16:01:42 <tflink> #startmeeting f20alpha-blocker-review-2 16:01:42 <zodbot> Meeting started Wed Aug 28 16:01:42 2013 UTC. The chair is tflink. Information about MeetBot at http://wiki.debian.org/MeetBot. 16:01:42 <zodbot> Useful Commands: #action #agreed #halp #info #idea #link #topic. 16:01:42 <tflink> #meetingname f20alpha-blocker-review-2 16:01:42 <tflink> #topic Roll Call 16:01:42 <zodbot> The meeting name has been set to 'f20alpha-blocker-review-2' 16:01:54 * pwhalen is here 16:02:10 * seasc looks around 16:02:43 * kparal here 16:02:45 * mkrizek is here 16:02:58 <tflink> any volunteers for secretary duty? 16:03:11 * nirik is lurking 16:04:20 <kparal> tflink: me 16:04:24 <tflink> kparal: thanks 16:04:46 * kparal pokes pschindl 16:04:51 * pschindl is here 16:04:54 * spstarr_work looks around 16:05:01 * roshi is lurking 16:05:12 * ignatenkobrain here 16:05:12 <kparal> roshi: welcome 16:05:17 <tflink> looks like we have plenty of people to get started with the always exciting ... boilerplate! 16:05:20 <spstarr_work> good timing after just getting lunch :) 16:05:28 <ignatenkobrain> hey kparal tflink nirik 16:05:28 <tflink> #topic Introduction 16:05:33 <roshi> thanks kparal 16:05:33 <tflink> oh 16:05:43 <tflink> #chair kparal 16:05:43 <zodbot> Current chairs: kparal tflink 16:05:50 <tflink> Why are we here? 16:05:50 <tflink> #info Our purpose in this meeting is to review proposed blocker and nice-to-have bugs and decide whether to accept them, and to monitor the progress of fixing existing accepted blocker and freeze exception bugs. 16:05:58 <tflink> #info We'll be following the process outlined at: 16:05:58 <tflink> #link https://fedoraproject.org/wiki/QA:SOP_Blocker_Bug_Meeting 16:06:05 <tflink> #info The bugs up for review today are available at: 16:06:05 <tflink> #link http://qa.fedoraproject.org/blockerbugs/current 16:06:11 <tflink> #info The criteria for release blocking bugs can be found at: 16:06:11 <tflink> #link https://fedoraproject.org/wiki/Fedora_20_Alpha_Release_Criteria 16:06:16 <tflink> #info Up for review today, we have: 16:06:23 <tflink> #info 7 Proposed Blockers 16:06:23 <tflink> #info 4 Accepted Blockers 16:06:23 <tflink> #info 0 Proposed Freeze Exceptions 16:06:23 <tflink> #info 2 Accepted Freeze Exceptions 16:06:35 <tflink> if there are no objections, we'll dive into the proposed blockers 16:07:15 <tflink> #topic (1000889) AttributeError: 'DeviceFormat' object has no attribute 'mount' 16:07:16 * handsome_pirate waves 16:07:18 <tflink> #link https://bugzilla.redhat.com/show_bug.cgi?id=1000889 16:07:21 <tflink> #info Proposed Blocker, anaconda, NEW 16:08:22 <tflink> it sounds like this might not be just in the partitioning screen, but there appear to be 3 reproductions with TC1 16:08:25 <tflink> +1 blocker 16:08:28 <ignatenkobrain> +1 from me 16:08:29 <kparal> +1 16:08:31 <pschindl> +1 16:08:37 <spstarr_work> +1 16:08:49 <mkrizek> +1 16:09:04 <tflink> proposed #agreed 1000889 - AcceptedBlocker - Violates the following F20 alpha release criterion: "The installer must be able to complete an installation to a single disk using automatic partitioning." 16:09:09 <tflink> ack/nak/patch? 16:09:12 <ignatenkobrain> ack 16:09:21 <kparal> ack 16:09:23 <mkrizek> ack 16:09:25 <handsome_pirate> ack 16:09:27 <tflink> #agreed 1000889 - AcceptedBlocker - Violates the following F20 alpha release criterion: "The installer must be able to complete an installation to a single disk using automatic partitioning." 16:09:35 <tflink> #topic (1000927) 20 Alpha TC1 i386 install image (netinst or DVD) freezes in the first step 16:09:38 <tflink> #link https://bugzilla.redhat.com/show_bug.cgi?id=1000927 16:09:40 <tflink> #info Proposed Blocker, anaconda, NEW 16:09:58 <handsome_pirate> +1 blocker 16:10:19 <tflink> +1 here as well 16:10:40 <pschindl> +1 16:10:42 <ignatenkobrain> +1 16:10:46 <spstarr_work> certainly +1 16:10:48 <pschindl> what a nice blockers :) 16:10:48 <mkrizek> +1 16:10:48 <kparal> +1 16:11:14 <seasc> +1 16:11:16 <tflink> proposed #agreed 1000927 - AcceptedBlocker - Violates the following F20 alpha release criterion for 32bit installs: "When using the dedicated installer images, the installer must be able to complete an installation using the text, graphical and VNC installation interfaces." 16:11:29 <handsome_pirate> ack 16:11:29 <tflink> yeah, I like the clear cut blockers :) 16:11:31 <mkrizek> ack 16:11:33 <pschindl> ack 16:11:35 <tflink> must be early in the release 16:11:38 <ignatenkobrain> ack 16:11:41 <tflink> #agreed 1000927 - AcceptedBlocker - Violates the following F20 alpha release criterion for 32bit installs: "When using the dedicated installer images, the installer must be able to complete an installation using the text, graphical and VNC installation interfaces." 16:11:49 <tflink> #topic (994180) Boot-time LUKS passphrase input *always* defaults to en-us 16:11:52 <tflink> #link https://bugzilla.redhat.com/show_bug.cgi?id=994180 16:11:54 <tflink> #info Proposed Blocker, anaconda, NEW 16:12:00 <seasc> +1 16:12:34 <ignatenkobrain> +1/-1 16:13:05 <kparal> ignatenkobrain: you can't vote +1 blocker and -1 FE 16:13:10 <kparal> doesn't make sense 16:13:16 <handsome_pirate> Isn't this a Beta? 16:13:21 * handsome_pirate looks at criteria 16:13:25 <spstarr_work> doesn't feel blocker to me, but i am locale bias 16:13:29 <spstarr_work> -1 16:13:48 <ignatenkobrain> kparal: I meant "refrain" 16:13:49 <tflink> handsome_pirate: alpha criterion is quoted in c#11 16:14:01 <handsome_pirate> Ah 16:14:09 <spstarr_work> hmm 16:14:15 <tflink> how have we handled stuff like this in the past? 16:14:26 <tflink> I know that system passwords have to be ascii 16:14:38 <handsome_pirate> I don't recall hitting this in the past 16:14:46 <tflink> oh, this isn't ascii - it's keymap 16:14:47 * handsome_pirate is leaning +1, though 16:14:52 <handsome_pirate> Aye 16:14:54 <tflink> handsome_pirate: do you have any non-us keyboards, though? 