**** BEGIN LOGGING AT Thu Mar 26 02:59:58 2015 Mar 26 06:48:00 Herrie: what is the output for ofono/connman in journal after you tried to connect the cellular context? Mar 26 07:09:52 DougReeder: ping Mar 26 07:33:53 morphis: Let me try Mar 26 07:33:55 Just rebooted Mar 26 07:55:15 morphis: TP is acting up now... Can power modem it seems and have a mbm_1 now as well... Mar 26 07:55:27 good Mar 26 07:55:32 Just behavior doesn't always seem to be consistent Mar 26 07:58:23 Can't power modem I was supposed to write Mar 26 08:00:35 hm Mar 26 08:00:47 so sometines you can, sometimes you can't? Mar 26 08:01:12 Yeah yesterday I could power it today not it seems Mar 26 08:01:28 Yesterday I didn't have mbm_1 only mbm_0 Mar 26 08:03:57 that should be fine Mar 26 08:04:05 mbm_1 would be second modem but you only have one Mar 26 08:05:45 Problem is that I cannot get adb to connect to TP4G on work laptop and haven't been home much :s Mar 26 08:07:18 :) Mar 26 08:09:07 Let me try another USB driver for the TP :s Mar 26 08:09:13 Because N4 works fine :s Mar 26 08:09:18 Herrie|Veer: ok Mar 26 08:09:28 Herrie|Veer: btw. do we have a backwards arrow somewhere in our assets? Mar 26 08:10:41 morphis: Need to check, I guess so, what you need it for? Mar 26 08:11:15 want to replace the cancel button in the messaging app Mar 26 08:13:02 Does DougReeder's commit work? I recall it didn't work still because of a missing id or so? Mar 26 08:13:13 Or maybe I'm confused with another PR Mar 26 09:01:57 morphis: is there a built image with Qt 5.4.1 ? maybe on luna-unstable ? Mar 26 09:27:45 morphis: http://pastebin.com/htj2b2AN Mar 26 09:28:00 Seems ADB works now on laptop with different drivers :) Mar 26 09:28:09 Hope N4 still works too LOL Mar 26 09:36:43 Herrie|Veer: according to the logs the modem itself connected Mar 26 09:37:10 but then it gets deactivated again Mar 26 09:37:45 Herrie|Veer: can you also activate the debugging mode for connman and paste it's log while running connmanctl connect .. Mar 26 09:40:03 How to do that again? Mar 26 09:40:19 add -d in the connman.service file Mar 26 09:42:46 Which was located where again sorry don't have log archive here :P Mar 26 09:42:48 Tofe: not sure, need to ask JaMa Mar 26 09:43:08 Tofe: but let me rebuild later today without 5.4.1 to make sure I don't see it with 5.4.0 Mar 26 09:54:15 ok Mar 26 09:57:12 Tofe: one other thing: do you had a chance to look at the card resize thing? Mar 26 09:58:31 morphis: Where's the connman.service file located? Mar 26 09:59:08 Herrie|Veer: in /lib/systemd/system Mar 26 10:02:09 OK done Mar 26 10:02:14 Should I reboot TP? Mar 26 10:02:19 Or there's a quicker way? Mar 26 10:02:25 yes Mar 26 10:04:32 Ok rebooting Mar 26 10:04:54 How's the bluetooth on N4 by the way? Working & lacking UI integration or? Mar 26 10:05:58 yeah Mar 26 10:08:13 morphis: card resize? you mean for alert cards for example? Mar 26 10:08:24 no, for app cards Mar 26 10:08:41 they have the wrong size/aspect when for example a notification comes in or a dashboard gets opened Mar 26 10:10:44 morphis: I adressed it in tofe/orientation:https://github.com/webOS-ports/luna-next-cardshell/commit/73ef6042e1eee26687325259bd1e643e00633e0b this should be importable into master I guess, but didn't try yet Mar 26 10:12:12 morphis: for some reason if I push a .service file it seems to get a 0 byte size after reboot :s Mar 26 10:12:21 Had this with ofono.service before too :s Mar 26 10:12:59 Tofe: ah, let me try that then Mar 26 10:13:14 Herrie|Veer: then edit the file manually on the device Mar 26 10:13:33 morphis: hmmm Mar 26 10:13:51 Tofe: