**** BEGIN LOGGING AT Thu Jan 12 03:00:01 2017 Jan 12 03:15:20 Herrie, this pic should make clear the change I’m proposing: https://hominidsoftware.com/non-public/MsgAddrPicker-move-search.png Jan 12 03:16:28 Separately, I’m experimenting with a Mojo-style list: https://hominidsoftware.com/non-public/MsgAddrPicker-Mojostyle.png Do you like it? Jan 12 03:16:58 The Mojo-style list is a little more compact than the flat list. Jan 12 05:58:46 DougReeder: Fine with me to move it! Jan 12 05:58:52 I like the Mojo style list! Jan 12 08:06:21 granpc: ok, the mystery deepens Jan 12 13:06:53 Herrie, Mojo-style list PR is https://github.com/webOS-ports/org.webosports.messaging/pull/49 Jan 12 14:31:54 DougReeder: Thanks! Jan 12 14:39:48 How difficult would it be to get a "transport picker" in the actual chat thread now? Currently this has 2 dummy values. Jan 12 14:51:46 I'm not sure. Jan 12 14:53:23 I guess the alternatives in the list depend On What mesg addr the contact has. Jan 12 14:53:53 I never used that on webOS. Jan 12 14:54:51 But the alternatives are the user's account, not the recipient's? Jan 12 14:56:36 We probably need to have the app store a list of the user's accounts somewhere. Jan 12 15:06:34 Herrie|Pre3: I just ordered a N5 for me :p Jan 12 15:08:02 DougReederPhone: It basically shows the different "IM" options for that contact.... Jan 12 15:08:48 I.e. you have a chat thread with me that contains all your conversations via SMS, Skype, ICQ with me. By setting this the next message would then be sent via this service. Jan 12 15:09:41 Tofe: NICE! Jan 12 15:10:20 Herrie|Pre3: what is the current status for that port? Jan 12 15:11:25 Tofe: I made an initial start with updating repos to working CM11 bits based on our porting guide but got stuck there. I guess a 12.1 fresh port is easier to pick up. Jan 12 15:13:36 DougReederPhone: Normally the transport picker would pick the same service to reply, but let's say I get a Skype msg from you and I want to reply but noticed you went offline and I want to send SMS instead.. Jan 12 15:14:24 Herrie|Pre3: ok, so we don't have anythhing "hammerhead" so far in meta-smartphones, for example Jan 12 15:14:39 Not really Jan 12 15:14:50 I might have some things locally still need to check Jan 12 15:15:01 Fine, I'll get prepare the field from a mako copy Jan 12 15:15:25 But started to try to understand by replicating Mako & Maguro build but those were broken Jan 12 15:15:53 So I fixed those but never really got to work much on Hammerhead and push much tbh Jan 12 15:16:18 ok :) Then it'll be a real life test for the brand new hybris thingy Jan 12 15:17:05 The trickiest might be the kernel, as we don't have any working example I guess... Jan 12 15:21:43 Tofe: any ideas? Jan 12 15:22:09 I was starting to think that maybe the mediatek drivers were using TLS directly as in this issue -- https://github.com/mer-hybris/libhybris/pull/5 Jan 12 15:22:35 but since it didnt crash on ubuntu, probably not. (unless ubuntu isn't using TLS, which it probably is) Jan 12 15:24:05 btw, I tried using the system partition from ubuntu on luneos, but that just crashed immediately & failed to link against libc Jan 12 15:24:11 granpc: no new ideas so far, on my side... Jan 12 15:24:16 so not quite sure what's up with that Jan 12 15:24:27 I'd assume they should be intercompatible if they're more or less the same thing Jan 12 15:25:08 yes, it's what is the most surprising here, there shouldn't be so many differences Jan 12 15:26:21 ah wait, you mean you swapped /system with ubuntu's ? Jan 12 15:29:03 a failure to link with libc means there is a hybris-non hybris mismatch in the stack Jan 12 15:29:36 like a patch libc used with a native libhardware, things like that Jan 12 15:29:40 patched* Jan 12 15:29:48 that doesn't make much sense though, if it worked in ubuntu Jan 12 15:30:00 hmm Jan 12 15:30:08 depends what executable you start Jan 12 15:30:12 i'm gonna try that again Jan 12 15:48:27 Herrie, it sounds like that would be less work than the msg addr picker. Jan 12 15:49:16 Sounds like that's more important than moving the search field. Jan 12 16:41:26 DougReederPhone: Yeah, let me post some screenshots when I'm home later