16:15:13 <handsome_pirate> Keymap could screw up your password, badly 16:15:16 <seasc> i am swiss and using ch/sg 16:15:26 * handsome_pirate is +1 blocker 16:15:28 <handsome_pirate> tflink: No 16:15:28 * kparal still reading 16:15:30 <tflink> I feel like we've dealt witht his before, don't recall if we took them as blockers, though 16:16:03 <seasc> basic usage blocker? - user looses all his data because he cant uncrypt his luks because of wrong keymap? 16:16:14 <spstarr_work> well, that is bad.... 16:16:25 <seasc> and not an user error imho 16:16:32 <handsome_pirate> If your keymap is borked and suddenly becomes fixed, then that screws up your passwords since you can't see what you're typing 16:16:52 <handsome_pirate> +1 blocker 16:16:53 <kparal> it's definitely a final blocker. now let's look at the milestones criteria... 16:17:19 <tflink> for f18, a similar bug was rejected as final blocker: https://bugzilla.redhat.com/show_bug.cgi?id=875567 16:17:23 <kparal> well, the Alpha criterion says it's covered by Alpha 16:17:31 * kparal looking 16:17:39 <tflink> kparal: but luks works ... if you're using a en-us keyboard 16:17:53 * tflink isn't saying it isn't a bug, just not sure it's alpha blocker material 16:17:54 <spstarr_work> tflink: locale bias 16:18:11 <tflink> maybe 16:18:14 <kparal> tflink: but you can't decrypt it 16:18:15 <seasc> it should not be a criteria to force the user using us-en? 16:18:21 <pschindl> But if you can encrypt your disk with czech layout but you can't decrypt it, it is bug 16:18:23 <tflink> but we have to draw the line somewhere 16:18:31 <tflink> I didn't say it wasn't a bug 16:18:43 <tflink> I said I wasn't sure it was severe enough to block _alpha_ 16:18:50 <pschindl> Anaconda should reflect what is used during booting 16:19:18 <tflink> and if we've released with a similar bug in the past ... 16:19:23 <handsome_pirate> tflink: Bad juju to knowingly bork people's data 16:19:23 <spstarr_work> pschindl: though at this early stage is that 'ok' to let some people suffer until things are flushed out more? 16:19:24 <pschindl> tflink: I agree that it may not be alpha 16:19:37 <tflink> handsome_pirate: but it's _alpha_ 16:19:41 <spstarr_work> i mean it is locale bias otherwise 16:19:47 <tflink> it's not eating data 16:19:56 <handsome_pirate> tflink: And, there's an _alpha_ criteria covering it 16:20:07 <tflink> handsome_pirate: point me to the locale criteria for alpha 16:20:09 <kparal> "However, if you've configured the 'gb' layout for the installation process and also in the installer for the newly installed system and the LUKS passphrase entry on reboot uses the 'us' layout altough the kernel parameter is set correctly to 'vconsole.keymap=gb', feel free to reopen this bug, assign it to dracut and change summary to reflect that it is a dracut bug." 16:20:13 <kparal> https://bugzilla.redhat.com/show_bug.cgi?id=875567#c20 16:20:45 <kparal> so, vpodzime says that you need to change layout in the system _and_ in anaconda, and it should work 16:21:23 <tflink> can you change the keymap @ boot time post-install? 16:21:34 <kparal> probably by adjusting kernel boot line 16:21:51 <kparal> yeah 16:22:46 <spstarr_work> workaroundable... not alpha blocker.... of course user needs to know how to do it in the meantime... 16:22:48 <tflink> -1 blocker 16:22:56 <tflink> not sure about FE 16:23:10 <tflink> the workaround isn't bad ... if you know about it 16:23:23 <spstarr_work> -1 its like setting the kernel option to reduce memory usage when we use debugging in anaconda on a 1GB machine it fails if you dont set slub_debug=- for example in boot command during install 16:23:47 <kparal> we will need to check that the workaround works 16:24:09 <tflink> kparal: which workaround, changing locales post-install or configuring the live env pre-install? 16:24:21 <kparal> adjust the kernel boot line 16:24:31 <handsome_pirate> With workaround, I am -1/+1 for Alpha, +1 for Final 16:25:00 <kparal> I'd punt and test the workarounds. if they work, I agree with -1 Alpha 16:25:01 <spstarr_work> +1 for final for sure 16:25:15 <kparal> and +1 Final right away 16:25:53 <tflink> I'm not even sure about final, we've released with similar bugs before - not sure it passes the "last blocker at go/no-go" test 16:26:17 <tflink> thoughts on punt vs. -1? 16:26:21 <tflink> for alpha 16:26:23 <kparal> more investigation needed, I think 16:26:27 <spstarr_work> well, for final maybe we can fix it for good? :) 16:26:36 <spstarr_work> instead of punting it over and over? 16:26:49 <tflink> spstarr_work: sure, feel free to submit patches :) 16:26:54 <kparal> we released similar bugs, but IIRC those bugs were "console keymap != X keymap", which is a different cup of tea 16:26:55 <spstarr_work> heh 16:27:11 <spstarr_work> tflink: its easy for me to say that w/o code to back it up, sure :) 16:27:14 * jreznik is here, sorry for being a bit late, meetings... 16:27:25 <kparal> jreznik: welcome to yet another meeting! 16:27:26 <tflink> jreznik: welcome to the party! 