**** BEGIN LOGGING AT Fri Sep 14 02:59:59 2012 Sep 14 03:00:06 what, that putExtra crashes? Sep 14 03:00:45 As soon as I move to the new activity it crashes Sep 14 03:00:52 I believe so yes Sep 14 03:01:08 there is no "believe so" - there is "in my logs, XXX was reported" Sep 14 03:01:26 Ok let me ask you this Sep 14 03:01:42 erm, no... go check your logs Sep 14 03:01:43 In my new activity, will this call the tag of the extra, getResQuestions = extras.getString("index"); Sep 14 03:03:50 and is that the line which crashes? Sep 14 03:04:10 Ok the issue is that the value I am trying to receive in my new activity is null Sep 14 03:04:23 Just did some quick logs and thats the issue Sep 14 03:04:41 pastebin the actual code and exception, please Sep 14 03:05:54 https://gist.github.com/a5a68700a06752834fbb Sep 14 03:06:14 I included the first activities intent creation and the second activities intent grabbing Sep 14 03:06:39 so resQuestions is an integer in the first activity, but you try to retrieve a String tagged as "index" in the second activity? Sep 14 03:07:06 I wanted to pass the integer Sep 14 03:07:20 I thought the first parameter was the identifier for the value Sep 14 03:08:40 it is Sep 14 03:08:53 let me make this really really simple for you Sep 14 03:08:56 So whats causing the issue then Sep 14 03:08:59 since hinting doesn't appear to work Sep 14 03:09:12 are you storing one type in the intent, then trying to retrieve a different type? Sep 14 03:09:21 I shouldnt be Sep 14 03:09:44 you are Sep 14 03:10:09 you're storing an integer and trying to retrieve a string Sep 14 03:10:19 integers and strings aren't the same thing Sep 14 03:10:21 So how do I receive the integer? Sep 14 03:10:44 you read the fucking manual Sep 14 03:10:46 http://developer.android.com/reference/android/content/Intent.html#getIntExtra(java.lang.String, int) Sep 14 03:11:16 getResQuestions = extras.getInt("index", resQuestions) + Sep 14 03:12:12 ah wow.... Sep 14 03:12:20 yeah... wow... documentation Sep 14 03:13:00 hi. I am trying to populate a gridview with a number of textviews. this is the code http://pastebin.com/ghvY6xbh . I am getting null pointer exception only when i add the onItemClickListener. Sep 14 03:13:19 so what should I do? Sep 14 03:15:03 i was able to display the grid without the clicklistener Sep 14 03:17:11 JakeWharton, you'll be happy to know that multi-select long-press thing is not an ABS bug. -> https://snipt.net/tweakt/-809/ Sep 14 03:17:34 I think there's some other step I need to do -or- it's a rather bad platform bug? Sep 14 03:19:22 yep, I need to setMultiChoiceModeListener() Sep 14 03:22:32 Is there a reason that the number for the variable i is not being displayed --- listItem.caption = "Index " +i; Sep 14 03:23:29 clipping? Sep 14 03:23:43 gonna need a little more context ;-) Sep 14 03:24:37 so can someone tell me whats up with the maps api pan feature? Sep 14 03:24:52 i can pan but its really irratic Sep 14 03:25:16 or it pans or zooms way to far if i tap the screen Sep 14 03:27:27 Has anyone used ActionMode / choice mode multi-modal? Sep 14 03:27:45 * mrenouf digs for sample code Sep 14 03:29:39 I have a listview in a fragment, and when the user clicks an item in the list i want it to trigger stuff in another fragment - should i send a callback to the parent Activity and let the activity decide what to do? Sep 14 03:31:31 ok... android-15/HoneycombGallery/src/com/example/android/hcgallery/ContentFragment.java seems to illustrate it's use Sep 14 03:32:03 speakingcode, that's best option Sep 14 03:33:08 cool just want to make sure i'm following best practice Sep 14 03:33:17 have the have the fragment(s) implement an interface that defines what you need to do Sep 14 03:33:35 gahh errors more.... Sep 14 03:33:44 then use those. onFragmentAttached() { if (fragment instanceof HandlesFooEvents) ... Sep 14 03:33:58 then that is a something the activity could publish events to Sep 14 03:34:30 if you want to get more sophisticated, look into EventBus Sep 14 03:34:41 still talking to me mrenouf ? Sep 14 03:34:56 speakingcode, basic publish/subscribe or observer pattern. (yes) Sep 14 03:35:01 yeah Sep 14 03:35:17 i am quite familiar with callbacks and observer pattern Sep 14 03:35:55 so that will work fine. thanks :-) Sep 14 03:37:54 question i have now tho is how do i trigger it from the onClick within the view returned by getView in my custom array adapter Sep 14 03:38:49 oh, call it on the context i guess? Sep 14 03:41:18 onCLick() { if (context instance of IMyFragmentClickHandler) { ((IMyFragmentClickHandler)context).clickCallbackFunction()/..... Sep 14 03:41:26 something like that i guess? Sep 14 03:47:19 speakingcode, yeah that's the idea Sep 14 03:48:35 speakingcode, relevant thread: https://groups.google.com/forum/?fromgroups=#!topic/android-developers/NBlMJnMaGbo Sep 14 03:52:41 thanks Sep 14 03:53:39 cool. pretty much how i imagined, time to make it Sep 14 03:54:12 hmm Sep 14 03:54:22 a fragment is a context yeah? Sep 14 03:55:05 oh, no, nm Sep 14 03:55:06 ok Sep 14 03:58:59 getActivity() Sep 14 04:01:27 heh, this google / acer / alibaba / aliyun thing is interesting Sep 14 04:04:40 if I needed to add a dropshadow image along the bottom of a linearlayout Sep 14 04:04:57 best way is probably to set its background to some Drawable defining a border or something right? Sep 14 04:05:10 "Aliyun is an OS created around the idea of pirating Android applications." oh, i get it now Sep 14 04:07:10 HorizonXP: right. like a 9-patch Sep 14 04:07:20 "Motorola CloudBB HMC3260" … haha wut the ffuuu Sep 14 04:13:31 g00s: not sure how i'd do that Sep 14 04:13:44 JakeWharton, the multi-model selection is actually part of AbsListView which talks to ActionBar. (API 11) Sep 14 04:14:30 would be awesome to have that, but it's not going to be transparent anyhow (would need a special SherlockListView, etc) Sep 14 04:15:27 the hard way is onItemLongPressListener -> getSherlockActivity().startActionMode(..) Sep 14 04:16:04 but not having the onItemCheckChanged (MultiChoiceModeListener) is a complicating things further Sep 14 04:18:06 how can i have a drawable PNG at the bottom edge of a view? Sep 14 04:27:39 i want to uninstall some apk using adb cmd, but i'm not sure what the exact name of the apk, is there any cmd like 'adb list apk' ? Sep 14 04:29:42 adb uninstall your.package Sep 14 04:30:48 if you don't know the package name, make sure the app is running, then go to settings/applications/running on your phone Sep 14 04:30:59 it should be there Sep 14 04:31:45 zhangxaochen, do "adb shell" then "cd /system/app" then "ls *apk" Sep 14 04:32:36 so I have this image: http://i.imgur.com/VMoea.png How can I have that along the bottom of a LinearLayout? Sep 14 04:32:50 so there is no single cmd that could do that? Ologn Sep 14 04:33:05 "adb shell pm list packages" will list all the installed packages on the system Sep 14 04:33:18 but under what circumstances wouldn't you know the name of your own package? Sep 14 04:33:45 might not be his Sep 14 04:34:10 add "-3" to the end to only list third-party (i.e. non-system) packages Sep 14 04:34:44 sampullman: I think it's a fair assumption that a developer would be working on their own package Sep 14 04:35:27 (oo-er, etc) Sep 14 04:35:31 Leeds, when eclipse is not on and i just forget the package name.. Sep 14 04:36:15 Leeds, what's the meaning of 'pm'? Sep 14 04:36:21 zhangxaochen: still, you don't know the name of your own package? Sep 14 04:36:33 yes i know Sep 14 04:36:40 (I call my package 'Bradford'... little Yorkshire humour there) Sep 14 04:37:06 but now curious about 'pm' Sep 14 04:37:26 pm... can be post meridian (i.e. after noon)... can be private message... can be project/program manager... or as in this case, is obviously package manager Sep 14 04:38:20 sorry for asking another stupid question.. Sep 14 04:38:41 i should have seen it myself Sep 14 04:41:59 aapt d badging <.apk> <-- will show details of an apk, including package Sep 14 04:43:07 mrenouf: so will "cat AndroidManifest.xml" Sep 14 04:43:33 or: head -10 AndroidManifest.xml Sep 14 04:43:34 I'm assuming he's got only the binary Sep 14 04:43:45 otherwise ... Sep 14 04:43:48 yeah Sep 14 04:43:48 assuming info is in the top ten lines, of course Sep 14 04:47:15 do i have to have eclipse to do a logcat? Sep 14 04:47:40 nope Sep 14 04:48:40 how would i do logcat without using eclipse? Sep 14 04:49:38 adb logcat Sep 14 04:49:41 or ddms Sep 14 04:50:01 what is better ddms or logcat or are they the same? Sep 14 04:50:23 adb logcat is just.. logcat Sep 14 04:50:28 ddms is logcat and a bunch of other stuff Sep 14 04:50:44 If you need "other stuff", ddms is better Sep 14 04:51:17 Cause im trying to get my phone to work with CM10 and i need to logcat or ddms the touchscreen Sep 14 04:51:47 #android-root Sep 14 04:52:02 oh thought i was in android-root sorry lol Sep 14 04:52:38 but thank you for helping with that Sep 14 04:58:30 Does anyone know why my (Fragment)PagerAdapter's getItem isn't being called again after destroyItem is called on the item (i.e. when the user flings back to a previously viewed page)? It's causing all sorts of trouble Sep 14 04:59:31 Because it doesn't actually destroy the fragment, it's still attached to the activity. It then finds it with findFragmentByTag, and skips getItem Sep 14 05:00:20 oh. that makes sense, thanks Sep 14 05:01:22 so if the fragments have large bitmaps I have to manually recycle/un-reference them if I don't want to trigger an oom error? Sep 14 05:04:59 If you're keeping references, yes Sep 14 05:06:43 great, totally solved my issue. thanks a bunch Sep 14 05:07:12 onDestroyView will be called after destroyItem is.. That would be a good place to do it Sep 14 05:08:34 seems to work in destroyItem, but I'll check that out too Sep 14 05:09:30 I just figured you had your bitmaps in your fragments Sep 14 05:10:02 does anyone here have any experience with the maps api? Sep 14 05:10:28 the panning on my mapiew is not working like it should Sep 14 05:12:11 They're not referenced in my code, but through myImageView.setImageBitmap Sep 14 05:16:38 It should be GCd along with the view then Sep 14 05:16:56 Also, don't keep references to views after onDestroyView is called Sep 14 05:18:03 Also, you say "large bitmaps" .. you should probably pre-scale your bitmaps before displaying them Sep 14 06:35:19 howdy. that's a lot of signoffs, heh Sep 14 06:38:11 SimonVT or anyone around? Sep 14 06:44:01 hi guys i am trying to add certificate to my android app. I have done it also but when i run my app some error is being thrown java.io.IOException: SSL handshake failure: I/O error during system call, Connection reset by peer.can anyone tell me what this error is all about . here is my question on stackOverflow http://stackoverflow.com/questions/12405902/adding-certificate-for-android-2-2.please help Sep 14 06:48:54 is your app connecting to a webservice or pulling data from something onlnie Android03 ? Sep 14 06:49:38 speakingcode : yes Sep 14 06:53:55 why there is still such error although there is an local.properties file with its content "sdk.dir=D:\\Android\\android-sdk" Sep 14 06:54:03 speakingcode : What happened Sep 14 06:54:04 error message: C:\cocos2d-2.0-x-202\cocos2dx\platform\android\java\build.xml:46: sdk.dir is missing. Make sure to generate local.properties using 'android update project' or to inject it through an env var Sep 14 06:54:05 ? Sep 14 06:54:43 does anyone have some ideas? Sep 14 06:54:51 thx in advance~ Sep 14 06:57:30 zhangxaochen: I'm not sure, but it seems like the sdk can't be found. are you sure the sdk directory is in your environment path? Sep 14 06:58:04 do i have to? Sep 14 06:58:12 do you use eclipse? Sep 14 06:58:18 or ant to build Sep 14 06:58:32 since there is already "sdk.dir=D:\\Android\\android-sdk" in local.properties... Sep 14 06:58:42 i use ant Sep 14 06:58:57 i'm not sure Sep 14 06:59:19 i have this line in my .bashrc file Sep 14 06:59:21 export PATH=$PATH:~/android-sdk-linux/platform-tools:~/android-sdk-linux/tools Sep 14 06:59:37 I assume I added it for a reason Sep 14 07:00:28 ok let me try Sep 14 07:00:58 I can't remember though..and I don't understand the build process well enough to know exactly how local.properties is used Sep 14 07:02:45 neither do i... Sep 14 07:03:41 hi guys i am trying to add certificate to my android app. I have done it also but when i run my app some error is being thrown java.io.IOException: SSL handshake failure: I/O error during system call, Connection reset by peer.can anyone tell me what this error is all about . here is my question on stackOverflow http://stackoverflow.com/questions/12405902/adding-certificate-for-android-2-2.please help it has alreadyt eaten up Sep 14 07:07:12 what has it eaten? Sep 14 07:07:33 sampullman, why when i open that project with eclipse, there is no android libraries in the project? Sep 14 07:08:04 i don't use eclipse, but i bet you have to import them into eclipse somehow Sep 14 07:08:22 did my solution not work? Sep 14 07:09:08 nvm, just realized that wouldn't have solved it at all Sep 14 07:09:19 still a good thing to have on your path, though Sep 14 07:10:30 hey guys. If i initialise a view everytime a method is called , will it increase the size of buffer and would not destroy that object ? Sep 14 07:11:03 imageView = (ImageView) view.findViewById(R.id.page) Sep 14 07:11:15 sunny_slls: you're going to have to explain that a bit better. Sep 14 07:11:28 suppose if this statement is defined in method Sep 14 07:11:36 and i call this method several times Sep 14 07:11:48 what would happen Sep 14 07:11:57 does it increase the size of buffer Sep 14 07:11:58 ? Sep 14 07:12:13 if you keep reference to the created objects, the GC will not "destroy" them Sep 14 07:12:23 if you do not keep refs, it will Sep 14 07:12:28 mostly Sep 14 07:12:33 its a local reference Sep 14 07:12:33 what do you mean by "buffer"? Sep 14 07:12:47 ImageView imageView = (ImageView) view.findViewById(R.id.fullpageAds) Sep 14 07:12:52 sampullman: means the cache Sep 14 07:13:47 Android03 sory Sep 14 07:14:20 hmm. i cant resolve R in eclipse Sep 14 07:14:32 i really hate eclipse sometimes Sep 14 07:14:36 if you attach the imageview to something, then it will not get GC'd until that thing is destroyed. otherwise, it will get GC'd some time after the method is finished Sep 14 07:14:41 speakingcode : never mind Sep 14 07:14:48 You will have no problems with it taking up more space sunny_slls. You will however waste a lot of cpu time on locating a view, when you could just store that as a member variable. Sep 14 07:14:58 k Sep 14 07:14:58 not sure what the cache has to do with this.. Sep 14 07:16:24 kingargaon: check the answer here http://stackoverflow.com/questions/885009/r-cannot-be-resolved-android-error Sep 14 07:16:40 kingargon: * Sep 14 07:16:44 speakingcode : do u have any clue about my problem Sep 14 07:16:44 ? Sep 14 07:18:46 sorry, no. i don't know much about certificates Sep 14 07:20:04 speakingcode: ok Sep 14 07:28:41 is there anyone having good knowledge of adding sslcertificates in android 2.2 devices Sep 14 07:28:41 ? Sep 14 07:31:52 Anyone know if the Android VPN API supports Suite B? Sep 14 07:34:00 hey everyone Sep 14 07:34:11 hi Sep 14 07:34:56 I wish to know whether some of the view animations equivalent to iphone are available Sep 14 07:35:42 http://pastebin.com/H6KhRJLP Sep 14 07:36:27 How to code view layout modifications & animations equivalent to this iphone code?? Sep 14 07:36:31 in andorid?? Sep 14 07:36:35 android* Sep 14 07:43:36 vanson2012: you might have a better chance at a response if you post videos/images Sep 14 07:45:17 #android Sep 14 07:46:08 hi. I tried the code in the guide here about gridview and it craches : http://developer.android.com/guide/topics/ui/layout/gridview.html?q=tables# . this is the second code I face with mistakes. Sep 14 07:47:06 what kind of crash? Sep 14 07:47:15 http://www.raywenderlich.com/2454/how-to-use-uiview-animation-tutorial Sep 14 07:47:36 ixc, NPE Sep 14 07:47:38 just like this effect of 2 images moving up & down respectively Sep 14 07:48:24 move x y, set width, set height Sep 14 07:48:25 superlinux-hp: and what is nullpointer? i think its not by guide :) Sep 14 07:48:41 and settings animations for the movement too Sep 14 07:48:54 how to achieve the same in android? Sep 14 07:49:20 ixc, if I remove the click event from the code, everything is fine. Sep 14 07:49:30 vanson2012: this might be a good start: http://developer.android.com/guide/topics/graphics/view-animation.html Sep 14 07:49:39 i can display the images. Sep 14 07:49:52 superlinux-hp: u replace HelloGridView.this for YourActivityName.this? Sep 14 07:49:58 probably not as easy as it is on iphone, unfortunately Sep 14 07:50:22 oh my gawd, hard coded xml again Sep 14 07:50:22 ixc, yes I did! Sep 14 07:50:34 and i think vogella.com also has nice animation guide Sep 14 07:50:43 superlinux-hp: which line has the npe? Sep 14 07:50:52 vanson2012: http://www.vogella.com/articles/AndroidAnimation/article.html Sep 14 07:51:49 vanson2012: anything that can be done in xml can be done programatically Sep 14 07:51:51 Hello. Could you explain me why localization might not work? I try to repeat example application ToDoList. I did two folders: values-fi and values-ru. Both locales (finnish and russian) failed to run in SDK and in Hardware Device. Here is the code: https://bitbucket.org/cyberorg/todolist/changeset/5a5310043e0d22697eb5d343730aa0c12e1c3f26 Sep 14 07:51:52 as far as i am aware Sep 14 07:52:25 superlinux-hp: looks only toast like can have npe, and it looks ok Sep 14 07:52:37 ixc, at the line where the onCreate() is Sep 14 07:52:39 oh god sampullman Sep 14 07:52:57 you must put your values-fi/ru in the res folder... Sep 14 07:53:11 superlinux-hp: pastebin your activity and logcat output pls Sep 14 07:53:21 ok Sep 14 07:53:58 sorry sampullman... Sep 14 07:54:01 wasn't you Sep 14 07:54:04 Ivru: i didn't say it's good to do everything programatically, what are you oh godding? Sep 14 07:54:06 oh Sep 14 07:54:13 can I "lock" a SurfaceView so that android.view.View.draw() isn't called until the surface got "unlocked"? Sep 14 07:54:19 X-Raimo: your values-fi/ru must be in the res folder Sep 14 07:54:21 ... Sep 14 07:55:10 rigid: to what purpose would that serve? Sep 14 07:55:41 sampullman: removing/adding a surface to a new viewgroup while the surface is alive Sep 14 07:55:50 thank you all I should stick on dynamic code-based android attempt Sep 14 07:56:30 you re right, android xml codes should be able to be expressed in java codes Sep 14 07:56:41 Ivru: Thanx to point it out. It helped. Exactly my fail. Sep 14 07:57:07 I facepalm'd a bit Sep 14 07:57:53 ixc, http://pastebin.com/ghvY6xbh this is the same thing but using TextViews. very same problem . and this the logcat http://pastebin.com/p3mZHbKa Sep 14 07:58:29 rigid: So it throws some sort of threading/memory access violation when you do it without "locking"? Sep 14 07:58:36 would maybe (View) view.getHolder().lockCanvas() work? Sep 14 07:58:36 I should think it would work as is.. Sep 14 07:58:48 sampullman: null pointer exception tbe, yes Sep 14 07:59:33 huh Sep 14 07:59:55 you could try the lockCanvas thing Sep 14 08:01:08 but im confused about the null pointer exception...what exactly is null? Sep 14 08:01:17 superlinux-hp: your variable is 'alphabet_gridview' and later you try to youse 'alphabet_gridView' Sep 14 08:01:38 java is case sensitive Sep 14 08:01:41 DAMMM Sep 14 08:02:02 mate I know already it's sensitivE Sep 14 08:02:55 ixc, see what name completion can do! Sep 14 08:03:04 it made me loose days! Sep 14 08:04:03 rigid: i don't see why you can't move a surfaceview to a new viewgroup, i bet it's something else Sep 14 08:05:03 sampullman: actually it's not a ViewGroup where I get the exception, it's the WindowManager... Sep 14 08:05:33 ixc, THAAANKS! Sep 14 08:05:42 ...the NullPointerException gets thrown at android.view.SurfaceView.updateWindow(SurfaceView.java:479) Sep 14 08:06:05 rigid: maybe you pass a null variable? Sep 14 08:06:18 ixc: i doubt it Sep 14 08:07:55 hey, ListFragment.setListShown(false); gives me an indeterminate progress indicator with the text "Loading..." Sep 14 08:08:03 is it possible to make the java code and the native code communicate in both ways? Sep 14 08:08:06 how do I change that text? Sep 14 08:08:23 setEmptyText doesn't seem to be what i'm looking for Sep 14 08:12:32 argh, i can't lock the canvas of the surface since it's SURFACE_TYPE_PUSH_BUFFERS >.