**** BEGIN LOGGING AT Thu Nov 08 03:00:02 2012 Nov 08 03:00:13 DrPoole: are you sure you aren't caching the registerId in prefs or something ? Nov 08 03:00:38 I'm able to send a message manually to both register_id's Nov 08 03:00:43 to a single device. Nov 08 03:01:19 hm. not sure if that was a proper response to your question. heh. Nov 08 03:02:04 i'd prefer it something simple as it using an invalid register_id (that's old because of cache) but ... the fact that GCM has both of these id's tied to a single device is really hurting my feelings. Nov 08 03:02:21 Is it possible to develop an Android application that only uses the NDK - with no Java code at all? Nov 08 03:08:58 lolcathost: that is like asking..."I can I write a program in C++ without using C++?' Nov 08 03:11:08 KNERD: Why so? If I write, say, a regular GTK+ application, I can do it in whatever language I like - I am not forced to use C. Nov 08 03:12:30 lolcathost: Well using the NKD..yes you can use another languag..sorry..i was not thinking Nov 08 03:13:28 KNERD: Yeah, but the examples I found online always involved calling C code from Java, hence my question. Nov 08 03:13:57 the SDK you must use Java Nov 08 03:14:47 lolcathost: try looking up NativeActivity, I haven't used it, but I believe it's what you're looking for. C++ based Activity Nov 08 03:15:00 how do i implement a ContentProvider that provides a stream? the docs only seem concerned with CP's that provide Cursors. Nov 08 03:15:13 kevinb: Let me see. Thanks! Nov 08 03:20:10 I am getting a strange error! ---> This Android SDK requires Android Developer Toolkit version 20.0.0 or higher..Current version is 18.0.0xxxxxxxx Please Update to current version... No Updates available! Nov 08 03:21:34 KNERD: updates are definitely available.. latest is 20.0.3 Nov 08 03:22:32 tophyr: how about implementhing this Nov 08 03:22:33 I think it is because the Android SDK URL is not included in the updates..I am trying the nVidia Tegra setup Nov 08 03:22:34 http://developer.android.com/reference/android/content/ContentProvider.html#openFile%28android.net.Uri,%20java.lang.String%29 Nov 08 03:22:49 tdignan: looking at that now Nov 08 03:22:55 i can only guess that's how it's implement Nov 08 03:22:56 ed Nov 08 03:23:38 it looks like you can use openFileHelper to do it as well Nov 08 03:23:39 but it doesn't really make sense since ParcelFileDescriptor is truly a FD, it doesn't look like it has any way to take an actual stream Nov 08 03:25:06 openPipeHelper maybe? that what yo umeant? Nov 08 03:25:15 http://developer.android.com/reference/android/os/ParcelFileDescriptor.html#fromSocket%28java.net.Socket%29 Nov 08 03:25:41 I don't know much about parcel file descriptor.. but it seems it can wrap up a lot of basic stuff Nov 08 03:25:59 socket != stream ;) but yes if i really wanted i could hack that together Nov 08 03:26:12 just doesn't seem like the right way.. if it's hacked, it's usually wrong Nov 08 03:26:39 i have the source sitting right here.. maybe i should just go look @ Email or something Nov 08 03:27:19 do you want like a buffered file? like stdio? Nov 08 03:28:03 the provider will download a file over http(s) and stream it to the receiver Nov 08 03:28:30 Is this correct? https://dl-ssl.google.com/android/eclipse/ Nov 08 03:28:42 I am getting an error..."not valid software location" Nov 08 03:29:45 looks like openAssetFile is what i need to implement to get this going Nov 08 03:29:47 Oh, i see Nov 08 03:29:54 http://developer.android.com/reference/android/content/ContentResolver.html#openAssetFileDescriptor(android.net.Uri, java.lang.String) Nov 08 03:30:40 KNERD: that's for the ADT Nov 08 03:30:46 sdk is different Nov 08 03:31:37 tophyr: Visit the URL and it says Android SDK Plugin Nov 08 03:32:14 "To install this plugin, please use the Update mechanism from within Eclipse and provide it with the following URL: " Nov 08 03:32:21 KNERD: exactly Nov 08 03:32:25 that's the Eclipse plugin for the SDK Nov 08 03:32:30 tophyr: are you sure assets are writable at runtime? Nov 08 03:32:33 it's saying you need the actual SDK version 20+ Nov 08 03:32:46 I didn't think they wer Nov 08 03:32:48 e Nov 08 03:32:54 tdignan: i think in this case it means Asset as in whatever i'm providing Nov 08 03:32:59 not assets/ Nov 08 03:33:05 i don't think those are writable either Nov 08 03:33:28 ie, in this case the asset concept will be my app's Document object and its underlying data Nov 08 03:33:31 oh yes...ADT..man this headache is making me se the wrong stuff Nov 08 03:33:50 tophyr: no, the docs say AssetFileDescriptor is gained from the AssetManager, which manages the applications /assets Nov 08 03:34:13 i think it's a one way thing :|\ Nov 08 03:34:27 sonofabitch Nov 08 03:34:30 you're right Nov 08 03:34:39 maybe openPipeHelper is the way to go then? wtf docs Nov 08 03:34:39 i think it's about time to blow this off and store the data as a blob in sqlite Nov 08 03:35:21 you could pass the socket from the http request straight into that ParcelFileDescriptor.fromSocket() call Nov 08 03:35:30 but I doubt that would be intuitive to use Nov 08 03:35:34 IPC chokes at about 1MB in most devices Nov 08 03:35:38 docs are ~20mb Nov 08 03:35:40 :P Nov 08 03:35:56 and the socket itself wouldn't work, as a) it's inaccessible and b) https would make it useless anyway Nov 08 03:36:06 i see b Nov 08 03:36:12 but I don't see why its inaccessible :) Nov 08 03:36:31 i know this is possible.. i just need to go digging to see how the provided apps do it i guess Nov 08 03:36:33 why can't ClientApp just make the request :)? Nov 08 03:36:59 tdignan: lol good point.. i did just yesterday force HttpURLConnection to do my bidding and send a body via a DELETE request, via some extensive reflection hacking Nov 08 03:37:09 tophyr: I have to change it from HTTPS to HTTP Nov 08 03:37:09 no reason i couldn't dig thru and get the Socket as well lol Nov 08 03:37:34 KNERD: you shouldn't have to.. is the https connection getting rejected or something? Nov 08 03:37:54 i wouldn't waste much more time on contentprovider... imagine if you were going to die tomorrow. would you want contentprovider to be your last memory Nov 08 03:38:20 tophyr: Well it's working now...but the Adroid instructions tell me to do this..."If you have trouble acquiring the plugin, try using "http" in the Location URL, instead of "https" (https is preferred for security reasons)." Nov 08 03:38:22 i would like to die knowing my work lives on Nov 08 03:38:31 if i write a shitty api, nobody will use it... so it will die with me :X Nov 08 03:38:43 KNERD: weird :X Nov 08 03:39:03 well, I am at McDonalds right now..so that could be it Nov 08 03:40:44 tdignan: hah! bingo. openTypedAssetFile that returns a new AssetFileDescriptor(openPipeHelper(...)) Nov 08 03:41:00 interesting Nov 08 03:41:18 openPipeHelper returns a ParcelFileDescriptor, which i can provide a PipeDataWriter to "write" to Nov 08 03:41:24 Okay...nVidia Tegra fully working..I wonder what it can do Nov 08 03:41:25 damn that's convoluted Nov 08 03:42:02 wtf did they not just do public OutputStream openGoddamnDataStream(...) Nov 08 03:45:02 aaaaaand it's api 11. WTF Nov 08 03:48:02 oh, ok.. only the *Helper methods are 11+. openFile is api 1 and openAssetFile is api 3 Nov 08 03:48:12 making this a PITA but still doable Nov 08 03:55:53 * jasta scratches his head Nov 08 03:55:54 hmm Nov 08 03:56:11 have i just found an incredibly subtle bug in ScaleAnimation? Nov 08 03:56:46 looks like when the animation completes, floating point rounding error can cause a final value in the transformation that doesn't match the requested final value Nov 08 03:57:00 i found a set of values that reproducibly show this to be true Nov 08 03:57:02 oooh, fun Nov 08 03:57:11 what are they? Nov 08 03:58:47 still working on my unit test to prove it for sure, but the values look to be mFromX=1.6460052, mToX=0.40816328 Nov 08 03:58:56 the final value is 0.4081633 Nov 08 03:59:10 tdignan: also, i was dead wrong about ParcelFileDescriptor being a pain.. it has a ParcelFileDescriptor.createPipe() method, which is exactly what i need. i return one of them and write to the other. Nov 08 03:59:21 the problem may not be in ScaleAnimation tho but rather in some other animation layer Nov 08 03:59:28 maybe even my code still Nov 08 04:02:20 * jasta scratches his head Nov 08 04:03:09 hmm.. looking at the code for ScaleAnimation it doesn't seem to me like that'd happen.. or that it'd be ScaleAnimation's fault, at least Nov 08 04:03:16 (1.6460052f + (0.40816328f - 1.6460052f) * 1.0f) = 0.4081633 Nov 08 04:03:28 hah yeah oyu're at the same place i am Nov 08 04:03:37 i just executed that expression in the debugger Nov 08 04:03:40 and got the wrong answer Nov 08 04:03:48 oh Nov 08 04:03:59 well that's just float err in the JVM Nov 08 04:04:02 jasta, do you know how floating point works? Nov 08 04:04:07 yes, it is. but it should be fixed. Nov 08 04:04:20 canadiancow: of course i do. that makes it ScaleAnimation's bug though. Nov 08 04:04:29 they're assuming that floating point error doesn't exist in the way this was coded Nov 08 04:04:36 * tophyr shrug.. not much to fix it with other than using double :/ Nov 08 04:04:40 it's an easy fix if you just special case interpolatedTime==1 Nov 08 04:05:00 oh, i see what you're saying Nov 08 04:05:01 oh, i see Nov 08 04:05:03 tophyr: sure you can. if (interpolatedTime == 0) { return mFromX; } else if (interpolatedTime == 1) { return mToX; } else { ... } Nov 08 04:05:15 that would fix ScaleAnimation Nov 08 04:05:24 is it just ScaleAnimation? Nov 08 04:05:28 or is it all the animations Nov 08 04:05:41 fix ALL the animations!! Nov 08 04:05:42 canadiancow: anywhere floats are used. Nov 08 04:06:04 the real implication here is that you can't easily test for this other than to try to estimate floating point rounding error in your code. Nov 08 04:07:02 in my specific case, i can probably just work around it by maintaining state independent of the actual animated result, which is probably the right thing to do anyway :) Nov 08 04:07:12 but still, very interesting case here :) Nov 08 04:07:54 fix and contribute to aosp.. get that karma goin Nov 08 04:08:44 actually, on second thought i think in my particular case i'm better off testing for some margin. Nov 08 04:09:45 what i have here is a double-tap to toggle zoom case. i was previously (with it working in all my test cases for a really long time actually) implementing it to just say if (currentScale != defaultScale) { zoomTo(defaultScale); } else { zoomTo(magnifiedScale); } Nov 08 04:10:04 this was failing in a weird way because zoomTo(defaultScale) would actually zoom to something just really close to defaultScale. Nov 08 04:10:27 but thinking through this more carefully, if (Math.abs(currentScale - defaultScale) > 0.01f) or something would actually be better logic. Nov 08 04:10:28 ahh Nov 08 04:10:42 what would you desire to happen if someone double tapped during the animation? Nov 08 04:10:49 or even better yet, (defaultScale / currentScale) > ... Nov 08 04:10:58 jasta: that's the one Nov 08 04:11:21 kevinb: stop the animation then start that logic again, just like it should. Nov 08 04:11:31 i love finding bugs like this actually Nov 08 04:11:37 this code was in production for like 6 months Nov 08 04:11:40 never noticed it. Nov 08 04:11:51 legitimately never had this rounding error reveal itself. Nov 08 04:11:58 mother f'er... while you're over there debugging and being smart, i'm trying to go to the bathroom and have lost my shoe Nov 08 04:12:04 wtf Nov 08 04:12:29 talk about "what am i doing with my life" moments Nov 08 04:12:32 tophyr: you're going to the bathroom wrong Nov 08 04:12:40 i had a little animation that sometimes would cancel at the wrong time (maybe because of a similar error). didn't notice it, and no user noticed it, but one day i was comparing screenshots checking something else and saw that something was off and finally found it Nov 08 04:13:47 pragma-: no.. going into that bathroom missing a shoe would definitely be doing it wrong Nov 08 04:14:10 catch foot herpes or something Nov 08 04:14:28 nice Nov 08 04:15:18 note to self: sync android source more regularly Nov 08 04:20:25 fungus is likely Nov 08 04:38:55 for those of you that have apps on the market.. how many apps do you just write as experiments to see if anyone will take the bait before you continue developing further? Nov 08 04:39:28 I've put a massively stupid app on the market before 'magnifying glass' that was just the camera with a zoom button, and it got thousands of downloads in a matter of a few days Nov 08 04:39:41 otoh, I've spent lots of time working on good things, and no one gives a shit about them Nov 08 05:07:52 that's normal Nov 08 05:08:12 music, art, comedy, writing, software. Nov 08 05:08:36 tdignan: It depends what you want. If you want something that will make money, doing it the Eric Ries Lean Startup minimal viable product way makes sense Nov 08 05:08:43 the stupid, obvious, simple, offhand thing becomes massively popular. Nov 08 05:09:02 the stuff you slave away at for years gets released to yawns. Nov 08 05:09:18 Like this app of mine for instance... Nov 08 05:09:19 https://play.google.com/store/apps/details?id=com.panaceasupplies.android.games.lilyhop Nov 08 05:09:35 It's a port of an open source program which took years to develop and draw Nov 08 05:09:58 The play is decent as well, although the movement is not 100% intuitive initially, although it's obvious enough Nov 08 05:10:40 I release it in June, still less than 500 downloads. With advertising. Eight five star ratings, one four star and one one star who says it's too hard. Nov 08 05:10:58 Ologn, let me give you an article i read yesterday Nov 08 05:11:27 Ologn, but do you use virtual currency in the app to hook your users!?! Nov 08 05:11:38 It's no big deal for me though...I spent less than a week porting it...it just is a case in point Nov 08 05:12:04 Ologn, www.smashingmagazine.com/2012/11/07/succeed-with-your-app/ Nov 08 05:12:07 also she needs to be able to jump off and explode, gotta have explosions. Nov 08 05:12:08 ;p Nov 08 05:12:09 lasserix: well one problem is it is mostly in C++, so that makes all the shims for Tapjoy and whatnot more difficult. Plus the game has no virtual currency. Nov 08 05:12:21 a really important article ^^ Nov 08 05:12:30 I was just shooting the shit Nov 08 05:12:32 afk Nov 08 05:12:45 Ologn: :| Nov 08 05:13:27 looks like a decent game Nov 08 05:14:07 i had an app making little money each month but I didn't keep updating it with the features it needed to please the users Nov 08 05:14:11 tdignan: The other thing is I ported a number of games to Android, but had no idea how to monetize them like I do my non-game apps Nov 08 05:14:13 so it kind of dwindled off Nov 08 05:14:26 but you're monetizing your non-game apps well eh? Nov 08 05:14:41 there is a channel, #makingmoneywithandroid for discussing monetization if you want to join us there Nov 08 05:15:09 tdignan: certainly. I made $570 last month. But I should easily push that up to $1000 in the next few months. Nov 08 05:16:01 Sorry to repeat. I'm stuck. I have an app that uses GCM to get notifications. I've come across a problem where if I uninstall my app, and then days later install it again i will get a new register_id from GCM but ... the old register_id still works. Anybody ever seen this? Nov 08 05:16:30 Ologn: nice Nov 08 05:16:53 DrPoole, not sure exactly, but a lot of apps don't completely uninstall when you "uninstall" Nov 08 05:17:15 problem is ... i still have the old reg_id in my subscribers database. Nov 08 05:17:20 they dont Ologn ? Nov 08 05:17:21 and since both reg_id's work ... Nov 08 05:17:24 they get 2 messages :/ Nov 08 05:17:57 canadiancow, I mean some apps put things on the SD card and so forth which don't get removed on an uninstall. Nov 08 05:18:11 Ologn: i don't see any paid apps on your pub page. is it all ad money? Nov 08 05:18:19 if they put them where they're supposed to they get removed T_T Nov 08 05:18:19 i remember a bug - was it samsung - that when you uninstalled and reinstalled, you were left with the previous crap Nov 08 05:18:25 tdignan, yes. All ads. Nov 08 05:19:03 im more surprised that google is allowing the message to go through. Nov 08 05:19:20 once the user gets a new reg_id with that app_id ... surely they'd remove any previous ones Nov 08 05:19:25 :) Nov 08 05:19:26 $$ Nov 08 05:21:20 hello,do we need forex card to create a merchant account ??? Nov 08 05:25:36 wtf is a forex card Nov 08 05:33:15 * t0mless gets paid only by clipper card Nov 08 05:41:19 t0mless, really? Nov 08 05:43:32 how do I use ebt on Google Play? Nov 08 05:45:26 do we need a forex card to create a merchant account ?? Nov 08 05:50:32 Lol no Nov 08 05:50:36 I get paid in beer. Nov 08 05:51:20 i have a clipper card Nov 08 05:51:24 i tried to use it on bart and it wouldnt work Nov 08 05:51:27 any idea why? Nov 08 05:51:33 because you suck? Nov 08 05:51:38 I use mine on bart all the time Nov 08 05:51:40 u_u Nov 08 05:51:49 and you get a discount on ac transit from bart Nov 08 05:51:56 it's only like 25 cents Nov 08 05:52:16 I think you get it on muni also Nov 08 05:52:22 ac transit? Nov 08 05:52:34 oakland bus system Nov 08 05:52:54 * canadiancow looks at a map to figure out where that is Nov 08 05:52:57 ok im not that bad Nov 08 05:56:49 lol... modern family is win Nov 08 06:00:08 Morning folks Nov 08 06:12:03 evening Nov 08 06:13:12 JakeWharton: Evening Nov 08 06:17:14 afternoon Nov 08 06:21:07 midday Nov 08 06:21:38 guys why i get so many compiling errors, i installed android r20.3 windows and also sdk following the android sdk manager. also i upgraded to jdk7 and using eclipse with adt plugin Nov 08 06:21:46 the compilation error are mainly Description Resource Path Location Type The method format(String) of type new Formatter(){} must override a superclass method MECARDContactEncoder.java /com.google.zxing.client.android.CaptureActivity/src/com/google/zxing/client/android/encode line 35 Java Problem Nov 08 06:21:55 i googled it. and some said it's because of incorrect compiler compliance level. I confirmed my project setting is 1.7 and i'm sure there are errors with no relationship to the compliance level(either 1.5 1.6 1.7) i noticed there're hints to use android tools to fix project properties. I run the android tool, but it didn't help Nov 08 06:22:11 use 1.6 Nov 08 06:22:19 i set it to 1.6 Nov 08 06:22:22 didn't help Nov 08 06:22:31 how odd Nov 08 06:22:50 dex never works for me with jdk7 or 8 Nov 08 06:22:58 at least I think it's dex Nov 08 06:23:11 i usually don't even read the error and just switch to jdk 6 Nov 08 06:23:13 how can i check the detail output via eclipse? Nov 08 06:23:28 oh so u mean the jdk is *too* new? Nov 08 06:23:34 dang Nov 08 06:23:40 not sure Nov 08 06:23:47 let me retry Nov 08 06:23:48 but I use jdk6 for compilation Nov 08 06:23:52 and I sure don't use Eclipse Nov 08 06:24:08 how do you compile ... Nov 08 06:24:15 i don't know how to run the commandline Nov 08 06:24:20 Maven and/or IntelliJ IDEA Nov 08 06:24:30 maven? Nov 08 06:24:32 intellij Nov 08 06:24:36 i'm new to java dev Nov 08 06:24:38 or Gradle, if I'm feeling adventurous Nov 08 06:24:38 anything else would be uncivilized Nov 08 06:24:43 why there's no standard project file Nov 08 06:24:58 if there were any like vs, it would be way easier to get hands on java projects Nov 08 06:25:10 davidleon: if you are using jdk7, you will have to go through some hoops http://www.informit.com/articles/article.aspx?p=1966023 Nov 08 06:25:17 Eclipse has project files Nov 08 06:25:31 they're these ugly, awful XML things you should never touch Nov 08 06:26:28 At present Android doesn't have a build system so you're stuck with two completely disjointed systems. I'd recommend just sticking with getting Eclipse to work Nov 08 06:26:36 you can move on to greener pastures once you've mastered that Nov 08 06:29:12 greener pasture <-- what do you mean by Nov 08 06:29:44 Maven, Gradle, and/or IntelliJ IDEA Nov 08 06:30:02 davidleon: there is Eclipse plugin for example to make APK through GUI interface Nov 08 06:30:02 or ANT Nov 08 06:30:06 JakeWharton: so your suggestion is firstly get jdk6 and play around for a while before using maven etc? Nov 08 06:30:22 I would never in good faith recommend Ant for anything Nov 08 06:30:40 davidleon: yes Nov 08 06:30:41 g00s i can't quite understand that article. can i get jdk7 and eclipse and