16:27:31 <handsome_pirate> +1 Punt, +1 Final 16:27:42 <handsome_pirate> .moar bacon jreznik 16:27:42 <zodbot> here jreznik, have some more bacon 16:28:03 <kparal> actually let's just punt and test is properly 16:28:18 * tflink is -1 final right now - too early for that 16:28:32 <tflink> it sounds like we're mostly "punt" - any objections? 16:28:39 <tflink> who's going to do the investigation? 16:28:53 * kparal is going to find some victim 16:29:17 <spstarr_work> :) 16:29:23 <kparal> #action kparal to make sure 994180 is investigated 16:29:34 <tflink> action kparal to investigate (or get someone to investigate) workarounds listed in rhbz#994180 16:29:37 <tflink> oh 16:29:41 <tflink> i forgot the # anyways 16:30:34 <tflink> proposed #agreed 994180 - This doesn't seem like an alpha blocker with the existing workarounds but we want to make sure those workarounds actually _work_ before rejecting. Will revisit at the next blocker review meeting 16:30:52 <spstarr_work> (see punt) 16:30:54 <handsome_pirate> ack 16:30:55 <mkrizek> ack 16:30:58 <kparal> ack 16:31:25 <tflink> #agreed 994180 - This doesn't seem like an alpha blocker with the existing workarounds but we want to make sure those workarounds actually _work_ before rejecting. Will revisit at the next blocker review meeting 16:31:41 <tflink> #topic (1000715) TypeError: Argument 2 does not allow None as a value 16:31:44 <tflink> #link https://bugzilla.redhat.com/show_bug.cgi?id=1000715 16:31:47 <tflink> #info Proposed Blocker, anaconda, NEW 16:31:47 <spstarr_work> +1 16:32:45 <tflink> hrm, not sure this is the same bug pwhalen is hitting 16:33:04 <tflink> nvm, misreading stuff 16:33:18 <handsome_pirate> tflink: Same root cause 16:33:28 <tflink> handsome_pirate: but the fix doesn't work 16:33:49 <handsome_pirate> tflink: What fix? 16:34:00 <tflink> setting /etc/hostname 16:34:35 <pwhalen> creating the file didnt fix it, on reboot it hung 16:34:52 <handsome_pirate> Aye 16:35:14 <handsome_pirate> But, it seems that there is *something* going on here 16:35:22 <pschindl> isn't it same bug as https://bugzilla.redhat.com/show_bug.cgi?id=679486 ? 16:35:46 <handsome_pirate> And, arm hit it with initial setup because arm doesn't do Anaconda; x86 hits it at Anaconda and never gets to Initial Setup 16:35:49 <spstarr_work> with kickstart you can get it to work without /etc/hostname being there.... 16:36:07 <handsome_pirate> spstarr_work: Sort of 16:36:18 <spstarr_work> but i had to add a hostname to my DHCP IP address 16:36:24 <kparal> if this is really an /etc/hostname bug, X needs to be restarted to take that into account 16:36:37 <tflink> pschindl: the hostname fix doesn't work, though 16:37:23 <tflink> it feels like there's a blocker issue in there somewhere, but these reports also feel a bit muddled and/or confused 16:37:26 <spstarr_work> handsome_pirate: yeah, sort of 16:37:39 <kparal> so does this occur even without changing keymap/locale? 16:38:06 <handsome_pirate> kparal: Aye 16:38:13 <Viking-Ice> did not that fracking bug die already 16:38:32 <tflink> Viking-Ice: which one? 16:38:32 <kparal> handsome_pirate: what's the reproducer then? 16:38:38 <Viking-Ice> tflink, 679486 16:38:51 <tflink> Viking-Ice: sounds like the patch didn't make it past f19 16:39:34 <handsome_pirate> kparal: It's been done on x86 and arm 16:40:41 <kparal> LC_ALL=en_GB.UTF-8 LANG=en_GB.UTF-8 liveinst 16:40:42 <kparal> crashes 16:40:45 <kparal> reproduced 16:40:54 <tflink> kparal: but it requires setting lang? 16:41:12 <kparal> tflink: that's equivalent to changing language in system settings 16:41:23 <tflink> pwhalen: were you setting lang in your arm testing? 16:41:24 <kparal> which is the only supported way to install anaconda with a different lang/keymap 16:42:09 <pwhalen> tflink, no - this is a precanned image 16:42:47 <tflink> thoughts on blockery-ness? I'm not all that thrilled about accepting this bug as an alpha blocker as is 16:43:03 <spstarr_work> i cant do my tests .. 16:43:03 <handsome_pirate> tflink: It's def a blocker 16:43:05 <spstarr_work> it blocks me 16:43:20 <spstarr_work> not without workaroundery 16:43:23 <handsome_pirate> tflink: It's borking the install 16:43:34 <pwhalen> and locks folks out of the arm images, requires they mount to edit passwd 16:43:34 <tflink> spstarr_work: sure, the issue you're hitting sounds like a blocker 16:43:48 <tflink> this report is just so vague, I'm not liking the idea of accepting it 16:43:51 <spstarr_work> actually let me power on the VM remotely and remove the DNS host... 16:43:59 * handsome_pirate is +1 blocker 16:44:01 <spstarr_work> it should fail with it set or not 16:44:02 <tflink> I suppose we could just reject it later 16:44:14 <tflink> if it turns out to be a different issue 16:44:23 <tflink> other thoughts? 16:45:29 <kparal> so, actually, I can add additional keymaps in anaconda 16:45:42 <kparal> I just can't use them for installation (e.g. for passwords) 16:45:56 <kparal> do we have any locale criteria at all? we don't, do we? 16:46:02 <tflink> kparal, Viking-Ice: thoughts on bocker for this bug? 16:46:12 <Viking-Ice> +1 16:46:14 <tflink> kparal: are we sure it's a locale issue? 16:46:33 <tflink> not everyone is using the liveinst or changing locale 16:46:56 <Viking-Ice> we need to kill this bug 16:46:57 <spstarr_work> handsome_pirate: doing a quick sanity tet ip=dhcp and DHCP with IP having a host assigned... no kickstart... if crashes its for sure a +1 16:46:57 <kparal> I don't really care about the underlying issue. it impacts anyone non-US who wants to have special characters in password/some other text field 16:47:07 <kparal> which is not _that_ problematic for Alpha 16:47:14 <tflink> ? 