[00:00:13] <cpu> agl: sweet
[00:00:18] *** MikeSmithXX has quit IRC
[00:00:29] *** isherman has joined #chromium
[00:00:29] *** ChanServ sets mode: +v isherman
[00:01:15] <agl> cpu: the previous suppression omitted some functions from the bottom. Is that valid?
[00:01:19] *** isherman has quit IRC
[00:01:28] <agl> cpu: (i.e. fun:_ZN54_GLOBAL__N_base_worker_pool_linux.cc_00000000_67754AA812WorkerThread10ThreadMainEv)
[00:01:39] *** Venom_X has quit IRC
[00:01:53] <jamesr> you only have to match the top of the stack in a valgrind suppression
[00:02:01] <agl> jamesr: cool, thanks.
[00:02:04] <cpu> agl: no clue, try and see
[00:02:10] <jamesr> there's a description of the syntax somewhere
[00:02:17] *** isherman has joined #chromium
[00:02:17] *** ChanServ sets mode: +v isherman
[00:02:19] <cpu> agl: I mean try minimal set and take it from there
[00:02:27] *** csilv has quit IRC
[00:04:15] <agl> cpu: http://codereview.chromium.org/2832092
[00:05:47] <cpu> looks like sargrass broke the tree
[00:05:53] *** hclam has joined #chromium
[00:05:58] *** hclam_ has joined #chromium
[00:06:27] <evmar_sheriff> cpu: i tried to ping them
[00:06:32] <stuartmorgan> He's afk
[00:06:33] <evmar_sheriff> cpu: i'm tempted to revert because i see no try job
[00:06:36] <stuartmorgan> so just revert
[00:06:41] <cpu> yes revert
[00:07:11] <evmar_sheriff> reverted
[00:07:33] <stuartmorgan> He had a try job for a recent revision; I'm guessing it was a bad merge
[00:07:52] <evmar_sheriff> it's a missing file i think
[00:07:56] <stuartmorgan> Since chromeos_accounts_add_user_overlay.html wasn't part of his change IIRC
[00:07:57] *** isherman has quit IRC
[00:08:22] <cpu> agl: done
[00:08:32] *** thakis__ has quit IRC
[00:08:40] <cpu> agl: you can check in
[00:08:44] *** tonyg-cr has quit IRC
[00:08:49] <stuartmorgan> evmar_sheriff: I think that include came from a merge, as opposed to being intended as part of his change
[00:09:14] *** lindasexihot has quit IRC
[00:09:16] <agl> cpu: done
[00:10:05] *** fta has quit IRC
[00:10:18] *** monreal has quit IRC
[00:11:37] <cpu> agl: thanks
[00:11:57] <cpu> evmar_sheriff: I went to his code review to do the same :)
[00:12:26] *** nebula has joined #chromium
[00:13:38] *** alokp has quit IRC
[00:14:39] <trungl-bot> Tree closed by cpu at google dot com: Tree is closed (revert is in |  continuing issues: http://bit.ly/bsyE8W)
[00:15:40] *** hrna has quit IRC
[00:16:46] <cpu> ice cream run, brb
[00:21:26] *** FullFlannelJacke has quit IRC
[00:22:54] *** FullFlannelJacke has joined #chromium
[00:26:40] *** markmentovai has quit IRC
[00:28:28] *** kjg has quit IRC
[00:29:53] *** awolfson has quit IRC
[00:32:27] <nickcarter_> evmar: successful builds on all three platforms, is the tree green enough to reopen?
[00:32:39] <evmar_sheriff> sure
[00:32:51] <trungl-bot> Tree opened by evan at chromium dot org (:evmar): Tree is open (continuing issues: http://bit.ly/bsyE8W)
[00:32:59] *** ebraminio has quit IRC
[00:35:01] *** dumi has joined #chromium
[00:35:01] *** ChanServ sets mode: +v dumi
[00:37:39] *** [Nef] has quit IRC
[00:41:59] *** tedoc2000 has quit IRC
[00:42:58] *** jrmuizel has quit IRC
[00:47:12] *** tedoc2000 has joined #chromium
[00:48:22] *** tedoc2000 has quit IRC
[00:49:20] *** js2 has quit IRC
[00:49:49] *** agl has quit IRC
[00:50:10] *** homata_ has joined #chromium
[00:51:00] *** SogarGofin has joined #chromium
[00:51:21] <SogarGofin> Hello all. Is anyone available for a quick question regarding the IPC of chrome?
[00:52:05] *** phajdan-jr is now known as phajdan-jr|afk
[00:52:40] <jamesr> SogarGofin: probably better to just ask your question
[00:53:33] *** nebula has quit IRC
[00:53:44] *** [Nef] has joined #chromium
[00:54:47] <SogarGofin> heh ok. I'm wondering if I can get any advice on what Chrome does to solve the problem of "how to deep copy from other processes"
[00:55:07] <evmar_sheriff> SogarGofin: i don't know what that means
[00:55:10] <jamesr> SogarGofin: we have explicit marshalling for the types we want to send across processes. we don't support sending arbitrary object trees
[00:55:35] *** ctruta has quit IRC
[00:56:17] <eglaysher> SogarGofin: this is actually the region of code I'm working on right now. You'll want to take a look at (for example) chrome/common/render_messages_internal.h for an example of what our macros that define our marshaling classes look like.
[00:56:18] <SogarGofin> Ok, that's interesting. Does that mean that you have specific marshalling code per-type?
[00:56:20] *** thakis_ has joined #chromium
[00:56:20] *** ChanServ sets mode: +v thakis_
[00:56:40] <SogarGofin> oh really? Wow thanks =)
[00:56:42] <eglaysher> (Though beware that I'm making major changes in this area, so treat it as accurate for now)
[00:56:43] <jamesr> SogarGofin: that's exactly right. see the file eglaysher referenced
[00:56:49] <eglaysher> (and not in a week or so)
[00:57:05] <jamesr> eglaysher: are you doing something about the multi-include pass stuff?
[00:57:09] <SogarGofin> I can describe my problem a bit. What we want to do is have the equivalent of a function call interface, where the function executes in a different process, and might return a value
[00:57:17] <eglaysher> jamesr: no. that's too well ingrained.
[00:57:23] <eglaysher> jamesr: I'm actually adding a fourth pass.
[00:57:27] <jamesr> eglaysher: ><
[00:57:33] <eglaysher> jamesr: but the gain is that we get to forward declare *everything*.
[00:57:49] <eglaysher> Including render_messages.h will no longer include every header in the project.
[00:57:57] <jamesr> eglaysher: ooh that is nice
[00:58:16] <SogarGofin> hehe
[00:59:03] <SogarGofin> Have any of you seen/written code that might marshall something like a complex object that has pointers from Process A Virtual Space into Process B Virtual Space in a generic way?
[00:59:06] <stuartmorgan> sheriffs/troopers: is anyone looking into where the two missing bots are? They don't seem to be listed in the continuing issues doc
[00:59:20] <eglaysher> jamesr: it seems that all the slowest files in the project (19 of the top 20 files by compile time, and spot checking all the way up to 50th slowest file) include render_messages.h
[00:59:24] <jamesr> SogarGofin: that's not a problem that we try to solve in a generic way in chrome
[00:59:39] <eglaysher> jamesr: so breaking this dependency should be the biggest compile speed gain I've done yet.
[00:59:39] <jamesr> SogarGofin: you might want to look at something like DCOM
[01:00:06] <trungl-bot> Tree closed by buildbot at chromium dot org: Tree is closed (Automatic: "compile" on "Chromium Linux Builder x64" from 54056: sargrass at google dot com)
[01:00:10] <jamesr> eglaysher: yeah, and it'll hugely cut down on the average incremental rebuild time
[01:00:23] <SogarGofin> I see, thanks. Yeah, I've come across a bit of DCOM in my past work but I was hoping to avoid it this time around, heh
[01:00:25] <jamesr> eglaysher: since touching a header that happens to be in render_messages.h will not mean rebuilding the world any more - right?
[01:00:41] *** alokp_ has quit IRC
[01:00:43] <eglaysher> jamesr: theoretically. it works in my prototypes.
[01:00:47] *** BUGa_grumpy has left #chromium
[01:01:13] <jamesr> SogarGofin: i think that not trying to solve this problem generically and instead being careful about what goes cross-process has worked well for chromium
[01:01:44] <SogarGofin> I guess that was my next question. All signs tell me that this problem might be a headache to solve generically
[01:01:52] *** xiyuan has left #chromium
[01:02:02] <stuartmorgan> evmar_sheriff/cpu: See my question above (in case it got lost in the stream)
[01:02:03] *** BCalvignac2 has joined #chromium
[01:02:10] <evmar_sheriff> argh build breakage
[01:02:15] *** BCalvignac3 has joined #chromium
[01:02:20] <thakis_> get used to it already
[01:02:56] <evmar_sheriff> SogarGofin: i think trying to solve it generically is a bad idea
[01:03:08] <evmar_sheriff> stuartmorgan: not sure, i haven't looked
[01:03:18] <thakis_> what's wrong with gmail today
[01:03:23] <jamesr> SogarGofin: if you go that route you end up with DCOM
[01:03:32] <SogarGofin> I figure that the beauty of a generic solution then easily isolate untrusted 3rd party libraries and still get results & work from them
[01:03:39] <evmar_sheriff> is anyone near sargrass?
[01:03:46] <SogarGofin> haha, yes I guess so
[01:03:49] <stuartmorgan> evmar_sheriff: I am; fixing now
[01:03:53] <evmar_sheriff> stuartmorgan: ok
[01:04:00] *** BCalvignac has quit IRC
[01:04:02] *** BCalvignac1 has quit IRC
[01:04:09] <SogarGofin> thanks very much for the reference and advice, I appreciate it!
[01:04:16] <evmar_sheriff> stuartmorgan: can you kindly tell him how the trybots work?  i suspect he might be new and unfamiliar with it
[01:04:22] <SogarGofin> It's an interesting problem for me
[01:04:36] <evmar_sheriff> SogarGofin: most libraries aren't written in a way that makes isolating them easy
[01:04:36] <leiz> evmar_sheriff: yes, he's our intern
[01:04:51] <stuartmorgan> evmar_sheriff: Yeah, we're discussing that.
[01:05:00] <stuartmorgan> evmar_sheriff: fix will be landed momentarily
[01:05:01] <evmar_sheriff> SogarGofin: e.g. they return an opaque pointer and have separate methods to get each field.  that means you need to IPC for each bit of data
[01:05:12] *** isherman has joined #chromium
[01:05:16] *** ChanServ sets mode: +v isherman
[01:05:33] <evmar_sheriff> stuartmorgan: which bots are missing?
[01:06:18] <rohitrao> mac 10.5 perf (2)
[01:06:26] <stuartmorgan> Oh, one was just mega-slow, so just that one
[01:06:36] <stuartmorgan> Its column has been empty all morning
[01:06:37] <SogarGofin> Yeah. Although in my case, returns might not be 100% necessary. More important is the ability to put a specific library into a process, let it do its thing over there, and receive from us instruction where we tell it, "call your function with this object, containing a pointer with this data, this value, etc"
[01:06:48] <evmar_sheriff> stuartmorgan: ah!  i only had been looking at the top row
[01:07:31] <rohitrao> memory_tests should not take 15+ hours :)
[01:07:47] <evmar_sheriff> i have no idea what to do other than ping nsylvain though :(
[01:09:19] <nsylvain> what's up?
[01:09:35] <evmar_sheriff> http://build.chromium.org/buildbot/waterfall/builders/Mac10.5%20Perf(2) has been running one test for 15 hours
[01:10:07] <nsylvain> ok, fixing
[01:11:21] *** pomaxa has quit IRC
[01:11:55] <cpu> SogarGofin: the other point is that when sending an object across the wire you have a security problem if you send the internal representation
[01:12:11] <cpu> well, at least a browser does
[01:12:49] *** rsimha has quit IRC
[01:13:36] <stuartmorgan> evmar_sheriff: okay, should cycle green. If not, there will be reverting and flogging
[01:14:28] <leiz> evmar_sheriff: from time to time, mac perf bots die, and tend to not get noticed for 10+ hours
[01:14:46] <leiz> mmm, flogging, that's short for file logging, right?
[01:15:39] *** Adys has quit IRC
[01:18:10] <jhawkins> evmar_sheriff: ok to check in valgrind suppression?
[01:18:31] *** rsesek has quit IRC
[01:19:01] <SogarGofin> cpu-> when you say security, do you mean a permission issue, or more like 'possibility of hacking'?
[01:19:16] <evmar_sheriff> jhawkins: yes, always :)
[01:19:23] <evmar_sheriff> jhawkins: but better to fix the valgrind issue ;)
[01:19:32] <jhawkins> ha, you're funny
[01:19:55] <evmar_sheriff> jhawkins: i've been watching the memory bots, i think we're mostly on a good trend except for CrOS which seems super screwed
[01:20:18] <trungl-bot> Tree opened by evan at chromium dot org (:evmar): Tree is open (continuing issues: http://bit.ly/bsyE8W)
[01:20:28] <jhawkins> evmar_sheriff: we've had a slew of recent issues that haven't been fixed though :( still working on finding appropriate owners
[01:20:33] <cpu> SogarGofin: I mean for example the concept of a string, which has at least 3 internal members
[01:20:52] *** Aferlak12 has quit IRC
[01:22:39] <cpu> SogarGofin: say size, capacity and the ptr to the data
[01:22:40] *** kjg has joined #chromium
[01:23:39] <cpu> SogarGofin: you don't send those, you send something that can be safely re-constructed into a string on the other side
[01:23:45] <atwilson> estade: (or anyone else) - is it possible/supported to do chrome-linux devel on gLucid?
[01:23:53] <estade> atwilson: yes
[01:23:53] <evmar_sheriff> atwilson: yes
[01:23:54] <atwilson> I have a noogler who was given a gLucid machine.
[01:24:02] <evmar_sheriff> better to use glucid, in fact
[01:24:21] <eglaysher> I build on glucid locally, and use try bots for ghardy
[01:24:22] <trungl-bot> Tree closed by buildbot at chromium dot org: Tree is closed (Automatic: "compile" on "Chromium Linux Builder (dbg)" from 54058: apatrick at chromium dot org, fbarchard at chromium dot org, jhawkins at chromium dot org, rohitrao at chromium dot org, sargrass at google dot com, victorw at chromium dot org, vrk at google dot com)
[01:24:30] <jamesr> atwilson: should work a-ok
[01:24:43] <jamesr> been using that setup for a few months now for primary dev
[01:24:58] *** paul_irish has quit IRC
[01:25:01] <SogarGofin> Hmm I see. My goal is to come up with an automated way to safely get that object reconstructed on the other side. That's what I'll be working on for the next few days =) Thanks again guys for pointing me at the file in chrome that does the IPC
[01:25:46] <SogarGofin> I actually have to run, but I'll be around lingering here for the next couple of days... hopefully not asking too-annoying questions =) thanks again!
[01:28:26] *** wash has joined #chromium
[01:28:55] *** Adys has joined #chromium
[01:29:28] *** paul_irish has joined #chromium
[01:29:29] *** ChanServ sets mode: +v paul_irish
[01:30:53] *** Beetny` has quit IRC
[01:32:00] *** exxe has quit IRC
[01:32:26] <trungl-bot> Tree opened by evan at chromium dot org (:evmar): Tree is open (continuing issues: http://bit.ly/bsyE8W)
[01:32:27] <rickcam> evmar_sheriff: I'm a noogler on gLucid, doing first Chrome build and getting SEGV in ncdecode_table and ncdecode_tablegen.  Ok to paste in a few lines of error log?
[01:32:53] <bryner_> windows trybot shows a bunch of compile warnings like "Failed to evaluate dependencies of source file:  <blah>, out of memory" -- known problem?
[01:33:11] <cpu> rickcam: maybe some NaCl guys here?
[01:33:26] <evmar_sheriff> rickcam: i know the issue, sec
[01:33:31] <rickcam> k
[01:33:57] <evmar_sheriff> rickcam: actually, i just wrote a chagne to fix it
[01:34:22] <evmar_sheriff> rickcam: you need to rm bin/ld from your chrome dir, and then 'sudo apt-get install binutils-gold'
[01:34:58] <rickcam> ok, I've previously swapped in gold in place of the default linker.  Still do that?
[01:35:17] <jamesr> there's a script on the binutils-gold that swaps out your ld
[01:35:18] *** rohitrao has quit IRC
[01:35:31] <jamesr> so it should take care of it for you. ld --version to be sure
[01:35:46] <rickcam> GNU gold (GNU Binutils for Ubuntu 2.20.1-system.20100303) 1.9
[01:37:12] <jamesr> odd. i don't have the trailing 1.9
[01:37:13] <rickcam> hrm, actually, doing ./ld --version in the chrome bin dir gives: GNU gold (GNU Binutils 2.20) 1.9
[01:37:24] <jamesr> yeah delete that one
[01:37:29] <rickcam> k
[01:38:29] *** FullFlannelJacke has quit IRC
[01:39:02] *** seventh has joined #chromium
[01:39:04] <rickcam> apt-get says its a no-op.  Just cp from /usr/bin/ld to .../chrome/bin/ld ?
[01:39:23] <jamesr> no, just rebuild
[01:39:30] <rickcam> k
[01:40:09] *** FullFlannelJacke has joined #chromium
[01:41:28] <rickcam> same error
[01:41:53] *** thakis_ has quit IRC
[01:42:10] <evmar_sheriff> rickcam: you have to delete the bad binaries from your build dir
[01:42:10] *** RT|Chatzilla has joined #chromium
[01:43:18] <evmar_sheriff> rickcam: easy way: rm out/Debug/*
[01:43:22] <evmar_sheriff> rickcam: note *not* -rf
[01:43:59] <rickcam> k.  will do.
[01:44:26] <evmar_sheriff> rickcam: sorry for the mess, lucid is still new :)
[01:45:13] <rickcam> np.  might as well face it now :-)
[01:48:45] *** d0k has quit IRC
[01:49:42] *** hbono has joined #chromium
[01:49:46] *** MX80 has quit IRC
[01:49:53] *** hbono has quit IRC
[01:50:35] <bryner_> nsylvain: i see you were clobbering the windows try bot -- do you know if it's ok now?
[01:50:37] *** hbono has joined #chromium
[01:51:14] <nsylvain> i think most of them are ok. Some of them might still be broken.  but you should try
[01:51:51] <bryner_> i have a job that's been compiling for quite awhile with no progress in the log
[01:52:55] <bryner_> (http://build.chromium.org/buildbot/try-server/builders/win/builds/42210)
[01:52:59] <nsylvain> that's a good sign ;)  Most likely linking a test executable
[01:53:09] <nsylvain> which would be after the webkit failures.. i'll look
[01:53:31] <bryner_> thanks
[01:53:36] *** pomaxa has joined #chromium
[01:53:38] *** paul_irish has quit IRC
[01:54:24] <nsylvain> oh well. looks likle it's broken
[01:54:44] <nsylvain> this one is a different problem though
[01:55:31] <bryner_> do you think it's specific to that build slave, i.e. should i just try submitting again?
[01:55:39] <nsylvain> yes, try again
[01:56:02] *** Sarten-X has quit IRC
[01:56:08] <bryner_> ok done
[01:57:28] *** ScriptRipper has quit IRC
[01:58:50] *** yutak_home has joined #chromium
[01:58:50] *** ChanServ sets mode: +v yutak_home
[01:59:37] <rickcam> evmar_sheriff: build completed!  thanks.
[02:00:05] <evmar_sheriff> rickcam: good to hear, to my knowledge you're the second person to ever run that script on a lucid box
[02:00:15] <rickcam> :-)
[02:02:32] *** kerz has joined #chromium
[02:02:32] *** ChanServ sets mode: +v kerz
[02:02:39] <kerz> ananta, ping?
[02:02:48] *** kerz has quit IRC
[02:02:56] *** kerz has joined #chromium
[02:02:56] *** ChanServ sets mode: +v kerz
[02:03:01] *** ejat has quit IRC
[02:08:09] *** chrisccoulson has quit IRC
[02:11:25] *** erikkay has quit IRC
[02:12:05] *** fsamuel_ has joined #chromium
[02:12:57] <ananta> kerz: yes?
[02:14:37] <evmar_sheriff> victorw: page cycler moz regression, maybe webkit roll?
[02:16:11] *** OtherJakeSays has joined #chromium
[02:17:12] <kerz> ananta, it looks like the official builders are having problems with your latest checkin
[02:17:19] <kerz> i think it's yours at least
[02:17:45] <kerz> both win and lin64 are failing to build
[02:17:48] *** stuartmorgan has quit IRC
[02:17:50] <victorw> evmar_sheriff: looking
[02:18:05] *** JakeSays has quit IRC
[02:18:37] *** japhet1 has joined #chromium
[02:18:37] *** ChanServ sets mode: +v japhet1
[02:19:32] *** leiz has quit IRC
[02:19:36] *** leiz has joined #chromium
[02:19:36] *** ChanServ sets mode: +v leiz
[02:19:51] *** RobWork has quit IRC
[02:20:12] *** japhet has quit IRC
[02:20:15] *** RobWork has joined #chromium
[02:20:28] <victorw> evmar_sheriff: not see any thing obvious, maybe just flaky?
[02:20:56] *** paul_irish has joined #chromium
[02:20:56] *** ChanServ sets mode: +v paul_irish
[02:21:00] <evmar_sheriff> victorw: yeah, not sure :(
[02:21:26] *** ScriptRipper has joined #chromium
[02:21:42] *** ScriptRipper has quit IRC
[02:21:42] *** ScriptRipper has joined #chromium
[02:21:56] *** awolfson has joined #chromium
[02:22:28] *** arv__ has quit IRC
[02:22:35] <victorw> evmar_sheriff: webkit canary xp perf is ok
[02:28:23] *** pomaxa has quit IRC
[02:28:25] <nirnimesh> nsylvain: is there a way to tell my local buildbot master to run a  separate port?