so how far is https://github.com/webOS-ports/luna-next-cardshell/pull/201 ? Mar 26 10:14:34 morphis: should be ok if we don't integrate the QPA PR Mar 26 10:14:49 ok Mar 26 10:15:06 that one still crashes the compositor? Mar 26 10:15:24 yes, still the issue with last client exiting Mar 26 10:15:33 ah Mar 26 10:15:36 also, we get a black screen on TP, that's quite a mystery Mar 26 10:15:59 because everything looks good in the logs Mar 26 10:16:01 when do we get the black screen? Mar 26 10:16:16 with QPA + orientation patch in cardshell Mar 26 10:16:22 directly after startup? Mar 26 10:16:32 yes Mar 26 10:16:49 I don't understand it yet Mar 26 10:16:54 from what I've seen we need to add display management somehow in sensorfwd Mar 26 10:17:10 and from what I've heard there are problems with using sensors and screen blanking Mar 26 10:17:38 Ah ok, I didn't think it would be that way, I thought it was the sensors that didn't work anymore Mar 26 10:17:56 TP also still gives black screen on screenshots. Not sure that's somehow related but might be... Mar 26 10:19:26 How to edit on device? I don't like vi much :P Mar 26 10:19:59 ed ? :) Mar 26 10:20:00 Herrie|Veer: vi :) Mar 26 10:21:14 Can you mail me connman.service for tp? Seems I don't have an image here and mine is 0 bytes now :s Mar 26 10:24:55 morphis: On the cardshell rest of rotation lock stuff is already implemented. Just pending Tofe's PR & QPA fix :) Mar 26 10:25:16 Just need to work out after reboot orientation stuff still Mar 26 10:25:22 But that's not critical for now Mar 26 10:25:30 Herrie|Veer: http://files.gravedo.de/connman.service Mar 26 10:25:34 Just need to research legacy a bit more for that Mar 26 10:27:47 ok Mar 26 10:27:53 next we need to fix sensorfwd Mar 26 11:13:42 morphis: Finally managed to get the file to device. Edited with hex editor pushed via webOS and copied over to LuneOS mount :s Mar 26 11:14:07 It appears as mbm_1 again now but other then that OK Mar 26 11:14:52 ok Mar 26 11:14:53 Sent you output by email because of 512k limit on pastebin Mar 26 11:15:13 hope rar is OK or you want another format? Mar 26 11:15:21 no that is fine Mar 26 11:15:35 but do you ever executed the connmanctl connect ... command? Mar 26 11:21:51 Yes connmanctl connect cellular_204049010165541_context1 Mar 26 11:24:47 don't see anything which could be the cause of that Mar 26 11:25:03 ah sorry Mar 26 11:29:07 It acts weird, I tried connect again, now have a mbm_1 and a new mbm_2 which has context and operator Mar 26 12:31:39 Herrie|Veer: ? Mar 26 12:59:07 morphis: Not sure what's up Mar 26 12:59:34 It doesn't want to connect to the cellular all other bits seem fairly OK :) Mar 26 13:08:40 Do we have the ofono-test scripts available on device? Mar 26 13:14:01 It could be it's my Ericsson card is faulty but seems to work in webOS so doubt that Mar 26 13:22:57 Seems my Bearer is set to none in org.ofono.ConnectionManager Mar 26 13:23:31 While I have umts available as technology with org.ofono.NetworkRegistation Mar 26 13:30:04 morphis, Herrie, pong. Mar 26 13:33:05 DougReeder: With the merged PR some more work needed to be done to send SMS directly to a number right? Mar 26 13:33:31 Oh, messaging. Mar 26 13:33:44 Yeah, you can’t initiate a conversation. Mar 26 13:34:09 We need to settle responsibilites between the app and the serviece Mar 26 13:34:54 IMNSHO, the app should initiate a conversation or reply to a message by calling the service, not by accessing the DB directly. Mar 26 13:38:35 Agree preferably not done by app itself. morphis your opinion? Mar 26 13:49:21 DougReeder: so why can't we