**** BEGIN LOGGING AT Thu Jul 02 02:59:59 2015 Jul 02 02:59:59 snowkidind: nope. not at a computer at the moment Jul 02 03:00:22 doubt that would be available to a un rooted install Jul 02 03:03:42 huh. there's a recursive form... http://source.android.com/reference/com/android/tradefed/util/FileUtil.html#recursiveHardlink(java.io.File,%20java.io.File) Jul 02 03:09:39 this tutorial instructs to add an EditText variable, find the id and call .setText on it, but it only says to do so in the onCreate method http://www.raywenderlich.com/56109/make-first-android-app-part-2 Jul 02 03:10:11 but if i do that i can't access the variable from my onClick method Jul 02 03:10:23 the variables need to be global and outside the onCreate method right? Jul 02 03:11:08 erm, no Jul 02 03:12:27 james0r: He's assigning a field. Jul 02 03:12:50 That code doesn't look good, though, because he's not even using the `private` modifier. Jul 02 03:15:06 Actually, I take that back -- I was looking at one of the comments. Ctrl+F fail. Jul 02 03:15:17 declaring the variable right beneath my class declaration seems to work fine though. bad practice? Jul 02 03:16:06 Using a field (a member variable) is fine for this. Jul 02 03:18:29 TacticalJoke, k thx Jul 02 03:21:52 I take back my takeback. He really is forgoing the `private` modifier. Jul 02 03:22:36 `TextView mainTextView;` should be `private TextView mainTextView;`. It doesn't make sense to use package-private visibility (IOW, what you get with no modifier) without good reason. Jul 02 03:25:13 TacticalJoke, so 'TextView mainTextView' would become 'private TextView mainTextView' ? Jul 02 03:25:40 Yeah. That's the most sensible thing to do here. Jul 02 03:27:33 TacticalJoke, k. yeah i'm still pretty newb, but i wanna create good habits along the way. still not so sure what scope i want my variables to have in some instances. Jul 02 03:28:27 Generally, the best scope is the most limited one. Jul 02 03:30:41 TacticalJoke: Actually it makes sense if you use Dagger Jul 02 03:31:53 Yeah, but he's not doing that. All he's doing, from what I can see, is teaching n00bs bad habits. Jul 02 03:33:42 In part three of the tutorial, he even does it with a LayoutInflater: `LayoutInflater mInflater;` Jul 02 03:34:06 mAxflater Jul 02 03:45:31 hm, an open source mapping thing associated with samsung https://mapzen.com/ Jul 02 03:46:35 i guess they got some apple guys to work on it Jul 02 03:50:05 join us now and share the software ~ Jul 02 03:51:15 Anyone ever used google api sign in before? Jul 02 03:51:32 Having a weird issue where my onConnected method is called, but the bundle its being passed is null Jul 02 03:51:46 Weird because onConnected usually means success Jul 02 03:58:27 i have a JNINativeMethod and populated it with(javafuncchar, signiture, cvoidfuncpointer) i have 2 entries to the array how in java the libs are loaded and it calls the c native functions how does it now which index method_table[] to get? Jul 02 03:59:33 how == now Jul 02 04:02:50 does jni miracle happens with it Jul 02 04:03:01 should i ponder on how jni gets it Jul 02 04:05:38 i got it its only 1 way Jul 02 04:06:01 do we really benefit from native functions Jul 02 04:06:34 well i im doihng the onload material Jul 02 04:06:40 jni onload Jul 02 04:07:39 if you have an array of JNINativeMethods the only way to access it is through index logn Jul 02 04:08:06 now if it is huge n would be huge Jul 02 04:08:10 now Jul 02 04:08:27 please hear me out Jul 02 04:09:24 should i create my own data structure of JNINativeMethods to speed things up if i have a a large number of native functions Jul 02 04:13:23 should i leave it to find class Jul 02 04:13:35 registerNatives Jul 02 04:13:37 please Jul 02 04:47:23 Should the values that can be calculated (very simple mathematical) be stored in