< Sep 14 08:14:49 ixc, because of what just happened to me, how can disable naming suggestions while keeping name completion? Sep 14 08:15:04 how can i Sep 14 08:15:29 "Intel has confirmed that it will not provide support for Linux on its Clover Trail Atom chip." Sep 14 08:15:41 bah Sep 14 08:15:47 superlinux-hp: i dont know, i use vim Sep 14 08:15:55 ah! gosh! Sep 14 08:16:30 well I cannot cos of I cannot guess the names of methods and interfaces Sep 14 08:18:42 you can use 'auto completion' example with vjde Sep 14 08:18:53 does anyone know how I can set localLOGV to true in classes? Sep 14 08:19:16 what is locallogv? :) Sep 14 08:19:42 ixc: a widely used private boolean used in android Sep 14 08:19:53 when true, local logging is enabled Sep 14 08:22:29 rigid: http://stackoverflow.com/questions/2018263/android-logging/2019563#2019563 ? :) Sep 14 08:24:38 ixc: i only used adb logcat Sep 14 08:25:32 hey Sep 14 08:25:46 anyone here already sent push through GCM? Sep 14 08:26:46 rigid: oh sorry i just now check this locallogv a local variable that you setup in your class Sep 14 08:27:21 ixc: yes, it's used inside the android os Sep 14 08:27:30 i guess it's compiled in Sep 14 08:27:59 rigid: if i recall, there is nothing special about it. just convention. and if you look through the code, its used quite differently in many places Sep 14 08:28:13 i prefer to just use proguard to strip out log calls Sep 14 08:28:20 "This level will be logged, even on release builds, and is required to be surrounded by an if (LOCAL_LOG) or if (LOCAL_LOGD) block, where LOCAL_LOG[D] is defined in your class or subcomponent, so that there can exist a possibility to disable all such logging" Sep 14 08:28:25 http://source.android.com/source/code-style.html#log-sparingly Sep 14 08:28:55 g00s: nah, i want to enable the logging output from the OS and I can't find where the variable is written :-/ Sep 14 08:29:18 'enable logging output from the OS ?' Sep 14 08:30:01 nobody enables anything. they compile it one way with '1' for debugging and '0' for release Sep 14 08:30:06 in fact, its stupid as hell Sep 14 08:30:06 g00s: in android.view.SurfaceView there is quite some useful Log.v() output, which i'd like to enable Sep 14 08:30:19 oh dear Sep 14 08:30:27 ah, well, you just need to flip that bit then Sep 14 08:30:34 and then recompile Sep 14 08:30:39 and of to the races you go Sep 14 08:31:04 well, i thought there would be an easier way than recompiling the OS & hacking my phone :) Sep 14 08:31:14 bwahaha Sep 14 08:31:19 wth :D Sep 14 08:32:07 ah... "static private final boolean localLOGV = DEBUG ? true : Config.LOGV;" Sep 14 08:32:21 :D Sep 14 08:32:44 how does that help ? Sep 14 08:32:48 you have to change DEBUG Sep 14 08:33:14 yep Sep 14 08:33:44 When the user navigates to a certain website, I want to redirect input to my app, similar to wikipedia's app. http://pastebin.com/qbqgU51N is my manifest, but it's not triggering the intent. What am I doing wrong? Sep 14 08:37:40 freeone3000: for it you need to setup category too (fixme) Sep 14 08:45:47 Hello, is it a proper way to ensure data consistency during local sqlite database to remote database syncing, by generating and comparing Sha1 hashes? Sep 14 08:46:26 hi all Sep 14 08:46:41 can anyone help me out with a small monkey runner problem Sep 14 08:47:20 works fine when connected to my desire z but when connecting to an xperia i get a load of errors Sep 14 08:47:32 ixc: Added that. Still launching it in a browser. Sep 14 08:47:35 xperia t even not i Sep 14 08:47:53 freeone3000: what category(ies) you added? Sep 14 08:49:02 ixc: Sep 14 08:51:24 it is android.intent.ategory.BROWSABLE and add default too Sep 14 08:52:09 java.lang.NullPointerException: java.lang.NullPointerException is the error im getting Sep 14 08:52:31 but it works when connected to other phones Sep 14 08:52:44 ixc: Ah. Of course, thanks. Sep 14 08:57:35 what does BluetoothSocket.getRemoteDevice() return if the socket isn't connected? Sep 14 08:59:51 is readyfor4GB not compatible with intel i3 ? Sep 14 09:00:10 sorry .. wrong channel Sep 14 09:00:47 flam_: As the documentation states, an InputStream that doesn't work until a device is connected. Sep 14 09:02:32 uhm, the documention just states that it returns the the remove device, not InputStream: http://developer.android.com/reference/android/bluetooth/BluetoothSocket.html#getRemoteDevice%28%29 Sep 14 09:05:01 can I just use "LinearLayout l = new LinarLayout(this);" to have a "dummy layout" where I can "l.addView(myView)" later? Sep 14 09:05:56 i would "someView.addView(l);" while it's still empty but I get OutOfResourcesException creating surface Sep 14 09:08:38 never mind it was a driver issue, sorted now Sep 14 09:09:59 YEAH! i f***in' can! :) Sep 14 09:11:34 The documentation says "This method shouldn't be used for [...] elapsed time measurements, as changing the system time can affect the results." in System.currentTimeMillis(). Which method would be the right one then to measure elapsed times? Sep 14 09:18:31 does anyone here have 'kindle for mac' installed from the App Store ? Sep 14 09:18:39 (on their mac :P ) Sep 14 09:24:34 EPG: i guess there's a function that returns "seconds-since-last-boot" or something, use that Sep 14 09:26:21 thx rigid Sep 14 09:26:48 SystemClock.elapsedRealtime() returns the milliseconds since boot Sep 14 09:27:01 (including time spent in sleep) Sep 14 09:27:53 does anyone know whether android:alwaysDrawnWithCache will affect ViewGroups that hold a PUSH_BUFFER surface? Sep 14 09:35:44 i see that while recording a video using MediaRecorder every frame is logged with "V/QualcommCameraHardware( 9035): releaseRecordingFrame E" along with other frequent output... Isn't that a perfromance issue? Sep 14 09:36:39 Since videos recorder on that low-end phone like to contain lags/errors... it just works if nothing else CPU hungry runs on the phone Sep 14 09:36:57 so if there's an incoming call -> recorded video disrupted Sep 14 09:44:47 is it possible to 'hook onto' a handler using reflection or something? i really want to intercept/destroy messages Sep 14 10:49:54 hm, seems that to many messages in logcat drive java crazy Sep 14 10:50:37 *too Sep 14 10:51:22 EPG: that's why there's filtering Sep 14 10:53:21 https://code.google.com/p/android/issues/detail?id=4352 Sep 14 10:53:35 guess this is the same problem Sep 14 11:13:23 funktronic: filtering doesn't remove the massive performance impact appereantly Sep 14 11:13:39 ah yeah be careful with logging then Sep 14 11:14:05 running logcat in a separate shell (and closing it in eclipse) is fine Sep 14 11:14:45 i could live with that, now that i know where the strange problem originates :) Sep 14 11:15:12 i always run it in intellij it works fine Sep 14 11:16:18 EPG: eclipse has problems with "too much logging"? Sep 14 11:17:19 dunno if eclipse or somehow java Sep 14 11:17:41 i doubt java has problems with large loads of strings... Sep 14 11:18:14 then eclipse :) Sep 14 11:18:47 i don't know how to find the root of the problem/debug eclipse or something so i'll just stick to the external logcat for now Sep 14 11:19:16 and try intellij Sep 14 11:48:16 hi, i'm working on an opengl app for android 2.3, i'm searching the way to handle a movie inside the opengl context. does anyone can help me? Sep 14 12:08:21 hey guys Sep 14 12:10:32 i am using a Gallery with an ImageSwitcher. i get out of memory exception when switching from one GalleryItem to another. is there a way to recycle the Bitmap from the ImageSwitcher? Sep 14 12:26:16 hi, i'm using the ndk to compile libav. But my compile fails because ndk-build append an include path that I don't want. I checked all my LOCAL_C_INCLUDES and no one does include $(LOCAL_PATH). But when I do ndk-build -n to see the command line of the compilation. It's like it appends -I$(LOCAL_PATH) Sep 14 12:26:51 anyone already encountered this problem ? or know from where it can come Sep 14 12:40:03 why why WHY is AbsListView.performLongPress PACKAGE private?!?! Sep 14 12:40:07 grr Sep 14 12:40:40 * mrenouf is implementing choice action mode support for pre-Honeycomb Sep 14 12:41:14 mrenouf: if tit is package private, you can just create the same package hierarchy, put a dummy class there with just a public method forwarding to the AbsListView.performLongPress Sep 14 12:41:24 s/tit/it Sep 14 12:42:30 olivier_b, yep. it just means I can't do it right now dude to complicating factors with our build system, it needs to be in a library now Sep 14 12:43:08 s/dude/do Sep 14 12:43:35 * mrenouf suffers from hypo-caffeination typing syndrome Sep 14 12:44:21 hey, i'l looking for a solution to shut of the screen when the phone is on the ear. I get that event with the proximity sensor, but how do i shut off the screen like the telephony app does? Sep 14 12:44:27 basically it will need to land in ABS, and use one of Jake's _TrojanHorse classes. (love that name) Sep 14 12:47:04 zorzar, maybe try grabbing a partial wake-lock, and then call PowerManager.goToSleep(...) Sep 14 12:47:37 oh wait, nm. that method overrides wakelocks Sep 14 12:49:49 You can't do what olivier_b said with framework classes, when building against the sdk Sep 14 12:50:06 they are probably signed Sep 14 12:50:33 PowerManager -> public static native int setScreenState(boolean on) Sep 14 12:50:37 hmm Sep 14 12:51:13 should compile, it's just not documented (@hide) Sep 14 12:51:14 No, the methods just don't exist in the android.jar we build against Sep 14 12:51:26 oh :-( Sep 14 12:51:39 didn't knew that Sep 14 12:52:22 It's nothing more than a bunch of stubs to keep size down (and to stop people from using non-public methods I assume) Sep 14 12:52:31 yep Sep 14 12:53:23 wait, there should be a standard way to say... "if proximity sensor triggered -> turn off screen" Sep 14 12:53:46 mrenouf: i hope so :) Sep 14 12:53:56 seems generally useful to any app Sep 14 12:54:26 though, how many apps are used, active and in the forground, while on a call. I guess you must be making a VoiP app, am I right? Sep 14 12:54:37 yeah ;) Sep 14 12:54:43 ;-) Sep 14 12:55:21 zorzar, isnt there a way to hook into the dialer? Like how Google voice does it? Sep 14 12:55:35 that way the dialer would be the active activity Sep 14 12:55:43 er, the "phone" activity Sep 14 12:56:35 mrenouf: i don't think so, but i don't know all details of the framework :/ Sep 14 13:27:03 I'm having an issue with SQLite DBs, I'm using two, one which is a static users database and one which contains all my data, the login works fine and I've done a db.close() and a c.close (close curser) then on my next activity i have a function which calls the other database file, but it falls over and it appears to still be trying to use the users database, anyone got any hints on how i can fix this? Sep 14 13:27:59 i've posted the code on stack overflow (http://stackoverflow.com/questions/12424506/android-sqlite-using-wrong-database) if anyone can help Sep 14 13:36:56 I'm writing what I thought was going to be a simple app for work. I can't seem to wrap my head around Fragments. I have created ActionBar tabs using Fragments and that works. Now I want to put a Custom ListView inside of one of my Fragments. Can someone help me wrap my head around this. Sep 14 13:37:10 mrenouf: any idea= Sep 14 13:37:20 ? Sep 14 13:38:17 Squintz, just like using an activity. main thing is override onCreateView replaces onCreate() { setContentView(..) } Sep 14 13:38:29 and you inflate the content yourself using the provided inflater and return it Sep 14 13:38:48 mrenouf: how to either pass the voip call to the android phone app, or disable the screen Sep 14 13:38:53 as a shortcut though, you can use a ListFragment (fragment analog of ListActivity) Sep 14 13:39:20 zorzar, no, sorry. I would go digging around for more. or ask during office hours Sep 14 13:41:23 mrenouf, do you know of any examples that show this? Sep 14 13:42:35 Squintz, second google hit: http://stackoverflow.com/questions/7533934/android-listfragment-with-a-custom-view-hierarchy Sep 14 13:42:42 c'mon man ;-) Sep 14 13:44:52 public View onCreateView(LayoutInflater inflater, ViewGroup container, Bundle savedInstanceState) { return inflater.inflate(R.layout.custom_layout, container, false); } Sep 14 13:44:56 thats basically it Sep 14 14:00:21 hey gusy, i have a function that checks the Internet connection state, just wondering, what do you guys do if the user isn't connectd? Sep 14 14:00:41 Pop an alert? Close the app? Open the network connection settings ? Sep 14 14:02:42 Halt and Catch Fire Sep 14 14:03:15 most apps say 'oops no connection' and do nothing Sep 14 14:03:28 and let you retry Sep 14 14:03:39 Snuffel: so what would you prefer to see? Sep 14 14:03:40 yeah, it becomes a question of what SHOULD you do Sep 14 14:03:44 what does your app do? Sep 14 14:03:50 Does the app need to terminate if the network connection is losT? Sep 14 14:03:59 sounds terrible Sep 14 14:04:01 Can the user use other parts of the app? Sep 14 14:04:11 my facebook app seems to break my wifi sometimes Sep 14 14:04:12 lov: no, it's all pulled from a web API Sep 14 14:04:18 if I just wait a bit and then retry it works Sep 14 14:04:22 i wouldn't want to restart Sep 14 14:05:06 and if I want to go to the settings that's easily done Sep 14 14:05:09 HorizonXP: "it's" is what? Sep 14 14:05:15 is your app basically a webview? Sep 14 14:05:16 i wouldn't want my app to always go there for me Sep 14 14:05:46 lov: no, it's not a webview, but all the data for everything being displayed is pulled from our servers. Sep 14 14:06:16 lov: as in, sure, the app *could* work, but there would be no items displayed, completely blank. Sep 14 14:06:41 So put "no internet connection" as an empty message or something Sep 14 14:07:02 or make it an element in the view Sep 14 14:07:04 Make a toast if data is displayed, but further refreshes fail Sep 14 14:07:06 a huge element Sep 14 14:08:05 Just don't use dialogs ;O I hate dialogs Sep 14 14:08:14 ok, thanks guys Sep 14 14:08:22 i'll do either a Toast or an Item Sep 14 14:08:31 and just for SimonVT's version, I'll put a dialog Sep 14 14:08:32 and a retry button :) Sep 14 14:08:33 :-P Sep 14 14:08:39 Snuffel: yes, and retry button Sep 14 14:08:47 :) Sep 14 14:09:05 i am going to take a nap Sep 14 14:09:16 since our child is not at home today Sep 14 14:09:22 Enjoy Sep 14 14:09:35 she'll be home in 2 hours :) Sep 14 14:09:45 so it'll be a short nap Sep 14 14:12:28 mrenouf: office hours are a perfect hint thanks! Sep 14 14:34:27 where can I submit patch for ndk-build ? Sep 14 14:35:30 http://android-review.googlesource.com/ Sep 14 14:35:32 r.android.com probably Sep 14 14:35:44 thank you Sep 14 14:36:00 documentation is on http://source.android.com/ Sep 14 14:46:05 hey, take a look at this code http://pastebin.com/KTVbsiZW. and let me know what you guess about the reason my app Activity disappears on line 105 and the previous activity is brought to front Sep 14 14:48:33 alertDialog is null? Sep 14 14:49:22 joel135: look at line 22 Sep 14 14:50:02 d4gg3r, And there's no mentioning about it in your logcat? Sep 14 14:50:24 no Sep 14 14:50:33 propose a filter Sep 14 14:50:43 so that I see what you want from logcat Sep 14 14:51:25 what is the best way to get data to android? I'm creating an application where the application will have an id number, and i'm going to need to get a name and phone number back from the server, what would be the best protocol to use? Sep 14 14:52:19 i've used an http get previously for another application, but not sure if that is the best thing to use in this regard. Sep 14 14:52:28 d4gg3r, grep GameActivity Sep 14 14:52:29 joel135: when I am debugging, the line that makes the Activity disappear is 125, and looks like the previous activity is not brought to the front, it is actually replicated! Sep 14 14:54:33 Hi everyone! Is it possible to create custom device which will be connected through 3.5 Jack? Sep 14 14:55:10 joel135: look at this http://pastebin.com/9shsrRhx, activity timeout? :| Sep 14 14:58:14 boo1: you might want to look into the APK Sep 14 14:58:20 er, ADK Sep 14 14:58:27 You probably shouldn't keep the wakelock until onDestroy is called.. Sep 14 14:58:42 Leeds, thanks, so it's possible, right? Sep 14 14:58:55 boo1: ADK is for USB devices Sep 14 15:00:00 It's possible Sep 14 15:00:06 Thanks a lot) Sep 14 15:03:07 joel135: no idea? Sep 14 15:08:57 d4gg3r, no Sep 14 15:38:05 hello all Sep 14 15:38:49 can i use Intent.CATEGORY_ to add category to Intent for own use? Sep 14 15:39:04 yes Sep 14 15:39:28 and system garantees not adding category? Sep 14 15:40:22 on its source, depending on the circumstances Sep 14 16:04:10 does anyone have good example code how to