**** BEGIN LOGGING AT Fri Feb 05 02:59:59 2016
Feb 05 08:19:13 <Tofe> Morning
Feb 05 08:20:13 <Tofe> Garfonso: I agree, but we should also keep the compatibility with a QML loading
Feb 05 08:20:33 <Garfonso> morning Tofe
Feb 05 08:21:06 <Tofe> Which isn't necessarily an issue, if we organize the files correctly and we point to the loadable-frameworks from QML
Feb 05 08:23:40 <Garfonso> maybe a package.json with the right content is already sufficient: https://nodejs.org/api/modules.html#modules_folders_as_modules
Feb 05 08:23:42 <Tofe> (and of course you're completely correct about the numberParamType typo)
Feb 05 08:24:04 <Garfonso> ;)
Feb 05 08:24:29 <Garfonso> just saw it in the diff and thought that's probably an oversight that can cause a hard to find bug later...
Feb 05 08:24:48 <Tofe> yep
Feb 05 08:25:07 <Tofe> anyway I didn't really want to merge it as-is, it's more a WIP
Feb 05 08:27:27 <Herrie|Veer> morning
Feb 05 08:34:10 <Herrie|Veer> Tofe: So for now we should include Garfonso's PR but not yours since it's WIP?
Feb 05 08:35:23 <Garfonso> Tofe: not sure what you are doing there. You try to replace legacy phonenumber handling with something new?
Feb 05 08:36:21 <Herrie|Veer> Garfonso: Yeah
Feb 05 08:37:10 <Herrie|Veer> Because legacy is outdated with it's numberplan & mapping & limited to 20 or so countries. While new one is up2date based on Google's work & 180 countries or so?
Feb 05 08:40:32 <Herrie|Veer> Garfonso: This is for the geolocation lookup. So +49 30 will show Berlin etc
Feb 05 08:42:16 <Herrie|Veer> Garfonso: And it was quite a pain to get g11n working in QML due to the dependencies as well ;)
Feb 05 08:44:25 <Herrie|Veer> So a simple straight forward library is the way to go, both myself and Tofe thought and we'll plug that into the Enyo/legacy bits to replace the outdated one there.
Feb 05 08:47:44 <Tofe> Herrie|Veer: not only geolocation, but also normalizing of phone number, which I need in the phone app
Feb 05 08:49:10 <Herrie|Veer> Tofe: Right, but those 2 go hand in hand, don't they? You can't really do proper normalization without the geolocation, or can you?
Feb 05 08:55:24 <Tofe> well, the common part is the geocoding, where the +46 will get you to sweden and then you'll decode the phone number as a swedish one
Feb 05 08:55:51 <Tofe> But I'm being picky; we do agree.
Feb 05 08:56:11 <Herrie|Veer> Tofe: LOL :P
Feb 05 08:59:45 <Herrie|Veer> Tofe: So your PR is not merge worthy yet?
Feb 05 09:05:11 <Tofe> Herrie|Veer: no, not at all
Feb 05 09:05:41 <Tofe> and I doubt this "fallback on old g11n" thing is a good idea
Feb 05 09:05:59 <Tofe> better only have one normalize format
Feb 05 09:09:00 <Herrie|Veer> Tofe: Yeah I agree
Feb 05 09:09:10 <Herrie|Veer> Just need to know what to include in release :P
Feb 05 09:09:21 <Herrie|Veer> We're now still at the 5s phone app loading :P
Feb 05 09:10:20 <Tofe> Herrie|Veer: oh, are we ? but phone depends on phonenumber.js now
Feb 05 09:11:19 <Herrie|Veer> Tofe: hmmmz
Feb 05 09:11:29 <Herrie|Veer> Let me see maybe I forgot to bump something
Feb 05 09:11:40 <Tofe> luneos-components maybe
Feb 05 09:12:16 <Herrie|Veer> Nope
Feb 05 09:12:29 <Herrie|Veer> Both phone & luneos-components are at latest
Feb 05 09:15:06 <Herrie|Veer> Tofe: Must have been first boot
Feb 05 09:15:10 <Herrie|Veer> On 2nd boot it's OK
Feb 05 09:15:35 <Tofe> ok
Feb 05 09:15:49 <Garfonso> Tofe: yes, remove the fallback... also I'm very much for using something new for phone number normalization and get rid of contacts lib all together
Feb 05 09:15:58 <Garfonso> (maybe keep the vcard handling)
Feb 05 09:16:39 <Tofe> Garfonso: I don't have the necessary knowledge to handle a removal or big rework of contacts lib, but I can replace the normalize handling
Feb 05 09:17:11 <Tofe> I'll remove the fallback tonight, then
Feb 05 13:52:52 <Herrie|Veer> JaMa: ping
Feb 05 13:56:55 <JaMa> pong
Feb 05 13:58:23 <Herrie|Veer> Had a Grouper testing build fail but we didn't change anything specific for Grouper. Can you have a look? Maybe out of space again?