[02:36:13] <trungl-bot> Tree closed by buildbot at chromium dot org: Tree is closed (Automatic: "compile" on "Chromium XP" from 54078: dtseng at chromium dot org, evan at chromium dot org (:evmar), imcheng at chromium dot org, jhawkins at chromium dot org, michaeln at chromium dot org, pkasting at chromium dot org, sergeyu at chromium dot org, victorw at chromium dot org)
[02:36:59] <jhawkins> we can't go 30 minutes w/out a tree closure :-/
[02:38:01] <evmar_sheriff> 1: main.obj : error LNK2019: unresolved external symbol _av_register_protocol referenced in function "bool __cdecl `anonymous namespace'::InitFFmpeg(void)" (?InitFFmpeg@?A0x3eb144bd@@YA_NXZ)
[02:40:23] <pkasting> It's imcheng
[02:40:32] <pkasting> Is he on here under another name?
[02:41:30] <evmar_sheriff> i can't blame them too much, it passed the trybots
[02:42:23] <akalin> gah
[02:42:41] <akalin> does anyone else have problems with heapcheck not catching leaks that are known to happen on the buildbots?
[02:43:05] <jhawkins> akalin: did you remove the suppression before checking locally?
[02:43:10] <akalin> jhawkins: yeah
[02:43:10] <jhawkins> if there is one
[02:43:36] <jhawkins> I don't know about heapchecker, but at least for Valgrind, you have to build the chrome and test binary with a different configuration
[02:43:42] <akalin> hmm
[02:44:08] <akalin> oh duh
[02:44:10] <akalin> yeah, i'm dumb
[02:44:34] <evmar_sheriff> pkasting: i'm reverting
[02:45:57] <evmar_sheriff> but that doesn't help the perf regression
[02:46:45] <evmar_sheriff> and why did all of these tests suddenly fail for the first time
[02:46:46] <evmar_sheriff> http://build.chromium.org/buildbot/waterfall/builders/Interactive%20Tests%20(dbg)/builds/27828
[02:46:51] <earyoyo> in windows chrome build, why I had "IDS_L10N_OFFSET_AR undeclared identifier"?
[02:47:08] <earyoyo> sounds like a resource to me
[02:47:31] *** byte has left #chromium
[02:48:51] <evmar_sheriff> fbarchard: you caused perf regression again
[02:48:56] <evmar_sheriff> fbarchard: are you fixing or shoudl i revert?
[02:49:27] *** Sarten-X has joined #chromium
[02:50:18] <fbarchard> evmar_sheriff: are you sure?  The regression we saw before was fixed
[02:50:24] <jamesr> http://build.chromium.org/buildbot/perf/xp-release-dual-core/moz/report.html?history=150&rev=-1&graph=vm_peak_r
[02:50:25] <evmar_sheriff> http://build.chromium.org/buildbot/perf/xp-release-dual-core/moz/report.html?history=150&rev=-1&graph=vm_peak_r
[02:50:29] <trungl-bot> Tree closed by evan at chromium dot org (:evmar): Tree is closed (compile fail reverted, still investigating perf http://bit.ly/bsyE8W)
[02:50:38] <evmar_sheriff> the bump appears identical to the last one
[02:51:13] <fbarchard> version 54053 bumps to ffmpeg 54040 which saves 1.6 MB of memory
[02:52:05] <jamesr> so what's the bump?
[02:52:42] <fbarchard> details on the problems/fixes are here http://code.google.com/p/chromium/issues/detail?id=50507
[02:53:50] *** estes has quit IRC
[02:53:53] <jamesr> it really looks like the fix did not fix it
[02:55:05] <fbarchard> jamesr: the virtual memory after the roll had gone up by 1.6 MB.  I found a static array in webm that accounted for 1.6 MB increase in bss
[02:55:29] <ananta> kerz: sanjeev reverted that cl.
[02:55:30] <jamesr> fbarchard: but do you see the graph? the vm_peak_r went up after your roll
[02:55:42] <evmar_sheriff> fbarchard: i am inclined to try reverting and see if it helps
[02:55:54] <fbarchard> jamesr: we do expect the size to go up, as theres a lot of changes in 4 months of ffmpeg
[02:56:21] <kerz> ananta, yeah, he came by and let me know
[02:56:22] <kerz> thanks
[02:56:30] <fbarchard> the current threshholds are right on the edge, and likely dont account for webm being added to ffmpeg
[02:57:01] <jamesr> i would expect the increase would be smaller with this second roll, then, but it's definitely not any smaller
[02:58:39] <fbarchard> the way to detect the issue I fixed is fixed or not is size avcodec-52.dll.  The bss should be about 300k, not 1.8 MB
[02:59:06] <evmar_sheriff> maybe the builder doesn't get a new dll?
[02:59:39] <fbarchard> for some reason, after aglcient sync and build, I still have the old dll.  I'm removing my src/third_party/ffmpeg to force a sync from deps
[02:59:41] *** hbono has quit IRC
[03:01:39] <evmar_sheriff> don't we check in the dll or something?
[03:02:03] <evmar_sheriff> in any case, even when you fix this it will be another commit, so that can include this change again
[03:02:12] *** hbono has joined #chromium
[03:02:21] <fbarchard> the dll itself is checked into deps/third_party/ffmpeg.  in src/DEPS we select a version to copy into src/third_party/ffmpeg
[03:02:49] *** hebz0rl_ has joined #chromium
[03:02:56] <fbarchard> ack.  I'm reproducing a problem... the dll is still large
[03:03:17] <fbarchard> bss=1878448.  its okay in chrome, but not chromium
[03:04:02] <fbarchard> deps\third_party\ffmpeg\binaries\win\avcodec-52.dll is the correct version
[03:04:04] <bryner_> evmar_sheriff: fyi i'm looking into the heapcheck failure from my checkin
[03:05:02] <bryner_> let me know if i should revert in the meantime
[03:05:05] <fbarchard> I had the wrong CL.  it should be 54050
[03:05:29] <fbarchard> I can have a quick fix in about a minute
[03:05:50] <evmar_sheriff> i can do 54050 right now if you wan
[03:05:51] <evmar_sheriff> t
[03:06:14] <evmar_sheriff> say the word and i commit
[03:06:50] *** ctruta has joined #chromium
[03:06:53] *** hebz0rl has quit IRC
[03:07:00] <fbarchard> its okay... almost done
[03:07:11] *** ctruta has quit IRC
[03:07:32] <evmar_sheriff> i just need to press enter
[03:07:38] <evmar_sheriff> ok, we'll just wait.
[03:08:46] <trungl-bot> Tree closed by evan at chromium dot org (:evmar): Tree is closed (compile fail reverted, perf fbarchard http://bit.ly/bsyE8W)
[03:10:04] *** dmaclach has quit IRC
[03:10:18] <akalin> wait
[03:10:30] <akalin> so apparently MessageLoop::DeletePendingTasks() only deletes tasks under purify/valgrind?
[03:11:12] <akalin> is there a check to see if something is running under heapcheck?
[03:11:17] <bryner_> jhawkins: yt?
[03:11:20] <akalin> since it doesn't delete tasks, heapcheck is complaining
[03:11:33] <evmar_sheriff> akalin: that sounds pretty busted, we should have same behavior in both cases
[03:11:47] <akalin> There's a TODO by someone named darin
[03:11:48] <trungl-bot> Tree closed by fbarchard at chromium dot org: Tree is closed (compile fail reverted, XP perf should cycle green)
[03:11:51] <akalin> "Delete all tasks once it is safe to do so"
[03:12:04] <akalin> "Until it is totally safe, only delete them under Purify and Valgrind."
[03:12:48] <akalin> i bet there's a bug on this somewhere
[03:13:18] *** ojan has quit IRC
[03:13:35] *** anicolao has joined #chromium
[03:13:36] <akalin> http://code.google.com/p/chromium/issues/detail?id=2734
[03:13:48] <anicolao> evmar_sheriff: can you help me land https://bugs.webkit.org/show_bug.cgi?id=43102
[03:14:09] <anicolao> evmar_sheriff: I am setting up a leopard system but meanwhile would be nice to get this patch underway as suggested by Ojan.
[03:14:47] <evmar_sheriff> anicolao: i am perhaps not the right person, i don't have a mac
[03:15:17] *** OtherJakeSays is now known as JakeSays
[03:15:35] <anicolao> evmar_sheriff: :-) ... you are, however, the current chromium sheriff? I am happy to watch the bots and follow up promptly with patches as needed. Or we can wait until my other system is set up.
[03:15:36] <eggy> i have mac!
[03:15:39] <eggy> err a&
[03:15:45] <eggy> and I use CHrome on it
[03:16:01] <fbarchard> With CL54086 and gclient sync size on src\third_party\ffmpeg\binaries\chromium\win\ia32\avcodec-52.dll confirms the right dll.  Hopefully squeeks thru XP perf now
[03:16:07] <anicolao> eggy: volunteering to patch this patch and run the layout tests?
[03:16:13] <evmar_sheriff> anicolao: i think victorw is your man
[03:16:29] <anicolao> evmar_sheriff: ok thanks.
[03:16:34] <anicolao> victorw: here?
[03:16:42] <victorw> anicolao: hi
[03:16:55] <evmar_sheriff> akalin: http://code.google.com/p/chromium/issues/detail?id=6532 also, it would be nice to unify this forest of bugs
[03:16:58] <anicolao> victorw: https://bugs.webkit.org/show_bug.cgi?id=43102
[03:17:38] *** rohitrao has joined #chromium
[03:17:39] *** ChanServ sets mode: +v rohitrao
[03:17:52] *** trungl has quit IRC
[03:18:02] <akalin> evmar_sheriff: i just marked my bug as a dupe of 2734; should I mark this one as a dupe, too?
[03:18:35] <akalin> or maybe 2734 as a dupe of this one
[03:18:39] <evmar_sheriff> akalin: i dunno, i suggest reading them carefully and then deciding yourself ;)
[03:18:46] <akalin> mm
[03:19:40] <victorw> anicolao: are you webkit committer?
[03:19:42] <akalin> evmar_sheriff: i'll make it delete tasks under heapchecker, too
[03:20:04] <evmar_sheriff> we should just always delete tasks
[03:20:11] <evmar_sheriff> bah, we need willchan back to spend a month on this
[03:20:19] <anicolao> victorw: no. if I was, I could commit the patch and watch the bots as ojan suggests.
[03:20:21] <akalin> evmar_sheriff: i can change it to delete tasks
[03:20:26] <akalin> if you promise to deal with the fallout ;)
[03:20:54] <akalin> i think we should just do it
[03:20:55] <evmar_sheriff> i'm not sheriff tomorrow, so go ahead :P
[03:21:07] <akalin> okay
[03:21:09] <akalin> for great justice!
[03:21:13] <victorw> anicolao: ok. I can land the patch for you. Will it affect other non-chromium bots?
[03:21:42] <anicolao> no, the only code change adds PLATFORM(CHROMIUM) to a #if test
[03:21:57] <anicolao> possibly layout test expectations are wrong, but still should only affect chromium.
[03:22:08] *** isherman has quit IRC
[03:22:35] <victorw> anicolao: ok. I can land it for you and let's watch the webkit canary bots and rebaseline if needed
[03:23:32] <anicolao> victorw: sorry forgot to prefix the above for your attn
[03:23:33] *** johnny_g has quit IRC
[03:23:45] <anicolao> victorw: thanks
[03:23:46] *** paul_irish_ has quit IRC
[03:24:04] <victorw> anicolao: will you be around for a while?
[03:24:21] <victorw> anicolao: it swould be nice if you are around while I am doing all these
[03:25:39] *** kerz has quit IRC
[03:27:58] <anicolao> victorw: here for at least 3h
[03:28:05] <anicolao> victorw: longer if necessary.
[03:28:17] <evmar_sheriff> ok people, i'm off.  please be kind to the tree
[03:28:20] *** evmar_sheriff is now known as evmar_afk
[03:28:25] <akalin> d'oh
[03:28:31] <akalin> i was going to ask you to look at http://codereview.chromium.org/3076015
[03:28:42] <evmar_afk> akalin: definitely have darin review it
[03:29:02] <akalin> evmar_afk: ok
[03:29:03] <victorw> anicolao: cool. for mac, do we need diff results for leopard and snow leopard?
[03:29:23] *** phajdan-jr|afk has quit IRC
[03:29:41] <anicolao> victorw: I am not expecting to, except that the fonts antialias differently so any test that we rebaseline that involves fonts needs to be done twice.
[03:29:55] <anicolao> victorw: I'd like to fix that but haven't had time to look into what's wrong/what the best fix is.
[03:30:10] *** estes has joined #chromium
[03:30:17] <cpu> rdsmith: ping
[03:30:23] <anicolao> victorw: in fact my understanding was that due to this issue we weren't running layout tests on snow leopard, but I cuold be completely wrong.
[03:30:47] <jamesr> we do not run layout tests on snow leopard
[03:30:48] <victorw> anicolao: ok. will start soon
[03:30:53] <cpu> bah I think he is gone
[03:30:57] <jamesr> put leopard results only into chromium-mac
[03:31:12] <anicolao> victorw, jamesr: thanks jamesr
[03:31:16] <victorw> jamesr: yes. understand
[03:31:51] <rohitrao> akalin: I was trying to decide if my prefs change could cause a crash, because now we're falling back to using the prefs object instead of localstate
[03:32:13] <rohitrao> akalin: but then I decided that without my change, if we had no localstate, we would've just NPEd
[03:32:22] <rohitrao> so I think it's ok :)
[03:35:14] <akalin> rohitrao: ah
[03:35:23] <akalin> trungl-bot: whois evmar
[03:35:23] <trungl-bot> akalin: It looks like evmar could be: evan at chromium dot org / :evmar / evanm at google dot com.
[03:35:29] <anicolao> victorw: which bot is furthest ahead/which should I be watching for the earliest warning? (Even better, is there a command line that will answer this for me in general?)
[03:36:13] <victorw> anicolao: http://build.chromium.org/buildbot/waterfall.fyi/waterfall?branch=&builder=Webkit+(webkit.org)&builder=Webkit+Linux+(webkit.org)&builder=Webkit+Mac+(webkit.org)&builder=Webkit.org+Builder&committer=&reload=none
[03:42:17] <cpu> dinner time
[03:42:26] *** cpu has quit IRC
[03:43:35] *** yutak_home has quit IRC
[03:43:57] <victorw> anicolao: landing
[03:44:33] *** hbono has quit IRC
[03:44:40] *** MikeSmithW3C has joined #chromium
[03:46:42] *** bent-mozilla has quit IRC
[03:47:09] <anicolao> victorw: ok
[03:49:01] *** trungl has joined #chromium
[03:49:01] *** ChanServ sets mode: +v trungl
[03:49:32] <victorw> anicolao: landed: http://trac.webkit.org/changeset/64254
[03:49:44] <anicolao> victorw: yep saw it in #webkit
[03:49:53] <victorw> anicolao: let's watch webkit canary bots. normally win and linux pick up first
[03:50:11] *** hbono has joined #chromium
[03:50:23] <anicolao> victorw: these are the ones at the URL you already sent me, right?
[03:50:35] <victorw> anicolao: yest
[03:54:11] *** MikeSmithW3C has quit IRC
[03:56:24] *** dr_win has quit IRC
[03:56:44] *** trungl has quit IRC
[03:57:13] *** isherman has joined #chromium
[03:57:13] *** ChanServ sets mode: +v isherman
[03:58:55] *** rniwa has quit IRC
[03:59:21] *** rniwa has joined #chromium
[03:59:22] *** ChanServ sets mode: +v rniwa
[04:01:29] <victorw> anicolao: no test fail on linux
[04:02:28] <victorw> anicolao: your patch only affect mac
[04:03:03] <victorw> anicolao: mac webkit is little slow. have not pick up the patch yet
[04:03:12] <anicolao> victorw: yes the code change only affects chromium/cg
[04:03:46] *** dr_win has joined #chromium
[04:04:39] *** hbono has quit IRC
[04:06:54] <anicolao> victorw: any trivial way to tell what changelist it is compiling from the waterfall?
[04:08:16] <victorw> yes. click on the "build xyz" link, you should see Revision. For example, this link shows it is building 64252: http://build.chromium.org/buildbot/waterfall.fyi/builders/Webkit%20Mac%20(webkit.org)/builds/22786
[04:08:29] <victorw> anicolao: the revision could be chromium revision too
[04:08:43] *** hclam_ has quit IRC
[04:08:43] *** hclam has quit IRC
[04:10:01] *** MikeSmithW3C has joined #chromium
[04:10:43] *** mrossetti has quit IRC
[04:10:51] <victorw> anicolao: now,it picks up chromium patch 54086. think it will get yours in next run.
[04:10:53] *** pathorn has quit IRC
[04:11:21] <victorw> anicolao: it may take about 35 minutes for next run. so if you would like to have dinner, it's time to do it now
[04:11:31] <anicolao> after 10PM here :)
[04:11:49] <anicolao> victorw: thanks for the thought though :)
[04:12:10] <victorw> anicolao: oh,so you are in EST?
[04:12:31] <anicolao> victorw: yep, Waterloo ON
[04:13:21] <victorw> anicolao: I see. thanks for stay late to work this out
[04:13:58] <anicolao> victorw: no problem, it's my patch after all!
[04:15:32] *** estes has quit IRC
[04:17:19] <akalin> are we still keeping the tree closed?
[04:17:24] <akalin> the compile fail wa reverted, right?
[04:17:32] <jamesr> git cl upload did not do at all what i expected
[04:17:32] <victorw> anicolao: I am going to grab some dinner and back
[04:17:47] <jamesr> how does it pick a baseline to diff against?
[04:17:52] <jamesr> i thought it would just diff against trunk
[04:18:05] <anicolao> victorw: ok see you later, I'll pay attention meanwhile
[04:19:50] *** estes has joined #chromium
[04:19:51] *** Aria has joined #chromium
[04:21:21] *** jrmuizel has joined #chromium
[04:21:33] *** jrforbes_ has joined #chromium
[04:24:10] *** leeight has joined #chromium
[04:26:28] *** leeight has left #chromium
[04:28:21] *** dpranke has quit IRC
[04:34:41] *** fedorauser162 has joined #chromium
[04:35:39] <fedorauser162> can anyone tell me why after restarting chromium i get a print dialog repeatedly popping up?
[04:36:12] *** Adys has quit IRC
[04:37:21] *** Adys has joined #chromium
[04:37:31] *** Aria has left #chromium
[04:37:59] *** gonzojive has quit IRC
[04:38:01] *** gonzojive has joined #chromium
[04:38:45] *** johnny_g has joined #chromium
[04:40:03] *** trungl has joined #chromium
[04:40:03] *** ChanServ sets mode: +v trungl
[04:42:13] *** paul_irish_ has joined #chromium
[04:42:43] *** trungl has quit IRC
[04:43:44] *** bweinstein has quit IRC
[04:44:02] *** bweinstein has joined #chromium
[04:45:21] *** dmaclach has joined #chromium
[04:48:37] *** leeight has joined #chromium
[04:48:39] *** leeight has left #chromium
[04:49:46] <victorw> anicolao: there are 18 test failed, should I rebaseline all of them?
[04:50:03] <anicolao> victorw: my change isn't in that build, is it?
[04:50:35] <anicolao> victorw: unless I'm reading it wrong, this is the one _before_ the one with my edit :(
[04:50:51] <victorw> anicolao: I thought it is the case. If you look at the next run. your patch is there. but I think it does get your patch in somehow
[04:51:00] *** rniwa has quit IRC
[04:51:25] <anicolao> victorw: any way for me to easily see the source of ImageBufferCG.cpp as compiled in the build with the 18 failures? because tehse failures make sense with my change, if present.
[04:51:26] <victorw> anicolao: I think your patch does get included in this run
[04:51:52] <victorw> anicolao: http://build.chromium.org/buildbot/layout_test_results/webkit-rel-mac-webkit-org/54086/
[04:52:08] <victorw> anicolao: see layout-test_results.zip.
[04:52:19] <anicolao> victorw: looking now
[04:52:26] <victorw> anicolao: or you could rebaseline them and see the diff
[04:52:38] <anicolao> victorw: would like to understand why I thought my change wasn't in too though.
[04:53:25] <jamesr> anicolao: http://trac.webkit.org/log/?verbose=on&rev=64255&stop_rev=64253
[04:53:33] <jamesr> anicolao: that's the blamelist for the rev with the failures
[04:53:39] <jamesr> i see your change
[04:53:51] <victorw> anicolao: now, I think your change is actual in. probably when it pick up chromium change, it also took yours
[04:54:02] <jamesr> to see that, go to the dashboard
[04:54:02] <jamesr> (http://test-results.appspot.com/dashboards/flakiness_dashboard.html#useWebKitCanary=true&tests=svg%2FW3C-SVG-1.1%2Ffilters-composite-02-b.svg%2Csvg%2FW3C-SVG-1.1%2Ffilters-conv-01-f.svg%2Csvg%2FW3C-SVG-1.1%2Ffilters-diffuse-01-f.svg%2Csvg%2FW3C-SVG-1.1%2Ffilters-displace-01-f.svg%2Csvg%2FW3C-SVG-1.1%2Ffilters-example-01-b.svg%2Csvg%2FW3C-SVG-1.1%2Ffilters-image-01-b.svg%2Csvg%2FW3C-SVG-1.1%2Ffilters-light-01-f.svg%2Csvg%2FW3C-SVG-1.1%2Ffilt
[04:54:03] <jamesr> ght-04-f.svg%2Csvg%2FW3C-SVG-1.1%2Ffilters-specular-01-f.svg%2Csvg%2Fcustom%2FfeDisplacementMap-01.svg%2Csvg%2Fcustom%2Fimage-with-transform-clip-filter.svg%2Csvg%2Fcustom%2Fmask-colorspace.svg%2Csvg%2Fcustom%2Frecursive-filter.svg%2Csvg%2Ffilters%2FfeLighting-crash.svg%2Csvg%2Ffilters%2Ffilter-on-tspan.svg%2Csvg%2Ffilters%2Ffilter-width-update.svg%2Csvg%2Ffilters%2FfilteredImage.svg%2Csvg%2Ffilters%2Fparent-children-with-same-filter.svg) an
[04:54:03] <jamesr>  on the blue line
[04:54:07] <jamesr> apologies for the massive link
[04:54:12] <anicolao> jamesr: great, thanks! that's waht I wanted to hear, now just to figure out why I was confused. will do that after I'm done with the rebaseline
[04:54:32] <anicolao> jamesr: np on the link :)
[04:54:49] *** tonikitoo| has joined #chromium
[04:55:28] <victorw> anicolao: I can rebaseline for you.
