**** BEGIN LOGGING AT Thu Feb 19 02:59:58 2015
Feb 19 06:07:16 <Garfonso> I see, we need more people in the US shift. :)
Feb 19 06:17:32 <DougReeder> Hmm?
Feb 19 06:34:21 <DougReeder> Okay, I have a fix for the CrossAppUI under Enyo 2.5
Feb 19 06:45:36 <morphis> DougReeder: yeah!
Feb 19 06:46:01 <DougReeder> Now I need to untangle git, so there’s a usable PR.
Feb 19 06:47:48 <morphis> DougReeder: was it the problem with the src?
Feb 19 06:48:03 <DougReeder> Yup.
Feb 19 06:48:12 <DougReeder> -               this.setSrc(targetPath);
Feb 19 06:48:13 <DougReeder> +               this.setAttribute("src", targetPath);
Feb 19 06:48:44 <morphis> ah
Feb 19 07:03:09 <DougReeder> https://github.com/webOS-ports/org.webosports.app.settings/pull/44
Feb 19 07:03:37 <morphis> DougReeder: thanks, can you PR those changes also directly to webos-lib?
Feb 19 07:03:45 * DougReeder nods
Feb 19 07:04:04 <DougReeder> I’ll roll them into the overlap removal
Feb 19 07:04:24 <DougReeder> … or, maybe not, since that will require app changes
Feb 19 07:29:16 <DougReeder> https://github.com/webOS-ports/webos-lib/pull/13
Feb 19 12:42:43 <EricBlade> data that might be worth knowing: on a webOS TV, it takes 0.7sec to start a dynamic node service
Feb 19 12:45:07 <Tofe> EricBlade: app startup, you mean ?
Feb 19 12:45:53 <EricBlade> well say you have a node.js service, com.webos.service.something .. the bare minimum time before your service code is running on a cold start is 0.7sec
Feb 19 12:46:19 <Tofe> ok
Feb 19 12:46:31 <EricBlade> so when someone calls palm://com.webos.service.something/doSomething '{}' .. it'll take 0.7sec to get the most minimal of action
Feb 19 12:46:54 <Tofe> only for the first call, for still
Feb 19 12:47:04 <Tofe> s/for/but/
Feb 19 12:47:23 <EricBlade> right. after that, if it remains running, that 0.7sec is gone
Feb 19 12:48:26 <EricBlade> so expecting a response or action from a dynamic service is a little problematic.  not nearly as bad as it was on a Pre :-) but problematic
Feb 19 12:52:22 <EricBlade> i don't know what the default service timeout is in Open webOS, but on TV it's a lot shorter than legacy was .. i think the problem is an entirely different classification now than it was with legacy
Feb 19 13:48:48 <cobblemonster> yes Garfonso  iam all lone ,old,slow and COLD  -23 this morring over here in the US
Feb 19 15:49:43 <morphis> EricBlade: afaik most services have 30s as default timeout but I think the startup time can be improved a lot
Feb 19 15:49:56 <morphis> also it looks like our service performance can be really improved
Feb 19 15:50:36 <morphis> EricBlade: see http://issues.webos-ports.org/issues/699
Feb 19 16:04:48 <EricBlade> that doesn't sound very good.. although it's quite difficult to compare node 0.2, 0.4, and 0.10 they are very different beasts
Feb 19 16:05:59 <EricBlade> i was speaking purely of time it takes to get to code execution, though.  although if Doug's service is actually getting stopped and restarted frequently that could be a huge bottleneck
Feb 19 16:42:48 <DougReeder> I don’t think my service gets stopped and started - I checked the logs.
Feb 19 16:44:24 <EricBlade> that's certainly odd, though.  i wouldn't expect it to be that much slower.
Feb 19 16:44:33 * DougReeder nods
Feb 19 16:46:05 <EricBlade> without having any idea what it does, all i can think of is to add a bunch of logs to it at various points, and see if you can figure out what part of it takes so much longer
Feb 19 16:47:22 <DougReeder> When we get serious about performance, it will probably make more sense to write separate tests for processing bound and i/o bound tasks.
Feb 19 16:48:05 <DougReeder> My service has a lot of moving parts.
Feb 19 16:49:13 <EricBlade> i wonder if activitymanager doesn't already have something hidden somewhere in it to handle this case of wanting to start up a service before making a real call to it.
Feb 19 16:50:07 <DougReeder> If response time is a concern, one could call a NOP call beforehand.
Feb 19 16:51:18 <EricBlade> yeah my current idea is to make a call that will just adopt an activity that flags it interactive
Feb 19 16:51:57 <DougReeder> But, what’s the use case?