add or remove a rawcontact to a contact using contactscontract? Sep 14 16:04:52 I'm currently doing this: http://pastie.org/private/dluo0gmuwwbezyobsfpha with (id1 is the rawid of the rawcontact i want to join/seperate, id2 is the first rawid in the contact in question) Sep 14 16:05:04 it randomly works sometimes, sometimes it doesn't Sep 14 16:05:16 that's not really satisfactory Sep 14 16:05:22 does anyone have a better solution? Sep 14 16:05:51 hey, I have just published my first app Sep 14 16:06:01 How long it takes to appear in play? Sep 14 16:06:03 torniker: congrats... I hope Sep 14 16:06:04 congratulations! Sep 14 16:06:06 JakeWharton, I've got a reasonably workable ListViewCompat put together that works with MultiChoice ActionMode Sep 14 16:06:11 a few hours to a day max Sep 14 16:06:20 anything from a few seconds to a few hours Sep 14 16:06:34 is it a paid app using widgets or accounts? Sep 14 16:06:40 if yes, have fun with jelly bean users Sep 14 16:06:44 Leeds, mots thank you :) I can't wait to see my app in play :) Sep 14 16:06:51 torniker: so what is it then? :) Sep 14 16:07:20 Leeds, it's just free app Sep 14 16:07:30 torniker, what does it do? :) Sep 14 16:07:32 Leeds, for watching tv on the phone Sep 14 16:07:36 oh Sep 14 16:07:41 that actually sounds awesome Sep 14 16:07:42 how? Sep 14 16:07:45 iptv? Sep 14 16:07:50 mots, over hls Sep 14 16:08:03 where does it stream from? Sep 14 16:08:12 from internet Sep 14 16:08:16 nice Sep 14 16:08:35 have some servers they encode channels to hls and you can receive it on phone Sep 14 16:08:45 so... pirate streams? Sep 14 16:09:05 no all the tv stations know it Sep 14 16:09:39 so it's coming straight from the TV stations' servers? Sep 14 16:09:45 I have a website I stream them on the website, and now I stream on mobile Sep 14 16:10:16 the website is? Sep 14 16:10:18 he sends a gift basket every holiday Sep 14 16:10:20 so its cool with them Sep 14 16:10:20 myvideo.ge Sep 14 16:10:43 yes I only stream channels who gives me permission to stream Sep 14 16:11:09 mostly public channels Sep 14 16:11:50 ok... well, good luck anyway :) Sep 14 16:12:32 Leeds: you don't happen to have lots of knowledge on contactscontract? Sep 14 16:12:42 mots: nope, never touched it Sep 14 16:12:48 I'm digging through the source of the people app right now, but it's a huge clusterfuck Sep 14 16:14:19 west coast coders probably not awake yet ;-) Sep 14 16:29:04 if not here, where can i advertise contract jobs? Sep 14 16:29:42 i wouldn't advertise here... Sep 14 16:29:59 maraz, right, it says so in the topic Sep 14 16:30:09 yes, well Sep 14 16:30:23 that's one thing Sep 14 16:31:20 the rule is a bit peculiar in the tech community, but i suppose if it were a problem... Sep 14 16:33:01 hi, I have this problem http://stackoverflow.com/questions/8450539/images-taken-with-action-image-capture-always-returns-1-for-exifinterface-tag-or does somebody have a solution? Sep 14 16:35:10 The accepted solution isn't good enough? Sep 14 16:36:25 SimonVT, no, I have tryed them all but the orientation returned is always 0 Sep 14 16:48:33 Ok guys quick question, I found some open source code online that will dynamically create some forms but I need some help with my labels. Sep 14 16:49:12 Basically I need some help so that I can increase the value of the label as it loops through each time, the issue is that where the text is declared, it is after my loop' Sep 14 16:50:15 pastebin Sep 14 16:50:55 https://gist.github.com/4ac867a24570a76ecb48 Sep 14 16:51:13 Line 48 is where the loop is that creates the forms and lines 86 is where the text is set Sep 14 16:51:22 Hey guys, bit of a "best practices" question for you: what is the typical way to set up properties of a DTO in Java? As an example, I do a REST API call to get a "User" json. I want to then make a User DTO in my Java code. Should I pass the JSON to the User constructor to set up the props, have a whole bunch of constructor args for each prop, make setters and do u = new User(); u.setName(..), do a hashmap param....? Again, Sep 14 16:51:23 looking to play nicely in the Java world, thanks in advance Sep 14 16:51:27 I need a way to incorporate line 86 inside my for loop Sep 14 16:52:01 hmm, I am liking learnopengles.com Sep 14 16:52:24 @snwspeckle why not global var and ++ Sep 14 16:53:14 wurde: The loop is dynamic and is adjusted based on values from the previous activity Sep 14 16:53:32 The issue is that the original code generates the text in the labels after my loop Sep 14 16:53:56 Should I just wrap everything in my getView which is used to create the objects? Sep 14 16:54:06 or not creates, but sets them Sep 14 16:54:13 i realize that it is an adapter. why not move the loop then Sep 14 16:54:40 Move the loop inside the getView? Sep 14 16:54:55 try it. Sep 14 16:55:13 and i dont fully understand why you would call notifydatasetchanged in an adapter. Sep 14 16:55:14 Should I move the whole MyAdapter? Sep 14 16:55:34 wurde: Like I said, the original code was from somewhere online Sep 14 16:55:46 I didnt know how to do this so I found some existing code Sep 14 16:56:12 ja, it seems questionable. Sep 14 16:57:30 I get this error with public MyAdapter: Syntax error on token "public", new expected Sep 14 16:57:50 I made a revision on gist if you care to see Sep 14 17:00:01 wurde1: see what I am saying? Sep 14 17:02:50 wurde: AFK mate? Sep 14 17:03:13 fakingfantastic: that's absolutely one approach Sep 14 17:03:58 fakingfantastic: You could also have your User class handle that json itself Sep 14 17:05:23 e.g. public User(String json) { parseJSON(json); } private void parseJSON(String json) { /* parse the json string here, handle assignment of members */ } Sep 14 17:06:43 lov: that's what I'm doing now, but i didn't know if that's an anti-pattern. Should the user be able to sniff out json Sep 14 17:09:28 fakingfantastic: The implementation is really totally up to you. Sep 14 17:09:41 I think the only poor design decision would be to hold onto the JSON string after you're done consuming it Sep 14 17:10:11 Having the User class handle consumption by itself means that subclasses of User can parse other tidbits that may be present Sep 14 17:10:24 but you could always have some sort of Factory method to parse the content and return the right kind of User Sep 14 17:10:33 s/Factory method/Factory class/ Sep 14 17:11:19 mrenouf|work: awesome. there's a ticket for ABS support of that feature Sep 14 17:11:25 what does it do on pre-ICS? Sep 14 17:20:49 Alright I need some help, I still cant get it to work :( Sep 14 17:22:41 wurde: Yea there? Sep 14 17:29:57 Hey guys I need some help making some dynamic TextViews Sep 14 17:31:34 Come on anyone :( Sep 14 17:34:37 Anyone using the Facebook Api? My app fails saying Facebook.R class is not found but the Facebook class is there. Sep 14 17:34:48 any help would be appreciated Sep 14 17:35:20 I am using Eclipse (STS) to do this. Sep 14 17:37:13 is the facebook R file present in gen? Sep 14 17:37:55 clean both project and build again? Sep 14 17:37:59 *projects Sep 14 17:39:17 lov: yes it is there, i can open the source Sep 14 17:39:52 f2prateek: did that - didn't help Sep 14 17:42:19 where is the import pointing to? Sep 14 17:45:32 Can anyone help me dynamically change this TextView, I am having issues... Sep 14 17:45:44 snw, what's up? Sep 14 17:46:27 One moment Sep 14 17:48:52 lasserix: Ok so I took the idea from yesterday and re-structured my code to create forms, the issue I am having right now is that the labels above each EditText field are not changing based on its status in the for loop Sep 14 17:49:23 Basically the label should say "Index 1", "Index 2" and it increments the # in each pass of the loop Sep 14 17:49:30 can you paste the code? Sep 14 17:49:59 https://gist.github.com/872eb94a320a988bb9d7 --- and brb going afk for a sec Sep 14 17:51:20 f2prateek_: These are my imports:import com.facebook.android.AsyncFacebookRunner; Sep 14 17:51:20 import com.facebook.android.DialogError; Sep 14 17:51:20 import com.facebook.android.Facebook; Sep 14 17:51:20 import com.facebook.android.Facebook.DialogListener; Sep 14 17:51:20 import com.facebook.android.FacebookError; Sep 14 17:52:10 there is no Facebook.R import Sep 14 17:53:18 Does anyone have an idea why my calendar (created in my application) is being deleted when I reboot my phone? (it is also being deleted after some random time, but always when I reboot) Sep 14 17:57:08 Snwspeckle: unless I'm going blind, I don't see anywhere in getView where you setText to indexText--you initialize it, and set a tag but don't actually setText... I don't understand what the loop function is doing s Sep 14 17:58:07 since the textview "dynamic" comes from the adapter binding the views Sep 14 18:01:58 f2prateek_: Its inside the call to fb.authorize() that it needs this class - looking for R.drawable Sep 14 18:03:07 looks like jsr310 (date & time) replacement will make it into java 8; i wonder if this can be backported to android Sep 14 18:03:27 f2prateek_: nvm I am going to start with an example and work from there - thanks Sep 14 18:08:04 lasserix: Yea there Sep 14 18:08:19 but that's not how the listview adapter works, Sep 14 18:08:31 you need to put your arraylist elements into some corrospondance with the listview elements Sep 14 18:08:40 Good afternoon all. I wonder if someone could take a quick look at my layout XML and give me an ice why each ID field throws an error today, when they didn't yesterday? http://pastebin.com/EGeGkxvT Sep 14 18:09:06 usually this is done on the basis of position, ie position 1 in the listview corrosponds to the list[1] element Sep 14 18:09:16 lasserix: I found most of this code online, I didnt know how to go about it so I am using this as my structure Sep 14 18:09:34 well forgot the loop the loop has nothing to do with populating the listview Sep 14 18:09:49 the adapter's getView method is what will set the view's text information Sep 14 18:10:12 Ok but how can I increment the label without being in the loop? Sep 14 18:10:31 you don't increment the label Sep 14 18:10:46 It should be noted that I had a technical failure so had to recover my layout by decompiling the latest version of my app, from the APK on my test phone. Would the resulting layout be any different once run through apktool? Sep 14 18:10:53 the getView method of the adapter will set the label, like i said, based on some corrospondance with your arraylist data Sep 14 18:11:03 Well everytime the label is in a new row it should be increased Sep 14 18:11:12 right Sep 14 18:11:24 @hillct: Nope Sep 14 18:11:30 So how do I make it so when it sets it, it will be 1 higher Sep 14 18:11:31 so you'd do something like indexText.setText(ArrayListData[position]) Sep 14 18:11:42 hillct: Remember to use @+id when defining new IDs Sep 14 18:11:46 before returning the convertView Sep 14 18:12:28 SimonVT: I made that change in one test and it made no difference. Will try it again in a new test project Sep 14 18:12:40 for now just trying adding the line indexText or holder.indexText.setText(position) so you see what's happening Sep 14 18:12:56 There's a difference Sep 14 18:12:58 or setText("Index : " + position) Sep 14 18:13:06 Unless you defined the id's elsewhere, you can't use @id/ Sep 14 18:13:10 Ok I made the change, heres a reference, https://gist.github.com/872eb94a320a988bb9d7 Sep 14 18:13:17 SimonVT: do you have an idea why my calendar (created in my application) is being deleted when I reboot my phone? (it is also being deleted after some random time, but always when I reboot) Sep 14 18:13:22 no Sep 14 18:13:46 Ok so right now what it does is that it sets the text after the loop Sep 14 18:13:51 so the label is not dynamic Sep 14 18:14:02 holder.indexText.setId(position); this is weird Sep 14 18:14:22 don't bother setting ids unless you need them later Sep 14 18:14:39 you won't need them later since you will use the position argument based in the onitemselected listener Sep 14 18:14:46 @Snwspeckle: it better to build you dataset outside the adapter.. Sep 14 18:15:00 how do i calculate the height of the status bar in a native activity, or otherwise get the native activity to make the parent surface sized so that it won't overlap? Sep 14 18:15:37 Snwspeckle, like I said FOR NOW just use this line holder.indexText.setText("Index " + position); instead of holder.indexText.setText("Index " +indexVal); Sep 14 18:15:43 to see how the getView method is working Sep 14 18:16:02 SimonVT, tofra: OK, this is screwy. On a lark, ran the file through the source-> format option in the context menu, and suddenly all errors disappeared Sep 14 18:16:20 alright one moment Sep 14 18:16:32 Alright it setup them like you said, wow Sep 14 18:16:34 I fail Sep 14 18:16:47 and as tofra said you want to populate the arraylist in, for instance, the oncreate method and then pass it into the constructor of the adapter Sep 14 18:16:52 Do you see how it is working? Sep 14 18:17:05 Yes Sep 14 18:17:10 so now you can use the position arg as a reference handle to select which element of the arraylist you want and set up the rest as needed Sep 14 18:17:11 Position of the listview Sep 14 18:17:30 Quick one, do you know how to add a button to the bottom of the listview? Sep 14 18:17:44 at the bottom of the entire listview or in each row? Sep 14 18:17:53 Entire listview Sep 14 18:18:03 when clicked I will have it return to the previous activity Sep 14 18:18:11 you don' Sep 14 18:18:21 you don't really need that since that is what the back button is for Sep 14 18:18:38 But its not convenient for use Sep 14 18:18:40 correct! Sep 14 18:18:44 How can I set it back tot he activity then Sep 14 18:18:46 but if you want to implement it you want to use I think a relative layout with the button aligned to bottom and the listview aligned above Sep 14 18:18:58 It's not OK to put a back button on the screen.. Sep 14 18:19:19 then how can I let my users go back Sep 14 18:19:21 Use your back button, or the back in your actionbar.. Sep 14 18:19:25 Maybe on a dialog but the whole point of the physical/tablet back button is for that reason Sep 14 18:19:25 using the back button would be weird Sep 14 18:19:29 or have a button in the last item of the list if you want it to scroll with the list Sep 14 18:19:46 Alright I gtg, I will be back later Sep 14 18:19:48 Snwspeckle: this is phones Sep 14 18:19:53 not windows or desktops Sep 14 18:20:01 the backbutton is the convention for navigating back Sep 14 18:20:06 different paradigm Sep 14 18:20:31 why would back button be weird? thats all it does Sep 14 18:20:40 Although if it makes you feel better you "could" use a "confirm" button Sep 14 18:20:53 I will be back, sorry for cutting you off short, class just ended Sep 14 18:21:01 no worries Sep 14 18:21:34 important thing is you got the just of the listview :) Sep 14 18:22:05 so if we're using something like the save and discard in the actionbar, what should the back button be doing, save or discard? Sep 14 18:22:50 I think the important thing to take away from this is that you should watch the google io listview talk before using listviews Sep 14 18:23:14 tbh I think he'd be better off with a page viewer Sep 14 18:23:27 whats was he doing? Sep 14 18:23:28 @SimonVT: GREAT ANSWER :-) Sep 14 18:23:41 homework Sep 14 18:23:47 inputting some kind of test scores hah ha yeah Sep 14 18:24:21 f2prateek_ back button should do niether Sep 14 18:24:25 *in that case? Sep 14 18:24:34 SimonVT, did you saw my problem? Sep 14 18:24:53 because the equivalent buttons would be "submit" and "reset" Sep 14 18:25:49 hi all Sep 14 18:26:08 having an issue with a fragment transaction Sep 14 18:26:16 Can the nexus 7 be powered thru usb to computer? Sep 14 18:26:21 yes Sep 14 18:26:32 Ahh that must be nice to develop with then ;p Sep 14 18:27:00 basically i have a two-screen workflow, and pressing back at any time should exit out of it Sep 14 18:27:43 lasserix: well i was little confused since I'm not sure what the difference between pressing the back button vs pressing the done button is Sep 14 18:27:45 Ah, another android guru online :-) Lov, do you saw my question about dissapearing calender? Sep 14 18:27:58 thats in the stock alarm clock application on jb Sep 14 18:28:12 oh i guess I can check it in the source actually Sep 14 18:28:14 the first screen i show by calling replace and addToBackStack Sep 14 18:28:25 f2prateek_ done would verify that the inputs have been inputed, back would just cancel all action Sep 14 18:28:29 for the second screen, i call replace again but not addToBackStack Sep 14 18:28:57 pressing back on the second screen doesn't work properly Sep 14 18:29:09 i still see what's on that second screen Sep 14 18:29:21 until i hit back again, but that takes me back too far Sep 14 18:29:55 lasserix: the stock clock actually does the exact same thing with done and back. (try putting a label for an alarm, press back, alarm is assigned label, presssing done will do the same thing) Sep 14 18:30:29 lasserix: what you said was what I thought should have happened Sep 14 18:30:43 Oh yeah! well I guess if your forms are more complicated, I was just saying hypothetically and for the sake of someone who wants the ux confirmation Sep 14 18:31:18 oh i'm not even developing anything, that thing about the back button in the listview made me think Sep 14 18:31:30 karakuri I'm not sure but why aren't you adding second frag to backstack? Sep 14 18:32:31 our use case doesn't allow the user to go back to that screen Sep 14 18:32:44 they either complete the whole flow or start from the beginning Sep 14 18:33:07 so when they are in the second screen you want to application to finish or go back to some intro screen? Sep 14 18:33:16 can i get maven to not deploy to all connected devices? i have emulator in my config, and it only runs tests on the emulator, but i dont want it deployed to every device i have plugged in at the time Sep 14 18:33:23 yeah, let me rephrase things Sep 14 18:33:36 there are three screens Sep 14 18:33:53 So you just want to know how to move back two fragments in one back button press? Sep 14 18:33:54 you can go from A to B and B to C, but not C to B Sep 14 18:34:01 right Sep 14 18:34:13 pressing back from B or from C needs to go to A Sep 14 18:36:16 Uhh seems like there must be an easy to do this Sep 14 18:36:48 what i tried was addToBackStack when going to A to B, and then not doing that from B to C, but that seems to mess up Sep 14 18:36:58 simple when using the up action, just set the up target Sep 14 18:37:00 the view hierarchy doesn't change back to A when i hit back Sep 14 18:37:16 you can set b