**** BEGIN LOGGING AT Wed Sep 11 02:59:59 2013 Sep 11 03:02:56 did you plug it in properly Sep 11 03:03:39 sometimes you gotta give windows a kick after you've installed the drivers Sep 11 03:06:04 or maybe try adb kill-server then adb start-server Sep 11 03:09:59 Hey guys, I am having troubles with loading more than one webview in a single thread sitting Sep 11 03:10:12 romainguy: I wonder if theres a way to do additive mask? Sep 11 03:10:34 basically build a alpha mask then paint texture to image using it Sep 11 03:40:43 any good alpha mask paint examples out there? Sep 11 03:42:41 something like this https://www.dropbox.com/s/yowmu6aco5e44sm/SampleImage.jpg Sep 11 03:42:59 i'm getting the import com.googlecode.android.wifi.tether.R can not be resolved... i got it from svn, but the R didn't come with it...? Sep 11 03:43:23 hmm isnt R files usually generated? or am I confusing it with something else? Sep 11 03:43:33 yeah i think so Sep 11 03:43:43 can i generate it somehow? Sep 11 03:44:23 dunno not expect Sep 11 03:44:25 expert Sep 11 03:45:22 do you remember where it's supposed to be located? Sep 11 03:45:33 nope. i'm an hack lol Sep 11 03:45:55 supposely I know as I have cs degree but its been years and not keeping skill up much Sep 11 03:46:36 no job or something? Sep 11 03:46:43 sadly no Sep 11 03:47:04 sucks Sep 11 03:47:10 woot http://docs.coronalabs.com/guide/media/imageMask/index.html finally good example I think Sep 11 03:47:33 cool Sep 11 03:47:51 its not android bah Sep 11 03:48:07 eh? why use an image mask? Sep 11 03:48:18 Can't you just use the images alpha channel for what you're doing? Sep 11 03:48:37 that dont work as I want two textures intact Sep 11 03:48:58 generate one? :P Sep 11 03:49:06 yeah I am trying heh Sep 11 03:49:15 I dont understand this enough yet Sep 11 03:49:36 I have code to make mask but dont understand how to use it yet Sep 11 03:50:49 suppose I have image img1, mask drawmask, and result bitmap Sep 11 03:51:14 how do I draw img1 to result, masked by drawmask (which is argb_8 bitmap) Sep 11 04:30:50 yo Sep 11 04:31:07 so i need to connect to the database on my server from my android app Sep 11 04:31:16 any one has experience on this Sep 11 04:31:25 i need something to get started Sep 11 04:32:19 you don't Sep 11 04:32:24 you use a web service or something Sep 11 04:33:17 okay web service Sep 11 04:33:23 what object is that Sep 11 04:33:30 not an object Sep 11 04:34:21 seaworthy willing to spend money ? Sep 11 04:34:36 you could try firebase; i think mongodb client works on android Sep 11 04:34:49 (mongodb client is free) Sep 11 04:35:04 meh Sep 11 04:35:10 i want to right my own Sep 11 04:35:16 i have a postgres site Sep 11 04:35:38 i want users to login to it and get info Sep 11 04:35:44 just dont know how to do it Sep 11 04:36:15 doing something like ../getDataForUser?emal=blah@blah.com&password=blah Sep 11 04:36:21 seems kinda weird Sep 11 04:37:29 i basically need a pro to tell me how its done Sep 11 04:37:48 * pfn snickers Sep 11 04:37:51 use oauth Sep 11 04:39:06 aight ill check it out Sep 11 04:39:07 thanks Sep 11 04:39:09 use a challenge/response protocol that does not involve sending a password-equivalent over the wire Sep 11 04:39:48 okay i dont understand how it keeps the server and the app in sync Sep 11 04:39:59 like you can't just enter the password once Sep 11 04:40:05 and forget about it Sep 11 04:40:55 gonna read on auth thanks Sep 11 05:11:22 * g00s wishes there was a ConcurrentByteBuffer or something :| Sep 11 05:21:44 how would it behave... Sep 11 05:30:41 pfn would be nice to access the bytebuffer from multiple threads Sep 11 05:31:42 then just synchronize it Sep 11 05:32:18 that doesn't define how it should behave in the face of concurrency, though Sep 11 05:32:37 well, each thread could get a slice maybe Sep 11 05:32:52 you can do that already Sep 11 05:33:48 there's a slice method that gives you a buffer slice... Sep 11 05:34:57 Hi Sep 11 05:35:30 How do I tell ndk-build not to link Bionic and add it to include directories? Sep 11 05:35:53 I'm using custom libc (with things such as 16-bit wchar_t) in my project Sep 11 05:38:57 found it, I think Sep 11 05:38:59 TARGET_LDLIBS := -lc -lm Sep 11 05:39:25 very slowly getting to where I want to be on painting Sep 11 05:43:42 what is the best way to save app config information? Sep 11 05:44:49 using preferences or a database Sep 11 05:45:26 sorry I meant if I was concerned about security... user preferences is pretty secure, no? Sep 11 05:46:01 depends on your requirements Sep 11 05:46:37 lets say an oauth2 token, username, the nounce Sep 11 05:48:54 Could anyone tell me if Bitmap.createBitmap(Bitmap source, int x, int y, int width, int height, Matrix m, boolean filter); takes the origin of (x, y) before or after it applies the matrix? Sep 11 05:53:08 what is the name of that android cloud thing that I post data instead of directly to the phone? Sep 11 05:54:16 gcm nevermind :) Sep 11 05:56:52 my answer: before, its the coordinates of the source Sep 11 05:57:04 but it turns out that might not be the source of my problems anyway Sep 11 06:05:26 pfn still there? Sep 11 06:26:21 How can I add form data parameters with an HTTP POST request if I'm using HttpUrlConnection? Sep 11 06:33:51 Why somebody implements "LoaderCallback" interface in his Adapter ? Sep 11 06:38:44 How can I add form data parameters with an HTTP POST request if I'm using HttpUrlConnection? Sep 11 06:40:55 hello, i bought a motorola razr maxx (gsm) and it has android 4.0.4. 2 questions here, anybody knows where can i get android 4.1 update file for it? also second, how can i change its default system font? its a curly kinda font & ugly. i dont have any font style option in Settings/Display Sep 11 06:42:33 omid8bimo: Head to #android Sep 11 06:44:17 napster: ok Sep 11 06:54:15 ? Sep 11 06:54:36 Hello, can somebody explain me how to hand different strings to dynamically created fragments using viewpager, please? Sep 11 06:56:06 Hi, I'm developing a SIP application (using the sip api). I have an IncomringCallReceiver (broadcastreceiver). When a call comes in, I start a new IncomingCallActivity. When another call comes in before the current call ends, it has to be handled in the IncomingCallActivity and not in the IncomringCallReceiver. How can I do this? Do I have to unregister the receiver and register it again when a call is ended? Sep 11 06:57:56 hello ! Sep 11 06:58:21 I am on the search for an android icon for quite a while but I can't find it. Sep 11 06:58:43 How are apps like the system keyboard drawn? What is the Java analog to swing's Canvas? Sep 11 06:58:47 which icon? Sep 11 06:58:47 What I am looking for is the "+" icon in the default contact app of android. Sep 11 06:59:03 which can get used to add additional phone numbers, address sets, etc. Sep 11 06:59:05 have you find the zip with all the icons? Sep 11 06:59:27 I downloaded the AndroidSource git repository and checked packages/apps/Contact Sep 11 06:59:30 and ContactCommon Sep 11 06:59:41 what does the term SKU mean? Sep 11 06:59:55 kraftb: http://developer.android.com/design/downloads/index.html Sep 11 07:00:01 I also created a list of all icons and browsed thru them. but there are quite a few thousand images and I can't seem to find it. Sep 11 07:00:03 tried this? Sep 11 07:00:54 FrancescoV: Not until now. Thanks for the link. I am downloading it and will try to find it. Sep 11 07:01:18 But I fear that the icon is "constructed" ... is a layout on its own having some background and image/imagebutton overlay. Sep 11 07:02:30 I also downloaded AndroidStudio and tried to view the layouts of "Contact" and "ContactCommon" app ... but I tried AndroidStudio just for the first time and altough it showed some layouts it also reported problems with rendering them. And ... I couldn't fint the icon :( Sep 11 07:02:54 Well. I try the design pack. Maybe its there. Sep 11 07:02:58 thanks anyway Sep 11 07:08:53 np Sep 11 07:09:15 can a broadcastreceiver be active when the activity which has started the received is destroyed? Sep 11 07:14:57 Is there a way to programmatically move buttons around the screen Sep 11 07:26:07 Can you keep a broadcastreceiver alive when the app is destoyed? Sep 11 07:27:32 or do I need a service for that? Sep 11 07:33:12 FrancescoV: Can't you simply test it and try to access it after the app is destroyed by looking on adb logcat ? Sep 11 07:33:27 FrancescoV: depends on what you mean with app? if process then a service wont survive that either. Sep 11 07:33:42 But I assume when the app is destroyed all it's parts are inaccessible Sep 11 07:33:52 maybe other solution, I'm using this: http://developer.android.com/guide/components/services.html Sep 11 07:33:53 FrancescoV: of course, if the service is in the running state and sticky it'll get restarted later. Sep 11 07:35:05 but instead of making the listener in the receiver, I create the listener in a new activity. But maybe I need to create the listener in the receiver. But how can I update my UI if a new call comes in? Sep 11 07:36:02 So, I create a listener in the receiver, I start a new activity (IncomingCallActivity). When a new call comes in (handled with the listener), I need to update the UI of the current activity (IncomingCallActivity) Sep 11 07:37:46 FrancescoV: you could have your own broadcast that the activity listens to that you raise in your incoming call receiver. Sep 11 07:38:17 FrancescoV: or the activity could just listen for incoming calls itself in addition to the broadcast receiver Sep 11 07:39:14 appel1: 2nd solution: so I need to unregister the broadcastreceiver, then I handle a new incoming call in the activated activity. And when all calls are ended, I start the receiver again? Sep 11 07:40:05 FrancescoV: why do you need to unregister it? Sep 11 07:41:02 because when another call comes in the receiver wil start a new instance of the IncomingCallACtivity Sep 11 07:41:23 or do I need to check if an instance of IncomingCallActivity is active/alive? Sep 11 07:42:31 FrancescoV: perhaps I'm missing something, but how are your users supposed to answer their calls if you open an activity on top of the phone ui? :) Sep 11 07:42:32 what is an SKU? Sep 11 07:42:53 what does that name mean Sep 11 07:43:20 nb-ben: stock-keeping unit Sep 11 07:43:36 nb-ben: http://en.wikipedia.org/wiki/Sku Sep 11 07:44:24 lol Sep 11 07:44:55 so stock-keeping unit is the meaning of Sku in getSkuDetails() ? Sep 11 07:44:55 appel1: I just start a new activity whan has a answer/ignore button and extra widgets when call is answered. The activity is launched even when phone is locked Sep 11 07:46:29 FrancescoV: ok, didn't know an app could answer/ignore a call. Sep 11 07:46:43 appel1: it's sip Sep 11 07:46:45 (voip) Sep 11 07:47:15 that explains it then :) Sep 11 07:54:07 appel1: So first incoming call --> start IncomingCallActivity (which create a listener etc.), then 2nd incoming call --> onreceive in broadcastreceiver. How can I check if IncomingCallActivity is active & how can i pass data though it from receiver to the active activity? Sep 11 08:04:39 FrancescoV: don't know what is best but you could possibly have some global state that tells the receiver the activity is running and pass data to it using some separate broadcast.. or perhaps use FLAG_ACTIVITY_SINGLE_TOP when starting the activity so an existing instance is reused Sep 11 08:06:04 appel1: idd, now I can check if activity is launched already or not, now I need to figure out how this flag works Sep 11 08:07:32 appel1, do I need to set the flag on the Intent or inside the manifest? Sep 11 08:07:53 hi is there anyone that used maven on android? I am receiving dependency=[com.viewpagerindicator:library:apklib:2.4.1:compile] not found in workspace error in the pom.xml Sep 11 08:08:28 FrancescoV: haven't tried this myself, but perhaps this can help you: http://stackoverflow.com/a/11941272/1220644 Sep 11 08:09:50 wrong thing on clipboard :) ment http://stackoverflow.com/a/6857648/1220644 Sep 11 08:10:02 yo Sep 11 08:13:44 hi Sep 11 08:14:59 how does one go about haveing a free and apaid app in the store? Sep 11 08:15:17 the same app but with different features Sep 11 08:16:36 You could have your main app that is free and all the functionality, but some of it is disabled Sep 11 08:16:50 and then another paid package that "unlocks" the functionality in the other app Sep 11 08:17:09 in the main app, you would check if the other package is installed (and presumably signed with your certificate) Sep 11 08:18:28 javaDon, people tend to use IAP Sep 11 08:18:34 since it's annoying losing all data on upgrade Sep 11 08:18:41 ahh, IAP, good call :) Sep 11 08:19:00 IAP? Sep 11 08:19:03 much better way to handle it Sep 11 08:19:07 in app purchase Sep 11 08:19:46 appel1: tnx, it begins to work :) Sep 11 08:19:49 that dosnt explain how to manage the app Sep 11 08:20:42 there is one things that need to be handled more: when app is destroyed, the receiver must stay active. Do you know how to handle that? Sep 11 08:24:43 me? Sep 11 08:25:11 Francesco you talking to me? Sep 11 08:25:48 javaDon: that's a question that I need to figure out ;) Sep 11 08:28:00 FrancescoV: if it is a manifest registered receiver android will start your process and the receiver for you Sep 11 08:28:51 so here's a somewhat vague question Sep 11 08:29:18 why could code running in a separate thread just stop running on cursor.moveToFirst()? Sep 11 08:29:25 appel1: how can i do that? currently I have only this in my manifest file: Sep 11 08:31:38 FrancescoV: so what triggers that broadcast? Sep 11 08:32:17 I should not be trying to develop a TCP client Service at 03:30 Sep 11 08:32:20 In my activity something like this: this.registerReceiver(callReceiver, filter); Sep 11 08:33:17 FrancescoV: and what filter do you have? Sep 11 08:33:47 ah, my answer: because the cursor is null Sep 11 08:34:11 my custom action string: filter.addAction("android.SipDemo.INCOMING_CALL"); Sep 11 08:40:34 otiose: yep, that's what step through debugging is for Sep 11 08:45:28 it probably doesn't just "stop running". You have a debugging problem. You are not handling exceptions correctly. Sep 11 08:50:07 so guys how does one go about haveing a free app and a paid app and when user downloads the paid app all the prefrences get merged and the free app is replaced with the paid app? Sep 11 08:51:13 i know guys, thanks Sep 11 08:51:22 FrancescoV: hmm, might be that the sip api requires you to use a service.. Sep 11 08:51:40 FrancescoV: at least judging by a cursory look at projects like csipsimple and sipdroid Sep 11 08:51:46 javaDon: I don't think you can. You could have a free app with in-app-payment to makie it the paid version Sep 11 08:52:08 s/makie/make Sep 11 08:52:25 appel1: possible the only solution Sep 11 08:53:07 javaDon: another model ive seen, is to have the free app fully featured, and have it self-neuter itself, then the paid app does nothing but say 'its paid', unlocking the stuff in the main app Sep 11 08:53:18 there could be an action in free app that creates a bundle out of the settings and sends it back to whoever asks it Sep 11 08:53:22 they could use a shared userid to find eachother Sep 11 08:54:07 also the free app can identify that a paid version is installed and refuse to run afterwards, just silently launching the paid version. Sep 11 08:54:31 that method would double the used disk space, storing 2 copies of the app Sep 11 08:54:52 presumably the user would uninstall the free version soon after. Sep 11 08:55:00 if the paid one was just an empty shell, and it unlocked the free app, there wont be as much duplication and no need to transfer config Sep 11 08:55:10 * alankila shrugs Sep 11 08:55:16 the free app could send the user to the play store with instructions to uninstall it Sep 11 08:55:45 one case ive seen, holo launcher and holo launcher hd, are both free versions of the same app, one just has HD images in it Sep 11 08:55:46 tbh I still think that an in-app payment is better option Sep 11 08:55:58 and both use the same paid app to unlock it Sep 11 08:56:04 so one purchase covers both hd and sd versions of the app Sep 11 08:56:12 that's cool Sep 11 08:56:30 I've had those paid unlockers not work for me before, so I'm not mad about them Sep 11 08:56:48 i'm using a kindle, so the google services randomly stop working Sep 11 08:56:57 lol Sep 11 08:56:58 Hi All Sep 11 08:57:00 that must be fun Sep 11 08:57:01 and the license validation code will just throw its arms up and fail Sep 11 08:57:04 hey MickeySoFine Sep 11 08:57:16 how goes it today Sep 11 08:57:19 Ankhwatcher: ive gotten it to a point where i can run 2 shell scripts and re-root it easily Sep 11 08:57:23 that sounds like torture Sep 11 08:57:28 and in theory, i could maybe even do that from the tablet itself Sep 11 08:58:08 i am running an up in an android emulator Sep 11 08:58:13 Ankhwatcher: the trick, is that /efs/ isnt mounted with noexec, and without nosuid Sep 11 08:58:16 and anything i type it types in chinese Sep 11 08:58:21 how can i solve this issue? Sep 11 08:58:31 Ankhwatcher: so i can put a setuid root app in there, and regain root, even after system is nuked Sep 11 08:58:58 * alankila has seen that chinese issue and has to restart the emulator, never worked out how to restore normalcy Sep 11 08:59:38 it also tends to happen at perplexing regularity, suggesting that it's some fairly simple key combo that you can fat-finger in Sep 11 08:59:52 how often does that happen? Sep 11 08:59:56 im not trying to unlock it with in app payment Sep 11 09:00:24 i am trying to have a free and a paid verion of the app 2 seperate apks in the store Sep 11 09:00:44 "You can change the input method by long-pressing on an input field and choosing Input Method → Android Keyboard. I think you can also long-press on the bottom-left key on the virtual keyboard." Sep 11 09:00:48 so they will see myawesomeAppFree and myAwesomeApp Sep 11 09:00:58 the weirdchar is -> Sep 11 09:01:07 assumeing they do like most people would download the free version first Sep 11 09:01:17 so glad that's cleared up. Still have no idea how/why it goes into chinese mode Sep 11 09:01:38 Ankhwatcher: every 3 or 4 months Sep 11 09:01:47 when they download the meAwesomeApp and they had myAwesomeAppFree Sep 11 09:02:39 myAwesomeApp will delete the myAwesomeAppFree and take all of its preferences and Sep 11 09:02:41 there's permanent fix to AVD: just remove PinyinIME.apk from the emulator. It can never go into that pinyin mode afterwards. Sep 11 09:02:59 settings Sep 11 09:03:25 javaDon: the real sticking point is that you can't (probably