Feb 19 16:52:42 <EricBlade> user does something, and the ui is expecting a response from the service
Feb 19 16:53:18 <EricBlade> i can't get too much more detail :-)
Feb 19 16:53:31 <EricBlade> s/get/give
Feb 19 16:54:26 <DougReeder> ONly use case I can think of offhand is typedown search - and that would more typically use DB8, which is always running, yes?
Feb 19 16:55:41 <DougReeder> Without a real task that real users need to do, I suspect that falls under YAGNI.
Feb 19 16:55:42 <EricBlade> yeah, there's a service between the UI and db8 here, that handles the interactions .. i don't like having component A directly mucking around with a database that component B owns
Feb 19 17:54:50 <Herrie|2> morphis: http://pastebin.com/Rq0UggE6
Feb 19 17:54:57 <Herrie|2> Connman side seems OK for TP4G?
Feb 19 17:55:02 <Herrie|2> Still no bars though?
Feb 19 18:17:47 <Herrie|2> Tofe: https://github.com/webOS-ports/luna-next-cardshell/pull/190
Feb 19 18:18:14 <Herrie|2> Tofe/morphis: This improves the layout on TP a bit. On Digital Clock still no text. Not sure what's up with that... Did some testing but cannot figure it out... Tofe ideas?
Feb 19 18:26:19 <morphis> Herrie|2: and a connmanctl connect cellular_204049010165541_context1 doesn't work?
Feb 19 18:28:15 <Herrie|2> No get an "operation aborted" on that one
Feb 19 18:28:27 <Herrie|2> I confirmed the card works in legacy 3.0.5
Feb 19 18:28:34 <Herrie|2> Just to exclude that to be the problem
Feb 19 18:29:07 <morphis> Herrie|2: can you modify /lib/systemd/system/ofono.service
Feb 19 18:29:25 <morphis> to ExecStart=/usr/sbin/ofonod -n --nobacktrace
Feb 19 18:29:27 <morphis> add -d
Feb 19 18:29:31 <morphis> reboot and retry then
Feb 19 18:32:59 <Herrie|2> Ok rebooting
Feb 19 18:36:31 <Herrie|2> Hmmz now cellular is gone completely from connmanctl technologies and services
Feb 19 18:37:15 <Herrie|2> I have the following in ofono.service: ExecStart=/usr/sbin/ofonod -n --nobacktrace -d
Feb 19 18:38:02 <Herrie|2> Ah wait ofono.service = 0 bytes
Feb 19 18:42:13 <Herrie|2> Hmmz I can confirm it's 259 bytes when I put it
Feb 19 18:42:18 <Herrie|2> Then when I reboot it's 0 bytes
Feb 19 19:02:48 <Herrie|2> Ok needed to chmod
Feb 19 19:03:49 <Herrie|2> now mbm_0 shows
Feb 19 19:06:43 <Herrie|2> morphis: Log shows: Feb 18 06:56:47 tenderloin ofonod[1627]: create_ril: can't create unix socket: Too many open files (24)
Feb 19 19:07:09 <Herrie|2> It shouldn't do RIL for TP4G I thought?
Feb 19 19:10:22 <morphis> Herrie|2: ignore that
Feb 19 19:10:32 <morphis> it just tries to probe the ril but doesn't find it
Feb 19 19:10:50 <Herrie|2> OK
Feb 19 19:10:55 <Herrie|2> Let me reboot and pull a clean log
Feb 19 19:11:05 <Herrie|2> See lots of ofono stuff in log now
Feb 19 19:15:02 <Herrie|2> http://pastebin.com/z9zA9Fpd
Feb 19 19:16:46 <morphis> Herrie|2: can you filter that for ofono with
Feb 19 19:16:52 <morphis> journalctl --no-pager -u ofono
Feb 19 19:16:56 <morphis> and then do the connect
Feb 19 19:17:17 <Herrie|2> THis is what I get currently: http://pastebin.com/x1wJQF2F
Feb 19 19:19:08 <Herrie|2> http://pastebin.com/GKhjCAZ7
Feb 19 19:24:43 <morphis> Herrie|2: how does the log look?
Feb 19 19:27:15 <Herrie|2> https://bpaste.net/show/24d9cd3cb117
Feb 19 19:35:36 <morphis> Herrie|2: can you add a -P ril to the ofono cmdline?
Feb 19 19:44:53 <Herrie|2> To the ofono.service?
Feb 19 19:52:26 <morphis> right
Feb 19 20:03:59 <morphis> Herrie|2: you have that?
Feb 19 20:12:25 <Herrie|2> Let me do now
Feb 19 20:12:33 <Herrie|2> Little one needed a bath ;)
Feb 19 20:14:27 <morphis> ok
Feb 19 20:16:18 <Tofe> Herrie|2: testing Settings.height/width isn't the correct way to handle this, as this will not really handle screen orientation changes very well
Feb 19 20:16:59 <Tofe> but I understand there isn't much choice right now, as we don't have any formal screen orientation information
Feb 19 20:17:19 <Herrie|2> Yes it's a dirty hack for now
Feb 19 20:17:34 <Herrie|2> 2 analog clocks work, just text on digital is gone
Feb 19 20:19:09 <Tofe> Herrie|2: how do you usually test it?