16:47:21 <tflink> isn't that the last bug? 16:47:21 <handsome_pirate> kparal: You're on the previous bug 16:47:25 <tflink> or am I missing something 16:47:39 * kparal is talking about 1000715 16:47:55 <handsome_pirate> Wait, now I'm confused 16:48:07 * handsome_pirate fails to see how locale has *anything* to do with this 16:48:12 <spstarr_work> handsome_pirate: crashed 'An unknown error has occured' 'TypeError: Argument 2 does not allow None as value' 16:48:24 <handsome_pirate> kparal: No passwords are involved 16:48:33 <spstarr_work> it has nothing to do with 1000715 16:48:40 <kparal> once again, this crash occurs if you run liveinst with other than en_US locale 16:48:53 <tflink> are you guys sure that the issue you're talking about is the one originally reported as 1000715? 16:48:58 <kparal> that means, you can't use your native keymap for passwords and the like 16:48:59 <Viking-Ice> for the first we *always* seem to be hitting locale bugs in alpha 16:49:03 <tflink> spstarr_work, handsome_pirate ^^ 16:49:08 <Viking-Ice> someone has to just starting to talk about this with the anaconda team 16:49:10 <spstarr_work> append initrd=initrd-x86_64.img ip=dhcp inst.stage2=ftp://172.18.2.4/os-x86_64 slub_debug=- <----- 16:49:21 * jreznik is lost now... 16:49:21 <tflink> I'm wondering if the hostname stuff isn't actually a dupe 16:49:30 <spstarr_work> tflink: locale didnt change anything I didnt see any locale defaults 16:49:31 <tflink> this is starting to make more sense 16:49:39 <handsome_pirate> kparal: This crash occurs no matter the keymap 16:49:48 <tflink> handsome_pirate: no, i don't think it does 16:50:02 <Viking-Ice> handsome_pirate, any non us keyboard 16:50:02 <tflink> handsome_pirate: the issue you guys duped to this does, sure 16:50:04 * handsome_pirate is also confused now 16:50:04 <kparal> handsome_pirate: what what I've read, it occurred when somebody tried to change his locale 16:50:25 <tflink> handsome_pirate: I don't think that the issue you and spstarr_work are talking about is _this_ but 16:50:28 <tflink> bug 16:50:32 <spstarr_work> can I set the locale in dracut options/grub to confirm it is locale triggered causing /etc/hostname not to be created? 16:50:53 <tflink> spstarr_work: what makes you think the two issues are related? 16:51:00 <spstarr_work> i dont think they are 16:51:14 <kparal> this doesn't seem to be related to hostname 16:51:18 <spstarr_work> HOWEVER 16:51:35 <spstarr_work> in my kickstart i have 'lang en_US.UTF-8' 16:51:49 <spstarr_work> if we discount anything else 16:52:00 <kparal> let's not discuss kickstarts here. this is Live image 16:52:02 <tflink> spstarr_work: was c#22 abrt? 16:52:05 <spstarr_work> that worked but i cannot tell if its due to kickstart code path or locale set 16:52:35 <kparal> ah, I see, boot.iso as well. ok 16:52:37 <tflink> let's leave the hostname stuff for later triage and focus on the original locale issue 16:52:44 <spstarr_work> kparal: yes and PXE now also 16:52:53 <tflink> if they end up being the same issue, that can be handled later 16:53:26 <spstarr_work> +1 and if its not then -1 later anyway, but something is blocking me from testing 16:53:48 <tflink> the original issue is that liveinst crashes when non-en lang is selected before running, right? 16:53:48 <kparal> so, once again, this bug 1000715 is affecting anyone who would like to use non-us keymap during the _installation_ 16:53:51 <spstarr_work> i will now go off and see if locale has anything to do with 1000715 and report in channel later 16:54:48 * handsome_pirate knows now what happened 16:54:53 * handsome_pirate screwed up 16:54:57 <spstarr_work> kparal: is locale/lang en_US.UTF-8 != non-us keymap ? 16:55:14 <tflink> the bug sounds like: if you're using or set a non-en-us keyboard during liveinst, the install crashes before welcome screen 16:55:19 <handsome_pirate> I saw that the same error message was being emmitted and did not read through the comments as well as I should have 16:55:21 <tflink> am I understanding correctly? 16:55:28 * handsome_pirate puts on the Hat-O-Shame 16:55:30 <kparal> tflink: yes 16:55:36 <tflink> ok, in that case +1 blocker 16:55:41 <handsome_pirate> kparal: tflink: Apologies 16:56:01 <tflink> .fire handsome_pirate 16:56:01 <zodbot> adamw fires handsome_pirate 16:56:08 <tflink> :-D 16:56:12 <tflink> other votes? 16:56:17 <spstarr_work> +1 16:56:30 <Viking-Ice> +1 16:56:37 * handsome_pirate goes to do some quick bugzilla-foo to fix his screw up 16:56:39 <handsome_pirate> Oh 16:56:40 <handsome_pirate> +1 16:56:42 <kparal> +0 16:56:58 <pwhalen> +1 16:57:01 <tflink> proposed #agreed 1000715 - AcceptedBlocker - Violates the following F20 alpha release criteria for installs using a non-en-us keyboard: "The installer must run when launched normally from the release-blocking images." 16:57:11 <Viking-Ice> ack 16:57:22 <tflink> kparal: +0? 16:57:29 <handsome_pirate> ack 16:57:49 <tflink> the argument being that we don't need non-en-us keyboards working, or did I miss a workaround in all the mess? 16:57:53 <kparal> tflink: I don't think this is a widely used feature 16:58:08 <kparal> I don't see a workaround 16:58:26 <kparal> you can configure the system to use en_GB by default, but the installation has to be performed in en_US 16:58:40 <spstarr_work> i note 1000715 is not the liveinst one... 16:58:48 <kparal> but since we almost rejected locale-related issues in LUKS for Alpha, accepting this won't make much difference 16:58:48 <spstarr_work> that one is 679486......... 