initiate a conversation? Mar 26 13:50:01 Some record for the conversation needs to be set up first, IIRC. Mar 26 13:52:06 so the service creating a chat thread after the app has created the message doesn't work? Mar 26 13:53:50 That flow might be possible, but would be fragile. Mar 26 13:54:49 Watches are really for loosely-coupled processes. Tightly-coulpled processes should use direct calls. Mar 26 13:57:51 Watches don’t allow returning an error when something goes wrong. Mar 26 13:58:25 morphis: Any idea for ofono/connman? I can swap in my Sierra some time to see what it does and if it's Ericsson specific? Mar 26 14:01:22 Herrie|Veer: yeah, please Mar 26 14:01:42 DougReeder: ok, so what if the app creates the chatthread and the message Mar 26 14:01:55 OK will try to see if I can do that tonight Mar 26 14:02:02 the chatthreader wouldn't kick in if the message akready has a thread assigned Mar 26 14:07:33 Herrie|Veer: the bearer gets only set when a connection is established Mar 26 14:08:31 You need a software layer to manage the chat thread and message records. If it lives in the app, that’s problematic for incoming messages. If it’s in a service everything is copacetic. If you have it in both, you’re duplicating code, which leads to bugs. Mar 26 14:09:02 OK any other suggestions for the Ericsson currently? I don't see anything that seems really off currently Mar 26 14:09:28 Herrie|Veer: not really so far Mar 26 14:09:34 need to think a bit about this Mar 26 14:09:48 DougReeder: hm Mar 26 14:12:42 DougReeder: what do you prefer? Mar 26 14:13:45 It seems to me, that the chatthreader service should handle creation and updating of chat threads and messages. Mar 26 14:14:09 so we would create a method createThread which we would simply call then? Mar 26 14:14:13 The Messaging app could have read-only access to the DB records. Mar 26 14:14:38 I haven’t thought about the API in detail. Mar 26 14:15:35 The call might have the arguments of messaging addr and message body. Leave thread management to the service. Mar 26 14:16:01 ok Mar 26 14:16:03 So the chatthreader could attach the message to an existing thread, or create a new one. Mar 26 14:16:23 so when it attaches it to an existing thread how would the app behave then? Mar 26 14:16:28 switch to the attached thread? Mar 26 14:16:50 The call would return the thread… Mar 26 14:17:02 morphis: that's what legacy currently does I think Mar 26 14:17:46 ok Mar 26 14:17:50 then why not Mar 26 14:18:35 Yeah, I think for new conversations, the app would start keeping track of the thread. Mar 26 14:19:17 btw. I did some visual fixes for the messaging app Mar 26 14:19:56 see https://github.com/webOS-ports/org.webosports.messaging/tree/morphis/work Mar 26 14:20:23 Some thought should be given to replying. The app would know of a thread. I’m not sure, offhand, what to do if the app has a “bad” thread or the “wrong” thread. Though that should be rare. Mar 26 14:21:39 Herrie, does legacy ever start a new thread for the same person? Mar 26 14:22:28 You could have a thread for one IM addr, and a different thread for another IM addr, then edit contacts so they were the same person. Mar 26 14:23:35 So, morphis, I think you’re right: the chatthreader should return the “right” thread, and the app should check. If the app was using the “wrong” thread, it should load the new thread and messages. Mar 26 14:24:04 Almost always, the chatthreader would return the same thread that the app was using. Mar 26 14:24:13 DougReeder: I actually send a SMS directly in a new thread yesterday. I didn't remember I already had a thread. After it