**** BEGIN LOGGING AT Sun Jan 24 02:59:59 2016 Jan 24 03:55:02 Herrie: If you get a chance, could you test this please? https://github.com/webOS-ports/org.webosports.app.pdf/pull/12 Jan 24 03:55:23 debug.html doesn't pull up a PDF, so at the moment, I can't really test it to make sure it works. Jan 24 04:19:32 Well, this at least seems to fix the debug.html functionality by adding some working mock data: https://github.com/webOS-ports/org.webosports.app.pdf/pull/13 Jan 24 04:20:41 And with that, I should be going. Jan 24 05:26:38 Herrie, I’m not sure what to do with #743. I can’t tell from the description whether he’s talking about actual behavior or expected bahvior. Jan 24 07:57:32 morning Jan 24 07:57:50 DougReeder: OK I'll ask Matt :) I somehow assumed this was one of the bugs you opened LOL Jan 24 09:28:59 Morning! Jan 24 09:29:12 Herrie|Laptop: ping Jan 24 09:30:58 Herrie|Laptop: don't you think the db8 configuration files (kinds and permission) should be in the "configuration" folder for the phone, like what has been done for the browser? Jan 24 10:22:38 Tofe: Morning Jan 24 10:22:47 Tofe: pong, ehm let me see Jan 24 10:22:58 I think I took it from how it's done with other components ;) Jan 24 10:23:44 I have already the commit corresponding to the move, if needed Jan 24 10:24:56 Tofe: https://github.com/webOS-ports/mediaindexer/tree/master/files for example Jan 24 10:25:04 And I know there were more like that Jan 24 10:25:44 but then, what about the browser? Jan 24 10:26:10 Like: https://github.com/webOS-ports/luna-universalsearchmgr/tree/webOS-ports/master/files Jan 24 10:26:49 Tofe: We should be consistent ;) Jan 24 10:26:55 So I guess we'd need to move it for some Jan 24 10:27:10 From OWO seems to follow the files folder pattern for most but also not for all Jan 24 10:27:21 https://github.com/webOS-ports/sleepd/tree/webOS-ports/master/files Jan 24 10:27:33 https://github.com/webOS-ports/activitymanager/tree/webOS-ports/master/files Jan 24 10:27:37 For me it's not really about consistency, it's about having it working :) as it uses predefined macros to handle the packaging, I guess only one way works Jan 24 10:28:33 Tofe: Well consistency is good, just various bits would need to make sure it ends up at the right place :P Jan 24 10:28:42 This one is wrong from OWO for example: https://github.com/webOS-ports/app-services/tree/master/com.palm.service.contacts.linker Jan 24 10:28:46 So, what about the browser ? Why does it work ? Jan 24 10:29:14 Tofe: The location of the files doesn't matter as long as it ends up at the proper location on device Jan 24 10:29:59 I think it MIGHT come not from Browser recipe but from elsewhere :P Jan 24 10:30:04 LEt me see Jan 24 10:30:36 Herrie|2: well, at some point, the build must know where to look for these files Jan 24 10:31:47 Tofe: Yeah Jan 24 10:31:59 Well it seems it's not in CMakelist.txt nor in recipe Jan 24 10:32:05 So I reckon they come from elsewhere Jan 24 10:32:07 Let me check Jan 24 10:32:41 Also OWO sometimes uses DB folder, sometimes DB8 Jan 24 10:32:47 I reckon we should use DB8 Jan 24 10:34:45 Herrie|2: I guess the recipe includes another recipe for handling this kind of things Jan 24 10:36:01 In the meantime I've found what's wrong; it's the Db8Model C++ implementation Jan 24 10:37:20 Tofe: https://github.com/webOS-ports/meta-webos-ports/blob/fido/meta-luneos/classes/webos_application.bbclass Jan 24 10:37:26 That's where it comes from Jan 24 10:37:47 Could be the location is different for apps & system components Jan 24 10:37:52 I'll dig a bit into it Jan 24 10:39:32 Yeah seems apps use configuration\db folder like https://github.com/openwebos/core-apps/tree/master/com.palm.app.calendar and