16:58:56 <Viking-Ice> so you have to revert to en_us before you start install right 16:58:56 <tflink> yeah, that's a good point 16:59:00 <spstarr_work> you might want to fix that 16:59:07 <tflink> Viking-Ice: pretty much 16:59:26 <tflink> but since we didn't take the last luks-in-non-en-us bug as a blocker ... 16:59:45 <spstarr_work> we were discussing 1000715 = 'TypeError: Argument 2 does not allow None as a value' .... 679486 == 'Liveinst doesn't start of hostname changes' 16:59:53 <handsome_pirate> Okay, I did some quick bugzilla-foo 17:00:21 <spstarr_work> so I am +1 for 1000715 for the OTHER issue 17:00:48 <kparal> spstarr_work: I don't know who started talking about hostname, but it's not relevant to this bug 17:01:05 <handsome_pirate> .bug 995268 17:01:09 <zodbot> handsome_pirate: Bug 995268 TypeError: Argument 2 does not allow None as a value - If no hostname is set by DHCP - https://bugzilla.redhat.com/show_bug.cgi?id=995268 17:01:22 <tflink> stop it, we're not talking about that bug right now 17:01:24 <handsome_pirate> I just reopened that one for the hostname issue and set it as a blocker 17:01:30 <tflink> this has been confusing enough as it is 17:01:32 <spstarr_work> ahh thank you 17:01:35 <spstarr_work> now im unconfused 17:01:37 <spstarr_work> :D 17:01:42 <pwhalen> yea, wow 17:01:54 <handsome_pirate> tflink: Just reporting what I did 17:02:19 <tflink> just so that we're clear, we're talking about the liveinst keymap issue here :) 17:02:28 <spstarr_work> tflink: +1 for the liveinst keymap issue 17:02:29 <kparal> I think this should get accepted if LUKS-with-gb-keymap gets accepted. otherwise, /me shrugs 17:02:47 <spstarr_work> well... 17:02:50 <spstarr_work> kparal: you are right on that 17:02:59 <spstarr_work> it is exactly locale specific 17:03:01 <spstarr_work> so -1 17:03:15 <tflink> but it's a crasher vs. having to change locale post-install 17:03:29 <spstarr_work> tflink: is it workaroundable if not then +1 17:03:31 <spstarr_work> if it is... -1 17:03:43 <tflink> well, you can install w/ en-us keymap 17:03:47 <Viking-Ice> accept them both sas I said before we always have locale issue ( non us related ) with the installer @ alpha and have been having those kind of problems during alpa for as long as I can remember and we should not have to deal with these issues again and again 17:03:57 <spstarr_work> which is the same with the LUKS keymap bug... 17:04:00 <kparal> true, this one can't be worked around. but I don't see why anyone would like to do so, without LUKS 17:04:19 * Viking-Ice propose to outsource anaconda to brno 17:04:22 <kparal> if we _had_ some locale criteria, I'd fine with having them affect Beta 17:04:34 <Viking-Ice> then we never should have non us locale issues again ;) 17:04:36 <tflink> Viking-Ice: yeah, it'd be nice to have fewer of these repeat-feeling locale bugs, but do we usually take locale issues as alpha blockers? 17:04:39 <handsome_pirate> Viking-Ice: LOL 17:04:46 <spstarr_work> maybe we should make that a new criteria item then :) 17:04:47 <kparal> Viking-Ice: many of the developers are in Brno 17:04:54 <Viking-Ice> tflink, no we dont we always let them slide to beta 17:04:59 <Viking-Ice> ( or usually ) 17:06:14 <Viking-Ice> but as I said we should not have to be dealing with this stuff repeatedly 17:06:36 <kparal> so, who wants to propose a locale criterion? 17:07:03 <tflink> would we have support from devs if we did have a locale criterion? 17:07:15 <Viking-Ice> ? 17:07:24 <kparal> maybe we would get some input on that 17:07:28 <tflink> unfunded mandates are unlikely to succeed 17:07:29 <Viking-Ice> bit odd for us to have them dictate the criteria right 17:07:36 <kparal> come on, let's move. no workaround, +1 17:07:48 <spstarr_work> +1 for no workaround 17:07:48 <kparal> we can re-assess if we have new info 17:07:52 <tflink> proposed #agreed 1000715 - AcceptedBlocker - Violates the following F20 alpha release criteria for installs using a non-en-us keyboard: "The installer must run when launched normally from the release-blocking images." 17:07:52 <Viking-Ice> however the locale criteria could be a bit of pandora box 17:07:58 <handsome_pirate> ack 17:08:00 <kparal> ack 17:08:06 <Viking-Ice> ack 17:08:20 <tflink> #agreed 1000715 - AcceptedBlocker - Violates the following F20 alpha release criteria for installs using a non-en-us keyboard: "The installer must run when launched normally from the release-blocking images." 17:08:43 <tflink> any volunteers to look into proposing a locale criterion? 17:09:08 <spstarr_work> wouldn't that locale be weighted though? 17:09:18 <kparal> spstarr_work: sure 17:09:56 * tflink takes that as a 'no volunteers' 17:10:24 <tflink> moving on 17:10:34 <Viking-Ice> tflink, I'm not touching that hot potato ( which country to exclude ) 17:10:47 <Viking-Ice> ;) 17:10:56 <spstarr_work> heh, i am en-us bias 17:11:03 <tflink> #info there was some discussion around whether we should have locale criterion or not but more investigation would be needed before any propoal went forward 17:11:08 <tflink> Viking-Ice: yeah, agreed on that one 17:11:15 <tflink> #topic (1002055) [abrt] gnome-shell-3.9.90-1.fc20: g_variant_valist_new_nnp: Process /usr/bin/gnome-shell was killed by signal 5 (SIGTRAP) 17:11:18 <tflink> #link https://bugzilla.redhat.com/show_bug.cgi?id=1002055 17:11:20 <tflink> #info Proposed Blocker, gnome-shell, NEW 17:12:03 <kparal> tldr: TC1 i386 Live is DEAD 17:12:08 <kparal> doesn't work even in VM 17:12:19 <kparal> gnome-shell doesn't start 17:12:25 <tflink> sounds like a pretty clear blocker 17:12:31 <handsome_pirate> +1 blocker 17:12:32 <Viking-Ice> yup 17:12:32 <mkrizek> +1 blocker 17:12:38 <kparal> +1 17:12:39 <ignatenkobrain> +1 17:13:17 <jreznik> +1 blocker 17:13:25 <pschindl> +1 17:13:29 <tflink> proposed #agreed 1002055 - AcceptedBlocker - Violates the following F20 alpha release criterion for the 32bit desktop spin: "Release-blocking live images must boot to the expected boot menu, and then to a desktop or to a login prompt where it is clear how to log in to a desktop." 