[04:55:39] <leiz> hey, what's going on with the tree?
[04:55:51] <anicolao> victorw: am verifying outputs, you can rebaseline in parallel if you want
[04:56:03] <anicolao> victorw: only 18, I can verify them all by hand.
[04:56:15] <victorw> jamesr: it is little confusion. the build log says his change is in current run. but it was actually picked up by previous one. http://build.chromium.org/buildbot/waterfall.fyi/builders/Webkit%20Mac%20(webkit.org)/builds/22788
[04:57:12] <leiz> fbarchard: did r54053 break the Google Chrome Linux 'official' builder?
[04:57:29] <jamesr> i think it flaked out
[04:57:49] <jamesr> i don't see how  his patch could have broken browser_tests ReloadIndependently
[04:58:42] <leiz> pkasting: did r54073 break linux views?
[04:58:51] <jamesr> that test has some spurious failurs in the past
[05:00:44] <leiz> chromeos bots are also unhappy
[05:01:09] *** FullFlannelJacke has quit IRC
[05:01:49] *** FullFlannelJacke has joined #chromium
[05:02:33] *** trungl has joined #chromium
[05:02:33] *** ChanServ sets mode: +v trungl
[05:02:55] <rohitrao> is someone looking at interactive_ui_tests?
[05:03:28] <trungl> not me
[05:03:35] <rohitrao> hm
[05:03:39] <rohitrao> who can i blame?
[05:03:48] <trungl> yourself?
[05:03:55] <trungl> thakis?
[05:03:57] <trungl> society?
[05:04:07] <rohitrao> leiz: thanks for adding the "latest build" part to http://build.chromium.org/buildbot/waterfall/builders/Interactive%20Tests%20(dbg)/builds/27828/steps/extract%20build/logs/stdio :)
[05:04:29] <trungl> an act of god?
[05:04:33] <leiz> rohitrao: you're welcome! that saved all of us from a lot of confusion
[05:04:45] <leiz> I've updated the sheriffing notes
[05:04:46] <trungl> rohitrao: sergeyu?
[05:04:58] <rohitrao> trungl: nah, see the extract build warning
[05:05:08] <rohitrao> dtseng
[05:05:36] *** paul_irish has quit IRC
[05:06:20] <trungl> ah
[05:06:26] <rohitrao> revert, I guess?
[05:06:49] <trungl> is dtseng around?
[05:07:03] <rohitrao> not even sure who that is
[05:07:06] *** FullFlannelJacke has quit IRC
[05:07:10] <rohitrao> who/dtseng seems wrong?
[05:07:17] *** FullFlannelJacke has joined #chromium
[05:07:32] <trungl> urm, yeah
[05:07:33] <trungl> lemme see
[05:07:41] <trungl> crap
[05:07:54] <trungl> let me go on vpn first, and then I can check the committers list
[05:07:55] *** akalin has quit IRC
[05:07:56] *** awong has quit IRC
[05:08:00] <leiz> I'd help, but I've had a couple beers :)
[05:08:01] <trungl> back in a minute
[05:08:02] <rohitrao> what is this list?
[05:08:14] <trungl> leiz: that can only make things better!
[05:08:17] <rohitrao> leiz: that's the best time to revert
[05:08:22] <trungl> it's the canonical commiters list!
[05:08:31] *** akalin has joined #chromium
[05:08:34] <rohitrao> feel free to IM me a link :)
[05:08:35] <jamesr> check_perms failures on media/mf/* stuff
[05:08:37] *** awong has joined #chromium
[05:08:40] <trungl> but back in a sec, as I said, at which point I'll be able to tell you
[05:08:41] *** trungl has quit IRC
[05:08:41] <jamesr> on a try job
[05:08:46] <jamesr> has than been fixed in trunk?
[05:08:47] *** paul_irish has joined #chromium
[05:08:48] <leiz> jamesr: I can fix the check_perms failures!
[05:08:56] <jamesr> i dunno what bots would run check_perms
[05:09:12] <leiz> jamesr: only the main linux bot
[05:09:27] <leiz> jamesr: I'll fix it since I added it
[05:09:54] <anicolao> victorw: they all LGTM
[05:10:01] <jamesr> leiz: http://build.chromium.org/buildbot/try-server/builders/linux/builds/40195
[05:10:05] <anicolao> victorw: how's rebaseline going?
[05:10:14] *** trungl has joined #chromium
[05:10:17] <jamesr> leiz: since it's just a try run i dunno if there actually is a problem in ToT or not
[05:10:22] *** ChanServ sets mode: +v trungl
[05:10:33] <victorw> anicolao: doing it now
[05:10:43] <leiz> jamesr: ah, try job url?
[05:10:52] <jamesr> leiz: that's what i sent ya. http://build.chromium.org/buildbot/try-server/builders/linux/builds/40195
[05:11:55] <leiz> jamesr: it's fixed on trunk, but lkgr has not caught up yet, so don't worry about it
[05:12:13] <jamesr> cool. i do too little chromium stuff these days
[05:12:22] <jamesr> what's the story with the tree? i has a patch
[05:13:29] <leiz> I went out for drinks, came back and now it's red
[05:13:31] <anicolao> jamesr: victorw is rebaselining a bunch of mac tests for me and we'll be back in a minute, if you're talking about mac being red.
[05:13:36] *** bryner_ has quit IRC
[05:13:46] <jamesr> anicolao: nah, i mean the chromium tree not the webkit canaries
[05:14:07] <anicolao> jamesr: ah, then I am happily blameless ;-P
[05:14:43] <jamesr> it looks like a lot of flake
[05:14:53] <jamesr> is there a sheriff or is this "you break it, you buy it"?
[05:14:55] <victorw> anicolao: are all these baselines same as those in platform/mac-leopard? If so, just delete them should work
[05:15:28] <rohitrao> jamesr: no sheriff, I don't think
[05:15:36] <rohitrao> we're going to revert dtseng
[05:15:43] <rohitrao> trungl: so revert? :)
[05:15:57] <trungl> rohitrao: are you sure it's him?
[05:16:02] <trungl> rohitrao: I didn't actually look
[05:16:09] <rohitrao> double-check
[05:16:16] <rohitrao> but it looks like new tests he added are failing
[05:16:21] <rohitrao> wonder why they passed on the trybots
[05:16:25] <trungl> in that case, revert
[05:16:46] *** FullFlannelJacke has quit IRC
[05:17:10] <jamesr> it didn't pass trybots?
[05:17:17] <jamesr> i see a compile failure on the last win try job from that patch
[05:17:25] *** FullFlannelJacke has joined #chromium
[05:17:28] <victorw> anicolao: my rebaseline tool tells me they are the same as those in platform/mac-leopard,
[05:18:01] <anicolao> victorw: so they should be but I didn't verify it. I verified the results visually with pixie.
[05:18:38] <anicolao> the code edit should bring chromium on mac in line with webkit on mac where they previously differed.
[05:18:58] <victorw> anicolao: I trust my tool. so basically we don;t need to update new baselines. My tool just deletes them as they are the same as those in fallback path
[05:19:05] <victorw> anicolao: is this expected?
[05:19:12] <rohitrao> jamesr: second-to-last patchset passed on windows
[05:19:21] <jamesr> rohitrao: bizarre
[05:19:23] <anicolao> victorw: yes, because the change makes us use the same colorspace as webkit mac
[05:19:24] <rohitrao> I guess there were substantial changes between 8 and 9?
[05:19:32] <anicolao> victorw: so the results should match.
[05:20:06] <anicolao> victorw: this, I think, is progress ;-)
[05:20:26] <anicolao> victorw: (though technically webkit-mac has them wrong too...shhhh)
[05:20:28] <rohitrao> no, no changes between 8 and 9
[05:21:12] <anicolao> victorw: (they're much closer to correct than before)
[05:21:13] <jamesr> difference between trybots and real bots?
[05:22:25] <trungl> that may be the case
[05:22:27] <trungl> who can tell
[05:22:57] <jamesr> the shadow knows!
[05:24:25] <rohitrao> akalin: heh, of course that PrefsControllerTest doesn't crash on the real official bot :)
[05:24:35] *** cpu has joined #chromium
[05:24:35] *** ChanServ sets mode: +v cpu
[05:26:34] <anicolao> victorw: you happy or worried?
[05:30:48] *** bryner_ has joined #chromium
[05:30:49] <victorw> anicolao: not sure:). almost all baselines are same as those in platform/mac or platform/mac-leopard. the only one may be different is mask-colorspace. it has different checksum file
[05:31:12] <anicolao> victorw: mask-colorspace is my test. the result is wrong, but close enough.
[05:31:23] <rohitrao> how did someone manage to break kiosk mode on linux views?
[05:31:28] <rohitrao> cpu: I reverted dtseng
[05:32:08] <cpu> thanks
[05:32:17] <rohitrao> cpu: the chromiumos bots look really unhappy
[05:32:24] <rohitrao> and linux (views) is failing consistently
[05:32:28] <rohitrao> but I don't see why
[05:32:31] *** bryner__ has joined #chromium
[05:33:04] <victorw> anicolao: could you check if you have the right checksum for this one? you can check agaist the archived results from webkit canary bot
[05:34:24] <victorw> anicolao: I guess the one checked in could be wrong.
[05:34:42] <victorw> anicolao: will this test run non chromium port?
[05:35:08] *** bryner_ has quit IRC
[05:35:10] <anicolao> victorw: will do whatever needs to be done, but have to first confess I don't understand the checksum files. the checked in checksum for mask-colorspace.svg was generated by me on snow leopard, so it is more likely wrong than the one from the bot.
[05:35:42] <jamesr> .checksums are always associated with .png
[05:35:48] <jamesr> the .checksum next to a .png should match that .png
[05:35:56] <jamesr> so if you generated a .png and .checksum on snow leopard, then the two should go together
[05:36:05] <cpu> rohitrao: well at least some seem to be related to the cl you reverted
[05:36:07] <anicolao> victorw: mask-colorspace.svg is a valid test for all ports
[05:36:34] <anicolao> jamesr, victorw: but when the leopard bot regenerates the png and checksum, then they could be different than my snow leopard ones? and we are meant to run the tests only on leopard?
[05:36:38] <victorw> anicolao: ok. I will add this one to platform/chromium-mac.
[05:36:47] <victorw> anicolao: I think we should be good
[05:36:49] <jamesr> anicolao: the leopard results go in a different place from the snow leopard results
[05:37:08] <jamesr> anicolao: you should run on both. snow leopard results go into platform/mac. leopard results go into platform/mac-leopard. chromium results go into platform/chromium-mac
[05:37:20] <jamesr> you never have a .png without a .checksum or a .checksum without .png
[05:37:33] <jamesr> so if the leopard ones are different then they both go into a different folder
[05:38:07] <anicolao> jamesr: but if I generate a png on linux and it's identical to the png on mac, you'd expect the checksums to also match?
[05:38:11] *** wash has quit IRC
[05:38:14] <jamesr> yes
[05:38:18] <jamesr> if they are bit-for-bit identical
[05:38:19] <victorw> jamesr: I think we just need to add his new leopard result to mac-leopard
[05:38:25] <jamesr> victorw: that sounds right
[05:38:36] <anicolao> victorw: OK
[05:40:25] <cpu> I am tempted to think that the CrOS redness is  xiyuan@
[05:40:30] <rohitrao> cpu: oh, kiosk failure is 54073
[05:41:35] <anicolao> jamesr: and if the rendered page is pixel-for-pixel identical, you'd expect the PNGs to be bit-for-bit identical?
[05:41:44] <jamesr> anicolao: i would ask 'diff'
[05:42:04] <jamesr> it's never lied to me
[05:42:31] <anicolao> jamesr: I guess people won't be interested in changing the format of the pixel tests from PNG to something more basic.
[05:42:37] <jamesr> why?
[05:42:42] <jamesr> they take up gigantic amounts of space already
[05:43:00] <jamesr> and we have tools that can handle PNGs
[05:43:44] <anicolao> jamesr: because the same pixel results do not seem to always generate the same PNGs. Or possibly I'm doing something wrong. I will have to do more of these to understand why identical results onscreen can create different results.
[05:44:09] <jamesr> anicolao: i think your eyes are deceiving you
[05:44:17] *** estes has quit IRC
[05:44:27] <jamesr> anicolao: very very similar results != identical results
[05:44:40] <jamesr> you can try using the image diff tool to highlight subtle differences
[05:44:41] <anicolao> jamesr: possibly I've been sloppy and not examined every antialiased pixel. However, I am using pixie not my eyes.
[05:44:45] <jamesr> maybe roundoff or antialiasing differences
[05:44:48] <jamesr> yeah, don't use pixie either
[05:44:53] <jamesr> use diff if you want to know if they are the same
[05:44:58] <jamesr> regardless PNG is not the problem
[05:45:25] <anicolao> jamesr: OK. I will chalk it up to antialias differences and check more thoroughly next time.
[05:45:38] <jamesr> but you don't have to check yourself
[05:45:41] <jamesr> the tools do that stuff
[05:45:54] <anicolao> victorw: back to today's topic: are we good to go on the rebaseline change?
[05:46:23] <anicolao> jamesr: no, in the case of these tests where the colorspace is subtly off, I need to rebaseline them wehn it's fixed but I am not fixing antialiasing. so the tests may vary but be correct, manual inspection is all I've got.
[05:46:42] <rohitrao> cpu: should I revert pkasting's change?
[05:46:53] <rohitrao> it has kiosk in the cl description, so I'm suspicious :)
[05:46:55] <jamesr> anicolao: you have to manual inspect to see if the results are correct, but the rebaseline tool will take care of duplicates for you
[05:47:43] <anicolao> jamesr: fair enough.
[05:48:52] <anicolao> jamesr: just dreading if/when I figure out what should be changed to fix CG's sRGB->Linear conversion, which will break all these once more :)
[05:49:20] <jamesr> anicolao: we also have 2800 tests that need snowleopard image baselines generated
[05:49:43] <anicolao> jamesr: what is the plan for that?
[05:49:53] <jamesr> 300 at a time, i guess
[05:49:59] <anicolao> jamesr: can we get rid of irrelevant text antialiasing while we're at it?
[05:50:05] <victorw> anicolao: ok. landing the patch. remove chromium specific baselines and add leopard result to platform/mac-leopard for your new test
[05:50:06] <jamesr> not really
[05:50:23] <anicolao> jamesr: sad :( why not?
[05:50:33] <jamesr> anicolao: well, we need to make sure we antialias text correctly
[05:50:39] <anicolao> victorw: thanks for your patience/help!
[05:50:53] <jamesr> and right now we don't have any way to say for which tests the text antialiasing is part of the test
[05:51:00] <anicolao> jamesr: no, I mean irrelevant text antialiasing. as in so many SVG tests whose purpose is to test something else but have text commentary in the output
[05:51:04] <jamesr> or just some descriptive text that's not what is actually being tested
[05:51:08] <jamesr> yeah
[05:51:11] <jamesr> but how do you know that?
[05:51:15] <anicolao> jamesr: turning off antialiasing for those could make them much more robust.
[05:51:23] <anicolao> jamesr: because I've read the svg files?
[05:51:39] <jamesr> and have you encoded that knowledge in some machine-readable way? wanna do that for all 20000 tests?
[05:51:43] <victorw> cpu: are we going to open the tree soon?
[05:51:51] <jamesr> also i'm not sure we can just turn off text antialiasing so easily
[05:51:56] <jamesr> across all ports/platforms
[05:52:02] <victorw> cpu: would like to roll webkit
[05:52:33] <anicolao> jamesr: for SVG, which is mainly what I'm thinking of here, it should be possible to turn off in the files themselves.
[05:52:37] *** dhollowa_ has joined #chromium
[05:52:50] <jamesr> anicolao: we could just delete text from layout tests, sure
[05:52:58] <anicolao> jamesr: however, the encoding/scale problem remains. but it's reasonable to do it while regenerating their snow leopard baselines.
[05:53:03] *** bryner__ has quit IRC
[05:53:08] <jamesr> anicolao: or put it in a comment or what have you. not clear that it's any easier than mass rebaselining
[05:53:39] <jamesr> go for it if you want :). i'm gonna gtfo, looks like waiting for the tree to open might have been a mistake
[05:53:40] <rohitrao> victorw: I reverted a bunch of changes, but I have no idea what's wrong with the cros builders on the right
[05:53:41] *** jamesr has quit IRC
[05:53:57] <anicolao> jamesr: we can set text-antialiasing: false on the irrelevant text in those files while we're rebaselining them.
[05:54:02] *** paul_irish has quit IRC
[05:54:18] <anicolao> jamesr: ah well thanks for the discussion, gl with the tree in the morning.
[05:54:26] <victorw> rohitrao: ok.
[05:57:55] *** bweinstein_ has joined #chromium
[05:57:55] *** bweinstein has quit IRC
[05:58:22] *** bweinstein_ has quit IRC
[05:58:59] <anicolao> victorw: 22790 still doesn't have the rebaseline?
[05:59:15] <victorw> anicolao: still landing, it is slow
[05:59:39] <anicolao> victorw: OK going to take a brief break and come back - 15m
[05:59:46] *** paul_irish has joined #chromium
[05:59:49] <victorw> anicolao: sure
[06:00:32] <cpu> victorw: go ahead
[06:00:37] <rohitrao> cpu: I think cros might be a bot problem? I'm not sure
[06:00:47] <victorw> anicolao: landed
[06:00:51] <victorw> cpu: thx
[06:00:56] <rohitrao> although hitting two bots at once is strange
[06:01:24] <victorw> cpu: done
[06:01:37] <cpu> rohitrao: earlier we had issues and a reboot seemed to work, but it was more clear from the logs
[06:02:08] <victorw> anicolao: http://trac.webkit.org/changeset/64264/trunk
[06:02:09] <rohitrao> I know the libcros warnings show up in the successful runs too
[06:02:11] *** mirandac_ has joined #chromium
[06:03:07] <cpu> rohitrao: from my perspective I am mostly concerned with bad check-ins, if a piece of infraestructure goes bad is less of a concern
[06:03:27] <cpu> of course then it masks bad checkings, sigh
[06:04:02] <rohitrao> :)
[06:05:18] <maruel_ooo> I'm sending clbber try jobs to the win try slaves
[06:05:37] <trungl-bot> Tree closed by cpu at chromium dot org: Tree is closed (victorw on webkit roll)
[06:06:57] *** isherman has quit IRC
[06:08:22] <rohitrao> maruel_ooo: can we try rebooting the two cros bots on the right?
[06:08:41] <rohitrao> we're not really sure if these failures were caused by a checkin or something else
[06:08:48] <rohitrao> although two bots breaking at once is suspicious
[06:11:36] <maruel_ooo> rohitrao: sure
[06:11:39] <maruel_ooo> we = you?
[06:11:54] <cpu> me too
[06:12:07] <cpu> I mean, I don't have good leads
[06:12:44] <cpu> emailed  xiyuan@ but that is a long shot too
[06:13:45] <maruel_ooo> cpu: I mean s/can we try/can rohitrao/
[06:14:03] <maruel_ooo> I don't feel like connecting to the slaves tonight :)
[06:14:19] <trungl> hey cpu, feel like doing a (fairly trivial) review?
[06:14:21] <cpu> ah
[06:14:28] <rohitrao> i should learn how to mess with the bots
[06:15:05] <cpu> trungl: shoot
[06:15:25] <trungl> cpu: http://codereview.chromium.org/3073008/show
[06:15:34] <trungl> it's fairly trivial assuming it compiles
[06:16:27] <cpu> when is messin' with browsermain trivial
[06:16:38] * cpu has been duped
[06:16:51] <trungl> it's trivial on wednesday nights?
[06:17:03] <trungl> wednesday night is "browsermain night"
[06:17:04] *** Adys has quit IRC
[06:17:06] <trungl> drinks are on the house
[06:17:30] * trungl goes to pour himself a drink
[06:17:41] *** loislo has joined #chromium
[06:18:47] <cpu> it is puzzling that ssl needs to be initialized so early
[06:18:48] *** loislo has quit IRC
[06:18:56] *** ebraminio has joined #chromium
[06:19:08] <trungl> cpu: possibly it can be moved later
[06:19:17] *** paul_irish has quit IRC
[06:19:34] <trungl> though that sort of change would be much scarier
[06:19:40] *** paul_irish has joined #chromium
[06:20:08] <cpu> BrowserMainSexyParts
[06:20:52] *** fsamuel_ has quit IRC
[06:21:23] <trungl> haha
[06:23:22] <trungl> cpu: thanks
[06:24:45] <trungl-bot> Tree opened by cpu at chromium dot org: Tree is open (poke victorw if webkit fails, continuing issues: http://bit.ly/bsyE8W)
[06:25:30] <cpu> love the test=everything works
[06:25:53] <trungl> cpu: for some reason, that test always fails
[06:25:56] <maruel_ooo> rohitrao: cpu: so you want me to reboot them or not?
[06:26:00] <maruel_ooo> and which one?