database for quick access? Would it matter except complicating database table. Jul 02 04:50:13 balance between file access and memory Jul 02 04:50:29 you dont want to be loading xml files and sprites at the same time Jul 02 04:52:13 sudhirkhanger usually i would leave derived data out of a database Jul 02 04:52:30 unless that data is given to another system, and you need to audit Jul 02 04:55:14 Yeah, it's basically. A listview and some extra info in detailed activity. And maybe sharing that extra info intents. Jul 02 04:58:26 damn it, there are too many good books to read :( Jul 02 05:03:34 hmm, is registerNatives worth doing? I've never bothered Jul 02 05:13:21 ""It turns out that automated vehicle technology — unlike humans — abides by the law. And that's bad news for local government revenues," lol - nm jobs, they are worried about speeding tickets Jul 02 05:18:26 dragorn http://www.wired.com/2015/07/online-anonymity-box-puts-mile-away-ip-address/ Jul 02 05:23:42 so much for red-light revenue Jul 02 05:26:31 money is money Jul 02 05:26:40 doesn't matter where it comes from Jul 02 05:26:51 it matters that it's gone Jul 02 05:27:19 whether from jobs or law abiding robots, it's all gonna be a money drain Jul 02 05:55:40 need some help with using android studio to commit a project to an existing respo like this https://github.com/mr-fool/Android Jul 02 05:55:55 I try right clicking the project name tab on android studio Jul 02 05:56:08 but the github options overwhlemed me Jul 02 05:56:48 hi my eclipse tools doesnt show android 5.0. builds Jul 02 05:56:56 how do i fix this Jul 02 05:57:03 horny-sama: Learn to do it from the command line, or use SourceTree. IDEA is really not great at git Jul 02 05:57:30 CedricBeust: I don't think window works well with command line either :P Jul 02 05:57:40 idea is fine at git when you know how git works (by using the commandline :) Jul 02 05:57:50 horny-sama: I use it every day Jul 02 05:57:58 was pushing things just 1/2 hr ago from the command line with git Jul 02 05:58:20 you can get msysgit, or at least my preferred cygwin Jul 02 05:58:56 yes, msysgit or mwing are both great, work right out of the box Jul 02 05:59:02 bash, git, ssh, everything Jul 02 05:59:44 isntalling msygit Jul 02 05:59:53 honestly I am not a fan of github Jul 02 06:00:46 cvs and 8tracks for life Jul 02 06:01:15 now that mysgit is installed Jul 02 06:01:19 how do I use it Jul 02 06:01:22 I don't mind GitHub, but Git is no fun compared to Mercurial. Jul 02 06:01:31 GitHub really needs Mercurial support. Jul 02 06:01:53 Mercurial is dead, no point really Jul 02 06:02:07 opened git gui Jul 02 06:02:10 It's not dead. It's in active development. Jul 02 06:02:13 it's not dead, but it's basically the same as git to use Jul 02 06:02:26 but I still want to push without dling the respo Jul 02 06:02:30 It's dead in use and adoption, it's just going to shrink Jul 02 06:02:32 ive used both, but can't really see any advantage to either Jul 02 06:02:38 It's as powerful as Git without the horrible UI. Jul 02 06:03:21 "ui" :D Jul 02 06:03:38 git's ui is horrendous but despite that, git is much more powerful and it won for that reason Jul 02 06:03:47 deebo: People often object to that term when talking about command-line UIs, but I don't understand why. Jul 02 06:03:58 It's a UI, no question Jul 02 06:04:07 CedricBeust: I don't think Git is more powerful. Jul 02 06:04:28 SCM wars Jul 02 06:04:59 anyone want to give me a hand on using git in conjunction with my android project that I created with android studio? Jul 02 06:05:19 horny-sama: Aren't those separate topics? Jul 02 06:05:21 There is as much a war between git and Mercurial as there is a war between humans and cockroaches Jul 02 06:05:33 CedricBeust: What can Git do that Mercurial can't? Jul 02 06:05:42 TacticalJoke: not