can't) uninstall the original app Sep 11 09:03:47 fudge Sep 11 09:03:54 thats messed up stuff man Sep 11 09:04:04 are your sure? Sep 11 09:04:33 javaDon: on android, user has certain control and it tends to extend to things like what apps are installed. Applications can't even ask themselves to be uninstalled. Sep 11 09:04:40 javaDon: http://stackoverflow.com/questions/891141/is-it-possible-to-programmatically-uninstall-a-package-in-android Sep 11 09:05:17 or well, they can ASK but user has the final control Sep 11 09:05:42 yeah it's a bad idea don, it's just going to make it confusing for the user Sep 11 09:06:20 I bought the premium version but then it told me to uninstall wtf? Sep 11 09:10:48 so whats a better way of handleing this? Sep 11 09:11:09 i have to make some money from mya app and i also want to offer it for free Sep 11 09:11:18 javaDon: http://developer.android.com/google/play/billing/billing_overview.html Sep 11 09:16:57 appel1: have you used the sip api before? Sep 11 09:17:01 or somebody else? Sep 11 09:25:33 looks like i will be doing the in app billing after all looks pretty simple Sep 11 09:25:41 simpler then handleing 8 .apk files Sep 11 09:25:43 :) Sep 11 09:26:06 anhoo im still stumped though on how i can filter out devices based on there screen size Sep 11 09:26:22 since there are some screen sizes that are 1 size and some that are another Sep 11 09:26:35 but are both classified as normal screens Sep 11 09:28:43 ok i gotta run out for ait be back l8r guys Sep 11 09:38:10 How can I add form data parameters with an HTTP POST request if I'm using HttpUrlConnection? Sep 11 09:41:06 napster: use the application/x-www-form-urlencoded mime type and take the key, escape it accoring to url escaping scheme, place =, take the value, escape it too, and if you have more parameters, place &, and repeat. Sep 11 09:42:22 alankila: ok, looks not very easy. Will digg into it. Thanks. Sep 11 09:42:42 or may be for now, using Apache's stack is easier. Sep 11 09:42:44 There is no utility method that I'm aware of which would do this for you Sep 11 09:42:52 alright Sep 11 09:42:59 but it's like 5 lines of code Sep 11 09:44:01 pastebin.com/xG9jz2A9 Sep 11 09:45:28 this one is a bit different in that there's a message authentication code in the system, but other than that... Sep 11 09:46:21 alankila: ive seen a util for it recently Sep 11 09:47:13 alankila, napster: https://code.google.com/p/android-remote-stacktrace/source/browse/src/com/nullwire/trace/ExceptionHandler.java#192 Sep 11 09:47:31 exactly what you waited, name/value pairs in a post body Sep 11 09:47:32 thanks alankila Sep 11 09:47:58 yeah but that uses the apache httpclient piece of shit Sep 11 09:48:26 it is sort of deprecated Sep 11 09:48:32 ah Sep 11 09:48:58 i didnt notice that, i just remember having to add a field to that lib recently Sep 11 09:49:06 sure, the stuff in httpclient *is* more conveniently engineered, no question about that Sep 11 09:49:32 but if you know how http requests are structured then you can generate the request just the same and it's not even hard. Sep 11 09:49:44 ive been using URLEncoder.encode in my other code Sep 11 09:49:58 yeah Sep 11 09:50:03 iirc apache httpclient 4 went off the deep end so thoroughly that it become actually hard to use it to do anything Sep 11 09:50:10 must be second system syndrome or something Sep 11 09:50:33 all of the code i wrote uses URLEncoder and HTTPUrlConnection Sep 11 09:51:49 hmm, thats an odd place for a null pointer error, *opens code* Sep 11 09:56:00 Actually I preferred to use HttpUrlConn api itself. But these little things does matter. Since google guys are putting effort into the HttpUrlConn iself, we might get this soon. Sep 11 09:56:30 * alankila fixes security issue with his accountapiquery... Sep 11 09:57:23 strange, somebody on 2.3.4 just had another NPE in a weird place Sep 11 09:57:27 * clever reopens code Sep 11 09:58:38 I wish URLEncoder.encode() accepted a Charset though Sep 11 09:58:52 ah the joys of NPE Sep 11 09:59:04 *sigh* so much fail because of old java APIs Sep 11 09:59:23 Ankhwatcher: git reveals its this line this time, handler.post(new Runnable() { Sep 11 09:59:41 but i set that in Session.onCreate Sep 11 09:59:58 so there is no way for it to be null at this stage Sep 11 10:01:31 oh, hmmm, maybe if the service was stopped?, *looks* Sep 11 10:02:03 onDestroy does null it out, but the thread should be stopped by then, i'll add a boolean to detect it Sep 11 10:05:20 * alankila idly wonders if result of (char) (byte) 128 is some value like 0xff80 Sep 11 10:05:29 I wish I had simple java command line interpreter to test stuff like this Sep 11 10:06:20 time to try out beanshell Sep 11 10:07:13 yep, ff80 came out Sep 11 10:07:15 piece of shit Sep 11 10:09:38 well, byte has a max value of 127 Sep 11 10:10:20 All true. Sep 11 10:10:44 was just kinda hoping that because char is unsigned 16-bit type the cast would have been unsigned Sep 11 10:16:55 but that would be going against the documented behaviour, i guess Sep 11 10:18:12 sure, bytes have a particular meaning and they are signed type. I have nothing really against this, I guess. Sep 11 10:23:57 any idea why i get on some devices when starting a service 09-11 12:09:49.078 E/AndroidRuntime(4855): java.lang.RuntimeException: Unable to create service .modules.MQTTService: java.util.MissingResourceException: Error locating the logging class Sep 11 10:24:29 service is in the manifest like all the other services which works well Sep 11 10:36:14 Heya guys, quick, albeit weird question, I think I'll b hitting a dead end on this real soon Sep 11 10:36:50 On one of our testing devices (An Xperia with Android 4.0) there is a semi-frequent occurrence of complete insanity when the user switch away and back to the application. Sep 11 10:36:59 Namely, all the background drawables disappear Sep 11 10:37:12 Not the imageviews, nothing is broken, just everything is background-less. Sep 11 10:37:23 All of the application remains this way until force closed. Sep 11 10:48:19 morning Sep 11 10:48:28 Barely. Sep 11 10:48:47 indeed Sep 11 11:00:13 how can I make a textview keep x and y even while scaleing? Sep 11 11:10:00 anyone knows how to resolve merge conflicts in android studio?? Sep 11 11:18:43 hi Sep 11 11:21:09 I want to show Standings (soccer). I get the data via json. I have to calculate the standings by myself. What is the easyest way to do that? Sep 11 11:23:03 here is the link to the json data: https://openligadb-json.heroku.com/api/matchdata_by_league_saison?group_order_id=3&league_saison=2013&league_shortcut=bl1 Sep 11 11:23:39 anyone used twitter4j before? is that the easiest to use for searching? Sep 11 11:31:17 do I have to have an access token/secret as well as a consumer key/secret when using twitter api? Sep 11 11:48:16 Just ran into odd behaviour. Basically nothing gets saved to shared settings or file with my galaxy S4. Other phones it works fine. Is there some S4 speciality that I need to take care of? Sep 11 11:48:21 Hello everyone. How can i catch the extra arguments that i pass in am (I am using am to open my app) in my aplication Sep 11 11:48:50 am start ... [-e ] Sep 11 11:49:05 ivali, it's probably in intent extras Sep 11 11:49:13 getIntent() Sep 11 11:49:26 thank you! Sep 11 11:55:43 Hey, I'm having a problem with a popup menu. Basically, it doesn't actually pop up when I click the button. I've looked at several examples and they all do it the same way I did. One thing I can think of is that it is being showed behind some of my other stuff on the screen? Anyway, here's the code: http://pastebin.com/2ZAtynWB Sep 11 11:57:21 If you need the xml files as well, let me know. Sep 11 12:00:57 When you say popup menu, do you mean a context menu? Sep 11 12:01:18 i.e. http://farm5.static.flickr.com/4132/5028762301_9bcd9963d4.jpg Sep 11 12:02:17 Tee_Pee: No, a popup menu. Like this example: http://www.pavanh.com/2013/05/android-popup-menu-example.html - Although it looks similar to what you posted. Sep 11 12:03:08 ah, my bad Sep 11 12:03:32 Hi everybody. Someone knows how to do that https://developer.android.com/design/media/progress_activity_do.png - I just can't find the style/image/animation of that :/ Sep 11 12:03:46 Would you recommend using a context menu instead? Sep 11 12:04:39 Nah, you don't need to Sep 11 12:04:42 the issue with your code is simple Sep 11 12:04:45 you never inflate a menu Sep 11 12:05:00 final PopupMenu popup = new PopupMenu(this, popup_but); Sep 11 12:05:15 this is fine, but you'e missing Sep 11 12:05:25 popup.getMenuInflater().inflate(R.menu.popup, popup.getMenu()); Sep 11 12:05:36 So something like this: popup.getMenuInflater().inflate(R.menu.popup, popup.getMenu()) ?; Sep 11 12:05:37 Oh Sep 11 12:05:38 Yeah Sep 11 12:05:49 and obviously, you need your popup.xml Sep 11 12:06:12 Where should I inflate it? Before the onclick? Sep 11 12:07:05 Tee_Pee: Ah it works now. Thank you! :) Sep 11 12:07:25 No problem :) Sep 11 12:07:38 anyone know any goof twitter sdks? Sep 11 12:07:43 *good Sep 11 12:11:49 is it possible to change the layout (xml) from an activity to another xml-layout? Sep 11 12:14:44 FrancescoV, sure but you would have to re-create the Activity. Sep 11 12:15:00 better to use Fragments Sep 11 12:20:03 anyone worked with twitter4j before? Sep 11 12:24:26 i just voiced my opinion on how the new iphone 5s is looking like sh it Sep 11 12:24:31 in iphonedev Sep 11 12:24:40 theres nothing new about it Sep 11 12:24:44 im totally angry Sep 11 12:24:51 i was thinking of geting that phone Sep 11 12:25:00 looks like i will have to cough up 800$ Sep 11 12:25:03 for a lg g2 Sep 11 12:25:37 someone knows how to do that https://developer.android.com/design/media/progress_activity_do.png - I just can't find the style/image/animation of that :/ Sep 11 12:26:41 Zirkome, Widget.ProgressBar.Small Sep 11 12:31:52 pfn, that's not that i want but I'll investigate. It's for an activity circle like in the Google PlayStore Sep 11 12:32:00 what* Sep 11 12:32:45 man, what the fuck HTC. Sep 11 12:33:25 hi guys, what should i be using to calculate distance between two points(very short distance) i tried working with GPS but that does not work in the short distance, so any sensors that is built in to measure short distance? a link regarding the topic would be fine, thanks in advance Sep 11 12:34:33 Zirkome, sounds exactly like what you want Sep 11 12:35:41 meadhikari: some numbers would be nice. Are we talking about like 1 metre distance? Sep 11 12:35:54 alankila, ya around 1 meters Sep 11 12:36:35 it is clear that you can't reliably estimate the distance when the inaccuracy in the positioning system itself is larger than the distance you want to measure. Sep 11 12:37:10 pfn, no which one i see is rotating left AND right :p Sep 11 12:37:44 GPS may reach precision around 2 metres if you have extremely good conditions. If you have ground-based gps helper stations, the precision can be further improved to something like 10 cm if your device knows how to use them. Sep 11 12:38:18 Zirkome, backport it, that style came in 3.0 and up Sep 11 12:38:29 not sure if any phones have any of the hardware and software for those ground based reference stations... nor if they are commonly deployed anywhere Sep 11 12:39:25 Zirkome, add a Holo in the widget name somewhere and it'll probably show what you want Sep 11 12:39:38 style name Sep 11 12:40:11 pfn, it's better but it's not that i think i'll give up :p Sep 11 12:43:57 Hey. So I got the popup menu to work. I'm using the popup menu to switch between layouts using setContentView. My problem is that after switching, even though the new layout has buttons with the exact same ID, it doesn't work there. I tried changing the IDs and do an entirely new popup menu for the new layout, but it crashes. Am I going about this wrong? Sep 11 12:44:57 Moopz: if you replace the button instances, you must set all handlers anew too Sep 11 12:46:41 alankila: Handlers? You mean the click listeners etc? Sep 11 12:47:02 and any objects you may have cached on the activity from the previous layout Sep 11 12:48:38 can't solve the problem. How can i recreate a fragment if it has been inactive for a while Sep 11 12:49:09 alankila: Please bare with my feeble mind. So I have this code: http://pastebin.com/z92NSRcv - How would I go about it? I tried overwriting the instance of my button, but it didn't work. Sep 11 12:50:04 now it crashes on OntaskCompelete Sep 11 12:50:13 or view and returns nullpointer Sep 11 12:50:54 Moopz: as I expected, your setContentView is followed by nothing Sep 11 12:50:57 Dominee: do extensive research on activity/fragment lifecycle (honest answer) Sep 11 12:51:08 alankila: What would I follow it with? Sep 11 12:51:14 yes i did this morning Sep 11 12:51:15 so all your views in the activity are now stale, they don't refer to the currently active objects. Sep 11 12:51:27 Dominee: that's not extensive Sep 11 12:51:40 Moopz: you sort of need to call onCreate again after setting the correct contentview Sep 11 12:51:55 split the method onCreate to two, one which sets a layout and then updates all objects from the currently active layout Sep 11 12:51:56 hahaha ok, Sep 11 12:52:09 oncreate and oncreateview... Sep 11 12:52:14 then in onCreate just call setNewLayout(R.layout.loco_control); Sep 11 12:52:34 in any case I suspect you are using android wrong though Sep 11 12:52:48 you are probably planning to make these layouts show completely different things, not just same objects in new configurations Sep 11 12:52:54 Dominee: it takes a while to understand, take your time, it's not something you want to breeze over Sep 11 12:53:14 true, i will do some research Sep 11 12:53:15 you should use either fragments or activities, give each a single UI and single purpose Sep 11 12:53:20 it takes time indeed Sep 11 12:53:32 alankila: That's right. I read somewhere that views was faster than activities, which is why I did this. Sep 11 12:53:46 Moopz: forget about crazy micro-optimizations like that. Sep 11 12:53:47 Dominee: i am still struggling with it 2 years later... to some degree :) Sep 11 12:54:18 Just do it like every android tutorial tells you to do it, declare separate activities and switch between them via startActivity(), and use utility classes for common functionality you need in all activities Sep 11 12:54:19 so i have a long time to go ;) Sep 11 12:54:29 or be modern and consider single activity and multiple fragments Sep 11 12:54:39 Dominee: well, it's a good chance you are smarter than me :) Sep 11 12:55:03 just coding java for 1 year now Sep 11 12:55:14 alankila: I suppose I'll read up on fragments. It seems doing Activities would require some reworking of my code. Sep 11 12:58:36 Dominee: to answer your original question briefly, Android OS might restore your fragment instance for you but your code is responsible for restoring the component and its UI to the desired state Sep 11 12:59:02 Dominee: knowing when and how to do that is what you need to read up on Sep 11 13:00:31 its recreating things simultaneously Sep 11 13:00:46 i think thats no the right thing Sep 11 13:00:48 t Sep 11 13:01:15 when I try this it works fine: standings[i] = Integer.valueOf(e.getString("team_id")); Sep 11 13:01:16 but this do not work, why? standings[Integer.valueOf(e.getString("team_id"))] = 0; Sep 11 13:01:23 Dominee: well presumably its restoring an activity and at least 1 fragment Sep 11 13:02:08 I want to set the index by the team_id and not from 0-17 Sep 11 13:02:26 I have a FragmentActivity with one Fragment. If something happens, I would like to change the UI to a viewpager (already in xml-layout). Can I use the first active fragment as the first fragment in the viewpager and add pages later too? Sep 11 13:02:59 Uh, can I step through the different states/images in a DrawableAnimation? Sep 11 13:03:09 AnimationDrawable* Sep 11 13:03:21 I'm creating a hangman game, I thought using an animation would be easiest Sep 11 13:04:17 FrancescoV: sounds possible to me Sep 11 13:04:50 gonna look into it today, its nice to learn new things about the lifecycle Sep 11 13:05:45 gives me a better view how an android app works Sep 11 13:06:04 Dominee: you don't want to see the sausage being made, it'd ruin the magic Sep 11 13:11:07 FrancescoV, there is no reason why you can't, but you will have to be careful because you will presumably want to be saving state when it is outside the pager (e.g. using ids/tags) as well as inside the pager (using fragment pager adapters), which may cause a gotcha since they do their own funky storage on your behalf Sep 11 13:11:31 I'm not sure if it will cause any problems, but it is worth being aware that you may be storing your fragment twice Sep 11 13:11:59 gah, now eclipse suddenly does not find my phone -_- Sep 11 13:12:57 or do I need a Map to do this? Sep 11 13:14:28 funkbox: haha, you have been there? :P Sep 11 13:14:37 sebton, Kegsay, tnx both. It's for my sip/voip app. When there a 2 incomming calls (1 already active), I would like to visualize them in a viewpager ;) Sep 11 13:16:05 a viewpager seems a little odd for that use case (I would expect a less intrusive notification or small window to say "oi! You have someone else ringing you", unless you simply plan on utilising the swipiness of pagers so you can say 'swipe to view the other call'? Sep 11 13:17:36 Does anyone have a better answer for this? http://stackoverflow.com/questions/8448496/how-to-set-a-specific-frame-in-android-animation selectDrawable() appears to do absolutely nothing Sep 11 13:18:07 currently I have a little layout that comes visible with info about the new incoming call but with a viewpager, the user can switch easily from callX to callY (a call can be set on hold) Sep 11 13:22:38 hey guys... when the user try to activate my app, can i check which google account he is using? Our projects here will sell bunches of thousands specific apps and it would be a pain in the ass to insert keys one by one or any other manual registration. Allowing X activations for the same google account would be faster i think... Sep 11 13:26:11 got dc Sep 11 13:26:12 hey guys... when the user try to activate my app, can i check which google account he is using? Our projects here will sell bunches of thousands specific apps and it would be a pain in the ass to insert keys one by one or any other manual registration. Allowing X activations for the same google account would be faster i think... Sep 11 13:31:58 luist: http://developer.android.com/reference/android/accounts/AccountManager.html Sep 11 13:37:20 got dc again when someone answered me… can i