[06:26:01] <trungl> I've never seen *everything* work
[06:26:38] <cpu> maruel: any on the far right that you fancy
[06:27:01] <cpu> at this time I don't really expect any serious effort
[06:27:17] *** 84XAAPVWF has joined #chromium
[06:28:11] *** dmaclach_ has joined #chromium
[06:28:26] *** homata_ has quit IRC
[06:28:58] <cpu> in fact I am going bye bye
[06:29:03] *** rickcam|home has quit IRC
[06:29:08] <trungl> 'night cpu
[06:29:16] *** Adys has joined #chromium
[06:29:17] <cpu> :)
[06:29:18] *** cpu has quit IRC
[06:29:27] <anicolao> why does git clone of webkit take forever within corp?
[06:30:09] <trungl> does it not take forever outside of corp?
[06:31:01] <anicolao> trungl: I'm talking 68 KiB/s ... as in, I hope it'll be done by morning.
[06:31:10] <trungl> wow, that is slow
[06:31:20] <trungl> but it might just be an upstream problem
[06:31:26] *** dmaclach has quit IRC
[06:31:27] *** dmaclach_ is now known as dmaclach
[06:31:32] <anicolao> trungl: no, if I go home it'll go 10x the speed.
[06:31:47] *** arv__ has joined #chromium
[06:31:48] <trungl> anicolao: right now? do you know this for a fact?
[06:32:50] <anicolao> trungl: mmm, what's the easiest way to verify I wonder. my house is locked down so I can't literally try it now.
[06:33:22] *** tonikitoo| is now known as tonikitoo
[06:33:24] *** hebz0rl_ has quit IRC
[06:33:45] <anicolao> trungl: but I could try from a local university. one second...
[06:35:56] *** dhollowa_ has quit IRC
[06:37:53] <trungl> anicolao: fwiw, I get > 6 MiB/s (from corp, but mtv)
[06:38:00] <trungl> it could be a local connectivity issue
[06:38:52] <anicolao> trungl: git turns out not to be installed at the university
[06:39:11] <trungl> how annoying
[06:39:22] <anicolao> trungl: is there a trivial way to tunnel git through ssh - I could ssh into MTV and then git over the tunnel to see if that alters things.
[06:39:52] *** arv__ has quit IRC
[06:39:57] *** arv__ has joined #chromium
[06:40:10] <rohitrao> trungl: in retrospect, i should've just gone to sleep hours ago
[06:40:45] <trungl> anicolao: trivial? not that I know of
[06:40:52] <trungl> (I mean, you could port forward, I guess)
[06:41:00] <trungl> but
[06:41:03] <trungl> uhm
[06:41:04] <trungl> hrm
[06:41:52] <trungl> anicolao: if you have a laptop, you may also be able to vpn to a particular location
[06:42:10] <maruel_ooo> trungl: vpn will be slower
[06:42:18] <anicolao> trungl: oh well don't worry about it, I am not here for long enough for this to be worth optimizing. my laptop is snow leopard, which is the whole reason I'm setting up a leopard environment :)
[06:42:27] <maruel_ooo> anicolao: you can port forward but that will probably not help
[06:42:50] <trungl> anicolao: if you have an existing repo, you could clone from that instead
[06:43:08] <trungl> rohitrao: sleep is more mere mortals!
[06:43:17] <trungl> rohitrao: I mean, sleep is for mere mortals!
[06:43:23] <rohitrao> apparently so is proper grammar
[06:44:36] <trungl> proper grammar is just a bourgeois conspiracy anyway
[06:44:37] <maruel_ooo> ahah
[06:45:10] *** 36DAABBGH has quit IRC
[06:50:01] *** pfeldman has joined #chromium
[06:50:18] *** terinjokes has joined #chromium
[06:51:18] <terinjokes> hey guys, I'm taking a look at a bug in chromium, to see if I can't find my way around and potentially fixing it ;)
[06:53:16] *** jrmuizel has quit IRC
[06:54:56] <rohitrao> which bug?
[06:55:20] <trungl> terinjokes: sounds good, once you really get going you may want to post something on the bug to make sure someone else isn't also working on it
[06:55:32] <terinjokes> checking out the sources now
[06:55:47] <terinjokes> rohitrao: it's a bug with the developer tools
[06:56:01] <rohitrao> do you know which bug #?
[06:56:28] <terinjokes> #49502
[06:56:49] <anicolao> trungl: hmm. when I try git clone git://git.webkit.org/Webkit.git WebKit from paddington.mtv I get the remote end hung up unexpectedly.
[06:56:56] <anicolao> trungl: presumably this works for you?
[06:57:06] <trungl> anicolao: that worked fine for me
[06:58:04] <trungl> anicolao: more accurately, I let it clone to 25%, then killed it since I don't really want to put extra load on the server
[06:58:22] <rohitrao> terinjokes: this seems complicated enough that once you have a plan, you should post it to the bug and get feedback
[06:58:29] <anicolao> trungl: sure, but I don't get it. I'm running the same command, from the same host, and it won't even connect.
[06:58:34] <anicolao> trungl: I mean it hangs up immediately.
[06:58:48] <anicolao> trungl: Initialized empty Git repository in /tmp/wkfoo/third_party/WebKit/.git/
[06:58:48] <anicolao> fatal: The remote end hung up unexpectedly
[06:59:10] <rohitrao> trungl: why won't this test fail locally?
[06:59:19] <rohitrao> it doesn't seem fair that the bots get to have all the fun
[06:59:37] <trungl> anicolao: okay, now I get the same thing as you
[06:59:44] <trungl> anicolao: it worked previously
[06:59:58] <anicolao> trungl: odd. well I'll leave my slow clone running
[07:00:20] <anicolao> trungl: and maybe play with silly options tomorrow to prove to techstop that there is a problem.
[07:00:48] <trungl> anicolao: I'm getting the same problem with another mtv machine
[07:00:59] <trungl> anicolao: so I'd bet on a problem on the server side
[07:01:02] <terinjokes> rohitrao: i've always been told, "if there's a will there's a way"...
[07:01:04] *** bers has joined #chromium
[07:01:13] <FullFlannelJacke> latest Chromium build 6.0.478.0 (53895) Ubuntu 10.04 is having some wacky page rendering.  The fonts and page alignment are all screwy.  I have to reload page to fix it
[07:01:19] <anicolao> trungl: yes the hangup problem is new, seems to happen everywhere at the moment.
[07:01:37] <rohitrao> there's always a way.  the trick is making sure it's the right way ;)
[07:01:54] <trungl> rohitrao: what's the test that's failing?
[07:02:06] <rohitrao> FullFlannelJacke: if it's reproducible, filing a bug with a test case would be really helpful
[07:02:14] <rohitrao> trungl: TabContentsTest.WebKitPrefs
[07:02:25] *** mirandac_ has quit IRC
[07:02:29] <trungl> rohitrao: which bot is it failing on?
[07:02:32] <rohitrao> looks like we override some prefs in the test, but the official builds don't see that
[07:02:33] <trungl> (or did it fail on?)
[07:02:39] <rohitrao> trungl: all of the official bots :)
[07:03:20] *** peepsalot has quit IRC
[07:03:23] <anicolao> victorw: your rebaseline is finally underway...thanks!
[07:03:23] <trungl> rohitrao: really? where?
[07:04:22] <victorw> anicolao: still need to watch the output on webkit canary mac to make sure they pass.
[07:04:43] <anicolao> victorw: yes, but the ETA is only 2m. so we're close.
[07:04:55] <anicolao> victorw: and what could possibly go wrong? :-D
[07:05:21] <FullFlannelJacke> rohitrao: Most pages I visit, especially web forums are screwed up.  Everything is bunched up to the left side of page.  I have to reload page to finally get it right..
[07:05:43] <FullFlannelJacke> just started happening after my daily Chromium PPA update
[07:05:45] <trungl> rohitrao: you're testing 472 locally?
[07:05:59] <rohitrao> FullFlannelJacke: file a bug with a url and a screenshot, I guess
[07:06:00] <victorw> anicolao: the baseline should be fine. the mask-colorspace results should be the same on chromium leopard and non-chromium leopard, right?
[07:06:01] <anicolao> victorw: oh dear, unexpcted passes
[07:06:05] <anicolao> victorw: looking now.
[07:06:16] <rohitrao> trungl: yeah, it passes on a local official build
[07:06:36] <anicolao> victorw: yes, I think so.
[07:06:36] <trungl> rohitrao: huh
[07:06:41] <rohitrao> trungl: but it's failing pretty convincingly on the release bots :)
[07:06:43] <victorw> anicolao: that's fine. we should rejust remove them from test expectations
[07:06:46] <trungl> rohitrao: are you sure you're grabbing all the right deps and all?
[07:06:52] <rohitrao> trungl: who knows
[07:06:58] <rohitrao> everything compiles, though
[07:07:04] <rohitrao> and I get the colorful icon
[07:07:11] <victorw> anicolao: it can be done when you have time tomorrow. the build looks ok
[07:07:13] <trungl> rohitrao: and also the right webkit?
[07:07:21] <anicolao> victorw: yes, I will, after taking a quick look @ the output.
[07:07:29] <trungl> (well, we should have a branch of that)
[07:07:36] <rohitrao> trungl: I'm just using ToT
[07:07:40] *** peepsalot has joined #chromium
[07:07:43] <rohitrao> i didn't pull the branch or anything
[07:07:56] <victorw> anicolao: I think that's expected. we have the right baselines now so the tests pass
[07:08:08] <trungl> rohitrao: but the release builders are building 472 (mostly? only?)
[07:08:15] <rohitrao> but this is failing on all the bots, including the nightly version increment builds
[07:08:22] <anicolao> victorw: I'm sure it is, but I will still take a minute to look @ the baselines.
[07:08:37] <anicolao> victorw: only wish the previous build had told us, so that we could have fixed it in one go.
[07:08:57] <trungl> rohitrao: oh
[07:08:59] <trungl> where?
[07:10:41] <victorw> you could upload a patch to remove them, and get it landed tomorrow with commit queue
[07:11:33] *** path[l]_ has joined #chromium
[07:11:55] *** path[l] has quit IRC
[07:11:55] *** path[l]_ is now known as path[l]
[07:13:27] *** arv___ has joined #chromium
[07:13:38] *** xji has quit IRC
[07:13:49] *** n3v3le has joined #chromium
[07:16:33] *** arv__ has quit IRC
[07:17:14] *** pfeldman has quit IRC
[07:19:21] *** paul_irish has quit IRC
[07:19:55] *** paul_irish has joined #chromium
[07:20:38] *** jrforbes_ has quit IRC
[07:24:00] *** Engin has joined #chromium
[07:25:21] *** dmaclach_ has joined #chromium
[07:28:06] *** rohitrao_ has joined #chromium
[07:28:06] *** ChanServ sets mode: +v rohitrao_
[07:28:14] <anicolao> victorw: are you a webkit reviewer?
[07:28:30] <victorw> anicolao: no
[07:29:07] *** dmaclach has quit IRC
[07:29:36] *** awolfson has quit IRC
[07:32:10] *** rohitrao has quit IRC
[07:32:11] *** rohitrao_ is now known as rohitrao
[07:33:16] <m0> .
[07:33:35] <anicolao> victorw: OK nobody is around to review the patch. I'm going to head home and let the cq take care of it in the fullness of time - probalby one of the germany guys can r+/c+ it in a few hours.
[07:33:46] <anicolao> victorw: thanks again for your help & patience. ttyl.
[07:34:32] <victorw> anicolao: np. you can cc jorlow he is the webkit gardener in a few hours and he is a reviewer
[07:35:50] *** estes has joined #chromium
[07:36:57] <victorw> sheriffs: ping
[07:37:23] *** anicolao has quit IRC
[07:38:08] <victorw> any sheriff around?
[07:40:43] *** johnny_g has quit IRC
[07:43:39] <trungl-bot> Tree opened by victorw at chromium dot org: Tree is open (continuing issues: http://bit.ly/bsyE8W)
[07:44:05] <victorw> I am going to be offline. for webkit rolling issue, ping jorlow who will be online in a few hours
[07:46:26] *** victorw has left #chromium
[07:46:37] <terinjokes> does a checkout of chromium normally take this long to run post-process stuff?
[07:47:21] *** arv___ has quit IRC
[07:49:44] *** bers has quit IRC
[07:59:23] *** anicolao has joined #chromium
[08:00:53] *** anicolao has quit IRC
[08:01:04] *** anicolao has joined #chromium
[08:01:05] *** ChanServ sets mode: +v anicolao
[08:04:53] *** victorw has joined #chromium
[08:09:51] *** markusheintz_ has joined #chromium
[08:17:16] *** Zaba has quit IRC
[08:18:35] *** Zaba has joined #chromium
[08:18:59] <trungl-bot> Tree closed by buildbot at chromium dot org: Tree is closed (Automatic: "test_shell_tests" on "Webkit" from 54100: akalin at chromium dot org)
[08:24:31] *** arv__ has joined #chromium
[08:25:09] *** akalin2 has joined #chromium
[08:25:10] <akalin2> hey
[08:25:16] <akalin2> is the build failure a flake?
[08:25:39] <akalin2> boo, any sherriffs around
[08:26:02] <trungl-bot> Tree opened by akalin at chromium dot org: Tree is open (test_shell_tests failure looks like a flake)
[08:28:04] *** kjalarr has quit IRC
[08:29:02] *** arv__ has quit IRC
[08:30:32] *** kjalarr has joined #chromium
[08:32:54] *** arv__ has joined #chromium
[08:35:10] *** pdelgallego has joined #chromium
[08:44:07] *** FullFlannelJacke has quit IRC
[08:44:13] *** Erkan_Yilmaz_ has quit IRC
[08:44:22] *** arv__ has quit IRC
[08:44:29] *** General13372 has joined #chromium
[08:47:06] *** General1337 has quit IRC
[08:47:13] <trungl-bot> Tree closed by buildbot at chromium dot org: Tree is closed (Automatic: "compile" on "Chromium Linux Builder x64" from 54102: thestig at chromium dot org (:leiz))
[08:48:01] *** FullFlannelJacke has joined #chromium
[08:59:34] <jochen__> leiz: ping?
[09:06:50] <leiz> jochen__: waiting for green
[09:07:27] <leiz> (and guten morgen!)
[09:10:28] *** paul_irish has quit IRC
[09:14:03] *** jianli has quit IRC
[09:15:21] *** star-affinity has quit IRC
[09:15:31] <trungl-bot> Tree opened by thestig at chromium dot org (:leiz): Tree is open (cycling green; CrOS still busted; continuing issues: http://bit.ly/bsyE8W)
[09:15:39] <jochen__> guten morgen :)
[09:16:29] *** peepsalot has quit IRC
[09:16:34] <terinjokes> excuse my english, but that looks German for "Good Morning" ;)
[09:16:47] <leiz> usually the next thing I do is ask where you speak english or not, but most people in munich do :)
[09:17:38] <terinjokes> (checked Google Translate, i guess that little bit of German roots are showing through)
[09:20:32] *** homata_ has joined #chromium
[09:20:54] *** Precea is now known as Precea[BNC]
[09:20:58] *** peepsalot has joined #chromium
[09:21:54] *** Beetny has joined #chromium
[09:22:27] *** 84XAAPVWF has quit IRC
[09:24:51] *** peepsalot has quit IRC
[09:26:30] *** ph0nk has quit IRC
[09:34:36] *** plantain has quit IRC
[09:37:37] *** paul_irish_ has quit IRC
[09:40:12] <MikeSmithW3C> ukai: is Roland Steiner on vacation or something? or just not using IRC these days?
[09:43:00] *** nirnimesh has quit IRC
[09:43:27] <terinjokes> if i want to be working on the developer tools, is there a certain project i should be building (instead of building all)
[09:47:18] *** tittiatcoke has joined #chromium
[09:50:06] *** trungl has quit IRC
[09:53:34] *** MikeSmithW3C has quit IRC
[09:56:10] *** MikeSmithW3C has joined #chromium
[09:57:50] *** mnissler has quit IRC
[09:58:16] *** Paddy_NI has joined #chromium
[10:06:15] *** FullFlannelJacke has quit IRC
[10:06:37] *** nickcarter_ has quit IRC
[10:08:33] *** maikmerten has joined #chromium
[10:09:58] *** JodaZ has quit IRC
[10:13:08] *** JodaZ has joined #chromium
[10:13:28] *** path[l] has quit IRC
[10:13:30] *** mnissler has joined #chromium
[10:13:30] *** ChanServ sets mode: +v mnissler
[10:15:19] *** FullFlannelJacke has joined #chromium
[10:19:37] *** akalin2 has quit IRC
[10:22:35] *** General13372 has quit IRC
[10:24:53] *** victorw has quit IRC
[10:26:32] *** Paddy_NI has quit IRC
[10:30:41] *** ROBOd has joined #chromium
[10:31:27] *** hrna has joined #chromium
[10:35:52] *** path[l] has joined #chromium
[10:40:14] *** fta has joined #chromium
[10:40:21] *** fta has joined #chromium
[10:43:09] *** pomaxa has joined #chromium
[10:45:49] *** ejat has joined #chromium
[10:48:54] *** estes has quit IRC
[10:53:36] *** path[l] has quit IRC
[10:54:01] *** path[l] has joined #chromium
[10:55:27] *** fta_ has joined #chromium
[10:56:05] *** leeight has joined #chromium
[10:56:12] *** leeight has left #chromium
[10:56:15] *** FullFlannelJacke has quit IRC
[10:56:50] *** fta has quit IRC
[10:56:56] *** fta_ is now known as fta
[11:03:42] *** rohitrao has quit IRC
[11:03:42] *** JakeSays has quit IRC
[11:03:58] *** JakeSays has joined #chromium
[11:04:40] *** up365 has quit IRC
[11:05:22] *** up365 has joined #chromium
[11:05:47] *** ph0nk has joined #chromium
[11:08:27] *** chrisccoulson has joined #chromium
[11:10:49] *** FullFlannelJacke has joined #chromium
[11:12:53] *** up365 has quit IRC
[11:26:16] *** up365 has joined #chromium
[11:27:27] <anicolao> any webkit reviewers here?
[11:27:43] *** glider has joined #chromium
[11:27:43] *** ChanServ sets mode: +v glider
[11:28:52] *** fedorauser162 has quit IRC
[11:29:41] <glider> good morning, trungl-bot!
[11:29:41] <trungl-bot> glider: Good morning!
[11:30:16] <anicolao> glider: good morning.
[11:33:08] <glider> Is anyone taking care of the waterfall at the moment?
[11:33:27] <anicolao> glider: you mean the fact that linux & windows are red on convolution-crash.svg?
[11:34:00] <glider> anicolao: maybe
[11:34:08] *** gavin has quit IRC
[11:34:20] *** romainhuet has joined #chromium
[11:34:28] *** gavin has joined #chromium
[11:34:29] *** gavin has joined #chromium
[11:34:39] <glider> anicolao: I've just came and wanted to know if any bots are taken by other sheriffs
[11:35:15] <anicolao> glider: I haven't seen anybody in the channel for several hours, so far as I know you are the only active sheriff
[11:35:30] <glider> anicolao: oh, I see
[11:35:37] <anicolao> glider: are you also a webkit reviewer?
[11:36:41] <glider> anicolao: nope unfortunately
[11:37:15] <anicolao> glider: OK well once a reviewer comes along it should be possible to fix webkit with the patch on https://bugs.webkit.org/show_bug.cgi?id=43182
[11:37:34] <anicolao> [at least, that will fix convolution-crash.svg layout]
[11:38:15] <glider> anicolao: great, thank you!
[11:38:45] *** [Nef] has quit IRC
[11:39:04] <anicolao> glider: well as I broke it I'm not sure I deserve thanks. unfortunately for me it is almost 6AM so I need to call it a day, but tomorrow I will investigate why this test needs different pixel results for linux.
[11:39:52] <anicolao> glider: for now submitting any working baseline is the right answer since the point of the test is to validate a fix for a crash bug, and that it does no matter what the pixels look like.
[11:41:02] <glider> anicolao: okay, I'll tell that to any Webkit reviewers as soon as they arrive
[11:41:34] <glider> anicolao: is there a Chromium patch at codereview.chromium.org fixing that?
[11:42:05] <anicolao> glider: no it's a webkit patch.
[11:42:32] *** MikeSmithW3C has quit IRC
[11:53:06] <trungl-bot> Tree opened by glider at chromium dot org: Tree is open (cycling green; CrOS still busted; continuing issues: http://bit.ly/bsyE8W, Reliability -> need help)
[11:55:52] *** path[l] has quit IRC
[11:59:10] *** ukai has quit IRC
[12:00:14] *** [Nef] has joined #chromium
[12:01:41] *** fta_ has joined #chromium
[12:02:59] *** path[l] has joined #chromium
[12:03:12] *** fta has quit IRC
[12:03:25] *** fta_ is now known as fta
[12:08:02] *** d0k has joined #chromium
[12:20:02] *** fsamuel_ has joined #chromium
[12:20:53] *** [Nef] has left #chromium
[12:21:03] *** jeremyselier has joined #chromium
[12:31:44] *** hebz0rl has joined #chromium
[12:33:57] *** homata_ has quit IRC
[12:39:40] *** awolfson has joined #chromium
[12:45:11] *** Adys has quit IRC
[12:46:20] *** Adys has joined #chromium
[12:46:35] *** monreal has joined #chromium
[12:50:44] *** oal has joined #chromium
[12:51:49] <oal> Any chances we'll see functionality like Opera has, to put tab bar and navigation down to the bottom of the browser?
[12:52:17] <selckin> no
[12:52:36] <oal> Ok
[12:52:37] *** pdelgallego has quit IRC
[12:54:07] *** tbassetto has joined #chromium
[12:58:28] *** exxe has joined #chromium
[12:59:01] *** pdelgallego has joined #chromium
[13:11:54] *** fta_ has joined #chromium
[13:13:14] *** fta has quit IRC
[13:13:25] *** fta_ is now known as fta
[13:19:56] *** n00p has quit IRC
[13:20:22] *** n00p has joined #chromium
[13:23:48] *** Emperorlou has joined #chromium
[13:24:41] *** Emperorlou has left #chromium
[13:40:35] *** andrix has joined #chromium
[13:43:36] *** hebz0rl has quit IRC
[13:44:44] *** tonikitoo has quit IRC
[13:45:18] <ebraminio> hi everybody, ctrl+drag cannot clone a text in chromium, there is any chance to fix it?