android work together well? Nov 08 06:31:13 intelliJ++ Nov 08 06:31:55 davidleon: you should install JDK, install Android SDK, install Eclipse and than install Eclipse plugin Android Developer Tools if you want to use Eclipse Nov 08 06:32:03 why intelliJ++, many are using eclipse Nov 08 06:32:21 can anyone help me in broadcast receiver? Nov 08 06:32:29 hs_sergey: sure Nov 08 06:32:33 hs_sergey: i can use eclipse to debug my example proj Nov 08 06:32:37 * capella found it buggy .... slow on my VM Nov 08 06:33:10 hs_sergey: just i don't know the mechanism of android UI. so try to get some other example project to run and find out why i can deal with it Nov 08 06:33:18 I have WIN host, and intelliJ is more natural feeling - good IDE Nov 08 06:33:36 capella: is it feature-rich as eclipse? Nov 08 06:33:41 +1 for intelliJ, also netbeans is good too... Nov 08 06:33:43 davidleon: there are several examples in Android SDK folders Nov 08 06:35:09 dunno about for Android yet but for everything else intellij is much better than Eclipse, even for technologies like GWT where the authors have put lots of work into an Eclipse plugin. Nov 08 06:35:09 hs_sergey: yes, some of support addon proj(can't remember the exact name) can't be compiled correctly Nov 08 06:35:14 I had Eclipse (Juno i believe) ... it's got a "CLEAN" button for when it mucks things up, and you sometimes hove to manually remove error message it generates in previous builds ... again, struck me as buggy Nov 08 06:35:36 but yes lots of others love it :) Nov 08 06:36:11 intelliJ++ is in java, is it? Nov 08 06:36:21 It is. Nov 08 06:50:35 yep jdk6 solves majority of problems Nov 08 06:50:53 now it reports error of IntentIntegrator Nov 08 06:51:22 what is an IntentInterator?? Nov 08 06:51:34 is is newer API level than 2.3.3? Nov 08 06:52:39 that's in the ZXing lib Nov 08 06:52:51 barcode processing Nov 08 06:52:52 How to restrict viewPager swipe-able on one direction? Nov 08 06:53:02 s/on/in Nov 08 07:04:12 doogan: yep, i'm compiling it, but failed Nov 08 07:04:38 napster: you could store the previous page and stop the pager from switching if the new page is < the prev Nov 08 07:04:53 davidleon: API 16 is the newest released Nov 08 07:04:56 which is Android 4.1 Nov 08 07:05:12 but i still want the app be able to run on 2.3.3 Nov 08 07:05:22 that's what minSdkVersion is for Nov 08 07:05:54 err it doesn't depend on api level? Nov 08 07:05:59 that sounds odd for me Nov 08 07:06:02 no Nov 08 07:06:17 you should always be compiling with the latest APIs and setting a reasonable minimum API that you support Nov 08 07:06:30 sampullman: but the pager drags almost making the next page visible, right? Nov 08 07:06:42 read this: http://simonvt.net/2012/02/07/what-api-level-should-i-target/ Nov 08 07:07:09 I was trying to not make any response to a left-to-right swipe on the ViewPager sampullman Nov 08 07:07:39 napster: that's right. what i did was clear the prev. page Nov 08 07:07:52 so it swipes halfway, but just shows black Nov 08 07:07:54 minSdkVersion="7" targetSdkVersion="16" is quite reasonable Nov 08 07:07:57 it's rather nice, i think Nov 08 07:07:59 with a build target of 16 Nov 08 07:08:09 hmm got it sampullman Nov 08 07:08:39 Is it possible to intercept the motion event rather? Nov 08 07:09:51 you may be able to stop all back scrolling with onPageScrolled Nov 08 07:10:03 i've never tried, but is seems like that should work Nov 08 07:10:42 let me check that one Nov 08 07:11:22 hi all. In the surfaceview we need to create a thread which controls the canvas drawings. So while i was going through the example, the thread created is always extends Thread class and doesn't implement runnable interface Nov 08 07:11:31 can someone please explain why Nov 08 07:12:16 like in this example http://www.droidnova.com/playing-with-graphics-in-android-part-ii,160.html Nov 08 07:13:03 why do i have to create thread by extending it to Thread class and not implementing the Runnable interface ? Nov 08 07:13:24 even in the jetboy example its done like that Nov 08 07:19:10 |/ns markcapella foofoo| ... but thats not the same as what I type manually Nov 08 07:19:11 |/msg nickserv identify markcapella foofoo| Nov 08 07:20:23 sunny_slls you can implement a runnable if you want http://stackoverflow.com/questions/9201558/thread-or-runnable-when-using-canvas-and-surfaceview-in-android Nov 08 07:22:20 test Nov 08 07:22:40 lasserix: ok but is there any reason behind using a thread like that. do we need a unique object associated with each thread while doing the game programming ? Nov 08 07:24:15 lasserix: is it a good standard way of creating thread if the frames are more ? Nov 08 07:25:30 runnable is prefered, i believe Nov 08 07:25:42 why does updating a rom suck so hard? Nov 08 07:25:52 Idk why extending thread is so apparent on the web Nov 08 07:26:07 but the author of replica island uses runnable Nov 08 07:26:27 lasserix: never use java threads Nov 08 07:26:31 sunny_slls, I'm not sure what you mean by unique object Nov 08 07:26:54 as far as frames go, that can be a very complicated question you might google "fixing your timestep" Nov 08 07:28:33 monsti: ha yeah that is what I learned after much trial and error Nov 08 07:28:58 lasserix: same here :) Nov 08 07:30:24 man, this threaded tile renderer is kicking my ass. estimating 1 day in my LoE was a terrible mistake :) Nov 08 07:31:01 sunny_slls, idk how complicated your game is but the set of videos by this guy http://www.youtube.com/watch?v=U4Bk5rmIpic are useful for gaming stuff Nov 08 07:31:11 time to sleep. i'll take another approach tomorrow :) Nov 08 07:31:27 * monsti is going to kick som motorola devs Nov 08 07:31:29 maybe it'll come to you in your dreams like the inventing of the sewing machine Nov 08 07:31:31 +e Nov 08 07:31:37 :) Nov 08 07:31:44 updating my xoom2 fails Nov 08 07:32:57 test more Nov 08 07:34:14 I have an imageview that i want to take up most of the screen, with a textview below it, is there anyway to garuntee the textview will stay on the screen (ie, if i rotate to landscape, the imageview will push the textview off the screen)? Nov 08 07:39:37 Is there a way to set the maximum size of a dialog fragment? Nov 08 07:41:27 Anyone using sbt? I want to add an Android library project to my project but I have a hard time. I would like to use this https://github.com/cyrilmottier/Polaris . Nov 08 07:41:57 I have some sample text with two buttons that call setFontSize on the sample text in a dialog fragment. Problem is on smaller devices, the dialog fragment will grow beyond the bounds of the screen on repeated calls to settext(greater value)... is there a way I can set max size of the dialog fragment? Nov 08 08:00:36 if I am using static instance getter for my dialogfragments, do I have to later set that dialog fragment to null to prevent leaks? Nov 08 08:06:03 if you keep a Fragment around after an activity has come and gone, you'll leak the activity instance. Nov 08 08:06:15 that's why you are given facilities for this, like setRetainInstance() Nov 08 08:06:25 using a static var to save a fragment is definitely wrong. Nov 08 08:06:43 err I use the way its done on the google dev site Nov 08 08:06:47 Oh? Nov 08 08:06:49 show me Nov 08 08:06:57 static DialogFragmentSettings newInstance() { return new DialogFragmentSettings(); } Nov 08 08:07:03 to get it, then show Nov 08 08:07:14 show me the google dev page Nov 08 08:07:24 http://developer.android.com/reference/android/app/DialogFragment.html Nov 08 08:07:31 * monsti shows the motorola devs the dev pages Nov 08 08:07:56 lasserix: oh Nov 08 08:08:05 that won't cause any problems. Nov 08 08:08:13 Sorry I understand static in theory, but I still don't quite grasp the practice Nov 08 08:08:16 it creates a new instance every time you call it Nov 08 08:08:29 how does it know to derefrence? Nov 08 08:08:50 dereference what? It says "return new MyDialogFragment()" Nov 08 08:09:01 it's just a wrapper around the constructor. Nov 08 08:09:02 a factory Nov 08 08:09:19 so i call DialogFragmentSettings fragPage = DialogFragmentSettings.newInstance(); in another fragment Nov 08 08:09:24 you made it sound like you had some sort of sick singleton :) Nov 08 08:09:38 what happens to fragpage? Nov 08 08:09:42 no, fragments can't be like, fragmentception Nov 08 08:09:46 you can't have a fragment in a fragment Nov 08 08:10:02 no i'm declaring an instance and then using show Nov 08 08:10:03 tdignan: sounds like challenge Nov 08 08:10:11 p_l|work: good luck, i tried it :) Nov 08 08:10:22 it might be workable but it's not supported Nov 08 08:10:27 I have a settings fragment, if the user clicks such and such i call DialogFragmentSettings fragPage = DialogFragmentSettings.newInstance(); fragPage.show() Nov 08 08:10:55 you have to add it to the UI first Nov 08 08:11:03 oh wait, it's a dialog fragment Nov 08 08:11:14 yeah Nov 08 08:11:20 ok, so it works, right? Nov 08 08:11:25 it's all working already :) im just wondering about leaks Nov 08 08:11:26 yeah Nov 08 08:11:37 nah there's nothing about the static method that would leak Nov 08 08:11:49 right Nov 08 08:11:50 i see Nov 08 08:11:54 thanks! Nov 08 08:11:56 you'd need to keep a reference to something after it's supposed to be gone to leak Nov 08 08:12:24 like if i did static DialogFragmentSettings fragPage = DialogFragmentSettings.newInstance(); fragPage.show() Nov 08 08:12:27 that might leak? Nov 08 08:12:29 i always kill dialogs with dispose Nov 08 08:12:40 there's really no reason to make that thing static Nov 08 08:12:49 i created a custom activity and every dialog is registered to the base activity Nov 08 08:12:55 I don't think you can initialize it until after onCreate anyway Nov 08 08:12:56 right i was just being hypothetical Nov 08 08:12:58 when it's disposed it also disposes subdialogs Nov 08 08:13:48 dispose? Nov 08 08:15:07 evey dialog has a .dispose() Nov 08 08:15:29 So I have another kinda weird problem, in this dialog fragment I have a sample content with buttons to change the text size of the sample content using setTextSize... this causes the sample grow which causes the dialog fragment to grow which will eventually grow beyond the bounds of the height of the display (but not the width) is there any way I can fix this easily? Nov 08 08:15:58 *ie fix the maximize size of the dialog fragment so it wont grow beyond the display Nov 08 08:23:39 is there some recommended class for drag'n'drop sortable list? Nov 08 08:26:06 honest advice... expandable listview Nov 08 08:26:08 discuss Nov 08 08:26:13 seems like a terrible UI concept Nov 08 08:26:43 Pitel_IPEX: in the old days, you would look at the music player Nov 08 08:26:53 man that code was fucking gross Nov 08 08:27:09 i'm not sure if the technique has changed ics+ Nov 08 08:27:58 https://github.com/bauerca/drag-sort-listview this seems to be most linked on google Nov 08 08:28:11 yeah, or https://github.com/commonsguy/cwac-touchlist Nov 08 08:28:39 ron_frown: expandable ListView isn't _that_ bad of a UI concept if it animated. any idea if it works with android:animateLayoutChanges? Nov 08 08:29:05 just seems like a shit user experience concept now especially in the wake of android on the tablet Nov 08 08:29:19 seems like list fragment / detail fragment makes more sense Nov 08 08:30:36 it all depends on how much data you want to show in the child lists. a lot of data would warrant master/detail view as with expandable works well with smaller data sets so you don't jump to a new Activity with < 5 items Nov 08 08:32:46 does gridview have dnd reordering ? i'm so stuck in froyo Nov 08 08:33:33 Hello Nov 08 08:33:38 Anybody here Nov 08 08:33:39 ? Nov 08 08:33:55 nope Nov 08 08:33:59 g00s: what's dnd? Nov 08 08:34:06 drag n drop Nov 08 08:34:18 duke nukem dorever Nov 08 08:34:33 I have a quick question about textview in fromhtml Nov 08 08:34:40 s/rever/never/ ;) Nov 08 08:35:02 i am using fromhtml to style my text in my app Nov 08 08:35:19 during newline, i used
Nov 08 08:35:41 is there any way that we can control the height set in
tag? Nov 08 08:35:42 g00s: i actually _just_ found this for myself: http://blahti.wordpress.com/2011/10/03/drag-drop-for-android-gridview/ Nov 08 08:35:50 or even in

tag Nov 08 08:35:52 ? Nov 08 08:35:57 junnel1001: css? Nov 08 08:36:16 grantland: ahh yes. i'm amazed this stuff wasn't in android 1.0 Nov 08 08:36:17 grantland: can we have css in our textview? Nov 08 08:36:32 * Gumboot thinks he could probably have helped with BadMrsFrosty's problem... Nov 08 08:37:01 grantland: in stackoverflow, they said only html tags are accepted Nov 08 08:37:22 junnel1001: oops thought you were using a WebView. Why not just use \n instead of html? Nov 08 08:37:41 g00s: seriously D: Nov 08 08:38:04 grantland: they have a content that has many styles (e.g new line, bold, italic, etc) Nov 08 08:38:25 grantland: I searched it and they said to use html.fromhtml() Nov 08 08:38:43 junnel1001: ah sorry. don't know much about specifying style with html Nov 08 08:39:44 grantland: it looks now ok but in having a
tag, i can't control its height-spaced created Nov 08 08:39:49 grantland: ah ok Nov 08 08:41:28 can somebody help me about this problem? Nov 08 08:44:51 junnel1001 unless you can tie some css to it, you wont be changing that Nov 08 08:45:04 and even then Nov 08 08:45:10 its not so much the br but the lineheight Nov 08 08:45:13 line-height: 1000% Nov 08 08:47:14 ron_frown: I am now using

tag, but is there any way to control the vertical space of the content? Nov 08 08:47:56 you can set line height in html, but whether the fromhtml understands any of that (and I'd be surprised if ti did) Nov 08 08:48:00 remains to be seen Nov 08 08:48:37 if you can use css, this is all very simple.. Nov 08 08:48:46 otherwise, it's close to impossible. Nov 08 08:48:48 ron_frown: you're up late these days Nov 08 08:49:07 sampullman: can we use css in html.fromhtml() ?? Nov 08 08:49:22 I managed to fall asleep at 8pm tonight Nov 08 08:49:26 sampullman: they said not.. Nov 08 08:49:27 i have no idea wtf you are doing, sorry Nov 08 08:49:28 woke back up at midnight =) Nov 08 08:49:43 why are you writing an android app in html? Nov 08 08:49:49 ron_frown: on a biphasic sleep pattern now :) you're best thinking will be between sleep :D Nov 08 08:50:30 junnel1001: have you tried replacing all
's with \n? Nov 08 08:50:31 I dont know what my deal was about a year ago, but I would go to sleep at 10 or so, and wake up at 1am, completely wide awake and totally refreshed Nov 08 08:50:38 and I'd go about my day Nov 08 08:50:53 for about 1yr straight Nov 08 08:50:58 sampullman: we have a content needs to be added in the app, it has different style(italic, bold, with newline, etc) Nov 08 08:51:29 sampullman: it is not good to use many textviews in order to get the style they want Nov 08 08:51:56 like i said, i have zero idea what you are talking about Nov 08 08:52:21 sampullman: ok Nov 08 08:52:26 What is the entry for the checkbox when I am setting its style in themes? ie style Nov 08 09:00:42 hey folks - is it possible to programatically get the ACTUAL position of an item in a gridview outside of the touch listener? getChildAt only works on the elements that are currently being rendered, so even if I'm halfway down the list the top one will be 0 Nov 08 09:01:30 I am creating a gson object in order to pass a tablerow through the android client to the java server. All I do is Gson gson=new Gson(); Type type = new TypeToken(){}.getType(); String x=gson.toJson(myTableRow, type); When I do this I get the exception java.lang.IllegalArgumentException: class android.text.BoringLayout declares multiple JSON fields named mPaint... I cannot find any help in google. Can you help me p Nov 08 09:02:20 weird I was using irc.freenode.net but now xchat is saying unknown host for that server name Nov 08 09:05:07 Tee_Pee: assuming you're generating the GridView from a list of some sort, couldn't you get the item displayed at the top of the GridView, find the index of that in the list, and then offset the actual child you want from that index? Nov 08 09:06:04 unfortunately not, but I've found a solution - since the views I generate via the adapter are custom views, I'll just store their position within them and retrieve that. Nov 08 09:06:28 Tee_Pee: perfect Nov 08 09:16:00 anyone please?.. Nov 08 09:17:47 dioHere - really couldn't help you without any code... a stacktrace, some samples etc. Nov 08 09:18:12 Tee_Pee: You need to see the code of the tableRow? Nov 08 09:18:20 dioHere: nothing to talk to you about but what on Earth is going on with that 'type' variable? You make an anonymous inner class instance, throw it away, and ask for its type? Nov 08 09:19:39 Hmm, I think I see what is going on. The anonymous inner class probably causes the compiler to record the type parameter in the class's declaration Nov 08 09:20:47 alankila this is the way that gson works. See the answer to this post http://stackoverflow.com/questions/3458930/deserializing-json-array-using-gson Nov 08 09:22:15 Hmh. I guess some people just do their best to make java look better than it is. Nov 08 09:22:37 like generics actually _worked_ in it. Nov 08 09:23:54 hahah Nov 08 09:23:57 generics work in java Nov 08 09:24:02 they just work shitty =) Nov 08 09:24:21 I dont get why you cant deal with the types passed in as the generic parameter Nov 08 09:24:25 like c# Nov 08 09:24:26 so nobody has worked with gson before? Nov 08 09:24:36 plenty of people have Nov 08 09:24:51 I just manually parse my JSON responses via JSONArray and JSONObject Nov 08 09:25:02 just wait =0 Nov 08 09:25:12 gson actually has performance enhancing stuff in it Nov 08 09:25:25 notably json streaming, and caching Nov 08 09:25:53 I know, but for the purposes of my app I do not need that particular feature Nov 08 09:25:57 ron_frown: it can't deal with it because the type vanishes during the compilation step, of course. Nov 08 09:25:59 as my JSON handling is minimal Nov 08 09:26:10 yep Nov 08 09:26:11 I get that Nov 08 09:26:30 I tend to use gson in places, jsonobject in places Nov 08 09:27:10 ron_frown: generics are broken (erasure) in java, largely because of sun's marketing dept Nov 08 09:27:34 and they were never fixed, largely … uh, because of suns marketing dept :) Nov 08 09:27:36 I think two features that should be required in any language Nov 08 09:27:54 proper generics like c# has Nov 08 09:27:59 and reflection like c# has =) Nov 08 09:28:19 The only thing that I can think that would fail with gson is when the object has some paint "properties" like color etc. I really don't have a clue but I'm trying to figure out how I am creating multiple json fileds named mpaint Nov 08 09:28:24 I'm personally slightly ambivalent about these type erasure based generics. They are fundamentally kinda useless, but they do allow catching some mistakes at compile time Nov 08 09:28:31 i'm reading the first revision of MS Press Windows 8 Internals book Nov 08 09:28:41 first, its nice they have an internals book Nov 08 09:28:49 second, in many ways, wp8 is … quite ahead of android Nov 08 09:28:51 but at source code level they largely replace (Foo) with at a different place, and require a few @SuppressWarnings here and there Nov 08 09:28:52 they are more open than before Nov 08 09:29:06 alankila for stuff I do, I hate having to do shit like this Nov 08 09:29:10 and it's not "fake" like Apple Nov 08 09:29:15 get(whatever, User.class) Nov 08 09:29:29 to actually do things with the "user" type inside the get method Nov 08 09:29:35 i really like that in wp8, you can create your own custom voice user interface for your app. Nov 08 09:30:09 looks like I won't have my freebie WP8 phone for a few weeks, at least :) Nov 08 09:32:21 ot, but i'm also reading "The Cambridge Handbook of Expertise and Expert Performance" … this book is incredible Nov 08 09:33:02 being good at deliberate practice is one of those meta skills, thats really really important to being the best you can be Nov 08 09:33:09 I have a listadapter that loads additional data when scrolling stops. This works fine. However, it also has a filter (for search), and I can't figure out how to listen for the filter to have taken effect. Any clues? Nov 08 09:33:31 annoia: holy cow, its been a long time since i've seen you here Nov 08 09:33:37 g00s: you may practice practice, but how do you practice practice practice? Nov 08 09:33:49 :) Nov 08 09:34:45 g00s - I haven't done a lot of spare time Android development lately... I mostly