Feb 19 20:22:23 <Herrie|2> On TP just
Feb 19 20:22:33 <Herrie|2> I don't know how to trigger it easily in QT Creator ;)
Feb 19 20:22:37 <Herrie|2> N4 looks OK
Feb 19 20:22:44 <Herrie|2> I might need to adjust scaling a little
Feb 19 20:22:50 <Herrie|2> Clock is a bit bigger compared to legacy
Feb 19 20:26:41 <Herrie|2> hmmz my ofono.service gets to 0 bytes after I push it :S
Feb 19 20:27:05 <elvispre> Hi. "I have read the rules."
Feb 19 20:27:25 <elvispre> I have been working on Calculator lately.
Feb 19 20:27:42 * elvispre uses the name corpuscle on GitHub.
Feb 19 20:27:47 <Herrie|2> :D
Feb 19 20:27:48 <GodGinrai> elvispre: give it a TI-83 mode
Feb 19 20:27:50 <GodGinrai> :P
Feb 19 20:27:50 <Herrie|2> Good stuff!
Feb 19 20:27:57 <Herrie|2> Looks really good :)
Feb 19 20:28:02 <Herrie|2> I've been reviewing and merging!
Feb 19 20:28:19 <elvispre> I thought I might call it good enough for now and have a look at something else...
Feb 19 20:28:24 <Tofe> Herrie|2: in LunaService.qml, in the Tests, change the default value of currentLockStatus to "dockmode"
Feb 19 20:28:39 <Herrie|2> Tofe: :D
Feb 19 20:28:53 <Herrie|2> now changed with Notepad instead of NOtepad++ maybe that helps
Feb 19 20:29:24 <Herrie|2> elvispre: How about: http://issues.webos-ports.org/issues/845
Feb 19 20:30:26 <Tofe> Herrie|2: I've got text in the digital clock in qtcreator; well, it's images, but still, it's there
Feb 19 20:30:43 <Tofe> I see the : if that's what you meant
Feb 19 20:32:31 <elvispre> Herrie: Where would I start?
Feb 19 20:33:55 <elvispre> GodGinrai: :-) Maybe later.
Feb 19 20:38:24 <elvispre> OK. Issue 845: Maybe just a standalone app to )configure and) make those calls. That would be a start. I'll see what I can do.
Feb 19 20:47:06 <elvispre> Patch the Settings app, in fact...
Feb 19 20:59:03 <Herrie|Veer> elvispre: If you want something a bit easier to start with, maybe you can add timezone selection to the Settings app
Feb 19 20:59:29 <Herrie|Veer> I did the initial Timezone settings in org.webosports.app.firstuse recently in QML
Feb 19 20:59:51 <Herrie|Veer> Should give you some pointers for what calls to make in Enyo etc :)
Feb 19 21:00:47 <elvispre> Herrie: I'll take a look at the Settings app and have a play. Thanks for the suggestions.
Feb 19 21:01:39 <Herrie|Veer> For Timezone on 3.x it opens a new page (panel) when you press on it.
Feb 19 21:02:08 <Herrie|Veer> This could be replicated in LuneOS and would allow you to develop it quite independently :)
Feb 19 21:02:27 <Herrie|Veer> It's in Date & Time app in 3.x
Feb 19 21:05:58 <Herrie|2> This is how it looks in 3.x: http://tinypic.com/r/4tq0q1/8
Feb 19 21:08:16 <Herrie|2> And the work I've done in QML: https://github.com/webOS-ports/org.webosports.app.firstuse/blob/master/qml/TimeZonePage.qml
Feb 19 21:15:48 <Herrie> The settings app is available online via: https://webos-ports.github.io/org.webosports.app.settings/debug.html
Feb 19 21:39:54 <elvispre> Herrie: So it is. I have it locally now too. Good that it runs in the browser.
Feb 19 21:53:00 <cobblemonster> Herrie more work pushed to the calendar  :)
Feb 19 21:53:09 <cobblemonster> your that is
Feb 19 23:33:47 <Garfonso> We got someone working on calendar? :)
Feb 19 23:35:34 <Garfonso> Issue 699 puzzles me a bit. C+Dav always felt a lot faster on LuneOS. Maybe that's my "new is good" bias? :-)
Feb 19 23:36:19 <Garfonso> Actually everything is better than node 0.2 when it comes to network...
Feb 20 00:41:23 <cobblemonster> Yes Garfonso  iam picking at a littel ever day working on the week view
**** ENDING LOGGING AT Fri Feb 20 02:59:59 2015