[13:52:22] <trungl-bot> Tree closed by glider at chromium dot org: Tree is closed (CrOS still busted; continuing issues: http://bit.ly/bsyE8W)
[13:55:39] *** __ed has quit IRC
[14:00:06] *** awolfson has quit IRC
[14:05:34] *** tittiathome has joined #chromium
[14:05:50] *** tittiatcoke has quit IRC
[14:09:13] *** Aferlak12 has joined #chromium
[14:11:42] *** bauerb has joined #chromium
[14:11:42] *** ChanServ sets mode: +v bauerb
[14:16:06] *** tittiathome is now known as tittiatcoke
[14:26:27] *** jrmuizel has joined #chromium
[14:27:23] *** aroben has joined #chromium
[14:27:23] *** aroben has joined #chromium
[14:32:56] *** tonikitoo has joined #chromium
[14:34:22] *** d0k has quit IRC
[14:40:10] *** zloidemon has joined #chromium
[14:40:36] *** d0k has joined #chromium
[14:41:57] *** zloidemon has quit IRC
[14:47:52] *** jrmuizel has quit IRC
[14:55:23] *** zloidemon has joined #chromium
[14:55:25] *** satish_ has joined #chromium
[14:55:39] <satish_> anyone looking at the webkit chromium port build failure?
[14:56:05] <satish_> http://build.webkit.org/console?category=core - the failure happens from my patch onwards, but it looks unrelated to my change. any suggestions how to find/fix?
[14:57:04] *** ejat has quit IRC
[14:58:54] *** lorenzo has joined #chromium
[15:01:35] *** __ed has joined #chromium
[15:04:52] <jorlow> glider: what's up witht he tree closure?
[15:05:20] <anicolao> ok I'm awake again, did the convolution-crash fix get put int?
[15:05:39] <jorlow> anicolao: yeah...but it wasn't quite correct
[15:05:43] <jorlow> so i had to add another baseline
[15:05:45] <jorlow> but it's rolled in
[15:06:19] <anicolao> jorlow: oh :-( sorry
[15:06:26] <jorlow> no big deal
[15:06:38] *** thomasvl has joined #chromium
[15:06:38] *** ChanServ sets mode: +v thomasvl
[15:07:48] <anicolao> jorlow: well I'm annoyed that I made an error in the first place, frustrated that my fix didn't work, and unclear on why this pixel test should vary at all when I can't find the difference with pixie, and operating on just under 4h of sleep. So it seems like a bigger deal just now but I'm sure I"ll feel better later :)
[15:08:33] <jorlow> ha...well, my suggestion is to take a nap  :-)
[15:09:22] <mnissler> anyone working on the official builder breakage?
[15:09:59] <jorlow> does anyone here know how to kick a webkit.org builder?
[15:10:06] <jorlow> (i.e. not a canary)
[15:14:22] *** fta_ has joined #chromium
[15:14:41] <anicolao> jorlow: the fallback order is linux -> win -> mac?
[15:14:59] <jorlow> linux looks at win
[15:15:13] *** dmaclach has joined #chromium
[15:15:17] <jorlow> neither look at chromium mac
[15:15:26] <jorlow> maybe not even upstream mac
[15:15:27] <jorlow> not even sure
[15:15:29] <jorlow> it's complex
[15:15:33] *** fta has quit IRC
[15:15:37] *** fta_ is now known as fta
[15:15:44] <jorlow> i just do what the rebaseline tool tells me to  :-)
[15:16:23] <glider> (back)
[15:16:25] <anicolao> jorlow: :-D OK. where should i read to learn to use the rebaseline tool?
[15:16:46] <jorlow> http://trac.webkit.org/wiki/Rebaseline
[15:17:10] <anicolao> jorlow: thanks for the link & for cleaning up after me.
[15:17:21] <jorlow> no worries
[15:17:29] <glider> jorlow: started reverting r53988
[15:17:59] <glider> jorlow: sorry for the delay -- was trying to figure out what's wrong
[15:18:07] <jorlow> glider: no prob
[15:18:21] <jorlow> wow..it's been failing that long?
[15:20:22] <hwennborg> jorlow: the prototype-inheritance.html layout test is now failing after my patch. would you like me to do the suppression?
[15:21:14] *** lorenzo has quit IRC
[15:21:14] <jorlow> hwennborg: where can i see the failure?
[15:21:30] <hwennborg> on the chromium waterfall, webkit linux fails
[15:22:14] <jorlow> yes...please add to the test_expectations downstream
[15:22:25] <jorlow> lets not leave it in for more than a day or two tho
[15:22:33] <jorlow> er...ideally less
[15:22:38] <jorlow> that's a very important test
[15:23:03] <hwennborg> ok, will make it quick
[15:26:13] *** lorenzo has joined #chromium
[15:26:52] <mnissler> any sheriff in? I suggest a revert of r53969 (if possible) since that seems to have caused the official builders breakage, at least on mac.
[15:28:35] *** dmaclach has quit IRC
[15:28:58] *** leeight has joined #chromium
[15:29:08] *** leeight has left #chromium
[15:30:15] <jorlow> mnissler: glider is it
[15:31:18] <glider> mnissler: are you sure that's 53969?
[15:34:14] <mnissler> well, I have a local official build that compiles before 53969, but not after...
[15:34:26] <mnissler> but maybe a better fix is to just fix the include...
[15:34:35] <mnissler> glider: let me test that first...
[15:34:58] <glider> mnissler: yes, please
[15:37:15] *** nssy has joined #chromium
[15:38:15] *** awidegreen has joined #chromium
[15:40:00] *** nssy has quit IRC
[15:42:42] *** innervision has joined #chromium
[15:42:49] <innervision> hello everyone
[15:46:50] *** nssy has joined #chromium
[15:47:18] *** pinkerton has joined #chromium
[15:47:19] *** ChanServ sets mode: +v pinkerton
[15:47:56] <mnissler> glider: http://codereview.chromium.org/3032037
[15:48:22] <mnissler> glider: That restores the build for me, but I've only compiled the chrome app, not all tests...
[15:48:24] *** anicolao has quit IRC
[15:48:49] <mnissler> glider: doing so at the moment, will report back.
[15:49:51] <innervision> hey, just a question. does anyone get this problem of the TEXTAREA field being *really* slow when editing a text with linebreaks?
[15:50:29] <glider> mnissler: I've just reverted another CL, so let's wait till the change propagates
[15:53:47] *** awidegreen has quit IRC
[15:56:07] *** path[l] has quit IRC
[15:56:55] <mnissler> glider: complete mac official build suceeded.
[15:56:58] <terinjokes> chrome isn't building here
[15:57:09] <terinjokes> (mac build)
[15:57:15] <mnissler> glider: I you want this submitted, ping me.
[15:57:21] *** anicolao has joined #chromium
[15:57:21] *** ChanServ sets mode: +v anicolao
[15:57:53] *** anicolao_ has joined #chromium
[15:57:54] *** path[l] has joined #chromium
[15:58:29] *** dmaclach has joined #chromium
[15:58:29] *** ChanServ sets mode: +v dmaclach
[15:59:31] *** mnissler_ has joined #chromium
[15:59:49] *** awolfson has joined #chromium
[16:00:22] *** fta has quit IRC
[16:00:44] *** path[l] has quit IRC
[16:01:38] *** ChrisPartridge has quit IRC
[16:01:39] *** anicolao has quit IRC
[16:02:24] *** fta has joined #chromium
[16:02:26] *** path[l] has joined #chromium
[16:02:43] *** xdgc has joined #chromium
[16:02:51] *** path[l] has quit IRC
[16:04:14] *** path[l] has joined #chromium
[16:07:09] *** dmaclach_ has joined #chromium
[16:07:25] *** dmaclach__ has joined #chromium
[16:09:55] *** shreyas_ has joined #chromium
[16:10:14] *** dmaclach___ has joined #chromium
[16:10:14] *** dmaclach has quit IRC
[16:10:14] *** ChanServ sets mode: +v dmaclach___
[16:10:14] *** dmaclach___ is now known as dmaclach
[16:10:18] *** cypriendecouty has joined #chromium
[16:10:37] *** innervision has left #chromium
[16:10:39] *** rsesek has joined #chromium
[16:10:39] *** ChanServ sets mode: +v rsesek
[16:10:47] *** dmaclach_ has quit IRC
[16:12:16] *** stalled has quit IRC
[16:12:19] *** cypriendecouty has quit IRC
[16:12:38] *** paul_irish has joined #chromium
[16:12:49] *** malavv has quit IRC
[16:12:52] *** malavv has joined #chromium
[16:12:53] <xdgc> Hi, I have a weird problem with a popup window for an extension I'm writing: with every whitespace in the html the popup window gets bigger, vertically. Is this a know problem? (I couldn't find anything on that on Google)
[16:13:53] *** dmaclach__ has quit IRC
[16:14:03] *** ejat has joined #chromium
[16:14:03] *** ejat has joined #chromium
[16:17:05] <xdgc> For instance, when I write <span>one two three</span> the popup window size will increase. But not with <span>one&nbsp;two&nbsp;three</span>.
[16:18:19] *** anicolao has joined #chromium
[16:20:06] *** nebula_ has joined #chromium
[16:21:13] *** Emperorlou has joined #chromium
[16:21:16] *** anicolao_ has quit IRC
[16:21:28] *** anicolao_ has joined #chromium
[16:22:02] <xdgc> OK, why is it that I always find the problem as soon as I ask people on IRC? The weird behaviour was caused by a missing doctype.
[16:22:04] *** dhollowa_ has joined #chromium
[16:23:06] *** anicolao has quit IRC
[16:23:50] *** stalled has joined #chromium
[16:25:44] *** Emperorlou has quit IRC
[16:26:14] *** jrmuizel has joined #chromium
[16:33:31] *** Kunalagon has joined #chromium
[16:37:03] *** thakis_ has joined #chromium
[16:37:03] *** ChanServ sets mode: +v thakis_
[16:37:59] *** seventh has quit IRC
[16:41:41] *** nebula_ has quit IRC
[16:41:56] *** Sonderblade has joined #chromium
[16:42:15] *** stuartmorgan has joined #chromium
[16:42:16] *** ChanServ sets mode: +v stuartmorgan
[16:43:40] *** dhollowa_ has quit IRC
[16:48:06] *** xdgc has quit IRC
[16:50:43] *** rohitrao has joined #chromium
[16:50:43] *** ChanServ sets mode: +v rohitrao
[16:51:45] *** paul_irish has quit IRC
[16:52:25] *** fta_ has joined #chromium
[16:53:03] *** erikkay has joined #chromium
[16:53:03] *** ChanServ sets mode: +v erikkay
[16:53:05] <malavv> Wohh, when did we start needing the DirectX sdk?
[16:53:05] *** mnissler_ has quit IRC
[16:53:36] *** fta has quit IRC
[16:53:51] *** fta_ is now known as fta
[16:55:33] *** MikeSmithW3C has joined #chromium
[16:55:44] <pcgod> http://groups.google.com/a/chromium.org/group/chromium-dev/browse_thread/thread/731c5e5790671c03
[16:56:44] *** BHSPitCSP has joined #chromium
[16:56:45] *** mrossetti has joined #chromium
[16:57:18] <BHSPitCSP> Is there a way for me to see outgoing POST data in chromium when I submit a form?
[16:57:24] *** Adys has quit IRC
[16:57:35] <BHSPitCSP> Trying to automate some things
[16:58:46] *** Adys has joined #chromium
[17:00:12] *** nssy has quit IRC
[17:00:29] <terinjokes> BHSPitCSP: is it not in the developer tools?
[17:00:48] <BHSPitCSP> terinjokes: if it is, I'm asking how :)
[17:00:55] <jorlow> whoa....the tree is sad
[17:01:09] <BHSPitCSP> part of the problem is that this form submission doesn't return a new page, it returns a download
[17:01:12] <terinjokes> BHSPitCSP: resources tab, the name of the page you're loading
[17:01:27] <BHSPitCSP> and you can't inspect a page that doesn't actually appear
[17:01:28] <terinjokes> there should still be a header sent
[17:01:34] <terinjokes> what's the URI?
[17:01:35] <glider> jorlow: I've committed the fix.. probably
[17:01:54] <BHSPitCSP> right, but the sent headers it shows is for the page that is up
[17:02:17] <BHSPitCSP> chromium doesn't open a new tab or show a blank page with the download, it just downloads
[17:02:25] <jorlow> sweet
[17:02:28] <BHSPitCSP> (It's an export operation in phpmyadmin)
[17:02:30] <jorlow> looking forward to the pig turning green
[17:02:54] <terinjokes> BHSPitCSP: haha, you can automate using SQL
[17:02:56] <terinjokes> :P
[17:03:11] *** e-jat has joined #chromium
[17:03:11] *** e-jat has joined #chromium
[17:03:15] <BHSPitCSP> terinjokes: not if it's a database you only have access to via PMA.
[17:03:34] <terinjokes> BHSPitCSP: sorry, my sarcasm came out
[17:03:44] <BHSPitCSP> oh :P
[17:04:09] *** Engin has quit IRC
[17:04:20] <terinjokes> BHSPitCSP: i just tried a file download, and I can see the headers
[17:04:54] *** ejat has quit IRC
[17:05:26] *** trungl has joined #chromium
[17:05:26] *** ChanServ sets mode: +v trungl
[17:05:27] *** paul_irish has joined #chromium
[17:06:34] <BHSPitCSP> weird, it worked the second time
[17:06:46] <terinjokes> yay!
[17:13:20] *** lorenzo has quit IRC
[17:13:58] *** path[l] has quit IRC
[17:14:43] <terinjokes> ok, trying to get bootstrapped, and getting confused about what each of these devtools_ files actually does, can I get a quick rundown?
[17:17:35] <trungl> 'morning, Chromium
[17:17:35] <trungl-bot> trungl: Good morning!
[17:17:47] *** ph0nk has quit IRC
[17:19:02] *** Emperorlou has joined #chromium
[17:20:16] *** rfulton has joined #chromium
[17:20:58] *** Emperorlou has left #chromium
[17:21:11] *** solsTiCe has joined #chromium
[17:21:35] <terinjokes> i think devtools_http_protocol_handler is what i'm interested in
[17:22:03] <solsTiCe> hi. I don't know if it's a chromium bug or www.cuil.com bug but middle click open link in same tab instead of opening a new one
[17:22:30] <rfulton> is there a way to build the visual studio project files for just one project? I've been trying to build skia alone, but I'm not getting anywhere. I have the /src/skia files, the necessary /src/base and /src/build files, and the /src/third_party/skia files, but I can't get gclient to process the /src/skia gyp.
[17:25:18] <pcgod> solsTiCe: http://code.google.com/p/chromium/issues/detail?id=255
[17:26:25] *** n00p has quit IRC
[17:26:25] *** n00p has joined #chromium
[17:27:59] *** thomasvl has quit IRC
[17:28:14] <solsTiCe> ok
[17:29:35] <solsTiCe> ah webkit upstream bug ?
[17:30:15] *** dave_levin has quit IRC
[17:30:36] *** solsTiCe has left #chromium
[17:31:52] *** rubenbb has quit IRC
[17:33:18] <brettw> glider: I'm also sheriff today, let me know if you need anything.
[17:33:20] *** rubenbb has joined #chromium
[17:33:24] <brettw> Looks like your revert broke something?
[17:33:48] <glider> brettw: oh, nice
[17:34:11] <jorlow> dglazkov: ping
[17:34:14] *** markmentovai has joined #chromium
[17:34:14] *** ChanServ sets mode: +v markmentovai
[17:34:50] *** mnissler_ has joined #chromium
[17:35:29] <glider> brettw: I've fixed the breakage
[17:36:28] <glider> brettw: but reverting that CL didn't fix CrOS
[17:36:49] <glider> I'm trying 53990 now
[17:36:52] *** tittiatcoke has quit IRC
[17:42:17] *** trungl has quit IRC
[17:43:26] *** Venom_X has joined #chromium
[17:44:33] <brettw> glider, ok
[17:44:48] *** thakis_ has quit IRC
[17:45:01] <dglazkov> good morning, Chromium!
[17:45:01] <trungl-bot> dglazkov: Good morning!
[17:45:05] <dglazkov> jorlow: png
[17:45:23] <jorlow> dglazkov: http://test-results.appspot.com/dashboards/flakiness_dashboard.html#showExpectations=true&useWebKitCanary=true&tests=svg%2Fcustom%2Fuse-events-crash.svg%2Csvg%2Fhixie%2Ferror%2F017.xml%2Csvg%2Fcustom%2Fgradient-cycle-detection.svg%2Csvg%2Fcustom%2Frecursive-clippath.svg%2Csvg%2Fcustom%2Frecursive-filter.svg%2Csvg%2Fcustom%2Frecursive-gradient.svg%2Csvg%2Fcustom%2Frecursive-mask.svg%2Csvg%2Fcustom%2Frecursive-pattern.svg
[17:45:25] <jorlow> third down
[17:45:30] <jorlow> and fourth
[17:45:38] <jorlow> looks like the checked in baselines are just plain wrong
[17:45:47] <jorlow> i thought we were never supposed to do that...?
[17:47:49] <dglazkov> oops! :)
[17:49:13] <jorlow> k...just double checking
[17:49:58] *** e-jat has quit IRC
[17:51:56] *** dmaclach has quit IRC
[17:52:15] *** path[l] has joined #chromium
[17:53:18] *** paul_irish has quit IRC
[17:54:27] *** thomasvl has joined #chromium
[17:54:27] *** ChanServ sets mode: +v thomasvl
[18:00:25] *** Engin has joined #chromium
[18:01:26] *** RT|Chatzilla has quit IRC
[18:02:58] <glider> brettw: 53990 didn't help as well
[18:03:40] *** mnissler_ has quit IRC
[18:03:43] <jorlow> dglazkov:  https://bugs.webkit.org/show_bug.cgi?id=43194
[18:04:56] <brettw> glider: what problem are you trying to solve?
[18:05:12] *** paul_irish has joined #chromium
[18:05:13] *** ChanServ sets mode: +v paul_irish
[18:05:18] *** hasimo has quit IRC
[18:05:38] *** hasimo has joined #chromium
[18:05:49] *** rfulton has quit IRC
[18:05:56] *** Engin has quit IRC
[18:06:03] <glider> brettw: well, the CrOS bot started failing somewhere between 53988 and 53994
[18:06:26] <brettw> glider: oh, the UI tests
[18:06:31] *** Engin has joined #chromium
[18:06:41] *** victorw has joined #chromium
[18:06:41] *** ChanServ sets mode: +v victorw
[18:07:23] *** trungl_afk is now known as trungl
[18:07:29] <brettw> glider: I was just looking at the JavascriptHistory test on the Vista Tests (UI test), and it looks like it's been flaky a while. Am I reading the flakiness dashboard properly?
[18:08:00] <markmentovai> glider: the 54137 revert turned mac compile red?
[18:08:10] <markmentovai> so now we have broken chromeos and broken mac too?
[18:08:22] <glider> markmentovai: yes, sort of
[18:08:29] <glider> markmentovai: reverting it back
[18:08:34] *** arv__ has joined #chromium
[18:08:35] <markmentovai> rerevert
[18:08:39] *** victorw1 has joined #chromium
[18:08:39] *** ChanServ sets mode: +v victorw1
[18:08:41] <rohitrao> reunrevert
[18:09:59] <brettw> I'm not sure the CrOS problems are checkin related
[18:10:21] <brettw> it says Cros library has not been loaded and then fails doing simple stuff (initial loads)
[18:10:25] <trungl> unrevert?
[18:11:02] <mnissler> markmentovai: any chance you could take a quick look at http://codereview.chromium.org/3032037/show?
[18:11:09] <mnissler> markmentovai: it fixes the mac official build for me.
[18:11:17] *** BCalvignac2 has quit IRC
[18:11:43] *** froek has joined #chromium
[18:11:44] <glider> brettw: looks like the JS test is flaky
[18:12:15] *** lorenzo has joined #chromium
[18:12:26] *** vt100 has quit IRC
[18:12:30] <froek> Chromium team, looking for help in --show-eula commandline param.  When I run the mini_installer.exe with this, I get an empty blank modal dialog.
[18:13:13] <froek> Inside ShowEulaDialog(), it calls GetLocalizedEulaResource() which returns a res:// to the resource it is supposed to load
[18:13:25] <markmentovai> mnissler: i will look
[18:13:35] <mnissler> markmentovai: thanks!
[18:13:54] <froek> I've verified that it returns the res:// correctly, but when it calls EulaHTMLDialog::ShowModal it just paints a blank window.
[18:14:26] <brettw> glider: I'll mark that JS test and then bug some CrOS people
[18:14:31] *** johnny_g has joined #chromium
[18:15:35] *** agl has joined #chromium
[18:15:36] *** ChanServ sets mode: +v agl
[18:16:23] *** dhg has joined #chromium
[18:16:23] *** ChanServ sets mode: +v dhg
[18:16:30] <agl> glider: the tree is still closed? I'll revert my Skia roll unless you have plan.
[18:16:45] <agl> glider: (personally, I would revert everything in the window)
[18:17:02] <dhg> maybe there's something wrong with the bots?
[18:17:12] *** vt100 has joined #chromium
[18:17:21] <markmentovai> eek! the official build is broken! why didn?t anyone tell me? :)
[18:17:49] <dhg> i'm not sure that reverting is the answer.