to have no history Sep 14 18:37:16 and pressing back a second time takes me to whatever i was doing before A Sep 14 18:37:22 you can override the back button Sep 14 18:37:24 just dont do any addToBackStack? Sep 14 18:37:32 you can have b call finish() when c launches Sep 14 18:37:47 these are fragments, not activities, so no finish Sep 14 18:37:55 yeah i just realized that :) Sep 14 18:38:02 you can just clear the stack and replace to intro fragment in onback? Sep 14 18:38:07 karakuri: oh just realised you tried that Sep 14 18:38:53 so, backstack manipulation, where do i start? Sep 14 18:39:08 The question is how to pop two fragments off the backstack Sep 14 18:39:33 Seems like something reasonable so someone must know Sep 14 18:39:40 override the back behavior or dont add b to the back stack Sep 14 18:40:42 karakuri: look at popBackStackImmediate Sep 14 18:41:23 or popBackStack(int id, int flags) Sep 14 18:41:30 and just use id of intro fragment Sep 14 18:41:42 I get "Error - Package file was not signed correctly.". I had to reinstall my PC. Maybe I was using JDK6 and now I have JDK7. Sep 14 18:42:00 in my user profile, there's a "public key" - is that of any use? Sep 14 18:42:21 karakuri: there might be something in here http://developer.android.com/reference/android/app/FragmentManager.html#popBackStack() Sep 14 18:42:35 or its family Sep 14 18:42:50 yeah, looking at that as we speak Sep 14 18:42:54 i'll figure it out Sep 14 18:43:00 or i'll get our designers to change it Sep 14 18:50:27 anyone here own an Asus TF700? Sep 14 18:50:34 thinking about getting one Sep 14 18:53:06 QUERY: I'm fresh-caught, and I have a rather specific itch to scratch. I use the eBay mobile app, and eSnipe, which has a mobile site, but not an app. ISTM it should be possible to create something that would be.. I gather it's a SEND_ACTION target app, which would accept eBay's share, regex the text it got to extract an item number, and create a URL with that and a pre-programmed text string... and then hand that back to Andr Sep 14 18:53:15 Pointers? Sep 14 18:54:32 CURSOR: You could probably use BeautifulSoup along with the ASE to whip something up, but bear in mind that your app is liable to break for any reason due to eSnipe changing. Sep 14 18:54:41 karakuri i have the older model, the keyboard is really nice but since I've heard you can use wireless keyboard and mouse with nexus 7 who knows? although the double battery life i really, really nice. you might try #android Sep 14 18:55:17 thinking about nexus 7 as well Sep 14 19:01:25 hey, is it possible to do opengl processing and not render to the screen? Sep 14 19:01:36 instead output to a buffer or even a file Sep 14 19:14:48 back Sep 14 19:15:12 hey guys i follow this tutorial and the pan feature is way too sensitive Sep 14 19:15:14 https://developers.google.com/maps/documentation/android/hello-mapview Sep 14 19:15:31 the map works but it does not pan like the doc says it should Sep 14 19:16:13 i touch the screen and zooms out like crazy Sep 14 19:16:34 i pan and it jumps to the ocean Sep 14 19:16:38 greets... so I enter some stuff into some EditTexts and then hit a button to "get results".. I'm trying to get that button to also hide the soft keyboard when it's pressed, but this: http://stackoverflow.com/questions/1109022/close-hide-the-android-soft-keyboard (using the getWindow().setSoftInputMode() method) doesn't seem to work...any tips? Sep 14 19:18:30 tnzr, getSystemService(Context.INPUT_METHOD_SERVICE)... imm.hideSoftInputFromWindow(tetEditInstance.getWindowToken(), 0); Sep 14 19:18:46 (cast service to InputMethodManager) Sep 14 19:18:56 you prob need to call this on all your fields to ensure success Sep 14 19:20:41 I am putting a button under a listview in relative layout but its not showing on my screen? Sep 14 19:21:37 listview is probably taking the whole screen? Sep 14 19:22:02 So how can I extend the button so I can scroll to it? Sep 14 19:22:14 is the only time that your app will be killed after an activity swap is if the OS/new activity needs the resources currently being used by your activity? Sep 14 19:22:26 not necessarily Sep 14 19:22:34 gutano: awesome thanks Sep 14 19:22:44 it may just decide that there are too many processes running, even if there's enough resources Sep 14 19:22:57 I THINK that there's an LRU that handles this Sep 14 19:22:59 but don't quote me on this. Sep 14 19:23:13 Snwspeckle: let the button have whatever height you want, have the scrollview fille the remaining screen Sep 14 19:23:16 *fill Sep 14 19:23:24 so, I should always save the data during onPause? Sep 14 19:23:30 if I need it later at all Sep 14 19:23:34 ok wtf Sep 14 19:23:47 eclipse, I got in my view xml and its randomly changing values Sep 14 19:23:49 wtf is going on Sep 14 19:23:50 *listview not scrollview Sep 14 19:24:03 restart eclipse Sep 14 19:24:09 grrr Sep 14 19:25:31 snwspeckle button must have alignparentbottom=true and listview must have layout_above=button Sep 14 19:26:24 my eclipse is messed up Sep 14 19:26:34 I click in the xml editor and code appears randomly Sep 14 19:26:40 and disappears.. Sep 14 19:26:50 poltergeist Sep 14 19:26:53 Noxz: you should be storing and restoring your state anyway as part of the activity cycle Sep 14 19:26:54 yup, it happens Sep 14 19:27:06 its annoying as hell my god Sep 14 19:27:10 you should always assume that your application is liable to be killed at any time that you're not directly in the foreground Sep 14 19:27:13 GOOGLE MAKE YOUR OWN IDEA -.- Sep 14 19:27:20 IDE* Sep 14 19:27:40 they are kinda busy building a one to one mapping of the earth Sep 14 19:27:53 lov# does that apply to a Preference screen in your same application? Sep 14 19:27:55 Mehh Google is failing with Android imo Sep 14 19:28:03 the OEM's make shit phones every year Sep 14 19:28:08 where onPause is called when going to the Prefs screen Sep 14 19:28:10 nothing 'game changing' Sep 14 19:28:13 uh Sep 14 19:28:23 going to the preferences screen is really just a special case of your activity getting backgrounded. Sep 14 19:28:28 the user could go to preferences and then press home Sep 14 19:28:33 Snwspeckle: what would you rather be seeing on Android phones? Sep 14 19:28:37 ohh Sep 14 19:28:45 Snwspeckle: err, it's opensource? Google doesn't actually have the power to make them do much of anything? Sep 14 19:29:15 ctate: I would get 10 erections if google withheld market for emitting error level logging for things that aren't errors Sep 14 19:29:27 my god it makes debugging an utter fucking chore Sep 14 19:29:32 Meh it more the fact Google needs their own phone and not all these crappy repetitive phones. If Google can show OEM's they need to make their phones cleaner then they will change Sep 14 19:29:40 ha ha ha Sep 14 19:29:47 that's adorable. Sep 14 19:29:48 All the OEM's care about is money and they just dish out the same basic phone 6 times a year' Sep 14 19:29:49 The nexus line is pretty good for that Sep 14 19:29:50 you know about the nexus phones right? Sep 14 19:29:53 also, overshare Sep 14 19:29:57 Snwspeckle: also, there have been much more evolution there than in iOS land... while BB is playing catch-up, and Windows Phone has some crucial problems (at least from EU-centric pov) Sep 14 19:30:20 Ok motodev is pissing me off now Sep 14 19:30:35 R cannot be resolved error now, da hell... Sep 14 19:30:44 motodev Sep 14 19:30:48 there's your first mistake. Sep 14 19:30:54 I like it Sep 14 19:30:56 you dont? Sep 14 19:33:19 you were just talking about OEMs crapping things up...? ;) Sep 14 19:33:49 They are terrible Sep 14 19:33:56 OEM's are a joke and ruin Android Sep 14 19:34:15 Google + Google Designed Phone + Google not taking crap from carriers = My Dream Sep 14 19:34:45 carriers (especially in USA) will give you crap anyway Sep 14 19:35:01 that's what apple is doing... google is winning (market share, perhaps not $$$) by doing the opposite and catering to everyone Sep 14 19:35:02 Duh, all they care about is money, greedy american ethics Sep 14 19:35:23 Google has really not had any trouble with the whole 'acquiring $$$' thing... Sep 14 19:35:34 apple<->carrier was a both-way crapfest (been working for a carrier when it introduced iphone 3g... was lulzy time) Sep 14 19:35:41 all business enterprises must be greedy by defination Sep 14 19:35:42 no, cetainly not... but it's not due to android phone sales Sep 14 19:36:06 Android phones just lack design and quality, all they do is dish out phones every 2 months to turn around a profit Sep 14 19:36:59 Snwspeckle: Android is about choice. There's still a fair number of people who clearly like the OEM customizations Sep 14 19:37:00 Well back onto programming Sep 14 19:37:26 I need some assistance plotting my theory Sep 14 19:37:48 a pretty big complaint I received lately (after updating a phone to CM10) was... lack of TouchWiz Sep 14 19:38:13 So I have dynamically created EditText widgets but now I need to take the data from each one, plug it into a formula, get a result, put it in an array and keep doing this till all the values are complete Sep 14 19:38:26 p_l|backup: i did that for a friend once, and they were lost without touchWIz Sep 14 19:38:27 Snwspeckle: understand that, no matter how much google markets a nexus phone, the carriers will spend even more marketing their craptastic android phones with bloatware. Sep 14 19:38:42 seriously? there was never a way to add a listener for onContextItemSelected? you *have* to override the Activity method??? Sep 14 19:39:00 g00s: Its so sad its true Sep 14 19:39:39 the thing is, Android from beginning provided places to customize by OEM, although some of them went and put crap in other places. Hell, I suspect that some stuff was missing on purpose expecting OEMs to put their own twist Sep 14 19:39:47 But can anyone possibly help me with my theory Sep 14 19:40:52 I basically need to construct a foreach loop for each EditText widget, get its data from it, run it through an algorithm, get the answer into an array and then keep looping till all are complete Sep 14 19:41:36 Snwspeckle: can't you simply copy the values (through a loop), put it into formula then display? Sep 14 19:42:02 snwspeckle you don't need a loop Sep 14 19:42:05 Well the issue is that the EditText widgets are displaying dynamically and not static Sep 14 19:42:12 you are using the listview to accompilish the task Sep 14 19:42:15 if an activity is left I should use removeCallbacks(Runnable r) in onPause() to ensure that no more posts of r will be run, right? Sep 14 19:42:36 so? Trigger the evaluation on change of contents of any of those Sep 14 19:42:38 So lasserix, just go through the listview by there id? Sep 14 19:43:14 the user will see the list, input answer to each row, when done they click confirm Sep 14 19:43:29 Exactly how I have my layout atm Sep 14 19:43:55 i think there are two ways you can get the input, either by row or by all rows Sep 14 19:44:29 So how do I do all the calculations at once without a loop? Sep 14 19:44:40 use the loop after to calculate Sep 14 19:44:45 you don't need a loop to display Sep 14 19:44:57 that is the point of using a listview, since it displays all at once Sep 14 19:45:06 *loop after to iterate through your input values Sep 14 19:45:28 Yea thats what I am saying Sep 14 19:45:47 So when the user hits confirm the loop begins and pulls from row 1, calculates, inputs in array then loops again Sep 14 19:47:09 http://stackoverflow.com/questions/10536128/how-to-get-checkbox-edittext-with-listview-values-in-android-while-clicking-a-bu Sep 14 19:48:44 and http://www.mindfiresolutions.com/Keeping-the-data-in-ListView-with-interactive-items-while-scrolling-in-Android-1621.php Sep 14 19:48:51 Ad then use position as the counter right? Sep 14 19:49:11 the user will go through each row adding their input Sep 14 19:50:05 vListSortOrder=l1.getChildAt(i); --- Whats this doing? Sep 14 19:50:13 iterating through each row Sep 14 19:50:25 l1 is the listview getchilding ie getting all rows Sep 14 19:50:52 someone else will have to chime in because i dont know that the edittext once changed will be preserved by the listview when the view gets recycled Sep 14 19:51:07 ah true Sep 14 19:52:22 repo --version shows a 1.8.2 version from git-repo googlecode project but the latest version available on project's page is 1.7 . Where's 1.8.2? Sep 14 19:52:44 probably not strictly IT since it's not about application framework Sep 14 19:53:03 anyways you'll have to save the edittext value once inputted somewhere, and then at the end list through whatever is saving all those values and do your caclulation Sep 14 19:53:45 i would be ofmore help but it is like 110 degrees Sep 14 19:57:07 god damn eclipse is being 100% dumb Sep 14 19:57:11 throwing random errors Sep 14 19:59:11 Snwspeckle: that's what Eclipse does... Sep 14 19:59:26 I just recently had the "magical invisible character that makes everything around it have inexplicable syntax errors" feature Sep 14 19:59:31 submit.setOnClickListener(new View.OnClickListener() { -- says submit doesnt exist, its my onClick parameter.... Sep 14 19:59:51 what is 'submit'? Sep 14 20:00:26 onClick parameter of my button Sep 14 20:01:19 you call setOnClickListener() on a Button Sep 14 20:01:33 yea Sep 14 20:02:41 nvm restarted eclipse Sep 14 20:02:42 fixed Sep 14 20:05:08 Ok what kind of variable are row positions as? Sep 14 20:08:30 Arggg.. I'm getting crazy.. I'm creating an calendar, and when i reboot my phone, it's gone.. Sep 14 20:10:20 @Snwspeckle: int ? Sep 14 20:12:06 Can someone look at my code here http://pastebin.com/56fFyELe and tell me what's wrong with my onPause implementation? Sep 14 20:12:18 lasserix: Ok so now the issue is that position is not accessible since its outside the getView Sep 14 20:12:43 @qkzoo1978: what is wrong? Sep 14 20:13:33 It crashes when it should be quietly destroyed. I'm a java/android Noob so I'm sure I goofed. Sep 14 20:14:01 what's the error message? Sep 14 20:15:03 "unfortunately, fLash has crashed" Sep 14 20:16:22 Whoops, "unfortunately, qFlash has stopped." Sep 14 20:16:58 note the imgur link in the channel topic. qkzoo1978 Sep 14 20:17:14 when it crashes, it logs a stack trace of where it crashed Sep 14 20:17:20 pastebin is yor frend :) Sep 14 20:18:01 ctate: that image is a gift that keeps on giving Sep 14 20:18:08 that is why it is still in the channel topic Sep 14 20:18:10 say I have a "Logout" button on the action bar (not a tab in the stacked mode, but on the actual action bar) .. what part of the theme would one override to change the text color on something like that? Sep 14 20:18:14 it is full of truth Sep 14 20:18:27 and by "button" I should have probably said "TextView" Sep 14 20:19:49 Oh, okay, umm, I'm not using eclipse, is it still present? Sep 14 20:21:24 * lov glares at qkzoo1978 Sep 14 20:21:26 adb logcat *:E Sep 14 20:21:27 go Sep 14 20:22:38 This is all onboard my phone, don't have A pc. Sep 14 20:22:48 Aide. Sep 14 20:23:21 how are you doing development? Sep 14 20:23:28 are you somehow compiling on your phone? Sep 14 20:25:13 lov: ... yes Sep 14 20:25:45 Yes, using Aide. It works well so far, albeit does have some restrictions and limitations, but pretty robust otherwise. Has it's own ide, completion, etc. Sep 14 20:27:41 oh. neat! Sep 14 20:28:17 well, find some logcat app, and run it at error level Sep 14 20:29:38 Anyone care to help on how to get the row id Sep 14 20:30:04 Is this recommended www.bugsense.com/ ? Sep 14 20:30:39 are these options available to all devices out there? -march=armv6 -marm -mfloat-abi=softfp -mfpu=vfp Sep 14 20:31:51 yay, multi choice modal listview / action modes working on Gingerbread ;-) Sep 14 20:32:05 thanks to ABS Sep 14 20:32:07 qkzoo1978: if you don't know how to do it yourself, sure, why not. Sep 14 20:32:10 pay out the moola Sep 14 20:32:16 and the last 12hr of my hacking ;-) Sep 14 20:32:18 note that google play includes crash reporting. Sep 14 20:32:24 Heya. I am new to android development. Where is a good place to start? Sep 14 20:32:27 the metrics may not be as in depth or detailed, however. Sep 14 20:32:58 Theres just ONE glitch. On touch end, after long press, the selected item is unchecked, and ends the action mode. Sep 14 20:33:41 Ok someones gotta help me here lol Sep 14 20:34:50 Get the row id of what? Sep 14 20:34:51 Cannot refer to a non-final variable position inside an inner class defined in a different method --- huh?? Sep 14 20:35:09 for (int i = 0; i < position.getCount() ; i++) { -- part of my loop which grabs the row id Sep 14 20:35:16 or total number of rows Sep 14 20:35:24 if you're using eclipse, Ctr-1 for help Sep 14 20:35:33 it will suggest to set the offending variable to final Sep 14 20:35:40 therianthrope: The Google pages on it are actually fairly nice. Sep 14 20:35:48 therianthrope: Do you have the development environment set up yet? Sep 14 20:35:58 pastebin please Sep 14 20:36:14 so the mmssms.db was corrupted and removed. Im planning on rebooting into recovery and dumping everything - but is there anything else I should try before rebooting? Is it possible to dump the entire ram using adb? Or is there a way to rebuild the database? Sep 14 20:36:41 tofra: https://gist.github.com/986b98dd752aa77ca972 Sep 14 20:37:02 Hodapp: yes Sep 14 20:37:16 tavelram: pull mmssms.db *and* the journal to your computer, try using sqlite3 till you arrive at working copy Sep 14 20:37:33 p_l|backup, thanks, where is the journal? Sep 14 20:37:34 then copy it back and kill messagin app Sep 14 20:37:40 tavelram: in the same directory as db Sep 14 20:38:22 I once had phone app not loading because the journal was corrupted and couldn't be replayed, which led to failure in opening the database Sep 14 20:38:26 p_l|backup, .db-shm or .dm-wal? Sep 14 20:38:29 db Sep 14 20:38:44 tavelram: db-wal, I think. That's Write-Ahead Log Sep 14 20:38:53 First, we told you to move the creation of myItems out of the adapter.. it's still there.. :-) Sep 14 20:38:57 not sure about db-shm, it might be shared memory thing Sep 14 20:39:05 And what do you want to happen when you click on an item? Sep 14 20:39:07 just in case copy it as well Sep 14 20:39:14 ok, already copied all of it. Sep 14 20:39:44 sometimes it will work if you delete the auxiliary files and force it to run with db Sep 14 20:40:06 Its when you click the button Sep 14 20:40:12 p_l|backup, but the db got corrupted because it was unable to resize to above 1MiB. And now it's 1024 byte instead, and wal is 45KiB.. Sep 14 20:40:17 gah, I am really not liking this min3d library. Sep 14 20:40:30 tofra: I was just AFK for a little so I probably