17:13:33 <mkrizek> ack 17:13:41 <pschindl> ack 17:13:44 <kparal> ack 17:13:56 <Viking-Ice> ack 17:13:57 <tflink> #agreed 1002055 - AcceptedBlocker - Violates the following F20 alpha release criterion for the 32bit desktop spin: "Release-blocking live images must boot to the expected boot menu, and then to a desktop or to a login prompt where it is clear how to log in to a desktop." 17:14:02 <tflink> #topic (1001081) anaconda unable to select INSTALLATION SOURCE (Exception yum.Errors.RepoError...) 17:14:05 <tflink> #link https://bugzilla.redhat.com/show_bug.cgi?id=1001081 17:14:08 <tflink> #info Proposed Blocker, pykickstart, MODIFIED 17:14:23 <tflink> another clear blocker, should be fixed in TC2 17:14:25 <tflink> +1 17:14:30 <spstarr_work> reading... 17:14:32 <mkrizek> +1 17:14:51 <tflink> proposed #agreed 1001081 - AcceptedBlocker - Violates the following F20 alpha release criterion: "When using the DVD image, the installer must be able to use the DVD as a package source." 17:15:00 <spstarr_work> +! 17:15:03 <spstarr_work> +1 17:15:07 <Viking-Ice> ack 17:15:12 <mkrizek> ack 17:15:13 <handsome_pirate> ack 17:15:17 <pschindl> ack 17:15:18 <pwhalen> ack 17:15:20 <tflink> #agreed 1001081 - AcceptedBlocker - Violates the following F20 alpha release criterion: "When using the DVD image, the installer must be able to use the DVD as a package source." 17:15:30 <tflink> #topic (679486) Liveinst doesn't start if hostname changes 17:15:30 <tflink> #link https://bugzilla.redhat.com/show_bug.cgi?id=679486 17:15:30 <tflink> #info Proposed Blocker, xorg-x11-xauth, ASSIGNED 17:15:41 <pschindl> +1 blocker 17:15:46 <spstarr_work> +1 17:15:50 <spstarr_work> if no workaround 17:15:50 <pschindl> this is the same think we had in f19 17:16:02 <tflink> as a final blocker, sure 17:16:27 <tflink> +1 FE, not as sure about alpha blocker 17:16:29 <spstarr_work> actually .. why would hostname 'change' 17:16:33 <spstarr_work> -1 17:16:33 * jreznik already saw dgilmore's commit 17:16:43 <tflink> spstarr_work: certain dhcp environments 17:16:55 <jreznik> so +1 FE as it should be fixed already for the next TC 17:16:57 <kparal> this bug occurs whenever you change a network connection 17:17:00 <Viking-Ice> -1 17:17:02 <Viking-Ice> for alpha 17:17:05 <kparal> so for example you disable wifi and enable wired 17:17:07 <spstarr_work> tflink: more edge/corner case though? unless dhcp lease expiry more often 17:17:07 <tflink> Viking-Ice: FE? 17:17:12 <Viking-Ice> sure 17:17:16 <spstarr_work> its still a lesser case 17:17:28 <spstarr_work> -1 for alpha 17:17:30 <tflink> I'm seeing mostly -1/+1 17:17:36 * kparal thinking 17:17:49 <handsome_pirate> -1/+1 17:18:00 <spstarr_work> i have to think of who will use this DURING ALPHA how many people will be impacted? in % 17:18:10 <spstarr_work> if its high then +1 if its quite low then -1 til final 17:18:19 <dgilmore> spstarr_work: hostname if not set comes up as loclahost, if you have reverse dns up and dhcp gets a ip hostname changes 17:18:32 <kparal> spstarr_work: anyone in RH office, since it also affects dhcp hostnames :) 17:18:37 <spstarr_work> hmm 17:18:48 <dgilmore> i have working reverse dns here at home 17:19:12 <spstarr_work> but will this impact the average user however? 17:19:20 <spstarr_work> at least in the early alpha 17:19:29 <kparal> do average users use Alpha? :) 17:19:34 <spstarr_work> lol 17:19:35 <spstarr_work> :) 17:19:39 <tflink> more likely to affect the average alpha user than the average final user, I'd say 17:19:43 <kparal> anyone, I'm +0 here, +1 FE 17:19:47 <kparal> *anyway 17:20:05 <spstarr_work> I am leaning +0 then 17:20:11 <dgilmore> whats an average user 17:20:14 <kparal> you can't all be +0! 17:20:20 <dgilmore> +1 blocker 17:20:22 <kparal> that's my number! 17:20:22 <tflink> proposed #agreed 679486 - RejectedBlocker AcceptedFreezeException - This is a reoccurance of a bug that was accepted as final blocker for F19 but the fix didn't make it into F20. It doesn't violate any F20 alpha release criteria but a tested fix would be considered after freeze. 17:20:28 <handsome_pirate> ack 17:20:30 <dgilmore> ack 17:20:32 <pschindl> ack 17:20:41 <kparal> ack 17:20:45 <tflink> #agreed 679486 - RejectedBlocker AcceptedFreezeException - This is a reoccurance of a bug that was accepted as final blocker for F19 but the fix didn't make it into F20. It doesn't violate any F20 alpha release criteria but a tested fix would be considered after freeze. 