https://github.com/openwebos/core-apps/tree/master/com.palm.app.clock Jan 24 10:39:54 Email & memos too Jan 24 10:40:05 But components seem to use the files/db structure Jan 24 10:40:22 so, what about phone ? Jan 24 10:43:33 I guess we should keep it configuration/db for apps and just use files/db for system components, services etc Jan 24 10:43:48 So phone should be in configuration/db like browser Jan 24 10:44:14 Ok, I'll do a PR; but first I'll check with my build Jan 24 10:44:17 I have another question: do we need to call putKind at some point ? Jan 24 10:44:49 Tofe: AFAIK not Jan 24 10:45:08 ok Jan 24 10:45:52 configurator scans on boot: Jan 23 15:37:17 qemux86 configurator[425]: [] [pmlog] Configurator CONFIGURATOR_WARNING {"directory":"/etc/palm/db_kinds","parent":""} Failed to open directory: /etc/palm/db_kinds, under Jan 24 10:45:57 And will create them Jan 24 10:46:04 At least that's what I assume Jan 24 10:46:42 ok, so just a reboot and then it's fine Jan 24 10:46:43 Yup ;) "Configurator is a userspace service that creates the database schema (webOS Db8 kinds) for Applications. It scans the /etc/palm/db/kinds directory to read the App's Db8 schema files to create the db kinds in the Db8." Jan 24 10:46:57 https://github.com/openwebos/configurator Jan 24 10:47:31 good Jan 24 10:49:44 A lot of things in webOS just make sense and work automagically :) Jan 24 10:56:07 Herrie|2: there must still be something missing in the recipe; I'll compare with browser Jan 24 11:02:55 Herrie|2: any special reason why the appinfo isn't at the root of the project? Jan 24 11:10:32 Tofe: For Phone you mean? https://github.com/webOS-ports/org.webosports.app.phone/blob/master/CMakeLists.txt#l18 Jan 24 11:10:42 Herrie|2: I think I've fixed it Jan 24 11:11:01 In Browser it's done differently: https://github.com/webOS-ports/org.webosports.app.browser/blob/master-webengine/CMakeLists.txt#L24 Jan 24 11:11:07 Just different approaches to the same Jan 24 11:11:13 Though both from morphis i suspect :P Jan 24 11:11:47 no, not the same, because when we tell bb that the phone is an app, then it expects appinfo to be at the root Jan 24 11:11:54 you'll see with my coming two PRs Jan 24 11:12:43 Tofe: Well CMakelist.txt seems to take care of the appinfo.json for both apps Jan 24 11:12:56 Maybe the bb will just continue even when there's no appinfo.json? Jan 24 11:15:04 Anyway I'll be away for a bit, doing weekly grocery shopping with the little one ;) Jan 24 11:16:56 Ok Jan 24 11:17:09 Two PRs done, so that you'll see what I was talking about Jan 24 11:17:36 No hurry to merge, I've got my own build Jan 24 11:18:08 So build is working? Jan 24 11:18:21 yup, with the two of them Jan 24 11:18:36 _and_ the db things are now in the package Jan 24 11:19:05 (thus the need for inherit webos_application) Jan 24 11:19:20 (thus the need for having appinfo.json at the root) Jan 24 11:19:27 OK :) Jan 24 11:19:34 Makes sense to have consistency across apps Jan 24 11:19:43 Like a true Dutch we'll go by bike :P http://www.hetzwartefietsenplan.nl/6050-large/urban-arrow-tent.jpg Jan 24 11:20:17 Don't forget your 8 layers of clothes! Jan 24 11:21:19 LOL ;) Jan 24 11:21:21 It's +10 ;) Jan 24 11:21:48 Ah wait +7 feels like +3 ;) Jan 24 11:21:51 Nothing too bad :D Jan 24 11:22:00 Little one is covered completely, no wind nothing :D Jan 24 11:22:07 And enough space for groceries LOL :P Jan 24 11:22:15 oh, positive, I didn't know it had warmed up to much Jan 24 11:23:45 It was around 0 for 2 days LOL Jan 24 12:30:28 mmh is there some db8 expert around here? Jan 24 12:32:42 I'm trying to mimic the bahavior of Legacy's phone app, which is to force the _id of the created callgroup object, but I'm being rebuked by db8, saying my _id is