**** BEGIN LOGGING AT Fri Jan 25 02:59:59 2013 Jan 25 03:03:33 actually I think gson will work pleasantly Jan 25 03:03:40 now that I re-evaluate my requirements Jan 25 03:15:30 why would one choose binders over sockets for IPC? Jan 25 03:15:33 or sockets over binders? Jan 25 03:16:35 i thought sockets are pretty high overhead for ipc Jan 25 03:17:30 yeah I am not sure Jan 25 03:17:41 I am trying to argue to my team to pick one over the other Jan 25 03:17:44 just not sure which one is best Jan 25 03:17:53 we have 3 native daemons that all need to communicate with each other Jan 25 03:19:07 oh, native … no idea if binders would be of any use in that scenario Jan 25 03:19:39 get some data ;) Jan 25 04:27:04 IGUAY Jan 25 04:30:19 I think I've spent a total of 6 hours writing tests today Jan 25 04:30:36 how to i stop my shared preference strings from returning "java.long.nullpointerexception" because obvioulsy if no preferences have been set yet then it is going to return null. http://pastie.org/5853207 Jan 25 04:31:34 JakeWharton I've spent less than 1hr writing test in my entire 2.5yr at current employer Jan 25 04:31:37 lang not long Jan 25 04:31:40 and I am actually pissed abotu that Jan 25 04:31:46 haha that scares me! Jan 25 04:31:59 they dont believe in unit testing Jan 25 04:32:12 no matter how much I beat a dead horse about it Jan 25 04:32:30 I dont know how many times I've said I dont mean to say I told you so but... Jan 25 04:34:25 str8uplinux: firstly, your strings don't return an exception... until you try to access them without checking for null Jan 25 04:35:00 lets just say I've got a lead on new work at a new company Jan 25 04:35:09 secondly, if the preference doesn't exist, it will return the default value you supply - "" in the paste - not null Jan 25 04:35:24 undoubtedly less interesting thn some of the stuff I get to work on at the day job Jan 25 04:39:09 can anyone recommend a quality, free OCR lib for android that will perform "fast enough" for most users on your typical phone Jan 25 04:39:29 I use Leadtools OCR for Windows, but doesn't seem to have a version for android Jan 25 04:39:53 ideally it would support many different fonts Jan 25 04:40:14 wtf Jan 25 04:40:28 you think android hardware is gonna have the hp to do that sort of analysis? Jan 25 04:40:39 actually, even more ideally, it would work on pictures captured from the phone's camera Jan 25 04:41:41 well, if the answer is "there is no such thing", then that would be useful to me - it certainly isn't obvious that your typical phone CPU is incapable of OCR Jan 25 04:41:53 at least not to me.. hardware is pretty impressive these days, even for phones Jan 25 04:42:37 ARM has high clock frequencies (ghz range now) but performance is pretty arse.. Jan 25 04:43:17 I actually cant wait to see whats possible for ubuntuphone Jan 25 04:43:25 if it 1) uses wayland/weston Jan 25 04:43:33 and 2) development isnt limited to a specific platform Jan 25 04:43:43 yeah, they are definitely one of the first to skate where the puck is going - but, sigh Jan 25 04:43:44 I think its gonna have a mildly decent chance Jan 25 04:44:00 I dont care about a phone is a computer when you plug it into x or whatever Jan 25 04:44:08 I just think that as mucha s google tries to avoid it Jan 25 04:44:14 dalvik and java are slowing shit down Jan 25 04:44:22 I've seen os internals ad naseum Jan 25 04:44:31 and the architecture to just make android work at all Jan 25 04:44:32 si Jan 25 04:44:36 ron_frown: have you followed the gnome mailing lists over the years ? Jan 25 04:44:37 overhead to say the least Jan 25 04:44:47 years, as in, last decade maybe ? Jan 25 04:44:48 ron_frown: with ubuntu phone be x11? Jan 25 04:44:58 * cantido thought it would be some DirectFB thing Jan 25 04:45:13 well its just running straight ubuntu apps Jan 25 04:45:26 wayland should still be better than directfb afaik Jan 25 04:45:49 the examples I've seen and the sdk stuff for ubuntu mobile is just qt Jan 25 04:45:55 hm; Things 2 on sale $24 Jan 25 04:46:07 ? Jan 25 04:46:30 a to-do app for