do it at work now, and usually I'm running around between too many tasks to manage IRC too. Nov 08 09:35:29 IRC is the bastion for the useless, like me. Nov 08 09:35:38 haha Nov 08 09:35:43 * g00s == alankila Nov 08 09:35:58 :D Nov 08 09:36:43 I really don't want to do a sleep, but I can't find a callback :( Nov 08 09:37:04 annoia: hum, sleep and callback? you mean like handler's postDelayed? Nov 08 09:38:48 alankila: When I do a search I perform a query on the sqlite in the background. Only thing is, while I know when it's done, I don't know when it is effectuated in the listview. Nov 08 09:40:04 so you need to wait until the UI thread has finished drawing the (replaced) contents of the adapter? Nov 08 09:40:19 So after the search I can do the extremely ugly thing and wait a bit, and do the long query I usually do when onscroll stops. But I would much rather wait for a callback Nov 08 09:40:22 Yes Nov 08 09:40:38 is this like throttling in a Loader ? Nov 08 09:41:01 It's a lazy list Nov 08 09:41:08 * g00s runs … realizes he just put his pineapple in the freezer for too long Nov 08 09:41:14 ? Nov 08 09:41:27 g00s: Did it explode? Nov 08 09:42:04 Basically my problem right now is that the lazy list is too lazy :P Nov 08 09:42:13 it seems to me that thread synchronization tricks should work for this, but one will have to form a theory about how the android UI thread does the painting. Nov 08 09:42:41 oh what the fuck this is fucked. Nov 08 09:42:44 for instance, one could post a delayed immediately after the adapter's notify of new data occurs, and assume that when ui thread gets around to handling that post, the data is also displayed Nov 08 09:43:01 alankila: Well, the background thread is an API call: runQueryOnBackgroundThread Nov 08 09:43:25 In the SimpleCursorAdapter (probably BaseAdapter) Nov 08 09:43:39 I'm kinda curious... I've programmed some thingies in other OS before. But never on Android. If i were to pick up one specific Programming Language to make some Adroid stuff, what language would that be? Nov 08 09:44:03 PsychoX75: If games, then C++, if anything else, then Java Nov 08 09:44:16 hmm. Curious. Why have I missed the existence of that method until now. Nov 08 09:44:25 annoia: can i have a word with you in private? Nov 08 09:44:30 Sure Nov 08 09:44:50 alankila: It's only used for when you filter your list Nov 08 09:47:39 Hmh. Anyway, I'm running out of ideas, because I haven't actually used any of that stuff. I've always used arrayadapter and dealt with my database in some component that feeds the results to it. Nov 08 09:48:53 that cursoradapter contains methods like getDropDownView suggests that this class resembles a garbage heap of features. Nov 08 09:49:21 hmm I guess they all do. Nov 08 09:49:37 oh gods, this is too messy. Why did I even look at this class. Nov 08 09:50:40 Haha Nov 08 09:51:28 alankila: android is like DOS in 2012 , but implemented in terms of linux Nov 08 09:51:42 C:\ > Nov 08 09:51:44 :D Nov 08 09:52:34 and DOS took over the world, even though there was OS/2 :P Nov 08 09:53:04 It's all about getting an easy to pronounce name :) Nov 08 09:53:27 OS/2 had great geek appeal due to the slash, but DOS sounds better :) Nov 08 09:53:56 ibm and their slashes Nov 08 09:53:57 DOS/2 wouldn't have been so great, it's literally just half of DOS. Nov 08 09:54:32 System/360 , etc … thats 360'th of the system Nov 08 09:54:38 :) Nov 08 09:55:26 s/390 is then a really, really small machine, right? ;) Nov 08 09:55:34 :) Nov 08 09:56:01 or the "s" stands for "suck" and it means it only sucks very, very little. Nov 08 09:56:57 though I confess, I don't understand the appeal of placing slashes and random numbers. Sounds like someone named products after some internal cataloging scheme. Nov 08 09:57:10 g00s: actually, OS/2 lost to Windows 95, not DOS Nov 08 09:57:26 or did it lose to windows 3.11? in any case. Nov 08 09:57:26 p_l|work: That's even more embarassing Nov 08 09:57:44 well, they called it a better dos than dos and a better windows than windows Nov 08 09:57:48 :) Nov 08 09:58:28 annoia: no, it really isn't - 95 was *huge* *marketing* success, leading to quickly forming market of win95 applications, and I don't know if OS/2 supported Win32s extension to Win16 Nov 08 09:58:34 i'm reading about gtk3 and gnome, man this is depressing ! wtf is going on Nov 08 09:59:09 yea... i avoid gnome3 Nov 08 09:59:31 I think gnome folks try to solve issues relating to lack of manpower by doubling down on process and vision Nov 08 09:59:32 staying on gnome2 until everything completely breaks, in which case i'll jump over to some other dm Nov 08 09:59:49 i liked xfce but havent used it in a while Nov 08 10:00:06 I'm actually pretty happy with Gnome 3 Nov 08 10:00:11 delete all the features! break everything 3rd party on every version! but largely I think because the sane people left already Nov 08 10:00:33 Not too happy about Ubuntu though. It seems like everything breaks on every updaste Nov 08 10:00:36 update Nov 08 10:00:38 that ship has a hole in the bottom, but it has taken a long time to fill up with water and go down entirely Nov 08 10:00:58 i'd rather stay on gnome2 and patch bugs myself than upgrade to gnome3 Nov 08 10:01:03 hello, i am trying to find a way to position a widget at arbitrary locations. Basically i am looking for a way, to have a canvas and draw on it, but keep all the benefits that the UI widgets offer. Does anybody have an idea, how this could be done Nov 08 10:01:04 ? Nov 08 10:01:21 * hs_sergey moved to KDE Nov 08 10:01:33 KDLOL Nov 08 10:02:19 had to patch some stuff when they broke the "zoom" mode for wallpapers for some stupid reason. i read the ticket comments that resulted in the change, made no sense Nov 08 10:02:21 ok, so this isn't the year of linux on the desktop Nov 08 10:02:27 2013 will be ! Nov 08 10:02:27 no, definitely not Nov 08 10:02:29 and it'll never be Nov 08 10:02:38 anyway qt has manpower advantage of something like 100:1 over GTK+ Nov 08 10:02:51 who ows qt now >.> Nov 08 10:02:55 *owns Nov 08 10:03:02 logica? Nov 08 10:03:10 trolltech? Nov 08 10:03:15 Digia Nov 08 10:03:23 oh digia Nov 08 10:03:30 fffuuuuuuu Nov 08 10:03:47 well, the next time somebody wants to sell it, i'll buy it for $10 Nov 08 10:04:39 I'll bid $10.01 Nov 08 10:04:50 $10.02 Nov 08 10:04:50 this is going to be messy Nov 08 10:04:56 I'll bid $5 and a copy of openmotif Nov 08 10:05:01 * Frigolit pokes Zider in the ribs Nov 08 10:05:18 * Zider ribs Frigolit in the poke Nov 08 10:05:23 :D Nov 08 10:05:28 i see you're here as well Nov 08 10:05:31 actually, no, I withdraw that bid - what would I do with qt? I've got no space at home to store it... Nov 08 10:05:32 where?! Nov 08 10:05:40 oh shit! where are we? :E Nov 08 10:05:54 nooo idea Nov 08 10:07:45 rox: anyway, what sort of widget? Nov 08 10:08:00 If I am not using the dialog fragment alert dialog builder, but just inflating my own view, how do I change the theme styles of the dialog window? Nov 08 10:09:20 new Thread(new SearchUpdatedTimer()).start(); // TODO Eeeeeeeew... Find a callback! Nov 08 10:09:28 I'm a sad panda now :( Nov 08 10:10:44 do dialog fragments implicitly inherit the default alert dialog attributes? Nov 08 10:11:28 there are more questions about dialog fragments ... Nov 08 10:11:37 ? Nov 08 10:11:44 what are they deprecated now? Nov 08 10:12:08 no ... Nov 08 10:12:29 oh what's this there are more questions about dialog fragments ... signify? Nov 08 10:13:14 well, if i was google, i would have a bot going through all the questions here and flagging apis getting the most questions ... Nov 08 10:13:25 maybe those apis suck because they are confusing or something Nov 08 10:13:53 and google groups, and other forums Nov 08 10:13:58 And instead they choose to make it even worse with fragments? Nov 08 10:14:39 do dialog fragments implicitly inherit the default alert dialog attributes? Nov 08 10:14:56 you can withhold your +1 vote on the page with the api :) Nov 08 10:17:32 what is onCreateView from Activity for? is not mentioned in the activity's lifecycle Nov 08 10:40:13 anyone noticed the issue that SkDynamicMemoryWStream is removed from libskia.so in 4.1? Nov 08 10:49:38 Why would the title disspear in a dialog fragment after i set a custom theme? Nov 08 10:58:39 hi Nov 08 10:59:53 anyone active at present? I got some issue i require help with.. Nov 08 11:01:21 spew your question, someone will run with it if they know something... Nov 08 11:01:34 yop.. Nov 08 11:01:43 s/will/may Nov 08 11:02:10 i need the minimum video resolution for android devices for 4 general ratios Nov 08 11:02:45 1.78, 1.66, 1.6 and 1.5 Nov 08 11:03:38 like i know HTC legend forms the lowest for 1.5, that is 240X320 Nov 08 11:04:12 basically i need to run a fullscreen video on multiple devices Nov 08 11:05:58 Clandestine: are you using VideoViewer with RelativeLayout or something else ? you can get the screenHeight and width at onCreate with metrics Nov 08 11:06:26 yes i am using VideoViewer Nov 08 11:07:06 but i need to create these videos beforehand ... so i need the minimum false-safe resolutions for different aspect ratios Nov 08 11:07:32 like i said, for the aspect ratio 1.5, 240x320 is the minimum resolution Nov 08 11:07:43 what about 1.78, 1.66 and 1.6 Nov 08 11:08:59 you can set only width or only height and the rest will set correct aspect ratio with VideoView if you use RelativeLayout around it Nov 08 11:09:27 hi... does Google provide us with statistics about which languages people use on their phone? Nov 08 11:09:59 I found statistics about which languages are chosen for localization, but that's not the same Nov 08 11:10:43 hello all. I have a question. Is it possible to install a company-wide app market on a device and completely remove every other possibility to install an app (play store + manual installation) ?? Nov 08 11:11:45 Pinas: you need some devicemanagement for that Nov 08 11:12:08 Pinas: query? Nov 08 11:12:08 monsti - yea right. Can i do this device management without root ?? Nov 08 11:15:32 no as I know Nov 08 11:15:58 play store is preinstalled and can't be simple removed Nov 08 11:18:25 I do wonder about the multi user accounts in 4.2 Nov 08 11:18:45 it's at least plausible that might work. Nov 08 11:19:24 if users have differing permosions Nov 08 11:24:52 How to play a video fullscreen on multiple devices? Nov 08 11:24:54 how do you debug native applications ? I've followed this guide: http://tools.android.com/recent/usingthendkplugin but the debugger only breaks in the breakpoints set in java code, not in C code Nov 08 11:28:01 Clandestine: you would detect screen height and screen width and compare that aspect ratio to your video aspect ratio to determine if you need to set only height or only width with RelativeLayout to the detected screenHeight or width accordingly Nov 08 11:29:34 that would fullscreen to fit without losing aspect ratio Nov 08 11:29:54 on any resolution automatically Nov 08 12:06:02 I need help with implementing leadbotl push notifications Nov 08 12:06:07 leadbolt* Nov 08 12:07:07 hey in many programs there's menu you can pop up by pushing one button from phone. How does one create such a menu? Nov 08 12:08:00 you mean the menu button? Nov 08 12:08:31 tneva82 override onOptionsCreatedMenu Nov 08 12:09:11 Hm, Google Play updates apps a lot slower than it used to do :/ Nov 08 12:10:00 how to implement Leadbolt push notifications into app? Nov 08 12:10:13 ok thanks Nov 08 12:11:48 Amex: geez man, just read their docs :/ Nov 08 12:12:15 HDroid: I did, and Dispay adds are working just fine, but Notifications won't Nov 08 12:12:59 Then help the people here figure it out by showing what you've done. Nov 08 12:53:32 Don't use the menu button ! Nov 08 12:53:36 its evil and I mean that Nov 08 12:55:23 Hi there. I'm trying to send data to a server, I'm following http://www.softwarepassion.com/android-series-get-post-and-multipart-post-requests/ (HTTP POST) and it compiles and and runs fine. But I can se that app never reaches line 11. Any ideas? Nov 08 12:58:12 Anyone? Nov 08 12:59:20 Is it okay to access a database in a class extending AppWidgetProvider if it is a simple query? Nov 08 12:59:35 Err do I have to use a service? Nov 08 13:02:23 wroos: check connection to server and server response code Nov 08 13:02:50 hs_sergey: Do you know how I can do that? Nov 08 13:03:56 wroos: send request from browser. use for example FireBug or simple create test html page with apropriate form Nov 08 13:08:45 a short question Nov 08 13:09:04 can i use api level 16 to develop a 2.3.3 app? Nov 08 13:09:34 davidleon: as long as your minSdkLevel is low enough ofc Nov 08 13:10:01 hs_sergey: I know that the post function working on the serverside...working from the browser anyway. Nov 08 13:10:23 timroes: so why api level 10 still exists? Nov 08 13:10:37 i mean why it's still on the list of sdk Nov 08 13:10:55 davidleon: yes if you set min sdk version for example 8 Nov 08 13:11:07 i don't get your question davidleon Nov 08 13:11:17 ah okay i guess i get it :) Nov 08 13:11:22 wroos: add sysout and see whats happening through adb logcat Nov 08 13:11:32 why on earth a developer want to restrict himself to apilevel 10 for example, while he can actually compile against api level 16? Nov 08 13:11:44 how to set minsdklevel Nov 08 13:11:45 davidleon http://developer.android.com/guide/topics/manifest/uses-sdk-element.html#ApiLevels Nov 08 13:12:15 davidleon: i do this for example to see if i really have no higher level api in the code that should run on every platform Nov 08 13:13:01 and the question is, why wouldn't it be there? i am pretty sure then someone would search it, and you would find it somewhere around the internet to download Nov 08 13:13:43 hs_sergey: How do I sysout? If you dont mind me asking. Nov 08 13:13:45 http://simonvt.net/2012/02/07/what-api-level-should-i-target/ useful. Nov 08 13:14:02 timroes: he can safely develop app on apilevel 16 and set his minsdkversion to 8 Nov 08 13:14:17 wroos: add System.out.println("variable = " + variable); Nov 08 13:14:30 timroes: also wouldn't it bloat the app if you widen the gap between minsdkversion and the actual apilevel you use? Nov 08 13:14:31 Ok. Nov 08 13:15:09 what's the shortcut for project wide search in eclipse? Nov 08 13:15:12 wroos, sysout is useful if you can read "stdout", otherwise you can use Log.v("yourtag","yourdebug info"); Nov 08 13:15:13 davidleon: in what way would it bloat my app? Nov 08 13:15:28 wroos, Log.e, Log.w, Log.d and Log.v for verbosity levels Nov 08 13:15:46 you mean because I need backward compatible alternative code for older devices? Nov 08 13:16:14 timroes: e.g. i uses a 16 specific feature, it surely has some code to deal with that in the case that sdk version 8( i would assume be a framework on the device) Nov 08 13:16:15 timroes: yep ,exactly Nov 08 13:16:25 wroos: I like to use Log.e("tag", "error to report") because it will make your Logcat window appear even if it is not showing Nov 08 13:16:27 http://www.softwarepassion.com/android-series-get-post-and-multipart-post-requests/ wich variabel should I test? Nov 08 13:16:41 eh yeah ofc it will, but therefor you support down to api level 8 Nov 08 13:16:44 timroes: and api 16 would be a different framework from api 10 Nov 08 13:16:59 timroes: ok, get it Nov 08 13:17:05 ader10: dlaroche: I have use Toast before... Nov 08 13:17:06 thats your decission if you want to have a high backward compatibility, or just use every new feature without making it backward compatible Nov 08 13:17:13 I'm new to this... Nov 08 13:17:50 wroos: I guess that would work in most situations, but if you are trying to log a bunch of data then a bunch of flashing Toast messages wouldn't be that useful Nov 08 13:18:13 wroos: we were all new to this at some point Nov 08 13:18:21 dlaroche: Thats true. Nov 08 13:22:37 Hey is it safe to access a database via content provider in on the onUpdate method of an AppWidgetProvider subclass? Nov 08 13:22:51 *by safe I mean I don't have to use a service to do this if the query is simple Nov 08 13:55:49 After implementing onTouch() I seem to have lost the ability scroll the webView, is there a method to re-enable scrolling, this is the code for the touch: http://pastebin.com/yKrtb9yw Nov 08 13:57:11 docmur: Are you sure that onTouch() doesn't override touch events on your webview? Nov 08 13:58:11 quick question - I want to fire a method each time a fragment comes into view (view pager)... onStart() and onResume() are only run when the fragment is instantiated, and I need this to run EVERY time a fragment is starting to show on the screen. Nov 08 13:58:23 Hey is it safe to access a database via content provider in on the onUpdate method of an AppWidgetProvider subclass? Nov 08 13:59:04 Tee_Pee doesn't onpageselectedlistener have a scroll dx method? Nov 08 13:59:44 it does, but I was hoping that there was something internal to the fragment... for some reason running this from a listener gives me a NPE. Nov 08 14:00:13 oh, i see Nov 08 14:00:30 i dont know lifecycle of fragments in pagers but do the normal life cycle stops happen? Nov 08 14:00:35 ie on attach on create etc Nov 08 14:00:41 as for your contentprovider question - I believe it is, but I could be wrong. Nov 08 14:00:57 I honestly have no idea... hold on, let me fetch the relevant code Nov 08 14:01:42 i think there is a isvisible method of the fragment pager adapter Nov 08 14:04:19 http://pastebin.com/eUUiZvx9 Nov 08 14:04:26 there we go Nov 08 14:06:34 Tee_Pee you find the fragment by a tag not that Nov 08 14:06:56 http://stackoverflow.com/questions/6976027/reusing-fragments-in-a-fragmentpageradapter check penos answer Nov 08 14:07:28 Hi, is there anyone out there who can help me via TeamSpeak screen share/control? I'm trying to send data from my activity to a widget. Simple stuff if you know how. I can gift something via PayPal for your time (not a job... just a gift) for helping me out in this way. Thanks. Nov 08 14:07:30 then do fragmentmanger findfragmentbytag and do what you need to do Nov 08 14:07:52 fluxg10 what's the problem? Nov 08 14:08:11 i am actually just reimplementing my widget into my app as we speak :P Nov 08 14:09:20 no exact problem just want someone to show me how to send to get a list from an activity class and get it showing inside a widget. I'm new to android. Nov 08 14:10:00 listsin app widgets are only supported in api >=3.0 Nov 08 14:10:05 *so you know Nov 08 14:10:24 is there no way to get an app widget in 2? Nov 08 14:10:33 not with a listview Nov 08 14:10:36 lasserix - the problem is that this is a fragmentstatepageradapter Nov 08 14:10:40 it doesn't assign tags Nov 08 14:10:53 are you sure it doesn't automatically assign tags? Nov 08 14:10:59 yup Nov 08 14:11:13 but doesn't matter, I think I got a different solution, let me try Nov 08 14:12:00 Tee_pee why dont you override insantiateitem and assign the tag with the position? Nov 08 14:12:13 then in on page selected you know exactly what the tag is Nov 08 14:12:52 flux810: you could use a multiline textview with buttons to scroll up and down, but listviews in app widgets don't happen until >= 3 api Nov 08 14:13:49 Sorry my IRC session froze, so I have onTouch() detecting X gestures but this has somehow disabled the ability to scroll vertically in the same view, is there a way to re-enable that, Or can I handle the scrolling in onTouch() this is the onTouch code: http://pastebin.com/yKrtb9yw Nov 08 14:14:22 I don't know what I'm doing so I need someone to show me if possible Nov 08 14:14:24 docmur i think you want to return false on ontouch detecting x gestures so the touch event is passed down the hiearchy Nov 08 14:14:33 flux: what do you want to do? Nov 08 14:14:36 is there any way to disable apps from running in background? Ie always start from scratch? Nov 08 14:14:51 lasserix - I am thinking of getting a reference to the fragment from the pager directly rather than the adapter Nov 08 14:14:54 via getChildAt() Nov 08 14:15:17 I hope the answer is that easy, I thought returning false threw the touch out aka didn't process it Nov 08 14:15:19 oh then iterate through getchildat and set up all your tags Nov 08 14:15:22 then I can find the fragment via findViewById Nov 08 14:15:52 docmur: you process what you need before and then return false so it passes down the line, i believe Nov 08 14:16:10 well I'm trying it in a sec Nov 08 14:16:30 I have a method in my activity that retrieves data from a database