missed that Sep 14 20:40:38 I've used it to load a 3ds model, and all of the vertices are now hidden inside this faux-OO bullshit. Sep 14 20:40:43 tavelram: ... sounds bad Sep 14 20:40:45 I can't access them; they're package-private. Sep 14 20:40:58 faces are accessible, colors are accessible, vertices... are not. Sep 14 20:41:04 tavelram: try things on PC, but some data might be lost Sep 14 20:41:11 where is the submit button? Sep 14 20:41:16 on every row? Sep 14 20:41:18 p_l|backup, yeah. is there a way to dump the entire ram? Sep 14 20:41:42 At the end, the listview is above it Sep 14 20:41:42 tavelram: well, you can use gdb to dump the whole process Sep 14 20:41:57 and I think you can dump stuff through /proc/ Sep 14 20:42:13 then move the part for the submit button out of the adapter... Sep 14 20:42:33 In the adapter you only want stuff which is related to the items in the list, or the list it self.. Sep 14 20:42:34 But I need to get the total number of rows from the list then loop threw them Sep 14 20:42:38 move it to onCreate.. Sep 14 20:43:15 Ok so then how do I access position? Sep 14 20:43:18 Anyone else have a 3DS or STL reader they can recommend? Sep 14 20:43:45 What you can do, is: create a variable: int mSelectedIndex; (on the top of your code) Sep 14 20:44:49 But what happens when they button is clicked is it goes through the rows and calculates for each row Sep 14 20:44:52 eh.. nvm.. let me rethink.. Sep 14 20:45:24 So when the user hits confirm the loop begins and pulls from row 1, calculates, inputs in array then loops again Sep 14 20:45:46 What is your goal? have a list with entries, when clicking on an entry nothing should happen, but when clicking on the submit, some magic happens? Sep 14 20:46:06 yes Sep 14 20:46:17 p_l|backup, ok, how do I do that? I tried dd:ing /proc/pid/mem but only got 0 bytes. Sep 14 20:46:18 the fields are just for user input and they are dynamic Sep 14 20:46:36 tavelram: just checked that and it doesn't work Sep 14 20:46:41 the getView is called when a row is displayed.. So you are searching (findViewById) the submit button on every row.. Sep 14 20:46:45 tavelram: I think it can be only mmaped Sep 14 20:47:02 you only want to do it once.. as you only have one button.. Sep 14 20:47:04 tavelram: however you can run gdbserver on the phone, and use adb to forward the connection to gdb on the PC Sep 14 20:47:10 No I only have one button Sep 14 20:47:17 My button is statically placed Sep 14 20:48:10 So what you should do, create a global variable, in the getView create view.setOnClicklistener, and that listner sets the global variable.. Sep 14 20:48:47 I think I got it Sep 14 20:48:50 Ugh, will that adb *:E command work through any emulators? Sep 14 20:49:13 p_l|backup, does gdbserver only work on a new process? Sep 14 20:49:23 the submit button should not be in the adapter.. it should be in onCreate.. Sep 14 20:49:45 Broke here: String temp = edit.getText().toString(); Sep 14 20:49:45 tavelram: gdbserver --pid= as root Sep 14 20:50:22 tofra: Heres a revision I am currently using, https://gist.github.com/986b98dd752aa77ca972 Sep 14 20:51:24 your onclick of your submit contains the for loop.. Sep 14 20:51:45 not needed, in the onclick you should read indexVal.. Sep 14 20:51:57 Not emulators, I mean terminal emulators, onboard, or does it need to be run off device with cable? Sep 14 20:52:28 tofra: No I shouldnt? Sep 14 20:52:32 move the filling of myItems also to the oncreate.. Sep 14 20:52:37 indexVal is used to create the forms Sep 14 20:53:39 ok.. create a new variable mSelectedIndex.. Sep 14 20:53:58 btw: android sugests variables which are global start with m... Sep 14 20:53:58 :D Sep 14 20:54:28 ok, good tip :) Sep 14 20:56:01 then in de adapter, before you return convertView, add: convertView.setOnClickListner.. In this listener do mSelectedIndex = position.. Sep 14 20:57:55 although creating a different onclick listner for each row isn't the most efficient way, but I don't have an example available now.. It's late already :D Sep 14 20:58:26 tofra: Thing is I can grab it based on my listview, I am crashing on my getText Sep 14 20:58:32 EditText edit = (EditText) findViewById(R.id.ItemCaption); Sep 14 20:58:36 I keep crashing here Sep 14 20:59:25 update your github Sep 14 20:59:59 revised Sep 14 21:00:02 https://gist.github.com/986b98dd752aa77ca972 Sep 14 21:00:28 what's the message? Sep 14 21:00:44 NullPointerException... Sep 14 21:01:08 can't be on that line.. Sep 14 21:01:27 also, you don't want to search a view.. Sep 14 21:01:32 Thats all that logcat tells me Sep 14 21:01:48 whatcha mean search the view? Sep 14 21:01:50 .. in a for/loop.. Sep 14 21:03:00 How do I see with logcat what the error is.. Sep 14 21:03:40 you probably get it in line 54.. Sep 14 21:03:54 It didn't find the ItemCaption.. Sep 14 21:04:26 Why wouldnt it Sep 14 21:04:28 its a layout.. Sep 14 21:05:16 I'm still not clear on what you actually want, but I think it's pretty basic.. So with any basic tutorial you should be able to create a list view, and show the selected index.. Sep 14 21:06:09 oh the ItemCaption is not the view my onCreate is in Sep 14 21:06:20 so I need to search for that item in a different layout Sep 14 21:06:33 you should not search...... Sep 14 21:07:01 k but I still need to search for it Sep 14 21:07:11 within the getView, the onClicklistner should set the mSelectedIndex.. Sep 14 21:07:33 in the submit.setOnclicklistner, you should read the mSelectedIndex.. Sep 14 21:08:44 The problem is that I am not sure how to get the id of a widget in a layout not called in onCreate Sep 14 21:09:45 eh, how do you mean? Sep 14 21:09:57 well.. Sep 14 21:10:14 outside the adapter, you should not search for a view.. Sep 14 21:10:14 The EditText widget is in a different layout, I inflate that layout with my listview to make it dynmically Sep 14 21:10:36 the object in myItems should contain all data which you want to have, and need to build the view.. Sep 14 21:10:40 can put View to Intent as Extra? Sep 14 21:10:40 Regardless of my bad programming habits, its erroring because it cannot find it Sep 14 21:10:53 Serializable Sep 14 21:10:54 ? Sep 14 21:11:01 p_l|backup, oh, I haven't got gdbclient. Is there a way to dump the entire ram from cwm I guess it would survive? Sep 14 21:11:19 tavelram: gdb can attach to gdbserver Sep 14 21:11:26 get gdb from ndk Sep 14 21:11:50 ok, thanks Sep 14 21:13:09 I don't remember the syntax from memory, been some time since I did remote debugging Sep 14 21:14:12 Ah it worked! Sep 14 21:14:17 Nice, partially got it working Sep 14 21:14:39 But 1 problem Sep 14 21:16:13 hello guys Sep 14 21:16:35 somebody know somthing about bluetooth? Sep 14 21:16:44 programming Sep 14 21:20:36 p_l|backup, ok, np :) Sep 14 21:21:26 i need catch volume up and down buttons of bluetooth Sep 14 21:22:42 p_l|backup, can I dd the system partition while in android? Sep 14 21:22:49 tavelram: yes Sep 14 21:22:58 just take care that the partition might in some lulzy fs Sep 14 21:23:23 ok, yeah Sep 14 21:23:31 have some deleopers here :D Sep 14 21:24:00 demonoid_com, I've got no idea, sorry. Sep 14 21:24:11 tavelram: I think one of the OSes on phones when queried with file would give you "VMS Alpha Executable" :D Sep 14 21:24:25 tavelram np :) Sep 14 21:24:32 p_l|backup: a lot of files will give you that Sep 14 21:24:52 zinx: in practice, I found only RFS disk images to have that signature Sep 14 21:24:52 p_l|backup: the magic for that is way too generic Sep 14 21:24:58 heh Sep 14 21:25:44 zinx: the only files I *had* which reported that were Samsung RFS images Sep 14 21:25:59 though I think .COM is worse Sep 14 21:26:30 * p_l|backup had seen a HTML file *run* after renaming it to .COM Sep 14 21:28:40 p_l|backup: it's not hard to make a .com that runs Sep 14 21:28:51 is there a way that I can epoert two separate apps Sep 14 21:29:02 p_l|backup: they are literally just loaded in to memory verbatim Sep 14 21:29:05 by changing the package name in AndirodManifest.xml ? Sep 14 21:29:11 p_l|backup: at 0x100 or so Sep 14 21:29:14 zinx: yeah, I know Sep 14 21:29:24 zinx: now try figuring a magic value for that :) Sep 14 21:29:37 p_l|backup: file attempts it :P Sep 14 21:30:09 (the funny thing is that it was a random HTML file with error message that got downloaded instead of .COM... and not only it ran, it ran and printed gibberish to screen in endless loop) Sep 14 21:30:45 hehe Sep 14 21:34:57 Ok, is onPause the proper place to remove a update request? Does onPause Always get called when an app quits? Sep 14 21:35:10 p_l|backup, :) Sep 14 21:35:18 GPS update request. Sep 14 21:35:33 qkzoo1978: it gets called for a lot of reasons Sep 14 21:35:41 qkzoo1978: read the activity lifecycle page Sep 14 21:35:45 tavelram: so, did you succeed in dumping memory? Sep 14 21:36:27 dd:ed the datadata while I was waiting for the unzipping to finish. Sep 14 21:36:36 (btw, you can try dumping memory of whole device with /dev/mem or /dev/kmem... but it might lead to HIDDEN FUN STUFF) Sep 14 21:36:40 qkzoo1978: http://developer.android.com/guide/components/activities.html and http://developer.android.com/training/basics/activity-lifecycle/index.html Sep 14 21:37:25 Thank you, reading up on it now. Sep 14 21:38:59 p_l|backup, yeah, Ill probably try that first then actually Sep 14 21:39:01 that is a bad idea Sep 14 21:39:01 p_l|backup can you help me ? i need find where is my bluetooth in /dev/? Sep 14 21:39:07 if you hit iomem you're going to crash the phone Sep 14 21:39:16 well, i should say, WHEN you hit iomem Sep 14 21:39:36 demonoid_com: probably it's not in /dev/ Sep 14 21:39:57 p_l|backup: wow and where is Sep 14 21:40:13 zinx, ok, could I start at an offset or something? Sep 14 21:40:56 p_l|backup i need sniff bluetooth traffic Sep 14 21:41:09 demonoid_com: nowhere, because BlueZ (the bluetooth stack used on Linux) doesn't work like that. You'll first need to bind a connection to rfcomm device, which will then show up in /dev/ (not guaranteed it will work this way, direct bt access isn't supported afaik, only through the official java apis) Sep 14 21:41:29 tavelram: eh, just cat /proc/iomem and don't dump anything that's actually io Sep 14 21:41:40 tavelram: i.e., feel free to dump System RAM Sep 14 21:41:50 tavelram: but don't dump something that says, for instance, that it's for video or sound :P Sep 14 21:42:07 demonoid_com: ... you're on your own. You'll have to at least root the system, then find how bluez supports sniffing (if at all) Sep 14 21:43:10 p_l|backup: ok and how i can catch bluetooth buttons for volume up and down ? in API 8 did you know? Sep 14 21:43:39 demonoid_com: I think they're converted to normal input events by framework Sep 14 21:44:09 p_l|backup;i read all and there only say onReciver but this is only for media buttons play,stop next ... Sep 14 21:44:33 demonoid_com: not sure, they might be forwarded only to mixer (and not always) Sep 14 21:44:39 p_l|backup: but volume up and down isn't it :( Sep 14 21:46:24 p_l|backup and i think i need android NDK and write my own event Sep 14 21:47:07 demonoid_com: I suspect you won't be able to catch it Sep 14 21:47:57 p_l|backup:mmm why? i need only catch index of volume Sep 14 21:48:12 zinx, ok, thanks :) Sep 14 21:48:19 p_l|backup:bluetooth in call i need Sep 14 21:48:52 demonoid_com: there's no guarantee you'll get the event without some hackery that might easily broke, and then there's not always a guarantee the volume messages are sent to device at all Sep 14 21:50:07 p_l|backup:i read log of my eclips and there send bluetooth commands like AT+GTV=12 :) Sep 14 21:51:33 p_l|backup:every time when i press button for volume up on bluetooth send some commands AT+... Sep 14 21:52:04 well, you can try hacking the way the control stream is used Sep 14 21:54:53 p_l|backup:yes maybe this is one way :( Sep 14 21:57:11 p_l|backup:the good programmers told:"If you wanna make some good program use c/c++" :DDDD Sep 14 21:59:11 untrue on Android Sep 14 21:59:18 and I generally avoid C++ Sep 14 22:00:04 p_l|backup why? Sep 14 22:00:47 c++ is epicness Sep 14 22:01:16 C++ is not that great Sep 14 22:01:23 from a language perspective Sep 14 22:01:26 C++ is a patchwork of badly implemented kitchen sink Sep 14 22:01:58 a lot of its features are poorly designed and clumsy - such that you can do the same thing in C without any more work, given the proper library for it Sep 14 22:02:16 ==zinx Sep 14 22:02:27 especially the total mishmash that is C++x11 Sep 14 22:02:30 and the higher level stuff is better done in other languages with quite often a not so big performance penalty Sep 14 22:02:31 lol yeah Sep 14 22:02:37 that causes me so much trouble Sep 14 22:02:44 and if you are just crunching numbers, go for modern fortran Sep 14 22:02:49 because i need complex numbers, and C++'s complex numbers don't mesh with C99's Sep 14 22:03:08 * ctate was paid, in the late 80s and early 90s, to write FORTRAN IV code Sep 14 22:03:26 I'm currently considering generating C/C++ code from another language, to use sensible stuff where I can't fit it normally Sep 14 22:03:27 zinx: doesnt' boost have some complex stuff? :) Sep 14 22:03:33 p_l|backup: i'm not really a fan of fortran, and the thing about it being better for number crunching is a myth pretty much Sep 14 22:03:35 p_l|backup: yes, but you're a MADMAN. Sep 14 22:03:37 ctate: yes it does Sep 14 22:03:39 QED. Sep 14 22:03:51 ctate: but then you have to use C++'s complex numbers, which don't support some of the functions i need Sep 14 22:04:02 ctate: functions which are part of C99 Sep 14 22:04:11 zinx: mostly the fortran for numbers thing, these days, boils down to array ordering Sep 14 22:04:17 ctate: yeah Sep 14 22:04:25 fortran multidimensional arrays are column-ordered; C's are row-ordered Sep 14 22:04:31 which is dumb because you can order them how you want in C Sep 14 22:04:33 but a lot of the algos are column oriented Sep 14 22:04:44 right, you just have to pay attention Sep 14 22:04:47 yup Sep 14 22:04:57 but people are used to their [x][y] etc Sep 14 22:04:59 zinx: not really a myth - there are two things: the compilers are even better (for HPC stuff), and due to certain things it's easier to autoparallelize fortran than C/C++ (that is, without extra work on programmer side) Sep 14 22:05:03 ctate: *sigh* Sep 14 22:05:17 p_l|backup: modern C compilers are a lot smarter than you might think Sep 14 22:05:17 also yes the fact that fortran has no aliasing lets the compilers do more Sep 14 22:05:27 (and have a lot of the aliasing stuff too) Sep 14 22:05:41 zinx: it was based on comparison between modern compilers Sep 14 22:05:44 the author of the code has to pay attention to the aliasing stuff and use the pragmas and shit Sep 14 22:05:46 zinx: in HPC environment, mind you Sep 14 22:05:51 p_l|backup: wich what options? Sep 14 22:06:01 but the authors of the code are like particle physicists or molecular biologists, in practice Sep 14 22:06:04 so that's kind of a lost cause :) Sep 14 22:06:05 p_l|backup: if you do things like tell the C compiler anything can alias, of course it's going to make suboptimal code :/ Sep 14 22:06:11 heyup Sep 14 22:06:21 p_l|backup: and if you code them exactly the same, you're messing up your array ordering Sep 14 22:06:32 i'm trying to save a file, in c, using fwrite Sep 14 22:06:46 zinx: seriously, the assumptions about the CS chops of the people writing the code are not warranted here Sep 14 22:06:52 also, what ctate said ;) Sep 14 22:06:58 the FORTRAN IV work i was doing, that i mentioned? that was at NIST. Sep 14 22:07:07 working as teh computer jock for polymer chemists Sep 14 22:07:18 who would be the ones maintaining the code long after i'd left Sep 14 22:07:23 yeah Sep 14 22:07:32 this is where i acquired my habit of writing TONS OF COMMENTS Sep 14 22:07:35 hehe Sep 14 22:07:36 self-defense Sep 14 22:07:48 less calls you get the better eh; ) Sep 14 22:07:58 the call to fwrite returns, with number of data packets written, and apparently it writes okay, but then, can't find the file in the filesystem! Sep 14 22:08:00 ctate: You can accomplish the same thing by baking the information into the code itself. Sep 14 22:08:05 zinx: the thing is, Fortran as a language gives certain constructions that even non-CS people can get performance similar to well-done C code... and then there are tools which are feasible for Fortran but which require crazy CS skills in C Sep 14 22:08:06 And by using simple solutions. Sep 14 22:08:09 java on android is like c# on windows ;) Sep 14 22:08:15 MrQwak: you do need to fclose() the file Sep 14 22:08:15 rking: in practice this does not work well Sep 14 22:08:26 demonoid_com: you mean kicking the shit out of C? :P Sep 14 22:08:33 rking: because the people who will be hacking on the code six years later are not going to be proficient at reading it Sep 14 22:08:34 zinx: yup, i did do the fclose Sep 14 22:08:37 * rking ← Totally disagrees. Sep 14 22:08:44 this was my observation at the time Sep 14 22:08:48 ctate: Do you have a Github ID? Sep 14 22:08:56 MrQwak: it's probably not where you are expecting it to be, then, but there Sep 14 22:08:59 p_l|backup :DDDD Sep 14 22:09:01 (.NET on windows is much better choice, most of the time, than most other popular options) Sep 14 22:09:07 i would hope that things will shift ove rtime, but i note that most of hte scientists i was working with there in 1990 are still there Sep 14 22:09:08 that is, for windows developement Sep 14 22:09:15 rking: nope. why? Sep 14 22:09:24 ctate: I wanted to see some of your code. Sep 14 22:09:31 ctate: well, a lot of good went into updating fortran :) Sep 14 22:09:34 *that* code, or my current code? Sep 14 22:09:37 zinx: path is something like \data\data\com.company.app\files\myfile.txt Sep 14 22:09:38 my current code is in AOSP :) Sep 14 22:09:43 rking: scavenge android source? :) Sep 14 22:09:49 rking: efficient number crunching can have some odd stuff in it that needs to be explained in comments, though i can't comment on what ctate was working on specificly Sep 14 22:09:53 ctate: Anything that conforms to your standards of commenting. Sep 14 22:10:12 mm yeah most of my current stuff isn't commented that heavily, since it's for a different audience. Sep 14 22:10:17 MrQwak: well, i see at least two major problems there Sep 14 22:10:21 if you're reading, say, ActivityManagerService Sep 14 22:10:23 MrQwak: the first is that \ is not a path separator Sep 14 22:10:24 and you find some comments Sep 14 22:10:37 oh Sep 14 22:10:39 MrQwak: the second is that you should not hardcode /data/data/com.company.app/, as it may not be where