invalid (too long)... I think mojodb is correct, as my _id isn't a 96-bit base64 id like it would like to have, but how did Legacy succeed ? Jan 24 12:33:33 Looking at a Doctor image, it looks like it simply does a "put", like me Jan 24 12:41:36 Ok, I think it simply got restricted by this commit https://github.com/openwebos/db8/commit/080a3bc9fcea2c913faf219735f62de082d7979c which means Legacy phone app doesn't even run on OWO... Jan 24 12:44:33 I'll have to make up another solution, I guess. Jan 24 13:08:05 Tofe: feniksa is DB8 expert when around Jan 24 13:09:37 Tofe: We have forked DB8 but currently still using upstream, we can switch and use our own if needed Jan 24 13:09:44 If you want to modify DB8 Jan 24 13:10:52 No, I don't :) Jan 24 13:11:34 Tofe: OK ;) Jan 24 13:11:46 WEll you can modify the put in some kind to make sure it works or... ? Jan 24 13:11:47 The restrictions they've put on the _id seems completely ligit to me; it's the way the legacy phone app tinkers with the db that is edgy Jan 24 13:11:58 Tofe: Ah OK Jan 24 13:12:05 Might be the 3.0.x code is different Jan 24 13:12:09 Did you check that or 2.x? Jan 24 13:12:13 I'll just use another property as the index Jan 24 13:12:22 it was 2.x Jan 24 13:12:26 Ah ok :) Jan 24 13:28:39 OK little one is asleep :) Jan 24 13:28:51 Lunch cooking so got some time to look into things now as well :) Jan 24 15:00:51 nizovn: ping Jan 24 15:05:11 Herrie|Laptop: pong Jan 24 15:16:23 Herrie|Laptop: I have now something that works a bit, but there are issues in underlying layers (voicecall maybe) preventing it to work well Jan 24 15:16:39 for instance, the status of the ended voicecall is quite wrong Jan 24 15:16:58 Maybe it would be time to bump all the stack? Jan 24 15:20:25 nizovn: Have you seen the comments from elvispre about the BT scanning? Jan 24 15:22:22 yes, they sound possible, i will debug plugin a bit later Jan 24 15:22:42 Herrie|2: https://github.com/webOS-ports/org.webosports.app.phone/pull/26 tested on device, though still bugged Jan 24 15:24:07 Tofe: Yeah why not Jan 24 15:24:21 Can you fix the audio at least before upgrading? Jan 24 15:26:11 nizovn: Latest nightly should have the BT work from elvispre in Settings :) Jan 24 15:26:31 ah ! I knew I had forgotten something! Jan 24 15:27:15 Herrie|Laptop: No, I don't think it does. You merged the icons stuff to master but Bluetooth is on its own branch. Jan 24 15:28:04 Herrie|Laptop: I have just merged those together a minute ago if you want another PR. Jan 24 15:28:39 Quite sure I merged it already: https://github.com/webOS-ports/org.webosports.app.settings/pull/61 Jan 24 15:29:29 That's Bluetooth to Bluetooth branch. Isn't nightly building the master branch? Jan 24 15:29:59 Ah yeah Jan 24 15:30:09 You're right Jan 24 15:31:09 Herrie|Laptop: Top of my Bluetooth branch is now merged with master. I can do a PR to master from there if that's what you wanted already? Jan 24 15:31:55 elvispre: It's pretty separate from rest of Settings, so why not? Jan 24 15:32:02 Will you add an icon too? Jan 24 15:32:09 Done :-) Jan 24 15:32:24 OK :D Jan 24 15:34:17 Herrie|Laptop: Should have a PR now. Jan 24 15:42:46 elvispre: merged Jan 24 15:46:23 Herrie|Laptop: I think I've discovered an bug between Db8Model and LunaService Jan 24 15:59:31 Tofe: Oh joy :D Jan 24 16:59:02 * elvispre Has found a way to get Bluetooth properties out of navigator.BluetoothManager and into the Settings app. Jan 24 17:08:16 Herrie|2: my LS2 call for setCallMode doesn't work, but says it works... Jan 24 17:08:24 but it doesn't. Jan 24 17:12:14 https://github.com/webOS-ports/luneos-components/pull/30 <-- tested on N4, the Db8Model works, and I get no more errors on the logs. No regression for other