file, but I want my widget to do that aswell, how do I do that? Nov 08 14:16:37 Tee_Pee you can't assign ids since ids refer to xml layouts? you have to use tags Nov 08 14:17:09 fluxg10 ha i am doing the same thing, I've seen on the web you can make a instance of your database helper in your update method Nov 08 14:17:13 or you can set up a content provider Nov 08 14:17:27 or you can set up a service that does the query and updates the widget through a callback Nov 08 14:17:41 i think the first option is not recomended since broadcast recievers have limited lifespan Nov 08 14:17:48 *or the onupdate method does Nov 08 14:19:30 /quit Nov 08 14:19:54 Okay, I have a fresh installation of the android dev environment. I loaded up the SoftKeyboard sample in the dev environment and it starts with half a dozen errors involving InputMethodSettingsFragment. Nov 08 14:20:00 Are these normal, or have I fucked something up? Nov 08 14:21:05 Basically it complains about the @Override directive on a bunch of methods it implements from the InputMethodSettingsInterface Nov 08 14:22:06 guys? Nov 08 14:22:30 I'm installing the Android SDK package on http://developer.android.com/sdk/installing/installing-adt.html#Download Nov 08 14:22:56 Where do i " specify the location of your Android SDK directory:" ? Nov 08 14:23:06 inside eclipse..? Nov 08 14:23:12 in eclipse Nov 08 14:24:09 yeah i know. But where in it? The walkthrough is talking about a 'welcome window'. I dont get any welcome window. So it seems i must find my way in the menu bar at the top. Nov 08 14:24:24 "In the "Welcome to Android Development" window that appears, select Use existing SDKs. Nov 08 14:24:38 Kinda stuck at that part. Nov 08 14:25:54 Is your computer on? Do you see the "Welcome to Android Development" window? Nov 08 14:26:28 Im inside Eclipse. I'm nlooking at a Ecplipse window Nov 08 14:26:33 So I tried returning false but that didn't work, so I still can't scroll Nov 08 14:26:54 Ah, turns out I have Java 5 installed Nov 08 14:27:00 not 6 Nov 08 14:27:09 so you can't @Override an interface Nov 08 14:27:22 PsychoX75: you should be able to set the SDK location with Window->Preferences->Android Nov 08 14:27:42 ah. thanks. (and no I dont get a 'welcome to android' window at all) Nov 08 14:27:44 Hi, is there anyone out there who can help me via TeamSpeak screen share/control? I'm trying to send data from my activity to a widget. Simple stuff if you know how. I can gift something via PayPal for your time (not a job... just a gift) for helping me out in this way. Thanks. Nov 08 14:31:35 sup everyone, i'm writing an app that has two fragments in my MainActivity, one fragment is kind of a tab selector, the other displays the content of the selected tab. now, when I override my mainactivitys onDestroy method to do some cleaning up when the app closes, the onDestroy will only be called when im in the first of four tabs.. what could be the reason? Nov 08 14:36:22 fluxg10: do you have a database helper class? Nov 08 14:39:29 fluxg10: just make an instance of your sqlite helper class in your onupdate and access what you need Nov 08 14:40:03 alternatively you can set up a content provider and use the context from the parameter list of onUpdate Nov 08 14:41:29 I get a bunch of Call Requires API Level 11 or 14 errors in the sample code. Is it better to annotate the methods or change the minimum API level? Nov 08 14:41:34 hi Nov 08 14:42:13 and if I change the minimum API level, how do I do that? Nov 08 14:42:22 sorry, /noob here Nov 08 14:42:36 jeffheard: either don't use those methods, install the compatibility layer, or make your minimum requirement higher (which will exclude older devices) Nov 08 14:42:53 jeffheard: you set the minSdk value in the manifest to change your minimum api level Nov 08 14:43:04 thanks, folks Nov 08 14:43:15 jeffheard compability should already be installed tho Nov 08 14:43:16 if you wrote your code in a backwards compatible manor where you're sure that API levels that can't use those methods won't ever call them, use the @TargetApi annotation Nov 08 14:43:54 lov: this is simply the SoftKeyboard example from the Android samples Nov 08 14:44:28 I've not written any of my own code yet. I just created a New… Other… Android Sample project in Eclipse Nov 08 14:44:50 and I'm faced of course with a ton of errors that I have to clear up before I even start coding Nov 08 14:44:54 do you have to set your jdk level to 1.6 or something? Nov 08 14:45:08 lasserix, yes, I'm developing on Lion. Nov 08 14:45:38 I just have the default JDK installed Nov 08 14:45:59 Is there another way to enable scrolling? For anyone that wasn't here, I have onTouch() filtering some X coordinates but now it seems my scrolling ability has been truned off in my webview, this is the touch code: http://pastebin.com/yKrtb9yw Nov 08 14:46:01 If it's important I can install 7, but it just makes for some minor errors as far as I can tell Nov 08 14:50:44 I am looking at getting into android development as I want to pick up a Nexus 4 next week. I don't really know much about android development though. From what I understand development is primarily done using Java however there are options for C++ via the "native sdk", is that correct? Nov 08 14:51:11 Woohoo! I managed to run my first "Hello World" App on my emulator. (i know it's nothing, but for me it means allot!) Nov 08 14:52:10 Congratulations :) Nov 08 14:52:52 ditoa: the C++ NDK stuff is almost always used for high performance stuff, such as games Nov 08 14:53:05 Or interoperability with existing C++ libraries Nov 08 14:53:12 :) Nov 08 14:53:22 so in reality Java is the main development language? Nov 08 14:53:33 Yes Nov 08 14:53:52 How is development for it? are the tools good? Nov 08 14:54:09 intellij idea is excellent for it ;) Nov 08 14:54:14 Sorry my old IRC session crashed again Nov 08 14:54:25 so is it possible Nov 08 14:55:32 ditoa: you can use C++ if you want, but the expected path is to use Java Nov 08 14:55:34 ditoa: tools are imho better then on C++ or C, since you have really good debugging and stuff. Most ppl use Eclipse, cause the official android developer tools are for eclipse, Intellij IDE has also pretty nice support. I prefere using NetBeans, though its Android support is imho the worst of all of these three main IDEs Nov 08 14:55:46 ditoa: see http://developer.android.com/tools/sdk/ndk/index.html Nov 08 14:55:50 ditoa: the tools are decent Nov 08 14:56:19 thanks timroes and lov Nov 08 14:56:23 How do I run the SoftKeyboard example in an emulator? Nov 08 15:02:39 returning false from everything I'm reading should work but just doesn't Nov 08 15:11:25 If I have app widgets that start an activity, how should I flag the activity / intent for the back stack? Nov 08 15:11:33 *the activity is just to customize the widget Nov 08 15:27:23 Wrapping the WebView that lost scroll from onTouch in a ScrollView doesn't fix the issue either :S This is weird Nov 08 15:33:46 Under eclipse I changed android:completionThreshold="1" in an auto complete textview (before it was 3). When rerun the app it works as if it was still 3? is there a way to "force" changes or maybe I'm doing something wrong? Nov 08 15:37:31 displaying a toast doesn't require any manifest change? Nov 08 15:39:36 errr bad shortcut, so abnout my toast question? Nov 08 15:39:38 is there a way to create custom AVDs? Nov 08 15:39:50 baptistem: no it doesnt Nov 08 15:39:56 sleepster: in what way custom? Nov 08 15:40:23 TachyonDev, create them to match an exact hardware spec for an exact phone Nov 08 15:41:05 sleepster: eh no? I mean you can use an X86, or Arm, or MIPS target, set a resolution, and try to match the settings Nov 08 15:41:09 TachyonDev: ok thx, my no-toast error may have an other origin Nov 08 15:41:12 but an exact hardware emulation model? Nov 08 15:41:21 baptistem: .show() ? Nov 08 15:41:22 :P Nov 08 15:41:34 Thanks TachyonDev Nov 08 15:41:37 I see your point Nov 08 15:42:25 also, assuming I can change all aspects of the Android OS from the linux kernel up to the apps in the Dalvik VM... I need to create a guaranteed secure channel between a linux server and a system app Nov 08 15:42:36 what ways can I do this? Nov 08 15:42:45 heh Nov 08 15:42:48 err. I mean linux service Nov 08 15:42:52 not server Nov 08 15:42:55 ooooh! I was using .displayalittleblackboxwithafunkymessageinside()... dammit :) Nov 08 15:42:58 so from a linux service to a system app Nov 08 15:43:15 I was looking into DBus since I am familiar with it Nov 08 15:43:37 this is sounding less and less like -dev and more and more like -root Nov 08 15:44:06 oh my project is purely a research project Nov 08 15:44:18 I work for a small startup and we are just experimenting Nov 08 15:44:44 I was hoping to avoid encryption/decryption to keep messages secure Nov 08 15:44:50 if you have full control, why not just compile it together Nov 08 15:44:54 due to the processing power required Nov 08 15:45:06 i guess the question is a bit too abstract Nov 08 15:45:08 TachyonDev: ok I got it, "can't create handler inside a thread ... blabla no looper.preparer()" it wasn't .show() :p Nov 08 15:45:18 baptistem: gah Nov 08 15:45:32 baptistem: i don't even want to see what you are trying to do Nov 08 15:46:30 TachyonDev: please don't ! I'm not proud of what I'm asking to do :) Nov 08 15:46:41 heh Nov 08 15:46:55 hi! Nov 08 15:46:59 ZOMG HI! Nov 08 15:47:11 .. too much coffee today Nov 08 15:52:05 never too much coffee :-) Nov 08 15:52:21 I have an activity, where there are no tabs on the screen, but I want to start an intent to call a tab activity...or fragment, how can I specify that specific tab? So far I can call the activity just fine, but it loads the first tab Nov 08 15:54:49 Hi, anyone got to successfuly set a bp in jni code using command line tool only Nov 08 15:55:42 it remind me a function in some api that was like ".setdecaffeinated()" Nov 08 15:55:48 s/remind/reminds/ Nov 08 15:57:13 Hi, anyone got to successfuly set a bp in jni code using command line ndk-gdb tool only Nov 08 16:07:53 hkhalid: viewpager, tab fragments, etc? Nov 08 16:08:44 TachyonDev, I am using the template Android provides for tabs and swipes for the content. Do you think it has to do with using an extra parameter when starting the intent? Nov 08 16:10:07 hkhalid: not sure i follow you, which template are you using? Nov 08 16:10:43 TachyonDev, when you make a new project with eclipse, you can have an option to have it setup tabs/swipable content for you Nov 08 16:12:08 any java dev here? Nov 08 16:12:17 lol dctrd Nov 08 16:12:32 dctrd: go to ##java Nov 08 16:12:44 hkhalid: so you are having problems adding more tabs? Nov 08 16:12:48 dctrd: no, go away. Nov 08 16:12:56 dctrd: we're all too busy partying Nov 08 16:13:07 purpose? Nov 08 16:13:15 dctrd: fun Nov 08 16:13:25 dctrd: the purpose of partying is to enjoy ones self Nov 08 16:13:26 is there any beer? Nov 08 16:13:32 * lov toots a party horn Nov 08 16:13:37 TachyonDev, I am having trouble calling a certain tab, instead now, the intent starts the activity which has tabs in the layout, but it starts with the first tab, I want it to load the 2nd tab Nov 08 16:13:41 dctrd: only the best craft beer money can buy Nov 08 16:15:05 hkhalid: o_O, the intent isnt doing that .. if you just have a beginning sample project with a tab swipe system .. Nov 08 16:15:14 if you want to switch the current tab, just do it Nov 08 16:16:12 TachyonDev, I have an activity with buttons, I have it where if you tap any of them, they load an activity that has tabs, and it loads that activity with the first tab selected. What I want insteadis to have the ability to specify which tab is loaded from the activity w/o the tabs Nov 08 16:16:44 so pack in an extra into the intent Nov 08 16:16:52 thanks TachyonDev :) Nov 08 16:17:11 then just pull it out in the tab activity and mViewPager.setCurrentItem(x) Nov 08 16:17:38 yarp Nov 08 16:21:32 why does eclipse have to be so god damn slow Nov 08 16:24:10 Hi Nov 08 16:24:43 Ho can I know if an external activity launched from my activity has crashed? Nov 08 16:25:31 external? eh .. cross process exception handling? remote exceptions? .. still dont think android supports it Nov 08 16:26:57 if you have launched another activity, your activity is paused.. you wont know if something is wrong until you come back to your activity .. then you have to check results Nov 08 16:29:10 Can anyone here recommend a 360 pano image generator that works on android. Using pano2vr, and it's super buggy on android? Nov 08 16:37:37 fakingfantastic: #android Nov 08 16:38:07 evancharlton: really, Im talking about a generator to be used in an android app I'm building Nov 08 16:38:15 if I'm still in the wrong place, just let me know Nov 08 16:38:39 fakingfantastic: oh, it sounded like you were looking for an app; my apologies Nov 08 16:38:48 yeah no, that was me, looking over, totally unclear Nov 08 16:39:31 is it possible to launch a searh dialog over a ScrollView / WebView Nov 08 16:40:25 I enabled deafult_searchable in my manifest but when I call OnSearchRequest() the bar doesn't appear Nov 08 16:43:00 does anyone know of some really good free online training videos? Nov 08 16:43:56 sleepster: khan academy? Nov 08 16:45:24 sleepster: http://m.youtube.com/watch?v=bjQKCLvtLN0 Nov 08 16:46:16 watching videos is a really inefficient way to learn APIs, I'd say Nov 08 16:46:31 indeed Nov 08 16:46:41 TachyonDev, so from the fragment activity, its there that I should do the tab change correct? Nov 08 16:47:01 TachyonDev, its sucha pain to convert fragments to fragment activities, or even listfragments Nov 08 16:47:38 well they are kind of different things Nov 08 16:48:43 hkhalid: if you are launching fragmentactivities with viewpagers in them and want to specify the tab at creation, you would pack an extra into the intent, pull it out in the oncreate in the viewpager activity, and tell the viewpager to switch to it Nov 08 16:48:49 brb Nov 08 16:56:27 hi guys. i'm using actionbarsherlock and i want to make the title of my actionbar clickable. is there any way i can do this without hacking workarounds ? Nov 08 16:56:59 to add: i want to display the app logo and the title in the top left corner and both should be clickable Nov 08 16:57:12 just clicking the logo to go back is too picky Nov 08 17:01:41 Is there a way to over ride the default action bar to show a custom icon instead of the launcher icon, I want to load a banner into the top of the bar and then the icons into the lower bar Nov 08 17:02:23 docmur: did you consider using a customView for that ? Nov 08 17:02:53 As in an Image View Nov 08 17:03:48 docmur: you can set a customView using ActionBar.setCustomView(View view); Nov 08 17:03:56 Any idea why making a SSL connection to my site works on ICS in my app and my browser, but on 2.3.3 it works in my browser but on my app it throws a "11-08 12:02:42.237: W/System.err(4476): javax.net.ssl.SSLHandshakeException: org.bouncycastle.jce.exception.ExtCertPathValidatorException: Could not validate certificate signature." Nov 08 17:05:25 How would that work with a top and bottom action bar, Can I sent what appears in either top or bottom Nov 08 17:06:01 docmur: i ran into the exact same problem. customViews can only be displayed in the top actionbar Nov 08 17:06:19 you can workaround this by hacking your own fake bottom action bar Nov 08 17:06:34 or by adding a view to an menu item in the bottom action bar Nov 08 17:06:53 If I root my evo 3d and ics comes out in a week, will that unroot it? Nov 08 17:17:04 I made a layout with an image view in it and assigned the image inside the image view. When I do actionBar.setCustomView(R.layout.topbar) the image isn't there Nov 08 17:19:54 mcurran: If I'm understanding your inquiry correctly, the answer is a definite maybe. ;-) If you *just* root it, and run a carrier ROM, then it will probably still participate in OTA updates, possibly asking you first. If you install an alternate ROM/image like Cyanogen, then no, I believe not. Nov 08 17:27:28 i think if you modify the system image you're not going to get OTA updates Nov 08 17:27:36 regardless, this is a discussion for #android-root Nov 08 17:34:33 I have an xml pull parser, I can read one attribute of the tag correctly but not the second one? what could I be missing? Nov 08 17:39:49 TachyonDev, back? Nov 08 17:40:13 yeah im back Nov 08 17:40:23 dealing with jelly bean notifications Nov 08 17:41:15 going to put a like button in the notification .. not sure what else to put there though Nov 08 17:43:27 hi guys. i'm using actionbarsherlock and i want to make the title of my actionbar clickable. is there any way i can do this without hacking workarounds ? to add: i want to display the app logo and the title in the top left corner and both should be clickable, just clicking the logo to go back is too picky Nov 08 17:46:18 TachyonDev, sounds nifty, wish I had JB for my phone :) Nov 08 17:47:04 hkhalid: im itching to order the nexus 4 tues morning Nov 08 17:47:12 and a nexus 10 Nov 08 17:49:12 nexus everything Nov 08 17:49:55 TachyonDev, split the difference and get nexus 7 Nov 08 17:50:28 Sculptor: heh already have one Nov 08 17:50:35 high five Nov 08 17:50:38 :P Nov 08 17:50:51 i have one Nov 08 17:50:52 16gb Nov 08 17:50:55 i'm ordering 32gb Nov 08 17:50:59 and giving away my 16gb Nov 08 17:51:13 i have the 8gb from IO Nov 08 17:51:26 i want me google glasses Nov 08 17:51:27 price is $350, $100 more than in the us of a Nov 08 17:52:17 8gb is too small, considering that you get ~ 5gb after a factory reset Nov 08 17:53:09 im surving just fine with 8GB on my phone Nov 08 17:53:15 which is ~5 usable Nov 08 17:53:28 i made an app that uses a date-picker, its defined in the xml. im not sure how to get the data from it though Nov 08 17:53:42 cus its in the xml thing and no where in my java code Nov 08 17:54:02 TachyonDev, dang, I want one too :) Can I pay you a fee to help me getting this extra working? I added the putextra, and then in my fragment activity, I have the bundle, but that seems to get my app to crash as it doesnt like that part, http://pastie.org/5346848 the 2 sections Nov 08 17:54:27 hey? Nov 08 17:54:34 Reaga, have you ever heard of findViewById(int) ? Nov 08 17:54:36 Reaga: give it an id with android:id="@+id/mydate" in xml, then use findViewById(R.id.mydate); in your code to get a reference to it Nov 08 17:54:39 if not, please read android fundamentals Nov 08 17:54:42 ahh thanks Nov 08 17:55:41 hkhalid: Bundle extras = getIntent().getExtras(); Nov 08 17:56:22 the bundle in the oncreate is for the saveinstance of the current activity Nov 08 17:57:14 TachyonDev, where else could it be placed though? the oncreate is the only spot I have seen for it Nov 08 17:57:27 that line of code goes in the oncreate Nov 08 17:57:40 dont use the protected void onCreate(Bundle savedInstanceState) Nov 08 17:58:01 protected void onCreate(Bundle savedInstanceState) { Nov 08 17:58:01 super.onCreate(savedInstanceState); Nov 08 17:58:01 final Bundle extras = getIntent().getExtras(); Nov 08 18:01:04 hkhalid: make sense? Nov 08 18:02:39 TachyonDev, I missed that Bundle part for the saveInstanceState, but wasnt my Bundle totally seperate? I tried your new method, which is just making the var final correct? it still crashes, im checking again Nov 08 18:03:12 Did you read the stack trace? Nov 08 18:04:16 Im not sure if its eclipse, but I used the error filter for logcat and its staying blank Nov 08 18:04:41 Well, get the stack trace before you start randomly debugging Nov 08 18:08:00 I want to allow users to log into my app using openID, Facebook, or Twitter. Is there a library that has all of these? Nov 08 18:09:28 first google hit http://code.google.com/p/socialauth/ Nov 08 18:09:49 Thanks, what did you search on? Nov 08 18:11:07 using openID, Facebook, or Twitter android library Nov 08 18:12:59 i have to themes right now ... one's parent is @style/Theme.Sherlock and the other one's parent is android:Theme ... do you guys know how i can combine these both into one style ? Nov 08 18:13:12 there has to be some misunderstanding, why can I not getAttributeValue in my xmlpullparserr Nov 08 18:13:40 I can get one attribute, not the second one Nov 08 18:13:50 two* Nov 08 18:19:03 * kbs doesn't trust most OAuth implementations on android Nov 08 18:20:14 I think for facebook at least, it's less worse to use the facebook sdk than other alternatives. The typical problem is that OAuth clients embed a