your app's data is Sep 14 22:10:42 i should use /? Sep 14 22:10:47 it's not unlikely that i wrote them, since the original authors wrote basically none :) Sep 14 22:10:48 MrQwak: yes, / is the path separator Sep 14 22:10:55 it's not hardcoded Sep 14 22:11:00 MrQwak: ah, ok Sep 14 22:11:01 MrQwak: use teh APIs that tell you the path to your app's data Sep 14 22:11:06 ah ok Sep 14 22:11:31 yeah, i used the java api, to get the path. i'll double check the direction on the slashes Sep 14 22:11:34 but yes, make sure you're using the standard API functions to get the path Sep 14 22:11:35 [also use the APIs that tell you where shared/external storage lives. don't hardcode /sdcard or /mnt/sdcard. seriously.] Sep 14 22:11:43 :D Sep 14 22:11:51 * p_l|backup recently was lost trying to find where the sdcard was Sep 14 22:11:54 (there are /existing/ phones that put things in different places) Sep 14 22:11:56 pathUserData = (m_Context.getFilesDir()).getPath(); Sep 14 22:12:02 cool Sep 14 22:12:05 zinx: I have one that has /mnt/sdcard0 and /mnt/sdcard1 Sep 14 22:12:07 what zinx said Sep 14 22:12:18 does that sound about right zinx? Sep 14 22:12:20 yeah Sep 14 22:12:39 there are also phones for which Context.getDatabaseDir() is not a subdirectory of Context.getFilesDir(). Sep 14 22:12:43 be told! Sep 14 22:12:46 yeah Sep 14 22:12:50 unfortunately the paths returned by API changed between 2.3.3 and 4.1 so I had one application confused about its content Sep 14 22:12:54 have to use the most specific path function Sep 14 22:13:04 (at least the paths for this specific phone) Sep 14 22:13:43 ctate: Hrm. Well, I definitely don't believe in explaining the language via comments. Stopping that is step #1 in overcoming an addiction to commenting. Sep 14 22:14:00 rking: it's all about your target audience Sep 14 22:14:01 it's hard to comment too much Sep 14 22:14:05 it's very easy to comment too little Sep 14 22:14:12 (i'm looking at you, linux kernel) Sep 14 22:14:23 (i'm also looking at you, ActivityManagerService :) ) Sep 14 22:14:23 ctate: Totally disagree, as well. Sep 14 22:14:31 zinc, it's forward slashes, like '/' are you sure that's not right? Sep 14 22:14:33 Comments are a redundancy, and redundant information goes out of sync. Sep 14 22:14:37 no Sep 14 22:14:40 rking: but fwiw, algorithms can easily be complex enough that they warrant comments, and have nothing to do with the language Sep 14 22:14:51 comments are communciation of information that is not expressed directly in any given small bits of code Sep 14 22:14:57 zinx: Show me an example of a comment that justifies itself, and I'll stop being so general. =) Sep 14 22:15:05 the code will pretty much never tell you *why* something. Sep 14 22:15:10 ctate: That's what "Extract Method" is for. Sep 14 22:15:16 rking: ==ctate Sep 14 22:15:55 but that only applies when you're talking about a scope so limited that a single method can encapsulate the topic Sep 14 22:16:29 an example of comments that i think are seriously, seriously valuable Sep 14 22:16:47 are ones that describe any complex invariants that must be maintained throughout operation Sep 14 22:16:51 ctate: If a single method cannot encapsulate it, then there is a much bigger problem. Sep 14 22:17:22 Java is a bad example, by the way, of a system where it's easy to encode all knowledge into the system. It pushes against you when you try, but you have to push back. Sep 14 22:17:50 hang on Sep 14 22:18:28 when i browse the filesystem in DDMD, it's like /mnt/sdcard/data/data/com.company.app/files/... Sep 14 22:18:46 do i need to prepend /mnt/sdcard ? Sep 14 22:19:08 MrQwak: use path returned from API Sep 14 22:19:10 the have getPath was just returning /data/data/com.company.app/files/... Sep 14 22:19:13 or variants: "we would ideally encapsulate all of this under a single mutual exclusion region, but we can't do that because there is implicit reentrancy inside some of these outcalls. however, we also have guarnatees from [xyz] that it will deal with concurrency/contention on its own, and redundant initialization of [qrx] we can guarantee to be benign ourselves, so we leave the mutex *here* and can rely on [xyz]'s to manage contention safely Sep 14 22:19:28 that's the gist of a comment i wrote recently Sep 14 22:19:34 the path from the API does not exist on the filesystem Sep 14 22:19:40 that is somethign that is very much not expressible *in code* Sep 14 22:20:08 ctate: Have you looked at any of the languages that has deeper concurrency support? Sep 14 22:20:23 and whoever comes along in three years and wants to learn that code so they can work on it productively will thank me for it Sep 14 22:20:38 this is really odd Sep 14 22:20:53 rking: let's say you want the phase derivative with respect to time, of a windowed signal. let's say you choose to calculate it by convolving the signal with the derivative of the window you're using (which means, element by element multiplication in the time domain by the derivative of the window, then a fourier transform of the signal), which is a non-obvious way of doing it, but more efficient than the obvious way Sep 14 22:21:06 i tried saving, at '/mnt/sdcard/data/data/..' and that worked Sep 14 22:21:12 rking: how do you encode the information of /why/ what you're doing returns the results you're getting in the code itself? Sep 14 22:21:22 ctate: I'm in 100% agreement that that information must be included. What I doubt is that a comment is the best way to do so. Sep 14 22:21:31 so the API is not returning the correct path in the filesystem Sep 14 22:21:42 rking: only a little bit. my impression is that they do a reasonable job of handling the semantics of concurrency within a single program, but i don't really work on single programs; i work on *systems* of programs, with very different sorts of concurrency and interoperation behaviors Sep 14 22:21:54 or, the returned path, is relative to the sdcard? Sep 14 22:22:03 rking: the actual algorithm involves a simple element-by-element multiplication, then a fourier transform which is generally a single function call Sep 14 22:22:14 rking: but it does nothing to explain what you're actually calculating Sep 14 22:22:14 MrQwak: are you calling the wrong method to find the location on the sdcard? Sep 14 22:22:36 MrQwak: for the sdcard you need to use the "External" variants Sep 14 22:22:55 thanks ctate Sep 14 22:22:55 rking: and, more importantly, how Sep 14 22:23:11 it's just user data (player profile, options etc) for a game Sep 14 22:23:14 or rather why, i guess i should say Sep 14 22:23:16 where is the best place to save it? Sep 14 22:23:27 zinx: An easy way is to write a version that works but is slow, and then later move that to your Unit Tests. Those tests will then 1) compare slowImplementation() vs. speedyImplementation() for correctness, and 2) benchmark both. At that point, someone reading the main codebase will see the speedyImplementation() method and have the whole story as well as a 2nd algorithm that does the same effect. Sep 14 22:23:43 rking: but they're going to be different because of floating point variances Sep 14 22:23:53 zinx: =| Sep 14 22:23:54 have been using getFilesDir() Sep 14 22:24:10 Context.getFilesDir() gives you your app-private data directory Sep 14 22:24:15 (as you discovered :) ) Sep 14 22:24:51 that's not on the sdcard? Sep 14 22:24:52 rking: (you can even get different results running the same algorithm if the fft library you're using decides to use a slightly different algorithm to calculate the fft) Sep 14 22:24:57 MrQwak: correct Sep 14 22:25:01 ah Sep 14 22:25:04 because sdcards are world-readable Sep 14 22:25:09 zinx, p_l|backup, so if Ive got this line in iomem "30000000-34ffffff : System RAM", this should be correct?: adb shell su -c "dd bs=4096 skip=196608 count=217088 if=dev/mem of=mnt/emmc/0.mem" Sep 14 22:25:12 where would that be in the file system? Sep 14 22:25:12 rking: this happens a lot in octave for example, which creates a new fftw plan for each fft :P Sep 14 22:25:14 zinx: http://kentbeck.github.com/junit/javadoc/latest/org/junit/Assert.html#assertEquals(double,%20double,%20double) Sep 14 22:25:21 MrQwak: it's on the internal flash disk somewhere Sep 14 22:25:30 is there a path to it? Sep 14 22:25:38 yes, that's what getFilesDir() gives you Sep 14 22:25:48 oh, actually not Sep 14 22:25:53 ah Sep 14 22:25:56 other apps cannot read/write it, including adb Sep 14 22:26:05 i can't browse to it though? Sep 14 22:26:07 rking: now, consider something else Sep 14 22:26:11 zinx: There isn't even an assertEquals() for doubles that *doesn't* include a Δ. Don't let these trivial speedbumps, like how do I test floats, be the reason why you don't follow a superior process. Sep 14 22:26:12 correct; you cannot browse it Sep 14 22:26:13 is it only accessible from my app? Sep 14 22:26:15 right Sep 14 22:26:26 rking: to people knowledgable in the field, the mechanism used is likely to be obvious and understood Sep 14 22:26:29 it bing private, kinda makes sense! Sep 14 22:26:32 :) Sep 14 22:26:34 :) Sep 14 22:26:36 being* Sep 14 22:26:43 rking: do i really want to spend extra time implementing the more complicated version? Sep 14 22:26:51 rking: rather than simply adding a comment Sep 14 22:26:57 okay, that's great, thanks, you've been a great help :) Sep 14 22:27:23 rking: fwiw Sep 14 22:27:36 zinx: Wait, you're suggesting leaving the simple version and not writing the advanced one? Sep 14 22:27:36 rking: the more complicated version will not be at all obvious to someone not experienced in the field either Sep 14 22:27:45 rking: because it still uses fourier domain for computing derivatives Sep 14 22:27:46 so the files i was saving, were actually there after all Sep 14 22:27:52 rking: "advanced"? Sep 14 22:27:53 I'm honestly not seeing the value here of avoiding comments for the sake of avoiding comments. Sep 14 22:27:56 rking: there's nothing more advanced about it Sep 14 22:28:05 rking: in fact, the opposite is true, it's much less refined Sep 14 22:28:06 is there no way i can get to see them? (other than from my own app) Sep 14 22:28:10 zinx: Complication that buys speed isn't being "advanced"? Sep 14 22:28:25 rking: the method i described is simpler, not more complex Sep 14 22:28:36 like some super-all-privilages file browser? Sep 14 22:29:32 zinx: I just fail to see what about this can't be easily added to a method name and backed strongly by the Unit Tests. Sep 14 22:29:39 rking: the more complex version offers nothing, uses more memory, and is no more clear Sep 14 22:30:01 rking: I fail to see why you should avoid comments here for the sake of avoiding comments. Sep 14 22:30:26 Hodapp: Attempting to avoid them pushes the knowledge deeper into the code. Sep 14 22:30:38 rking: i don't know why you would write more code that's more complex that isn't any more clear, simply to avoid writing a comment. Sep 14 22:31:28 rking: note the point that the more obvious version is not going to be clear to anyone not in the field. Sep 14 22:31:31 zinx: I'm assuming there will be tests, for other reasons (mostly regression testing and teasing out a better design). After you have them for those reasons, it happens that they also document the code. Sep 14 22:31:31 Job Security? Sep 14 22:31:48 rking: those tests will use the simpler version - there's no reason for them not to Sep 14 22:31:55 rking: The claim that any knowledge that a comment can contain, can be contained simply in the code, is begging the question. Sep 14 22:32:17 "Real Programmers don't comment their code. It was hard to write, it should be hard to read" Sep 14 22:32:18 rking: The claim that knowledge 'deeper in the code' is preferable is, too, begging the question. Sep 14 22:32:34 zinx, if the area is 30000000-34ffffff, am I allowed to grab 30000000-0x35000000 ? with bs=4096 the cound would be 20 479.9998 otherwise... Sep 14 22:32:41 count Sep 14 22:32:48 how does one create a search/filter widget, which refreshed the results on each key pressed (just like in contacts app) - is this possible using the standart approach? Sep 14 22:32:52 * rking should've known better to enter in a code æsthetics discussion with a bunch of Java programmers. =) Sep 14 22:32:56 tavelram: only 30000000-34ffffff, 0x35000000... are out of bounds Sep 14 22:33:16 tavelram: the memory may or may not be ok to access - if you access it, do know that it can crash the device Sep 14 22:33:32 rking: i'm not a java programmer, i'm a C programmer Sep 14 22:33:37 Hehe Sep 14 22:33:38 rking: I consider Java a necessary evil of Android programming. I avoid using it otherwise. Sep 14 22:33:50 rking: I mostly prefer Scala, Clojure, C, and Python. Sep 14 22:33:59 it seems to me that to argue against writing comments, one has to argue that of course everyone is competent to write code that is clearly comprehensible to all future readers & users thereof Sep 14 22:34:05 which i think is perhaps an unwarranted claim :) Sep 14 22:34:20 i think it's highly unrealistic Sep 14 22:34:39 i gave an example where one would need specific domain knowledge to understand any version of the code without comments Sep 14 22:34:51 i am making an app, using the google maps api, that provides the user with data based on where they are zoomed in. the data is fetched whenever the user moves to a different geographical location. should i keep an open connection between each client and the server while the client is running the app, or should i open up a new connection each time the client needs new data and close it as soon as all the new data has been sent? Sep 14 22:34:54 i mean, i agree that people shouldn't go mucking with code that htey don't understand Sep 14 22:35:01 Hodapp: There's a difference between begging a question and throwing a challenge out there to see what happens if it's lived up to. Sep 14 22:35:17 but i don't agree the argument that it is *bad* to provide additional information to meak the code easier to learn & understand Sep 14 22:35:28 rking: There's a difference between phrasing a challenge as a challenge and phrasing a challenge like it's already settled fact. Sep 14 22:35:28 Hodapp: A decade ago I ran across some dudes that were into writing expressive code instead of verbose comments. I tried to be like them. I don't regret it. Sep 14 22:35:29 and where the obvious solution is more complex than the non-obvious one, so it is a burden to implement the more complex one, not even taking in to account that the more complex version is no more clear to someone not familiar with the specific domain Sep 14 22:35:52 Hodapp: I'm still waiting to see a URL of some code that has a meritorious comment. Sep 14 22:35:59 (also, i totally get that what and how to comment is its own skillset) Sep 14 22:36:09 * zinx watches rking ignore explained example Sep 14 22:36:26 rking: What is the benchmark for a meritorious comment? Sep 14 22:36:33 zinx: I need to see code, not just a description of a vague thing. Sep 14 22:36:37 anyway, i described a comment i wrote recently that i thought was a reasonable example of useful information for future readers of the code but is not directly expressible *in* hte code Sep 14 22:36:46 Hodapp: One that I cannot more profitably integrate into the system itself. Sep 14 22:36:48 rking: no, actually -- that is exactly my point Sep 14 22:36:56 rking: it wasn't terribly vague, and i can't link you any code to this Sep 14 22:37:05 that the code itself *is not expressive* of the material covered by the comment Sep 14 22:37:07 rking: what i told you want highly specific Sep 14 22:37:10 erh, was Sep 14 22:37:13 rking: How does one measure profitability of said integration? Sep 14 22:37:16 Hodapp: BTW I *do* write TODO/XXX/FIXME comments. Though even those, I'm starting to write into more mechanical places (such as a "pending" unit test) Sep 14 22:37:20 rking: that you think it was vague explains the need for a comment, which i never gave Sep 14 22:37:48 [speaking of my example at @ 20 minutes ago] Sep 14 22:37:57 Hodapp: It varies, because this isn't a solitary case. Sep 14 22:38:10 ANDROID SDK VS NDK? Some advice Sep 14 22:38:10 zinx, oh, actually, Im calculating it wrong. The block actually has 0x34ffffff-0x30000000 +1 bytes, right? Sep 14 22:38:20 tavelram: no Sep 14 22:38:26 tavelram: erh yes Sep 14 22:38:36 phew :) Sep 14 22:38:51 Hodapp: One simple example of where it matters is if you look at a backtrace and can see a symbol name that gives you the clues about what happened, rather than have to dig through source to find that same info. Sep 14 22:39:01 tavelram: [0, 9] is 10 bytes, that is the form it's giving you Sep 14 22:39:09 yeah :) Sep 14 22:39:09 zinx: If this is a recurring problem you should be able to find some source somewhere that demos it. Sep 14 22:39:17 tavelram: i.e., that both memory addresses are in the range Sep 14 22:39:26 yeah Sep 14 22:39:44 rking: phase derivative not something that's commonly used in open source programs Sep 14 22:39:55 rking: and the ones it's used in, it's unlikely to have any comments *OR* unit tests Sep 14 22:40:52 zinx: Well, it's just so hand-wavey. You're like, "I'm over here in this special domain, and there's this case where you can't understand it unless you knew more. But I'm right about it." Sep 14 22:41:03 zinx, and it's /dev/mem, right? Sep 14 22:41:08 zinx: Even if it's 100% true it's a terrible example to use to communicate. Sep 14 22:41:13 rking: well, i suppose so Sep 14 22:41:29 rking: the point was not about the specific example, though Sep 14 22:41:50 rking: i was using it because if you already know the domain in question, you're going to assume it's clear and thus shouldn't need a comment Sep 14 22:42:02 zinx: I'm just saying if there's validity to the theory that there's information whose best home is in comments, there should be a way to find one example of such on some Github repo, somewhere. Sep 14 22:42:13 rking: Do you feel that your own case, where you cannot define 'profitability' of one form over another, has sufficiently moved beyond hand-wavey? Sep 14 22:42:19 for write some process is more beter use native library or SDK? Sep 14 22:42:32 rking: as state, code explains the what you're doing, but not the why you're doing it Sep 14 22:42:36 erh, stated Sep 14 22:42:47 rking: sometimes it's pretty useful to know the why. Sep 14 22:43:18 Hodapp: Where do you decide that I "cannot define it". I gave a very clear example. If I embed my intent in the names of my methods, you can see them in backtraces. If I give them crappy names and make giant methods, you can't tell what happened until you go back to the pre-compiled code and read. Sep 14 