17:20:58 <tflink> OK, that's all of the proposed blockers for today 17:21:06 <spstarr_work> you might want to refresh page... 17:21:06 <tflink> there are no proposed FEs 17:21:10 <spstarr_work> the OTHER issue came back 17:21:41 <spstarr_work> 995268 17:23:12 <spstarr_work> im trying to debug that one now... as its a blocker 17:23:47 <tflink> #topic (995268) TypeError: Argument 2 does not allow None as a value - If no hostname is set by DHCP 17:23:50 <tflink> #link https://bugzilla.redhat.com/show_bug.cgi?id=995268 17:23:52 <tflink> #info Proposed Blocker, anaconda, ASSIGNED 17:24:15 <tflink> handsome_pirate: seriously, you don't know better than to propose blockers without citing criteria? 17:24:22 <spstarr_work> but i dont think it is DHCP releated, as i just triggered it now its trying to use locale: zu_ZA.UTF-8 by DEFAULT 17:24:48 <tflink> interesting, what happens when you use en_us? 17:24:51 <spstarr_work> we are doing auto locale discovery now? 17:24:59 <spstarr_work> tflink: i wonder if thats why my kickstart works 17:25:10 <spstarr_work> i set lang explicitly to en_US.UTF-8 17:25:19 <Viking-Ice> spstarr_work, you mean the peski geoip crap that really does not belong in installer mayber 17:25:24 <Viking-Ice> mean maybe 17:25:29 <spstarr_work> hehehe 17:25:36 <kparal> it can be connected to geoip, right 17:25:36 <spstarr_work> my machine at home is in Canada it SHOWS that 17:25:46 <spstarr_work> but its not properly finding something(?) 17:25:50 * handsome_pirate gives up 17:26:11 <tflink> LANG=en_CA.UTF-8 17:26:19 <spstarr_work> tflink: fails in grub 17:26:24 <kparal> spstarr_work: so, the reproducer is to use boot.iso with kickstart? 17:26:51 <spstarr_work> PXE it fails, boot.iso it fails... let me check error output to see if it is same stack... 17:26:59 * Viking-Ice cant for the love of all beer understand why team anaconda added geoip crap to the installer 17:27:20 <Viking-Ice> someone was not drunk enought when they decided that 17:27:30 <handsome_pirate> dcantrell: ^^ 17:27:41 <spstarr_work> testing 17:28:32 <spstarr_work> but this is till a blocker of some sort... because i keep crashing anaconda at the very start 17:28:38 <spstarr_work> regardless of whats triggering it 17:28:42 <tflink> not really 17:28:47 <kparal> spstarr_work: does it happen without a kickstart? 17:29:01 <spstarr_work> kparal: without a kickstart 100% 17:29:07 <pwhalen> this issue seems to prevent initial-setup-graphical from running as well 17:29:09 <spstarr_work> with one.. i can get around this 17:29:37 <spstarr_work> booting boot.iso now w/o any options just default... 17:30:21 <kparal> spstarr_work: 100% failure rate? 17:30:28 <spstarr_work> yes 17:30:43 <spstarr_work> X started now... anaconda trying to load... 17:31:00 <spstarr_work> failefd 17:31:07 * kparal booting x86_64 netinst 17:31:15 <spstarr_work> in a KVM 17:31:33 <spstarr_work> TypeError: Argument 2 does not allow None as a value: 17:31:38 <spstarr_work> Locale variables in innermose frame: 17:31:39 <kparal> works for me 17:31:47 <spstarr_work> locale: ZU_ZA.UTF-8 .. 17:31:52 <kparal> I guess it must be geoip stuff then 17:31:56 <spstarr_work> must be 17:32:07 <kparal> +1 blocker 17:32:19 <pwhalen> +1 17:32:31 <spstarr_work> which is why kickstart works since it forces locale 17:32:31 <kparal> sadly spstarr_work has to be the one to reproduce and test the fix :) 17:32:38 <spstarr_work> :) 17:32:39 <spstarr_work> sure 17:32:48 <spstarr_work> can we rename bug please? 17:32:57 <spstarr_work> its not DHCP or hostname issue 17:33:04 <kparal> you can 17:33:04 <tflink> spstarr_work: can't you rename it? 17:33:09 <spstarr_work> or i can ammend yeah i'll ammend the title 17:33:19 <dcantrell> win 31 17:33:31 <kparal> dcantrell: ? 17:33:44 <tflink> kparal: irssi command, I think 17:33:49 <tflink> just missing /win 17:33:49 <kparal> ah :) 17:33:51 <dcantrell> fatfingered 17:33:52 <dcantrell> yeah 17:34:24 <tflink> proposed #agreed 995268 - AcceptedBlocker - Violates the following F20 alpha release criteria for non-en-us locales or improperly detected locales: "Release-blocking dedicated installer images must boot to the expected boot menu, and then after a reasonable timeout to the installer." 17:34:28 <Viking-Ice> ack 17:34:38 <pschindl> ack 17:34:43 <spstarr_work> ack 17:35:04 <tflink> #agreed 995268 - AcceptedBlocker - Violates the following F20 alpha release criteria for non-en-us locales or improperly detected locales: "Release-blocking dedicated installer images must boot to the expected boot menu, and then after a reasonable timeout to the installer." 17:35:17 <tflink> ok, i think that's all the proposed blockers ... for reals this time :) 17:36:05 <spstarr_work> interestingly it shows a generated kickstart with 17:36:09 <spstarr_work> lang EN_CA.UTF-8 17:36:16 <spstarr_work> network --hostname=localhost.localdomain ... 17:36:27 <spstarr_work> but even with localhost as hostname its still a hostname set 17:36:28 <tflink> there are no proposed FE, so moving on to the accepted blockers 17:36:57 <tflink> #topic (997690) SizeNotPositiveError: bytes= param must be >=0 17:36:58 <tflink> #link https://bugzilla.redhat.com/show_bug.cgi?id=997690 17:36:58 <tflink> #info Accepted Blocker, anaconda, POST 17:37:59 <tflink> I don't see any huge progress, but it does appear to be on the anaconda devs' radar 17:38:36 <tflink> #info no overt progress inbug, new duplicate found, no action needed from us at this time 17:38:48 <tflink> #undo 17:38:48 <zodbot> Removing item from minutes: <MeetBot.items.Info object at 0x2639ca10> 17:39:03 <tflink> #info patch submitted, waiting for merge and new anaconda build 17:39:16 * tflink is actually reading and paying attention :) 17:39:19 <Viking-Ice> well I think with the suggested patch that bug is actually fixed ( but another has cropped up ) based on #17 17:39:22 <tflink> any other things? 