**** BEGIN LOGGING AT Mon Feb 16 02:59:59 2015 Feb 16 13:13:46 Herrie: ping Feb 16 13:14:29 morphis: ping Feb 16 13:14:45 Garfonso: pong Feb 16 13:15:33 morphis: c+dav activities never fire, because they have a requirement "internetConfidence": fair (or excellent). Feb 16 13:15:47 not sure who is meant to set that in activitymanager... Feb 16 13:15:49 oh Feb 16 13:15:59 it is up to the connection manager to provide that Feb 16 13:16:06 afaik webos-connman-adapter doesn't include that Feb 16 13:16:30 Garfonso: what is meant with "fair" or "excellent"? Feb 16 13:18:26 excellent is better internet connection than fair... but that is not 100% consistent over devices, it seems... for example: on a Pre3 excellent rarely happens. Don't know why... usually fair is the way to go and get a stable internet connection. Feb 16 13:18:54 I read somewhere that fair means internet connection is considered "stable" and excellent means it has some "high bandwidth", too. Feb 16 13:19:16 the requirement is added in the mojosync framework (like here: https://github.com/webOS-ports/mojoservice-frameworks/blob/webOS-ports/master/mojoservice.transport.sync/javascript/commands/synccommand.js#L298 ) Feb 16 13:19:30 morphis: pong Feb 16 13:19:34 so all synergy connectors in legacy used fair, from what I know Feb 16 13:19:51 Garfonso: what we can do for now is to provide "fair" by default Feb 16 13:20:33 yes, that would be good. :) Feb 16 13:22:58 Garfonso: no idea how we could measure different levels the best way Feb 16 13:25:47 morphis: yeah.. don't know either... also not sure what legacy did there. And as I said: this was never really specified anywhere. Feb 16 13:26:12 ok Feb 16 13:27:41 one heuristic might be to switch to excellent if wifi is connected or something. Feb 16 13:28:53 Garfonso: but that field seems to be only included if wifi is connected Feb 16 13:29:04 and we have it per wired, wan, wifi Feb 16 13:30:02 morphis: I'm around now Feb 16 13:30:29 Herrie|Veer: yeah saw that Feb 16 13:30:45 Herrie|Veer: there was somethign I wanted from you Feb 16 13:32:02 Garfonso: btw. you had time on the weekend for download manager or backup service? Feb 16 13:33:44 You want me to work on internationalisation stuff? QML set proper locales etc since we have that all available now? Feb 16 13:34:07 Date, time etc I can do now I guess in cardshell based on what we've set in FirstUse Feb 16 13:34:14 Herrie|Veer: not sure if QML sets the right locale already by default Feb 16 13:34:26 but you mean time etc., right? Feb 16 13:34:31 not translations? Feb 16 13:34:59 Time etc to start with Feb 16 13:35:05 good Feb 16 13:35:18 Herrie|Veer: for time there is also a preference key to respect Feb 16 13:35:23 For translations I can pull stuff from legacy so we have an idea what we need Feb 16 13:35:25 if we should 12/24 Feb 16 13:35:29 morphis: Yup Feb 16 13:35:33 Know that one Feb 16 13:35:45 good Feb 16 13:35:52 Herrie|Veer: let me write something down for translations Feb 16 13:40:20 Herrie|Veer: ah, I remember what I wanted from you Feb 16 13:40:31 the second clock in the dock mode screen is broken on the touchpad Feb 16 13:45:52 Ah let me check that Feb 16 13:45:58 I think I also noticed that Feb 16 13:46:14 It doesn't show the letters/numbers where it should right? Feb 16 13:46:20 ye Feb 16 13:46:25 s/ye/yes/ Feb 16 13:46:25 It's ok on N4 with me Feb 16 13:46:55 I guess I'll do some debugging when I got time and just implement different values for landscape and portrait Feb 16 13:47:01 That should solve it Feb 16 13:48:59 Herrie|Veer: it's only on the touchpad Feb 16 13:49:04 on the N4 its fine here too Feb 16 13:49:30 Yeah will have a look on TP Feb 16 13:49:37 Don't test much on TP lately tbh Feb 16 13:51:28 should be much better now with the memory leak fixed Feb 16 13:54:24 Is it properly gone now? Feb 16 13:57:05 completely Feb 16 14:44:31 Great :) You have a wifi or 4G tp? Feb 16 14:59:15 Herrie|Veer: wifi only Feb 16 15:06:23 OK, any idea how to