really you need version control? Jul 02 06:05:47 while programming Jul 02 06:06:29 As far as I remember, there is a single thing that Git can do which Mercurial can't (and this is because the author of Mercurial thinks it's a terrible idea). Jul 02 06:06:39 There are probably a few things Mercurial can do that Git can't (e.g., changeset evolution). Jul 02 06:07:14 I mean I have created an android proejct using android studio now I want to upload it to git and I am stuck Jul 02 06:07:20 to existing respo Jul 02 06:07:53 horny-sama: The instructions on github are crystal clear, it's just four commands Jul 02 06:08:26 CedricBeust: url? Jul 02 06:08:40 github.com Jul 02 06:08:49 horny-sama: Don't think of this as an Android-specific thing, because it's not. It's just a question about pushing to GitHub. Jul 02 06:08:54 CedricBeust: might as well told me it is on google Jul 02 06:09:06 it'sl iterally one click away Jul 02 06:09:41 I love how so many simple questions turn into debates :D makes things interesting Jul 02 06:10:30 horny-sama: Your question is "How do I push a bunch of files to GitHub". The answer is very easy to find. Jul 02 06:10:36 I mean "a bunch of changesets". Jul 02 06:10:40 TacticalJoke: no Jul 02 06:11:08 my question is how to add a bunch of files without having to create the respo folder on your pc so my android studio can still work on it Jul 02 06:11:51 horny-sama: AS doesn't care, it will keep working on it whether there's a repo or not Jul 02 06:12:03 I know how to If you really want to combine, you'll need to go to your copy of that existing repo, or clone it new. Then move your new files in, git add them, commit them, push them. Jul 02 06:12:11 but that's not what I want Jul 02 06:12:32 CedricBeust: I need to move it to the respo folder which .... Jul 02 06:13:12 horny-sama: http://goo.gl/Ce46x8 Jul 02 06:14:02 CedricBeust: that's not what I wnat Jul 02 06:14:03 lol Jul 02 06:14:11 I know how to do that alreadyt Jul 02 06:18:37 I'm using AlarmManager and I have a question about how it sees differences in Intents Jul 02 06:19:15 I'm setting an long extra on the Intent. Does AlarmManager see it as a different alarm because of that or are all alarms with that same Intent action the same to it? Jul 02 06:35:26 Xscreensaver 5.33 out, Android coolness Jul 02 06:37:10 i have no opinion on git vs hg, but find it interesting that both google and fb have engineers working on it Jul 02 06:37:59 They're huge companies, they have engineers working on everything Jul 02 06:38:25 i guess one question i have - would repo still be neccessary with hg Jul 02 06:38:41 splitting up all the git repositories that way for android - its kinda a clusterfuck Jul 02 06:38:49 yeah Jul 02 06:39:27 We'll probably never know because it would take Mercurial five times as long to sync this amount of code :) Jul 02 06:42:27 Im confused whether I should use full height drawer menu or a clipped one in my social/ecommerce app... Jul 02 06:43:27 CedricBeust actually if i had to guess, its exactly that problem that hg might be better suited for - hence the work from google/fb. but i guess we'll have to wait and find out. we've been at this long enough to see version control systems come and go. git won't br around forever either Jul 02 06:43:56 CedricBeust: I'm not sure about that. Facebook made contributions to Mercurial that put it above Git in performance in certain areas (I think it was mainly cloning and working-directory-state calculation). Jul 02 06:44:09 Sure, but pretty confident 1) it's here for a few years and 2) it won't get outclassed by Mercurial Jul 02 06:44:54 I would say that it's been well and truly outclassed by Mercurial already, if we're talking features. It just doesn't have anywhere near the market share. Jul 02 06:47:00 I doubt Mercurial and Git will be considered