17:39:45 <handsome_pirate> not from me 17:40:04 <tflink> Viking-Ice: it went to POST after c#17, though. i wonder if it's the same patch 17:40:52 <tflink> and that is in custom partitioning, which is not an alpha blocker 17:41:04 <tflink> it is probably the same patch, though 17:41:26 <tflink> but that's pretty much what you said, so moving on :) 17:41:41 <tflink> #topic (983319) Install gets stuck returning from Installation Destination spoke 17:41:44 <tflink> #link https://bugzilla.redhat.com/show_bug.cgi?id=983319 17:41:46 <tflink> #info Accepted Blocker, anaconda, MODIFIED 17:42:05 <tflink> I suspect this can be closed 17:42:23 <tflink> as I understand it, the blocking part of this issue has been fixed 17:43:05 <pschindl> I think so. 17:43:27 * kparal mutters something about too much text 17:43:28 <tflink> #info this has been fixed in anaconda-20.8-1 (F20 Alpha TC1) and can be closed 17:43:42 <tflink> kparal: ? 17:43:54 <kparal> too much text to read :) 17:44:02 <kparal> in those reports 17:44:24 <tflink> yeah, but we're almost done 17:44:31 <tflink> anything else on this bug? 17:45:16 <Viking-Ice> I guess not 17:45:20 <tflink> #topic (985342) illegal instructions with glibc-2.17.90 on armv7hl 17:45:21 <pschindl> tflink: I think it's fixed in 20.9 17:45:23 <tflink> #link https://bugzilla.redhat.com/show_bug.cgi?id=985342 17:45:26 <tflink> #info Accepted Blocker, glibc, MODIFIED 17:45:40 <tflink> pschindl: it's the bug we were waiting on a fix for for TC1, right? 17:45:56 <pwhalen> this is fixed with the latest version 17:46:01 <tflink> it's marked as fixed in 20.8-1 17:46:13 <tflink> ^^ the previous bug 17:46:18 <pschindl> no. It was this one: https://bugzilla.redhat.com/show_bug.cgi?id=998687 17:46:22 <pschindl> I think 17:46:39 <tflink> pschindl: 983319 is marked as fixed in anaconda-20.8-1 17:46:42 <handsome_pirate> tflink: I think it's fixed 17:46:49 <handsome_pirate> tflink: the glibc bug 17:47:08 <tflink> pschindl: do I need to go back? 17:47:09 <handsome_pirate> tflink: That's the fix pwhalen was wanting in RC2 17:47:16 <pschindl> tflink: You are right 17:47:39 <tflink> ok, just making sure I wasn't missing something 17:48:19 <tflink> #info there is a new build that should fix this issue on the affected armv7hl platforms 17:48:44 <pwhalen> build is now finished 17:48:52 <tflink> #info glibc-2.18-5.fc20 will be in TC2, will need testing to verify 17:49:09 <pschindl> not sure if it make to TC2 the build was made lately 17:49:32 <pschindl> dgilmore: did new glibc make it to TC2? 17:49:35 <tflink> #undo 17:49:35 <zodbot> Removing item from minutes: <MeetBot.items.Info object at 0x25fa9890> 17:49:37 <tflink> #info glibc-2.18-5.fc20 will be in TC2 or TC3, will need testing to verify 17:49:39 <pwhalen> I ping dgilmore, he hadnt started on the request, left the comment to ensure 17:50:22 <tflink> the last accepted blocker is a repoclosure auto-blocker, don't think it needs discussion yet 17:50:33 <tflink> which brings us to ... 17:50:34 <pschindl> it should be solved 17:50:38 <tflink> #topic Open Floor 17:50:49 <handsome_pirate> Yay 17:50:53 <tflink> Anything else that should be covered in-meeting? 17:51:22 <spstarr_work> :) 17:51:32 <Viking-Ice> dcantrell, what does it take for us to stop having to deal with locale issue at each development cycle? 17:51:42 <spstarr_work> handsome_pirate: it is locale related my issue, i confirm it in fact worse, setting to Arabic it doesnt even respect that in ks.cfg but it does not crash now 17:51:52 <spstarr_work> Viking-Ice: why did Anaconda 19 -> 20 break locales so badly now?? 17:51:52 <tflink> Viking-Ice: stop supporting non-en-us? 17:51:53 <Viking-Ice> let me rephrase how many bottles of brandy and whiskey does it take ;) 17:51:55 <spstarr_work> what did we do? 17:51:57 <dgilmore> pschindl: it did make it in 17:52:06 <tflink> :-D 17:52:07 <jreznik> just general reminder - alpha change deadline is coming in less than one week (sep 03) 17:52:11 <pschindl> dgilmore: cool, thx :) 17:52:24 <tflink> #info Alpha Change Deadline is 2013-09-03 17:52:35 <tflink> and they'll turn bodhi on and we can test the new blocker tracking app! 17:52:40 <pwhalen> I was on a call for the majority of this meeting, not sure the bug I hit in initial-setup is the same as the other mentioned - seems to be the missing hostname file preventing it from working 17:52:41 <dcantrell> Viking-Ice: just get everyone using english 17:52:52 <kparal> what an idea 17:53:15 <Viking-Ice> dcantrell, is wwoods hogging all that moonshine? surely he can slip the rest of the team some ;) 17:53:38 <pschindl> What about Esperanto? :) 17:53:58 <tflink> #info the next blocker review meeting will be 2013-09-04 @ 16:00 UTC in #fedora-blocker-review 17:54:14 <Viking-Ice> well we vikings have a saying English is to conquer not learn 17:54:38 <tflink> if there's nothing else ... 17:54:54 * Viking-Ice lits the quantum fuse 17:55:00 * tflink sets the patent-pending non-deterministic fuse for (0,5] minutes 17:55:05 <jreznik> fyi I'm going to be out for a few days, back on Tuesday 17:55:19 <Viking-Ice> same here archery tournament in sweden 17:55:21 <tflink> jreznik: cool, enjoy your time off! 17:55:31 <tflink> Viking-Ice: that sounds like fun 17:55:38 * tflink will also be gone fri-mon 17:55:39 <jreznik> thanks :) 17:56:00 <tflink> Thanks for coming everyone, we had a great turn out today! 17:56:07 * tflink will send out minutes shortly 17:56:11 <spstarr_work> :) 17:56:13 <tflink> #endmeeting