improve the data connection stuff on TP4G? Feb 16 15:06:49 I think we might need to tweak config a bit still Feb 16 15:07:12 right Feb 16 15:07:23 we just figured out how but never really configured those bits Feb 16 15:11:31 Herrie|Veer: you remember what we did? Feb 16 15:16:49 Ehm I can check the logs? Feb 16 15:23:15 would be great Feb 16 15:32:10 FWIW, characterizing a network connection turns out to be more difficult than they initially though. The standards are moving toward just saying what technology is being used: http://w3c.github.io/netinfo/#h-the-connection-attribute Feb 16 15:33:40 Given the vagarties of wireless, in practice you only have two states: “I think theres’ a connecting” and “There is definitly no connection”. Feb 16 15:35:22 DougReeder: stating that there is a connection is easy but the more complex thing is to say how fast is this connection Feb 16 15:53:14 Let's download a 1MB file every now and then ;) Feb 16 15:57:19 A ping might be a better idea LOL :P Feb 16 16:51:12 Tofe: I have another other request for you regarding the PIN window Feb 16 16:51:33 I need a way to bring into the foreground when needed Feb 16 16:51:51 morphis: download/backup: not yet.. plans for weekend changed a bit... :-/ But I managed to do some c+dav bug fixing. Feb 16 16:52:06 Tofe: looks like http://doc.qt.io/qt-5/qwindow.html#raise can do that but we don't handle that in the cardshell yet if I am right Feb 16 16:52:12 Garfonso: :) Feb 16 16:52:41 Garfonso: good that you've found out what is wrong with the activities Feb 16 16:54:52 morphis: that is what I see on a TP4G and Veer for internet requirement: https://bpaste.net/show/a6de0c341947 (while having wifi and 3G connection) Feb 16 16:55:19 3G allone leads to internetConfidence: excellent, too. Feb 16 16:55:38 Garfonso: ok, I have a patch pending marking it as "fair" for all by default Feb 16 16:55:45 but I can raise that to execellent as well Feb 16 16:56:13 maybe just make it excellent.. ;) Feb 16 16:57:30 good Feb 16 16:59:56 there is still a bug that makes c+dav lose the google oauth keys after a while. But I have not found that one, yet... had some guy in US mailing me logs the whole weekend already. :-/ Feb 16 17:47:54 Tofe: can you rebase https://github.com/webOS-ports/luna-next-cardshell/pull/187? Feb 16 17:59:05 Garfonso, https://help.yahoo.com/kb/mail-for-desktop/SLN22041.html?impressions=true says “When you sync, only your Yahoo Contacts will be added to your device's address book. Contacts on your device won't be added to your Yahoo Contacts.” Feb 16 17:59:49 So it looks like we shouldn’t expect contacts created on device to upload to Yahoo (and that is what I observe). Feb 16 18:00:26 We should think about how to tweak the Contacts app to guide users. Feb 16 18:01:08 I could simply exclude Yahoo from the list of accounts in the Create New Contact pane. Feb 16 18:37:38 Herrie|Veer: ping Feb 16 18:40:21 We would probably also need to display a message when a Yahoo account is created. Feb 16 18:41:34 pong Feb 16 18:41:44 Herrie|Veer: calendar app still breaks: https://bpaste.net/show/19e21ae2ca96 Feb 16 18:41:55 DougReeder: sounds like a plan Feb 16 18:45:12 morphis: ok seems I need to check stuff on my linux vm Feb 16 18:46:23 Herrie|Veer: :) Feb 16 18:46:29 Herrie|Veer: ah and for the timezone selection Feb 16 18:46:37 we need to improve that a bit Feb 16 18:46:51 I am feeling really lost with this list having so much information and being unsorted Feb 16 18:47:38 Herrie|Veer: also I am missing the default where the timezone + time settings should be automatically set Feb 16 18:47:49 only when you don't want that a page with the timezones should open Feb 16 19:13:55 DougReeder, Garfonso: what is the best way to find a person per phone number? Feb 16 19:17:30 morphis: The list is quite OK Feb 16 19:17:33 I could add a search Feb 16 19:18:00 It's sorted by timezone basically Feb 16 19:18:05 3.0.x has a search added to it Feb 16 19:21:04 Herrie: can't we