apps, it seems. Jan 24 17:13:50 Herrie|2: does this look wrong ? https://github.com/webOS-ports/org.webosports.app.phone/commit/6649e50f2840888257c51e8035946a201a681bcc Jan 24 17:14:20 I'm always going into the success callback, I checked Jan 24 17:40:23 Tofe what does the !! do? Never saw this before LOL Jan 24 17:56:28 Tofe: I see nothing wrong, but I'm not a big fan of the inline functions for success and failure. I prefer to have them seperate, just for easier reading Jan 24 17:56:39 It could be the audio service never returns a failure? Jan 24 18:08:47 the !! are there for being sure I'm passing a boolean Jan 24 18:11:34 As for the functions, I don't really need them, it's just for the logs Jan 24 18:26:31 Tofe when you have an empty function I think common practise is to use undefined instead? Jan 24 18:26:58 At least that's what I've seen often Jan 24 18:27:02 Not that it should matter Jan 24 18:34:57 Tofe: You might want to stringify as per: https://github.com/webOS-ports/org.webosports.app.browser/blob/master-webengine/qml/NavigationBar.qml#L573 Jan 24 18:35:04 That's how I did the calls with Browser Jan 24 18:35:31 I remember vaguely I had odd things happening otherwise sometimes Jan 24 18:37:04 nizovn: Any comments on this: https://github.com/webOS-ports/luna-webappmanager/pull/52/files Jan 24 18:45:06 looks good to me, a little cleaner api Jan 24 18:47:24 elvispre: just to clarify. that is what was causing problems with onpropertychanged? Jan 24 19:30:00 nizovn: Yes. With this patch, my onpropertychanged is invoked. Previously, there was a complaint that property changed was not handled at all. Jan 24 19:31:28 nizovn: I think it maybe should have been .arg(one).arg(two), but I couldn't get two args to work so I gave up on that. Jan 24 19:40:07 ok Jan 24 19:41:20 Regarding this bug: http://issues.webos-ports.org/issues/1055 Jan 24 19:42:38 What does it mean about adding rotation? From Chromium's device emulation, it looks like switching from landscape to portrait and back works just fine, so if it's not handling device orientation changes correctly, I'm thinking that may be more on the system side rather than on the app side. Jan 24 19:42:58 Herrie|2: interested, let me try the stringify thing Jan 24 19:54:33 Herrie|2: and now it doesn't even go in my callback :p Jan 24 19:56:37 I got it! I needed to use the private bus _and_ the stringify thing Jan 24 20:00:27 Tofe: Glad I was at least helpful for 50% of the solution :P Jan 24 20:01:15 PR is now updated with working setCallMode Jan 24 20:01:19 KyleMaas_Laptop: With rotation I mean to rotate the PDF document. It happens quite often to me that I get a PDF that's 180 degrees rotated (so up side down). We have device rotation, but when I would rotate the device 180 degrees, the PDF would do too ;) Jan 24 20:01:40 So the PDF should be able to rotate clock or counterclockwise without needing to rotate the device :D Jan 24 20:02:28 I know the PDF.js has an API for it :P Jan 24 20:02:38 I've seen it sometime when I was updating the library Jan 24 20:03:34 All in all, my PR is working pretty swiftly Jan 24 20:04:07 Tofe: :D Jan 24 20:04:15 KyleMaas1: ^ Jan 24 20:04:22 Sorry assumed you were still on your laptop nick :P Jan 24 20:06:19 KyleMaas1: https://mozilla.github.io/pdf.js/api/draft/shared_util.js.html#line909 Jan 24 20:07:14 Herrie|Laptop: we now have working favorites, "working" call history (but all the timestamps are wrong, set to jan. 1900), with almost 1:1 compatibility with legacy for the model Jan 24 20:07:42 And an ugly UI for these things. Jan 24 20:07:43 Tofe: Nice :) Let me merge all the good work from everyone and kick off some nightly ;) Jan 24 20:07:55 Tofe: UI we'll worry about later, that's "minor" :P Jan 24 20:08:10 Next thing to do is certainly to update our voicecall stack Jan 24 20:08:17 elvispre: I assume we need an updated Settings to go with luna-webappmanager? Jan 24 20:09:10 Tofe: voicecall itself should be fairly straight forward hopefully Jan 24 20:09:23 Just hope the telepathy changes don't get into the way Jan 24 20:10:00 Herrie|Laptop: Not entirely essential as it did not work anyway. I do have a matching version that does something with a received Property though. PR? Jan 24 20:10:16 elvispre: Why not? Jan 24 20:11:58 KyleMaas_Laptop: Hi ;) Jan 24 20:12:01 Herrie|Laptop: https://github.com/webOS-ports/org.webosports.app.settings/pull/64 Jan 24 20:12:03 Did you see my comments? Jan 24 20:13:07 Herrie|Laptop: Was just reading through the log, and yeah, I understand now. Jan 24 20:13:28 It would be a matter of adding the icon, and attaching the action to it Jan 24 20:14:01 Well probably 2 icons, 1 for clockwise and 1 for counterclockwise Jan 24 20:14:16 Single click would do 90 degrees rotation would make most sense I guess Jan 24 20:35:00 Herrie|Laptop: https://github.com/webOS-ports/org.webosports.app.pdf/pull/14 Jan 24 20:35:34 The phone app do_install is killing the build for me at the moment. No such file or directory as files/db8/kinds/*. Have we got mismatched REVNOs in the recipes? Jan 24 20:36:04 elvispre: hmmm Jan 24 20:36:07 let me check Jan 24 20:36:11 I used Unicode characters for the rotation arrows, so whether they show up or not will depend on font support. That PR will need to be tested on legacy and LuneOS to confirm that those characters are present. Jan 24 20:36:11 Tofe changed something today Jan 24 20:36:25 KyleMaas_Laptop: OK Jan 24 20:39:15 elvispre: Strange Jan 24 20:39:38 elvispre: They were moved as per: https://github.com/webOS-ports/org.webosports.app.phone/commit/2b5f59f525f0ab1174a58321fcf39c772c743828 Jan 24 20:39:58 Herrie|Laptop: It might just be me and the state of my sstate_cache. I'll try a cleaner build. Jan 24 20:40:19 (That'll shut me up for a few hours anyway :-) Jan 24 20:40:27 elvispre: And this should take care of it in the bb: https://github.com/webOS-ports/meta-webos-ports/commit/e6a3320c9b161bdeb5b2d5fc77823e58accf9b56 Jan 24 20:40:37 I'll kick off a new nightly anyway shorlty too Jan 24 20:40:43 So might be I'll run into the same Jan 24 20:40:48 Though Tofe said he tested it :P Jan 24 20:42:06 Don't worry. I should go and spend some time doing something else anyway. Jan 24 20:42:49 elvispre: Hehe ;) Jan 24 20:48:39 Another PDF app feature: https://github.com/webOS-ports/org.webosports.app.pdf/pull/15 Jan 24 20:49:31 Herrie|Laptop: The bug tracker could still use a few more app bugs/features. :) Jan 24 20:49:58 KyleMaas_Laptop: LOL ;) Yeah I need to think hard :P How is your C++? Jan 24 20:50:10 Or you prefer JS stuff? Jan 24 20:50:13 Rusty, but decent. Jan 24 20:50:25 I could use some help with the MediaIndexer I guess :P Jan 24 20:50:52 Used to love programming my Palm Zire 72 with OnboardC. :) Jan 24 20:51:30 Problem right now is that I still don't have a good dev environment and probably won't be able to download a nightly image for at least a few days yet. Jan 24 20:52:38 KyleMaas_Laptop: OK, anyway it's not much hurry ;) Jan 24 20:52:40 Herrie|Laptop: What's wrong with it? Jan 24 20:52:51 KyleMaas_Laptop: Well it lacks quite some things ;) Jan 24 20:53:08 https://github.com/webOS-ports/mediaindexer Jan 24 20:53:17 It currently only populates a few kinds and not all Jan 24 20:53:25 And it's still missing things like thumbnails Jan 24 20:53:39 It basically uses taglib underneath for extracting tags from files Jan 24 20:54:21 Oops, I missed smthg it seems Jan 24 20:54:24 let me read the log Jan 24 