22:43:26 rking: and it's either less clear or too cumbersome to describe the why in code a good amount of the time Sep 14 22:43:31 rking: Examples are not definitions. Sep 14 22:43:56 zinx: That's totally not my experience. I've found it very easy to put the "why" in, as I advance in skill at writing communicative code. Sep 14 22:44:08 rking: give an example of that :P Sep 14 22:44:24 rking: i suspect your code is not as obvious as you think it is to the uninitiated Sep 14 22:44:28 Hodapp: Because I say there are many reasons is no justification to assume I cannot define any of the reasons. Sep 14 22:44:54 rking: Oh, it quite is :) Sep 14 22:45:06 Hodapp: The one example I mentioned above is that there is a longer lifecycle of information that lives in the code's symbols, and more durable information is almost always better than more transient information. Sep 14 22:45:09 rking: And you've explained quite well why without me needing to add any further comment. Sep 14 22:45:17 (oh, and there's also the "what was i thinking 2 years ago when i wrote this?" factor to be considered. :) ) Sep 14 22:45:28 haha yeah Sep 14 22:45:36 i've some code quite like that Sep 14 22:46:27 rking: i have to be honest with you, too Sep 14 22:46:37 rking: if you are writing unit tests without comments on them, i think you are insane Sep 14 22:46:56 rking: as stated, the code describes what you are doing Sep 14 22:47:03 rking: what if that doesn't match what you want to do? Sep 14 22:47:12 rking: the only way to know what you want to do is via comments Sep 14 22:47:16 rking: comments lets you distinguish bugs from intentional edge case behaviors when refactoring Sep 14 22:47:26 Hodapp: "Oh, it quite is" ← "it" == __? Sep 14 22:47:37 rking: if the code doesn't match the comments, then THERE IS A PROBLEM. and that's a GOOD thing. Sep 14 22:47:52 rking: without the comment, the code could have a problem no one knows about. Sep 14 22:47:52 rking: https://github.com/android/platform_frameworks_base/blob/master/core/java/android/webkit/WebViewInputDispatcher.java Sep 14 22:48:05 zinx: 90% of the time, when that happens, the code is right and the comment is wrong. Sep 14 22:48:05 rking: comments make that very complex system understandable Sep 14 22:48:07 rking: just as, without the code, a comment could have a problem no one knows about. Sep 14 22:48:13 rking: it doesn't matter Sep 14 22:48:14 jreck: Thanks. Sep 14 22:48:23 rking: if they don't match, there's a problem, and one of them needs fixed Sep 14 22:48:29 thanks again for you help chaps Sep 14 22:48:31 night all Sep 14 22:48:37 rking: I'm capable of telling you what it is, but I don't need to. Sep 14 22:48:43 zinx: Yeah, you're right. The redundant-and-wrong comment needs to be deleted. Sep 14 22:48:50 rking: *sigh* Sep 14 22:49:01 rking: i hope i never have to work on any of your code :) Sep 14 22:49:36 rking: for mine, you can probably just filter out the comments :P Sep 14 22:50:02 well, i figure rking's code is probably rather good Sep 14 22:50:10 (though i'll admit to not putting enough effort in to maintainability in one-off projects) Sep 14 22:50:13 for its intended audience :) Sep 14 22:50:27 i mean, that's how the linux kernel folks get away with it Sep 14 22:50:31 hehe yeah Sep 14 22:50:42 they're explicitly excluding people who aren't yet competent to grok the whole thing Sep 14 22:50:59 Hodapp: If someone asked you why you love someone, and then you said, "Many reasons. For example ___.", how would it be a logical response to say, "Bzzzt. You didn't define it. And you said 'Many reasons' so that's reasonable for me to conclude that you cannot define any of them." Sep 14 22:51:15 and evne then, if you read between the lines on LKML, there's no inconsiderable amount of "wait this is based on untrue premises" that crops up Sep 14 22:51:27 post hoc Sep 14 22:52:56 ctate: Can you rephrase that last bit for me? I don't want to miss your point. Sep 14 22:53:24 ctate: I guess I just asked for a comment. =) Sep 14 22:53:53 ctate: having played muds recently, I fear that kernel code has nothing on them.. ;) Sep 14 22:54:09 rking: It's not necessary for me to give any reasons. They're just comments, after all. Sep 14 22:54:15 if you read between the lines on LKML, there seems to me to be a surprising amount of bugfinding and algorithmic discussion that boils down to the intent or even the deep behavior of the code *not* having been evident to the people working on it in succession, and people discovering this later on Sep 14 22:54:23 jreck: Lines 151, 160, 168, 180, etc., all 100% useless. Sep 14 22:54:25 (by way of rephrase) Sep 14 22:54:38 ctate: does this include the parts where Linus curb-stomps somoene? Sep 14 22:54:43 Hodapp: no Sep 14 22:54:50 rking: Agreed, but those aren't the important comments anyway Sep 14 22:55:04 ctate: and sometimes features just lie there for a looong time before being brought to light for different purposes (like per-process namespaces) Sep 14 22:56:00 rking: the huge comment at the top that describes the system? Invaluable in understanding it. The comment on line 860, for example? Super important code that would otherwise look like a bug Sep 14 22:56:04 Hodapp: I wasn't even talking about your reasons. (We're getting rather meta, here. I can hang if you want to, just noting.) I was referring to your criticism of my description of "better" housing of information, whwerein you said I didn't define it therefore I couldn't. Sep 14 22:56:33 rking: I'm not sure why you're asking for any further comment here. Comments, clearly, are simply not necessary. All of the relevant code has been given. Sep 14 22:56:40 Is any1 familiar with the nexus 7 toolkit here? Is it possible that the partition table for the 16gb device gets written to an 8gb device or would give this an error? Sep 14 22:57:08 ctate: I see. Though I would contend that for any one of those pieces of info, they could have been written clearer by the author, either in a comment or in code. If they could be written in code, I'd like to see why that's not superior in any given case over the comment equivalent. Sep 14 22:58:13 rking: re linux kernel? we're talking about the group of experts who are probably the world's top proponents of exactly the comment (or not) philosophy that you're espousing. Sep 14 22:58:56 jreck: BTW that's a TODO comment, which I advocate because it's a step toward a better system. Once they implement the deltion of the "WebViewClassic"s version of that code, the comment is obsolete. Sep 14 22:59:49 rking: The last part of that comment isn't for the TODO, it's the "we have two complex systems that this is sitting between with subtly different expectations for the same event" Sep 14 23:00:26 ctate: I don't think they have quite the same intent as I do. I believe the information should be there. Whenever I think I need a comment, I write one, but then I step back and look for a place to put that information in the source itself. Sep 14 23:01:39 jreck: So, if it's valuable functionality, it's encoded in a of Unit Tests, one for each of the variants of the condition. Sep 14 23:01:56 Or else it's floating without automated tests, which voids the rking warranty completely. Sep 14 23:03:01 rking: perfect unit tests and nothing but understandable, flawless code? Now you're arguing for an impossible idealism Sep 14 23:03:11 I'll be over here in the real world, trudging along with comments Sep 14 23:04:06 Hodapp: You're actually doing what I advocate against, which is withholding information. I'm saying don't do it in programming, but don't stop half way and leave it in those corpses: comments. Sep 14 23:04:15 Also, do you have unit tests that make sure your unit tests are working correctly and cover all input conditions? Sep 14 23:04:32 ctate: I hope Google helps the x-plane guys against that patent troll suing them for using the LVL Sep 14 23:04:33 jreck: That's called a coverage tool. Sep 14 23:05:10 absolute trust in tools will simply lead toward more subtle, hard to find bugs Sep 14 23:05:11 reminds me of Lodsys Sep 14 23:05:24 g00s: I'm tempted to see if one can organize some black PR campaign against patent trolls in their hometowns Sep 14 23:05:25 jreck: But as a pragmatic matter I agree, some code doesn't get written with the tests it should have. But the same thing is true of that line 860 comment — a lot of code is written with similarly subtle behavior that doesn't so much as get the comment. Sep 14 23:05:33 g00s: something about stealing jobs especially Sep 14 23:05:53 p_l|backup: :) Sep 14 23:06:04 The biggest predictor of code defect rate, it is worth noting, is the number of lines of executable code. Sep 14 23:06:15 jreck: My advocacy is simply that 1) that information should be in the system, and 2) in a comment is teetering on the edge of "not in the system". Sep 14 23:07:13 Hodapp: To make that statement true requires a host of constraints. Sep 14 23:07:24 rking: Empirically, it is true. Sep 14 23:07:30 g00s: imagine if their neighbours would start giving them an evil eye for their "work" ;) Sep 14 23:08:02 Hodapp: For example, it's definitely not referring to code in the tests. Sep 14 23:08:47 hi all Sep 14 23:08:49 Hodapp: also, the # of comments predicts defect rates :) Sep 14 23:08:58 Hey there. Got a quick question.... Sep 14 23:09:19 quick question Sep 14 23:09:24 Hodapp: Another 2 things it constrains is that it must refer to the *same* programmer, and also with a similar level of knowledge of the languages the code is written in. Sep 14 23:09:29 can android app connect to MS Sql SErver? Sep 14 23:09:35 g00s: Is that true? Sep 14 23:09:47 phong_: if you get apropriate client code, yes Sep 14 23:09:53 How woud I go about implementing the download button/progress wheel that Google has in their play apps? The pin icon/button specifically? Sep 14 23:09:54 what do you meant? Sep 14 23:09:57 rking: i thought i read that in uncle bob's book, can't citate original source though Sep 14 23:10:13 jreck: Lines 1084 and 1085 are just too cute. Sep 14 23:10:20 phong_: if there's some pure-java library to access SQL Server, or if you use some intermediary Sep 14 23:10:34 g00s: Interesting. His claim is that more comments ⇒ more defects? Sep 14 23:11:11 rking: lemme find the real source Sep 14 23:11:24 g00s: Please do. I appreciate it. =) Sep 14 23:11:28 p_l|backup, does this help ? import java.sql.*; Sep 14 23:12:04 phong_: no. There's no connector included. You can load in JDBC, but you'll still need pure-java JDBC connector for SQL Server Sep 14 23:12:14 there's none included Sep 14 23:12:15 rking: Actually, no, that constraint is not present either. Sep 14 23:12:23 rking: Apparently you're not aware of what "empirically" means. Sep 14 23:12:41 rking: I suppose I should have added a comment in there about it. Sep 14 23:12:52 Hodapp: You're quite the fountain of personal attacks in this conversation. Sep 14 23:12:52 p_l|backup, will this help Sep 14 23:12:52 Class.forName("net.sourceforge.jtds.jdbc.Driver"); Connection conn = DriverManager.getConnection( db_connect_string, db_userid, db_password); Sep 14 23:13:14 rking: Not really, no. But if you need to make that assertion to feel better about the high horse you're on, go ahead. Sep 14 23:13:19 jreck: Something about this big comment at the top: It was done in a single commit. It has not been maintained at all. Sep 14 23:13:21 phong_: ... you're not getting what I'm saying Sep 14 23:13:38 p_l|backup, i'm new , i never program android before Sep 14 23:13:42 i wanna learn ;) Sep 14 23:14:14 phong_: you'll have to a) first include in your app the JDBC framework (as I don't think Android includes it), then include the jtds driver (assuming it's pure java) and then you can program with JDBC Sep 14 23:14:38 p_l|backup, can you show me how to setup android programming environment? Sep 14 23:14:42 i'm not sure what i need to learn Sep 14 23:14:48 Hodapp: I know exactly what emperical means, and I meant what I said. The metric that more "lines of" code == more defects would require numerous assumptions to be measurably, demonstrably true. Sep 14 23:15:01 phong_: http://d.android.com is a good start Sep 14 23:15:16 rking: Apparently you don't know what 'exactly' means either. Sep 14 23:15:27 especially read the articles about application fundamentals, activity/service lifecycle etc. Sep 14 23:15:38 phong_: Unlocking Android was pretty good book, too Sep 14 23:15:49 though it might be a bit dated now Sep 14 23:15:50 ok how about any android book for noob? Sep 14 23:15:55 the book that show step by step Sep 14 23:16:17 phong_: If you know a bit of java, Unlocking Android was good start IMHO Sep 14 23:16:26 something like pre dummy - for - android Sep 14 23:16:44 'pre dummy' ? Sep 14 23:16:50 like pre "android for dummy' Sep 14 23:17:20 i want to knwo what i need to setup step by step Sep 14 23:17:24 a book to show that? Sep 14 23:17:52 rking: that file has been maintained, and the maintainer/submitter of that file (me) is not the one that wrote it Sep 14 23:18:09 phong_: http://d.android.com has *very* good tutorials, IMHO. And "Unlocking Android" was pretty easy to understand, even if it doesn't cover latest stuff Sep 14 23:18:56 rking: that comment also documents knowledge that was passed down through the years Sep 14 23:19:58 Hodapp: I'm glad I had good interactions with you in the past. Otherwise I'd have you chalked up as a doofus at this point. =) Sep 14 23:19:59 p_l|backup: i think their is a meap for unlocking android 3e Sep 14 23:20:17 *there :/ Sep 14 23:20:30 meap? Sep 14 23:20:39 'early edition' Sep 14 23:20:45 early access or something Sep 14 23:20:47 * p_l|backup used first edition Sep 14 23:20:54 good Sep 14 23:20:58 oh , heh. i have 2e, they are working on 3e Sep 14 23:21:05 rking: You'll find that people might pay a little more regard for your viewpoint if you get into the habit of showing a little bit of humility. Sep 14 23:21:06 1e was better than the others, one of the authors left Sep 14 23:21:10 the approach in 1e was pretty good, IMHO, including some of the rarer stuff Sep 14 23:21:29 now I'm reading Effective Java and finding out all the juicy stuff I missed Sep 14 23:21:37 Hodapp: Aha, and you're the example of humility in this thread? Sep 14 23:22:41 jreck: I mean it hasn't had the act of maintenance performed on it. No one has modified the comment as it stands. Perhaps this is because it is static knowledge with no need of updates, but seeing as it wasn't written too long ago I doubt that will be true. Sep 14 23:23:01 jreck: It's begging to go out of date, I think. Just one little tweak to the actual code, and it's invalidated. Sep 14 23:23:01 rking: "I know you are but what am I," is not a valid dismissal of what I said. Sep 14 23:23:48 Hodapp: How do you know I didn't take it to heart? Sep 14 23:23:58 p_l|backup: How would I go about implementing the download button/progress wheel that Google has in their Play apps? the pin icon/button specifically? Sep 14 23:24:16 p_l|backup: Any thoughts? :/ Sep 14 23:26:19 rynojvr: AsyncTask. Sep 14 23:27:15 hwrd|work: I was more specifically talking about how it shows the progress; it fills up blue as the download progresses. Sep 14 23:27:56 ...? Sep 14 23:28:10 rynojvr: android.widget.ProgressBar Sep 14 23:28:59 well, the current one I'm not sure. Some fiddling with how you generate the fill pattern Sep 14 23:30:26 I don't think it uses native progress bar Sep 14 23:30:31 hwrd|work: http://developer.android.com/design/media/progress_activity_custom.png Sep 14 23:31:01 rynojvr: oh, you want that specific type of bar? Sep 14 23:31:21 roman nurik has code for that Sep 14 23:31:24 hwrd|work: basically. Sep 14 23:31:33 yeah what f2prateek said. Sep 14 23:31:46 Also, I love Roman Nurik. He writes awesome shit. Sep 14 23:31:50 platform_frameworks_base? Sep 14 23:32:18 no Sep 14 23:32:24 check his google+ stream Sep 14 23:32:30 or wait a sec Sep 14 23:32:34 http://code.google.com/p/romannurik-code/source/browse/#git%2Fmisc%2Fpinprogress Sep 14 23:32:37 there you go Sep 14 23:32:41 ah. That was the only thing of his I could find on github.com Sep 14 23:33:04 The only thing I don't like is that he uses google code Sep 14 23:33:38 well some of them still use git so cloning is fairly easy Sep 14 23:34:37 p_l|backup, zinx, thanks for all the help! I grepped in the datadata and mem dumps and found alot of sms message data :) Sep 14 23:36:36 f2prateek: Thanks. This looks like it could work. I'll give it a shot. Sep 14 23:36:45 hwrd|work: thanks as well. Sep 14 23:36:49 ugh, i can not find this comment / defects correlation Sep 14 23:36:57 rynojvr: It is no problem, Sep 14 23:36:59 *. Sep 14 23:37:09 I have a dialog in a fragment that is going to be quite complicated. Is there a way to launch this dialog as its own activity thus cleaning up the code? Sep 14 23:37:48 sometimes, google search still sucks balls. keep pulling up blogs about defect density with "comments" section Sep 14 23:40:14 patrickod: write it as an Activity using the dialog window theme Sep 14 23:40:37 much, much better control of fancy behaviors etc Sep 14 23:40:44 ok. Sep 14 23:40:54 so just add in in the manifest file. Sep 14 23:41:03 how do I launch it from the other activity ? Sep 14 23:41:12 with startActivity() ? Sep 14 23:41:21 if you need results, startActivityForResult() :) Sep 14 23:42:47 and we all want results ! Sep 14 23:46:04 what's the lowest API version my app should support? Sep 14 23:46:26 whomp: whichever you can stomach supporting Sep 14 23:46:44 1.6 is afaik the limit for applications that want to support newer stuff, though Sep 14 23:47:00 (due to some changes regarding screen size differences) Sep 14 23:47:00 i'm looking for insight into when i should draw the line Sep 14 23:47:12 whomp: decide based on number of users, I guess Sep 14 23:47:21 for example, if i set it at 14, am i cutting off a lot of people? Sep 14 23:47:24 and whether users with older phones are important to you Sep 14 23:47:29 14? Huge amounts Sep 14 23:47:43 older users are just as important Sep 14 23:48:00 I'd say that 2.2 is pretty safe bet now Sep 14 23:48:23 phones that were last updateable to 2.1 are starting to break down now thanks to RoHS Sep 14 23:48:27 you think at least 3/4 of the phones in use out there are at 2.2 or higher? Sep 14 23:48:58 whomp: the data depends on area. In global terms? yes. Locally? Not necessarily Sep 14 23:49:19 the US is the main market i Sep 14 23:49:25 'm targeting Sep 14 23:49:42 well, I'd say 2.2 is safe bet Sep 14 23:50:09 for 75% of the users? Sep 14 23:50:48 well, check dashboards on d.android.com Sep 14 23:51:17 