Feb 05 14:00:48 <JaMa> sure
Feb 05 14:01:08 <JaMa> grouper or maguro?
Feb 05 14:01:35 <JaMa> there is ICE in latest maguro build
Feb 05 14:02:05 <Herrie|Veer> Sorry maguro :P
Feb 05 14:02:14 <JaMa> possibly hw issue
Feb 05 14:02:19 <Herrie|Veer> We didn't change anything machine specific afaik
Feb 05 14:02:29 <JaMa> I've retriggered th build
Feb 05 14:02:46 <Herrie|Veer> JaMa: OK if it fails again we need ka6sox?
Feb 05 14:03:18 <JaMa> yes, nothing interesting in log
Feb 05 14:03:28 <JaMa> but it could be faulty ram or something like that
Feb 05 14:03:39 <JaMa> I've seen couple ICEs lately on aruba as well
Feb 05 14:03:42 <Herrie|Veer> Or faulty HD?
Feb 05 14:06:00 <Herrie|Veer> JaMa: Unstable mako still bootlooping (as expected)
Feb 05 14:06:31 <Herrie|Veer> Tofe will look into the other issues for Tenderloin & Maguro (probably root caused by PulseAudio)
Feb 05 14:09:08 <Herrie|Veer> Audio doesn't work on those, lunawebappmgr crashes but seems linked to PulseAudio as well.
Feb 05 14:09:25 <Herrie|Veer> So might be that fixing PulseAudio will fix that too.
Feb 05 14:21:09 <Tofe> Herrie|Veer: speaking of which, I've seen similar pulseaudio complaints on my current N4 image; so maybe I was on the wrong track
Feb 05 14:22:55 <Herrie|Veer> Tofe: Yeah but since PA doesn't recommend it anyway it might be good to use their recommendation?
Feb 05 14:26:17 <Herrie|Veer> JaMa: Seems it failed again...
Feb 05 14:29:01 <Tofe> Herrie|Veer: yes, but not until we fix the issue, otherwise it will just introduce more differences with the working image
Feb 05 14:31:23 <Herrie|Veer> Tofe: OK.
Feb 05 14:32:06 <JaMa> now the error is different
Feb 05 14:32:09 <JaMa> strange one
Feb 05 14:32:12 <JaMa> | ERROR: Required kernel image is not available as fastboot image! /home/jenkins/workspace/luneos-testing/webos-ports/tmp-glibc/deploy/images/maguro/zImage-maguro.fastboot doesn't exist.
Feb 05 14:32:39 <JaMa> I'll clean the workspace that will fix it, but we probably miss the dependency
Feb 05 14:41:30 <ka6sox> what is the issue?
Feb 05 14:42:55 <JaMa> gcc is quite often failing with ICE
Feb 05 14:43:06 <JaMa> we haven't changed gcc lately
Feb 05 14:49:12 <ka6sox> what version are we using?
Feb 05 14:52:23 <JaMa> of gcc?
Feb 05 14:52:53 <JaMa> 49
Feb 05 14:59:23 <ka6sox> I've had issues with 5 but not 4.9
Feb 05 15:26:24 <JaMa> Tofe: your meta-webos-ports commit 923d9b0b breaks build with newer qtwebengine, was there some reason why you overwrite EXTRA_QMAKEVARS_PRE instead of appending to it?
