land and quickly sync it off somewhere that will be available for testing during the move	<a href="#t12:20" class="time">12:20</a>
wwoods	but we plan to start making RCs at the end of next week.	<a href="#t12:21" class="time">12:21</a>
wwoods	okay, so yeah. this week and next week: clean up blocker list, retest as much as possible, etc.	<a href="#t12:22" class="time">12:22</a>
wwoods	end of next week: RC1	<a href="#t12:22" class="time">12:22</a>
wwoods	that's when I'll make the testing matrix	<a href="#t12:22" class="time">12:22</a>
wwoods	and we'll shift priorities from Blocker Triage to Release Testing	<a href="#t12:23" class="time">12:23</a>
jds2001	and we'll still get accused of not testing :)	<a href="#t12:23" class="time">12:23</a>
wwoods	then we will summarily kick the ass of the testing and release Fedora 9 to deafening applause, choirs of angels, &c	<a href="#t12:24" class="time">12:24</a>
wwoods	ha. yes.	<a href="#t12:24" class="time">12:24</a>
jds2001	sounds great! :)	<a href="#t12:24" class="time">12:24</a>
wwoods	Fedora QA (which does not exist)	<a href="#t12:24" class="time">12:24</a>
wwoods	will test Red Hat's desktop product (which also does not exist)	<a href="#t12:24" class="time">12:24</a>
jds2001	lol	<a href="#t12:24" class="time">12:24</a>
wwoods	this sure is a lot of work for something that doesn't exist	<a href="#t12:25" class="time">12:25</a>
wwoods	but I guess it's OK 'cuz I don't exist either	<a href="#t12:25" class="time">12:25</a>
wwoods	anyway, enough nonsense. any other actual discussion topics?	<a href="#t12:25" class="time">12:25</a>
wwoods	otherwise: meeting ends in 60s	<a href="#t12:25" class="time">12:25</a>
f13	there is no fpoon	<a href="#t12:25" class="time">12:25</a>
wwoods	indeed	<a href="#t12:27" class="time">12:27</a>
wwoods	thanks all	<a href="#t12:27" class="time">12:27</a>

Generated by irclog2html.py 2.3 by <a href="mailto:marius@pov.lt">Marius Gedminas</a> - find it at <a href="http://mg.pov.lt/irclog2html/">mg.pov.lt</a>!