WebView, which means a nasty app can just snatch the password directly Nov 08 18:20:46 sweet phandroid picked us up Nov 08 18:20:48 http://phandroid.com/2012/11/08/groupme-4-0-for-android-brings-overhauled-experience/ Nov 08 18:29:11 TachyonDev, I got further, thanks dude! Nov 08 18:29:21 yeah np Nov 08 18:36:42 TachyonDev, DONE! Got it working completely Nov 08 18:36:46 thanks so much TachyonDev Nov 08 18:36:56 hkhalid: glad i could help Nov 08 18:44:41 Hey, why do variables have m's before them like mFinishIntent? Nov 08 18:45:29 Snicers-Work2: http://en.wikipedia.org/wiki/Hungarian_notation ? Nov 08 18:45:59 ist has nothing to do with hungarian notation? Nov 08 18:46:28 Snicers-Work2: its just the coding style guides of android. I don't like the M either :) http://source.android.com/source/code-style.html Nov 08 18:47:49 it's a handy indicator to see if a variable is a member variable or a local variable Nov 08 18:49:21 for some things it's obvious or not very relevant, like mContext is unlikely to indicate something different than context, but x versus mX you know mX is relevant to the whole object whereas x is possibly just some temporary variable or related to the current method Nov 08 18:52:00 There is nothing in the code style about member variables. Nov 08 18:55:40 the mVarName convention is pretty useful in multi-threaded code; you can quickly scan which variables have potential race conditions. Nov 08 18:55:49 Yes there is Nov 08 18:59:02 SimonVT : when I search for "member" I get nothing. Nov 08 18:59:18 How can I dynamically build a resource id? For instance, I am pass the string "pool", so I want to do imageView.setImageDrawable(R.id.pool_thumb) Nov 08 18:59:54 http://source.android.com/source/code-style.html#follow-field-naming-conventions Nov 08 19:00:00 fakingfantastic: if at all possible, you shouldn't do that Nov 08 19:00:01 They refer to them as non-public non-static fields. Nov 08 19:00:25 Snicers-Work2: which is pretty much the definition of a member variable... Nov 08 19:00:33 ok Nov 08 19:00:49 strange that they don't reference it as such is all I am saying. Nov 08 19:04:15 evancharlton: ok, i definitely will. Can you elaborate why? Nov 08 19:07:36 fakingfantastic: it's Resources.getIdentifier to get the id from a string, but it's slow which is one disadvantage. another disadvantage is you don't get compile time error checking, like if you delete pool_thumb because you thought it was no longer used, but your code referenced it directly, at compile time you'll get an error and can fix it. if you are referencing it dynamically, the error would happen at run time, possibly not by you Nov 08 19:07:36 an unsuspecting user. Nov 08 19:08:08 kevinb: really good stuff… thanks man Nov 08 19:08:42 fakingfantastic: i tried to do that and it works, but for some cases, like in txt files, the resources will be gzipped and hence you wont get the resoruce if with just "pool" Nov 08 19:15:54 hi Nov 08 19:16:27 has anyone ever debugged a native app ? I'm trying to debug one and the debugger starts but the breakpoint in native code is not hit Nov 08 19:16:43 If i set a breakpoint in java code it is hit Nov 08 19:17:18 I tried the Debug.waitForDebugger() trick but it didn't work Nov 08 19:18:25 xroberx: you can do some remote gdb stuff Nov 08 19:18:32 xroberx: otherwise it's pretty opaque Nov 08 19:19:16 dragorn: I've used gdb just a couple of times when I was learning C... what is that remote gdb stuff you're talking about ? Nov 08 19:19:32 xroberx: gdb from pc to android Nov 08 19:19:43 xroberx: search 'android ndk gdb' and you'll probabyl find something Nov 08 19:20:21 dragorn: you mean invoking ndk-gdb from the commandline while the app is running ? Nov 08 19:26:57 wow I think it's easier to cut some carrots with a chainsaw than using the debugger to debug native code Nov 08 19:27:33 *SNRK* Nov 08 19:29:14 cutting carrots with a chainsaw would be super easy Nov 08 19:29:23 now, cutting the in any sort of controlled fashion... Nov 08 19:30:00 evancharlton dem sparkles. That is all. Nov 08 19:30:14 5 days Nov 08 19:30:47 :) Nov 08 19:35:35 I love those Eclipse error dialogs: An internal error has ocurred "Null Pointer Exception" Nov 08 19:36:02 If hell exists, for sure they use Eclipse in there Nov 08 19:39:26 the enlightened surely live in the temple of saint ignucius Nov 08 19:39:46 hello, anybody knows about in app billing and public key association ? thanks Nov 08 19:42:30 from a datePicker, i use a a CalendarView cv.getDate() which returns a long value. i want to do this with a TimePicker but there is no .getTime() or .getDate(). only .getCurrentHour() and .getCurrentMinute() which return Integers Nov 08 19:43:42 TimePicker tp = (TimePicker) findViewById(R.id.timePicker1); Integer foo = tp.getCurrentHour(); Nov 08 19:43:50 i want foo to be a long of the milliseconds Nov 08 19:44:30 Touch luck Nov 08 19:44:30 looks like you get to make a long from integers Nov 08 19:44:34 Convert it yourself Nov 08 19:44:36 have fun in CS101 Nov 08 19:45:20 My public key has changed in last month… How it's possible ? Nov 08 19:45:26 why is it designed this way though? Nov 08 19:45:29 it seems so inconsistent Nov 08 19:45:43 it's not. A timepicker is to select a time in hours and minutes Nov 08 19:45:52 how do i get times from it Nov 08 19:46:26 just with a calendarview Nov 08 19:46:29 my in app billings not works before this change… ='( Nov 08 19:47:01 hmm ok Nov 08 19:47:20 why would you have a method in a class with the same name as the class? Nov 08 19:47:46 illusyon: I would ask /dev/null (ejem... Google) about it Nov 08 19:48:04 Snicers-Work2, those methods are called constructors Nov 08 19:48:16 and why you would have them, is quite obvious, isn't it?! Nov 08 19:48:32 Thought it was, this one is empty (picking through someone elses code) Nov 08 19:48:54 so the constructor has no logic to it, does the class execute onCreate then? Nov 08 19:49:02 i'm search it, but I don't get any specific results about it Nov 08 19:49:19 xroberx my problem is a little specific Nov 08 19:49:24 it will call the super-constructor. It is an activity I assume? Nov 08 19:50:06 It's a fragment Nov 08 19:50:46 xroberx i'm know if I change this key it will works, but I don't know how this public key has changed Nov 08 19:51:06 illusyon: it should never change Nov 08 19:51:29 illusyon: but good luck getting any kind of support from Google... Nov 08 19:51:38 Is there any reason I wouldn't be able to load a searchbar on top of a webview in a scrollview? Nov 08 19:53:08 Can anyone point me to a good tutorial on OpenGL ES 2? Nov 08 19:53:22 I'm trying to figure out how to use an array of colors instead of just one... Nov 08 19:53:23 I'm accessing a static variable of my Service from an Activity, and want to implement some sort of synchronization - doing: if (MyService.object != null) { MyService.object.run(); } isn't necessarily safe, right? Nov 08 19:53:46 Since the Service could have set the object static variable to null after the null check of the Activity. Nov 08 19:53:52 xroberx i'm agree, but my has changed ! On google i'm get some cases about it, of the public keys has changes, but nobody knows why =/ Nov 08 19:53:56 dammit Nov 08 19:54:02 I don't know if there is a best practise way to do this in Android though. Nov 08 19:54:40 thanks.. Nov 08 19:55:07 illusyon: yeah, you have been fscked up Nov 08 19:56:05 ejcweb: your activity and service have the same main thread. if you're not using asynctasks or Threads or similar then you don't need to worry about thread synchronization. if you are then you need to worry about it the same way any java app would Nov 08 19:57:29 kevinb: Great, thanks. That said, my null check above is still susceptible to throwing a NullPointerException if the static variable becomes null (as set from the Service) after the Activity has checked the null condition? Nov 08 19:57:57 I'm not sure if I should be doing something clever to prevent this from happening. Nov 08 19:59:15 assuming the doSomething is in the main thread and if you set object to null from another thread then you need to do something, if you only set it to null from the main thread you don't Nov 08 19:59:45 onClick onResume etc will all run in the main thread. Your service's onCreate and onStartCommand also all run on the main thread Nov 08 20:02:11 but if your service's onCreate does something like spawn an async task which in doInBackground might change "object", then you need something in your activity. A simple, but not thorough, test would be to add some SystemClock.sleep(1000) in places. You'll see that if your service sleeps for a second that your activities UI is not responsive for that second. and that if you sleep between the null check and doSomething you're still okay a Nov 08 20:02:11 other main thread code is running. Nov 08 20:02:17 Perhaps a classic question. Nov 08 20:04:14 is there a standard way of doing image manipulation on large bitmaps in android? Nov 08 20:04:18 I'm trying to run my first app which i'm learning through the 'building my first app' tutorial on developer.android.com site. I'm calling the following string: android:hint="@string/edit_message" but i get the error that he cant find the resource. Nov 08 20:04:20 One could learn a great deal just by reading everything here... Nov 08 20:04:29 like loading the images in chunks or something Nov 08 20:04:52 the google developer account are annual ? 25$ every year ? Nov 08 20:04:55 kevinb: Maybe I'm missing something. To be sure, if I do this: if (MyService.object != null) { // I do some stuff here for 100ms, but MyService.object might now equal null by the end of it } - even if I'm not doing anything async or in threads? Nov 08 20:05:06 PsychoX75: that string has to be defined in your strings.xml file in res/values Nov 08 20:05:12 sampullman: when you say manipulation, are you just trying to resize and stuff? Nov 08 20:05:18 and i did declare the strings in the strings.xml. At a certain point i literally copy pasted the codes to both files. Nov 08 20:05:23 kevinb: As in, I want to stop MyService.object being changed until I try to use it. Nov 08 20:05:32 Anyone had any luck with including google_play_services without the use of Eclipse? Running on Sublime Text and I'm not sure what to include Nov 08 20:05:44 so m activity_main.xml and strings.xml looks exactly as on the site. Nov 08 20:05:47 bluescales: no, but the effects should be possible to do on chunks of images at a time without losing info Nov 08 20:06:07 mostly feature detection along with affine transformations Nov 08 20:07:18 should i 'tell' activity_main.xml to use the strings.xml file somewhere? Nov 08 20:07:28 ejcweb: if you are not using any kind of threading, then all your code is blocking. so no objects can change out from underneath you. Services do not run in a background thread by default. Nov 08 20:07:32 sampullman: I know bitmapfactory.options lets you do all kinds of stuff to sample limitedamounts of image data, take a peek there maybe? Nov 08 20:08:00 sciepy, if you ever find one, let me know Nov 08 20:08:52 hmm.... i decided to put a "+" sign in it like this: android:hint="@+string/edit_message". Now it compiles but in the app where it's supposed to show the string, it says "FALSE". Nov 08 20:09:14 sampullman: u use Teamviewer? Nov 08 20:09:55 bluescales: yeah, i know how to load large bitmaps by sampling, but i want to do transformations on the hi res images and save them back to disk. i suppose i can do most of the computation with a sampled version, but stitching the hi res images together would still be an issue Nov 08 20:10:00 kevinb: Ah, I understand now. Does it block for the execution of a whole method then? Nov 08 20:10:10 PsychoX75: no, what's that? Nov 08 20:11:00 :S okays... Ehm it's a way for you to remotely view my screen. Something like PCAnywhere, but without the headaches of PortForwarding/Firewall tweakings. Nov 08 20:11:01 PsychoX75: if you post your code using the string and your strings.xml i might be able to help Nov 08 20:11:02 ejcweb: yes, it's more like there's an internal while loop that checks for events and dispatches calls to them. so it can't dispatch new events until your previous event handler finishes. Nov 08 20:11:36 no, i'm using opencv to play with some augmented reality stuff Nov 08 20:11:44 sec Nov 08 20:12:44 kevinb: Thanks for explaining, it makes sense now! :) Nov 08 20:12:49 sampullman: if you haven't seen it already there's BitmapRegionDecoder in API 10 http://developer.android.com/reference/android/graphics/BitmapRegionDecoder.html Nov 08 20:12:57 http://pastebin.com/6tVPhasb Nov 08 20:13:10 sampullman: yeah, I'm not sure about that, I'm afraid Nov 08 20:13:35 bluescales: no worries, thanks anyway Nov 08 20:14:16 kevinb: i think that's exactly what i need, cool. supporting api 7 was my goal, but that may not be possible Nov 08 20:14:20 anyone seen any strange activity on their developer console stats? Nov 08 20:15:07 strange Nov 08 20:15:12 I put my app on slideme at the start of last month, and since then i have a rapidly increasing number of installs on play from thailand, the majority of which are shortly uninstalled Nov 08 20:15:37 it doesnt really look realistic Nov 08 20:18:15 PsychoX75: it seems like you're referencing the string correctly. what's the issue again? is the app compiling? Nov 08 20:20:30 He's adding a string resource id without actually adding a string resource.. Nov 08 20:21:19 xrogerx i'm find my error ! thanks ! Nov 08 20:21:34 sampullman: the app is compiling if i Hardcode text in it. Nov 08 20:21:42 Like i did in the last sentence Nov 08 20:21:52 it's possible make in app billings from my account with package products of other account ? Nov 08 20:21:57 xrogerx Nov 08 20:22:00 SimonVT: it's in his strings.xml, though Nov 08 20:22:16 but as soon as i reference to it with @string like that, it gives me an error "No resource found" Nov 08 20:22:25 as if it doesnt realize that i defined the resource at all. Nov 08 20:23:18 try doing a project clean and then building again Nov 08 20:23:25 kk Nov 08 20:26:18 any coincidence the new oreilly book 'bad data' has a goose on it ? i doubt it :| Nov 08 20:27:30 hahah.... sorry... now i know what went wrong. I had more projects open. I was sitting and editing the strings.xml file of ANOTHER project. Nov 08 20:27:53 Eclipse opened the two files next to each other for me, and i automatically thought that they were related. :P Nov 08 20:27:56 thanks allot man Nov 08 20:28:26 now since I spend the last 2 hours trying to figure this out.... my braincells are busted. Gonna head to bed with a glass of whine and a nice movie. Nov 08 20:28:34 C ya all tomorrow ppl. Nov 08 20:31:30 hhmm I've got an ImageView, and I want it to scale the image to fill the View's width, maintain aspect ratio, and just crop nay height that doesn't fit. fitStart almost works, and centerCrop almost works, but they both arent quite what I need. any ideas? Nov 08 20:43:32 hi Nov 08 20:44:02 is there a preferred JDK version (6 or 7) for getting started with android dev? Nov 08 20:44:47 I'll say 6 Nov 08 20:45:03 Yeah Nov 08 20:45:03 thanks Nov 08 20:45:56 should i start with learning java or with an android dev book? is there a "best" way for a beginner in mobile app dev? Nov 08 20:47:54 id recommend starting by reading the documentation at developer.android.com Nov 08 20:48:57 anyone ever used a listfragment or fragmentactivity and it doesnt appear Nov 08 20:49:06 a blank activity w/o errors Nov 08 20:49:43 is it possible to center the actionbar icon Nov 08 20:49:46 You should definitely learn the java language first Nov 08 20:50:33 ditoa: do you have any previous programming experience? Nov 08 20:51:36 SimonVT, me? Nov 08 20:51:46 If you don't know java, sure Nov 08 20:52:03 maraz: only some C a decade ago Nov 08 20:52:08 ah Nov 08 20:52:20 then you should definitely read up on java; it is a very different language Nov 08 20:52:47 SimonVT, are there people here paid by Google to give support or no Nov 08 20:52:55 yeah Nov 08 20:52:55 no, there are not Nov 08 20:53:06 read the docs Nov 08 20:53:11 everything is there. Nov 08 20:53:17 maraz, calm down please Nov 08 20:53:19 then there's stackoverflow... and google Nov 08 20:53:25 * jakob| is waiting for compat lib 11 Nov 08 20:53:56 what is in compat lib 11? Nov 08 20:53:58 i am really tempted to pick up a nexus 4 on tuesday and get into app development Nov 08 20:54:08 the oreilly book learning java is brilliant imho Nov 08 20:54:14 just tempted? Nov 08 20:54:22 very tempted :P Nov 08 20:54:28 like 99% sure i will buy one Nov 08 20:54:43 hkhalid: No, they are not paid to give support Nov 08 20:55:09 They're here of their own free will Nov 08 20:55:22 how much reading should i do before starting my first app Nov 08 20:55:54 what a strange question Nov 08 20:56:26 if you never made an app before, then you have to do all the reading Nov 08 20:56:33 The paid component of support is that they pay someone to write the docs and examples Nov 08 20:56:36 what is "all the reading"? Nov 08 20:56:45 developer.android.com Nov 08 20:56:51 all of that Nov 08 20:56:57 Design guidelines, best practices. Nov 08 20:57:06 theTerry: thats a lot to read Nov 08 20:57:09 All the important things to consider before just starting to make an app Nov 08 20:57:13 every class? Nov 08 20:57:34 Not the API doc Nov 08 20:57:34 Reaga: start with the hello world and android example stuff on developer.android.com. Then figure out what you actually want to do. Nov 08 20:57:46 that is a good suggestion Nov 08 20:57:59 i did that Nov 08 20:58:10 but im stuck on something quite basic, and its making me question my ability to program Nov 08 20:58:31 any dutch developers here who have published their app and charging money for it? Nov 08 20:58:56 I have been programming for 15 years and I still question my ability to program Nov 08 20:59:17 im suck on making a list appear in a dialog Nov 08 20:59:22 a list that is dynamic Nov 08 20:59:31 i have to mess around with adapters and stuff to make it work Nov 08 20:59:40 ListAdapte Nov 08 20:59:41 r Nov 08 20:59:54 if you are making a list then messing around with adapter is what goes on Nov 08 21:00:03 but it seems so complex Nov 08 21:00:09 it's java Nov 08 21:00:14 nah its android Nov 08 21:00:52 well keep at it and good luck Nov 08 21:02:30 tbh i found the data models in swing more straight forward Nov 08 21:02:33 ArrayAdapters aren't that hard after you understood the basic concept once. Nov 08 21:02:35 wat Nov 08 21:02:40 String is an abomination Nov 08 21:02:43 *Swing Nov 08 21:02:56 your not wrong Nov 08 21:03:46 it took me ages to follow how list items were re-used, atleast with swing you write the renderer and it simply does what you tell it to Nov 08 21:05:01 mainly because you normally only have a single list item, granted :P Nov 08 21:05:36 romainguy gave a really explanation of it at I/O in 2009. It really should be part of the javadocs for ListView/Adapter Nov 08 21:06:04 im happy enough with it now, but might look out for that if its available online just to make sure Nov 08 21:06:40 sorry, it was 2010: http://www.google.com/events/io/2010/sessions/world-of-listview-android.html Nov 08 21:06:43 at least I'm pretty sure that's it Nov 08 21:06:43 ty Nov 08 21:07:05 oh yes, the PDF references Wave; it must be the right year Nov 08 21:07:52 Reaga, ever done a list in a fragment? Nov 08 21:12:08 another of my big take-aways from android development is that you need to pay a hell of a lot of attention to the hand-over between your app and various parts of the API, in a way that you wouldnt normally Nov 08 21:16:39 I recently learned the hard way that implementing onSaveInstanceState shouldn't be optional Nov 08 21:16:50 hkhalid: ive never done a fragment Nov 08 21:18:22 terry, its really quite an ugly mechanism isnt it Nov 08 21:18:23 wow an hour long video on listview? Nov 08 21:18:23 jesus chhrisst superstar Nov 08 21:18:47 patience, padawan Nov 08 21:18:52 you must have patience Nov 08 21:19:15 yer, err, what project do you book that to again? Nov 08 21:22:08 ehey Nov 08 21:23:38 hi Nov 08 21:23:41 finally Nov 08 21:23:46 had to ghost myself Nov 08 21:23:50 anywaaays. Nov 08 21:24:01 hi, i dont have file>new>android project Nov 08 21:24:09 i have new>project>Android/folder Nov 08 21:24:35 longstory short i've been trying to use phonegap Nov 08 21:24:55 i followed all the steps but still cant get it to work Nov 08 21:25:11 any help