get that answer again? Sep 11 13:37:37 14:31 < mmathis> luist: http://developer.android.com/reference/android/accounts/AccountManager.html Sep 11 13:40:17 hello Sep 11 13:40:40 is it safe to call getActivity() in Fragments onCreate() ? Sep 11 13:43:16 make it clickable! <== How do you at Square handle it? Avoid having the Action Overflow? Sep 11 13:44:23 duo_, no Sep 11 13:45:24 pfn: then why are google doing this in their samples?!! Sep 11 13:45:35 pfn: http://developer.android.com/training/displaying-bitmaps/manage-memory.html Sep 11 13:45:40 BitmapFun.zip Sep 11 13:45:41 what samples Sep 11 13:45:48 in ImageGridFragment.java Sep 11 13:46:12 xlq: thanks! Sep 11 13:46:25 not going to unpack and look Sep 11 13:46:35 why would you call getActivity in onCreate? Sep 11 13:46:38 pastebin then? ok, gimme a sec Sep 11 13:46:43 I can't think of a conceivable reason Sep 11 13:47:03 duo_, pfn https://github.com/android/platform_development/blob/master/samples/training/bitmapfun/src/com/example/android/bitmapfun/ui/ImageGridFragment.java#L78 Sep 11 13:47:27 And three more times after that. Sep 11 13:48:07 right Sep 11 13:48:09 it's either wrong, or there is a special circumstance in how the fragment's life cycle is managed Sep 11 13:48:24 ... Sep 11 13:49:08 Or it might just be a sample code error. Sample code is just that - samples. I've found syntax errors before etc. Sep 11 13:49:16 they're often incomplete Sep 11 13:49:21 Is it explicitly mentioned that getActivity() should not be called or is it implicitly because of the overlapping lifecycles? Sep 11 13:49:40 I mean, there' onActivityCreated() for that. Sep 11 13:49:53 the camerapreview sample for example is atrocious on its own Sep 11 13:50:22 i hate them, god how i hate them Sep 11 13:51:17 kakazza, look at lifecycle flow, it is only guaranteed non null after on activity created Sep 11 13:51:54 I think that it is ok to call get activity in oncreated Sep 11 13:52:21 However, be careful to call getActivity() only when the fragment is attached to an activity. When the fragment is not yet attached, or was detached during the end of its lifecycle, getActivity() will return null. Sep 11 13:52:40 so, wouldn't onattach be the point when you can call getactivity? Sep 11 13:52:49 onattach happens before on create Sep 11 13:53:19 you just can't assume that things in the activity's oncreate has finished Sep 11 13:53:31 until onactivitycreated Sep 11 13:53:51 "Caution: If you need a Context object within your Fragment, you can call getActivity(). However, be careful to call getActivity() only when the fragment is attached to an activity. When the fragment is not yet attached, or was detached during the end of its lifecycle, getActivity() will return null." Sep 11 13:53:57 From the docs. Sep 11 13:54:06 https://developer.android.com/guide/components/fragments.html#Lifecycle Sep 11 13:54:30 yes Sep 11 13:54:54 Which would make it safe in onCreate Sep 11 13:55:40 depends on what you're doing, but it's safe to get a non-null pointer Sep 11 13:55:54 it's not overall safe Sep 11 13:56:00 Elaborate Sep 11 13:56:18 read mmathis' description Sep 11 13:56:57 I have a new stupid question: So I've made a menu that can launch new activities using startActivity. My question is, does startActivity start a new instance of the activity, no matter if there's already an instance of it? If so, how would I go about detecting instances of activities and then switching to them? Sep 11 13:57:12 Moopz, android:launchMode Sep 11 13:57:17 set it on your activity to do the right thing Sep 11 13:58:00 pfn: So I'd use Single Instance? Sep 11 13:58:11 Moopz, I dunno, I suppose you would Sep 11 13:58:49 Would startActivity then just always switch to that activity? Sep 11 13:59:01 is that what it's documented as doing? Sep 11 13:59:07 Just a heads up men - if your PMSing wife says she wants a certain flavor of chocolate, you better fucking come back with that flavor of chocolate. Sep 11 13:59:21 Tee_Pee, you only just now figured that out? Sep 11 13:59:29 Tee_Pee: sounds serious :O Sep 11 14:00:02 pfn - nah, but it's healthy to remind myself from time to time, otherwise I get sloppy. Sep 11 14:00:06 Like just now... Sep 11 14:05:38 interesting, my ping to lol has gone to shit since switching to uverse :( Sep 11 14:05:48 45mbps down but 50ms to lol :( Sep 11 14:06:06 your ping to where? Sep 11 14:06:13 50ms isn't bad either Sep 11 14:06:23 League of Legends? Sep 11 14:06:37 not when I used to be sub20ms Sep 11 14:07:36 anyone with any exp. using app engine? i just set it up for the first time and experiencing latency ~160ms. wondering if its normal for a free user. Sep 11 14:08:28 balls2thewall, let it warm up Sep 11 14:08:31 first requests are always slow Sep 11 14:09:51 xlq: is there a way to have 2 accounts loged in on the device? it would be good for each user to have its personal account to use the google services, but still have the project account to get the apps updates automatically Sep 11 14:09:53 by first you mean the first day or just the first request that needs to launch the app. i set it up 3 hours a ago and 160ms is pretty consistent Sep 11 14:11:16 luist: I have no idea, I only repeated mmathis's reply. :P Sep 11 14:11:26 xlq: oh okay lol Sep 11 14:11:32 mmathis: ^ Sep 11 14:11:32 Would startActivity then just always switch to that activity? --- yes, and you should get onNewIntent(Intent) with whatever you put in the latest intent Sep 11 14:12:25 luist: you can be singed into multiple google accounts at the same time Sep 11 14:12:55 perhaps multiples of any kind of account Sep 11 14:13:00 but for sure multiple google accounts Sep 11 14:13:05 I have like 3 logged in on my phone Sep 11 14:13:32 though actually... as far as getting u pdates from the play store goes Sep 11 14:13:40 I'm not sure how that is handled with multiple accounts Sep 11 14:15:15 I don't really know though, I just googled "android check users google account" when I sent that answer to you tbh Sep 11 14:21:54 <\0x90\> a quickie. is int_value integer default '0' same as int_value integer default 0 when creating a database/column ? Sep 11 14:26:59 \0x90\: for sqlite, it is, because it's loose about types Sep 11 14:27:46 <\0x90\> ok cheers Sep 11 14:28:19 Hmm wait Sep 11 14:28:39 I'm reading about this and I thought I understood this but sqlite manual says that any column with exception of primary key can be used to store any value of any storage class Sep 11 14:29:35 No, I believe there will be conversion to the actual storage class. Sep 11 14:30:25 <\0x90\> i saw some examples (plain sqlite though.. not android) and it's done without '' Sep 11 14:31:06 The whole thing is really insane. Basically sqlite engine will perform an analysis of the string value and if it believes a lossless conversion is possible, it will do it Sep 11 14:31:52 * g00s waked up this morning to read CursorWindow :| Sep 11 14:31:56 but if not, it will just store whatever gunk into that row and column Sep 11 14:32:39 I generally think that engines that have even some modicum of strictness are preferable -- fail early rather than late. Sep 11 14:34:55 I hate systems where you could select an integer column and get text. I wonder what happens, will it throw an exception? Barf. Sep 11 14:35:43 <\0x90\> me too Sep 11 14:36:22 <\0x90\> and i'm already acquiring phobias in regard to exceptions and phone vibrating when it happens. it's reminding me of frequent disconnects on 33.6k dialup modems in the 90s. Sep 11 14:36:54 H2 is a nice engine, but probably too slow forwhat you need. and no cursor to send over CursorProvider Sep 11 14:37:17 you would have to use plain jdbc stuff which is not as optimized as the sqlite java bindings Sep 11 14:37:31 sqlite is awfully fast and loose Sep 11 14:38:01 I'm working on some code right now where the logfiles are sqlite databases. Even on the windows/osx code. Sep 11 14:38:02 i wish it had single precision float class :( Sep 11 14:38:12 sqlite is neat Sep 11 14:38:17 moreso not on flash Sep 11 14:38:35 "So how are you generating that" "Well we have no documented methods..." *sigh* Sep 11 14:38:45 sqlite is neat. This is a horriffic abuse. :P Sep 11 14:38:46 g00s: singe precision kinda sucks Sep 11 14:38:53 "single" Sep 11 14:39:00 oGMo not if you only need 5-6 sig figs Sep 11 14:39:10 a representative row from one of the logs I have to make: Sep 11 14:39:11 and sqlite has issues with double precision anyhow Sep 11 14:39:12 119|0|1377804488861|63:77;7>68:79;9404766?6497374<78655329428294422 Sep 11 14:39:22 except it goes on for 512 more characters Sep 11 14:39:30 what the hell is thta Sep 11 14:39:50 what is the stuff in the last col? Sep 11 14:40:00 oGMo: a big binary blob squooshed into a database column... 'cause... i don't know. :( Sep 11 14:40:08 ha Sep 11 14:40:10 Love how Fido calls 500 megabytes of data month, LOTS of data Sep 11 14:40:11 It's the raw 8bit RSSI signal levels from the USB device :P Sep 11 14:40:33 i think in sqlite blobs should be the first columns in the row Sep 11 14:40:34 dragorn: wowbad Sep 11 14:40:40 oGMo: yeeerrp. Sep 11 14:40:42 oh Sep 11 14:40:46 alankila: it seems that if you try to get an int from a column that has a string it will return 0 Sep 11 14:40:49 well that's possibly less bad Sep 11 14:40:51 Other parts of the db are the conversion math for it Sep 11 14:41:03 sqlite has an interesting notion of the conversion validity, if a numeric affinity column can represent the first 15 digits correctly it would use floating point storage type Sep 11 14:41:26 dragorn: if it's not something that should be indexed itself then sqlite isn't bad for structured storage .. seems really fast and efficient even for the blob crap Sep 11 14:41:27 What really blows my mind is I have config values for the device in Hz or Khz, like 338000 ... Which get stored in the dbase in khz or mhz as floats, like 338.1233222342322 ... .... *brain asplode* Sep 11 14:41:30 ive also noticed, that if you insert "15" by mistake, it wont match 15 Sep 11 14:41:30 mmathis: of course. Always good to fail silently in those circumstances. It's like PHP all over again. Sep 11 14:41:45 its stored as a string, and wont compare properly against a real number Sep 11 14:41:48 oGMo: it could be worse. it's still pretty heinous. Sep 11 14:42:06 alankila: you could use the above flaw to force your numbers to be stored as strings Sep 11 14:42:14 dragorn: it seems like it could be better, though if all that were expanded it might explode the size of the db, dunno Sep 11 14:42:16 dragorn: use whole integer values, they can represent numbers exactly up to something like 1 << 54 Sep 11 14:42:29 dragorn do you think a db in general was right for that? i had something similar but have been toying with just appending to a file Sep 11 14:42:30 alankila: yes, I know Sep 11 14:42:32 alankila: I didn't design the db Sep 11 14:42:36 oh. Sep 11 14:42:37 it would probably be worse if it was a custom format that was highly inflexible and had no commandline query tool Sep 11 14:42:47 alankila: This has to be compatible w/ commercial windows and mac products ;P Sep 11 14:43:01 alankila: so they made up this db format... I'm just the only person to implement it outside of the original designer ;P Sep 11 14:43:08 * alankila has databases which contain REAL columns in oracle too... Stores 5-6 digit integers not meant for any mathematical computation, they're just ID values Sep 11 14:43:23 and no, not my design either. I gacked briefly and decided best not care Sep 11 14:43:32 alankila: i hope you're not suggesting fixed-point math :p Sep 11 14:43:41 oGMo: there's no useful way to extract command line stuff from it Sep 11 14:43:53 oGMo: I mean, i could, because I've written the open source version of their device drivers too Sep 11 14:44:01 oGMo: but I kind of doubt anyone else would Sep 11 14:44:16 dragorn: well right i just mean for anyone developing with it Sep 11 14:44:20 oGMo i'm experimenting with converting all my sqlite REALS to small integers with known scale :P take up much less space Sep 11 14:44:22 oGMo: no... But if you want humans with their 10 fingers to agree with a numeric calculation value you need to be careful to round it by human expectations, or avoid rounding/representational errors altogether. Sep 11 14:44:36 g00s: that's fixed-point and rather painful Sep 11 14:44:41 i know :( Sep 11 14:45:09 its also painful that the engine is using 64 bits for a 5 sig-fig float tho :( Sep 11 14:45:38 g00s: I can hear the call for a custom flat file database format of fixed-length records... Sep 11 14:45:42 eh Sep 11 14:45:49 yeah, i have been thinking about that Sep 11 14:45:55 i rarely bother with single floats anymore Sep 11 14:45:58 alankila: i've got a history of having absurd database issues inflicted on me... at one point I had to wedge OCRed documents into a version of DB2 that couldn't do things like case-insensitive LIKEs Sep 11 14:46:10 alankila: nor could it handle any wacky OCR error characters Sep 11 14:46:16 alankila: and did I mention i had to convert it all to EBCDIC? :P Sep 11 14:46:19 dragorn: huh, what db2 is that? And you mean, lower(foo) like :foo wouldn't work? Sep 11 14:46:20 when your integers are 64-bit is it really saving space to use 32-bit floats? Sep 11 14:46:24 (no) Sep 11 14:46:24 then i decided, maybe i will try to split these tables; put each logging run in its own db; but SQLiteOpenHelper is a piece of shit there too Sep 11 14:46:41 alankila: DB2 on CMS on a s/390 around 1996 Sep 11 14:46:41 dragorn: argh, please no more. That ebcdic does it Sep 11 14:46:43 dragorn, case insensitive likes aren't really part of sql92 ... Sep 11 14:46:53 dragorn: haha awesome Sep 11 14:47:06 pfn: no, but explain that to management who "just wants them tobe searchable we already told people the website would be active" Sep 11 14:47:11 The website, by the way, was written in REXX. :P Sep 11 14:47:18 but in 1996, db2 had freetext search extensions Sep 11 14:47:25 ... and this is why the android API only makes me sort of angry, and not totally insane. :P Sep 11 14:47:26 I remember I wrote a few amigarexx scripts once Sep 11 14:47:28 very, very dimly Sep 11 14:47:36 i remember REXX on os/2 Sep 11 14:47:56 i remember it being kinda neat but i don't recall the oddities Sep 11 14:48:19 oGMo: rexx wasn't impossible to deal with. The wackass db2isms and CMSisms were pretty bad tho. Sep 11 14:48:25 dragorn: sure Sep 11 14:48:28 Like I figured out how to crash CMS. That was good times. Sep 11 14:48:29 oGMo the numbers are usually represented as 6 bytes :( sqlite stores 1,2,3,4,6,8 bytes i think depending on size Sep 11 14:48:36 sorry, 4 bytes Sep 11 14:48:40 dragorn, but yeah, freetext search was available as an extender in 1996 Sep 11 14:48:46 pfn: not for us :P Sep 11 14:48:47 though as always, case sensitive search when you downcase everything is the same as case-insensitive search ;) Sep 11 14:48:48 dragorn, I know, because I wrote a bunch of apps that used them :p Sep 11 14:48:58 pfn: on the cms variant of db2? Sep 11 14:49:05 pfn: There's a LOT of db2 versions Sep 11 14:49:23 dragorn, I suppose on z/os was different Sep 11 14:49:27 pfn: power series aix dbs != cms s/390 db2 != intel db2 != mvs/zos db2 Sep 11 14:49:33 pfn: yeah. We had zos db2 as well Sep 11 14:49:35 and it was ims or something Sep 11 14:49:40 pfn: but we didn't use that for this Sep 11 14:49:42 also, android, to keep things topical, now back to ibm stuff from the 90s Sep 11 14:49:51 oGMo: heh Sep 11 14:49:52 yeah sorry Sep 11 14:49:58 woke up angry I guess :P Sep 11 14:50:08 sqlite was originally written for missile guidance system, i thought ? Sep 11 14:50:15 hey i don't care, 90s computing was more interesting than web crap Sep 11 14:50:22 Well, we could keep it topical - lets port hercules to android :P Sep 11 14:50:28 bet we could get zOS booting Sep 11 14:50:36 networking would be a bit of an issue Sep 11 14:50:41 I think we should consider android framework on a C64 emulator. I look forwards to ACTIVITY.BAS Sep 11 14:50:57 I've got some z/os system images somewhere Sep 11 14:51:18 dragorn: hehe Sep 11 14:51:52 dragorn@mediawhore:~$ du -sm IBM\ ZOS\ 1.10/ Sep 11 14:51:52 16488 IBM ZOS 1.10/ Sep 11 14:51:53 :) Sep 11 14:51:55 Greetings everyone. Sep 11 14:52:06 hehe Sep 11 14:54:23 I was hoping someone could help me out :$ I have managed to get my StackView to work to demonstrate a widget with vertical gestures. Sep 11 14:54:52 so here's osmething topical - i know android studio is sort of the bastard stepchild, but here goes... When I make a bunch of drawable resources, and throw them in the appropriate directories and all, and then make an imageview - i can assign one to it w/ the standard image list picker. The previews show up in the image list fine. Then the layout rendering in android studio says it cant' convert the drawable. Sep 11 14:55:55 But i wanted to change to a way where only a single view is visible at a time. The docs made me believe an adapterviewflipper shouldve been able to help me out, but i can't seem to get any gestures to work and also noticed that it doesnt have a onGenericMotionEvent method... does anyone know another element that does have onGenericMotionEvents? Sep 11 14:55:58 hrm, so do people typically use rhino or what for JS app extensions? Sep 11 14:56:14 (Other then the Stackview, since apparently it's not collapsable to a single view) Sep 11 14:57:18 annihlator: I've got no idea, sorry Sep 11 14:57:52 oGMo i think pfn has. i wonder if it will ever be 'retired' by nashorn Sep 11 14:58:02 thats old code too Sep 11 14:58:09 probably rotting ;) Sep 11 14:58:36 anything newer? Sep 11 14:59:53 i dont think for js; i was looking into stuff like luajit / luaj but didn't get far (for lua) Sep 11 15:01:00 oh, v8 is already included heh Sep 11 15:01:26 oGMo on android ? Sep 11 15:02:14 ah cool Sep 11 15:02:38 g00s: right Sep 11 15:03:35 though my guess is it's a little raw .. i.e. not much api etc Sep 11 15:03:54 jni :( Sep 11 15:04:20 hey all Sep 11 15:04:32 ios dever here starting out in android Sep 11 15:04:45 g00s: well sure Sep 11 15:05:41 JeremieDeNoob: welcome to the good side Sep 11 15:05:53 lol not sure about that yet Sep 11 15:06:01 i think its beneficial to have both Sep 11 15:06:12 that way when one of them falls you still have the other Sep 11 15:06:14 hehe, the devs i know that do both often dont see android as the good side Sep 11 15:06:26 haha Sep 11 15:06:28 lol its not bad so far Sep 11 15:06:38 i did some blackberry 10 dev and it was very similar Sep 11 15:06:48 except it was with c++ instead of java Sep 11 15:06:54 but the general structure is the same Sep 11 