Feb 05 15:35:02 <Tofe> JaMa: one moment
Feb 05 15:36:38 <Tofe> JaMa: no, it's just a mistake
Feb 05 15:40:22 <JaMa> ok, will push a fix
Feb 05 15:41:24 <Tofe> ok thanks :)
Feb 05 15:51:16 <Herrie|Veer> nizovn: ping
Feb 05 15:51:55 <nizovn> Herrie|Veer: pong
Feb 05 15:52:26 <Herrie|Veer> nizovn: You need any more info for mediaindexer?
Feb 05 15:52:38 <Herrie|Veer> Or you're OK with what you have now?
Feb 05 15:52:55 <Herrie|Veer> For testing you can sideload legacy music app IPK and see if it works :)
Feb 05 15:53:09 <Herrie|Veer> Just put some MP3 files on media/internal :P
Feb 05 15:55:06 <Herrie|Veer> And the kinds will be pretty much instantly populated
Feb 05 15:55:32 <nizovn> i'm ok with that, for now i'm thinking how to find offset of apic in mp3 file
Feb 05 15:55:54 <Herrie|Veer> nizovn: The code I quoted in issue should be able to do that?
Feb 05 15:56:01 <Herrie|Veer> Quite sure taglib somehow supports it
Feb 05 15:56:44 <Herrie|Veer> The rajeevandlinux link
Feb 05 16:00:10 <nizovn> i can't find offset in docs, maybe looking in the wrong place
Feb 05 16:08:34 <Herrie|Veer|2> nizovn: Yeah it's not easy to find :P
Feb 05 16:09:14 <Herrie|Veer|2> Tofe: did small luna-systemui PR to fix error in log. Basically calling a method that wasn't released in OWO :P
Feb 05 16:27:17 <Herrie|Veer|2> nizovn: Ah seems taglib doesn't provide the offset?
Feb 05 16:27:26 <Herrie|Veer|2> I guess that would need to be hacked in somehow?
Feb 05 16:28:30 <nizovn> probably, i will read more about id3v2
Feb 05 16:31:49 <Herrie|Veer|2> nizovn: You should check attachedpictureframe.cpp
Feb 05 16:32:01 <Herrie|Veer|2> parseFields has a pos
Feb 05 16:32:11 <Herrie|Veer|2> Which is probably what you're looking for
Feb 05 17:53:02 <Tofe> Garfonso: ping
Feb 05 18:05:59 <nizovn> Herrie: am i right that we need something called "extractfs"? seems it's used in com.palm.app.musicplayer/utility/utilities.js in getTrackImage function
Feb 05 18:29:53 <Herrie> nizovn: I haven't been that far yet
Feb 05 18:29:55 <Herrie> Let me check
Feb 05 18:45:19 <Tofe> Herrie: I'm going to setup a bunch of PR for phonenumberlib, but I'd like to wait for Garfonso to tell me if it seems ok; and I still need to test it properly on device
Feb 05 18:46:03 <Herrie> Tofe: OK
Feb 05 18:46:21 <Herrie> I'm looking into some warnings & errors in logs and clearing those out
Feb 05 18:48:39 <Garfonso> Tofe: Pong
Feb 05 18:50:46 <Tofe> Garfonso: I've sketched up a "phonenumberlib" loadable framework
Feb 05 18:50:54 <Tofe> one moment, I'm going to push it to GH
Feb 05 18:52:17 <Tofe> Garfonso: I've now update my PR: https://github.com/webOS-ports/loadable-frameworks/pull/5
Feb 05 18:52:45 <Tofe> should it work ? especially the require in contacts' prologue.js
Feb 05 18:54:26 <Tofe> Herrie: my PRs are now in place (meta-webos-ports, loadable-frameworks and luneos-components)
Feb 05 18:55:17 <Tofe> Herrie: it's completely not tested :p
Feb 05 19:13:14 <Garfonso> the require should work, from what I know.