sent it would add it to the existing thread and switch there Mar 26 14:24:52 So, the API call would pass IM addr and message body, and return thread. Mar 26 14:26:30 yeah Mar 26 14:28:56 This API also handles edge cases like an message comming in before the app finds out about it. Mar 26 14:30:49 I think the same code in the app would handle a new thread (initiating conversation) as changing thread (some edge case). Mar 26 14:35:52 good Mar 26 14:35:58 So, the API call just needs to return the thread ID. Mar 26 14:36:01 DougReeder: so you're going to implement this in the service? Mar 26 14:36:18 Heh. I’ve got a lot on my plate. Mar 26 14:36:41 Do we want to defer the remaining work on Contacts? Mar 26 14:39:19 what is missing on contacts currently? Mar 26 14:39:34 on how much time do think it would need to implement what we discussed for the messaging app? Mar 26 14:41:27 There’s a large number of smaller things that need to be done in Contacts: pictures, Reminders, initiatin calls, emails, & IMs, etc. Mar 26 14:42:31 There appears to be some issues with contact sync ATM, so even with editing implemented, I dont’ think one can properly beta-test it. Mar 26 14:43:04 DougReeder: I think it would be good to have this relatively minor fix in for messaging together with db watches Mar 26 14:43:28 right Mar 26 14:43:28 Right now, I’m updating Serene Notes and adpating it to LuneOS. Mar 26 14:44:10 I can start work on Messaging after I get Serene Notes up to speed. Mar 26 14:44:19 good Mar 26 14:44:53 I haven’t looked at the chatthreader at all, so I can only guess this could take weeks. Mar 26 14:45:13 hm Mar 26 14:45:24 do we have a list of things we need to do for contacts? Mar 26 14:46:02 There’s a few items on the issues site, and the README has a list of smaller things. Mar 26 14:46:31 can you update that with what is missing? Mar 26 14:47:13 Together, they’re complete with everthing I know of. Mar 26 14:47:46 Update which? Mar 26 14:49:06 the list in the README, but if it is complete Mar 26 14:50:28 So, the unimplemented features are all in https://github.com/DougReeder/org.webosports.app.contacts/blob/master/README.md and issues.webos-ports has bugs. Mar 26 14:51:33 Herrie|Veer: morphis: I've built 5.4.1+testing only locally and it will probably be already outdated a bit Mar 26 14:51:43 I can send PR in github if you want to test it Mar 26 14:51:50 Have you had time to look at https://github.com/webOS-ports/org.webosports.app.contacts/pull/16 ? Mar 26 14:55:55 JaMa: already picked a few changes from jansa/dizzy locally Mar 26 14:56:04 DougReeder: looks good Mar 26 14:56:17 DougReeder: fine for me Mar 26 14:56:25 * DougReeder nods Mar 26 14:59:18 DougReeder: merged Mar 26 14:59:39 good :) Mar 26 14:59:58 Thank you. Do you want to update meta-webos-ports, or shall I? Mar 26 15:00:05 DougReeder: will do that Mar 26 15:02:23 JaMa: so far I only saw one regression with the keyboard Mar 26 15:13:10 We won’t be able to Synergize WhatsApp: http://mashable.com/2015/03/25/whatsapp-developers-api/?utm_cid=mash-com-Tw-main-link Mar 26 15:14:15 :) Mar 26 15:18:39 There are some good ideas on communicating hidden gestures to users in the 2 min video at https://www.youtube.com/watch?v=3mbt6oDpHuc Mar 26 15:24:00 DougReeder: They never had a public API. Everything that was out there including Yowsup, Yapari, MojoWhatsUp is based on reverse engineering the protocol. Mar 26 15:25:19 IMNSHO, reverse engineering is generally a waste of time, in the end. Mar 26 15:27:09 Look at Palm and iTunes. Mar 26 15:27:31 eh? Mar 26 15:28:21 DougReeder: Not a big fan of it but we're