15:07:03 Can I put the button inside the text of TextView ? Sep 11 15:07:04 well android is mostly java so what do you want .. in the end the goal's to make your app available to more paying customers Sep 11 15:07:53 yep but i like it better than the BB because its way better implemented with eclipse Sep 11 15:08:15 heh Sep 11 15:08:16 not as buggy, no warnings to ignore, and setting up test devices (virtual and real) is a breeze) Sep 11 15:08:30 blackberry always seemed painful ;/ Sep 11 15:09:03 its kind of buggy. its like android just really shitty Sep 11 15:09:04 lol Sep 11 15:09:11 hehe Sep 11 15:09:29 like android (not in the shitty way i mean like the setup is the same, except its shittily implemented) Sep 11 15:10:03 ah, and now im reading about dialogs and toasts... it is the same... Sep 11 15:10:23 even the colour scheme is the same Sep 11 15:10:29 with the black and blue thing Sep 11 15:10:40 its like rihanna Sep 11 15:12:16 are there any favored methods for uploading data from the device to a server? Are people using FTP, POST, etc? Sep 11 15:12:49 JeremieDeNoob after they decided to refocus from making android a camera OS, didn't they decide to copy BB ? Sep 11 15:13:19 <\0x90\> two_tired, depends on the data? Sep 11 15:13:41 \0x90\>: GPS data Sep 11 15:14:03 <\0x90\> well i'm doing that right now, but i made my own python daemon and using sockets Sep 11 15:14:10 <\0x90\> i wouldn't use ftp for that though. Sep 11 15:14:27 You are uploading in realtime or batch? Sep 11 15:15:06 <\0x90\> i'm uploading every X seconds. but my android is always connected to a python server Sep 11 15:15:14 <\0x90\> so it just sends.. updates. Sep 11 15:15:18 OK Sep 11 15:15:29 <\0x90\> if you don't have any service on the other end, you could use http Sep 11 15:15:49 My criteria is daily batch uploads, with some kind of authentication Sep 11 15:17:33 <\0x90\> and in what format is your data, on the phone ? i personally would make me own daemon for that on the other end. Sep 11 15:17:52 just CSV.bz2 data Sep 11 15:18:29 what would be the advantage of writing a daemon over using HTTP PUT or FTP? Sep 11 15:19:03 <\0x90\> well if you're doing it like that, then its different Sep 11 15:19:13 <\0x90\> http is better i guess Sep 11 15:19:35 one other option is to send directly to an AWS S3 bucket Sep 11 15:19:49 maybe just put it on RabbitMQ Sep 11 15:19:52 <\0x90\> ftp is known to suck in many situations (behind nat/firewalls) so i would stay away from it. Sep 11 15:21:03 I want it to ultimately be in an S3 bucket, the question is if it's less work to have the App do an HTTP POST and save to S3 on the backend, or bypass the server altogether Sep 11 15:21:52 <\0x90\> i never used s3 so not sure, but if you can use s3 without any difficulties than it's better to bypass http post to some server... if ultimately you want it on s3 Sep 11 15:22:12 aws's java lib is reeeeeally nice Sep 11 15:22:13 yeah, there is just the question of authentication with S3 Sep 11 15:22:27 may not want to have our AWS key hard-coded into the App Sep 11 15:22:41 so the App could connect to server, get key, then use that to upload... Sep 11 15:23:05 two_tired: do the files need to be private? Sep 11 15:23:22 fdans:yes Sep 11 15:23:43 <\0x90\> two_tired, maybe a combination of http request to some server, get the encrypted key.. decrypt it, and use it to authenticate then upload to s3 ? Sep 11 15:23:48 you definitely don't want your key hard-coded Sep 11 15:24:03 you'd also better have some way to authenticate that http req Sep 11 15:24:53 yeah, I was hoping there were good examples of implementing this stuff Sep 11 15:25:18 oGMo, rhino works fine for app extension Sep 11 15:25:32 oGMo, who cares if the code is old... people care too much about that, fix stuff as necessary Sep 11 15:25:36 <\0x90\> well it's usually hard to find the specific stuff.. stuff that u need. but figure out what you want.. and how you want it to work, then go step by step. Sep 11 15:25:38 oGMo, and for that matter, rhino is more featureful than v8 Sep 11 15:25:47 oGMo, and yes, java bindings into v8 are poor Sep 11 15:25:58 pfn: sure though v8 is probably massively faster Sep 11 15:26:04 <\0x90\> so in your case .. it would be.. do the auth, get the key.. use the key to auth, upload file. Sep 11 15:26:11 and bindings aren't terribly difficult to accomplish Sep 11 15:26:46 two_tired: security is hard. So - if you don't do SSL (properly!) your logins are exposed for the HTTP fetch of the key. If you store it on external storage, anything can read it. if you store it on internal storage, a user w/ a rooted phone can read it Sep 11 15:27:57 two_tired: w/out knowing your full needs/security model, I'd say you could do cert pinning to validate the SSL exchange w/ the server (i assume you know that) and then could either use the android keychain API for storing the key in secure storage (with some user annoyance of requiring a pin/etc on the device), or consider using a reflector you host into S3, so you never actually give the S3 key to the end device Sep 11 15:28:37 two_tired: if you've got one S3 key for all users, you're going to HAVE to run a reflection service of some sort, since you can't trust a user device to be secure against that user. If you have per-user keys you could maybe get away with internal storage, tho I'd still be a little twitchy about it Sep 11 15:29:47 is there any way to upload an apk to Play and specify that you don't want users' Play-auto-update to work in this case? Sep 11 15:29:53 yeah, it sounds like it would be more secure to just do the S3 transfer from the server Sep 11 15:30:16 two_tired: if you have a company S3 key and plan to upload all the user data to that one ID, you'll *have* to do it on a server Sep 11 15:30:28 oGMo, perhaps not Sep 11 15:30:36 depends on how v8 is exposed... Sep 11 15:30:42 I see Sep 11 15:30:43 oGMo, it's not an official cts library or anything Sep 11 15:31:00 two_tired: because you can't trust an end-user device to keep that secure. If messing w/ my phone gets me access to EVERYONES data? That's a real problem / incentive to someone malicious Sep 11 15:31:08 v8 through a c++ lib it looks like .. simple enough api Sep 11 15:31:17 two_tired: an incentive which isn't there if messing w/ my phone only got me into my own data, if you follow Sep 11 15:31:27 i am no fan of c++, but then i'm not a fan of java either Sep 11 15:31:43 two_tired: so you're going ot have to do something where you log into your own server, and give the user a token, then use that token to auth them doing things Sep 11 15:31:45 but whatever gets the job done Sep 11 15:32:12 two_tired: so then they send you the file and you send the file to S3 Sep 11 15:32:22 OK, then I just need a way for the App to authenticate with the server so we don't get fake uploads Sep 11 15:33:19 then sticking stuff in S3 may be as simple as using s3cmd Sep 11 15:34:17 two_tired: depending on how the crypto works, you may be able to ask the server to sign a request (and it validates the path), then the phone uses that signature to talk directly to s3, keeping the bandwidth off that middleman Sep 11 15:34:34 if the phone tries to access something it shouldnt, dont sign it Sep 11 15:34:35 so you want to avoid saving auth data on the phone (login/password). You also want to encrypt all comms to the server, and you *must* validate the SSL certificates somehow, either by having them be properly signed commercial certs, or by storing the fingerprints in your app and validating by fingerprint Sep 11 15:34:45 clever: the middle man will be on EC2 as well Sep 11 15:34:46 oGMo, where is this lib anyway? Sep 11 15:34:59 any guide that "solves" SSL by turning off fingerprint validation should be burned to the ground Sep 11 15:35:14 two_tired: yeah, but depending on the volume of traffic, it may be simpler to just go directly to s3, then to bounce it off ec2 Sep 11 15:35:44 and it depends on how complex the crypto is Sep 11 15:35:54 pfn: i think it's part of the web stuff, but as long as you link that you're good apparently. haven't tried it yet Sep 11 15:36:02 i like the idea of client ssl certs to validate the end-user device Sep 11 15:36:03 would probably be trivial to include your own build too Sep 11 15:36:04 I'd suggest the login happen once (or rarely), and that you give the user some sort of hashed key that gives them future access to your service, which you can then revoke Sep 11 15:36:13 every user gets their own 2048bit keypair Sep 11 15:36:16 oGMo, I wouldn't say trivial, but yeah, it should be reasonable Sep 11 15:36:38 dragorn: that can be done with client certs, login once, and bind the cert to an account Sep 11 15:36:49 clever: yeah Sep 11 15:36:54 clever: agreed Sep 11 15:37:25 re: direct comms to s3, I don't know if s3 lets you have brokered transactional authentication w/out the users master key? Sep 11 15:37:47 dragorn: i think it wants several of the headers, including Date: to be signed? Sep 11 15:38:02 and you could just have a middleman sign those for you, s3 wont know the difference Sep 11 15:38:04 if so, go for it - what clever said. Auth the transaction, give the device a key, and have it send it to S3 directly Sep 11 15:38:14 don't ever ever ever give the end user device your s3 data key tho Sep 11 15:38:18 clever: cool. I don't know jack about s3 Sep 11 15:38:29 let me find the docs Sep 11 15:38:29 OK, I'm pretty sure direct path to s3 is off the table Sep 11 15:38:38 two_tired: listen to clever - he might have solved it Sep 11 15:39:16 two_tired: device asks your server to allow it to do something, your server tells the device the signed headers to do that operation, your device does it to S3. At no point does the device have direct S3 access to do whatever it wants Sep 11 15:39:48 then you just hav eto solve the device-to-your-service auth :P Sep 11 15:39:56 security is hard. unfortunately. Sep 11 15:39:58 two_tired: http://docs.aws.amazon.com/AmazonS3/latest/dev/RESTAuthentication.html Sep 11 15:40:19 and http://docs.aws.amazon.com/AmazonS3/latest/dev/UsingHTTPPOST.html Sep 11 15:40:23 I see Sep 11 15:40:47 one thing I haven't mentioned is that this app has low expectations for success Sep 11 15:40:58 we don't expect much volume Sep 11 15:42:00 two_tired: hmmm, under the upload examples page, it shows creating a json object for the policy, base64'ing it, hmmm, odd, and the access key is a bare field in the post data Sep 11 15:42:04 so I'm not worried about bandwidth from ec2->s3 Sep 11 15:42:07 i think this is a dumbed down version, to support
Sep 11 15:43:02 if the volume is low, just use normal file upload posts to https, and then normal s3 libs server side Sep 11 15:43:21 clever: yeah, that seems like the way to go Sep 11 15:43:54 if we do a larger project with more potential, we'll probably want to consider s3 PUT/POST from device Sep 11 15:43:58 ah, here it is, http://docs.aws.amazon.com/AmazonS3/latest/dev/RESTAuthentication.html Sep 11 15:44:02 scroll down to 'example object put' Sep 11 15:44:31 it shows the http headers, the string they get formated into, and then that string is signed, and becomes an extra header Sep 11 15:45:35 and there is a header with the md5 of the content, so the content is also covered by the signature Sep 11 15:45:46 I see Sep 11 15:45:49 nothing says the phone has to do the actual signing of those headers Sep 11 15:46:13 https post that string to the server, it signs it with the secret key, then continue as the docs say Sep 11 15:47:42 hardest part is just getting the hmac stuff working right, one wrong byte at the input and the output is totally different Sep 11 15:48:27 one advantage of saving to s3 server-side is that it would be that much easier to switch to an alternate storage system/service Sep 11 15:48:37 ah yeah Sep 11 15:48:48 hide the s3 from the app, and then the app doesnt care Sep 11 15:48:59 http://docs.aws.amazon.com/AmazonS3/latest/dev/AuthUsingTempFederationToken.html Sep 11 15:49:10 then you wont need this special stuff to create temporary secret keys Sep 11 15:49:33 I see Sep 11 15:49:44 no Perl?!! j/k Sep 11 15:50:06 i think the federation stuff gives equal access to the account that made it, but you could have one bucket per user, but that could be messy Sep 11 15:50:20 i'm only using s3 for the cost allocation reports Sep 11 15:50:55 I use s3 for reports and storing backups with s3cmd, so far Sep 11 15:51:25 i checked the math, and uploading ~200 files daily to s3 would just cost too much Sep 11 15:51:34 simpler to just scp them into ec2 Sep 11 15:52:19 that may change if i check to see if the backup has even changed Sep 11 15:53:06 clever: and you can always archive to s3/glacier from ec2 or something Sep 11 15:53:29 s3fs if you're lazy though i couldn't quite get it to work right Sep 11 15:53:38 oGMo: yep, but the per-transaction costs are steep, its not cheap to upload 200 files a day with s3cmd Sep 11 15:53:53 (and it's not really lazy but you can't access it in parallel which i don't like) Sep 11 15:53:59 clever: exactly Sep 11 15:54:04 clever: but you can archive them and upload 1 Sep 11 15:54:10 I do tar.bz2 things up first Sep 11 15:54:22 yeah, tar them all up into one big lump Sep 11 15:54:25 clever: or better, archive 6000 and upload 1 a month ;) Sep 11 15:54:44 the daily backup is so if something does go wrong, i only loose a days worth of data Sep 11 15:54:50 ah Sep 11 15:54:58 once a month uploads defeats that Sep 11 15:55:23 well ec2's volumes are low fail arent' they? not the ephemeral stuff Sep 11 15:55:33 still Sep 11 15:55:36 1 a day shouldn't hurt much Sep 11 15:55:36 yeah, ive not had a failure yet Sep 11 15:55:47 biggest problem is just misplacing an entire database Sep 11 15:55:54 heh Sep 11 15:55:58 i cant figure out which host it was on! Sep 11 15:56:02 you also have to protect against rm -rf / Sep 11 15:56:15 not as bad as misplacing your server, but annoying Sep 11 15:56:17 yeah, thats the major benefit of s3 Sep 11 15:58:13 if i have 500 devices logged in the same google account, can i install an app in X of them based on the hardware specs? Sep 11 15:58:16 or the model Sep 11 16:02:25 two_tired: another neat benefit of s3 that i forgot about, they can run a bit torrent seeder for any file in the bucket! Sep 11 16:09:45 can an app configure an additional google account for the user? Sep 11 16:12:02 luist, no Sep 11 16:12:06 no, but the walrus has run through all of his pants Sep 11 16:12:10 well, maybe Sep 11 16:12:12 but don't Sep 11 16:12:34 pfn: errr Sep 11 16:12:41 its for a good reason ^^ Sep 11 16:12:47 no it's not Sep 11 16:13:15 its for automating rating apps 5*s in the android market Sep 11 16:13:21 neat, my chromeipass has 66k users on chrome and lastpass has 116k users Sep 11 16:13:25 interesting Sep 11 16:13:37 how to leverage that into sales for keepshare... Sep 11 16:14:30 hmm, lastpass has 270k on firefox and by comparison, passifox barely beats 1% of that :-/ Sep 11 16:15:10 pfn: well consider that all students in the public schools in a city are receiving a tablet to help in the studies, and there will be some specific apps for them in this project. A common account to send them all the apps and updates automatically would be very usefull Sep 11 16:15:15 make keepshare free Sep 11 16:15:21 and just borrow passwords from time to time Sep 11 16:15:29 pfn: and configuring this additional account for them automatically would be even better Sep 11 16:15:36 an MSDN login here, a porn password there, it all adds up to tremendous savings Sep 11 16:18:19 luist: you might be better off with a custom 'app store' type app thats preinstalled on everything Sep 11 16:19:09 luist: that app would provide a list of all of your custom apps, and notify them when updates come out, gcm for the notifications if you want to force updates Sep 11 16:19:37 clever: so you want me to implement a new app store and not use the default one which already works :T Sep 11 16:20:05 luist: your custom one would let you manage everything within your users, without having to publish every single app on the play store Sep 11 16:20:10 sharing a single google account across thousands of tablets is not the right approach Sep 11 16:20:17 what pfn said Sep 11 16:20:19 if you think it is.... Sep 11 16:20:22 yeah, you have problems Sep 11 16:20:47 is the app store open source? Sep 11 16:21:07 luist: the play store isnt, but its not a very complicated thing to do just what you need Sep 11 16:21:15 no Sep 11 16:21:41 well, you can't autoinstall... Sep 11 16:21:43 luist: http://stackoverflow.com/questions/9208573/android-install-apk-with-intent Sep 11 16:21:58 download the apk file, use this intent to start the install process Sep 11 16:22:04 notification manager to tell the user updates are ready, to start the above process Sep 11 16:22:26 listview/adapter to list all apps Sep 11 16:22:39 payment, skip that, you likely arent needing it Sep 11 16:22:46 I think F-Droid acts like its own Store. https://f-droid.org/ Sep 11 16:23:09 yep, that could also be used, but it doesnt check for updates automaticaly Sep 11 16:23:11 And is open source. Not sure if it has automatic updates, but it has its own repo where it lists apps. Sep 11 16:23:19 i understand your reasons.. but the thing to install remotely is so good Sep 11 16:23:21 Could be added, I guess. Sep 11 16:23:23 you need to start it and download the package list manualy Sep 11 16:24:01 luist: there should be a permission somewhere to let your app install things without the users consent Sep 11 16:24:09 it might be as simple as putting it in /system/app/ Sep 11 16:24:22 Who owns the tablets? I guess there's options like Freeflight or mobile device managment software to accomplish that, similar to what companies use. Sep 11 16:25:16 luist, you can always ship your own rom that allows auto-install Sep 11 16:25:23 lol Sep 11 16:25:30 That might be over doing things, no? Sep 11 16:25:38 you're gonna catch negative publicity flak for whatever way you decide to do Sep 11 16:26:29 heh Sep 11 16:26:52 pfn: im not responsible for the rom Sep 11 16:26:53 Can't the app itself just autoupdate? (I guess that was suggested already, didn't see the backlog) Sep 11 16:27:18 apps can't self-update Sep 11 16:27:21 less work if thats centralized into a single management app Sep 11 16:27:35 pfn: well…. play store can update stuff automatically when on wi-fe Sep 11 16:27:39 wi-fi* Sep 11 16:27:39 pfn: mine does it all the time, but its a bit messy, i just open an apk url in the default browser Sep 11 16:27:50 luist, that's play store and system signatures Sep 11 16:27:55 clever, that's not self-updating Sep 11 16:27:58 