mac; very nice Jan 25 04:46:33 ;) Jan 25 04:46:40 I wonder what they'll do to stop the classic linux issue of all the apps looking totally different Jan 25 04:46:42 I just use wunderlist Jan 25 04:47:03 cantido qt already has a very mature cross platform gui toolkit Jan 25 04:47:04 AND Jan 25 04:47:09 there is no window chrome Jan 25 04:47:10 $24 for a to-do app? Jan 25 04:47:18 desktop app Jan 25 04:47:20 ron_frown: that doesn't stop people from using all sorts of random toolkits Jan 25 04:47:35 well thats what ubuntu is pushing Jan 25 04:47:40 QML Jan 25 04:47:41 qtquick and qt Jan 25 04:47:51 no window chrome because no windowing, surely? Jan 25 04:47:58 leeds, you think? =) Jan 25 04:47:58 haha Jan 25 04:48:11 just saying thats one thing that frequently looks different in linux Jan 25 04:48:15 not gonna be a prob here Jan 25 04:48:36 ron_frown: windows chrome is a very small problem.. you can use whatever window manager you like Jan 25 04:49:03 I am very very familliar with all that stuff on linux Jan 25 04:49:10 you can have any gui tool kit anywhere Jan 25 04:49:20 and even with the same toolkit you can have vastly diff appearances Jan 25 04:49:23 ron_frown: not if an app is written for one toolkit Jan 25 04:49:39 qml is much like html Jan 25 04:49:46 a window manager doesnt need to be directly related to the application.. it just needs to be a window manager Jan 25 04:49:48 i wonder if the Ubuntu apps will be like BB10 apps, in that they all pretty must just keep running in the background like a regular desktop Jan 25 04:49:51 you could have images for buttons you could have a regular frame Jan 25 04:50:00 you could have a plain ass area with no borders Jan 25 04:50:12 g00s thats the impression I got Jan 25 04:50:31 the qt stuff they've released thus far, I didnt see anything specific to suspend as app is going sleep Jan 25 04:50:36 BB10 demo I saw looked really slick Jan 25 04:50:58 I wasnt that impressed Jan 25 04:51:07 better than bb whatever Jan 25 04:51:11 ron_frown: thats because you are the destroyer Jan 25 04:51:11 that came before it Jan 25 04:51:18 I *AM* the destroyer Jan 25 04:51:22 :) Jan 25 04:51:27 I dont even think ubuntu looked that impressive Jan 25 04:51:32 just know what it *COULD* be Jan 25 04:51:37 and only time will tell Jan 25 04:51:46 if its a "you can only dev on a linux box" Jan 25 04:51:54 I think thats going to severyly limit their customer base Jan 25 04:52:18 right now their docs say ubuntu 12.04 Jan 25 04:52:19 thats it Jan 25 04:52:24 no older ubuntu boxes Jan 25 04:52:29 no other dist boxes Jan 25 04:53:11 hmm Jan 25 04:53:21 I guess thats just to get the qml toolkit preview Jan 25 04:53:40 there is an interesting question - BB10 supports dev on Mac/Linux/Windows... Android supports dev on Mac/Linux/Windows... iOS supports dev on OSX... WP8/WRT supports dev on W8. Jan 25 04:54:27 in particular, I think saying you must have Windows 8 to develop for WP8 or WRT is going to be a mistake Jan 25 04:54:36 I think win8 is already a big mistake Jan 25 04:54:42 too little, too late Jan 25 04:54:46 bad implementation, ugly os Jan 25 04:54:57 really nice technology stack (.net etc) Jan 25 04:55:00 ron_frown: and you don't think ubuntu has totally missed the boat? Jan 25 04:55:17 cantido I think they have a few things working in their favor Jan 25 04:55:23 they are def late to the game Jan 25 04:55:35 if its just qt, qt is standard enough where people know it Jan 25 04:55:40 it comes down to whether shuttleworth can work with the carriers Jan 25 04:55:50 if they go about it like android free to implement, and put anywhere anyone wants Jan 25 04:55:55 I don't see it really.. they haven't made that much of a dent on the desktop and they're giving it away Jan 25 04:56:00 the lack of java vm is going to be a positive thing Jan 25 04:56:07 :) Jan 25 04:56:09 if I were android I'd expand on what was possible with ndk Jan 25 04:56:13 developers, developers, developers Jan 25 04:56:16 ron_frown: dalvik isn't a jvm ;) Jan 25 04:56:24 same fucking diff Jan 25 04:56:26 vm is vm Jan 25 04:56:46 