allowed in most of EU hence it happens. US is a bit stricter I believe... Mar 26 15:30:06 The problem is the originators are free to change anything at any time, without notice. So there’s down time until you catch up. Mar 26 15:32:03 http://www.techhive.com/article/181057/Palm_Lost_the_iTunes_Sync_Battle_Again.html Mar 26 15:34:20 Palm sunk a fair amount of engineering resources into iTunes sync, that could have been better employed elsewhere. Mar 26 15:34:40 ah Mar 26 16:08:19 JaMa: I wrote a small script recently which automatically bumps the SRCREV for our components Mar 26 16:08:34 what do you think about running that automatically and if any updates available trigger a nightly build Mar 26 16:45:32 Tofe: problem does only appear with 5.4.1 Mar 26 16:47:09 morphis: why not use small .inc file with AUTOREVs? Mar 26 16:58:53 JaMa: thought that we would have more control over what we build then Mar 26 17:03:39 it will be included only in this special nightly build, right? Mar 26 17:03:56 you mean the SRCREV bump Mar 26 17:07:09 I meant .inc file with AUTOREVs Mar 26 17:18:22 JaMa: but then a local clone would get out of sync with those nightlies Mar 26 17:37:30 Tofe: ping Mar 26 18:23:33 morphis: Hope to have some time later tonight or tomorrow morning to switch the Sierra for the Ericsson on the TP4G and will let you know results. Any logs you're interested in? Mar 26 18:23:51 Herrie|Veer: the usual ones Mar 26 18:24:28 OK so journalctl --no-page - u ofono & also for connman? Mar 26 18:24:45 I suspect somehow the Sierra will work it's less picky config wise Mar 26 18:34:59 Just to let you know, currently not much of a solution, but luna-next, maliit-server is currently running on my device (but without display output...) I now have to probe several things... Mar 26 18:37:53 sewziq: Still sounds like progress :) Mar 26 18:38:19 morphis or Tofe might have some ideas still :) Mar 26 18:39:14 morphis/Tofe: Aren't there any device specific patches applied for Mako and Tenderloin in past 6 months that would need to find their way to Grouper as well? Mar 26 18:41:19 I only tried some "dirty" steps, changed the calls to create the display etc. to the ones found in test_glesv2. (just to look if the egl-calls dont't crash. Now I will revert the calls and look into the parameters etc. ... Mar 26 18:42:08 will take some time - if I'm ever able... Mar 26 18:42:41 Good thing is morphis found his N7 :P Mar 26 18:43:04 I'm only happy to beeing at this point! :-) Mar 26 18:44:28 All this stuff is way too low level for me :P Mar 26 18:44:41 I'll focus on some other JS/QML stuffs :P Mar 26 19:52:55 * EricBlade looks at ragingmind Mar 26 19:53:25 * EricBlade notes the Michigan IP Mar 26 20:06:59 Herrie: all this stuff is too... "other" than the things I usually do... :-) Mar 26 20:11:36 morphis: Just swapped the Sierra in, now booting ;) Mar 26 20:19:24 Don't seem to get signal, might need to swap cable Mar 26 20:21:58 Ah forgot to connect red "main" cable :P Mar 26 20:22:06 THat migth explain why I didn't get signal LOL Mar 26 20:22:10 Connected now and booting Mar 26 20:23:44 OK signal now, trying to connect Mar 26 20:26:10 OK it connects, no signal bar Mar 26 20:27:04 And when I try to launch google.com in browser it doesn't connect Mar 26 20:28:57 Some output: http://pastebin.com/MvG4Kyp9 Mar 26 20:40:56 morphis: Sent you some logs :) Mar 26 21:12:33 morphis: http://pastebin.com/v96hEE8d Mar 26 21:24:21 Seems the price for the Urbante LTE is out: $ 590,- Mar 26 21:24:28 A bit steep, but quite sure it'll drop soon ;) **** ENDING LOGGING AT Fri Mar 27 02:59:59 2015