that's just self-downloading Sep 11 16:28:03 the user still has to manually press install Sep 11 16:28:08 yeah Sep 11 16:28:17 pfn: i dont press install to install any update Sep 11 16:28:28 pfn: everything just updates when i connect on wi-fi Sep 11 16:28:28 luist, this is really out of your league, please don't consider doing this Sep 11 16:28:37 pfn: http://stackoverflow.com/questions/9208573/android-install-apk-with-intent would this ask the user when ran from a /system/app/ package? Sep 11 16:28:38 pfn: on my HTC ONE Sep 11 16:28:58 clever, yes Sep 11 16:29:18 pfn: how does the play store install without asking then? Sep 11 16:29:23 clever, system signatures Sep 11 16:29:31 and that is? Sep 11 16:29:45 the signature allows it to Sep 11 16:29:50 What about http://www.androidcentral.com/google-releases-admin-app-managing-enterprise-education-and-government-accounts Sep 11 16:29:55 pfn: why is this bad? it seems pretty normal considering i allowed it Sep 11 16:30:41 pfn: i just jammed the play store apk into /system/app on my kindle, so i clearly done have the right signature Sep 11 16:30:53 clever, the play store apk has the right signature Sep 11 16:30:58 pfn: but its still able to update things in the background all the time Sep 11 16:31:12 so amazon just happend to leave the public key for that in the rom? Sep 11 16:31:14 I suppose it's possible it just accepts systemOrSignature Sep 11 16:31:20 may only need system app permission, rather than signature Sep 11 16:31:23 you'd have to look it up to verify Sep 11 16:31:43 clever, and yes, I would expect amazon to leave google permissions alone for intrepid hackers to install gapps Sep 11 16:31:46 yeah, i'm curious now Sep 11 16:31:48 ^^ Sep 11 16:32:07 pfn: the gcm stuff only works from /system/app, it fails hard if you install it as a normal app Sep 11 16:33:00 and amazon formats /system/ with every OTA update Sep 11 16:33:14 they havent heard of incremental updates Sep 11 16:33:18 apps can self update if they aren't installed through the play store Sep 11 16:33:35 like nova launcher on my phone updates itself Sep 11 16:33:43 i think thats just a policy thing Sep 11 16:33:58 the play store will refuse any app that has that feature Sep 11 16:34:13 since they cant review the future versions for safety Sep 11 16:34:30 mmathis, it updates itself? without root? without being installed in /system? Sep 11 16:34:35 well, nova launcher has a version on the app store that is the same as the "beta" version Sep 11 16:34:38 kakazza: is this opensource? Sep 11 16:34:39 Yeah, they went down hard on self updating apps this year. Sep 11 16:34:45 I guess it may requre root Sep 11 16:34:48 I'm not sure about that Sep 11 16:34:49 luist: Google's App? No, I doubt it. Sep 11 16:35:04 but, the beta version updates itself Sep 11 16:35:04 still good tho Sep 11 16:36:04 actuall I don't think it needs root Sep 11 16:36:09 I have no idea how they do it Sep 11 16:36:20 mmathis, by updating itself, it does it prompt-less? Sep 11 16:36:35 another possibility is using a custom dex/res-apk loader Sep 11 16:36:38 I just know that it was like, "if you want to use the beta/self updating version you have to download it from our site because the play store doesn't let apps update themselves" Sep 11 16:36:43 that would allow a silent update Sep 11 16:36:50 it does prompot Sep 11 16:36:52 mmathis, but does it actually update itself Sep 11 16:36:56 no, it doesn't, it just prompts, ok Sep 11 16:37:37 oh yeah I guess it doesn't do it automatically Sep 11 16:37:42 it prompts to download an apk Sep 11 16:37:45 and then installs that Sep 11 16:39:28 https://gitorious.org/f-droid/fdroidclient/source/862d7b11836d0d4f0501cf5b0ae017769a4c528b:src/org/fdroid/fdroid/AppDetails.java#L799 Sep 11 16:41:03 pfn: nothing special in this case, no sign of any flags to force it into the background Sep 11 16:41:10 ? Sep 11 16:41:18 clever, it starts the activity which results in a prompt Sep 11 16:41:38 yeah Sep 11 16:41:47 was wondering if it had code to make it not prompt Sep 11 16:41:54 no one has code to make it not prompt Sep 11 16:42:14 the play store is clearly capable of it, just wondering if it needs an extra flag Sep 11 16:42:28 it uses a different intent and permission Sep 11 16:42:38 do you know which intent? Sep 11 16:42:40 It's simple. Just ship code in javascript or similar form and make a fixed shim that passes everything to javascript which you can update any time you want Sep 11 16:42:53 alankila, or a custom dex/res-apk Sep 11 16:42:55 or compile position independent executable Sep 11 16:42:59 all you need is the basic Sep 11 16:43:03 then you can use a fixed shim to call into it Sep 11 16:43:08 alankila: thats just wrong, http://sealedabstract.com/rants/why-mobile-web-apps-are-slow/ Sep 11 16:43:13 launcher Sep 11 16:43:16 why idiotic blog post is idiotic Sep 11 16:43:32 alankila: it explains why mobile js performance is so bad Sep 11 16:43:42 seems reasonable to me Sep 11 16:43:50 mobile js performance Sep 11 16:43:54 yeah scripting language and gc is evil, or whatever Sep 11 16:43:54 just quit trying to use it for bullshit like animations and graphics Sep 11 16:44:12 but you cant get away from GC, proper android apps still have GC Sep 11 16:44:14 It's just fine when you can afford the overhead. Sep 11 16:45:08 the problem with mobile is that it's far closer to the threshold where it's not acceptable, though. Sep 11 16:45:27 where is the free lunch that make today's shitty bloated apps tomorrow's lightweight speed demons? Sep 11 16:46:15 it's as if we were hitting limitations on electronics we can make and deploy and the amount of computation per watt :-/ Sep 11 16:47:50 I for one sincerely hope we have a few breakthroughs undiscovered yet, perhaps changing from silicon to something else, or embracing probabilistic computing... Don't know, but would be nice to have something. Sep 11 16:47:50 pfn: oh, i just noticed, 'adb install' doesnt give a prompt! Sep 11 16:48:11 clever, that's surprising? Sep 11 16:48:24 adbd clearly has elevated privileges Sep 11 16:48:28 * alankila idly wonders if android app could contact adbd running on the device... Sep 11 16:48:40 pfn: yep, but thats a good place in the code to start looking, at how it does it Sep 11 16:48:44 and pretend to be a PC giving it commands. Most likely not. Sep 11 16:48:53 alankila, if you enable adb over wifi Sep 11 16:48:54 sure Sep 11 16:49:01 alankila: i have noticed one odd thing, if you run adb on the phone, it acts like adb on a desktop Sep 11 16:49:06 you can even do adb devices Sep 11 16:49:13 but it wont see itself Sep 11 16:49:26 the most you could maybe do, is tap into a 2nd phone over usb otg Sep 11 16:49:50 turning on adb over wifi requires root, or a real pc connected to usb Sep 11 16:49:54 Doesn't sound deployable anyway Sep 11 16:50:05 most phones don't run development mode Sep 11 16:50:25 in fact it's even a challenge to work out how to get into it since 4.2 or 4.3 Sep 11 16:50:32 yeah i'm doing that now with adb tcpip Sep 11 16:50:37 for some glass shit Sep 11 16:51:21 ? Sep 11 16:51:40 Does wifi direct on android support multiple, simultaneous connections? Sep 11 16:52:17 pfn: hmmm, the pm command appears to use IPackageManager.installPackageWithVerificationAndEncryption which isnt in the docs Sep 11 16:52:34 clever, of course, it's private/internal api Sep 11 16:53:03 can be accessed with reflection though, unlikely to do any good however Sep 11 16:53:26 or if it works out then you should report it to google as a potential security issue. Sep 11 16:53:41 reflection only works within your dalvik vm Sep 11 16:53:47 user is *supposed* to be in control of this shit Sep 11 16:53:58 this is likely using inter-process rpc to call functions on a central instance Sep 11 16:54:40 ive looked at the sqlite code before, it stores c pointers in java private ints, in theory reflection can change the pointer and exploit sqlite calls Sep 11 16:55:03 but those are done under the same uid as your code, and the NDK gives you the same access, much simpler Sep 11 16:56:02 yeah what we really need is calls to some central process of trusted nature that performs something unexpected, maybe by jumping into native code Sep 11 16:56:12 exploits like that are probably there but could be difficult Sep 11 16:57:09 i have found something similar Sep 11 16:57:26 the notifications are remoteviews rendered by the systemui process Sep 11 16:57:44 amazon modified them to break the time since notification (.setWhen) Sep 11 16:57:58 and something gmail does, causes a null pointer exception, in systemui Sep 11 16:58:02 boom, no home button, no status bar, no back button! Sep 11 16:58:48 alankila, it communicates through IBinder which will block on permissions, so no, using reflection to access will do nothing Sep 11 16:59:19 clever, that's amazon's fault Sep 11 16:59:34 pfn: yep, but i just didnt expect it to be able to crash such a major process like that Sep 11 17:00:10 pfn: it also refuses to tell me how many hours until an event occurs, just gives a date, 2013-09-11, even if its 1h 5mins away Sep 11 17:00:10 the whole remoteview thing is rather ugly chicken too Sep 11 17:00:49 alankila: DateTimeView looks like a great tool, but its not public, and amazon broke it Sep 11 17:01:05 in its stock form, its perfect for my app Sep 11 17:01:09 so I'm not too surprised if you can discover ways to crash stuff with it. It's just not a very elegant approach. Sep 11 17:01:37 damn, was hoping I could hax0r repo to work on windows Sep 11 17:01:38 guess not Sep 11 17:01:41 booo Sep 11 17:01:49 I wonder what would happen if I ran it over smb Sep 11 17:02:26 filteredFlags = flags | PackageManager.INSTALL_FROM_ADB; Sep 11 17:02:39 pfn: aha!, thats one way it could be forcing the prompt to go away Sep 11 17:02:58 if (uid == Process.SHELL_UID || uid == 0) { Sep 11 17:03:06 clever, yes, but what's the point of looking at this, the prompt won't go away for a non-system/signature app Sep 11 17:03:28 pfn: i'm wondering how a system app would do it Sep 11 17:03:38 * pfn shrugs Sep 11 17:03:43 if you're gonna write/ship a system app, ok Sep 11 17:05:45 drat, can't repo init over cifs either Sep 11 17:17:24 huh, Xpra is kinda cool. self tuning protocol for various latencies Sep 11 17:35:34 the current version of the maven android plugin can't reference aar right? Sep 11 17:35:44 cannot Sep 11 17:37:43 arrr curse you chris banes Sep 11 17:37:50 use sbt Sep 11 17:37:50 :p Sep 11 17:42:00 running ant after gradle feels like i'm using some highly optimized-for-speed thing written in hand-tuned assembly language Sep 11 17:42:24 myabe i should keep the ant build around :P Sep 11 18:00:34 How to set imageview Sep 11 18:00:49 To stretch like with 160 dpi screen Sep 11 18:02:49 Set width and height converting px to dp ? Sep 11 18:03:12 i mean 150x150 px picture i should set to 150dp x 150dp ? Sep 11 18:12:50 huh? Sep 11 18:13:05 if you want your view to be 150dp then set it to be 150dp, and set scale to fitXY Sep 11 18:17:16 hallo, i have a problem in my app infrastructure... can someone guide me in the right direction... my apprach -> broadcastreciever -> service -> Asynctask with REST-Call -> if response = true () -> Asynctask write response to sqllite. Why does my UI freeze ? Sep 11 18:22:06 perhaps the call isn't actually happening asynchronously? Sep 11 18:22:27 <_tree_branch> Hi, I am not sure if this is the correct platform to ask this question. I am trying to build a functionality where I can automatically play some prerecorded message when a call comes in. User does not need to answer the call. And this is obviously on a custom Android build. Any pointers? Sep 11 18:22:51 lambda forgo asynctask, just use intentservice :| Sep 11 18:23:25 _tree_branch: sure using an incoming call broadcast receiver, maybe a PhoneStateListener if necessary Sep 11 18:24:08 nseidm1: you are suggesting that it can play audio into the phonecall. Are you claiming that this is possible? Sep 11 18:24:09 <_tree_branch> nseidm1: that would only allow me to check the phone state like ringing, in call.. etc. What I would like is more like incoming call interceptor Sep 11 18:24:45 whats wrong with asynctasks? :/ Sep 11 18:24:48 _tree_branch: you can have something answer when it rings, but I am quite certain that you can not send audio into the call Sep 11 18:25:19 doing it via the speaker would cause it to be canceled out Sep 11 18:25:27 (the same reason speakerphones don't feed back) Sep 11 18:26:41 <_tree_branch> I guessed it would be a tricky thing, was not sure if it was impossible. I will try to explain the thing again Sep 11 18:26:59 the bottom line is youw ant it to send audio to the person on the other end, and you can't Sep 11 18:27:04 the rest is kind of moot Sep 11 18:27:50 root66: you can't abort an incoming call broadcast Sep 11 18:27:55 you can abort an outgoing call broadcast Sep 11 18:28:04 look, none of that matters Sep 11 18:28:13 you can't stream your audio, period Sep 11 18:28:38 if you are into electronics engineering, look into making a device that utilizes an earpiece Sep 11 18:28:39 stream audio over the phone call, lol, heck no Sep 11 18:28:58 have the device play audio, sure, why not Sep 11 18:29:00 that's what a ringtone is Sep 11 18:29:34 You said you wanted it to play a prerecorded message when someone calls, and the user does not have to answer the call Sep 11 18:29:41 so were you saying you want a prerecorded message played ON YOUR PHONE? Sep 11 18:29:48 you want a custom ringtone? Sep 11 18:30:28 <_tree_branch> no, I want that the call is not answered in the default manner. Instead the caller gets a prerecorded message Sep 11 18:30:37 not possible Sep 11 18:30:39 <_tree_branch> it is like voicemail, but not a voicemail by a carrier Sep 11 18:30:54 yeah that is completely impossible Sep 11 18:31:56 all calls have to be answered in order to play a recording, but you might be able to hack it to not ring/vibrate until it hears a certain tone or phrase in the answer. Sep 11 18:32:26 of course, this hack would be at the rom level rather than the app level Sep 11 18:32:28 you can't delay an incoming call Sep 11 18:33:08 you can use root and radiooptions to kill an incoming call Sep 11 18:33:29 you can use reflection on most manufacturers with TelephonyManager to kill an incoming call Sep 11 18:33:37 varies from manufacturer to manufacturer Sep 11 18:34:06 <_tree_branch> yes, I agree. I was not talking at the app level Sep 11 18:34:18 <_tree_branch> In other words, is such a thing possible at rom level? Sep 11 18:34:25 yes Sep 11 18:34:29 if you have enough access/source to be able to record the traffic/voice on the line, you can certainly apply speech recognition and choose when to trigger the phone's ringer Sep 11 18:34:58 and to otherwise record the traffic to a file Sep 11 18:35:29 _tree_branch: at the rom level you can answer the incoming call and have an app play audio over the call's channel Sep 11 18:36:37 you're heading down the road of a system level app, and depending on the mods you want to make to the framework root level shell commands may still be required Sep 11 18:37:32 <_tree_branch> nseidm1: I understand. any pointers in this direction? Will this involve working at the RIL level? Sep 11 18:37:58 likely the audio manager Sep 11 18:38:11 all shell commands access some kernel or framework api -- you can write code to access such apis yourself without shell commands... Sep 11 18:39:38 <_tree_branch> but I need to somehow cut off the usual path of an incoming call? just the audio manager may not work out Sep 11 18:39:45 RIL level, no, you can use TelephonyManager to answer the call. Everything else is at the AudioManager level. Sep 11 18:42:45 <_tree_branch> okay Sep 11 18:44:55 back to my question, why a asynctask with sqllite write blocks the ui-thread? makes no sense for me Sep 11 18:45:08 lambda it doesn't :) Sep 11 18:45:14 you did something wrong Sep 11 18:45:22 hello guys Sep 11 18:45:42 Is there a way to make a call to a method cause a compiler error? Sep 11 18:46:51 can a heavy db write use 100% cpu so the whole system freeze? Sep 11 18:46:56 jaynewstrom: sure, remove the method. Sep 11 18:47:43 or make it private or otherwise inaccessible Sep 11 18:48:11 well yeah, but I want the method only to be called dynamically at runtime. I don't want someone to accidentally call it. I'm using EventBus, so it needs to be "package-private" Sep 11 18:48:39 my action lasts ~ 5-6 seconds ... the strange thing: the ANR-Dialog doesn't appear Sep 11 18:49:30 I was thinking something like @Deprecated, but that causes an error, not a warning. Sep 11 18:50:51 lambda: if you write to db in your ui thread, then the ui thread can't do anything. the system as whole can still do other things. Sep 11 18:51:13 <_tree_branch> nseidm1: To make things more interesting, here is another question. When a user calls, the phone rings, it is answered automatically but the call is not answered in traditional way (nothing goes via mic). Is it still possible? Sep 11 18:51:38 _tree_branch, phone audio is not controllable via android Sep 11 18:51:40 you planning on manufacturing a custom phone in china? Sep 11 18:51:58 Hi all, I need to develop a dialog with a layout that is separated in 3 steps. Step 1 ..., Step 2... Step 3... Done. How can I do this? Sep 11 18:51:59 oh, you want to hack on the ril Sep 11 18:52:10 if you want to hack on the ril, it is outside the scope of #android Sep 11 18:52:43 FrancescoV: ViewAnimator Sep 11 18:53:01 ViewFlipper is an easy way Sep 11 18:53:22 <_tree_branch> nseidm1, pfn: not exactly a custom phone :) It is a crazy idea some of us had Sep 11 18:53:46 alankila: but i'm not writing from the ui... i have a service with recurring tasks -> all starts from the background / asynctasks :( Sep 11 18:53:46 so a custom rom. Sure you can do anything. You can make the phone give a full reacharound with enough mods Sep 11 18:53:51 * g00s is getting too comfy with genymotion; hopes it isn't too expensive once it goes out of beta Sep 11 18:54:08 tnx both :) but what the main difference between the 2? Sep 11 18:54:29 ViewFlipper is the super class Sep 11 18:54:33 ViewAnimator is a subclass Sep 11 18:54:37 essentially the same thing Sep 11 18:54:40 I thought it was the other way around Sep 11 18:54:46 something like that :-P Sep 11 18:54:47 thats really wierd Sep 11 18:54:55 how I can see witch lib is ministro going to download and install??? Sep 11 18:54:57 ViewFlipper is the subclass Sep 11 18:55:06 and you can define the android:inAnimation and android:outAnimation Sep 11 18:55:09 and just use showNext Sep 11 18:55:25 can