as best I can tell the only purpose for dalvik is to circumvent oracles shit Jan 25 04:56:57 and there are small jvm's for ARM etc machines, so I wouldnt be surprised if lots of ubuntu phone apps are ported android java apps Jan 25 04:57:22 cantido well I mean you have that with ios too Jan 25 04:57:30 i bet FirefoxOS has more potential than Ubuntu Phone at this point, just because they actually have carriers on board and real dev phones Jan 25 04:57:32 but look at how apps run with significantly less capable hardware Jan 25 04:57:42 g00s have you used firefox os? Jan 25 04:57:44 its pretty shitty Jan 25 04:57:49 i bet it is Jan 25 04:57:49 I was messing with it today Jan 25 04:57:57 html5 is - generally, pretty shitty Jan 25 04:58:17 I think you should say generally pretty shitty when attempting to replace native desktop apps with it Jan 25 04:58:32 there are some fundamental things that need to happen to HTML5 to make it better for mobile tcp connections Jan 25 04:58:32 I've built some nutso html5 shit Jan 25 04:58:39 I just think that unless you've got *big* marketing money, *mainstream* hardware support and *serious* carrier support... both Canonical and Mozilla are chasing up a dead end Jan 25 04:58:40 but its not designed to replace a REAL app Jan 25 04:59:06 ms has some pretty decent hardware support and it looks like win8 is doing no where near as good as they should have been Jan 25 04:59:29 Nokia should just give up and build some android phones Jan 25 04:59:35 if google just had a decent native gui toolkit available for ndk, I think I'd be pretty satisfied with google Jan 25 04:59:42 I know you can use them Jan 25 04:59:44 ron_frown: but the surface RT was kinda … a joke spec wise Jan 25 04:59:56 s/google/android Jan 25 05:00:01 i think half the storage wound up going to a student edition of Office Jan 25 05:00:04 qt requires neccesitas Jan 25 05:00:11 which is pretty burly library stuff for qt Jan 25 05:00:13 I think we can probably all agree that RT is a complete abortion Jan 25 05:00:37 i think MS is keeping it to hedge their bets against instel Jan 25 05:00:41 I really like clutter/mx but I think thats even less likely to fly for native toolkit Jan 25 05:00:54 java is super easy Jan 25 05:00:56 but Jan 25 05:00:56 Windows 8 will 'succeed' just because it's Windows, and there's still way too much inertia involved for it to 'fail' Jan 25 05:01:00 I am just hating it Jan 25 05:01:23 leeds numbers were just released today saying linux had 2x the market share of desktop os's than ms Jan 25 05:01:25 * capella decided to hate it 4 months ago Jan 25 05:01:29 I have no idea where they came up with those numbers Jan 25 05:01:46 WP8... could survive as a minority platform, but is incredibly likely to challenge either Android or iOS for any sort of dominance, ever Jan 25 05:02:05 ron_frown: released by who(m)? Jan 25 05:02:11 Leeds I'd say only because ms will continue to pump money into it forwever Jan 25 05:02:20 like ie Jan 25 05:02:27 nobody likes ie, and its a big as pos Jan 25 05:02:30 and ms keeps developing Jan 25 05:02:32 Leeds: i was reading yesterday, that BB10 actually had more interest than WP8 dev Jan 25 05:02:49 see this is what I personally dont like Jan 25 05:03:03 arrogant companies like bb and ms that thought they could just sit on the top forewver Jan 25 05:03:03 g00s: my personal feeling, based on hearing from MS and RIM developer advocates, is that BB10 is a more interesting platform for me Jan 25 05:03:09 they failed to make a move and they got burned Jan 25 05:03:21 now they are desperatly trying to suck whatever wang will get them market share Jan 25 05:03:26 but the ship has kinda just sailed Jan 25 05:03:49 btw, this is where I saw those numbers http://www.androidpolice.com/2013/01/23/developer-economics-2013-report-devs-want-a-third-platform-tablets-nearly-as-important-as-phones/ Jan 25 05:04:00 (i try not to pull stuff out of my ass;) Jan 25 05:04:04 ron_frown: thing is, what really happened was that a new market segment grew from nowhere (consumer media/app phones) which