[18:17:57] <dhg> but i could be wrong
[18:18:04] <terinjokes> I'm looking to see how the the Developer Tools works, if anyone has a moment to help me out
[18:18:22] <dhg> i talked to glider about this earlier... so maybe i missed to conversation in here about it
[18:18:27] <dhollowa> markmentovai: you mean you didn't hear miranda cursing yesterday? ;-)
[18:18:34] <dhg> but this is appearing on every failing test on cros:
[18:18:43] <dhg> "??chrome/installer/util/master_preferences.cc(167)] Master preferences file not found: /b/slave/chromium-dbg-linux-chromiumos/build/src/out/Debug/master_preferences"
[18:18:54] <markmentovai> i thought she was on the branch, and was cursing about something else
[18:19:04] <markmentovai> (but yes, i heard it 2,800mi away)
[18:19:14] <markmentovai> (what a set of pipes on that one!)
[18:19:29] *** craigsch has joined #chromium
[18:19:38] <dhollowa> markmentovai: yes, indeed.  branch cursing.  that's my fate today.
[18:22:03] *** bent-mozilla has joined #chromium
[18:22:16] *** Engin has quit IRC
[18:22:41] *** Engin has joined #chromium
[18:23:12] *** victorw has quit IRC
[18:25:43] *** ejat has joined #chromium
[18:25:43] *** ejat has joined #chromium
[18:26:00] <glider> dhg: http://build.chromium.org/buildbot/waterfall/builders/Linux%20Builder%20%28ChromiumOS%20dbg%29/builds/204/steps/ui_tests/logs/stdio is the lkgr in CrOS dbg
[18:26:15] <glider> there are tons of master_preferences errors, but the tests pass
[18:26:38] <dhg> well
[18:26:41] <dhg> nevermind :)
[18:27:05] <dhg> its odd.
[18:28:13] *** fqian has joined #chromium
[18:28:50] *** fta_ has joined #chromium
[18:29:07] *** andybons_ has joined #chromium
[18:29:24] *** EnginA has joined #chromium
[18:29:59] *** agl has quit IRC
[18:30:44] <glider> agl: we haven't got a plan yet
[18:31:13] <glider> something has affected server_->WaitForInitialLoads(), so it times out now
[18:31:14] *** fta has quit IRC
[18:31:19] *** fta_ is now known as fta
[18:31:37] *** taf2 has joined #chromium
[18:32:18] <brettw> glider: I kind of think we should restart the bot
[18:32:49] *** Engin has quit IRC
[18:33:44] <brettw> does anybody know how to do that?
[18:33:46] <glider> brettw: can we just reboot the slave?
[18:33:50] <brettw> yes, I want to do that
[18:34:18] *** xiyuan has joined #chromium
[18:34:19] *** ChanServ sets mode: +v xiyuan
[18:34:27] <brettw> Do we need to talk to sombody magic like nsylvain? I haven't tweaked the bots in 2 years.
[18:34:30] <rohitrao> brettw, glider: we restarted the bots last night
[18:34:35] <rohitrao> didn't seem to help
[18:35:30] <glider> rohitrao: is there a log of other actions you've tried?
[18:35:35] <brettw> rohitrao, did we try a clobber on that?
[18:35:44] <brettw> Maybe that will fix the "unable to load llibcros" problem
[18:35:55] <rohitrao> no idea, I just asked maruel to kick the bots
[18:35:59] <rohitrao> not sure what he actually did
[18:36:09] <rohitrao> so if you look at the CL description for one of those changes
[18:36:12] <glider> brettw: http://build.chromium.org/buildbot/waterfall/builders/Linux%20Builder%20%28ChromiumOS%20dbg%29/builds/204/steps/ui_tests/logs/stdio is again the lkgr
[18:36:20] <rohitrao> it says something like "this won't fix the libcros problem"
[18:36:28] <rohitrao> so I think the libcros error message is a red herring
[18:36:31] <andybons_> who should i bug about merges? kerz?
[18:36:39] <glider> brettw: and there are tons of libcros and everything worked
[18:36:44] <rohitrao> andybons: yeah, kerz and/or laforge
[18:36:45] <markmentovai> andybons_: you can notify kerz
[18:36:54] <thakis> if someone feels like reviewing a valgrind cl: http://codereview.chromium.org/2808079/show
[18:36:56] <brettw> glider, okay, that's kind of stupid :(
[18:37:23] <brettw> seems like somebody needs to log on to the bot and see what's up
[18:37:25] <thakis> does an s/it's/its/ do, so if you're into language, it's the PERFECT cl for you
[18:38:02] <rohitrao> thakis: unnecessary apostrophes cause memory leaks?
[18:38:14] <markmentovai> thakis: you missed an it's
[18:38:28] <nshkrob>
[18:38:31] <thakis>
[18:38:40] <rohitrao> you're both wrong
[18:38:46] <thakis>
[18:39:22] <glider> thakis: until this cros hell happened I was writing a letter about the suppressions I consider unnecessary
[18:39:25] * dglazkov is tempted to join this fascinating conversation
[18:39:27] <markmentovai> thakis: #include guards in that .h are wrong too
[18:40:13] <andybons_> pinkerton: my favorite rev didn't get in when they cut. i was so upset when i didn't see my white highlights under the tabs.
[18:40:23] <glider> thakis: "obj:*" is fragile -- why keep it at the bottom?
[18:40:34] <thakis> landmine get!
[18:40:35] <dhg> brettw: the unable to load libcros thing is fine... expected
[18:40:37] *** BCalvignac has joined #chromium
[18:40:50] <pinkerton> andybons_: i don't think you asked me to merge that one, or if you did, then drover lied to me
[18:41:21] <glider> thakis: and removing "obj:*" makes you remove the ellipsis from the bottom as well
[18:41:22] <andybons_> pinkerton: people make mistakes... ::cough:: you hack ::cough::
[18:41:37] *** BCalvignac has quit IRC
[18:41:39] <pinkerton> :|
[18:41:58] <rohitrao> thems fighting words
[18:42:07] *** rniwa has joined #chromium
[18:42:23] <thakis> rsesek: do we really want a suppression for everything that's statically initialized?
[18:42:48] <glider> so, is someone in a mood of looking at that cros bot?
[18:43:04] <glider> it's almost nine pm in msk
[18:43:14] *** markusheintz_ has quit IRC
[18:43:25] <rsesek> thakis: you tell me; you're removing the suppression. I guess the real person to ask is the one who put it there first.
[18:43:31] <mnissler> sheriffs: May I land a two-line fix for mac official? http://codereview.chromium.org/3032037/show
[18:43:43] <brettw> I'm going to try to log into the cros builder
[18:43:44] <andybons_> pinkerton: it was probably landed after i asked you to do it. no worries
[18:43:52] <thakis> rsesek: the suppression name suggests it's meant for the sandbox leak, so if this isn't enough there should be more focused suppressions
[18:44:01] <andybons_> pinkerton: but this time you're picking up the beer
[18:44:05] <pinkerton> andybons_: you've still ruined my self esteem for the day
[18:44:15] *** BCalvignac has joined #chromium
[18:44:29] <andybons_> pinkerton: coors helps with that, too.
[18:44:44] <pinkerton> it does. i have a 24pk waiting for me at home
[18:44:51] <glider> mnissler: I vote for it
[18:44:51] <pinkerton> vented wide mouth cans FTW
[18:45:20] *** hsuh has joined #chromium
[18:45:45] <andybons_> pinkerton: niiice
[18:46:26] <mnissler> glider: thanks! anyone else feel like voting?
[18:46:41] <thakis> pinkerton, andybons_: you're looking for #chromium-light-beer
[18:46:49] <brettw> anybody here ever log onto the buildbots? I don't have access to the password
[18:46:51] <rohitrao> hey, did anyone change anything on the official builders?
[18:46:56] <pinkerton> i don't drink coors light. what do you take me for.
[18:46:56] *** dhaffner has joined #chromium
[18:47:02] <rohitrao> wondering why WebKitPrefs started working all of a sudden
[18:47:13] <glider> brettw: I do have?
[18:47:16] <glider> oh
[18:47:21] <glider> brettw: I do have
[18:47:21] <thakis> regular coors counts as light beer too
[18:47:30] <glider> brettw: what do you need?
[18:47:37] <terinjokes> Anyone know how the Developer Tools work, and mind giving me a quick rundown?
[18:47:43] *** bauerb has left #chromium
[18:47:50] *** bauerb has joined #chromium
[18:47:50] *** ChanServ sets mode: +v bauerb
[18:47:53] <thakis> terinjokes: implementation-wise, or how they'r'e used?
[18:47:59] *** BCalvignac has quit IRC
[18:48:02] <andybons_> terinjokes: in what sense?
[18:48:07] *** bauerb has left #chromium
[18:48:45] *** bauerb has joined #chromium
[18:48:45] *** ChanServ sets mode: +v bauerb
[18:48:49] *** shreyas_ has quit IRC
[18:48:57] *** satish_ has quit IRC
[18:49:02] <hsuh> hey guys, I thought I had a memory leak because my interactive chart page was consuming more and more memory as I played with it - but then, when I reached almost 50mb (using the Heap snapshot feature) it stopped growing! does that have an explanation?
[18:49:35] <brettw> glider: either the password or for somebody to log on and look at it
[18:49:39] <terinjokes> implementation-wise, i want to see if I can't fix a bug, but with so many files, I gotta figure out how it works, and where to start ;)
[18:49:54] <brettw> glider, I suspect something is broken on the bot and reverting patches won't help
[18:49:59] *** BCalvignac has joined #chromium
[18:50:17] <rohitrao> brettw: but two different bots went down at once
[18:50:37] <rohitrao> not that doesn't mean the bots are broken, but it seems less likely :)
[18:50:51] <brettw> rohitrao, oh, both debug and release?
[18:50:51] <rohitrao> aren't
[18:51:09] <rohitrao> brettw: yeah, the two bots on the far right are both failing the same way
[18:51:20] <brettw> then maybe I'm wrong
[18:52:25] <markmentovai> what?s all of the xmlTextWriterWriteDocCallback error stuff in the cros test logs? red herring?
[18:52:47] <hsuh> i thought perhaps the garbage collector only starts collecting when a page reaches some limit, to avoid frequent interruptions
[18:52:52] *** fta_ has joined #chromium
[18:52:58] <markmentovai> probably, i see them in the previous run too
[18:53:01] <brettw> rohitrao, the chromeos builders are just chromeos builds of Chrome running on LInux, right?
[18:53:18] <brettw> Maybe I should do a checkout
[18:53:33] <rohitrao> brettw: you might be describing linux views.  I'm not sure what the difference is
[18:53:57] *** fta has quit IRC
[18:54:04] *** fta_ is now known as fta
[18:54:20] *** shreyas_ has joined #chromium
[18:57:29] *** tonyg-cr has joined #chromium
[18:57:29] *** ChanServ sets mode: +v tonyg-cr
[18:59:07] *** glider has quit IRC
[18:59:38] *** Adys has quit IRC
[18:59:53] <brettw> I'm going to do a local chromeos build to see if I see the problem
[18:59:58] <brettw> might take a while though
[19:00:43] *** Adys has joined #chromium
[19:01:03] <thakis> brettw: ok to land http://codereview.chromium.org/2808079 ? it's mac only, and mac looks healthy
[19:01:20] <brettw> thakis, sure
[19:01:28] *** Erkan_Yilmaz has joined #chromium
[19:03:04] <mnissler> glider: mac official fix is in. mac build seems to be broken by the revert though, but I guess you're already on top of that.
[19:03:19] *** shreyas_ has quit IRC
[19:03:36] *** tbassetto has quit IRC
[19:03:51] <mnissler> glider: never mind, I didn't see the revert revert...
[19:03:56] <thakis> terinjokes: most of the devtools stuff is in javascript files in webkit (third_party/WebKit/WebCore/inspector/front-end), the chromium-related plumbing is mostly third_party/WebKit/WebKit/chromium/public/WebDevTools*.h (and ...chromium/src/WebDevTools*.cpp), from there use grep to see how it all connects to the brorwser process
[19:04:14] *** froek has left #chromium
[19:05:03] *** aroben is now known as aroben|lunch
[19:05:05] <terinjokes> thakis: thanks... i was looking in the devtools_ files
[19:05:20] <thakis> which devtools_ files?
[19:05:35] <thakis> sounds correct, too :-P
[19:06:05] <terinjokes> "chrome/browser/debugger/devtools_"
[19:06:38] <thakis> everything in chrome/browser is browser-process stuff. what i mentioned above is renderer-process stuff
[19:07:49] <thakis> terinjokes: the super-high-level overview is that devtools are implemented in javascript (which is in ...WebCore/inspector/frontend) and rendered by a webcore webview. there's some special js support so that the devtools can do things normal js can't
[19:08:43] *** jeremyselier has quit IRC
[19:10:29] <terinjokes> i'll take a look at that, and see what I can't figure out
[19:11:16] *** hebz0rl has joined #chromium
[19:11:29] *** johnny_g has quit IRC
[19:12:11] *** js2 has joined #chromium
[19:13:55] *** bweinstein has joined #chromium
[19:15:33] *** fqian has quit IRC
[19:19:03] *** nirnimesh has joined #chromium
[19:19:03] *** ChanServ sets mode: +v nirnimesh
[19:19:18] <victorw1> nsylvain: ping
[19:19:21] *** rniwa has left #chromium
[19:19:38] <nsylvain> pong :(
[19:19:43] *** fta_ has joined #chromium
[19:19:46] <nsylvain> something else broken?
[19:20:05] *** phajdan-jr has joined #chromium
[19:20:05] *** ChanServ sets mode: +v phajdan-jr
[19:20:06] <victorw1> nsylvain: we are having issue in webkit canary linux that causes test shell test fail. looks like it can't create temp dir
[19:20:28] <nsylvain> this is the same problem as yesterday
[19:20:33] <nsylvain> it happened on the main waterfall too
[19:20:45] *** fta has quit IRC
[19:20:47] *** bers has joined #chromium
[19:20:49] *** fta_ is now known as fta
[19:21:04] <nsylvain> evan changed the code yesterday to fail the test, instead of just deleting random files.
[19:21:38] <nsylvain> i'll look at the tmp dir on thsi machine
[19:21:47] <victorw1> nsylvain: thx
[19:21:57] <phajdan-jr> nsylvain: is the problem a lack of disk space?
[19:22:06] *** tittiatcoke has joined #chromium
[19:22:25] <nsylvain> phajdan-jr: no, not likely
[19:23:42] <phajdan-jr> nsylvain: I see. Please let me know if we should change the test infrastructure and/or write the tests in a slightly different way to make such issues less likely.
[19:26:00] <victorw1> anicolao_: hi
[19:27:04] <andybons_> pinkerton: merged
[19:27:13] *** evmar_afk is now known as evmar
[19:28:15] <nsylvain> phajdan-jr: Looks like we have a LOT of temp files leaked in the tmp dir.  And we have code to remove them, but it looks like the pattern changed
[19:28:36] <nsylvain> phajdan-jr: it used to be "org.chromium.org.ABCDF"
[19:28:49] <nsylvain> and now it's ".org.chromium.org.ABCDF"
[19:28:57] <phajdan-jr> nsylvain: maybe that started happening after I've made ui_tests use a temp directory as the user data directory?
[19:29:01] <evmar> that seems like a definite mistake, tony and i were discussing it yesterday
[19:29:07] <evmar> we should go back to the dotless one
[19:29:13] <nsylvain> phajdan-jr: maybe
[19:29:18] <nsylvain> yeah
[19:29:20] <thakis> rsesek: is that update notifier supposed to be a black dot?
[19:29:39] <evmar> commit 3a09e24e418feff9ba3f0d7783bbf64f23f9d9d6
[19:29:39] <evmar> Author: dhg at chromium dot org <dhg at chromium dot org@0039d316-1c4b-4281-b951-d872f2087c98>
[19:29:41] <nsylvain> and I always type rm -rf * , and this does not get those files
[19:29:43] <evmar> Date:   Thu Jun 10 17:11:29 2010 +0000
[19:29:45] <evmar> Changing temporary files on posix to have a . in front so they are hidden on linux systems (i.e. ch
[19:29:48] <evmar> romeos)
[19:30:05] <evmar> http://codereview.chromium.org/2662005
[19:30:21] <evmar> why are our change descriptions so useless :(
[19:31:11] <phajdan-jr> no rationale for the change, right?
[19:31:17] <evmar> http://code.google.com/p/chromium-os/issues/detail?id=3110
[19:31:17] <phajdan-jr> can we ping dhg?
[19:31:52] <phajdan-jr> ah, I see
[19:31:59] <dhg> whats up?
[19:32:14] <phajdan-jr> dhg: we're discussing the issue of temp files gumming up the bots
[19:32:40] <phajdan-jr> dhg: it seems like just removing the dot in fron of the temp dirs we create would revert a bugfix for chrome-os
[19:33:13] <evmar> it seems to me a cros bug that it creates a temporary file when you click "save"
[19:33:22] <phajdan-jr> nsylvain: what are the drawbacks of updating the code to remove temp files to also remove .org.chromium.... directories?
[19:33:36] <nsylvain> no drawback .. just need to be updated
[19:33:46] <evmar> we should not do it! hidden directories are stupid
[19:33:47] <phajdan-jr> evmar: if it's  download code... everything is possible
[19:33:56] <evmar> phajdan-jr: hahahahaha :)
[19:34:07] <nsylvain> drawback is that evan does not agree with that idea ;)
[19:34:09] <dhg> hmm, it would result in the user seeing the temp directories on downloads
[19:34:12] <dhg> which was a bug
[19:34:15] <nsylvain> And i really don't mind , as long as it's not a problem anymore ;)
[19:34:37] <brettw> I ran the ChromeOS tests locally and it works
[19:34:39] <nsylvain> (We should still update the script for 1 run, so all the machines get cleaned up automatically)
[19:34:50] <brettw> I mean the UI tests on a ChromeOS build
[19:34:53] <evmar> dhg: why does this temp file get created anyway?
[19:35:12] <phajdan-jr> brettw: it may be that it only gums up when you run tests a lot
[19:35:18] <dhg> lesee.. I think its the save as, it creates a temporary download name until its named by the user
[19:35:21] <phajdan-jr> evmar: maybe dangerous download handling
[19:35:25] <dhg> if i remember correctly
[19:35:57] <evmar> ugh, my /tmp is filled with hidden .org.chromium spew, i never saw it before :(
[19:36:00] <dhg> same thing with the "dangerous" downloads
[19:36:17] <phajdan-jr> now we have dangerous tmp files ;-)
[19:36:21] <dhg> haha
[19:36:27] <evmar> why don't we put them in /tmp or something?
[19:36:33] <dhg> i dunno
[19:36:35] <evmar> i guess there's the filesystem-crossing problem
[19:36:41] <phajdan-jr> evmar: it may be a different volume
[19:36:41] <evmar> hmmmmmm
[19:36:47] <phajdan-jr> evmar: yes, I think that's it
[19:38:43] <evmar> jeez, file_util has a bunch of tempfile functions, it seems way too complicated
[19:38:53] *** moblin_ has joined #chromium
[19:39:21] <moblin_> congratulations(?) on 50,000 bug reports googlers :)
[19:40:47] <evmar> phajdan-jr: https://cs/codesearch?q=Create\w*Temp&hl=en&lr=&c=chrome&sbtn=Search
[19:42:29] * evmar wonders why we have eight functions for temp files in file_util
[19:51:57] *** taf2 has quit IRC
[19:52:15] <jorlow> nsylvain: http://buildbot.jail.google.com//buildbot/waterfall.fyi/builders/XP%20Perf%20(webkit.org)/builds/10826/steps/gclient/logs/stdio
[19:52:30] <brettw> I just ran the ChromeOS UI tests again and they passed again. I think we should open the tree
[19:52:38] <brettw> I can look at the bot asynchronously
[19:52:42] <brettw> unless somebody else is?
[19:52:55] *** rniwa has joined #chromium
[19:52:55] *** ChanServ sets mode: +v rniwa
[19:53:40] *** jlouie has joined #chromium
[19:53:55] <brettw> any objections to opening the tree?
[19:54:52] <rsesek> thakis: it'll be yellow in the next release
[19:55:02] <brettw> what does the reliability bot error mean "partial result of current build failed"?
[19:55:12] <thakis> rsesek: ok. is the menu item in the enxt release too?
[19:55:26] <rsesek> thakis: the menu item should already be there
[19:55:50] <thakis> isn't for me in 6.0.472.4 dev
[19:56:21] <rsesek> hm
[19:56:24] * rsesek shrugs
[19:56:42] *** jamesr has joined #chromium
[19:56:42] *** ChanServ sets mode: +v jamesr
[19:57:26] <trungl-bot> Tree opened by brettw at chromium dot org: Tree is open BE NICE (Brett looking at CrOS; continuing issues: http://bit.ly/bsyE8W)
[19:59:48] *** maikmerten has quit IRC
[20:00:31] *** moblin_ has quit IRC
[20:00:43] *** hsuh has quit IRC
[20:01:43] *** fantasticulous has joined #chromium
[20:02:30] <rsesek> thakis: open a new window
[20:02:38] <rsesek> thakis: the new window should have the menu item
[20:02:40] <rsesek> that may be a bug
[20:02:42] <thakis> restarted
[20:02:44] <thakis> already
[20:03:06] <rsesek> darn
[20:03:45] <rjkroege> what's the best version of windows for chrome development -- i.e. which version are used for windows baselines?
[20:03:56] <rsesek> thakis: I'm pretty sure there's a bug that we don't tell the wrench menu to rebuild
[20:04:14] <rohitrao> andybons: tab_view.mm build errors in the official builders ;)
[20:04:20] <thakis> rsesek: want me to file?
[20:04:24] <rsesek> thakis: plz
[20:04:31] <evmar> rjkroege: unless you have a good reason not to you should just use win7
[20:04:36] <jlouie> Hi. I have a question concerning building differential installers (mini_installer.exe).
[20:04:37] <jlouie> How do the differential files (packed_files.txt, chrome.7z, setup.ex_) get included into the mini_installer.exe once create_archive_installer.py is run?