the broadcastreciever be the problem? Sep 11 18:55:42 so viewflipper is easy to use? Sep 11 18:58:15 yes Sep 11 18:58:25 it's a ViewGroup Sep 11 18:58:32 Can anyone help me figure out what I'm getting a Stub! exception when running the uiautomator through the maven-android plugin? Sep 11 18:58:46 lambda: okay. In theory there is both disk and cpu contention from any activity but the relevant schedulers should insulate other tasks from feeling the contention too harshly. Sep 11 18:58:57 <_tree_branch> nseidm1, pfn, root66: Thanks for your inputs! Sep 11 18:59:33 lambda: so the answer is that the system should not appear to freeze even when doing heavy work but perhaps there is something wrong with the hardware or something. The insulation can never be perfect. Sep 11 18:59:33 achuinard, it's running on the jvm and not in android Sep 11 19:00:02 thank you... i think its not the hardware but me ... :/ Sep 11 19:00:37 anr dialog doesn't appear for non-activities by default Sep 11 19:01:26 if the work is not being done in the ui thread (and the UI doesn't force synchronization somehow), I can't say why the application would appear to freeze. Sep 11 19:01:49 pfn: It happens twice, once as a maven unit test but another time it says the device is found and that the RuntimeException occurred on the device (I believe). https://github.com/stephanenicolas/Quality-Tools-for-Android/issues/12 Sep 11 19:01:51 onPostExecute runs in all conditions at UI-Thread ? Sep 11 19:01:59 I do know of some people who launch asynctask just right but then immediately go and ask for the result of doInBackground on the UI thread, forcing UI thread to stop Sep 11 19:02:20 yes, onPostExecute occurs in ui thread. Sep 11 19:04:24 Why this https://gist.github.com/bluesm/6528244 creates super long view Sep 11 19:04:27 Without text ? Sep 11 19:05:13 I see Sep 11 19:05:23 layout_width="0dp" sorry. Sep 11 19:07:24 you forgot to put a weight Sep 11 19:07:46 also why are you using a width of 0? Sep 11 19:08:03 does anyone use maven as the project build tool ? is it good or was unnecesary ? Sep 11 19:08:17 cool. just learned about PureData. i guess you can use pdlib in android and iOS Sep 11 19:08:36 Tomasso gradle is the new h0t / suck Sep 11 19:09:32 We use maven at work. It is great for having the same build on multiple platforms and such, but it also takes forever on all platforms compared to ant Sep 11 19:09:48 thanks g00s le me check for that xD Sep 11 19:10:21 The xkcd about compiling makes much more sense after using maven =) Sep 11 19:10:25 Tomasso may as well use the official build system; everything else will be perpetually behind at some point Sep 11 19:10:37 build variants, product categories are nice Sep 11 19:10:56 i'm already going nuts with it :D Sep 11 19:11:00 and you use it for android ? says its groovy based Sep 11 19:11:00 Sep 11 19:11:43 Tomasso its the Official (â„¢) build system going forward Sep 11 19:11:58 http://tools.android.com/tech-docs/new-build-system/user-guide Sep 11 19:12:04 great Sep 11 19:12:11 As uses it by default Sep 11 19:12:23 will things still be compiled via ant? Sep 11 19:12:37 but the maven plugin has been around for a while and a lot of people use it Sep 11 19:12:46 doesnt support aar atm though Sep 11 19:13:13 tacticalPro you can use an ant task from gradle if you have some fucky thing you need, but otherwise no Sep 11 19:13:19 *funky Sep 11 19:13:21 O.o Sep 11 19:14:01 and there is also eclipse plugin, i have an android project so i can generate build script ? Sep 11 19:14:06 anyone have a favorite file to bitmap decoder queue? Sep 11 19:14:07 i see Sep 11 19:14:29 i would only use gradle by itself, or with AS atm; the android-plugin is still young Sep 11 19:14:38 i dont think ADT does anything except export gradle Sep 11 19:14:39 shouldn't use ant from gradle Sep 11 19:15:04 like maven, you can use ant tasks if you have something special though Sep 11 19:15:06 becausea nything you could do in ant is much more easily run from gradle Sep 11 19:15:07 but i can always use ant without gradle? ant will always be supported by the sdk? Sep 11 19:15:10 but yeah, shouldn't need it Sep 11 19:15:13 tacticalPro, no Sep 11 19:15:18 ant does not support aar libraries Sep 11 19:15:20 oh Sep 11 19:15:40 well, you can do extra work to convert the aar into a library project Sep 11 19:15:44 but that's not a happy medium Sep 11 19:18:45 does gradlework well without an ide? like just notepad++ and terminal Sep 11 19:18:50 work* Sep 11 19:18:54 tacticalPro, of course Sep 11 19:19:00 good to hear Sep 11 19:19:37 tacticalPro thats kinda the point of maven and gradle; the build should not be driven by the ide Sep 11 19:19:41 or ant Sep 11 19:20:14 i see. Sep 11 19:20:34 gradle plugin is missing one thing atm which i miss: lint support Sep 11 19:20:48 should be easy to add a task to run lint Sep 11 19:20:51 you just run lint... Sep 11 19:21:34 yeah, but i want it in AS ;) Sep 11 19:22:10 so i can click the warning and go there easy. sure i could hack something together but Sep 11 19:22:52 just run 'lint' from gradle, and it shows up in jetgradle Sep 11 19:22:57 the results should be clickable Sep 11 19:23:13 hm Sep 11 19:25:04 pfn: Can you help any more with this Stub issue? Sep 11 19:25:18 hi, i am using eclipse and i have an android project and another simple project that has one class only (the model), i added the project model to the android and already checked inside the Java build path ... > Order and Export, but i am still getting "Could not find class '...' referenced from method ... Sep 11 19:25:18 pfn: I'm really not sure where I went wrong. Sep 11 19:25:26 anyone know why ? Sep 11 19:25:34 no, I don't use maven, so I don't know why it isn't running your code on-device Sep 11 19:26:14 I would switch to Gradle but it sounds like there is no good uiautomator step in Gradle yet Sep 11 19:26:34 what's the uiautomator step? just an instrumented test? Sep 11 19:27:05 ya i'd like to be able to just have one step run a fresh uiautomator with a new app install and take all screenshots Sep 11 19:28:34 pfn new zebralight headlamp look sick (1000 lm) Sep 11 19:29:04 that's pretty badass Sep 11 19:29:09 Cree XM-L2 Sep 11 19:29:25 Hello there! How is it possible that I get a nullpointer on line 50? http://i.imgur.com/ye7cJA6.png Sep 11 19:29:48 Oh, because it isnt a string >< Sep 11 19:31:09 screenshot -.- Sep 11 19:32:02 Zharf: How else would I have shown the debugging information? >< Sep 11 19:32:23 why do you use netbeans? Sep 11 19:32:41 hahaha sucker :p Sep 11 19:32:52 under: Because I like it :) Sep 11 19:33:14 I tried importing this project to Android Studio but it couldn't handle the git submodules. Sep 11 19:33:29 Even tried converting it into a Gradle project, but boy ... that was even worse. Sep 11 19:33:47 sbt! Sep 11 19:33:56 would solve that problem right quick Sep 11 19:34:08 Hi, really inexperienced with the APIs of Android but I'm exploring the possibilities and I have a short question: Would it be possible for an app to gain access to the headphones (in/out/ctrls(?)) to use it for very basic serial communication? Sep 11 19:34:27 Zuchto, yes, square does it Sep 11 19:34:41 pfn: What's square? Sep 11 19:34:54 Zuchto: yeah - because the input is just microphone input Sep 11 19:35:07 ahhhhhhh don't let Jake hear that Sep 11 19:35:09 Zuchto: android also has a pretty solid bluetooth SPP implementation Sep 11 19:35:35 plain rfcomm for the win :) Sep 11 19:36:03 pfn: If I change the build manager, I'll change it to Gradle, when it is a bit more robust. Sep 11 19:37:25 dragorn: the reason I'm asking for headphones in particular is partly thematic for the app I have in mind so bluetooth isn't really an option even if it would be fun to play with too :) But there isn't any problem with android claiming the input/output of the headphones for other uses when a notification arrives or similar? Would one have to guard against those kind of things? Sep 11 19:37:40 Zuchto: you'll probably get audio glitches Sep 11 19:37:58 well, yeah, I guess that would just be expected :) Sep 11 19:37:58 Zuchto: tho it may be possible to request exclusive access Sep 11 19:38:25 Thanks for the input! Sep 11 19:38:30 I could sign 2 apps with the same certificate ? Sep 11 19:38:52 yes Sep 11 19:39:23 Zuchto: tehre are a bunch of audio-parasitic hacks out there Sep 11 19:39:31 Zuchto: planning basic serial encoding over it? Sep 11 19:39:45 pfn: But then if i lost I will lost access to both of those apps ? :] Sep 11 19:39:53 of course Sep 11 19:40:19 ugh UX design hell... too many top-level actions for the actionbar Sep 11 19:40:25 dragorn: I'm basically interested in the easiest possible serial communication between two devices that requires a cable connection :P Sep 11 19:40:48 printer + OCR via the camera Sep 11 19:41:05 and a lot of paper Sep 11 19:42:04 Jake, do you use JIRA at Square? Sep 11 19:42:08 yes Sep 11 19:42:30 Can I assign a background to more than the Activity's layout in the onCreate() handler of the Activity? Sep 11 19:42:33 For open sourced projects like OkHttp etc, do you sync between Github and JIRA? Sep 11 19:42:39 Also I would really like to know if things like the palm-to-silence gesture is vendor specific or if it uses conventional sensors. Anyone who knows/has a good idea? Sep 11 19:42:52 aried3r: i've thought about that. but no, all GitHub for that stuff Sep 11 19:43:56 I see. Because we're thinking about switching to JIRA, but I really don't want to abandon Github :\ Sep 11 19:44:11 And I'm not sure Github for "others" and JIRA for internal makes sense. Sep 11 19:44:55 the GitHub projects are small enough that its issue make sense. our internal stuff is 10x larger in scope and requires something more powerful Sep 11 19:45:17 I think the idea is to have some kind of Kanban-ish virtual whiteboard in JIRA for the tickets. Sep 11 19:45:39 What exactly are the features you like most about JIRA, or plugins you use? Sep 11 19:46:19 reports, components, filters, greenhopper, releases Sep 11 19:46:40 dashboards, of course Sep 11 19:47:30 i like issues Sep 11 19:48:51 dashboards/agile/greenhopper are about the only thing I like about jira Sep 11 19:48:56 I otherwise dislike the clunkiness of jira itself Sep 11 19:49:17 Yeah, we use Greenhopper, too. Sep 11 19:49:25 I wonder if it still works out with >100 tickets Sep 11 19:49:36 100 tickets is nothin' Sep 11 19:50:12 I know, I'd just think it would get pretty crowded. Sep 11 19:54:57 anyone use any 'personal' self-contained OSS bug tracker for private projects? i didn't like bugzilla Sep 11 19:55:06 trac was meh Sep 11 19:55:36 g00s, do recommend me one when you've found one Sep 11 19:55:47 redmine Sep 11 19:55:48 scarab Sep 11 19:56:19 and for projects that never see anyone but you working on them Sep 11 19:56:20 google tasks Sep 11 19:56:33 we use mantis with xmms2 but it's too much effort to set up... and it's php Sep 11 19:56:56 I want to host it myself -.- Sep 11 19:57:00 I'd choose redmine or scarab probably Sep 11 19:57:04 redmine is very trac-like Sep 11 19:57:22 bugzilla, jeez. has anyone ever liked bugzilla? Sep 11 19:57:26 i looked at scarab 5 years ago at least, i thought the project was abandoned. seems like its still going Sep 11 19:57:48 even if it was abandoned 5 years ago, as long as it has all the features you want Sep 11 19:57:50 what does it matter Sep 11 19:58:04 I might write my own bug tracker at some point Sep 11 19:58:16 Zharf but thats what everybody says :P Sep 11 19:58:26 famous last words... Sep 11 19:58:31 g00s, if you want it done properly, do it yourself! Sep 11 19:58:47 how hard could it be? its just a table in a db ... Sep 11 19:58:52 ;) Sep 11 19:58:54 :) Sep 11 20:00:18 for personal tasks, just use a task manager... really nothing complicated... Sep 11 20:00:29 it's only really useful once you start interfacing with other people Sep 11 20:00:40 pfn the intellij task manager ? Sep 11 20:00:55 I want someone report me bugs/feature requests though :) Sep 11 20:00:56 oh, 'a' task manager Sep 11 20:01:26 Zharf it would be cool if there was an activity in your app that had some up/down voting of issues / ideas :D Sep 11 20:01:34 there's lots of those publically Sep 11 20:01:37 getsatisfaction.com Sep 11 20:01:38 is one Sep 11 20:01:52 I want to host it myself :) Sep 11 20:03:34 omg, $425/mo Sep 11 20:03:42 pfn wut the hell were you thinking :P Sep 11 20:03:55 getsatisfaction? they have a free tier or something Sep 11 20:03:55 what android dev can afford that :) Sep 11 20:04:20 hmm, maybe not anymore Sep 11 20:04:22 https://getsatisfaction.com/corp/pricing/ Sep 11 20:04:44 it says request pricing Sep 11 20:04:49 maybe you can request free Sep 11 20:04:50 ;) Sep 11 20:05:20 ah, yes, uservoice.com is another one Sep 11 20:05:21 ha Sep 11 20:05:53 hi, i am trying to use an external project with my android project, i have tried everything, created a jar and added to android project and marked export checkbox, but nothing works..im still get that at run-time it can't find the class Sep 11 20:06:09 is there a way to do it manually? to link inside a XML or something Sep 11 20:07:23 is there some kind of gotcha with ArrayAdapters? I'm trying to do adapter.remove(object); adapter.notifyDataSetChanged() but it's just removing the last thing in the list Sep 11 20:07:58 is moving a view via translationY more efficient than setting the position in the parents onLayout and calling requestlayout? Sep 11 20:07:58 mmathis, make sure your object has an appropriate equals/hashCode Sep 11 20:07:59 How can I get an action bar button in my UiAutomatorTestCase? Sep 11 20:09:04 I can identify by text but that only works on landscape Sep 11 20:20:37 I supposed I could get the action bar and then find the first checked child. Sep 11 20:20:43 com.onelouder.baconreader:id/actionbar - an example of what UiAutomatorViewer shows me Sep 11 20:20:46 so I assume id/actionbar Sep 11 20:20:56 I can use my package name but I can use a UiSelector with a uiautomatorviewer resource-id? Sep 11 20:24:59 ugh Sep 11 20:25:09 I was making a copy of the list in the constructor of the adapter Sep 11 20:25:16 and that was messing something up somehow Sep 11 20:25:27 then you weren't actually doing adapter.remove... Sep 11 20:26:27 well, what I was doing at first involved keeping two lists the same Sep 11 20:26:36 so I can only assume that somewhere along they way Sep 11 20:26:44 something just got crossed somewhere Sep 11 20:27:02 mAdapter.remove( result ); was definitely being called though Sep 11 20:27:08 and it was not removing what I was expecting Sep 11 20:37:35 ah, what i've been waiting for. Intel Bay Trail running Adobe Ps nicely. I hope this is in the Surface 2. Finally, a non-toy tablet Sep 11 20:38:50 works well with Android too; Intel guy says $200 tablets coming by the end of the year with this kit. Sep 11 20:40:05 fan and poor batter killed Surface Pro among other things; i tried one though and really like it. not there yet, almost though Sep 11 20:40:09 there's any way I can prevent google from automatically install an app I have just downloaded from google play store? Sep 11 20:40:28 but you just downloaded it Sep 11 20:40:32 why wouldn't you install it.... Sep 11 20:40:40 because Sep 11 20:40:43 nosense Sep 11 20:40:48 pirate Sep 11 20:40:50 here come the sense Sep 11 20:41:05 invernomuto: sei ubriaco Sep 11 20:41:12 i'm making a program that automatically perform dinamic analisys of an app Sep 11 20:41:20 to prevent install malware Sep 11 20:41:34 i want to send the apk to my app Sep 11 20:41:35 then Sep 11 20:41:38 if it's clean Sep 11 20:41:42 i can install it Sep 11 20:41:50 it's impossible? Sep 11 20:42:01 hack your rom Sep 11 20:42:05 it's not possible as-is Sep 11 20:42:09 or do not download through play Sep 11 20:42:19 invernomuto i think samsung is starting to bundle AV with their Android devices now Sep 11 20:42:29 sounds like they're bundling lookout Sep 11 20:42:38 uhm Sep 11 20:42:55 the problem is that most android user dont know how to root their devices Sep 11 20:42:58 http://www.theverge.com/2013/9/4/4692856/samsung-takes-on-android-malware-with-bundled-enterprise-antivirus Sep 11 20:43:26 doesn't google scan apps? Sep 11 20:43:27 it will be great if i could do it without any modifications from the user's phone Sep 11 20:43:35 i don't think so Sep 11 20:43:51 google does a perfunctory scan Sep 11 20:43:54 under sei italiano? Sep 11 20:44:00 yes Sep 11 20:44:10 under ciao :) Sep 11 20:44:19 ciao! Sep 11 20:44:52 pfn what's a perfunctory scan? Sep 11 20:45:23 https://www.google.com/search?q=perfunctory&ie=utf-8&oe=utf-8&aq=t&rls=org.mozilla:en-US:official&client=firefox-a Sep 11 20:45:40 pfn: why the hell do you use firefox? Sep 11 20:45:54 because I <3 pentadactyl Sep 11 20:45:54 firefox is awesome :) Sep 11 20:46:09 what browser do you use under? Sep 11 20:46:16 chrome Sep 11 20:46:16 i use firefox too Sep 11 20:46:21 and chrome's extension model is pretty weak vis a vis firefox Sep 11 20:46:26 I use chrome, too Sep 11 20:46:33 (iceweasel, to be honset, but it's the same thing) Sep 11 20:46:39 firefox extensions > * Sep 11 20:46:45 I have all synched Sep 11 20:46:46 … inspired by Vim Sep 11 20:46:59 chrome can't even come close to competing with firefox with respect to extension capability Sep 11 20:47:02 which reminds me, i need to try Ctrl-P plugin in vim Sep 11 20:47:05 vimperator > * Sep 11 20:47:10 invernomuto, pentadactyl > vimperator Sep 11 20:47:23 i.e. pentadactyl is written by the original developer of vimperator, and has since improved it Sep 11 20:47:30 pfn really? Sep 11 20:47:36 yes Sep 11 20:47:42 invernomuto: hai pm Sep 11 20:47:48 pfn cool, im going to try it then Sep 11 20:47:55 hi Sep 11 20:48:29 man Sep 11 20:48:31 okhttp is gangster. Sep 11 20:49:06 jasta what was that technique you were using to interact with the service from the command line? just pm and Service.dump() ? Sep 11 20:49:29 hats off to square and co. Sep 11 20:49:29 er, am Sep 11 20:49:45 adb shell dumpsys activity service Sep 11 20:49:49 and yes, Service#dump Sep 11 20:49:56 ok, thx Sep 11 20:50:13 there's also adb shell dumpsys activity [ top | ] and Activity/Fragment#dump Sep 11 20:50:19 invernomuto, do you use keepass, too? ;-) Sep 11 20:50:21 * pfn needs customers Sep 11 20:50:31 adb shell dumpsys activity service also works, which dumps all services Sep 11 20:50:38 pfn no! Sep 11 20:50:46 g00s: i switched up my solution though and am now using the same trick that chrome uses for its remote debugging Sep 11 20:50:57 invernomuto, that's too bad, heh Sep 11 20:50:57 jasta what is that ? Sep 11 20:50:58 well, am planning to use (havent quite finished it yet) Sep 11 20:51:00 pfn going to check it tough Sep 11 20:51:36 LocalServerSocket on the device, and expressly securing it using LocalSocket#getPeerCredentials Sep 11 20:52:01 LocalServerSocket, isn't that a domain socket, and that's available on android? Sep 11 20:52:10 jasta are you going to blog about it or post something on github? Sep 11 20:52:10 oh, I guess it's android.net Sep 11 20:52:11 neat Sep 11 20:52:18 and using adb forward tcp: localabstract: Sep 11 20:52:28 i will probably blog about it, yes. Sep 11 20:52:33 cool Sep 11 20:53:11 then you could do something like, oh i dunno, websockets or an HTTP server over this LocalServerSocket and have some scripting glue which binds it all together Sep 11 20:53:27 that part i've already built. 