just do a simple Feb 16 19:21:06 City Feb 16 19:21:11 Zone Feb 16 19:21:14 listing Feb 16 19:21:17 so for example Feb 16 19:21:29 Wien Feb 16 19:21:32 GMT+1:00 Feb 16 19:21:37 much easier to read Feb 16 19:21:47 Ehm yes and no... Feb 16 19:21:57 WHen you have all african cities you have no clue where they are located Feb 16 19:22:01 and the user normally just looks for the city he is next too Feb 16 19:22:12 hm Feb 16 19:22:13 Yeah well... Let me work on it a bit Feb 16 19:22:24 This way it gives the most useful info imho Feb 16 19:22:28 Based on legacy as well Feb 16 19:22:38 hm Feb 16 19:22:44 We need at least 3 details, so we're pretty much stuck to 2 rows anyway Feb 16 19:22:54 Because it doesn't fit into a single on portriat Feb 16 19:23:06 Then I prefer to add a 4th value as well since we're anyway using 2 rows :P Feb 16 19:24:11 morphis, you mean, given a phone number, retrun a name for that phone number? Feb 16 19:24:45 In general, there could be more than one person associated with a given phone number. Feb 16 19:25:06 DougReeder: I know but I need something like that for the phone app Feb 16 19:25:15 and it just has the number of the incoming call Feb 16 19:25:59 Herrie: it's just that there are so many information one the whole screen so you get easily confused Feb 16 19:26:12 maybe adding a search and making the font size a bit larger helps Feb 16 19:26:12 I can rework it a bit Feb 16 19:26:24 Well making it larger makes the display problematic Feb 16 19:26:30 Already need to wrap for some countries ;) Feb 16 19:26:36 But I can try to tweak a bit Feb 16 19:27:29 To me it looks OK, but I'm happy to get feedback from others on it ;) Feb 16 19:27:34 DougReeder + Tofe have N4 as well? Feb 16 19:27:39 I can ask Benjamin as well Feb 16 19:28:21 I have an N4 Feb 16 19:29:43 I’ll poke around and see if people or contacts are indexed by phonenumber. Feb 16 19:33:47 DougReeder: thanks Feb 16 19:34:18 Herrie: yeah maybe he has a good idea Feb 16 19:35:02 Herrie: next builds will have a working PIN dialog Feb 16 19:35:11 no in firstuse yet Feb 16 19:35:24 Garfonso: next builds will also have the networkConfidenceLevel Feb 16 19:37:22 Herrie: ah btw. the enyo 2.5 update broke the wallpaper picker in settings Feb 16 19:38:06 morphis: Ah :s Feb 16 19:44:25 Tofe/DougReeder: This is the screenshot of what morphis meant with timezone clutter: http://tinypic.com/r/j9amqh/8 Feb 16 19:44:38 morphis: I could add a separator as well just like in legacy Feb 16 19:44:44 That might make it look more organized too Feb 16 19:45:21 Hmm, it does look a bit cluttered, but the solution is not obvious. Feb 16 19:46:45 Could the background color change for alternate rows? Feb 16 19:48:38 Not really Feb 16 19:48:45 I think a small divider might help already Feb 16 19:49:27 I think all the info needs to be there. Some modest tweak to the layout might help legibility. Feb 16 19:49:58 That's what I also think info wise Feb 16 19:50:06 Layout wise I'm not very strong :P Feb 16 19:50:12 So could tweak it a bit Feb 16 19:50:17 Will ask Benjamin for some input Feb 16 20:10:38 DougReeder: good idea Feb 16 20:11:20 Herrie|Veer: that is something we should really add to all lists in firstuse Feb 16 20:14:28 morphis, try luna-send -n 1 -a com.palm.app.contacts luna://com.palm.db/find '{"query": {"from": "com.palm.person:1", "where": [{"prop": "phoneNumbers.value", "op": "=", "val": "509-555-1212"}]}}' Feb 16 20:14:59 DougReeder: thanks a lot! Feb 16 20:15:15 There may be a better query. Feb 16 20:18:01 let me play with it Feb 16 20:20:35 There should be something in legacy? Feb 16 20:20:59 Herrie|2: for the person matching by phone number? Feb 16 20:21:01 A better query might be luna-send -n 1 -a com.palm.app.contacts luna://com.palm.db/find '{"query": {"from": "com.palm.person:1", "where": [{"prop": "phoneNumbers.normalizedValue", "op": "%", "val": "-2121555"}]}}' Feb 16 20:21:57 Note the value is reversed, with an extra dash - I’m not yet sure what