they just didn't play in, at all Jan 25 05:04:05 that will be the fate for apple too if the don't start fixing shit instead of breaking it Jan 25 05:04:29 windows mobile had a jump start of at least 10yr Jan 25 05:04:36 you could buy apps for it Jan 25 05:04:40 you could build/sell apps for it Jan 25 05:04:45 dominance within a segment doesn't change much - and doesn't change fast - but when a new segment grows it's a much more open market Jan 25 05:04:51 I am trying to use sharedpreferences, but everytime i try to set my preference to a variable, it throws a java.lang.nullpointereception on lines 13 and 14. If i hard set the variables everything works as expected. http://pastie.org/private/gbeytzupbmljr8xtpvfdda Jan 25 05:04:55 ron_frown: but it was never a consumer media platform Jan 25 05:05:12 str8uplinux: did you read what I replied to you about half an hour ago? Jan 25 05:05:27 Leeds do you really need to ask? Jan 25 05:05:28 haha Jan 25 05:05:33 and there is weird shit along the lines of Lenovo working with / buying RIM (but the canadian govt probably wouldn't go for the latter) Jan 25 05:05:34 well, no :) Jan 25 05:06:03 g00s: there's a point where they either let the jobs go to China/Korea/Japan or they let the ownership go Jan 25 05:06:22 I dont know what could save bb at this point Jan 25 05:06:31 you could ron_frown Jan 25 05:06:36 *MAYBE* just opening everything up so anyone could build devices with it Jan 25 05:06:41 but because they license qnx Jan 25 05:06:45 dont see that happening Jan 25 05:06:47 plus Jan 25 05:06:55 when there is something thats already more attractive Jan 25 05:06:57 WHY? Jan 25 05:07:14 Leeds: I did read that, but it doesnt really make sense because it is definitely dying on those two lines before i try to use the strings anywhere (also, i'm actually trying a different route now... i was just trying a simple test with the code i showed you before) Jan 25 05:07:27 str8uplinux: paste *actual* code and *actual* stracktrace Jan 25 05:08:21 str8uplinux: have you tried using the debugger? Jan 25 05:08:48 debugger.... what the FUCK is debugger? Jan 25 05:09:04 a lot of people have no idea it seems Jan 25 05:09:22 haha Jan 25 05:09:24 "I get null pointer, why I get null pointer, tell mez!" Jan 25 05:09:49 but it compiles? Jan 25 05:10:00 wabz: THEN SHIP IT Jan 25 05:10:02 althought it did have some warningz Jan 25 05:10:22 They should hide all of the SDK in an extra package called "debugger" to reduce the amount of crappy apps made by people that won't use the tools provided Jan 25 05:10:29 Leeds: The actual code is pasted just above. Jan 25 05:10:52 str8uplinux: and the stacktrace? Jan 25 05:10:58 getting it now Jan 25 05:11:16 haha Jan 25 05:11:39 i do c# professionally and vs.net is the best IDE I've had experience of woring with Jan 25 05:11:44 cantido: I've always thought that anything click-through shouldn't have "yes" or "no" answers - it should be something like "red',"green","blue", with a randomly-generated correct answer which is inserted into the middle of the text Jan 25 05:11:48 ron_frown: whoring with? Jan 25 05:11:51 it has a concept of this immediates window which is like beanshell but within the context of your code Jan 25 05:12:10 I rember showing my coworker the immediates window and the shit you can do in it Jan 25 05:12:25 he'd been a developer in vs.net since .net 1.1 which was like 2001-2002 Jan 25 05:12:32 and he had no idea that debugging feature existed Jan 25 05:12:53 * cantido must be only person that actually likes eclipse.. Jan 25 05:13:06 yes, you apparently are Jan 25 05:13:12 ignorance is bliss dude Jan 25 05:13:13 cantido: i use eclipse too :| Jan 25 05:13:18 if you are happy and dont plan on changing Jan 25 05:13:18 i think canadiancow also Jan 25 05:13:23 just the 3 of us now ... Jan 25 05:13:24 dont touch intellij Jan 25 05:13:32 ron_frown: can vs.net debug m68k code via gdb? Jan 25 05:13:41 no idea Jan 25 05:13:57 I dont think they even use gdb for their c++ stuff Jan 25 05:14:02 eclipse has some rough edges, but it can do a lot of shit