state of the art for long, though. Something way better will come along in the not-too-distant future. Jul 02 06:47:30 Stuff like this looks kinda interesting: http://codicesoftware.blogspot.com/2015/07/towards-semantic-version-control.html Jul 02 06:47:33 Mercurial's obviously no good because 'h' and 'g' are the same finger on a dvorak keyboard. Jul 02 06:47:53 I don't see git going anywhere for at least five years but we'll see Jul 02 06:52:49 die git die :/ Jul 02 06:53:13 * capella but onlu cause I learned mercurial first ;) Jul 02 06:54:31 my mistake is git add --all instead of git projectName/* Jul 02 06:57:56 hey guys anyone knows how to get text from a notification without using the extras in api v19? Jul 02 07:13:14 Is there a way to make widget update onClock tick? Jul 02 07:16:25 if you enable the network permission in an update, do the users need to accept it? even though it is no longer something that is needed to be accepted on first install? Jul 02 07:31:24 anyone online? Jul 02 07:33:52 nope Jul 02 07:38:38 I should have just made a recepies app Jul 02 07:43:11 guys a best practice question Jul 02 07:44:15 game, with its instruction, what is the best way? 1. Fragment for the game, fragment for the instruction 2. one Fragment two layouts file 3. one fragment one layout Jul 02 07:47:20 icemanbp: what kind of game ? is it a game with lots of UI widgets ? Jul 02 07:47:32 no Jul 02 07:47:37 show some shapes Jul 02 07:47:54 instruction tells which shapes you have to touch Jul 02 07:47:57 you can use no fragments and just canvases Jul 02 07:48:02 and layouts Jul 02 07:48:15 you have to use canvas Jul 02 07:48:25 in the app there's different of this games Jul 02 07:49:03 there's also analytics for the results Jul 02 07:50:58 hi. I made a navigationdrawer with multiple other objects. and declare the navigationdrawer in the .main.java. However the application doesnt work and give error. Here are XML : http://pastebin.com/61c3UCJW JAVA: http://pastebin.com/t5G0c7DP and ERRORS : http://pastebin.com/Bz1ZRRmf Jul 02 07:53:07 Sorry the question is it doesn work as a slidemenu. it appears without me need to swipe it Jul 02 07:55:07 it is working now Jul 02 07:55:08 sorry Jul 02 07:57:48 Activity A starts B, B starts C, then I start B using Intent.FLAG_ACTIVITY_REORDER_TO_FRONT. Now when I finish B I want the resultCode to be passed back to A. How can I do this? onActivityResult does nto work since B was started by C. Jul 02 07:59:02 I also tried FOWARD_RESULT, but without luck. Jul 02 08:01:07 One option would be to just run the code in onResume, but I don't think that's a very good solution. Jul 02 08:02:29 I think that onActivityResult pass results to the caller Jul 02 08:04:08 barq: a lot of people would advice just using EventBus Jul 02 08:06:15 hey, what's the order android is picking the layout version? Jul 02 08:07:07 Is there a way to color TextClock (color the hour differently than the minutes) for widgets? Jul 02 08:09:37 anyone? Jul 02 08:18:27 is it possible to have your own font in a app? Jul 02 08:18:41 if you enable the network permission in an update, do the users need to accept it? (as didn't google play change it so network priv wasn't asked?) Jul 02 08:20:10 what do you mean joroci? Jul 02 08:21:21 i'll start from the beginning i only used default layout folder for my app, i found out i have an issue for andriod 4.2.2 rtl Jul 02 08:21:43 if i used right or left in the layout some views were strethced Jul 02 08:21:56 if you search with google Jul 02 08:22:11 there's priority on tag for resources Jul 02 08:22:36 i think i didn't search for the right term Jul 02 08:23:14 http://developer.android.com/guide/topics/resources/providing-resources.html Jul 02 08:23:27 its begins... http://arstechnica.com/business/2015/07/man-killed-by-a-factory-robot-in-germany/ Jul 02 08:24:22 http://stackoverflow.com/questions/17742600/android-values-priority Jul 02 08:24:25 "the robot grabbed the worker and crushed him against a metal plate." Jul 02 08:26:01 it was an accident Jul 02 08:26:05 no terminator Jul 02 08:26:12 "accident" Jul 02 08:26:15 reported by Sarah O'Connor no less Jul 02 08:26:39 i think it's terminiator 5 producer Jul 02 08:26:44 I don't see any catastrofic AI Jul 02 08:27:35 icemanbp i'm intersted only in the vXX priority Jul 02 08:27:41 i don't see it there Jul 02 08:27:59 ok, and what is the problem? Jul 02 08:28:11 lets say i have a v17 and a default folder Jul 02 08:28:14 when it goes to check the version take the right one Jul 02 08:28:17 ok Jul 02 08:28:29 i use lollipop which folder will be taken? Jul 02 08:28:34 if mobile is v17 goes in, otherwise it use default Jul 02 08:28:45 so in case of lollipop the default Jul 02 08:28:49 hmm Jul 02 08:29:03 becouse lollipop is v21 Jul 02 08:29:12 how about lower version than v17? Jul 02 08:29:17 default Jul 02 08:29:32 read the page, it explain how it works Jul 02 08:29:48 is the same as if I define layout and layout-land Jul 02 08:31:15 let me test because i think it's taking 17 Jul 02 08:32:35 joroci, http://developer.android.com/guide/topics/resources/providing-resources.html#BestMatch Jul 02 08:32:44 That explains the algorithm. Jul 02 08:33:42 thanks Jul 02 08:33:50 for version it'll use the most specific / closest one to the actual API level Jul 02 08:34:34 icedp: I think that's a bit of an overkill just for one use-case. Jul 02 08:34:46 so if everything will use v17? Jul 02 08:34:51 so everything will use v17? Jul 02 08:35:04 if that's the only specific folder Jul 02 08:36:12 joroci, everything API 17+ Jul 02 08:36:42 so if the device is API16 it will use the default? Jul 02 08:45:59 what happens with layout-land, layout-v17 and default if i have an api level18-land device? :P Jul 02 08:45:59 I use Intent.FLAG_ACTIVITY_REORDER_TO_FRONT to start an activity that is already somewhere in the stack. How can I get the resultCode back to the starting activity? Jul 02 08:46:56 danijoo, read my link. Jul 02 08:53:10 When using reorder to front, the resultCode is sent back to the Activity that sent the original intent to start it. How can I get the resultCode delivered to the activity taht used reordertofront to bringt it back to front? Jul 02 08:53:26 danijoo: just layout-land :D Jul 02 08:54:54 g00s: here? Jul 02 08:58:25 hi,im using fragmentstatepager adapter within a list view that every row has a view pager Jul 02 08:58:46 the first 3 items are created but when i scroll down, it gives array outof bounds exception Jul 02 08:59:28 sci-fic: that is a weird eceprience Jul 02 08:59:34 *experience Jul 02 08:59:39 post your code and error Jul 02 08:59:59 ok Jul 02 09:00:34 a listview with a pager in every row? Jul 02 09:00:45 is this even possible performance-wise? Jul 02 09:01:54 also I dont think it can recycle a viewpager properly (which might be the reason for your error) Jul 02 09:02:24 http://paste.ubuntu.com/11808974/ Jul 02 09:02:41 yeah i told that my boss too but he didnt care -.- Jul 02 09:02:57 stacktrace too please :) Jul 02 09:03:05 copying Jul 02 09:04:42 http://paste.ubuntu.com/11808983/ Jul 02 09:07:20 whats CustomViewPager Jul 02 09:08:21 just extended support viewpager and added a method to listener.page change listener returns current viewpager Jul 02 09:08:25 sci-fic: dude u need to call setAdapter inside your ViewHolder at each bindData call Jul 02 09:08:39 Solved, just used CLEAR_TOP instead of reorder to front. Jul 02 09:08:51 I have implemented a Viewpager inside recycler view. No probs Jul 02 09:09:17 I think he needs that to "recycle" the viewpageradapter Jul 02 09:09:30 reusing the old one with the new adapter of the row Jul 02 