40 some odd lines of bash scripting does it. Sep 11 20:53:45 that's all the glue to run the app, start the server, forward, etc. Sep 11 20:53:59 it's pretty sexy Sep 11 20:54:16 LocalServerSocket is really key though because otherwise you can't secure your solution Sep 11 20:54:39 third party apps would be able to connect to it, and technically they still can with a LocalServerSocket, but you have the opportunity to reject them with getPeerCredentials after you accept the connection Sep 11 20:55:01 :) cool Sep 11 20:55:09 so you can basically take the uid and ask the system if that uid belongs to any apps. if no, you say great, and allow it Sep 11 20:55:30 dumpsys is definitely easier, but this solution is more flexible it hink Sep 11 20:55:50 dumpsys formats your output automagically for you. and isnt very scriptable (there's no way to "error out" and return a non-zero exit code) Sep 11 20:56:01 that last detail is very important for my needs which is why i went with this other solution Sep 11 20:56:18 jasta, what exactly are you building? a remote interface to send debugging requests to your service? Sep 11 20:56:20 sure Sep 11 20:56:21 ive now got it so that if the dump call throws, then the script exits non-zero Sep 11 20:56:44 pfn: basically a main(String[] args) method in an app, with a whole lot of fancy command-line glue on top to make it work well Sep 11 20:56:59 jasta, to what end? Sep 11 20:57:07 instrumentation of complex apps Sep 11 20:57:19 the same goal as dumpsys, but for my app Sep 11 20:57:21 so a poor man's jmx Sep 11 20:57:32 yeah, but also a light-weight JMX. Sep 11 20:58:57 some day, maybe, android will offer its own flavor of this but i wouldnt count on that happening soon Sep 11 20:59:03 so you connect to this local socket over adb then? Sep 11 20:59:10 hello I tried to do this code cited in the 1rst answer in this post http://stackoverflow.com/questions/7565356/get-first-monday-after-certain-date Sep 11 20:59:17 yes, adb forward knows how to work with the UNIX abstract domain sockets Sep 11 20:59:26 and there is a gap between the result i get and what i looking for Sep 11 20:59:40 i want the next monday → I get the next saturday. weird Sep 11 20:59:46 chrome actually has this really funky design for this where they expose a websocket server from the app and then forward you to the cloud which uses javascript voodoo to interact with the websocket for its data Sep 11 20:59:51 but all the UI programming and rendering is in the cloud Sep 11 21:00:19 jasta, so you're restricting access to adb access, ok Sep 11 21:00:27 its uhm, interesting. and probably the right thing to do so they can make the UI fancy without adding bloat or shipping new versions of the app. but for my purposes, i'd really rather stay the hell away from HTML and javascript :) Sep 11 21:00:35 pfn: yeah, thats what UNIX sockets give you Sep 11 21:00:57 yes, I know what the domain sockets give, I'm just trying to understand the parameters of your problem and what you're trying to solve Sep 11 21:01:02 otherwise you'd have to do some kind of key / session exchange based on a key generated by the app on install and yada yada. lame :) Sep 11 21:01:37 pfn: honestly, i've never felt that i needed this so critically in the past Sep 11 21:01:43 but working at Facebook now, ...i need it. Sep 11 21:01:51 is that where you are now Sep 11 21:01:59 I guess the facebook app sucks that much to need this ;-) Sep 11 21:02:00 an app of this size is very, very hard to grok. Sep 11 21:02:22 its actually not so much that (but yes) as its a function of its size Sep 11 21:02:26 well, it spans multiple dex files doesn't it Sep 11 21:02:37 and they do some nasty ass dex loading hackery to make it all function Sep 11 21:02:49 there are so many moving parts in an app this big (i imagine maps would be of a similar scope/scale to want this) that you really just need tools to cut through the layers and access non-UI pieces directly Sep 11 21:03:06 pfn: yes, according to our blog post. it's...well, i don't want to go there on IRC. Sep 11 21:03:09 indeed, makes sense and that sounds pretty cool Sep 11 21:03:36 my main use case is pretty relevant elsewhere though Sep 11 21:03:57 i want a way to visualize what the network layer is doing. dumping recent requests, current requests, stat counters on # of connections, total payload, etc. Sep 11 21:04:11 and having a way to both get and reset those stats for various instrumentation tests is _huge_ Sep 11 21:04:24 the facebook code must be completely enterprised-up in order to span so many dex files , in purely java, for that matter Sep 11 21:04:24 heh Sep 11 21:04:26 :) Sep 11 21:04:40 yep, a simple-jmx kind of interface sounds pretty neat Sep 11 21:04:53 its just command-line, so nothing that fancy Sep 11 21:05:09 wouldn't be so difficult to grow into something more fancy :p Sep 11 21:05:14 but its sufficiently feature rich to let modules participate in the system easily and have their own CLI options Sep 11 21:05:39 personally, i think command-line is the right tool for this job just because it is then so easy to compose and script these simple modules Sep 11 21:05:43 the UNIX way, as it were. Sep 11 21:06:11 going all web style or JMX console with it is an option on top in my mind, not a replacement. Sep 11 21:07:28 is not you mean Sep 11 21:07:29 ok Sep 11 21:08:11 http://youtu.be/chXdfjmzgPw Sep 11 21:08:14 so, a user sent this to me Sep 11 21:08:19 what in the world could possibly cause this Sep 11 21:08:51 doesn't happen on my GN, N7 on 4.2.2 and 4.3 respectively Sep 11 21:09:15 lol nice bug Sep 11 21:09:45 Ah... I guess I'll start my app again from scratch Sep 11 21:10:02 Too many errors at once Sep 11 21:10:25 screen rotation is fun, isn't it? :3 Sep 11 21:10:28 ah Sep 11 21:10:30 I just found it Sep 11 21:10:36 what was the issue? Sep 11 21:10:38 because he has force gpu rendering for 2d operations turned on Sep 11 21:11:13 why does that cause that? Sep 11 21:13:25 oh, hmm, no, it's not 2d force gpu Sep 11 21:13:27 nevermind Sep 11 21:13:35 I think I'm re-attaching the fragment when it doesn't need to be Sep 11 21:14:16 pfn: Yeah, I have something similar in my app when navigating between fragments as well. It happens very rarely, but I also suspect a fragment/lifecycle issue I haven't figured out yet... Sep 11 21:14:43 yeah, easy fix Sep 11 21:14:57 don't create a new fragment if a fragment by that tag already exists Sep 11 21:15:21 great, I'll see if that helps with my issue as well... Sep 11 21:17:26 val f = getFragmentManager.findFragmentByTag("settings fragment") Sep 11 21:17:26 if (f == null) { val tx = getFragmentManager.beginTransaction(); tx.add(android.R.id.content, new SettingsFragment, "settings fragment") Sep 11 21:17:26 ; tx.commit() Sep 11 21:17:26 } Sep 11 21:17:26 Sep 11 21:17:37 meh, stupid buffer Sep 11 21:17:44 <*flooddetach> Channel [#android-dev] was flooded, you've been detached Sep 11 21:17:49 lolol Sep 11 21:17:56 nice znc plugin Sep 11 21:18:02 yeah Sep 11 21:21:42 pfn: is https://play.google.com/store/apps/details?id=com.hanhuy.android.irc.lite yours? Sep 11 21:21:47 under, yes Sep 11 21:22:03 ok i'll give a try Sep 11 21:23:46 pfn: from screenshots looks very nice Sep 11 21:23:52 is this compatible with znc? Sep 11 21:23:55 under, yes Sep 11 21:23:58 I use it exclusively with znc Sep 11 21:24:09 ook Sep 11 21:33:02 I'm using light theme dark actionbar. where do I find the responsible setting in the sources? Sep 11 21:33:49 in styles.xml Sep 11 21:33:50 where else Sep 11 21:34:37 the manifest uses @style/AppTheme, which is blank with AppBaseTheme as parent, which in turn has android:Theme.Light as parent. I'm trying to understand where the dark actionbar comes from Sep 11 21:35:22 pfn, little bug in your irc client: When I change from day/night mode, or tilt the screen, the application briefly goes to the "servers" screen Sep 11 21:36:25 <|0xD34D|> Chais_: you may have multiple styles.xml for different API levels Sep 11 21:36:52 <|0xD34D|> Chais_: something like values-v14 and values-v11 Sep 11 21:36:58 I only have one. also I only support lvl 14+ for this tutorial app Sep 11 21:37:18 oh wait Sep 11 21:37:37 that was the hint ^^ thx Sep 11 21:38:35 sonOfRa, that's normal Sep 11 21:38:49 I guess I could fix it Sep 11 21:39:35 it's kind of irritating when first using it Sep 11 21:40:52 you rotate that often? Sep 11 21:40:55 I usually lock it Sep 11 21:41:20 I don't know if I will in the future, right now I was just doing my monkeytesting I do when first trying out an app Sep 11 21:41:28 I tend to try to break everything new I get into my hands Sep 11 21:43:41 Hey you guys, if you know anything about WebViews, maybe you can help me: http://stackoverflow.com/questions/18732285/android-webview-loadurl-in-for-loop-firing-before-first-url-finished-loading Sep 11 21:45:04 sonOfRa, at least it's not really broken :p Sep 11 21:45:20 Unicode is kind of broken though :/ Sep 11 21:45:24 öäü Sep 11 21:45:29 Slap some people here as well. Anyone got LG G2 or other Snapdragon 800 device yet? :D Sep 11 21:45:31 utf-8 works for me, I'm able to use wnn emoji ime Sep 11 21:45:35 get replaced with the unicode replacement chars Sep 11 21:45:44 I can SEND them fine, but if I receive them, they get replaced Sep 11 21:45:54 indeed Sep 11 21:45:58 what android version? Sep 11 21:46:20 4.3 Sep 11 21:46:54 k, <4.0 uses a different font Sep 11 21:47:23 Q48VW: you can't threadsleep the main thread and expect things to not freeze up, make a stack or similar, pop off one entry every time onPageFinished is called Sep 11 21:48:17 Thorbear, thanks for responding. It isn't the main thread of an application, but a custom object class I made, does the rule still apply? Sep 11 21:48:31 pfn: have the same issue of youtube video Sep 11 21:48:59 under, I know, I haven't published my fix yet Sep 11 21:49:04 under, only in settings Sep 11 21:49:09 Q48VW: unless you're specifically run that class in the background it runs on the main thread Sep 11 21:49:16 gotcha Sep 11 21:49:18 Anyone familiar with multi project setup for gradle? https://groups.google.com/forum/#!topic/adt-dev/UVIXYLwSQbw . Can't seem to figure out how to reference a library Sep 11 21:49:54 Q48VW: Wait, you're using a WebView to download HTML? Sep 11 21:50:02 RyanM, yes Sep 11 21:50:05 Don't do that Sep 11 21:50:26 use HttpClient or HttpURLConnection Sep 11 21:50:26 why not Sep 11 21:51:53 WebViews are designed for displaying web pages to users Sep 11 21:51:54 if all you care about is the code, you'll have a much easier time with one of those classes Sep 11 21:51:55 I can't do that either Sep 11 21:52:43 The situation is, is when a user punches in their password in the password field, it translates it to something or other, making those classes impossible. Sep 11 21:53:32 the site I try to log in to is http://powerschool.lovejoyisd.net/public/ Sep 11 21:53:38 try to put a password in that field Sep 11 22:05:52 Log in, watch with Chrome Developer Tools (Network Tab), copy cURL, profit? Sep 11 22:06:26 Unless of course it does something in the browser in JS, I didn't actually click it ;) Sep 11 22:11:05 anyone know how to publish an apk that overrides (as in, disables) user's Play Auto-Update even if permissions haven't changed ? Sep 11 22:11:38 I don't. But out of curiosity: why? Sep 11 22:12:03 the app is a beta; i have a warning that data could be lost on update - and i plan on migrating away from sqlite and don't feel like writing the migration code which will be a huge hassle. Sep 11 22:12:29 at the same time, i dont want the auto-update to just nuke data Sep 11 22:12:46 the user can archive it off with existing functionality and then update when ready Sep 11 22:16:33 btw is there somewhere I can download the android icon pack as svg file? Sep 11 22:17:41 inkscape can convert them, but packs alls icons to 0,0 so it's just one blob Sep 11 22:17:48 on d.android.com / design there are some templates Sep 11 22:17:54 but PSD Sep 11 22:18:04 * g00s is pissed at inkscape lately Sep 11 22:19:17 man inkscape is a love/hate relationship if ive ever seen one Sep 11 22:19:22 I understand, I get that too sometimes. eg when I just spent 2 hrs vectorising some image and then inkscape decides to crash Sep 11 22:19:57 I have a love/hate relationship with Eclipse Sep 11 22:20:06 but linux and adobe illustrater probably combine even worse Sep 11 22:20:30 Chais_ birbeck what OS are you running inkscape on ? Sep 11 22:20:34 RyanM: that's why I'm taking a look at android studio atm Sep 11 22:20:42 g00s: Archlinux Sep 11 22:20:44 Linux Sep 11 22:20:50 ah, you guys are lucky Sep 11 22:21:02 i went to #inkscape to ask them whats going on with the mac port Sep 11 22:21:06 http://www.computerjokes.net/128.asp joke about programs crashing Sep 11 22:21:11 christ, it has to be the worst run project evAR Sep 11 22:21:26 they haven't released anything since 0.48.2 Sep 11 22:21:29 probably inkscape isn't artsy enough for OSX ;) Sep 11 22:21:33 no eta or anything Sep 11 22:21:50 i can't even get them to say if they are supporting mac *at all* Sep 11 22:22:03 i mean, if they say, 'fuck it, we're not doing mac anymore' that would be fine Sep 11 22:22:22 SuD, that's so bad that it stopped being good and got bad again Sep 11 22:22:57 g00s: do you have illu on your machine? Sep 11 22:23:04 i can't afford it Sep 11 22:23:09 Can anyone help me with writing a ble app Sep 11 22:23:13 i was about to set up a linux VM to run inkscape Sep 11 22:23:19 fuuuu Sep 11 22:23:21 that doesn't always stop people ;) Sep 11 22:23:21 Having trouble writing to a characteristic Sep 11 22:23:36 Chais_ heh :) Sep 11 22:24:01 you have a mac and you can't afford some software? Sep 11 22:24:02 -.- Sep 11 22:24:23 hi everybody! I have a RelativeLayout with a GlSurfaceView and another SurfaceView on top. To assure the second SurfaceView is always visible, I set it to surfaceView.setZOrderOnTop(true); in the Activity's onCreate. But now I want to show an ImageView on top of the SurfaceView as well, but it's hidden due to the command. But I can't call the same method for the ImageView as well, because it's not defined in the ImageView class... anybody an idea on how Sep 11 22:24:42 Zharf: you make it sound like having a mac is a good thing Sep 11 22:24:46 Zharf: of course not. It all went into hardware ;) Sep 11 22:24:54 so prestigious Sep 11 22:24:59 birbeck, no, I'm making it sound like he should've thought about what he's buying :) Sep 11 22:25:17 im sure he didnt decide to buy it Sep 11 22:25:25 or he could also decide to install Linux on it Sep 11 22:25:39 i got a mac in 2009 to do iOS dev Sep 11 22:25:48 apple users seriously need financial advice Sep 11 22:25:48 ;) Sep 11 22:25:49 after 10 years of linux, it was the best decision i made for a desktop Sep 11 22:25:55 dont regret it at all Sep 11 22:26:11 I'm actually running linux on mac but only because I have proper os x on another, a bit faster mac Sep 11 22:26:14 :-p Sep 11 22:26:16 birbeck: which is a terrible hassle. at least if you're determined to have linux run directly on EFI, not through bootcamp Sep 11 22:26:28 i still use linux, but not for desktop ;) Sep 11 22:26:34 os x kind of struggles on this machine, so I handed it over to linux. Sep 11 22:26:52 should probably have kept it running snow leopard but baaaah Sep 11 22:26:58 I have linux on my dev machines and servers and windows on others Sep 11 22:27:33 alankila yeah i have 10.7; i didn't do 10.8 because i heard it killed the battery Sep 11 22:27:38 but i hear good things about 10.9 Sep 11 22:27:44 * g00s will wait until 10.9.2 Sep 11 22:27:45 :D Sep 11 22:27:57 haha, thats great, speaking of linux … same old shit Sep 11 22:27:59 10.8 was a bugfix release of 10.7, believe me. Sep 11 22:27:59 I've not noticed an issue with 10.8 with respect to battery at least. Sep 11 22:28:04 Intel tells canonical to fuck off with mir Sep 11 22:28:10 put support into wayland Sep 11 22:28:29 I agree with intel Sep 11 22:28:33 canonical ruined linux Sep 11 22:28:34 I would never go back to 10.7. But I am very much looking forward to OS X Mavericks. Sep 11 22:28:38 anyway anyone who could convert the .ai icons to .svg in a way that inkscape doesn't stack them? Sep 11 22:29:28 Zharf: no, canonical ruined ubuntu. and perhaps the approach everyones taking on DEs. they tend to get dumbed down lately Sep 11 22:29:42 ubuntu is canonical's to ruin Sep 11 22:30:10 Hey everyone. I am trying to write a plain Bluetooth LE app using v4.3, I have been using the function BluetoothGattCharacteristic.writeCharacteristic attempting to read with the corresponding function. I can pair with the target alright, but when I try to send single bytes I am not actually getting any data. Does anyone have any experience working with the BLE Gatt stack. I would really appreciate it Sep 11 22:30:10 I've no idea what they are doing nowadays. It sounds a bit like feeble attempts to make it profitable somehow using various harebrained schemes Sep 11 22:30:12 Chais_, silly people keep bringing the crap from ubuntu to other places Sep 11 22:30:32 alankila, that's exactly what it seems to me as well Sep 11 22:30:34 I'm still using ubuntu but the amount of crazy seems to increase over time Sep 11 22:30:55 I must begin considering a migration path to something else, maybe centos Sep 11 22:31:16 arch maybe Sep 11 22:31:34 LFS! Sep 11 22:31:36 win8? Sep 11 22:31:50 yeah, i must be the only one that liked win8 :D Sep 11 22:32:04 I used ubuntu for about half a year when learning for the LPIC-1. then I went nuts over where ubuntu behaves weirdly, tried arch and fell in love Sep 11 22:32:09 * g00s needs adobe Lr, so its win or mac Sep 11 22:33:29 g00s, meh, I've used win8 for a year, it's had a few issues compared to 7 but mostly it's fine... it even has some cool stuff Sep 11 22:33:52 I haven't tried 8.1 yet Sep 11 22:34:17 it's the cool stuff that is missing from win8, like start menu... Sep 11 22:34:29 haven't seen 8.1 yet either Sep 11 22:34:38 SuD, that's where I disagree, the start menu isn't missing Sep 11 22:34:39 Thorbear, but what's the difference of using a stack rather than just pulling the links from an arraylist Sep 11 22:34:44 i bought pro cd key from china Sep 11 22:34:45 i got kinda pissed that they would not let you install without an outlook account Sep 11 22:34:46 7$ Sep 11 22:34:47 it's right there, and it's exactly like it used to be -.