kk Sep 14 23:51:22 thx for the help :) Sep 14 23:53:12 p_l|backup: lol, RoHS Sep 14 23:53:25 g00s: there was at least one car crash caused by it, even Sep 14 23:54:03 if you are just starting an app now, and plan to be done with it in a few months, i would say minSDK = gingerbread Sep 14 23:54:41 well, depends what you're using Sep 14 23:54:43 its not just about what can be stomached, it also reduces support costs because different versions of android have different bugs, and that generation of phones kinda sucked Sep 14 23:55:00 2012 is the year of GingerBread :D Sep 14 23:55:11 * p_l|backup till recently had only 2.2. Then he gave up on being able to take photos and play games, now he has 2.3 Sep 14 23:55:29 well, i still have only 2.2 … but i would not support it now with a new app Sep 14 23:55:53 also, notice a disproportionate # of buyers with newest phones. i wonder what others' experience is in this area Sep 14 23:56:24 well, some cheap phones bought in last few months have ICS Sep 14 23:57:14 p_l|backup: i was contemplating putting jelly bean on my droid -1 :) Sep 14 23:57:28 g00s: heh. You'll have better luck than I with my Spica Sep 14 23:57:40 since you've PowerVR gpu Sep 15 00:01:36 someone wrote to android-platform@ this week talking about how they're spinning up a project to do a phone based on ICS Sep 15 00:01:49 and i'm all uh.... why ICS and not JB? Sep 15 00:01:50 srsly? Sep 15 00:03:12 ctate: simple answer: Key Lime Pie isn't out yet Sep 15 00:04:48 * SpeedEvil ponders next version? Sep 15 00:05:01 lemon drizzle cake? Sep 15 00:05:25 ctate: i did hear some phone are skipping ICS update and going straight to JB though Sep 15 00:05:54 nmdjdjdjdjdjdkdkdkdkdkdkdkdkdkdkdkdkkdkdkdkdkdkdkdkkdkdjdkdjdkdkkdkdjdjdnncncnchhjddndjdjdjddjjdjdjjdjjiifnnvid. JD JJ Sep 15 00:05:57 oops Sep 15 00:09:40 g00s: this was for a *new* phone Sep 15 00:14:55 ctate: I have the dialog launching in a new intent now but it's also showing a default empty dialog underneath the one I create and show in onCreate Sep 15 00:15:04 any ideas as to how I can get rid of the default one? Sep 15 00:15:21 don't create and show it Sep 15 00:15:32 ? Sep 15 00:16:17 I'm not explicitly creating it? Sep 15 00:17:01 I'm creating my own dialog but there's another empty one beneath it. I have to back out of 2 dialogs to exit the activity. Sep 15 00:17:09 sorry for the confusion, this is my first android app Sep 15 00:18:48 it's your code that is creating and showing an empty dialog, isn't it? Sep 15 00:18:52 somewhere Sep 15 00:19:15 I'm setting the content of my dialog and it's showing correctly. I'm only creating the one Sep 15 00:19:27 I can paste code and PM you with a URL if you want? Sep 15 00:19:41 not via PM please Sep 15 00:19:54 you were talking about having a dialog Sep 15 00:19:56 https://gist.github.com/27b66f4f1c86cd01a6db Sep 15 00:20:02 and i suggested that you use an activity *instead* Sep 15 00:20:43 okay, this is an Activity with no UI, which also displays a dialog on top of itself Sep 15 00:21:07 ah so it's filling in a default UI Sep 15 00:21:13 and then I'm displaying a dialog on top of that. Sep 15 00:21:26 well, you've got an activity running. it has to draw *something* Sep 15 00:21:34 so by default it gets a blank window Sep 15 00:21:38 right Sep 15 00:21:54 what i was suggesting was that you not show a Dialog *at all* Sep 15 00:22:15 that you instead write an Activity that uses the same layotu theme as dialogs, and implement your logic in that activity Sep 15 00:22:23 ah ok. Sep 15 00:22:33 * patrickod got confused, thought I was just moving the dialog call Sep 15 00:22:37 because you said you wanted the dialog to do something more sophisticated than "press ok or cancel" Sep 15 00:22:52 yeah, it was an "instead of" recommendation Sep 15 00:22:57 okay we're all sorted now :) Sep 15 00:23:29 thanks Sep 15 00:23:31 :) Sep 15 00:25:27 how can i download the sdk for android version 2.2? in eclipse, when i go Window -> Android SDK Manager, i see a bunch of groups, each labeled "API ". however, when i try to install one of them, it tells me 'Missing SDK Platform Android, API '. Sep 15 00:26:53 i'm adding a number of views, and i'm sporadically receiving a "child already has parent" error Sep 15 00:27:42 well, android 2.2 is api level 8, so that's the one to ask for Sep 15 00:27:48 thing is, the parents that these views should reside in, I've already called removeAllViews() on Sep 15 00:27:56 how do i "ask" for this sdk Sep 15 00:28:01 ctate: would it make sense in this context to use a dialogfragment instead? Sep 15 00:28:02 although the question is why you're developing with SDK 8 instead of just targeting 8 in yoru manifest Sep 15 00:28:14 or better, using minSdk 8 and targeting the most recent Sep 15 00:28:32 patrickod: i'm actually not familiar with using fragments Sep 15 00:28:38 not mostly a UI guy Sep 15 00:28:40 ctate: link to SimonVT's blog post about that, that should help him see why he should do what you're saying. Sep 15 00:28:43 maybe i'm going about this the wrong way. i want to make an android app that works on 2.2 and up Sep 15 00:28:49 ctate: no worries. Sep 15 00:29:02 whomp: right. you want minsdk = 8, and build against the *current* sdk Sep 15 00:29:05 targetting the most recent Sep 15 00:29:06 patrickod: what are you asking? i used dialogfragments Sep 15 00:29:06 all i need to do is change the version number in my manifest file and run the 2.2 avd? Sep 15 00:29:11 right Sep 15 00:29:30 whomp: http://simonvt.net/2012/02/07/what-api-level-should-i-target/ Sep 15 00:29:30 HorizonXP: I have a dialog that will have quite a bit of code in it and wanted to move it out of the activity that's calling it Sep 15 00:29:32 buidling against hte current SDK is important so that your app looks & acts correct on newer versions of the OS Sep 15 00:29:40 it was suggested that I should use a separate activity Sep 15 00:29:47 but isn't this the use case for fragments? Sep 15 00:29:48 patrickod: dialogfragment works just as well Sep 15 00:29:53 patrickod: yup Sep 15 00:29:59 that's what i did Sep 15 00:30:00 great. thanks Sep 15 00:30:01 they're great Sep 15 00:30:30 * HorizonXP never thought he'd see the day he'd be helping others in #android-dev Sep 15 00:30:45 aw, it's like his first day of school Sep 15 00:30:56 ctate, where would i find a 2.2 avd? Sep 15 00:30:59 little HorizonXP is all growed up now Sep 15 00:31:13 ctate: lol Sep 15 00:31:18 ctate: i big boy now! Sep 15 00:31:27 err, don't you just install it in your sdk and t hen create the avd? Sep 15 00:31:35 * ctate is also not familiar with the IDE integration tools.... Sep 15 00:32:09 whomp: ctate is right. just install the SDK then create the AVD with that version. Sep 15 00:32:15 i guess so? right now it only gives me android version 14 and 15 Sep 15 00:32:31 whomp: so download the other SDK versions you need. Sep 15 00:32:54 ok so i do need the sdk lol, bringing me back to my original question: how do i get ahold of a specific sdk? in my case, i need sdk 2.2 Sep 15 00:34:29 whomp: SDK manager? Sep 15 00:34:40 whomp: should let you download it, right? Sep 15 00:34:57 HorizonXP: it doesn't. Sep 15 00:35:42 whomp: ok, so Eclipse -> Window -> Android SDK Manager Sep 15 00:35:49 in the list, scroll to Android 2.2 Sep 15 00:35:58 check SDK Platform under it, and install. Sep 15 00:36:02 it's not there Sep 15 00:36:09 it simply doesn't show up Sep 15 00:39:50 rking: (sorry for late response, was AFK) That comment is still accurate. Changes have been made to the code that have no affect how the pieces fit together, which is what the comment is describing. The thread flow, interested party, and data is all the same Sep 15 00:40:33 whomp: do you have anything listed in there? Sep 15 00:41:03 whomp: you need to be more forthcoming with details. I can't spoonfeed everything for you. Google is your friend here too, you're not likely to be the only one to have encountered this. Sep 15 00:41:08 rking: and the data flow between the threads is the part that's complicated, as that varies from website to website, and can even vary from touch to touch Sep 15 00:42:27 as i already said, it gives a series of folders, each named API Sep 15 00:45:16 So Android 4.1 (API16), Android 4.0.3 (API 15), etc. ? Sep 15 00:45:27 and there's checkboxes beside each right? Sep 15 00:45:34 yes Sep 15 00:45:39 you know that you can click them to expand them right? Sep 15 00:45:45 see my original post concerning this Sep 15 00:45:47 yes Sep 15 00:46:01 what was your original post? Sep 15 00:46:17 "how can i download the sdk for android version 2.2? in eclipse, when i go Window -> Android SDK Manager, i see a bunch of groups, each labeled "API ". however, when i try to install one of them, it tells me 'Missing SDK Platform Android, API '." -whomp Sep 15 00:48:00 and you checked your source URLs? refreshed them? Google says nothing? Sep 15 00:48:17 i'm getting towards the end of my knowledge, because I've never encountered this. Sep 15 00:48:39 i'm going to switch to command line dev anyway Sep 15 00:48:55 thanks anyway Sep 15 01:16:01 hi guys Sep 15 01:16:14 how to install eclipse in ubuntu? Sep 15 01:17:15 phong_: download the eclipse package from their website Sep 15 01:17:22 the ubuntu repos have an outdated version Sep 15 01:17:25 where Sep 15 01:17:30 what website Sep 15 01:17:40 hold on let me login chat using ubuntu Sep 15 01:17:40 http://www.eclipse.org/downloads/ Sep 15 01:17:41 brb Sep 15 01:17:42 patrickod: pretty much all the distros have outdated versions, because of the way eclipse updates now Sep 15 01:18:07 zinx: what's annoying is that on ubuntu both eclipse and java packages are outdated Sep 15 01:18:09 i'm bcack Sep 15 01:18:10 what now Sep 15 01:18:13 patrickod: lulz Sep 15 01:18:17 http://www.eclipse.org/downloads/ Sep 15 01:18:23 patrickod: sure you're not supposed to use java-package? Sep 15 01:18:30 zinx: gentle troll is best :P Sep 15 01:18:47 patrickod, wow so many link Sep 15 01:18:52 just tell me what to download Sep 15 01:18:59 you want the classic version Sep 15 01:19:13 this ? Eclipse Classic 4.2, 182 MB Sep 15 01:19:39 must be Sep 15 01:20:00 ok Sep 15 01:20:04 thanks it's downloading Sep 15 01:20:31 after download it is in .tz file Sep 15 01:20:39 i meant .gz Sep 15 01:20:58 actually tar.gz not sure what do do Sep 15 01:22:57 are you there brother Sep 15 01:23:18 tar xzf $FILENAME Sep 15 01:23:20 phong_: extract it where you want it Sep 15 01:24:06 zinx, lol man Sep 15 01:24:27 phong_: i was being serious Sep 15 01:24:34 phong_: extract it, run ./eclips/eclipse Sep 15 01:24:54 phong_: that's what Eclipse expects you to do... Sep 15 01:26:13 ok Sep 15 01:26:19 it is in my home dir Sep 15 01:26:25 with folder name eclipse Sep 15 01:26:54 phong_: ./eclipse/eclipse Sep 15 01:27:03 it's running Sep 15 01:27:06 of course Sep 15 01:27:35 so evertime i run with ./eclipse Sep 15 01:27:36 cool Sep 15 01:28:01 phong_: http://developer.android.com/sdk/installing/installing-adt.html Sep 15 01:28:02 phong_: yes Sep 15 01:28:29 phong_: you can create a launcher icon or whatever for your WM/desktop environment/etc Sep 15 01:28:49 u know what Sep 15 01:28:51 nevermind Sep 15 01:28:53 i give up Sep 15 01:28:55 ? Sep 15 01:28:57 i dont want to be programmer Sep 15 01:28:59 :) Sep 15 01:28:59 lol Sep 15 01:29:05 lol too much stuff Sep 15 01:29:09 programming, it's not for everyone :) Sep 15 01:29:13 yeah Sep 15 01:29:30 zinx Sep 15 01:29:34 can u do me a favor? Sep 15 01:29:52 do u have repo yourself? Sep 15 01:30:12 yes, but the code i have checked out is likely not what you want Sep 15 01:30:24 and it takes a few hours to check out a new tree, and many gigs of disk space Sep 15 01:30:24 oh Sep 15 01:30:38 i use ubuntu subversion Sep 15 01:30:40 do u use that? Sep 15 01:30:45 not for Android Sep 15 01:30:50 why Sep 15 01:30:52 and for personal stuff i use git now too Sep 15 01:31:01 because all the Android repositories use repo/git :) Sep 15 01:31:06 subversion for all programming Sep 15 01:31:24 and as for personal stuff, i use git because it does stuff i need better than subversion Sep 15 01:31:33 what is git Sep 15 01:31:33 and better than mercurial for that matter Sep 15 01:31:35 how to get git Sep 15 01:31:44 can i get git? Sep 15 01:31:59 use google :) Sep 15 01:32:13 it's available in pretty much every distro, though, prepackaged Sep 15 01:32:21 but why would you google, when you can ask random people on irc!? Sep 15 01:32:23 not a clue Sep 15 01:32:32 oj Jesus Sep 15 01:32:40 I like orange juice. Sep 15 01:32:40 God Bless you Sep 15 01:32:59 * JesusFreke doesn't think he sneezed Sep 15 01:34:45 it's installing new software ->https://dl-ssl.google.com/android/eclipse/ Sep 15 01:35:26 is there emulator for android ? Sep 15 01:35:32 yes Sep 15 01:35:42 i finish install https://dl-ssl.google.com/android/eclipse/ this eclipse Sep 15 01:35:43 next? Sep 15 01:36:52 and i accept all...installing android stuff Sep 15 01:37:01 android sdk manager Sep 15 01:37:34 zinx, what program have you done so far? Sep 15 01:37:54 phong_: for Android? nothing you would have heard of Sep 15 01:38:29 zinx, why not? Sep 15 01:38:38 wow, u're in here and u dont do any android app? Sep 15 01:38:56 phong_: i do mostly lower level stuff, when i do Android stuff Sep 15 01:39:07 what is lower level stuff Sep 15 01:39:13 example what do u meant by that Sep 15 01:39:18 phong_: working on the kernel, framework, drivers, etc Sep 15 01:39:23 oh Sep 15 01:39:24 exploits ;) Sep 15 01:39:40 that meant you're way better then the app programmer Sep 15 01:39:58 zinx, is there to run android emulator in linux? Sep 15 01:40:05 phong_: yes, but it's very slow Sep 15 01:40:14 i have fast pc Sep 15 01:40:18 it emulates an ARM device, which is not too fast on even a fast PC :) Sep 15 01:40:22 16gb, i7 2600k Sep 15 01:40:59 zinx, is there any good thing in linux i need to know? Sep 15 01:41:07 phong_: not really Sep 15 01:41:36 phong_: i will say that if you've had no programming experience, it will be very hard for you to jump right in to Android app development Sep 15 01:41:52 i have web experience Sep 15 01:41:55 asp.net and vb.net Sep 15 01:41:58 phong_: just because there is a lot of stuff to learn at once Sep 15 01:42:01 does that count? Sep 15 01:42:05 phong_: yeah, somewhat Sep 15 01:42:20 phong_: Java is a very different language from those, though Sep 15 01:42:25 i know that Sep 15 01:42:33 more like c++ Sep 15 01:42:35 phong_: and Android applications are a very different kind of application Sep 15 01:42:44 phong_: nah, more like C# Sep 15 01:42:50 so cancel out the install? Sep 15 01:42:54 ? Sep 15 01:43:02 why? Sep 15 01:43:09 you scare me Sep 15 01:43:12 hehe Sep 15 01:43:18 wasn't meant to scare you Sep 15 01:43:26 but i am Sep 15 01:43:38 just saying that you shouldn't get discouraged mostly :) Sep 15 01:43:48 if it seems hard, it's because it really is Sep 15 01:46:09 zinx, android virtual slow shlt Sep 15 01:46:11 u're right Sep 15 01:46:16 yup Sep 15 01:46:16 boot up takes awhile Sep 15 01:46:41 use the x86 emu :) Sep 15 01:46:45 fortunately you don't have to boot it up again unless you close it Sep 15 01:46:54 yeah, or do what ctate said Sep 15 01:46:54 what ctate ? Sep 15 01:47:00 i dont get it Sep 15 01:47:02 how to do it Sep 15 01:47:35 ctate, please explain how to do it Sep 15 01:47:47 or where , how to do what u said Sep 15 01:48:20 hello? Sep 15 01:48:31 what is x86 emu? Sep 15 01:48:35 you meant on pc ? Sep 15 01:48:37 Settle down Sep 15 01:49:40 You download x86 images in the sdk Sep 15 01:50:26 And then you read this: http://developer.android.com/tools/devices/emulator.html#accel-vm Sep 15 01:55:23 ok let me enable it rebbot my vmware Sep 15 01:55:24 brb Sep 15 02:06:41 much bettah Sep 15 02:07:07 so, what's the magic incantation to build for the toro? Sep 15 02:08:38 cjac@calcifer:/usr/src/android$ BUILD=full_toro BUILDTYPE=eng make -j8 Sep 15 02:08:43 look right? Sep 15 02:24:53 Hi all. I am using startActivityForResult to take a picture and then load it into an ImageView on my app. The picture gets stored on the SD card(I checked that), but I'm unable to load it on my ImageView. Anyone had similar issues with this? Sep 15 02:25:27 I've checked the Uri that I'm using to resolve to the Bitmap, and it is correct, just that the Bitmap returned is null. Sep 15 02:27:20 Does your app have permissions to access sdcard? Sep 15 02:27:59 Yes, I have set WRITE_EXTERNAL_STORAGE and READ_EXTERNAL_STORAGE from my Manifest. Sep 15 02:28:39 The READ permission was set before the application definition, but the WRITE was set in the application meta data. Sep 15 02:29:33 I set the read after having write didn't work, I don't think it does anything though. The folder and the files have mode rw on them, so I don't see any permission problems with that Sep 15 02:30:00 When I remove the WRITE_EXTERNAL_STORAGE I get a permission denied error, so I'm probably accessing it :p Sep 15 02:30:10 ok. sounds reasonable Sep 15 02:30:22 but otherwise, no clue about your problem. sorry :) Sep 15 02:30:53 Thanks, JesusFreke :) Sep 15 02:35:56 i'm reading about FireFox OS, sounds pretty interesting. i thought this was originally based on Android, but it is not at all. Standard linux kernel. They are going for teh very low end market (700mhz single core, 256mb) … but i thought to myself, would't an HTML5 app run like crap on such a device ? Sep 15 02:36:21 but then Eich says "“In modern optimisations, Javascript beats Java. There’s been a tonne of innovation in the last few years." Sep 15 02:36:34 wonder if he is right Sep 15 02:37:32 anyhow, fun article http://www.techweekeurope.co.uk/news/firefox-os-html5-mozilla-eich-iphone-5-92412 Sep 15 02:38:28 this was also interesting "Tizen is using the same APIs as Firefox to control the phone hardware" Sep 15 02:44:49 g00s: sounds like a chromebook Sep 15 02:45:11 without the NaCL ;) Sep 15 02:45:27 but I *like* salt! Sep 15 02:47:16 g00s: android is based on standard linux kernel too :P Sep 15 02:47:32 pyjar: :P Sep 15 02:47:37 "HP CEO Meg Whitman says 'we have to ultimately offer a smartphone'" Sep 15 02:47:40 hehe Sep 15 02:47:51 "again" Sep 15 02:48:02 Kindle fire smartphone. "Punch the monkey to answer this phonecall" Sep 15 02:48:10 see also: Dell Sep 15 02:48:23 "The news comes as HP is coming off of a flat third quarter in which the company saw a drop in sales for its PCs and printers." Sep 15 02:48:53 already posted this on #android, but... hater comments are amusing: http:// www.theverge.com/2012/9/14/3335204/google-statement-acer-smartphone-launch-aliyun-android **** ENDING LOGGING AT Sat Sep 15 02:59:59 2012