would be appreciated Nov 08 21:34:32 veQue: maybe there is a channel for phonegap Nov 08 21:34:45 g00s: yup i'm there Nov 08 21:34:50 =/ Nov 08 21:35:03 ever google all day for an answer? Nov 08 21:35:07 i'm at that point Nov 08 21:35:34 might find a solution soon i hope Nov 08 21:36:54 File->New->Other->Android/ Nov 08 21:36:55 ? Nov 08 21:38:06 just found answer on stack overflow... that's a peice of my puzzle oslved Nov 08 21:38:18 btw had to change perspectives from javaee to java Nov 08 21:38:48 thanks for helping sonOfRa Nov 08 21:40:20 * hackkitten hands veQue an energy drink :) Nov 08 21:42:41 cheers Nov 08 21:43:49 is there a simple way to create an AVD? I just built the android OS from source and there is a tool called "emulator" but it is command line Nov 08 21:43:57 is there a GUI for it so that I can create an AVD? Nov 08 21:49:43 yup Nov 08 21:50:01 sdk/tools/android Nov 08 21:50:09 thanks sampullman Nov 08 21:50:57 is that in the out directory? Nov 08 21:53:54 I think I found it Nov 08 21:54:17 it is in out/host/linux-86/bin/android Nov 08 22:00:43 So, I added sliders to an app Nov 08 22:00:49 And now it just crashes when I start it Nov 08 22:01:19 I can't tell why, but it seems like simply trying to make an instant of them in the .java in order to read them automagically is causing instant crashing Nov 08 22:01:21 Any ideas? Nov 08 22:07:36 have you checked logcat already? Nov 08 22:08:15 I can't get the app to run at all in the emulator Nov 08 22:08:16 * ctate points to the imgur link in the channel topic. Nov 08 22:08:23 ? Nov 08 22:08:44 http://imgur.com/jacoj Nov 08 22:08:45 Ahhh Nov 08 22:08:47 :) Nov 08 22:08:51 * Striker121 got there a second before Nov 08 22:09:01 Wouldn't know how to.... :( Nov 08 22:09:14 Although, being far less generic, I found an issue with a specific thing. Nov 08 22:09:17 it's printed to the log when your app crashes Nov 08 22:09:24 How do I read the log? Nov 08 22:09:33 there's a view in the Eclipse integration stuff Nov 08 22:09:41 or from a terminal you can run "adb logcat" Nov 08 22:10:02 [is this not actually documented in the SDK?!] Nov 08 22:10:06 The thing is bluetooth so I need to run it from the actual phone Nov 08 22:10:16 I couldn't find a single adequate documentation of how to do the USB debugging Nov 08 22:10:31 But, with a specific one line issue, THIS crashes the app as well Nov 08 22:10:43 if you can install the app over usb, then you're already set up for usb debugging :) Nov 08 22:10:43 leftSlider.setProgress((int)(leftSlider.getMax() / 2)); Nov 08 22:10:57 i'm guessing that "leftSlider" is null. Nov 08 22:10:59 I don't know how to install over USB, I've been using dropbox Nov 08 22:11:04 It isn't Nov 08 22:11:23 It's created as a null when the app is started Nov 08 22:11:30 Then the onCreate makes it reference the slider Nov 08 22:11:46 leftSlider = (SeekBar)findViewById(R.id.leftSlider); (In onCreate) Nov 08 22:12:12 I suppose there's one thing though Nov 08 22:12:16 that can return null Nov 08 22:12:29 Hmm.... Nov 08 22:12:37 Gah, debug statements make everyhting so much frakking easier Nov 08 22:12:39 in particular, if the ID passed in was not actually declared by the layout you set as your content view Nov 08 22:12:47 fwiw, you really should read http://developer.android.com/tools/workflow/index.html Nov 08 22:12:49 I might leave this until I have a micro USB cable (IE: Back in my dorm) Nov 08 22:13:02 which includes e.g. how to debug on actual hardware Nov 08 22:13:08 Thanks Nov 08 22:13:31 my god, Amazon App Store For Android is a desert Nov 08 22:13:33 android:id="@+id/leftSlider" has it in the layout :( Nov 08 22:13:46 Why use Amazon over the default? Nov 08 22:14:20 I use both Nov 08 22:14:36 Welp, I hope the issue is with the slider declaration Nov 08 22:14:37 but the stats I see in Play outweigh those at Amazon's by ..mountains Nov 08 22:14:44 Because that would explain why quite a lot of stuff is crashing it Nov 08 22:14:50 No idea why it would return null though Nov 08 22:15:27 you'd need to put some code (your onCreate() code in particular) and the layout file on pastebin and show us Nov 08 22:15:39 usually the reason is that you didn't inflate the layout you thought you did Nov 08 22:15:44 or that layout botches the id declarations Nov 08 22:15:46 stuff like that Nov 08 22:15:46 when I am trying to run "android" to create an AVD, it is telling me that it cannot locate SWT.jar for x86_64 ... do I need to do something special Nov 08 22:15:52 to build SWT.jar for 64-bit? Nov 08 22:15:53 There's only one layout Nov 08 22:16:20 or e.g. you're trying to call findViewById() *before* calling setContentView() -- i've seen that more than once Nov 08 22:16:44 like i said, share some code and people can help :) Nov 08 22:17:00 Will do. Definitively did setContentView before Nov 08 22:17:20 And this is based off another app I did where it had a textbox declared so I can't imagine how the view would be incorrect :( Nov 08 22:20:18 http://pastebin.com/D03t8eRR Nov 08 22:20:26 and this http://pastebin.com/YuUMUjLt Nov 08 22:20:55 Interesting.... Nov 08 22:21:06 I just changed the name in the layout and in the .java to leftSlider1 Nov 08 22:21:15 And NOW it gives me an error saying it doesn't exit... Nov 08 22:21:59 I've compiled the AOSP from scratch, which gave me a system.img, a userdata.img, and a ramdisk.img. How do I test these img files in an emulator? Nov 08 22:22:04 you're using Eclipse? Nov 08 22:22:10 Nevermind, clean and rebuild fixed that particular error Nov 08 22:22:14 And yes, Eclipse Nov 08 22:22:19 it sounds like you're hitting a bug that people seem to hit a lot, where it forgets to rebuild everything Nov 08 22:22:23 so yeah, clean & rebuild Nov 08 22:22:34 i don't see anything obviously wrong, fwiw Nov 08 22:22:39 Hmm.... Nov 08 22:22:55 Damn, this laptop is dying, guess I'll just have to try getting usb debugging set up Nov 08 22:23:15 Gotta get it done for tomorrow. Nov 08 22:23:40 It still amazes me the quick code testing if you use Dropbox for the files Nov 08 22:23:48 So much freaking easier than iPhone programming Nov 08 22:24:23 See you guys later Nov 08 22:25:38 Does anyone know how to test compiled AOSP in an emulator? Nov 08 22:25:40 sleepster: i haven't built android from source; but fwiw, swt.jar is a standard download you ought to be able to get from http://www.eclipse.org/swt/ You ought to be able to drop it into the folder that shell script looks at. Nov 08 22:26:33 Oh wow Nov 08 22:26:35 ;lingin4oignoelnrs Nov 08 22:26:41 Per usually, keyboard mashing is my only response Nov 08 22:26:54 Naming the sliders "leftSlider" and "rightSlider" crashes it Nov 08 22:27:06 ... Nov 08 22:27:06 But naming them "leftSlider1" and "rightSlider1" does NOT crash it Nov 08 22:27:20 namespace collision? Nov 08 22:27:23 hunh Nov 08 22:27:35 Could the issue be with the object in the layout being named the same as the object in the .java? Nov 08 22:27:47 *conceivably* Nov 08 22:27:51 I didn't think that could happen. I thought eclipse would at least give a build error Nov 08 22:28:12 Eclipse seems to be really strange with giving and not giving errors for Android code..... Nov 08 22:28:23 Oh well, that's working for now. Nov 08 22:28:29 Just got to ge thte commented out stuff running Nov 08 22:28:31 it's very odd, yeah Nov 08 22:28:34 See you and thanks for the help. Nov 08 22:28:37 np :) Nov 08 22:28:38 Has anyone used face detection? I'm kind of suprised getMaxNumDetectedFaces() returns 0 on my Nexus 7 Nov 08 22:30:01 once we build android OS on a 64-bit ubuntu machine, are we expected to be using the 32-bit JRE to run the emulator/android tool? I am noticing that I can't run them without installing the 32-bit version of java Nov 08 22:30:13 im using face detection, but not through android api's Nov 08 22:30:24 I've compiled the Android source. Does anyone know how to test the output on an emulator? Nov 08 22:30:39 Jonathan_Eyre: I am at the same step :) Nov 08 22:30:48 Jonathan_Eyre: I believe you have to create an ADT first, then run the "emulator" tool on it Nov 08 22:30:51 run 'emulator' from the same shell where you ran the 'make' that built the platform Nov 08 22:31:02 Jonathan_Eyre: err. AVD Nov 08 22:31:04 I mean Nov 08 22:31:16 create an AVD and then run the emulator on it I believe Nov 08 22:31:20 sampullman: did you roll your own OpenCV solution or did you use something else? Nov 08 22:31:22 i do not know this newfangled avd thing :) Nov 08 22:31:35 opencv, yeah Nov 08 22:32:11 once you get it running it's just a fine lines of code to do face detection Nov 08 22:32:15 ctate: I think that emulator needs a virtual device name. Nov 08 22:32:15 *few Nov 08 22:32:23 ctate: sorry, maybe I've made a msitake :) I just finished building the android OS as well and when I run the emulator it is asking me to point it at an AVD Nov 08 22:32:53 which I've determined I guess needs to be created with the "android" tool Nov 08 22:33:42 huh, okay. there you go, then. Nov 08 22:33:59 Jonathan_Eyre: it didn't four years ago; my how things change :) Nov 08 22:34:40 sleepster: What is the 'android' tool? That's like saying that in order to launch a batch file, you need to run the 'microsoft' program. Nov 08 22:34:58 yeah that's confusing me too Nov 08 22:35:19 i'm used to being able to 1) repo sync the OS sources, 2) lunch the right thing, 3) run "make", 4) run "emulator" Nov 08 22:35:43 ctate: I am very jealous. Nov 08 22:35:47 for AOSP sources, the proper lunch target is probably "full-eng" or "full-userdebug" Nov 08 22:36:04 (or whatever hardware product you're building for) Nov 08 22:36:13 it is a complete surprise to me that this might not work any more, seriously Nov 08 22:37:08 note that you have to run 'emulator' from the root of the source tree, and be configured via 'lunch' for the target that you just built / want to run. Nov 08 22:37:17 ctate: should the 'run-as' command be available and run from any phone ? Nov 08 22:37:31 g00s: my impression is yes Nov 08 22:37:35 ok, thx Nov 08 22:37:47 but iirc you can't run it *from* the phone, you can only run it via adb Nov 08 22:37:54 yes Nov 08 22:37:54 What is a handle in Android? Nov 08 22:38:08 Snicers-Work2: Android doesn't really use that term much; do you have some context for the question? Nov 08 22:38:12 wouldn't it only work if adb was running as root? Nov 08 22:38:15 I would assume that, anyway Nov 08 22:38:22 A handle is usually a windows thing. Nov 08 22:38:24 evancharlton: no, the point of run-as is to do that *without* requiring root Nov 08 22:38:36 Or a windows programmer talking about an android thing. Nov 08 22:38:43 Jonathan_Eyre: oh sorry, there is actually a tool called "android" in the out/host/linux-x86/bin folder Nov 08 22:39:09 im bummed. i gotta finish an android app for class.....and another API mashup.... Nov 08 22:39:14 ctate: , I thought it was odd. The Eclipse help says "Return the handle to a system-level service by name." for activity.getSystemService Nov 08 22:39:25 like bing or google maps api with some dynamic data behind it (not just a buncha static pushpins ) Nov 08 22:39:26 ha ha wow, ancient comment Nov 08 22:39:42 it gives you the object, not a "handle" Nov 08 22:39:54 haha, I saw that and was instantly confused. Nov 08 22:40:09 So basically it just returns an object that you can base conditions off of? Nov 08 22:41:48 no, it returns the service interface object that you then interact with Nov 08 22:42:04 i dunno what you mean by "base conditions off of" Nov 08 22:43:07 I mean you can do a if(!network.isConnected()){logic here;} sort of deal. Nov 08 22:43:24 I am sorry ctate, so I think I might be making this much more complicated than it needs to be :) so I run "source build/envsetup.sh" to setup my environment. Then, from the root of the source tree, I run "out/host/linux-x86/bin/emulator -avd MyDevice", it tells me "PANIC: Could not open: /home/user/.android/avd/MyDevice.ini" Nov 08 22:43:38 you skipped a step Nov 08 22:43:47 ctate: oh this was already made Nov 08 22:43:51 why does android want to import R all of a sudden Nov 08 22:43:52 after sourcing the envsetup.sh file, you then run something like "lunch full-eng" Nov 08 22:43:57 ah Nov 08 22:44:08 'lunch' is the alias that configures your build to the right product Nov 08 22:44:20 if you run just 'lunch' you get an abbreviated menu of likely possibilities Nov 08 22:44:54 only after you've run the 'lunch' thing can you actually build the OS properly (after which you should be able to run 'emulator' and it'll know from the environment what parameters to use) Nov 08 22:45:00 ctate: I see. thanks. so then I do lunch full-eng Nov 08 22:45:10 yep Nov 08 22:45:14 and then emulator Nov 08 22:45:15 THEN you do 'make' Nov 08 22:45:19 note that I already made it Nov 08 22:45:19 k Nov 08 22:45:21 then you do 'emulator' Nov 08 22:45:25 I'll run that again Nov 08 22:45:28 did you make it after doing "lunch full-eng"? Nov 08 22:45:36 because if not, then you built the wrong configuration Nov 08 22:45:40 and will have to (mostly) start over. :( Nov 08 22:45:51 remember to use e.g. -j16 or whatever is suitable, when you build :) Nov 08 22:46:00 :) yeah it's a long build Nov 08 22:46:07 what is the current build time record for the latest aosp? Nov 08 22:46:25 we have some machines in house that can do a full build from scratch in ~ 5 minutes Nov 08 22:46:31 damn Nov 08 22:46:34 and that's of the full product, not just the AOSP bits Nov 08 22:46:41 thanks ctate for the help. so I built it before and everything completed successfully.. will I have to do those steps each time? Nov 08 22:46:41 hp z620 with tons of ram etc Nov 08 22:46:57 they run linux, unfortunately, but at least they're superfast Nov 08 22:47:04 i have a nice i5-3570k overclocked to 4ghz, 16gb ram, samsung 256gb ssd. 3 monitors Nov 08 22:47:06 i spoil myself Nov 08 22:47:12 sleepster: you always have to source envsetup.sh and do the 'lunch' thing Nov 08 22:47:15 my launch time is amazing for the android sdk Nov 08 22:47:21 yeah my work computer can do it in about 15 or so Nov 08 22:47:26 but if you just built it with that same lunch configuration, it'll still be up to date Nov 08 22:47:54 its a 2x6-core xeon, 48gb of ecc, 2xSSD in RAID0 Nov 08 22:48:03 my work machine is relatively slow to build by modern standards; it's only 6 cores / 12 gigs of ram [the ram hurts a lot] Nov 08 22:48:15 yeah no hot build for you Nov 08 22:48:19 right Nov 08 22:48:29 otoh i don't want to downgrade to linux, so i suffer. :) Nov 08 22:48:34 heh Nov 08 22:48:44 my speed was in a VM in Windows Nov 08 22:48:44 :P Nov 08 22:49:07 ah, we don't run Windows at work. Nov 08 22:49:15 http://xkcd.com/303/ Nov 08 22:49:17 yeah, well you know why we have it at my work >_> Nov 08 22:49:23 :) Nov 08 22:49:33 TheTrash: so true Nov 08 22:49:39 Indeed :D Nov 08 22:49:42 though we do get to choose a new work computer Nov 08 22:50:37 the samsung ativ smart pc pro 700t looks interesting Nov 08 22:51:10 http://www.samsung.com/us/computer/tablet-pcs/XE700T1C-A04US Nov 08 22:51:34 windows 8 is painful without a damn touchscreen Nov 08 22:52:00 ctate: don't suppose you gave our new beta a whirl? Nov 08 22:52:10 gnope Nov 08 22:52:30 sorry Nov 08 22:52:36 i hope its on the path to another feature :) Nov 08 22:54:03 how do i change the order in which pages in my app load first? Nov 08 22:54:09 i have a splash screen that isnt working right Nov 08 22:54:36 nvm i found it Nov 08 22:55:37 from a wiped 4.1 emulator I keep getting "Failure[INSTALL_FAILED_DEXOPT]", is there a known thing that can cause this? Nov 08 23:01:07 Assuming that the AOSP I compiled works splendidly in the emulator, can I just create a boot.img and flash it to a device? Nov 08 23:01:50 I have hijacked the home screen long press by auto closing it with ACTION_CLOSE_SYSTEM_DIALOGS when the activity onWindowFocusChanged. Problem is this also cancels the ability to power off the device. Is there a way to only target the home screen long press dialog? Nov 08 23:02:12 Jonathan_Eyre: you need drivers Nov 08 23:02:12 ctate: extract-files still the way to go, or downloading them, or what, its been awhile Nov 08 23:02:40 SpNg: the fuck.. Nov 08 23:02:53 TachyonDev: ? Nov 08 23:03:38 TachyonDev: ? Nov 08 23:03:39 Jonathan_Eyre: http://source.android.com/source/building-devices.html Nov 08 23:04:08 you need to extract the binaries for the device you are building for Nov 08 23:04:10 if you're asking about getting the source tree, "repo sync -j64" or whatever Nov 08 23:04:12 go nuts :) Nov 08 23:04:14 oh, that Nov 08 23:04:19 yar Nov 08 23:04:37 i know nothing about dealing with hardware where conveying the proprietary binaries is an issue, sorry Nov 08 23:05:10 in general i would assume that you're on safer legal ground if you extract from your own handset, but I'm by no means a lawyer Nov 08 23:05:14 Jonathan_Eyre: just follow the instructions and either get the binaries from the google nexus page or extract them with the script .. if you aren't using a nexus device, then yeah you are boned Nov 08 23:05:21 https://developers.google.com/android/nexus/drivers Nov 08 23:05:51 * TachyonDev is excited to order his Nexus 4 on tuesday Nov 08 23:05:59 if it's for our devices where we ship those binaries, go ahead and get them from official channels, sure Nov 08 23:07:05 alright im tagging out .. later gaters Nov 08 23:07:12 SpNg: no. Nov 08 23:08:24 Could I pull the kernel and hardware-specific libraries off the device, build them inot the ROM, and put the ROM on the device? Nov 08 23:08:25 ctate: ah I didn't know you can use multiple threads for syncing Nov 08 23:08:29 repo sync took forever! Nov 08 23:08:34 evancharlton: ok thanks. Nov 08 23:09:18 sleepster: I generally use -j8 for syncing Nov 08 23:09:27 I only have 4 cores Nov 08 23:10:35 presumably it has some hyperthreading abilities Nov 08 23:10:44 guess IO is usually the bottleneck anyway Nov 08 23:10:57 git isn't too cpu intensive Nov 08 23:11:30 angelos: yeah, writing to HDD is the suck Nov 08 23:12:17 I read it was possible to pull the kernel and manufacturer supplied libraries off the device,a nd then use those to compile. Nov 08 23:12:26 Is that not true? Nov 08 23:12:54 Jonathan_Eyre: sounds plausible (doesn't AOSP have some extract_binaries.sh script?) Nov 08 23:13:10 Jonathan_Eyre: the people in #cyanogenmod-dev or #android-root would know more Nov 08 23:13:12 Yes, it does. Nov 08 23:27:00 hhmmm not possible to create a dialog with one Activity and then display it on another huh? Nov 08 23:27:39 is there a quicker way to test my themes out then closing and relaunching the emulator all the time? Nov 08 23:27:47 Wavesonics: no reason to Nov 08 23:28:01 well there is. but might have to work aroung it. Nov 08 23:28:29 no, it doesn't make sense to do that. if you want to show the same dialog in a second activity, recreate it Nov 08 23:29:39 or even better, use a dialogfragment Nov 08 23:34:21 Is there a good resource for android training videos? Nov 08 23:34:32 the new boston on youtube is awesome Nov 08 23:34:34 just google it Nov 08 23:34:40 k Nov 08 23:36:04 Chainfire: you around? Nov 08 23:36:54 what dev environment do you guys use? Do you guys just use a Windows Host and then build Android in a VM? Nov 08 23:37:00 or do you guys build on linux Nov 08 23:37:06 i.e. the host is linux Nov 08 23:37:42 * ctate uses a Mac Nov 08 23:37:52 I am on Windows at the moment and test directly on my Android phone and tablet, I also use the AVD's to test different os versions and sizes. Nov 08 23:38:16 Snicers-Work2: are you build an app? or the entire OS Nov 08 23:38:24 I need to build the entire android source Nov 08 23:38:27 for app development, the vast majority of people use windows Nov 08 23:38:36 but for Android *OS* development, that is not an optino Nov 08 23:38:40 *option Nov 08 23:38:41 I see, yeah that makes sense Nov 08 23:38:45 Ah, I am just doing an app, I assume you are bulding from AOSP? Nov 08 23:38:57 yup Nov 08 23:39:06 AOSP = Android OS Phone ? Nov 08 23:39:08 Yeah, I would use Linux for that. Nov 08 23:39:13 Android Open Source Project. Nov 08 23:39:20 hah Nov 08 23:39:21 makes sense Nov 08 23:39:31 http://source.android.com/ Nov 08 23:39:40 * ctate works on Android at Google, so.... ;) Nov 08 23:40:04 I just hate working inside linux.. but even worse, I hate working on linux inside a VM Nov 08 23:40:18 why the linux hate? Nov 08 23:40:27 you should try git on windows, you'll love linux :P Nov 08 23:40:32 poor video driver