[20:04:38] <rohitrao> andybons: tab_view.mm:852: error: expected type-specifier before 'ViewID'
[20:04:51] <rjkroege> evmar: thanks
[20:05:10] *** estes has joined #chromium
[20:06:31] *** froek has joined #chromium
[20:08:17] *** fta has quit IRC
[20:08:23] *** fta has joined #chromium
[20:09:35] *** xji has joined #chromium
[20:09:35] *** ChanServ sets mode: +v xji
[20:10:48] *** apatrick_ has joined #chromium
[20:10:48] *** ChanServ sets mode: +v apatrick_
[20:11:51] *** thomasvl has quit IRC
[20:14:21] *** anicolao_ has quit IRC
[20:15:15] *** mnissler_ has joined #chromium
[20:16:45] <trungl-bot> Tree closed by buildbot at chromium dot org: Tree is closed (Automatic: "update" on "Chromium Linux Builder (dbg-shlib)")
[20:17:19] *** pinkerton has quit IRC
[20:17:53] <phajdan-jr> it breaks on my dev box too :(
[20:18:10] <phajdan-jr> grit_info.py returned exit status 1
[20:18:49] <terinjokes> everything seems to break here, but i just figured that was because I was on a mac
[20:19:06] <phajdan-jr> beng: could your change break grit?
[20:19:19] <phajdan-jr> beng: you have modified tools/grit/resource_ids, please take a look
[20:19:56] * phajdan-jr sets up a revert timer just in case
[20:20:17] <phajdan-jr> it's going to hork the tree a _lot_
[20:20:17] <xji> phajdan-jr: do you mean the build failed due to beng's change?
[20:20:28] <phajdan-jr> xji: that's extremely probable
[20:20:57] <markmentovai> yes, it is
[20:21:46] <markmentovai> beng: gfx/gfx_resources.grd missing from your chicken?
[20:22:38] <markmentovai> and maybe gfx/bitmap_brush_image.png too? no idea how those fell out, they were in your upload #3 on codereview, but are missing from #4
[20:23:00] <evmar> he checked in hte image separately iirc
[20:23:14] <phajdan-jr> I have pinged Ben on IM and he has just responded.
[20:23:26] <markmentovai> evmar: yes, you're right - it's just the .grd that's missing then
[20:23:26] <phajdan-jr> he's not on IRC :(
[20:24:14] *** beng_ has joined #chromium
[20:24:17] <beng_> hi
[20:24:25] <beng_> mmentovai: you had a thought?
[20:24:39] <xji> beng_, are you looking at the failures?
[20:24:43] <markmentovai> [2:25pm] markmentovai: beng: gfx/gfx_resources.grd missing from your chicken?
[20:24:50] <trungl-bot> Tree closed by phajdan.jr at chromium dot org (:phajdan-jr): Tree is closed (update horked -> beng, on it; ETA reopening 12:15 AM PDT)
[20:25:13] <beng_> oops
[20:25:15] <beng_> you're right
[20:25:21] <beng_> i forgot to add it when I merged the patch onto this machine
[20:25:23] <beng_> added and landed
[20:25:23] * phajdan-jr mutters something about being on irc and watching the tree when landing
[20:25:26] <beng_> should go green now
[20:26:03] <phajdan-jr> thanks!
[20:26:14] <phajdan-jr> I'll try on my machine...
[20:26:50] <phajdan-jr> beng_: worked for me, so should also be good on the main waterfall
[20:26:52] <phajdan-jr> xji: ^
[20:27:00] <beng_> cool thanks
[20:27:07] <beng_> sorry for the mistake
[20:27:11] <xji> phajdan-jr: thanks.
[20:27:31] <jorlow> nsylvain: webkit linux (webkit.org) and xp perf (webkit.org) both seem pretty sick
[20:28:01] <nsylvain> working on xp perf now, bad sectors on drive
[20:28:09] <nsylvain> will look at webkit linux again.
[20:28:18] <phajdan-jr> I'll update the status
[20:28:27] *** fta has quit IRC
[20:28:32] <markmentovai> sheriffs, ok to reopen now. no need to remain closed any longer for this until there are other failures.
[20:28:33] <nsylvain> jorlow: webkit linux seems better now
[20:28:52] <jorlow> nsylvain: yeah...looks like linux valgrind webkit.org had the same issue
[20:28:53] <trungl-bot> Tree closed by phajdan.jr at chromium dot org (:phajdan-jr): Tree is closed (update horked -> beng, on it; ETA reopening 12:15 AM PDT; xp-perf: bad sectors, nsylvain on it)
[20:28:55] <jorlow> but it was temporary
[20:28:57] <phajdan-jr> markmentovai: I'll re-open now
[20:29:05] <markmentovai> that's what i like to hear
[20:29:21] <nsylvain> jorlow: this is a real code breakage that got fixed when ben checked in
[20:29:57] <trungl-bot> Tree opened by xji at chromium dot org: Tree is opened (cycling green)
[20:30:00] <jorlow> oh, i see...sorry
[20:30:05] *** jrmuizel has quit IRC
[20:30:07] *** fta has joined #chromium
[20:30:13] <markmentovai> phajdan-jr: is brettw still looking at chromeos? that seems to have been lost from the status.
[20:30:31] <phajdan-jr> brettw: ^
[20:30:33] *** jrmuizel has joined #chromium
[20:30:41] <markmentovai> xji: is nsylvain no longer on xp-perf's bad sectors?
[20:30:49] <markmentovai> or was that an error?
[20:30:58] <trungl-bot> Tree opened by xji at chromium dot org: Tree is opened (cycling green; xp-perf: bad sectors, nsylvain on it)
[20:31:13] <markmentovai> xji: thank you
[20:31:29] *** jrmuizel_ has joined #chromium
[20:31:30] <nsylvain> wait.. the xp perf i am on is for the canaries
[20:31:41] <nsylvain> not sure other people than victorw1/jorlow care about that
[20:32:01] <trungl-bot> Tree opened by xji at chromium dot org: Tree is opened (cycling green; xp-perf: bad sectors, nsylvain on it; Brett looking at CrOS)
[20:32:06] <jorlow> yeah...doesn't need to be on the status line
[20:32:12] <markmentovai> you're correct, doesn't need to be in the status. we can put it in the new doc thing.
[20:32:58] *** ph0nk has joined #chromium
[20:34:03] <trungl-bot> Tree opened by mark at chromium dot org (:markmentovai): Open. Cycling green. ChromeOS ? brettw. Details: http://bit.ly/bsyE8W
[20:34:57] *** jrmuizel has quit IRC
[20:34:57] *** jrmuizel_ is now known as jrmuizel
[20:35:29] *** lorenzo has quit IRC
[20:38:50] *** mnissler__ has joined #chromium
[20:40:17] *** mnissler_ has quit IRC
[20:40:36] *** froek has left #chromium
[20:41:12] *** arv__ has quit IRC
[20:43:55] *** satish_ has joined #chromium
[20:44:26] <satish_> victorw1: ping
[20:44:54] <victorw1> satish_: hi
[20:45:22] <satish_> victorw1: g'morning :) is there a webkit roll coming soon?
[20:45:35] *** ph0nk has quit IRC
[20:46:21] <victorw1> satish_: not yet. webkit canary bots have been broken for a while. a couple of issues have beend fixed, but I still need to rollout a change that breaks the webkit unittest
[20:46:54] <victorw1> satish_: once that's done. we will make sure webkit canary bots are green before can do a roll
[20:47:07] <rohitrao> andybons: ping?
[20:47:50] *** zloidemon has quit IRC
[20:49:04] <xji> anyone know whom should be contacted to remove lock for "update scripts fail" in chromium reliability?
[20:49:43] *** satish_ has quit IRC
[20:49:58] <xji> ^chromium reliability^chromium reliability build
[20:50:40] *** victorw1 is now known as victorw
[20:51:15] <markmentovai> xji: where is htat?
[20:51:34] <xji> markmentovai: http://build.chromium.org/buildbot/waterfall/builders/Chromium%20Reliability/builds/11340/steps/shell/logs/stdio
[20:52:06] *** eroman has joined #chromium
[20:52:07] *** ChanServ sets mode: +v eroman
[20:52:23] *** tonyg-cr has quit IRC
[20:52:48] <markmentovai> ugh, what a pain to find out which machine that's on
[20:54:44] <markmentovai> xji: ok, ran svn cleanup
[20:55:23] <xji> markmentovai: thanks. where did you ran that?
[20:55:45] <markmentovai> this is chrome-eng@chrome-master
[20:56:27] *** hebz0rl has quit IRC
[20:57:01] <markmentovai> beng: you still there?
[20:57:17] <trungl-bot> Tree opened by mark at chromium dot org (:markmentovai): Open. ChromeOS ? brettw. Details: http://bit.ly/bsyE8W
[20:57:31] <markmentovai> beng: check deps failure on linux tests x64 is you
[20:58:06] <markmentovai> or are you beng_?
[20:58:20] *** satish_ has joined #chromium
[20:58:59] *** satish_ has joined #chromium
[21:02:11] *** ph0nk has joined #chromium
[21:02:44] *** ejat has quit IRC
[21:03:13] *** Viclame has joined #chromium
[21:05:58] *** estes has quit IRC
[21:09:23] *** dhg has left #chromium
[21:09:31] *** Adys has quit IRC
[21:14:29] <trungl-bot> Tree closed by buildbot at chromium dot org: Tree is closed (Automatic: "check deps" on "Chromium XP" from 54160: avi at chromium dot org (:motownavi), ben at chromium dot org, craig.schlenter at chromium dot org, dhollowa at chromium dot org, eroman at chromium dot org (:eroman), jamesr at chromium dot org, pkasting at chromium dot org, rch at chromium dot org, rohitrao at chromium dot org, wtc at chromium dot org)
[21:15:01] *** Adys has joined #chromium
[21:15:14] *** pinkerton has joined #chromium
[21:15:14] *** ChanServ sets mode: +v pinkerton
[21:15:20] <motownavi> It's ben's break
[21:15:33] <beng_> looking at checkdeps
[21:16:32] *** thomasvl has joined #chromium
[21:16:32] *** ChanServ sets mode: +v thomasvl
[21:16:35] <trungl-bot> Tree closed by xji at chromium dot org: Tree is closed ("check deps" failure --> beng. ChromeOS ? brettw. Details: http://bit.ly/bsyE8W)
[21:16:35] <beng_> fixed
[21:17:56] <beng_> don't think it requires close
[21:18:03] <beng_> I think it can be opened
[21:19:20] <markmentovai> that is correct. no need to close.
[21:19:37] <markmentovai> cue trungl-bot
[21:19:40] <trungl-bot> Tree opened by xji at chromium dot org: Tree is open (cycling green. ChromeOS ? brettw. Details: http://bit.ly/bsyE8W)
[21:20:12] *** tonyg-cr has joined #chromium
[21:20:13] *** ChanServ sets mode: +v tonyg-cr
[21:23:34] *** mrossetti has quit IRC
[21:25:14] *** mrossetti has joined #chromium
[21:27:16] <rohitrao> markmentovai: is andy nearby?
[21:27:33] <markmentovai> rohitrao: no
[21:27:39] <rohitrao> ok, thanks
[21:27:51] *** estes has joined #chromium
[21:28:14] *** Zaba has quit IRC
[21:28:33] *** alokp has joined #chromium
[21:31:53] <trungl-bot> Tree closed by buildbot at chromium dot org: Tree is closed (Automatic: "net_unittests" on "Modules XP (dbg)" from 54164: ben at chromium dot org)
[21:32:56] <phajdan-jr> net_unittests?
[21:33:08] <markmentovai> that's nice. it's not beng_, which is the only chicken in the window.
[21:33:30] <phajdan-jr> oh, our HighResNowSingleton doe again
[21:33:33] <phajdan-jr> *foe
[21:33:52] *** arv__ has joined #chromium
[21:34:45] *** chrisccoulson has quit IRC
[21:35:20] *** craigsch has quit IRC
[21:35:44] <phajdan-jr> re-opened the tree
[21:35:52] *** Zaba has joined #chromium
[21:35:56] <trungl-bot> Tree opened by phajdan.jr at chromium dot org (:phajdan-jr): Tree is open (cycling green. ChromeOS ? brettw. Details: http://bit.ly/bsyE8W)
[21:36:08] <phajdan-jr> I'm talking with cpu now to reduce the flakiness HighResNowSingleton causes us.
[21:36:35] *** bauerb has quit IRC
[21:38:01] <dhollowa> rohitrao: ping
[21:38:11] <rohitrao> dhollowa: hello
[21:38:47] <dhollowa> rohitrao: i'm seeing viewid compile error on 472 branch
[21:38:58] <rohitrao> dhollowa: andybons and I are on it :)
[21:39:15] <dhollowa> rohitrao: cool.  thx.  any eta?
[21:39:52] <rohitrao> dhollowa: few minutes, http://codereview.chromium.org/3028035/show
[21:40:56] <dhollowa> sweet, thx, that gets me rollin.
[21:44:54] <andybons_> dhollowa: yeah sorry about that. i knew right when i committed that it would cause issues. now i've just been waiting for the branch to compile on my local laptop
[21:45:02] <andybons_> but rohitrao beat me to the punch
[21:45:31] * rohitrao wonders if he should just commit or wait for these 15 test binaries to finish linking
[21:46:27] <pkasting> Does anyone build linux_view?
[21:46:38] <pkasting> I need help debugging a problem
[21:46:57] <rohitrao> pkasting: do you have a linux box? there's some env variable you can define to get a linux views build
[21:47:05] <pkasting> I do not
[21:47:15] <rohitrao> ah, that makes things harder :)
[21:47:28] <evmar> phajdan-jr: got a link to the highressingleton failure? i changed it to print more info
[21:47:47] <evmar> googlers: my windows box is currently grinding in the antivirus scanner.  why doesn't it run at night?
[21:48:22] <jamesr> evmar: because that would make sense
[21:48:38] <evmar> it appears i can stop the service, even though the sophos control panel is all grayed out
[21:48:42] <evmar> shrug
[21:48:42] <jamesr> you can kill it
[21:48:46] <jamesr> but i'll come back in a couple hours
[21:49:49] <rohitrao> pkasting: http://www.chromium.org/developers/build-instructions-chromeos , if you manage to find a linux box
[21:50:18] *** chrisccoulson has joined #chromium
[21:50:45] <fta> tables are badly rendered since the last 2~3 days, known?
[21:50:52] <pkasting> :/
[21:51:00] <fsamuel> fta: still?
[21:51:26] <dhollowa> andybons_:  np.  wonders about dedicated branch bots...
[21:51:27] <jamesr> fta: that should be fixed by now
[21:51:37] <jamesr> fta: on ToT as of a day or two ago
[21:52:18] <fsamuel> If anyone sees any table bugs while browsing from ToT, please yell at me :)
[21:52:47] <fta> fsamuel, jamesr: oh, good. the ubuntu build farm is overcrowded so i'm still waiting for yesterdays builds :(
[21:52:57] *** tfarina has joined #chromium
[21:52:59] <fta> http://people.ubuntu.com/~fta/ppa-dashboard/chromium-daily.html
[21:53:21] *** roc has joined #chromium
[21:53:53] <jamesr> i dunno how to read that. but it should be fixed Real Soon Now (TM)
[21:54:17] *** tonikitoo has quit IRC
[21:54:48] <brettw> I think I figured out the ChromeOS build redness
[21:54:53] <brettw> just reverted my culpret
[21:54:56] <rohitrao> do tell!
[21:55:29] <brettw> it was hung doing some audio thing
[21:56:39] *** dglazkov_ has joined #chromium
[21:56:40] *** ChanServ sets mode: +v dglazkov_
[21:56:40] <rohitrao> fingers crossed
[21:57:03] <dhollowa> rohitrao, andybons_ : i got clean compiles for Chrome and unit_tests, but get unit_test failure for TabContentsTest.WebKitPrefs
[21:57:08] *** Engin has joined #chromium
[21:57:13] <andybons_> dhollowa: awesome.
[21:57:17] <andybons_> kudos to rohitrao
[21:57:18] <rohitrao> dhollowa: that's what I'm looking at now
[21:57:39] <rohitrao> dhollowa: are you looking at failures in general? or those autofill tests?
[21:58:10] <trungl-bot> Tree closed by buildbot at chromium dot org: Tree is closed (Automatic: "net_unittests" on "Modules XP (dbg)" from 54167: alokp at chromium dot org)
[21:58:30] <dhollowa> rohitrao: i'm merging AutoFill CLs to the branch.  so wanted to know that i'm starting from a good place.
[21:58:49] *** dglazkov has quit IRC
[21:58:50] *** dglazkov_ is now known as dglazkov
[21:58:50] <alokp> looking at net_unittests
[21:59:02] <rohitrao> dhollowa: got it.  that failure is known.  so is a crash in SwitchToPage
[21:59:18] *** EnginA has quit IRC
[21:59:19] <fsamuel> are text boxes in webpages crashy for anyone?
[21:59:27] <dhollowa> rohitrao: ok, thanks for the warning.
[21:59:33] <fsamuel> [9893:9921:772998060588:FATAL:webkit/glue/webkitclient_impl.cc(283)] Check failed: false. Unknown image resource
[21:59:33] <fsamuel> Backtrace:
[21:59:33] <fsamuel> 	StackTrace::StackTrace() [0x1ac2462]
[21:59:34] <fsamuel> 	logging::LogMessage::~LogMessage() [0x1ad972f]
[21:59:34] <fsamuel> 	webkit_glue::WebKitClientImpl::loadResource() [0x24d5205]
[21:59:35] <fsamuel> 	WebCore::ChromiumBridge::loadPlatformImageResource() [0x28fb5e7]
[21:59:37] <fsamuel> 	WebCore::Image::loadPlatformResource() [0x2cc42cc]
[21:59:39] <fsamuel> 	WebCore::PopupListBox::paintRow() [0x2caa1c0]
[21:59:41] <fsamuel> 	WebCore::PopupListBox::paint() [0x2ca9659]
[21:59:43] <fsamuel> 	WebCore::PopupContainer::paint() [0x2ca84e5]
[21:59:45] <fsamuel> 	WebKit::WebPopupMenuImpl::paint() [0x28b32a1]
[21:59:47] <fsamuel> 	RenderWidget::PaintRect() [0x19a8ae3]
[21:59:49] <fsamuel> 	RenderWidget::DoDeferredUpdate() [0x19a94bb]
[21:59:51] <fsamuel> 	RenderWidget::CallDoDeferredUpdate() [0x19a8d85]
[21:59:52] <jamesr> woah
[21:59:53] <fsamuel> 	DispatchToMethod<>() [0x19afb0b]
[21:59:56] <jamesr> easy on the copy paste thee
[21:59:57] <fsamuel> 	RunnableMethod<>::Run() [0x19af9ec]
[21:59:59] <fsamuel> 	MessageLoop::RunTask() [0x1adc109]
[21:59:59] <jamesr> use a pastebin or something
[22:00:01] <fsamuel> 	MessageLoop::DeferOrRunPendingTask() [0x1adc1ec]
[22:00:01] *** ph0nk has quit IRC
[22:00:03] <fsamuel> 	MessageLoop::DoWork() [0x1adc7e0]
[22:00:05] <fsamuel> 	base::MessagePumpDefault::Run() [0x1ae36f2]
[22:00:07] <fsamuel> 	MessageLoop::RunInternal() [0x1adb8d4]
[22:00:09] <fsamuel> 	MessageLoop::RunHandler() [0x1adb780]
[22:00:11] <fsamuel> 	MessageLoop::Run() [0x1adb711]
[22:00:13] <fsamuel> 	base::Thread::Run() [0x1b0de94]
[22:00:15] <fsamuel> 	base::Thread::ThreadMain() [0x1b0dff3]
[22:00:17] <fsamuel> 	ThreadFunc() [0x1aeef1b]
[22:00:19] <fsamuel> 	start_thread [0x7fbd6e233a04]
[22:00:21] <fsamuel> 	0x7fbd6bf1480d
[22:00:31] <rsesek> pastebin please
[22:00:32] <jamesr> fsamuel: don't copy paste gigantic things into irc
[22:00:52] <alokp> sherrifs - net_unittests failure does not seem related to my CL
[22:00:53] <jamesr> and no, that looks like a regression
[22:00:58] *** monreal has quit IRC
[22:01:05] <dhollowa> fsamuel: this is known, jhawkins is on it.
[22:01:41] <fsamuel> sorry for the massive paste :(
[22:01:58] <rsesek> anyone on 10.5 and can test something for me?
[22:02:00] <dhollowa> fsamuel: for workaround comment out the NOTREACHED in loadResource()
[22:02:01] *** dcheng has quit IRC
[22:02:05] <jamesr> rsesek: sure whatup
[22:02:06] *** jrmuizel_ has joined #chromium
[22:02:50] *** Aria has joined #chromium
[22:03:46] <victorw> jorlow: unit_tests on webkit canary xp test crashes, any idea?
[22:03:59] <rsesek> jamesr: got a small test application. can you test click-hold-drag-release on the last button menu item (the rounded textured button) in the View menu? http://www.corp.google.com/~rsesek/chrome/MenuButton.app.zip
[22:04:43] *** jrmuizel_ has quit IRC
[22:04:58] <victorw> jorlow: crahsed at  RenderViewTest.ExtensionMessagesOpenChannel
[22:05:47] <jamesr> rsesek: ok
[22:05:47] *** jrmuizel has quit IRC
[22:05:52] <rsesek> thanks
[22:06:20] <jamesr> rsesek: which item? What you need?
[22:06:28] <rsesek> jamesr: find out what it means
[22:06:42] <rsesek> the other two sets of buttons should NOT work on 10.5. but the last one should work
[22:06:46] <fsamuel> dhollowa: thanks!
[22:06:55] <rsesek> (this may be the most ridiculous app I've ever written)
[22:06:57] *** tonikitoo has joined #chromium
[22:08:09] <jamesr> rsesek: if i drag and release inside the button it clicks
[22:08:17] <jamesr> rsesek: if i drag and release outside the button it don't click
[22:08:18] <rsesek> halleluja!