20:54:28 Wow. That is a bunch of nigh-undocumented code. Jan 24 20:55:08 KyleMaas_Laptop: Feel free to Doxygen it first :P Jan 24 20:55:16 Herrie|Laptop: you bumped the phone's recipe, right? Jan 24 20:55:34 Tofe: Yeah Jan 24 20:55:39 It could be elvispre sstate Jan 24 20:55:42 ok, so in that case, it should work Jan 24 20:57:27 Herrie|Laptop: Doxygen is great, but without some explanation of how this thing's supposed to work, it's not much more than a call graph. Jan 24 20:57:44 I assume MojoMediaDatabase is where the heavy lifting is done? Jan 24 20:57:47 KyleMaas_Laptop: Ehm yeah Jan 24 20:58:05 KyleMaas_Laptop: Yeah Jan 24 20:58:41 The MojoMediaObjectSerializer is for individual file types like audio, video, image Jan 24 20:59:03 The MojoMediaDatabase is for the collection kinds like album Jan 24 21:00:09 If it's helpful I can push some example outputs in the source code from legacy as comments in the source code so that will help a bit on what to build? Jan 24 21:00:31 We also have the legacy filenotifyd code available that was used Jan 24 21:00:42 Might help. I'm unfamiliar with db8. Jan 24 21:01:31 KyleMaas_Laptop: You don't really need to worry about db8. We have all the kinds already in place Jan 24 21:01:43 It's just making sure we populate the right data in the right place Jan 24 21:03:38 It would be helpful to know how it works. Without that, the code looks like it work work as-is. It's adding data from the MediaFile to the object. Jan 24 21:04:42 KyleMaas_Laptop: Yeah it works more or less, but some things are missing Jan 24 21:05:36 I.e. we only populate the com.palm.media.audio.file, com.palm.media.image.file, com.palm.media.image.album, com.palm.media.video.file and com.palm.media.misc.file Jan 24 21:05:43 And not even everything that should be there Jan 24 21:06:25 And no unit tests. Jan 24 21:06:28 We're missing com.palm.media.audio.album, com.palm.media.audio.artist and com.palm.media.audio.genre for example which should be generated based on the data in com.palm.media.audio.file Jan 24 21:07:02 Similar like com.palm.media.image.album gets generated based on data in com.palm.media.image.file Jan 24 21:07:36 I can provide what legacy outputs and what we have nowadays. But there are some FIXME's in the code. When I add legacy output as comments that should provide enough info? Jan 24 21:07:56 And, see, that's where having an idea of how db8 works would be helpful. I'm not entirely sure what a "kind" is, for example. Jan 24 21:09:11 I don't think I ever actually developed a webOS app past Mojo. There's a lot of stuff in LuneOS which I'm still getting used to. Jan 24 21:10:05 KyleMaas_Laptop: DB8 has been there since Mojo ;) Jan 24 21:10:13 A kind is basically a table definition Jan 24 21:10:51 https://developer.lge.com/webOSTV/develop/web-app/app-developer-guide/db8/webos-db8-basics/ Jan 24 21:10:55 THat's a very short intro Jan 24 21:11:00 Ah. Never had to use db8. Jan 24 21:13:11 But in general everything for the kinds and permissions is in place ;) Jan 24 21:13:27 KyleMaas_Laptop: I ran into an issue with DB8 just now :P Jan 24 21:13:54 Appearantly while I was playing with MediaIndexer I somehow broke the mechanism to search for extension :P Jan 24 21:13:58 I'm not seeing where any properties are actually defined for mediaindexer. All I see are indices. Jan 24 21:14:00 So no PDF files would be returned Jan 24 21:14:46 The thing with DB8 is that it doesn't NEED to be a hard definition of a table ;) Jan 24 21:14:55 It can be free format more or less ;) Jan 24 21:15:06 Which makes it a bit hard to grasp at times Jan 24 21:15:52 In the example they gave it had a hard definition but it doesn't need to have it :P Jan 24 21:16:04 k Jan 24 21:16:48 So how