09:10:07 yeah the source code says that during notifydatasetChanged, it doesn't change the entire data of the view Jul 02 09:10:13 the code is optimized Jul 02 09:10:24 u need to reset the data in the viewpager using setAdapter Jul 02 09:10:32 after changing the data of the adapter Jul 02 09:11:19 let me try Jul 02 09:12:37 where does the 11 and 9 in the viewpager come from? Jul 02 09:13:09 from what I see every viewpager adapter has a size of 3. Jul 02 09:17:03 i guess,the first 3 lines of my listview was created succesfully,so it makes 3x3. whe i scroll down it tries to create next 3 ones Jul 02 09:17:18 but you are setting new adapters Jul 02 09:17:22 hmm Jul 02 09:17:50 but this would explain the values of the numbers. yeah Jul 02 09:18:33 have you tried using a FragmentPagerAdapter instead? 3 Fragments isnt what a StatePagerAdapter is made for Jul 02 09:19:24 not sure if statepager makes some weird caching internally Jul 02 09:19:37 actually i tried pager adapter,frag pager and state pager adapter :D Jul 02 09:19:42 u should use a pager adapter instead Jul 02 09:19:48 wheres the code of CustomViewPager.java? Jul 02 09:20:10 i copied support library view pager Jul 02 09:20:15 and changed only listener Jul 02 09:20:23 sci-fic: dude I have implemented the exact same thing using pagerAdapter. Jul 02 09:20:38 ok l'll use pager adapter Jul 02 09:20:42 brb 20mins Jul 02 09:20:57 if I have a container, and a limited number of fragments that get attached to it on pressing different buttons Jul 02 09:21:12 should I instantiate each one of them everytime? or store in mem Jul 02 09:21:20 any suggestions Jul 02 09:22:04 ashwink005, i'd add them to the backstack and on every click check if theres still a version of it available Jul 02 09:22:13 and reuse it if yes Jul 02 09:23:12 so in the fragmentTrasaction.add method I give the reference to the fragment already instantiated? Jul 02 09:23:20 that won't create any weird errors? Jul 02 09:23:31 ashwink005, no. you make a replace() with addToBackStack() Jul 02 09:23:41 and check the backstack each time Jul 02 09:23:43 Hey is anyone else seeing an error like this one from Picasso? http://hastebin.com/napikozepi.avrasm Jul 02 09:24:05 ok so a replace call then an addtobackstack call Jul 02 09:24:11 danijoo: right? Jul 02 09:24:11 We're seeing it in our crashlytics logs, it only seems to happen to users once Jul 02 09:25:55 ashwink005, http://stackoverflow.com/questions/19614300/how-to-save-and-reuse-same-instance-of-fragments Jul 02 09:26:17 Ankhwatcher, seems to be an issue in picasso Jul 02 09:26:23 i'd fill an issue in github Jul 02 09:33:54 danijoo: In that example the guy is instantiating new fragments each time causing an inflate each time Jul 02 09:34:45 danijoo: hat is the reason why I want to cache the fragments, to avoid inflation. Jul 02 09:35:58 ashwink005, yes. additionally you need to check the backstack if theres already a fragment of the one you want Jul 02 09:36:16 if yes, get that one from the backstack and use it in replace instead of a new one Jul 02 09:36:59 hmm nice.. so I'll inflate only when needed Jul 02 09:37:05 yep Jul 02 09:37:09 cool Thanks brah! Jul 02 09:51:07 how would you stop an observable from emitting for a specific amount of time? Jul 02 09:51:48 ie im doing a search and dont want to start another one until the first once completes (and also dont want next searches to queue up) Jul 02 09:52:46 i could set a boolean and filter everything out while search is running, but is there a more functional way? Jul 02 09:57:21 ah. there is a skipWhile() :) Jul 02 10:24:29 does the dropbox support streaming files (videos) TO dropbox? Jul 02 10:25:19 like i would start recording video from my camera to dropbox path (or using dropbox api) and in case of internet cut, the video would be still there Jul 02 