Jan 12 16:41:31 So you have some visuals. Jan 12 17:35:02 this is weird Jan 12 17:35:17 I tried it with ubuntu's /system partition and updated libhybris with upstream patches Jan 12 17:35:31 it gets further than before but still crashes Jan 12 17:35:33 https://paste2.org/NKCK2G6t Jan 12 17:36:12 still works fine on ubuntu touch though Jan 12 17:36:12 https://paste2.org/A6CW6DjI Jan 12 17:36:19 although it's weird that it shows different api versions? could that be the issue Jan 12 18:03:57 granpc: btw do you see any errors in logcat? Jan 12 18:06:35 DougReederPhone: http://herrie.org/transport.png Jan 12 18:06:41 If you want I can paste the legacy code for this somewhere Jan 12 18:09:42 nizovn: I do, but I think they were the same errors from ubuntu touch Jan 12 18:09:49 just gps warnings Jan 12 18:10:01 and something called "6620_launcher", I'm not sure what it is Jan 12 18:10:07 mediatek crap Jan 12 18:10:18 ok Jan 12 18:11:23 hmmm actually Jan 12 18:11:26 you might be onto something Jan 12 18:12:11 http://paste2.org/fspxI51w Jan 12 18:19:10 Herrie, you have 5 phone numbers? Jan 12 18:26:54 No it was the Mrs with various international sims for when we were traveling to China & Russia LOL :P Jan 12 18:32:12 & Uzbekistan LO Jan 12 18:32:14 +L Jan 12 18:47:25 :-) Jan 12 18:48:14 Anyway, I think I grok it. Jan 12 20:14:37 DougReederPhone: This is the 3.x code for transportPicker.js: https://bpaste.net/show/29a1346ed838 Jan 12 20:20:03 granpc: can you check diffs between the defconfigs that are used? Jan 12 20:21:04 Herrie: I'm downloading the maguro testing image, to test it on my side Jan 12 20:23:15 But I've got a pretty good cold, so I'm not at 100% :p Jan 12 20:23:36 Tofe: https://paste2.org/PYgkHXpF Jan 12 20:23:39 seems minor changes Jan 12 20:23:56 I actually remember making all of those while reading the porting guide Jan 12 20:25:15 (if you're wondering about stuff like CONFIG_APPARMOR, I actually hadn't gotten around porting apparmor to the ubuntu rom yet) Jan 12 20:26:31 looks good, yes Jan 12 20:27:25 so, if the /dev isn't created by a kernel module, than it's maybe an android service that starts it ? You have the same init.rc for both ubuntu and luneos ? Jan 12 20:27:44 yep, same init.rc except for the on-the-fly modifications done by luneos Jan 12 20:28:01 The init.rc I propose in my fork was only tested on qualcomm devices, it might be incomplete/incorrect for others Jan 12 20:28:22 you mean the sed replacements? Jan 12 20:28:51 no, this: https://github.com/Tofee/android_system_core/commit/deddf2bf2262897bfe95382a88d5fc1e8ced099b Jan 12 20:29:10 oh right. hm Jan 12 20:29:11 it includes import /init.${ro.hardware}.rc so it should be fine, but who knows Jan 12 20:29:14 I don't think I'm using that at all Jan 12 20:29:33 Tofe: OK Jan 12 20:29:41 yeah, mine still even contains bootanimation Jan 12 20:30:15 ... and the bootanimation shows up when lxc starts ? Jan 12 20:30:47 no Jan 12 20:30:51 oh, ok Jan 12 20:30:54 it's disabled Jan 12 20:30:57 but not entirely removed like in your fork Jan 12 20:31:15 also I just noticed I added /system/bin/sensorservice to the list of disabled services in 20-remove-services Jan 12 20:31:24 which probably has a whole lot to do Jan 12 20:31:39 ah, yes, might explain partially the issue Jan 12 20:33:04 nah same thing. same messages even Jan 12 20:33:19 I'm gonna go check if the same messages get printed in ubuntu, they probably do Jan 12 20:56:28 Herrie: no bootloop here on maguro, all looks fine Jan 12 21:01:38 Tofe: same messages on ubuntu, no crash: http://paste2.org/fLyKYXNC Jan 12 21:01:50 it's really weird that in the luneos version it's crashing while trying to printf a float Jan 12 21:02:06 I'm still thinking of TLS issues but it does not make much sense at this point, if i'm using the same android fs as ubuntu and all Jan 12 21:02:41 ok Jan 12 21:03:01 also none of the sensors actually work Jan 12 21:03:05 when I try to poll events Jan 12 21:03:10 I'm not really overly concerned about that right now though Jan 12 21:32:33 Tofe: OK... I'm using cm-12.1-20160719-SNAPSHOT-YOG7DAO8DL-maguro.zip Jan 12 21:32:39 Might need to retry I guess **** ENDING LOGGING AT Fri Jan 13 03:00:01 2017