does com.palm.media.audio.album not get populated, then? Jan 24 21:17:16 That's stored completely separately from com.palm.media.file? Jan 24 21:17:21 Er, audio.file Jan 24 21:17:30 KyleMaas_Laptop: Yes Jan 24 21:17:37 Because we don't do anything to populate it Jan 24 21:17:54 This mediascanner service basically scans /media/internal for changes and updates the various kinds Jan 24 21:18:19 Well, that explains a few of the bugs I remember running into with webOS's media player app... Jan 24 21:18:29 However there's nothing defined in the MojoMediaDatabase to do this for the com.palm.media.audio.album ;) Jan 24 21:18:34 Only for the image.album Jan 24 21:18:58 Why wouldn't you store than in one DB and run queries against it? Purely for legacy support reasons? Jan 24 21:19:29 And I take it that db8 does not support anything like an SQL view? Jan 24 21:20:06 KyleMaas_Laptop: I'm also no db8 expert ;) Jan 24 21:20:20 But yeah some "view" imho would make sense Jan 24 21:20:33 But I guess having it in a db8 kind might be quicker somehow Jan 24 21:20:50 This is the legacy code for the audio.file ;) https://bpaste.net/show/b2d173393bf0 Jan 24 21:22:40 That's what I basically replicated now in the MojoMediaObjectSerializer Jan 24 21:23:07 The "field"-order doesn't matter when you put the data Jan 24 21:25:02 Doesn't look like db8 supports anything like an SQL view. Jan 24 21:25:59 Does anything directly access the db8 data in the auxiliary kinds like com.palm.media.audio.album? Jan 24 21:26:17 I mean, I assume so, but just to verify that I understand the problem. Jan 24 21:27:49 KyleMaas_Laptop: Yeah apps like com.palm.app.musicplayer for example Jan 24 21:29:01 Boy, that seems at first glance like a very poor design choice. But oh well, if it's been that way for years, I suppose it will need to continue to be that way. Jan 24 21:30:22 KyleMaas_Laptop: I guess it was due to the db8 limitations ;) Jan 24 21:30:30 It's called a NoSQL database e:P Jan 24 21:30:32 :P Jan 24 21:31:40 Not the NoSQL part, the using-multiple-kinds-for-exactly-the-same-data part. Jan 24 21:33:32 So, if I understand this correctly, all you'd need to do is extend this: https://github.com/webOS-ports/mediaindexer/blob/master/src/MojoMediaDatabase.cc#L427 Jan 24 21:33:45 ...to store more data records for different kinds, right? Jan 24 21:35:45 And from what I can tell, this: https://github.com/webOS-ports/mediaindexer/blob/master/src/MojoMediaDatabase.cc#L465 Jan 24 21:36:19 ...only generates/stores a thumbnail for ImageMedia files, which means that album art would not be stored or queryable. Jan 24 21:36:44 Herrie|Laptop: Is that an accurate assessment of the current state of things? Jan 24 21:44:55 Well, I should probably get going. Jan 24 23:07:33 Tofe: Hmmm http://jenkins.nas-admin.org/job/luneos-testing_mako/lastBuild/consoleFull Jan 24 23:28:47 Tofe, Herrie, the documented way to get a DB8 id is to let DB8 assign it, by putting without an id. the return tells you what id was assigned. I don’t know if that helps, Jan 24 23:31:01 Herrie, you can take advantage of “function hoisting” in JS - function definitions are implicitly moved to the beginning of the surrounding function or file. So you can define functions (with names) after they are called. Jan 24 23:38:22 KyleMaas, DB8 is a NoSQL database. The API is documented at http://www.openwebosproject.org/docs/developer_reference/service_apis/db8/#.VqVf01L8-wY I’m not sure if that’s the info you need, though. A “kind” corresponds to a “class” in JavaScript. Tells you what fields to expect. But it’s fairly lax. Jan 24 23:40:40 KyleMaas, DB8 is NoSQL, so there are no views, no joins, and data is denormalized. **** ENDING LOGGING AT Mon Jan 25 02:59:58 2016