10:25:53 on desktop dropbox just syncs a folder Jul 02 10:26:00 does it behave the same on android? Jul 02 10:26:04 then yes Jul 02 10:27:14 but it syncs new file when it is completed no? Jul 02 10:27:27 i think so yess Jul 02 10:28:02 or maybe theres another solution to my problem Jul 02 10:28:17 i want to make surveillance camera which comunicates with movement detector Jul 02 10:28:17 maybe, if you describe it.. :) Jul 02 10:28:41 if move occurs, it records 2.5 minutes video or longer if move continues Jul 02 10:29:15 i would like to stream it. in case of phone destroying to at least get the first part of video Jul 02 10:29:47 i dont want to have my own server for it Jul 02 10:31:45 no to mention streaming in android is pain Jul 02 10:32:32 i dont think you can do that with dropbox Jul 02 10:33:15 i thought so Jul 02 11:02:56 kroot, look at the topic, "no bots", you've got at least one here: http://logs.nslu2-linux.org/livelogs/ it's pretty easy to identify it by intersecting the channels where it logs - thank you Jul 02 11:16:42 Hello. Why is my play store listing showing wierd characters when I try to share on facebook: https://www.facebook.com/sharer/sharer.php?u=play.google.com/store/apps/details?id=com.instano.genie Jul 02 11:23:11 that looks indeed strange Jul 02 11:44:30 what... listview bounces back down if I give it enough speed to go up to the top -.- Jul 02 11:48:46 Zharf: Andoird bug Jul 02 11:48:49 Android* Jul 02 11:49:28 this annoys me a lot in reddit client. Jul 02 11:50:04 barq, is there a bug report? Jul 02 11:52:09 Zharf: There are/were a bunch of bugs related to fast scrolling. What api version are you using? Jul 02 11:52:18 22 Jul 02 11:52:33 or whatever is the latest, I forgot :p Jul 02 11:52:41 yeah 22 Jul 02 11:54:07 Zharf: See if this matches your behavior: https://code.google.com/p/android/issues/detail?id=159739 Jul 02 11:54:35 pretty much Jul 02 11:55:15 Zharf: Then feel free to add a comment to give the issue some attention. Jul 02 11:55:29 and star it Jul 02 11:55:33 that report indicates it's been fixed Jul 02 11:55:46 No Jul 02 11:55:49 oh no, I misread Jul 02 11:56:09 its fixed in aosp, but not published Jul 02 11:57:12 I'm kinda glad and sad that it's a bug Jul 02 11:57:26 https://i.imgur.com/bKGj4no.jpg Jul 02 12:03:39 Have you tested it in the M preview? Jul 02 12:05:07 me? Jul 02 12:05:59 No, Zharf Jul 02 12:06:11 How do you see it has been fixed in AOSP? Jul 02 12:07:11 I haven't tried M preview, no Jul 02 12:13:08 hey everyone Jul 02 12:13:18 does anyone know a good app for Android app regression testing? Jul 02 12:13:29 I would presume it's something you embed into your app. Jul 02 12:15:05 NetworkingPro: Define regression Jul 02 12:15:39 On what level do you want to test Jul 02 12:16:11 isnt that a specific kind of unit test? Jul 02 12:16:55 I thought it's more UI testing. Jul 02 12:17:10 As in manual clicking, but that can be automated Jul 02 12:17:17 So Espresso or UIAutomator. Jul 02 12:18:00 barq: bascically automated app testing. Something you can set to test all of your UI. Jul 02 12:18:03 barq: exactly. Jul 02 12:18:18 so are both of those pro level tools Jul 02 12:18:19 ? Jul 02 12:18:24 I dont mind paying for a good one. Jul 02 12:19:27 you can do that with what barq said, but its also possible to do it with the normal androidTests Jul 02 12:19:53 for automated testing you can use jenkins ci for example Jul 02 12:21:26 Thanks, guys! Jul 02 12:21:35 Does Jenkins CI actually do the clicking, danijoo ? Jul 02 12:21:52 barq, no. jenkins does the automatic testing for every build Jul 02 12:22:23 On code leve, so unit tests, right? Jul 02 12:22:34 Or you can launch Espresso/UiAutomator tests Jul 02 12:22:38 its a server that "listens" for changes in your repo. For every change, it runs the tests for you tells you if it worked Jul 02 12:22:40 So just manages the tools Jul 02 12:22:42 exactly Jul 02 12:22:53 So you need both, jenkins ci and one of the other tools. Jul 02 12:23:03 barq, you dont need tools to test your ui Jul 02 12:23:12 they just make it easier Jul 02 12:23:22 How do you test it then? Jul 02 12:23:28 with unit tests Jul 02 12:23:42 plain, boring android unit tests ;) Jul 02 12:23:53 Doesn't work if you do multi-apk projects Jul 02 12:24:06 why? Jul 02 12:25:37 what has multi apk to do with unit tests?