support mostly Nov 08 23:40:32 * Snicers-Work2 is jealous of ctate. Nov 08 23:40:57 I use multiple monitors so it is a pain getting linux to read them all Nov 08 23:41:08 radeon? Nov 08 23:41:16 nvidia Nov 08 23:41:22 i've never had video driver issues, but i guess i buy comps based on linux compatibility Nov 08 23:41:26 I just use nouveau now Nov 08 23:41:27 using distribution tools or nvidia-settings? Nov 08 23:41:38 nvidia settings Nov 08 23:41:49 uhm, you use nouveau and nvidia-settings? that can't work :P Nov 08 23:41:54 I have 2 x 22 inch monitors rotated vertically Nov 08 23:42:17 ah, angelos maybe that is my problem then :) Nov 08 23:42:37 prolly, nouveau+randr tools works fine, so does nvidia+nvidia-settings Nov 08 23:42:38 so now when I run linux, I just disconnected one Nov 08 23:42:47 the propriatary drivers? Nov 08 23:42:55 yes Nov 08 23:43:07 do they work better than nouveau Nov 08 23:43:20 depends on your definition of better Nov 08 23:43:38 they're faster but don't integrate with other tools like nouveau does Nov 08 23:43:59 also feature complete (I'm lookg at you, HDMI audio) and prolly more stable Nov 08 23:44:05 ah, well I don't require much to be happy :) I use xmonad as my window manager Nov 08 23:44:09 or fluxbox Nov 08 23:45:04 try lxrandr maybe, it should work better than nvidia-settings for nouveau Nov 08 23:45:09 or the CLI xrandr tool Nov 08 23:45:57 what's interesting is how all the reviews I read say that linux outperforms windows Nov 08 23:46:10 but things seem faster on windows Nov 08 23:47:08 I could give you some recent numbers from valve on that :P Nov 08 23:47:26 :) linux is faster huh? Nov 08 23:47:35 maybe it's just poor driver support than.. Nov 08 23:47:45 steam rocks Nov 08 23:48:00 strongly recommend ignoring directx Nov 08 23:48:00 ctate> whats up ? Nov 08 23:48:15 i heard ctate works for google Nov 08 23:48:15 they just ported left 4 dead 2 to linux, 16% fps gain Nov 08 23:48:21 Chainfire: i totally can't reproduce your aapt performance problems :) Nov 08 23:48:21 here, in the second answer, isn't an AsyncTask more fitting? Nov 08 23:48:24 http://stackoverflow.com/questions/3027066/how-to-send-a-json-object-over-request-with-android Nov 08 23:48:33 angelos what does l4d2 use on windows, directx? Nov 08 23:48:40 Chainfire: so i suspect that the slow part (that you were avoiding via cache) have been fixed directly Nov 08 23:49:09 Reaga: the use directx, not sure if you can choose to use opengl on windows too Nov 08 23:49:13 ctate> would the current build be in GIT, or private ? Nov 08 23:49:16 fwiw, the framework resources take on ~ 1 minute to build on my (relatively slow) system Nov 08 23:49:29 Chainfire: should be in aosp already; the performance work was done last March Nov 08 23:49:57 ctate> then you're not using the right combination of files to make it happen Nov 08 23:49:58 if you can provide a repro case for "omg aapt is slow" against current master, that'd be gold Nov 08 23:50:08 okay good to know. repro case please :) Nov 08 23:50:13 ctate> because I did sync and compile before I patched Nov 08 23:50:26 okay Nov 08 23:50:33 just checking Nov 08 23:50:43 is there a channel specific to AOSP? Nov 08 23:50:45 * ctate rewrote your change with a hash table Nov 08 23:50:47 or is this it Nov 08 23:50:48 I don't know the exact date, but it'd be somewhere September Nov 08 23:50:56 sleepster: well, #android sort of, but it's kind of gotten out of date Nov 08 23:51:05 guess is hould head over there Nov 08 23:51:32 ctate> I honestly have no idea how to make a bare testcase for this, but I'll look into it Nov 08 23:51:43 lots of other people have the same problem though :( Nov 08 23:51:51 even if it's a large-scale case that'd help Nov 08 23:52:40 same damn problem with a texture skew issue that I can reproduce inside DSLR Controller when using 16-bit bitmaps, and other people have also reported, that only happens on Jelly Bean, but I can't seem to build a test case that demonstrates it reliably :/ Nov 08 23:53:09 well DSLR COntroller is fairly large scale but I obviously can't give you the sources for it Nov 08 23:53:16 heh Nov 08 23:53:20 I'll do some testing over the weekend Nov 08 23:53:26 see if I can repro this specific case Nov 08 23:53:41 it's all XML compilation so it has to do with a large amount of XML, obviously :D Nov 08 23:54:06 factory80: ... the new boston has 200 android videos... I am going to spend the next month or so watching them all... I both hate it and love it. Nov 08 23:54:24 ctate maybe? my question is above about how AsyncTask seems more fitting to use for the 2nd answer of this post, http://stackoverflow.com/questions/3027066/how-to-send-a-json-object-over-request-with-android Nov 08 23:55:25 as a general rule, never do anything that would cause the main (aka "event" or "ui" or "main looper") thread to block Nov 08 23:55:38 this includes any sort of potentially long-running I/O like network activity Nov 08 23:55:50 and I don't know how the first answer was accepted so well Nov 09 00:01:33 Hey, I'm trying to figure out how to stream my screen on my phone to a PC wirelessly like a video. Do I really have to save out bitmaps/jpgs like they're frames? Nov 09 00:01:39 Chainfire: fwiw, i see 17k cache hits vs 8k misses when rebuilding the framework resources :) Nov 09 00:01:45 Or is there a different more efficient way? Nov 09 00:02:01 ctate> sorry, just realised, what are you compiling? I know it has little effect on full AOSP build times. You say you're building framework? As I stated before this compile difference primarily shows during Eclipse use (build, run, build, run, build, run...) Nov 09 00:02:07 gman23: once the N10 / N4 ship, wifi display Nov 09 00:02:14 sorry if I'm a little off it's 1 AM here, I'm not very sharp Nov 09 00:02:26 ctate: I need to do this off of legacy hardware though Nov 09 00:02:32 don't you guys build some major package in Eclipse you can test :) Nov 09 00:02:34 Chainfire: hmmm. yeah, i'm just looking at how long it takes to build the platform framework resources. Nov 09 00:02:40 no, we don't build anything in eclipse Nov 09 00:02:51 innnnnnnteresting Nov 09 00:03:07 What IDE do you use, ctate? Nov 09 00:03:09 (the under-the-covers aapt invocations should be basically the same, i'd think) Nov 09 00:03:20 ctate, any reason why there hasnt been any licensing of DisplayLink? Nov 09 00:03:20 gman23: i use Eclipse for now but i don't use it to build Nov 09 00:03:24 oh Nov 09 00:03:25 I'm afraid I don't know enough about the internals of building to say anything sensible about that Nov 09 00:03:32 Chainfire: no worries Nov 09 00:03:42 robb_www: note that Android basically doesn't license *anything* Nov 09 00:04:15 But I do know that custom ROM builders have told me that it doesn't make a noticable different on full framework build, while it's a godsent durign Eclipse use for us mere mortal app developers Nov 09 00:04:35 robb_www: DisplayLink is more computer oriented, unlike being mainly a video connector for consumer AV Nov 09 00:04:57 Chainfire: innnnnnnteresting Nov 09 00:05:00 thanks for the pointers Nov 09 00:05:10 fair points @ both Nov 09 00:05:35 k, need more ctates Nov 09 00:05:39 (how many televisions support DisplayLink?) Nov 09 00:05:42 robb_www: and (probably) most cheap chips don't have DL transceiver, while getting HDMI transceiver is probably easier despite license costs Nov 09 00:05:48 ctate: zero Nov 09 00:05:57 there you go :) Nov 09 00:06:08 anyways, I'm going to bed. I'm going to try and throw something together to demonstrate it over the next few days and hope I succeed. If you manage to replicate the issue before that, please let me know so I don't waste my time :) Nov 09 00:06:49 thx Chainfire Nov 09 00:06:50 ah i misread the original post, i assumed we were talking about streaming PC to tablet, in which case USB would be more useful to me than wifi Nov 09 00:06:51 g'night Nov 09 00:18:04 anyone familar with Android OpenCV? Nov 09 00:23:07 I want to send JSON to my localhost, the URL that's supposed to receive it is not ready yet, so I have to test here first. I'm using GSON to convert to JSON easily, but here in this code, I don't know what URL to feed it, http://pastie.org/5348582 Nov 09 00:23:12 line 16 Nov 09 00:24:11 it'd just be http://localhost/path/as/usual, wouldn't it? Nov 09 00:24:16 assuming you're running a test server Nov 09 00:24:22 maybe localhost:portnum Nov 09 00:24:26 localhost would refer to the device, no? Nov 09 00:24:48 in the emulator, there's a special IP used to refer to the host Nov 09 00:24:51 I am not running anything that would handle the information once it's sent, I just want to watch the calls Nov 09 00:24:56 that it in fact did attempt Nov 09 00:25:02 I'm testing on the phone Nov 09 00:25:14 just put example.com, then Nov 09 00:25:16 watch the 404 Nov 09 00:25:20 right.. Nov 09 00:26:13 well, alright :) Nov 09 00:30:18 well alright, so I get a permission denied on the HTTP request and it's in warning form, nothing crashes. I think this is an Apache issue Nov 09 00:30:50 Im going to write a small expressJS thing to handle the GET, should work just fine, dont have to wait for the URLs Nov 09 00:42:47 So, Nov 09 00:43:02 it's pretty odd that when i export to apk, the tool chain reports error of unresolved symbols, while there's no error in eclipse with auto compilation on, i.e. the code has been compiled correctly Nov 09 00:43:16 Not really. There's two compilation steps - javac, and dex. Nov 09 00:43:28 So, I have an animation on a custom view, but that view lives in a ScrollView... I only see what happens when the animation is completed... Do I need to enable/disable scroll view caching? Nov 09 00:43:48 sorry if I said that twice, I have to use telnet for the moment... Nov 09 00:44:03 davidleon: Incremental compilation with Eclipse will catch anything that breaks at `javac`. When dex attempts to rewrite symbols to dalvik bytecode, and it's missing one (HOW?), your code still breaks. Nov 09 00:44:11 davidleon: Probably a resource that didn't resolve or something similar. Nov 09 00:45:13 So, I've narrowed down a specific piece of code causing my app to crash Nov 09 00:45:14 http://pastebin.com/D03t8eRR Nov 09 00:45:31 It's the line in onCreate() that declares the txTimer.schedule and the method inside it. Nov 09 00:45:41 kamoricks how to fix it Nov 09 00:45:52 (Any code commented out is no longer commented, that was just already uploaded on pastebin so I relinked) Nov 09 00:46:01 davidleon: What's the exact error message? Nov 09 00:46:09 I think I'm misunderstanding how the uithread stuff works. Any ideas what's wrong with that? Nov 09 00:46:21 kamoricks: hold on Nov 09 00:46:46 http://pastebin.com/ds56kqKn Nov 09 00:46:55 hey guys, anyone work with app engine here? Nov 09 00:47:05 all are warnings Nov 09 00:47:11 don't get why it fails Nov 09 00:47:55 I keep getting connection refused, the expressJS is up and running and ready to receive on port 3000, but the Android app is not sending the HTTPRequest due to this exception, http://pastie.org/5348673 Nov 09 00:48:06 davidleon: Because you're not compiling all your classes. Nov 09 00:48:09 there were unresolved symbol reference at first , i have to include the other project to get javac compilation step correct Nov 09 00:48:28 kamoricks: how do i specify that? Nov 09 00:48:47 anyone have a proper tutorial for android and app engine with eclipse? I have been searching for something for close to a week now, and everything from google seems to be outdated and not working any longer Nov 09 00:49:15 davidleon: Uh. Fix how you're compiling. Nov 09 00:49:27 kamoricks: i added that required proj in eclipse to "required projects on build path" Nov 09 00:49:40 kamoricks: i use eclipse adt plugin Nov 09 00:49:56 davidleon: See eclipse documentation. Nov 09 00:50:10 it's astonishing there's no one integrated solution Nov 09 00:50:11 davidleon: I don't use eclipse, but your issue is that you're not telling the ADK to compile all the classes. Nov 09 00:51:22 kamoricks: why there's no standard way of doing compilation Nov 09 00:51:33 davidleon: I'm sorry? Nov 09 00:52:09 davidleon: You run javac, then dex, then apk, then align, then jarsigner. It's pretty standard. Nov 09 00:52:11 there're so many sub-par compilation solution, all of them are cumbersome. I try to compile zxing lib, the test proj requires me copy some part of the source to test proj source tree Nov 09 00:52:24 davidleon: What tool you use to do that really doesn't matter. Nov 09 00:52:50 kamoricks: umm, i think that's android unrelated Nov 09 00:52:58 davidleon: ...Sorry? Nov 09 00:53:26 kamoricks: how to know how eclipse generate those commands and basing on what Nov 09 00:53:40 davidleon: Dunno. See eclipse documentation. Nov 09 00:53:57 davidleon: If you want to know what commands you should run, d.android.com has a compilation guide. Nov 09 00:55:09 kamoricks: i think i would be crazy to feed tools with those args Nov 09 00:55:42 davidleon: You use strange definitions of words. Nov 09 00:55:59 e.g. Nov 09 00:56:33 AlexZan: given that App Engine is pretty much a host for web apps... how does Android and Eclipse have anything to do with it? Nov 09 00:56:50 it's a bit slow to connect to the website, 2ndly, i suppose hard to use command lines Nov 09 00:56:53 Anyone know if it's possible to repeat an AnimatorSet? Nov 09 00:57:18 3rdly, it's pretty frustrating to deal with all that, considering other development tools are way more mature Nov 09 00:57:26 davidleon: Use them, then. Nov 09 00:57:29 Sicp: this may be a dumb question, but you do have INTERNET permission, don't you? Nov 09 00:58:07 *sigh* Nov 09 00:58:54 i understand it's no android only problem, all java thingie experience is a mess u don't have a standard ide don't have a standard proj file. Nov 09 00:59:07 davidleon: So... you're against competition. Nov 09 00:59:10 just if android extends IDE, it should extend to that part Nov 09 00:59:16 davidleon: There isn't an android IDE. Nov 09 00:59:25 kamoricks: there're HTML standards Nov 09 00:59:41 davidleon: And there's no HTML IDE. Your argument doesn't quite work. Nov 09 00:59:45 I do, Leeds Nov 09 00:59:52 I am looking at this, though, http://stackoverflow.com/questions/6547318/connection-refused-when-trying-to-connect-to-my-rest-server-from-android-emulato Nov 09 00:59:53 trying now Nov 09 01:00:04 I was testing on the phone, but now I'm going to try the emu Nov 09 01:00:10 kamoricks: there's HTML *parser* and there's android ADT plugin for eclipse Nov 09 01:00:43 it's stupid to require different android developer to use different IDE Nov 09 01:00:50 davidleon: There isn't a standard HTML parser! Every web browser uses a different one! And that has nothing to do with development, anyway! Nov 09 01:00:58 as it's hard for others to reuse codes Nov 09 01:01:08 davidleon: Source-code is IDE independent. Nov 09 01:01:22 kamoricks: the parsed result is supposed to be standard. Nov 09 01:01:22 davidleon: All of your arguments are why Android *doesn't* have a dedicated IDE. Nov 09 01:01:27 davidleon: And it is. Nov 09 01:01:35 kamoricks: u can specify the format. implementation can be different Nov 09 01:01:44 davidleon: All APKs can be read by all android phones. Is that your argument? How does that have anything to do with how it was developed? Nov 09 01:02:33 kamoricks: nvm, if you keep argue like that, android can still keep its ass stay at current level Nov 09 01:02:41 Oh yes. WIth logic. Nov 09 01:03:06 kamoricks: your logic doesn't make sense Nov 09 01:03:20 kamoricks: do you mean android can't have a standardised project file? Nov 09 01:03:29 kamoricks: do you mean android shouldn't have one? Nov 09 01:03:38 davidleon: I may be confusing "project" with "output". Nov 09 01:03:42 if you mean that, i believe we shall end the talk Nov 09 01:03:43 davidleon: Android has standardized output. Nov 09 01:03:51 davidleon: Developers work in their IDE of choice. Nov 09 01:04:28 kamoricks: their IDEs can be developed to support the stsandardised project file Nov 09 01:04:41 davidleon: Why would that matter? Nov 09 01:04:56 davidleon: The way that android projects must be built is well-documented. Nov 09 01:05:03 kamoricks: don't you know what android tool chains are for? Nov 09 01:05:20 the whole purpose behind android tool chain is supposed to lower the requirement for a developer to join Nov 09 01:05:22 fwiw, there are at least three widely-used means for building Android software Nov 09 01:05:27 davidleon: What do you consider a toolchain? I consider it the stuff distributed as part of the APK. Nov 09 01:05:38 admittedly one of them is used mostly by people who are working on the platform itself, not on standalone apps Nov 09 01:05:41 davidleon: Have you *seen* this channel? We don't need lower requirements. Nov 09 01:06:05 ctate: which is that? Nov 09 01:06:06 kamoricks: you need to talk with android development team. Nov 09 01:06:17 why they keep evolve from api level from 8 to 16 Nov 09 01:06:28 i believe there's not only feature enhancement and bug fixes Nov 09 01:06:39 but also things that would ease developer's life Nov 09 01:07:03 davidleon: erm... Nov 09 01:07:08 ctate: Makefiles, Eclipse, and IntelliJ? Nov 09 01:07:17 yup Nov 09 01:07:23 ctate: is the build system that you guys use publicly available? I am looking for a build system for my project and it would be nice if there was a continuous integration as well Nov 09 01:07:23 makefiles suck so hard Nov 09 01:07:25 ..and ant, and maven. Nov 09 01:07:43 sleepster: you mean Makefiles? :-) Nov 09 01:07:44 the build system that we use is shipped with AOSP Nov 09 01:07:47 it's GNU make based Nov 09 01:08:08 ctate: ah I see.. so you guys don't have a special build system that you guys use additionally to that? Nov 09 01:08:13 Pretty sure you'd be able to set up a CI system pretty easily. Nov 09 01:08:15 correct Nov 09 01:08:15 and why ppl created cmake and make series things? Nov 09 01:08:25 I see. thanks guys Nov 09 01:08:27 aren't they stupid? why not use GNU make instead Nov 09 01:08:32 everything should be GNU make Nov 09 01:08:53 Mentioning low entry barriers....I need some help with this Nov 09 01:08:53 http://pastebin.com/ZdwRt51T Nov 09 01:08:55 it has its drawbacks Nov 09 01:08:58 ah! and everyone should use emacs? Nov 09 01:09:07 I can't figure out why the txTimer.schedule part is crashing the app Nov 09 01:09:09 I was hoping for a build system that would build, deploy on several different machines, run unit tests, and certify the build as good Nov 09 01:09:10 note that you cannot build Android itself under Windows Nov 09 01:09:18 sleepster: Hudson? Nov 09 01:09:20 I I imagine I'm failing to understanding some threading issue. Nov 09 01:09:30 Leeds: Notepad should be good enough for everyone Nov 09 01:09:46 Striker121: you need to pastebin the actual crash message from the device's log Nov 09 01:09:51 sleepster: What's your stacktrace? It should be available in logcat. Nov 09 01:09:52 yah, you finally get kamoricks's logic Nov 09 01:09:52 Striker121: exception and stacktrace Nov 09 01:10:05 (iirc he may not have a usb cable at hand) Nov 09 01:10:13 Oh, I had thought the crash would be something explicitly obvious Nov 09 01:10:15 ppl can be satisfy with current ant, maven, intellij, makefiles, eclipse Nov 09 01:10:16 And I can grab one Nov 09 01:10:23 davidleon: I'm saying we have a standardized build system. I'm not saying Eclipse ADT isn't terrible. It's why I use IDEA + maven. But that's also why it's not standardized. Nov 09 01:10:46 It was two lines and a single simple function call. And the function is self has no issues. Runs fine elsewhere Nov 09 01:10:52 ok Nov 09 01:10:57 Assumed I had done something really dumb unknowingly Nov 09 01:11:01 * Striker121 is grabbing cable Nov 09 01:11:09 kamoricks: i'll try IDEA + maven Nov 09 01:11:16 davidleon: ..Is that what you actually wanted? Nov 09 01:11:29 davidleon: Would've gone much quicker if you just asked for that. Nov 09 01:15:02 kamoricks: i'm a bit reluctant to try any other ide rather than eclipse, as its reputation in the past is enormous Nov 09 01:15:18 i've tried to stick with eclipse Nov 09 01:16:10 eclipse is total balls Nov 09 01:16:39 Leeds, App engine is just a host for web apps? Nov 09 01:16:42 its frankly disappointing how bad eclipse is Nov 09 01:16:44 where did you read that? Nov 09 01:16:54 AlexZan: what do you think it is? Nov 09 01:17:00 and if you cant, dont ever use idea Nov 09 01:17:13 because