Feb 05 19:37:51 <Herrie> nizovn: https://bpaste.net/show/63db47bba13b that's the contents of /usr/palm/services/filenotifyd-triton/motes/image.js on 3.0.5
Feb 05 19:43:40 <nizovn> Herrie: so we need to use extractfs? i think it will take some time to implement it
Feb 05 19:45:08 <Herrie> nizovn: It's basically a path
Feb 05 19:45:13 <Herrie> I think we might have it already
Feb 05 19:45:18 <Herrie> If not we can create it
Feb 05 19:45:29 <Herrie> Not sure what extractfs does as an app or service :S
Feb 05 19:46:08 <Herrie> nizovn: Ah it seems to be some embedded image extractor: Description: Palm embedded file extractor
Feb 05 19:46:16 <Herrie> So I guess taglib does the same already
Feb 05 19:46:31 <nizovn> Herrie: seems it's fuse filesystem, you can see it in mount output
Feb 05 19:47:19 <nizovn> extractfs on /var/luna/data/extractfs type fuse.extractfs (rw,nosuid,nodev,relatime,user_id=0,group_id=0)
Feb 05 19:48:03 <Herrie> nizovn: OK so that might prove tricky for now :S
Feb 05 19:48:25 <Herrie> Maybe focus on the other kinds first? Those should be fairly straigt forward right?
Feb 05 19:49:32 <nizovn> not sure, i haven't looked at them yet
Feb 05 19:49:57 <Herrie> nizovn: This is the legacy code that generates them: /usr/palm/services/filenotifyd-triton/filenotifyd_module_audio.js
Feb 05 19:49:58 <nizovn> but i'm newbie on db stuff
Feb 05 19:51:24 <Herrie> nizovn: db8 is pretty easy in general ;)
Feb 05 19:52:19 <Herrie> https://developer.lge.com/webOSTV/develop/web-app/app-developer-guide/db8/webos-db8-basics/
Feb 05 19:54:00 <nizovn> ok
Feb 05 19:55:35 <Herrie> We're basically putting a lot of strings in a db kind
Feb 05 19:55:47 <Herrie> Which you can see as a table in a SQL db
Feb 05 20:01:29 <Garfonso> nizovn, Herrie: It shouldn't matter where the thumbnail images reside. There is no real need for a new partition or similar.
Feb 05 20:04:51 <Herrie> Garfonso: Just legacy and some other apps might expect it at extractsfs location
Feb 05 20:04:56 <Herrie> Like music app appearantly
Feb 05 20:04:59 <nizovn> Garfonso: those thumbnails are embedded in mp3 file. and seems that virtual fs magically can retrieve them based on path, offset and size
Feb 05 20:20:28 <HoloIRCUser1> Herrie, Nizonn, FWIW, "Triton" was the name of the custom background-task-runner that was ditched in favor of Node. So, if a file has Triton in the name, check to see if it was really being used.
Feb 05 20:24:54 <Herrie> HoloIRCUser1: (AKA DougReeder :P): Yeah but it seems it was still used for filenotifyd (mediaindexer in 3.0.5)
Feb 05 20:25:03 <Herrie> But then as nodeJS variant ;)
Feb 05 20:27:22 <ka6sox> Herrie, did you solve the gcc ICEing up issue?
Feb 05 20:27:34 <Herrie> ka6sox: Rerunning it solved it
Feb 05 20:27:43 <Herrie> Strange it happens randomly lately though
Feb 05 20:27:58 <ka6sox> :/
Feb 05 20:28:52 <HoloIRCUser1> Ok, then :-)
Feb 05 20:35:18 <Garfonso> nizovn: yeah, but if it's easier for us to extract them and save as image files somewhere in var or so, then why not do that?
Feb 05 20:35:54 <Garfonso> Herrie: yeah, but that's bad design. Apps should only rely on the path for the thumb that mediaindexer gives them.
Feb 05 20:36:55 <Herrie> Garfonso: Yeah that's true
Feb 05 20:50:46 <nizovn> Garfonso: i agree, but we will not be compatible with some legacy apps then
**** ENDING LOGGING AT Sat Feb 06 02:59:58 2016