> Jul 02 12:27:57 danijoo: robolectric had problems with multi apk, but unit tests should be ok, I mixed that up Jul 02 12:50:54 https://scontent-mxp1-1.xx.fbcdn.net/hphotos-xpa1/v/t1.0-9/11050309_10153705750826840_1020499227546389496_n.jpg?oh=c178d65f16eba2ba5ba7ad8b8b363836&oe=562513F4 Jul 02 12:54:09 hah Jul 02 12:54:25 good one Jul 02 13:23:23 what are multi apk projects Jul 02 13:31:17 in Rx, what use is the .subscribeOn(thread) call ? Jul 02 13:32:37 osxorgate, it makes everything "above" in the chain running on the thread Jul 02 13:33:18 for example: getSomethingFromDatabase().subscribeOn(Schedulers.io()).observerOn(AndroidSchedulers.mainThread()).subscribe(.....) Jul 02 13:33:29 eh, it means the side-effect that occurs from .subscribe() occurs on the subscribeOn thread Jul 02 13:33:42 not the side-effect of the Action Jul 02 13:34:20 danijoo: so in your example the fetching of data from the db is on the io() thread Jul 02 13:34:35 osxorgate, dont listen to me Jul 02 13:34:41 a good example is rxandroid WidgetObservable Jul 02 13:34:42 pfn can explain this way better :) Jul 02 13:35:25 WidgetObservable.text(textView).subscribeOn(mainThread).observeOn(whatever).subscribe(observer) Jul 02 13:35:57 the .setTextChangeListener that results from .subscribe() occurs on mainThread, but anything in observer runs on whatever Jul 02 13:36:30 i have a weird case where i do a networkcall (retrofit) and transform the result, and want to stash it in a view. But when I add a .delay() between the networkfetch and the .map() nothing gets sent into the view Jul 02 13:36:35 a good usecase is if you do a search on an edittext, use the query to get data from the server and show that in a list Jul 02 13:36:49 you would need to do the "middle" in an io thread Jul 02 13:36:58 so im thinking a thread is off somewhere Jul 02 13:37:26 osxorgate, your delay is probably occurring on the main thread Jul 02 13:37:45 and therefor subscribe to the edittext on mainthread, but call subscribeOn(iothread) for the server part and then observer on main again Jul 02 13:38:39 osxorgate, why would you delay the fetch -> map? Jul 02 13:38:51 pfn, because i want to simulate network lag Jul 02 13:38:55 but its probably not the right way Jul 02 13:39:11 i am trying to inject data into a view in a recyclerview Jul 02 13:39:29 so i make a subscription and when the view is recycled i unsubscribe Jul 02 13:40:15 in the meantime i have a ReplaySubject that i subscribe to, and i was hoping it would supply me with the transformed network data Jul 02 13:40:46 so in onBindViewHolder() i subscribe to the stream of data i want Jul 02 13:41:19 and if it's available, i would expect the onNext to fire immediately Jul 02 13:41:36 and if not, it will happen later (unless the view is recycled, then the subscription is unsubbed) Jul 02 13:43:17 Hello, I have a String resource kind of "This is the __company__ application" , this __company__ part is a something that I'll receive later in my app. I was thinking if there is some way to add String resources at runtime, than I would be able to represente the string as 'This is the @string/company_name message'. What do you suggest to do something like that. Or should I just replace this company part thought code every time that I would us Jul 02 13:43:17 e this string. Jul 02 13:43:43