the complet algorithm is. Feb 16 20:23:37 ok Feb 16 20:23:49 but something to start with Feb 16 20:24:36 What do you get from the phone lower layers? A phone number with area code? Feb 16 20:26:48 If you get 10 digits to seach on, query like this: luna-send -n 1 -a com.palm.app.contacts luna://com.palm.db/find '{"query": {"from": "com.palm.person:1", "where": [{"prop": "phoneNumbers.normalizedValue", "op": "%", "val": "-2121555-905"}]}}' Feb 16 20:27:49 If someone can tell me what phone numbers look like in various countries, I can update the test account I use. Feb 16 20:30:31 I’m reading http://en.wikipedia.org/wiki/Telephone_numbers_in_Germany but I’m not clear how a German would write down a phone number, not what the low-level phone makes available to higher layers. Feb 16 20:31:45 DougReeder: in germany you have the prefix which is either a mobile one or an area code Feb 16 20:32:13 In NL too Feb 16 20:32:30 +31-(0)6-123456789 Feb 16 20:32:33 = mobile Feb 16 20:32:41 yeah Feb 16 20:32:45 +31-(0)20-4300000 = fixed Feb 16 20:33:01 When you don't use +31 you use the 0 in front of 20 Feb 16 20:33:06 So 020 = area code Feb 16 20:33:10 I think something we need in the in the is https://github.com/googlei18n/libphonenumber Feb 16 20:33:19 Amsterdam = 020 Feb 16 20:33:23 Legacy had some mapping Feb 16 20:33:26 Let me find it Feb 16 20:33:38 We may well need that somewhere. Feb 16 20:34:12 For lookup, we might only need some last number of digits dialed. Feb 16 20:35:34 For example: /usr/palm/frameworks/globalization/version/1.0/formats/nl.area.json Feb 16 20:35:35 If Alice calls Bob, what will Bob’s phone display as the calling number? Feb 16 20:36:45 Which basically has a lot of JSON stuff: Feb 16 20:36:46 http://pastebin.com/SKSEYxqf Feb 16 20:36:56 "20":{"sn":"Amsterdam","ln":"Amsterdam"} Feb 16 20:37:30 http://git.shr-project.org/git/?p=libphone-utils.git;a=summary might some lightweight alternative Feb 16 20:40:37 Herrie|2: btw. do you have an idea when we want to let the LEDs we have under our control blink? Feb 16 20:46:00 morphis: Ehm blink when a new notification arrives Feb 16 20:46:08 Need to see what legacy actually does Feb 16 20:48:36 The key point for phone number lookup, is that % is a prefix operator. So, searching for -2121555 will find entries from any country or area code that end with 555-1212 Feb 16 20:49:46 http://www.webosnation.com/disable-gesture-area-led-when-app-active Feb 16 20:49:52 http://www.webosnation.com/led-notifications-get-customization-patches Feb 16 20:49:56 Those should give some ideas Feb 16 20:50:02 Card open keeps it white on Veer Feb 16 20:50:10 I created a patch to disable taht sometime Feb 16 20:50:26 So, we should probably search with area code. If that finds no results, we might search without area code (although that has the distinct possibility of false positives). Feb 16 20:50:26 Or rewrote a another webOS version patch ;) Feb 16 20:55:32 If we don’t find a match in the phone book, we need something like libphonenumber, to tell the user what region the call is from. Feb 16 21:07:36 +31-(0)6-123456789 is normalized as -9876543216--13-+ and +31-(0)20-4300000 is normalized as -000003402--13-+ Feb 16 21:20:52 DougReeder: is this normalize algorithm documented somewhere? Feb 16 21:21:15 You’d probably have to check the code in the contact linker. Feb 16 22:10:26 DougReeder: let me look Feb 16 23:25:21 morphis: messaging service needs to do this.. let me check. ;) Feb 16 23:27:06 morphis: ah.. messaging service uses a function in loadable-framework/contacts. https://github.com/webOS-ports/org.webosports.messaging/blob/master/service/javascript/assistants/AssignMessages.js#L97 Feb 16 23:27:23 DougReeder: ^^ Feb 16 23:33:08 DougReeder, morphis: the whole algrithm for "phone number normalization" is here https://github.com/webOS-ports/loadable-frameworks/blob/master/contacts/javascript/properties/PhoneNumber.js#L162 and it is awkward.. **** ENDING LOGGING AT Tue Feb 17 02:59:58 2015