[22:08:22] <jamesr> also i see rounded corners but no texture
[22:08:32] <jamesr> actually there's a top-to-bottom gradient
[22:08:46] <rsesek> jamesr: cool. and I'm assuming that normal click open and then click on button works still?
[22:09:06] *** rniwa has quit IRC
[22:09:21] *** anicolao has joined #chromium
[22:09:31] <anicolao> duh ... what's the right way (tm) to rebaseline tests? I want something as easy as checking in new expectations that break a buildbot and then grabbing the results using the rebaseline tool to submit a patch that rebaselines, without breaking the bot.
[22:09:57] <jamesr> anicolao: you mean webkit?
[22:10:07] <anicolao> jamesr: yes, layout tests.
[22:10:24] *** feldstein has quit IRC
[22:10:25] <jamesr> anicolao: in the patch that you change the behavior mark the test as failing in test_expectations.txt
[22:10:42] <jamesr> anicolao: let the webkit canaries cycle once. they will fail on your test but it won't turn the bot red since you marked the tests as expected failures
[22:10:55] <jamesr> anicolao: then use the rebaseline tool to grab the new expectations and remove the failing expectations from test_expectations.txt
[22:11:03] <jamesr> anicolao: land the new baselines as a follow-up. problem solved!
[22:11:08] <jamesr> rsesek: yeah
[22:11:18] <rsesek> jamesr: awesome. thanks
[22:11:20] <rsesek> pinkerton: it looks like we may be able to get click-drag-release working on BOTH 10.5 and 10.6, without breaking accessibility. it will require killing the segmented control
[22:11:29] <jamesr> rsesek: actually i think all the buttons work ?
[22:11:35] <rohitrao> dhollowa: consider also testing with "GYP_DEFINES="branding=Chrome buildtype=Official" gclient runhooks"
[22:11:46] <rohitrao> cause clearly some autofill code only breaks in official builds :)
[22:11:50] <rsesek> jamesr: interesting. click-drag-release probably shouldn't work for the other two sets
[22:12:10] <dhollowa> rohitrao: will do.  thx.
[22:12:24] <rohitrao> dhollowa: ask if you need help doing a full checkout with the internal code too
[22:12:35] <pinkerton> rsesek: i'm ok with not using a segmented control. will it totally look like ass otherwise?
[22:13:01] <jamesr> rsesek: you just mean depressing the mouse inside the button bounds and then moving the mouse around while keeping the mouse depressed right?
[22:13:18] <anicolao> jamesr: I guess I'm doing something wrong, the tests are marked as failing and when I run the rebaseline tool it says there's no image in the result. but you're saying all the images should be there even if the test already fail?
[22:13:25] <rsesek> jamesr: no. the click-hold-drag-release-over-button should be broken for everything but the "find out what it means to me" button
[22:13:31] <dhollowa> rohitrao: any experience with "drover --local" ?  would like to test merges prior to commit...
[22:13:32] <rsesek> pinkerton: the segmented control is nice because it aligns the buttons all nice and pretty. I now get to write the code that does that because of l10n reasons
[22:13:39] <jamesr> rsesek: aaah i see
[22:13:46] <jamesr> rsesek: so click outside the button but release inside it?
[22:13:48] <rohitrao> dhollowa: nope, but it seems pretty straightforward
[22:13:56] <jamesr> anicolao: 'failing' and 'missing' are different in test_expectations.txt
[22:14:00] <dhollowa> rohitrao: except that it doesn't work...
[22:14:11] <jamesr> anicolao: you have to explicitly mark a test as 'missing' if there are no expectations for something
[22:14:31] <pinkerton> rsesek: i guess ultimately i dislike the idea of rewriting something that works perfectly well just because of 10.5
[22:14:32] <rsesek> jamesr: no, clicking the menu open and do not release, then drag the cursor over the button and release
[22:14:57] <jamesr> rsesek: ah yes. only the bottom button works for that
[22:15:03] <rsesek> pinkerton: this won't be just for 10.5 reasons, it's because only NSButton really lends itself to the new approach that works without breaking a11y
[22:15:03] <jamesr> 'find out what it means'
[22:15:16] <rsesek> jamesr: ok. then that does exactly what I think it should,then
[22:15:27] <trungl-bot> Tree opened by brettw at chromium dot org: Tree is open Details: http://bit.ly/bsyE8W
[22:15:36] <rohitrao> dhollowa: i got nothing then
[22:15:39] <pinkerton> rsesek: so a11y is broken on 10.6 too with segmented controls?
[22:16:00] *** xji has quit IRC
[22:16:13] <rsesek> pinkerton: with the approach I was taking, a11y is broken for everything. the nssegmentedcontrols are even further broken because I can't hit test individual segments. so I'd need to convert those to buttons to simply get 10.6 working without a11y
[22:16:34] <rsesek> moral of the story: don't put buttons in menus
[22:16:38] *** cying has joined #chromium
[22:16:48] <pinkerton> i want a11y to work so we can tell gruber to go f- himself.
[22:16:53] <rsesek> :)
[22:17:00] <pinkerton> was that my outside voice?
[22:17:10] <rohitrao> you were certainly outside
[22:17:28] <rsesek> agreed. so I'll use this new approach that works for 10.5 and 10.6, doesn't break a11y, but will require turning the segmented control into 3 buttons and then writing the l10n alignment code
[22:17:32] <dhollowa> rohitrao: correction, pilot error, you have to be cd'd to .../src for it to work.  happy now.
[22:17:45] <rohitrao> ah, makes sense, i guess
[22:17:53] <markmentovai> that is the worst kind of error.
[22:17:56] <rohitrao> but it should complain otherwise
[22:17:57] *** Strat has joined #chromium
[22:19:29] <anicolao> jamesr: I'm wandering around the wiki looking for a page that explains this distinction but no luck. You're saying if I add them as 'MISSING' it'll generate output from which I can then proceed with rebaselining?
[22:20:03] <jamesr> anicolao: this should be documented at the top of test_expectations.txt
[22:20:21] <jamesr> hm
[22:20:22] <jamesr> maybe i'm mistaken
[22:20:25] <anicolao> jamesr: it is not, I looked there too. should I submit a patch?
[22:21:28] <jamesr> we have = MISSING set for some tests. now i'm confused too
[22:21:33] <victorw> nsylvain: ping
[22:21:55] <anicolao> jamesr: it's nice to have company :)
[22:24:34] <trungl-bot> Tree closed by buildbot at chromium dot org: Tree is closed (Automatic: "compile" on "Chromium Builder" from 54175: brettw at chromium dot org, pkasting at chromium dot org, thakis at chromium dot org)
[22:24:42] <thakis> :-/
[22:24:43] <pkasting> what
[22:25:03] <thakis> windows failure
[22:25:04] <thakis> not it
[22:25:18] <thakis> brettw: you
[22:25:19] <jamesr> anicolao: ok, this is not documented but it should work
[22:25:26] <pkasting> 5>c:\b\slave\chromium-rel-builder\build\src\third_party\ppapi\cpp\font.cc(43) : error C2220: warning treated as error - no 'object' file generated
[22:25:26] <pkasting> 5>c:\b\slave\chromium-rel-builder\build\src\third_party\ppapi\cpp\font.cc(43) : warning C4717: 'pp::FontDescription::FontDescription' : recursive on all control paths, function will cause runtime stack overflow
[22:25:29] <pkasting> Not me
[22:25:43] <brettw> reverting
[22:25:46] *** romainhuet has quit IRC
[22:26:49] <brettw> reverted
[22:26:58] *** rdsmith has joined #chromium
[22:26:59] *** ChanServ sets mode: +v rdsmith
[22:27:03] *** roc has quit IRC
[22:27:16] <victorw> any trooper around can help with webkit canary bot?
[22:27:28] <nsylvain> i'm back for a few minutes, i can look
[22:27:32] <anicolao> jamesr: I think what I'll do is rebaseline on linux, check in the expectations for windows, mark windows as fail and then either rebaseline windows/linux if windows fails and linux passes or remove the expectation if windows unexpectedly passes. does that seem OK?
[22:27:39] <victorw> nsylvain: sure. thx
[22:27:41] <jamesr> anicolao: well, no
[22:27:42] <trungl-bot> Tree opened by brettw at chromium dot org: Tree is open (Brett reverted)
[22:27:45] <nsylvain> victorw: which one you want me to look at?
[22:27:51] <jamesr> anicolao: what you want to do is generate linux baselines, mark mac and windows as = MISSING
[22:28:01] <victorw> nsylvain: xp tests (webkit.org) and linux perf (webkit.org)
[22:28:04] <jamesr> anicolao: check it in let it cycle and pull the windows and mac expectations from the canaries
[22:28:11] <jamesr> you don't want to check in linux results into the windows directory
[22:28:16] <jamesr> that'll basically never be right
[22:28:33] <anicolao> jamesr: OK. it's only windows that's the problem the mac baselines are in.
[22:28:52] <jamesr> ok, then just mark windows as = MISSING
[22:29:06] *** andybons_ has quit IRC
[22:29:10] *** taf2 has joined #chromium
[22:29:19] <anicolao> jamesr: and I do actually expect windows to match linux, but I could easily be wrong, so I'll do it the MISSING way.
[22:29:25] <victorw> nsylvain: xp tests bot crashed on unit_tests and lots of ui tests faile. looks like started ~6:00am
[22:29:34] <jamesr> anicolao: do you have text?
[22:29:38] <nsylvain> are you sure they are not real regressiosn?
[22:29:44] <jamesr> if there is no text then they might match
[22:30:32] <victorw> nsylvain: I am not sure. but on xp tests bot, extract_build is yellow, do you see any problem there?
[22:30:33] <anicolao> jamesr: depends on the test. I'll follow your pattern it is understandable enough.
[22:30:35] *** Venom_X has quit IRC
[22:30:55] <victorw> nsylvain: the patch for that run should not cause this
[22:31:16] <victorw> nsylvain: http://build.chromium.org/buildbot/waterfall.fyi/builders/XP%20Tests%20(webkit.org)/builds/6380
[22:31:29] <nsylvain> yellow just means in this case that the blamelist might be wrong
[22:31:35] <victorw> nsylvain: jorlow just clobberred both bots but does not seem help
[22:31:58] <nsylvain> there was a lighttp process hung
[22:32:03] <nsylvain> i'm cleaning up, maybe this will help
[22:32:15] <victorw> nsylvain: good. thanks!
[22:32:17] <rohitrao> victorw: the extract build logs will tell you what revision was actually tested
[22:32:23] *** dglazkov has quit IRC
[22:32:42] *** tfarina has quit IRC
[22:33:28] *** Precea[BNC] is now known as Precea
[22:33:59] <nsylvain> victorw: but linux perf fails pretty much the same way as windows.. and it started pretty much at the same time.. so i doubt this is entirely a bot problem.
[22:34:04] <victorw> nsylvain: does lighttp hang on xp tests? or linux_perf
[22:34:17] <nsylvain> it was on xp tests.. but that might not be the problem
[22:34:32] *** xji has joined #chromium
[22:34:32] *** ChanServ sets mode: +v xji
[22:34:42] <victorw> nsylvain: I know. I reverted the change that cause this issue. we can wait for one more cycle to see if the problem is gone
[22:35:07] <nsylvain> yeah
[22:35:10] <nsylvain> it worked on linux
[22:35:15] <nsylvain> it's turning green now
[22:36:20] <victorw> nsylvain: ya. thanks
[22:36:59] <victorw> nsylvain: hopeful the xp tests issue can be fixed. we are behind on webkit roll. need to catch up
[22:39:09] <victorw> nsylvain: xp test picks up pretty old webkit patch, any way we can make it run from the latest patch?
[22:40:36] *** ojan has joined #chromium
[22:40:36] *** ChanServ sets mode: +v ojan
[22:42:51] *** xji_ has joined #chromium
[22:42:51] *** ChanServ sets mode: +v xji_
[22:43:15] *** Viclame has left #chromium
[22:45:53] <trungl-bot> Tree closed by buildbot at chromium dot org: Tree is closed (Automatic: "net_unittests" on "Modules XP (dbg)" from 54181: brettw at chromium dot org, phajdan.jr at chromium dot org (:phajdan-jr), rdsmith at google dot com)
[22:46:41] <rohitrao> nsylvain: update_scripts is failing on the google chrome beta bot with "Working copy 'C:\b\depot_tools\bootstrap' locked."  Is that bad?
[22:46:57] <xji> phajdan-jr: is this the same flakiness?
[22:47:14] <phajdan-jr> xji: yeah, please re-open tree
[22:47:23] <phajdan-jr> xji: I have a band-aid in the queue
[22:48:57] <trungl-bot> Tree opened by xji at chromium dot org: Tree is open
[22:50:36] <rohitrao> mnissler: ping
[22:50:40] *** rniwa has joined #chromium
[22:50:40] *** ChanServ sets mode: +v rniwa
[22:50:46] *** beng_ has quit IRC
[22:53:05] *** andrix has quit IRC
[22:54:50] *** MikeSmithW3C^ has joined #chromium
[22:55:24] *** tonikitoo has quit IRC
[22:56:06] *** dcheng has joined #chromium
[22:57:23] *** ROBOd has quit IRC
[22:58:57] <dcheng> Any idea why gclient sync keeps trying to use the GNOME keyring?
[22:58:59] *** MikeSmithW3C has quit IRC
[22:59:12] *** ananta has quit IRC
[22:59:15] <dcheng> I'm trying to checkout one of the repositories, and it just keeps failing with the error "GNOME Keyring is locked and we are non-interactive"
[23:00:14] <thakis> dcheng: i think evmar or leiz updated it to read your svn pw from some python crypto module which wraps gnome keyring on linux
[23:00:27] <phajdan-jr> xji: landed the band-aid for net_unittests. please ping me if it turns out not to help.
[23:00:35] <dcheng> thakis: Any idea how to make it work?
[23:00:44] <xji_> phajdan-jr: thanks.
[23:00:53] <thakis> dcheng: no, sorry
[23:00:55] <phajdan-jr> dcheng: kick it. always works. ;-) just kidding
[23:01:22] <victorw> nsylvain: xp tests unit_tests issue is gone, looks like lighttp is the issue.
[23:02:18] *** anicolao has quit IRC
[23:03:59] *** Adys has quit IRC
[23:04:04] <trungl-bot> Tree closed by buildbot at chromium dot org: Tree is closed (Automatic: "net_unittests" on "Modules XP (dbg)" from 54187: alokp at chromium dot org, jhawkins at chromium dot org, mrossetti at chromium dot org)
[23:04:07] *** isherman has joined #chromium
[23:04:07] *** ChanServ sets mode: +v isherman
[23:09:10] *** roc has joined #chromium
[23:10:31] <phajdan-jr> flakiness :(
[23:10:53] <selckin> open a ski slope
[23:11:12] <trungl-bot> Tree opened by phajdan.jr at chromium dot org (:phajdan-jr): Tree is open
[23:11:29] * phajdan-jr wonders why the sheriffs haven't done that
[23:11:32] <phajdan-jr> xji_: ^
[23:11:35] <phajdan-jr> brettw: ^
[23:11:44] <brettw> phajdan-jr, you mean open the tree?
[23:12:01] <phajdan-jr> brettw: yes. net_unittests was a flake, and nobody reacted it seems.
[23:12:13] *** krasimira has quit IRC
[23:12:14] <brettw> Was doing code reviews for a few minutes
[23:12:58] *** dpranke has joined #chromium
[23:12:59] *** ChanServ sets mode: +v dpranke
[23:13:13] <trungl-bot> Tree opened by xji at chromium dot org: Tree is open
[23:14:01] <xji_> phajdan-jr: was looking at whether it included your change already. looks like it is included your change.
[23:15:12] <bryner> am i supposed to reply somewhere if i was tbr'd on a review?  there's no review url listed for the change (rev 54112)
[23:15:31] <phajdan-jr> xji_: no, it didn't ;-p
[23:16:19] <phajdan-jr> bryner: if there's no review link, then I guess you can send an "OK" (or comments) e-mail and remind the patch author to use gcl/git-cl to upload things
[23:16:21] <brettw> Can somebody shed some light on this ARM build error? "chrome/browser/views/tabs/base_tab.cc:320: error: 'waiting_to_loading_frame_count_ratio.unstatic.2460' may be used uninitialized in this function"
[23:16:26] <brettw> it's a static initialized to
[23:16:26] <brettw> 0
[23:16:31] <bryner> phajdan-jr: ok
[23:16:44] <rsesek> shess: ping
[23:17:04] <shess> rsesek: was just working on the popup animation cl.  rly
[23:17:50] <rsesek> shess: :) glad to hear it. but I have a question: do you have any ideas on how to test |-mouseDragged:| on a button. I can synthesize the first NSEvent in a sequence but as soon as I send it to the button's |-mouseDragged:|, it appears to spin a modal runloop and then not do anything :(
[23:18:08] <fishd> tip-of-tree build is exceptionally crashy in the browser process
[23:18:21] <fishd> is this a known issue?
[23:18:41] <phajdan-jr> fishd: drive-by ping about my download code cleanup review ;-p
[23:18:52] <shess> rsesek: location_bar/autocomplete_text_field_unittest.mm has an example of a drag-select.
[23:19:02] <rsesek> ah ha! I'll check itout
[23:19:02] <rohitrao> mirandac: ping
[23:19:04] <rsesek> thanks
[23:19:29] <shess> rsesek: you have to post the events.  I'm often thinking of writing some utility fns to let you run interesting sequences (like click at point and drag to other point), but never got around to it.
[23:19:44] <shess> rsesek: note that if you want to cause drag-and-drop, then you're probably in for an entire additional level of pain.
[23:19:48] *** legion13 has joined #chromium
[23:20:05] <phajdan-jr> fishd: http://codereview.chromium.org/3071005/show
[23:20:07] *** Kunalagon has quit IRC
[23:20:25] *** anicolao has joined #chromium
[23:21:40] <rsesek> shess: thanks. I'll experiment. though this class is small enough if it's too much work I may just skip the test
[23:21:46] <rsesek> it's easy to test in person
[23:21:48] *** Engin has quit IRC
[23:26:41] *** dpranke has quit IRC
[23:27:25] <thakis> rsesek: http://codereview.chromium.org/3029041/show
[23:27:31] <thakis> pinkerton: ping for tabpose cl
[23:28:32] *** thomasvl has quit IRC
[23:28:52] <pinkerton> can i do it tomorrow?
[23:29:41] <thakis> sure :-/
[23:29:53] <pinkerton> it's late here, i was about to drop
[23:30:47] *** pinkerton has quit IRC
[23:31:17] <mnissler__> chrome mac official is back \o/
[23:31:29] *** mnissler__ has quit IRC
[23:34:13] *** alokp has quit IRC
[23:35:54] <dcheng> thakis: I think it's because I have Lucid on this machine.
[23:36:16] <dcheng> I had to add 'password-stores = ' to ~/.subversion/config to disable the GNOME keyring integration.
[23:37:00] <dhollowa> anybody know the plans for webkit and the 472 branch?
[23:37:34] <phajdan-jr> dcheng: may be worth documenting on the wiki; interesting that I've not hit this problem, also on Lucid
[23:39:16] * phajdan-jr lands a patch of death. please consider closing the tree for a while or do some throttling, etc
[23:39:58] <dcheng> phajdan-jr: Not entirely sure what's going on still. I'll document it once I get it completely figured out.
[23:40:14] <dcheng> phajdan-jr: The GNOME keyring wouldn't be so bad if I could actually figure out what the default keyring password is. Or if it worked.
[23:41:22] <phajdan-jr> oh, and now webkit roll ;-) it gets interesting
[23:41:33] <trungl-bot> Tree opened by phajdan.jr at chromium dot org (:phajdan-jr): Tree is open (phajdan.jr lands a patch of death; watch out for compile failures)
[23:42:02] *** js2 has quit IRC
[23:42:12] *** js2 has joined #chromium
[23:45:34] *** tittiatcoke has quit IRC
[23:48:54] *** isherman has quit IRC
[23:49:27] *** romainhuet has joined #chromium
[23:49:32] *** romainhuet has quit IRC
[23:49:57] *** dpranke has joined #chromium
[23:49:58] *** ChanServ sets mode: +v dpranke
[23:50:33] <thakis> dhollowa: in what sense?
[23:50:54] <thakis> dhollowa: 472 will has a branched webkit too
[23:51:01] *** BHSPitCSP has left #chromium
[23:51:13] <thakis> dhollowa: it's here http://src.chromium.org/viewvc/chrome/branches/WebKit/472/
[23:51:14] <dhollowa> in the sense that some of my merges rely on new bits in wk
[23:51:19] <thakis> dhollowa: you need to drover webkit changes to there
[23:51:38] <thakis> dhollowa: the drover docs mention how to merge stuff to webkit
[23:51:48] <dhollowa> thakis: not DEPS changes, but the specific wk changes, yes?
[23:52:02] <thakis> dhollowa: yes. we don't want to roll deps on the branch :-)
[23:52:18] <thakis> dhollowa: see "webkit merges" at http://sites.google.com/a/chromium.org/dev/developers/how-tos/drover
[23:52:24] <thakis> (first hit for "webkit drover")
[23:52:31] <dhollowa> thakis: thx.  i'll take a look at the docs.
[23:52:55] <rdsmith> Question on the MS6 builder: Should we expect the "updating" step (I think it's gclient sync) on Windows to take an hour?
[23:55:59] <thakis> trungl: feel like looking at http://codereview.chromium.org/3029041/show ?
[23:56:45] <rohitrao> arv: around?
[23:58:09] *** JakeSays has quit IRC
[23:58:24] <phajdan-jr> rdsmith: I expect all kinds of weird things to happen on Windows. Serioudly though, ping nsylvain or maruel.
[23:58:56] <phajdan-jr> the patch of death landed green?! I'm surprised. ;-)
[23:59:05] <rdsmith> phajdan-jr: :-}.  But will do.
[23:59:23] *** JakeSays has joined #chromium





top