it'll just ruin your dat when you have to use eclipse Nov 09 01:17:17 Leads, a solution for backend Nov 09 01:17:24 with front end tools Nov 09 01:17:28 ron_frown do you just deal with eclipse? Nov 09 01:17:30 AlexZan: How is that different from a host for webapps? Nov 09 01:17:32 some including android tools Nov 09 01:17:32 Could anyone point me towards a decent tututorial on getting a stacktrace through logcat? Nov 09 01:17:37 fuck no Nov 09 01:17:39 intellij Nov 09 01:17:44 Everything I seem to find is based off using the emulator Nov 09 01:17:50 Which won't work for this bluetooth based app Nov 09 01:17:54 eclipse basically grenades for me editing framework source Nov 09 01:17:59 ron_frown oh haha thought you didn't use it since you said Nov 09 01:18:02 intellij handles without a hiccup Nov 09 01:18:02 don't ever use idea Nov 09 01:18:14 I am saying... if you know what you COULD be using, and cant use it Nov 09 01:18:17 its like torture Nov 09 01:18:18 kamoricks, well for one, it goes beyond webapps Nov 09 01:18:23 Striker121: seirously, there's a view/pane for it in the Eclipse Android integration UI Nov 09 01:18:24 haha true Nov 09 01:18:27 Striker121: It works exactly the same. Emulator connects to the computer like anything else. `adb logcat` will get you current logs, including (probably) the item you want. Nov 09 01:18:28 if all you know is eclipse, eclipse is a bit more tolerable Nov 09 01:18:29 AlexZan: it's HTTP Nov 09 01:18:34 or, if you are hooked up via a cable, just 'adb logcat' from the command line Nov 09 01:18:35 I am planning to make changes to the AOSP source code and I was thinking of storing an entire snapshot of the source in my own SCM but the problem with that is that it changes so frequirently. How do people manage this problem? Nov 09 01:18:40 ron_frown: i actually did framework dev in eclipse for ages. to be clear, it sucked, but it iddn't suck any moer than doing regular app dev with it sucks :) Nov 09 01:18:42 Leeds, 90% of the things on the net are :D Nov 09 01:18:44 i tried it once, the tools are pretty behind. otherwise its a great editor Nov 09 01:18:55 although i only use it for apps, not framework Nov 09 01:18:55 sleepster: Branching from the current branch and merging in changes. Nov 09 01:19:07 how else would u communicate with a standard webservice Nov 09 01:19:21 kamoricks: ah I see. so it seems that since master changes so frequently, it would be a full time job merging in the changes Nov 09 01:19:23 the other day I tried to get eclipse going with jb 4.1.2 source and it was NOT happy Nov 09 01:19:24 why hxam requires such advanced CPU technology... qemu and vmware doesn't require that Nov 09 01:19:41 i will expect intel x86 emulator be way faster than arm emulator Nov 09 01:19:49 I had gingerbread and early ics source setup "ok" in eclipse Nov 09 01:19:56 anyone have experince with google plugin for eclipse and Android? Nov 09 01:20:22 sleepster: No, it's `git merge`. Nov 09 01:20:31 oh yeah, another issue x86 image doesn't start on my machine Nov 09 01:20:37 i can only run arm-emulator Nov 09 01:20:56 kamoricks: hmm that's true so it shouldn't be tough.. how about if I use a different SCM than git? Nov 09 01:21:09 sleepster: Then you'll need to figure out how to make that thing speak git. Nov 09 01:21:40 git-svn is pretty good, if you like subversion Nov 09 01:22:14 agreed. Nov 09 01:22:34 it's surprisingly functional. Nov 09 01:22:49 I would advise against it if you're not familiar with git though Nov 09 01:23:18 yeah, i would just use git Nov 09 01:23:33 yeah wait, the only reaosn git svn exists is when someone hates you enough to force you to use an svn server Nov 09 01:24:05 hahah Nov 09 01:24:17 I remember trying to convert my boss over to use git Nov 09 01:24:24 (i'm by no means an expert) Nov 09 01:24:41 i convert everyone i can Nov 09 01:24:43 the rest, i kill Nov 09 01:25:01 :) Nov 09 01:25:02 it was just a mind fuck straight up Nov 09 01:25:03 I like git Nov 09 01:25:04 for my boss =) Nov 09 01:25:09 I like what I know Nov 09 01:25:15 git won me over when i discovered interactive rebasing Nov 09 01:25:21 i melted in my chair Nov 09 01:25:23 i learned svn first, so i thought git was wack Nov 09 01:25:30 then i realized it was i who was wack Nov 09 01:25:42 hah Nov 09 01:25:42 sampullman: so did i, but i learned C first too. now i know C and svn are whack :) Nov 09 01:25:58 i can sort of respect svn Nov 09 01:26:01 there are questions I got about cherrypickin and some of how branching is handled Nov 09 01:26:08 Hi guys, quick question, if I wanted to submit a patch to Android, where would I get the code & submit? Nov 09 01:26:20 it was explicitly an attempt to replace cvs with something extremely similar in workflow, but which didn't suck Nov 09 01:26:20 repo init; repo sync Nov 09 01:26:28 budm: source.android.com Nov 09 01:26:30 and then what is it repo commit or repo submit? Nov 09 01:26:31 go forth and hack Nov 09 01:26:47 and i still claim git is one of the shittiest tools i've ever had to use Nov 09 01:26:59 ctate: didn't you watch linus' talk on this exact topic? "cvs that doesn't suck" is impossible. Nov 09 01:27:12 which is a huge pity given how super powerful it is Nov 09 01:27:25 jasta: if you think linus is right about everythign, we have nothing more to say to one another Nov 09 01:27:26 jasta: Thanks, hard to believe it was that easy. Makes you wonder how my patching skills are gonna be, huh? Heh. Nov 09 01:27:33 Strange. I seem to have two installs of the android SDK now Nov 09 01:27:45 ctate: yeah, git UI (yes, CLI can be a UI) is in some cases just awkward for awkward's sake. Nov 09 01:27:52 not merely awkward Nov 09 01:27:58 actively counterproductive and stupid Nov 09 01:28:04 how so? Nov 09 01:28:06 I couldn't find the one that eclipse plugin used (The folder ADB was supposed to be in was indicated as not existing anywhere on the computer) Nov 09 01:28:08 i didn't mean it was less than awkward Nov 09 01:28:19 i meant in some cases it is almost like it's doing it on purpose Nov 09 01:28:22 I've found git being very intuitive and quick for playing with android framework soruce Nov 09 01:28:25 like there was an obvious alternative that was missed Nov 09 01:28:27 ron_frown: is it the stage, or the index, or the cache? you have to know a priori what to call it when you invoke any given command Nov 09 01:28:37 haha ok that I get Nov 09 01:28:39 (for example) Nov 09 01:28:40 When you install it for Eclipse, is it suppsoed to be some kind of stinted version or something? The things to install/update are all labeled the same oddly... Nov 09 01:28:42 for example, why is git push's behaviour configurable? on one machine, if i type git push, i might go "OH SHIT OOPS!" Nov 09 01:28:46 on another, it might be myd aily flow Nov 09 01:28:48 that is not right Nov 09 01:29:22 ctate: the stage is the cache, what are you talking about? Nov 09 01:29:23 :) Nov 09 01:29:26 the reason 'repo rebase' exists is because it's so easy to cause MAJOR HARM if you get the direct git manipulations wrong. Nov 09 01:29:48 ctate: err, i'mg onna stop you right there Nov 09 01:29:52 jasta: except when it's the index, and if you say --staged when the command wants --cached, oops! Nov 09 01:29:59 repo rebase exists because repo is leaky. Nov 09 01:30:04 nope Nov 09 01:30:17 it's because git is so massive that its almost impossible to leak Nov 09 01:30:20 but thats why it exists Nov 09 01:30:28 almost impossible not to leak* Nov 09 01:30:29 repo rebase exists because people were doing "git rebase blah blah goog/master" when in fact they were on the gingerbread branch, not master upstream Nov 09 01:30:41 and forgot that, and didn't get it right Nov 09 01:30:55 and git won't let you know you screwed up Nov 09 01:31:05 ctate: what are you talking about, --staged and --cached are the same thing? seriously. Nov 09 01:31:06 this is *literally* the reason 'repo rebase' was implemented Nov 09 01:31:28 sorry, we're having two conversations here. Nov 09 01:31:37 jasta: with some git commands you call it the stage, with other commands you call *the same thing* the cache, or the index. you have to know which name goes with the command you're using. Nov 09 01:31:43 Even with repo you can still hose stuff if you do the wrong thing. Nov 09 01:31:46 but it's harder Nov 09 01:32:00 ryanm: sure; this is an example Nov 09 01:32:23 ctate: actually i really do not know what you're saying. i agree wholeheartedly about your comments in general, but on this one i'm baffled. and i consider myself to be a git expert but this has never caused me confusion nor have i ever even spent any significant time thinking about it. to me, the two concepts are identical. Nov 09 01:32:25 git checkout -- . is a great way to lose data Nov 09 01:32:42 unless you want to revert =) Nov 09 01:33:08 ryanm: does that fuck the reflog, too? Nov 09 01:33:14 evancharlton: there is no reflect. Nov 09 01:33:15 err Nov 09 01:33:16 reflog* Nov 09 01:33:28 who are you, that kid in the matrix? Nov 09 01:33:30 git will happily lose your shit if you haven't committed or stashed it. Nov 09 01:33:37 ah yes Nov 09 01:33:39 evancharlton: what he said Nov 09 01:33:43 git checkout . will just erase all your uncommitted unstaged crap Nov 09 01:33:47 in practice only a few things directly skip the reflog Nov 09 01:33:48 as it hsould, that's what you use it for. Nov 09 01:33:52 theree's no commit and therefore no ref Nov 09 01:33:57 but as you love god don't delete a branch unless you mean it :) Nov 09 01:34:02 also typing on a plane == laaaaaag Nov 09 01:34:06 ryanm: keep in mind if you staged that stuff it would not be lost. which makes sense. Nov 09 01:34:22 jasta: actually if you have staged but not yet committed changes in a branch that you then delete, it's gone Nov 09 01:34:24 ctate: oh seriously, that and i think git push are my two biggest gripes with git UI. Nov 09 01:34:43 ctate: i know, i was referring to git checkout . -- that will not deleted staged stuff. Nov 09 01:34:44 evn though in most cases staging is a safety-inducing thing Nov 09 01:34:46 right Nov 09 01:35:14 jasta: yep, I know Nov 09 01:35:22 oh, another thing? git happily proceeding with some garbly bullshit default user name and e-mail pisses me off. Nov 09 01:35:23 anwyay, the fact that any conversation about git quickly dives into minutiae about exact behaviors and how to do [xyx] task proves my point: it's a really shitty tool. Nov 09 01:35:26 my coworkers go months without noticing or changing it Nov 09 01:35:51 ctate: i take issue with your phrasing, but i agree with you 100% on the point itself. Nov 09 01:35:54 it's a supercharged hyperpowerful swiss army knife with ninety-three blades, each a foot and a half long and razor sharp, but with no actual handle. Nov 09 01:36:01 it's a really great tool. but it hates you. Nov 09 01:36:05 nope Nov 09 01:36:10 and it will never love you. Nov 09 01:36:15 being easy to use is part of being a good tool Nov 09 01:36:20 powerful != good Nov 09 01:36:29 well, actually that just doesn't maek any sense. Nov 09 01:36:32 [see also: Xwin] Nov 09 01:36:33 i mean, airplanes aren't very simple. Nov 09 01:36:36 I didn't like git until I started using it at work Nov 09 01:36:37 it sounds like it just needs some finishing touches Nov 09 01:36:42 airplanes also have really shit UX Nov 09 01:36:48 hehe, ok good point. Nov 09 01:36:52 even then I still have those "gosh dangit whyyyyy" moments Nov 09 01:37:03 but that basically puts the bar to clear of being a great tool somewhere around hammer or nail. Nov 09 01:37:07 or the "oh-god-how-do-I-recover-this" moments Nov 09 01:37:08 there's a reason being a commercial-grade pilot is *very very hard* :) Nov 09 01:37:31 They also have terrible WiFi. Nov 09 01:37:34 it wasnt even those complications Nov 09 01:37:34 jasta: but git doesn't even aspire to it Nov 09 01:37:40 ryanm: i gave a talk once titled "how to use git to exit a paper bag" Nov 09 01:37:52 it was more my boss didnt get what amounts to floating repos Nov 09 01:37:55 git has to be usable; there is at least one new tutorial a week saying how easy it is :P Nov 09 01:37:58 basically explaining that if you know git rebase --abort and git reset --hard, you can get out of anything :) Nov 09 01:38:02 ==g00s :) Nov 09 01:38:24 the fact that you can get by *only* with "git rebase" also proves my point :) Nov 09 01:38:44 that's sixty or sixty-five of those blades all by itself Nov 09 01:38:51 ctate: i still take issue with your phrasing. i would never dare say git is shitty :) Nov 09 01:38:56 i didn't say that Nov 09 01:38:58 it will only seek revenge Nov 09 01:39:00 i said it's a shitty *tool* Nov 09 01:39:07 shhh Nov 09 01:39:08 :) Nov 09 01:39:18 see also: radial arm saws. Nov 09 01:39:25 if you keep saying this, it will go and change its interface again to reset us Nov 09 01:40:30 ctate: actually i do think some of the best evidence of just how awkward it is comes from looking at its history of terrible UI back before the 1.6 series Nov 09 01:41:03 some of the early decisions were so clearly not designed for humans that it makes you wonder if this thing came from somewhere else in the galaxy. Nov 09 01:41:06 ctate: i found this to be the best summary of things i dislike about git; dont think they were every really refuted https://steveko.wordpress.com/2012/02/24/10-things-i-hate-about-git/ Nov 09 01:41:20 "only 10" Nov 09 01:41:24 :) Nov 09 01:42:41 oh man 3. is so true Nov 09 01:44:50 ctate: watcha working on lately Nov 09 01:44:57 g00s: Android. Nov 09 01:45:02 \o/ Nov 09 01:46:32 nite Nov 09 01:47:04 ctate, but git is the awesome! Nov 09 01:47:15 you're a perforce fan though, aren't you Nov 09 01:47:20 ew Nov 09 01:47:23 oh my god what? Nov 09 01:47:24 :) Nov 09 01:47:43 hey, in other news... Nov 09 01:48:29 i recently interviewed with a company that was predominanetly a .NET shop with almost all of the engineers using Linux or OS X natively Nov 09 01:48:34 i was so baffled Nov 09 01:48:39 lol Nov 09 01:48:41 wth Nov 09 01:48:44 i used P4 a lot and am comfortable with it. it has serious shortcomings w.r.t. git functionally, although its associated toolset is a billion times better than git's. Nov 09 01:48:45 it's like my brain seized up Nov 09 01:48:49 i couldn't stop asking questions about it Nov 09 01:49:04 I will pay HARD MONEY for a git tool that is comparable to P4's "Time Lapse View" tool. Nov 09 01:49:38 some of that hooker rainy day fund? :) Nov 09 01:49:39 hehe Nov 09 01:49:53 ... what? Nov 09 01:50:14 just seemed like you were really trying to emphasize that ... nevermind Nov 09 01:50:51 I've never used time lapse view Nov 09 01:51:08 I would pay good money for an equivalent to p4merge Nov 09 01:51:22 and that. that's the other one. Nov 09 01:51:28 can't you actually use p4merge? Nov 09 01:51:29 Specifically, I want a three-way merge tool that isn't kdiff3 Nov 09 01:51:31 as a git merge tool? Nov 09 01:51:36 i *think* you can probably rig git to use p4merge as an external tool Nov 09 01:51:44 might take a little massaging but ==jasta Nov 09 01:51:46 you might be able to, I had issues finidng a download Nov 09 01:51:53 *finding Nov 09 01:52:02 i think that's because p4merge is either non-free or windows only Nov 09 01:52:03 I think you can, in fct. Nov 09 01:52:16 i also failed to find a download once ages ago Nov 09 01:52:20 *fact Nov 09 01:52:27 P4V in general is quite good, and git gui.... isn't. Nov 09 01:52:27 but it could also be that their web site sucks, i dont know Nov 09 01:52:40 p4merge, like p4v, is written in Qt Nov 09 01:52:47 it's cross platform Nov 09 01:52:47 I use p4 on a regular basis Nov 09 01:52:55 I fucking despise it Nov 09 01:52:56 but kdiff3's interface is clear as mud Nov 09 01:52:56 apparently it's easy to accidentally the power cable when you try to unplug a usb cable Nov 09 01:53:02 ron_frown: ha ha Nov 09 01:53:45 I've seen weird stuff in how it does merging on occasion Nov 09 01:53:54 hmm you know, just occurred to me taht expression makes no sense. clear as a piece of perfectly transparent glass is actually very hard to visualize or understand the glass. Nov 09 01:53:57 but what really pisses me off, is the random gui tool layout Nov 09 01:54:02 but mud, that's quite obvious. Nov 09 01:54:05 one time you open it up Nov 09 01:54:10 Gah, I hate to be the idiot again, but is there anything special to get logcat working? I have USB debugging enabled on my phone, but it simply won't show as a connected device Nov 09 01:54:13 its like 3x the size of your window Nov 09 01:54:20 Is there anything special to do? Nov 09 01:54:22 next time you have to resize to be able to see branch merging Nov 09 01:54:29 Every example I find shows people just plugging it in and it working Nov 09 01:54:33 Striker121: okay, first thing to test Nov 09 01:54:41 in a terminal/shell window, type "adb devices" Nov 09 01:54:47 No devices Nov 09 01:54:50 okay Nov 09 01:54:53 First thing I did :( Nov 09 01:54:58 partly that was to verify that the 'adb' tool was in your path properly etc. Nov 09 01:55:05 windows, mac, or linux? Nov 09 01:55:10 (pre-emptive: jasta: You don't want tto see the glass, you want to see what's behind it Nov 09 01:55:13 striker121: you need the USB drivers, most likely Nov 09 01:55:29 Striker121: what ryanm said, if you're on windows Nov 09 01:55:33 I am Nov 09 01:55:40 Will do then Nov 09 01:55:49 err, pre-emptive was from something else, I'm just lagging. Nov 09 01:55:55 Striker121: read http://developer.android.com/tools/device.html Nov 09 01:55:57 It was showing up as a media device, assumed I had them. I suppose that's the generic non-frills one then Nov 09 01:56:07 Striker121: in particular, note 3. Set up your system to detect your device. Nov 09 01:56:08 etc Nov 09 01:56:23 "media device" is just the MTP/PTP interface Nov 09 01:56:42 read ze link :) Nov 09 01:57:22 (fwiw, i see P4V download links @ http://www.perforce.com/downloads/complete_list ) Nov 09 01:57:49 and P4V is the package that includes p4merge and time lapse view etc Nov 09 01:58:11 ctate, earlier, were you trying to say that the adb command should work from anywhere? Nov 09 01:58:30 * Striker121 had to move into the correct folder in the SDK install Nov 09 02:00:11 Also, despite having the driver in the folder, it won't recongnize it :( Nov 09 02:01:05 Striker121, put it in your path Nov 09 02:01:13 ? Nov 09 02:01:41 Striker121: that's a path thing yeah Nov 09 02:04:38 Wow the Samsung drivers take forever to install Nov 09 02:05:49 their software sucks almost as much as their non top notch hardware Nov 09 02:06:06 kies is crap Nov 09 02:06:08 Could be worse. At least they aren't Asus Nov 09 02:06:24 their usb drivers for win7 x64 never worked properly on me Nov 09 02:06:39 "Driver updates? What are those" As their soundcards fail since release and only get fixed when some random guy breaks there code and does it Nov 09 02:06:54 today they work, the very next day they refust to recognize the usb device, etc Nov 09 02:07:10 kies keeps losing a possibility to recognize the attached phone, etc Nov 09 02:07:20 refuse Nov 09 02:08:01 FINALLY, working debugging Nov 09 02:10:20 Crap, gotta do something for someone, I'll be back Nov 09 02:12:11 Here's the logcat results http://pastebin.com/55978NuF Nov 09 02:12:32 Unaught exception is listed..... Nov 09 02:12:43 I imagine I need to do something more involved now? Nov 09 02:19:45 Striker121: look, it's telling you exactly what the crash is Nov 09 02:19:49 11-08 21:09:12.762: E/AndroidRuntime(4568): java.lang.NullPointerException Nov 09 02:19:50 11-08 21:09:12.762: E/AndroidRuntime(4568): at ieee.bluetooth.ThinBTClient.ThinBTClient.sendSliders(ThinBTClient.java:240) Nov 09 02:20:21 which, going by your earlier paste, is this line: Nov 09 02:20:24 * ctate | outStream.write(msgBuffer); Nov 09 02:21:07 so maybe outStream is null? Nov 09 02:21:09 Ah :D Nov 09 02:21:10 Anyone have an example of how to handle text in a textview in an app widget that exceeds the size of the widget? Nov 09 02:21:12 etc Nov 09 02:21:12 I get it Nov 09 02:21:16 the truth is out there! Nov 09 02:21:27 I didn't iniate the slider values which make up the msgBuffer Nov 09 02:22:57 Thanks guys Nov 09 02:23:03 a public service announcement :) http://www.latimes.com/health/boostershots/la-heb-blood-pressure-brain-damage-20121101,0,3791840.story Nov 09 02:27:21 And upon looking back at it, the values WERE always initilized, oh well. Looking harder **** ENDING LOGGING AT Fri Nov 09 02:59:58 2012