- Sep 11 22:34:58 even on mac, iTunes account is different than admin account Sep 11 22:35:13 g00s, huh? I haven't got an MS account... Sep 11 22:35:21 Zharf O.o Sep 11 22:35:48 -.- Sep 11 22:36:04 was it something I said? Sep 11 22:36:13 Q48VW: popping items off a stack will give you the item and remove it from the stack. You *can* do the same with an arraylist, I just figured a stack would make it more obvious Sep 11 22:37:01 Thorbear, the problem is that in my for loop, it starts to load the next url in the list before the first one completes, so the only link ever fully loaded is the last one Sep 11 22:37:43 g00s, I login with just a regular username and password, like in previous windows versions :) Sep 11 22:37:46 that's why you do the popping when the page is fully loaded, and don't use a loop at all Sep 11 22:38:07 so just have a straight list of .loadurl calls? Sep 11 22:38:11 Zharf you were able to set up all your accounts without any email ? Sep 11 22:38:14 yep Sep 11 22:38:17 dafuq Sep 11 22:39:32 Q48VW: what you have a list of doesn't really matter to me =) the point is: start loading first element. onloaded: is there more elements? load next. Sep 11 22:40:58 Hey everyone. I am trying to write a plain Bluetooth LE app using v4.3, I have been using the function BluetoothGattCharacteristic.writeCharacteristic attempting to read with the corresponding function. I can pair with the target alright, but when I try to send single bytes I am not actually getting any data. Does anyone have any experience working with the BLE Gatt stack. I would really appreciate it Sep 11 22:41:24 * Zharf zzz Sep 11 22:45:00 ble_nub somebody was trying it with a Ti sensortag a few weeks ago and it worked, but was far slower throughput than iOS Sep 11 22:45:51 something seemed iffy Sep 11 22:47:30 g00s okay, can you point me to a converstation or do you recall details? Sep 11 22:48:39 ble_nub hm, can't remember the nick. i'll probably start doing ble soon too Sep 11 22:49:26 sensortag is great to have as a 'sanity' device Sep 11 22:56:01 thanks, I'll look into sensortag Sep 11 23:03:51 dragorn, maybe Sep 11 23:09:09 probably start doing it soon? Sep 11 23:09:13 that would entail actually buying a device :p Sep 11 23:17:52 Thorbear, I think I understand Sep 11 23:20:00 is there any possible way to utilize group by in a cursor loader? Sep 11 23:20:46 Hi, having a problem on the Nexus 4 using a TextWatcher. onTextChanged method does not fire in the emmulator or on the Nexus 4 but it does on other devices. Sep 11 23:21:02 it's a textchangedListener for an EditText BTW Sep 11 23:49:46 Weird thing setOnKeyListener fires, but addTextChangedListener doesn't on the Nexus 4. Sep 11 23:49:55 err the callbacks fire that is. Sep 11 23:50:45 Although using setOnKeyListener disrupts the edit text for some reason. No text gets inserted even though I return true. Sep 11 23:56:34 Anyone know any OpenGL ES 3 applications on the App store besides the Unity benchmark? Sep 12 00:00:18 I have an app where I'm allowing the user to select an existing video. I'm then uploading the video to the server. I need to ensure that the video is in a particular orientation. However, I'm finding it impossible for pre-API 17 to determine video orientation. I've tried using a MediaMetadataRetreiver.METADATA_KEY_VIDEO_HEIGHT and WIDTH and comparing the values, but regardless of the orientation of the video (landscape or portrait) the height and wid Sep 12 00:00:19 th remains the same! Any suggestions for determining orientation of video pre-API 17? Sep 12 00:04:44 excellent Sep 12 00:19:39 Thorbear, it's working man. I've been stuck on this for days. You're a saviour Sep 12 00:25:49 Pyuruku, why can't you group by in your query..., hmm is that even supported Sep 12 00:25:57 no it isnt Sep 12 00:26:05 which makes no sense Sep 12 00:26:16 they could at least provide a raw query function Sep 12 00:27:23 I created a separate class and is trying to get user preferences: SharedPreferences settings = getSharedPreferences(PREFS_NAME, 0); I am getting an error: cannot resolve method getSharedPreferences(). Is this method only accessible in an activity class? Sep 12 00:28:58 zquad, only in a context Sep 12 00:29:14 of course, methods don't appear out of thin air Sep 12 00:29:30 I don't understand how to do masking, here is my mask build and paint function at http://pastebin.com/gTF30Dfw Sep 12 00:31:11 Oh. thanks pfn Sep 12 00:31:49 anything stand out as being really wrong? Sep 12 00:32:15 basically it paints tiled texture into bitmap, build mask then applies it to first picture Sep 12 00:32:24 then paint result to final bitmap Sep 12 00:37:33 ok thats damn weird Sep 12 00:37:47 fix was changing drawmask to argb_8888 Sep 12 00:37:57 Hi all, can someone tell me how to get a "callback" happening whenever my app receives focus? I would've thought that my googling would've picked this answer u Sep 12 00:37:59 up Sep 12 00:38:02 how do I just use alpha_8 for mask? for lot less space Sep 12 00:38:24 some sort of oncreate on the application class? Sep 12 00:38:27 *subclass Sep 12 00:40:28 out for now Sep 12 00:56:57 back to software engineering Sep 12 01:00:08 pfn: do you know about the rotation bug while adding server? it also drive the app crazy Sep 12 01:00:56 while adding? I don't know, no Sep 12 01:01:04 I rarely add... Sep 12 01:01:12 oh activity gets retwerked Sep 12 01:01:19 I'll look at it sometime Sep 12 01:01:43 pfn, tomorrow I explain with video, now its very late Sep 12 01:03:03 if i dl an app i just keep rotating it until it crashes Sep 12 01:03:10 almost all do eventually ;D Sep 12 01:03:45 bb Sep 12 01:06:03 of course they crash. they got dizzy ;)] Sep 12 01:08:30 am I doing something wrong? Sep 12 01:08:32 drawnmask.drawColor(0); this does not work Sep 12 01:10:37 finally found it Sep 12 01:11:09 its real stupid bug, I painted it with color 0 Sep 12 01:11:15 thats alpha 0 too. no effct at all Sep 12 01:11:22 so I changed it to 0xff << 24 Sep 12 01:29:14 when using ViewPager, it's expected that when changing pages Android will call into onCreateOptionsMenu() for the newly visible page, yes? Sep 12 01:29:46 newly visible Fragment that is Sep 12 01:29:56 yes Sep 12 01:30:05 if you set has options menu Sep 12 01:30:45 i thought so, damnit if that's not what is happening... using v4 support + ABS + ViewPagerIndicator Sep 12 01:31:07 when changing pages, the Fragment that hosts the ViewPager gets the call to onCreateOptionsMenu Sep 12 01:31:30 the child fragments never get onCreateOptionsMenu called, yes i am setting has options menu Sep 12 01:32:09 after some tracing i feel like im either crazy or this is a bug in support v4 Sep 12 01:32:23 you're crazy Sep 12 01:33:22 plenty of bugs in support lib ;) Sep 12 01:33:22 any thoughts beyond that? Sep 12 01:33:28 ABS doesn't support child fragments having menus Sep 12 01:33:32 maybe you're not crazy Sep 12 01:33:47 oh JakeWharton god bless you, was hoping you'd show up Sep 12 01:34:00 so what im seeing, this is expected behavior then? Sep 12 01:34:42 yes. you can probably modify Watson.java to add child fragment lookup or something Sep 12 01:35:09 interesting Sep 12 01:35:37 thanks Jake, ive been stuck on that for a few days, couldn't find shit on Google Sep 12 01:35:48 so does that deviate fromsupport-v4 behavior, or does it exhibit the same Sep 12 01:36:19 pfn: you are still correct, im also crazy Sep 12 01:36:21 no clue. the behavior predates child fragments by over a year Sep 12 01:37:02 i haven't shipped this app yet, thoughts on moving to ActionBarCompat to get around this issue? Sep 12 01:37:16 minSdkVersion="14" would work around the issue Sep 12 01:37:21 i recommend that Sep 12 01:37:37 drop ABS completely Sep 12 01:37:39 can't swing it Sep 12 01:38:34 minSdk = 16! :P Sep 12 01:39:06 sorry Jake, to be clear -- ActionBarComapt would be included in your recommendation? Sep 12 01:39:28 assuming i can't do SDK 14 Sep 12 01:39:32 sabton who knows, but i wouldn'tbe surprised if google keeps moving stuff into the support lib / appcompat. may as well use it Sep 12 01:39:43 can't say. never used it, didn't bother investigating it. Sep 12 01:39:57 i'll do the leg work, just didn't want to lose anything over ABS Sep 12 01:40:25 presumably the whole point is they are both supposed to work the same though yes? Sep 12 01:40:29 more or less Sep 12 01:40:38 they are exactly the same Sep 12 01:40:52 well what about this child fragments + options menu thing? Sep 12 01:41:01 i meant in concept Sep 12 01:41:06 k :) Sep 12 01:41:13 sry just making sure i understand Sep 12 01:41:35 sabton are you sure you need nested fragments ;) Sep 12 01:41:41 i will try out ActionBarCompat then and see what it buys me, thx for showing up Sep 12 01:41:47 g00s im just using ViewPager is all Sep 12 01:42:06 sometimes i think "oh i can use nested fragment" and then think, oh shit, there is some weird bug . corner case - screw it Sep 12 01:42:24 i'm always here, just constantly trying to scrub my eyes with bleach from some of the things that go by every day Sep 12 01:42:38 +1 to eye-scrubbing Sep 12 01:42:49 except I'm usually scrubbing my brain with it Sep 12 01:42:51 bleach though - ouch Sep 12 01:43:08 how can I make keys to SharedPreferences global so that users do not have to pass strings? I thought about an enum or I can use public static final String on a class... what is the best approach? Sep 12 01:43:19 g00s: It hurts slightly less than reading that the solution to all your screen-rotation issues is android:configChanges Sep 12 01:44:08 zquad: public static final Strings are a pretty common solution Sep 12 01:44:10 zquad: use a Preference interface Sep 12 01:44:21 and then just String FOO = "foo"; Sep 12 01:44:33 prefs.getString(Preference.FOO); Sep 12 01:45:02 JakeWharton can you elaborate on the preference interface? Sep 12 01:45:43 https://gist.github.com/JakeWharton/75a0991fda681d5d871f Sep 12 01:46:03 ooooooh... I like that Sep 12 01:46:20 thanks RyanM, JakeWharton Sep 12 01:47:17 zquad, I usually do this https://github.com/pfn/advanced-keyguard-manager/blob/master/src/com/hanhuy/android/bluetooth/keyguard/Settings.java Sep 12 01:48:30 that's what I do as well, kind of Sep 12 01:48:35 consistent defaults, consistent types, no confusing the type of a property ever, etc Sep 12 01:48:39 then you can just @Inject prefs Sep 12 01:48:54 is there quick way to invert bitmap colors? Sep 12 01:49:14 yea pfn that is a really good idea Sep 12 01:49:31 open it in photoshop, option+command+i Sep 12 01:49:49 lol guess thats true but no im talking about by code Sep 12 01:50:02 yeah i know, just trolling my way out Sep 12 01:50:04 I do something similar in scala as well Sep 12 01:50:07 lol Sep 12 01:50:34 XXCoder, depends on your definition of invert Sep 12 01:50:50 you can just do 2's complement of all pixels Sep 12 01:51:21 hmm ok. well im trying to figure masking too lol Sep 12 01:51:30 there is no real good example out there Sep 12 01:51:36 you can apply a colormatrix color filter to invert a drawable. to do it to a bitmap you'd have to create a new bitmap and draw the old bitmap to it with the color matrix applied Sep 12 01:52:29 oh finally http://stackoverflow.com/questions/12614542/maskingcrop-image-in-frame Sep 12 01:53:39 kevin thanks Sep 12 01:54:09 pfn if you had a code change equivalent in magnitude to iosched 2010 -> iosched 2013 (adding action bar, tablet layouts, etc) would you refactor or start over ? Sep 12 01:54:59 g00s, I dunno what those apps look like, but if nothing can be reused, then start over Sep 12 01:55:40 especially if your design and architecture are bad Sep 12 01:55:55 it was fine for froyo :P Sep 12 01:55:58 joel spolsky says don't rewrite: http://www.joelonsoftware.com/articles/fog0000000069.html . but in that case he's also talking about a sufficiently complex/mature product Sep 12 01:56:07 daaaa somebody caught me Sep 12 01:56:14 kevinb i remember that one Sep 12 01:56:30 well, it definitely applies for stuff you never wrote to begin with :P Sep 12 01:56:43 'i'l rewrite this big legacy pos' Sep 12 01:57:38 I don't think I've ever agreed with anything joels has written Sep 12 01:57:43 haha Sep 12 01:57:49 kevinb hey, they are starting from scratch again right, with rust ? :P Sep 12 01:58:11 at whatever rendered rust is used in Sep 12 01:59:12 not sure. i think there are definitely times where rewriting makes sense and works. but it's also really easy to see a rewrite as less work than it truly is. Sep 12 01:59:35 it really depends Sep 12 01:59:57 most of the time, a rewrite is a migration proecess Sep 12 02:00:06 components from old to new Sep 12 02:00:53 i think the biggest hurdle for me still is conceptualizing how to split up my UI for Tablet - not from the /visual/ perspective, but from the activity / fragment partitioning perspective Sep 12 02:01:12 but if the app is small, then there's really no migration possible Sep 12 02:01:48 consider how you want to make it work on tablet vs phone Sep 12 02:02:12 yeah a rewrite of a small enough app, that isn't dealing with any legacy file formats or plugins or server communication is a lot more feasible than something with a lot of required legacy support (that probably isn't properly documented) Sep 12 02:02:15 then storyboard how it comes together Sep 12 02:02:21 i'm looking at iosched now; trying to figure out their ui/phone/tablet hierarchy Sep 12 02:02:37 yeah, from external architecture i know how it should look Sep 12 02:02:49 just still not sure how to organize the code Sep 12 02:04:24 which is one of my complaints about the tooling; if i go to : iosched2013/android/src/main/java/com/google/android/apps/iosched/ui … even looking at xmls, its pretty hard to comprehend the structure of things Sep 12 02:04:59 g00s: wasn't me, tho I dabble in other bt stuff Sep 12 02:05:09 g00s: mostly on the capture side tho Sep 12 02:05:12 oh hey dragorn Sep 12 02:05:54 * g00s can't wait to start playing with BLE Sep 12 02:07:43 g00s: It's pretty easy to sniff/spoof too :P Sep 12 02:07:51 g00s: tl;dr - don't use btle for anything that matters Sep 12 02:08:04 hello so im trying to figure out how to set a value in the parent fragment from a dialog. problem is my check to see if the value is true is called after the dialog.show even though the dialog is still visible. not sure why. http://pastebin.com/JhdMsZnq Sep 12 02:08:17 g00s: if you want some internals on it search for 'mike ryan shmoocon' Sep 12 02:08:34 cooldman224, dialog.show does not block Sep 12 02:08:36 dragorn oh noes, they messed up again heh Sep 12 02:08:38 g00s: infact, http://www.youtube.com/watch?v=AOqgH7dJWrQ Sep 12 02:08:50 pfn: ugh so what do you suggest? Sep 12 02:08:56 g00s: I show up briefly at some point in there. I think. Depends when they start recording Sep 12 02:09:01 cooldman224, callbacks Sep 12 02:09:26 pfn: i have a callback for the result. so what you mean? Sep 12 02:09:39 g00s: yeah; btle hops over half the bandwidth that bluetooth classic does, and they pooched the key exchange Sep 12 02:09:53 cooldman224, use the callback correctly Sep 12 02:10:12 pfn: whats the correct why Sep 12 02:10:15 way Sep 12 02:11:16 dragorn thanks for the heads up; i know BLE is being used for some ballistic computers; would be funny if somebody sniped the sniper :P Sep 12 02:11:17 cooldman224, moving your if into the callback Sep 12 02:11:52 dragorn from a security perspective, it BT 2/3 better or worse ? Sep 12 02:16:21 pfn: thanks got it. i thought it was more elegant before :/ Sep 12 02:17:12 it's not elegant either way :p Sep 12 02:18:55 g00s: so BT is only "good" because it's really hard to sniff, technically Sep 12 02:19:08 g00s: so there's not a lot of hackery types with the budget for the commercial sniffers Sep 12 02:19:16 g00s: but ppl are trying to fix that Sep 12 02:19:39 g00s: the main problem is even if the pin mechanism were secure (which I dont' think it is, particularly), all pins are 0000 or 1234 ;P Sep 12 02:19:57 g00s: I wouldn't really trust any of it for security ;P Sep 12 02:20:05 interesting Sep 12 02:20:27 It's hard to follow a connection because it hops at like 200uS over 79mhz Sep 12 02:20:35 but you CAN recover the hopping pattern Sep 12 02:20:53 i'm not super expert in the BT internals Sep 12 02:21:19 g00s: here; if you really care, http://www.youtube.com/watch?v=K-7r7TB4ukI Sep 12 02:21:25 dragorn thanks Sep 12 02:21:56 g00s: That's a relatively recent look at the state of BT intercept at affordable levels Sep 12 02:22:53 There's also gr-bluetooth which the same ppl (ossmann and dominic spill) are working on, which uses SDRs to do it; The hackrf (mike ossmann again) can do 20mhz at a time, and he's talking about doing a radio frontend for a board on another project which could do 80 or 160mhz at a time. If you can get 80mhz you just sit on the entire BT spectrum and tune w/ software, and then following connections gets, in theory, really easy Sep 12 02:24:23 if I use alpha_8 bitmap, what is proper way to draw to it? Sep 12 02:25:28 g00s: at some point i'll have the time to port the ubertooth code over to android :P Sep 12 02:25:40 ooh Sep 12 02:27:58 fun 'fact' if actually true : A single $800 Samsung Galaxy S4 has the same power as 375 IBM 3340 computers that cost a total of $33 million in 1973 Sep 12 02:27:58 pfn: very true! but still needs to happen Sep 12 02:28:34 g00s: I actually find that surprising, I would have thought the s4 would be more powerful. Sep 12 02:29:01 or more powerful than more of them Sep 12 02:29:19 or as powerful as more of then, rather. I'll spit it out eventually :) Sep 12 02:29:23 :D Sep 12 02:29:34 them* Sep 12 02:29:34 * JesusFreke sighs Sep 12 02:32:52 JesusFreke are you using qicr from your android tablet while monocycling ? Sep 12 02:32:57 :P Sep 12 02:33:38 g00s: not true Sep 12 02:33:45 g00s: considering a 3340 is a hard drive :P Sep 12 02:33:59 shit I think we had a 3340 at one of my old jobs ;P Sep 12 02:34:16 g00s: supplanted by the 3380, which we definitely had, and which modern systems still emulate :P **** ENDING LOGGING AT Thu Sep 12 02:59:59 2013