that makes my life easier Jan 25 05:14:12 again Jan 25 05:14:18 dont us anything else Jan 25 05:14:33 my boss ask me how "hard" mssql is (we're an oracle shop) Jan 25 05:14:42 I said are you planning on ever swapping our codebase over? Jan 25 05:14:53 no? then please for the love of god and all that is holy Jan 25 05:14:55 dont touch mssql Jan 25 05:15:28 using oracle and oracle's tools is just torture for me Jan 25 05:15:29 haha Jan 25 05:15:35 knowing what things COULD be Jan 25 05:15:41 and seeing what they actually are Jan 25 05:15:45 BAD BAD Jan 25 05:15:49 so you are saying ignorance is bliss ? Jan 25 05:15:54 precisely Jan 25 05:15:57 ron_frown: well, don't ever use lisp Jan 25 05:16:05 because then, everything will start to suck Jan 25 05:16:11 like, everything Jan 25 05:16:13 I take antidepressants because I just "dont want to give a fuck" Jan 25 05:16:53 i'm half joking ;( Jan 25 05:18:08 the more languages you use, i think the more frustrating any one becomes Jan 25 05:18:13 Leeds: here is ALL updated files including StackTrace http://pastie.org/private/heabeflsl3ggz0hlzf0qsq Jan 25 05:18:20 actually Jan 25 05:18:26 i wish i had feature 'X' from language Y now Jan 25 05:18:32 these languages I've done in this order Jan 25 05:18:57 c->c++->delphi->coldfusion->php->c#->f#->c++ again Jan 25 05:19:16 and honestly I'm finding c++ theraputic Jan 25 05:19:34 * cantido likes C + glib for anything that isn't android related.. Jan 25 05:19:51 and c++ 0x has modern languages features Jan 25 05:20:13 ron_frown: java 8 will have lambdas and default methods :) Jan 25 05:20:13 * cantido wouldn't touch C++ with a shitty stick Jan 25 05:20:33 and android will adopt java > than whatever version it has now, when? Jan 25 05:20:36 never. Jan 25 05:20:46 Stroustrup's 4e of the C++ ARM is now , 1300 fucking pages Jan 25 05:20:56 wtf is default methods Jan 25 05:22:03 what i'd love to see in java (and it may already be there) Jan 25 05:22:08 async/await shit Jan 25 05:22:17 str8uplinux: what led you to believe you should be using PreferenceManager.getDefaultSharedPreferences()? Jan 25 05:22:25 async programming in android can become quite the pita Jan 25 05:22:31 thats library stuff though, right ron_frown ? Jan 25 05:22:42 jsr166x and later, fork/join, Jan 25 05:22:54 maybe I'v enever messed with that crap Jan 25 05:23:11 my java stuff, I do in android only Jan 25 05:23:35 Leeds: I did a lot of reading, and I saw everyone using it in their examples. Is there a better route to accomplish this? Jan 25 05:24:06 str8uplinux: firstly, Context.getSharedPreferences() Jan 25 05:25:08 secondly, standard practice, if you're having a null problem, is to separate out calls and assignments - rather than doing foo.getBar().baz.biddle(), make it 3 or 4 lines - at least for debugging - so you can tell where the null actually comes from Jan 25 05:26:22 Leeds: so, Context.getSharedPreferences() -- are you refering to the PreferencesData.java class? Jan 25 05:26:25 ron_frown: i was reading an interesting story today about dianne h and the CM community Jan 25 05:26:37 CM = cyanogen mod? Jan 25 05:26:38 I'm referring to getting your shared preferences Jan 25 05:27:25 ron_frown: yeah; basically (and i agree with her) she mentioned that google might not allow play access from CM if they made certain changes to the UI Jan 25 05:27:35 I agree Jan 25 05:27:37 Leeds: so do away with the class i'm currently using (PreferenceData.java)? Jan 25 05:27:49 I honestly havent heard much from dianne Jan 25 05:27:52 str8uplinux: I'm not going to tell you how to write your code Jan 25 05:27:56 I kinda thought she may have parted ways from google Jan 25 05:28:02 actually, it would be cool if - as a dev, i could say 'allow downloads / ratings from non stock roms' Jan 25 05:28:20 ron_frown: really? you think she left ? Jan 25 05:28:23 no Jan 25 05:28:29 I just havent heard shit from her Jan 25 05:28:35 havent seen her participation Jan 25 05:28:36 :) Jan 25 05:28:45 maybe I'm not just hanging around in the right circles Jan 25 05:28:49 Leeds: i understand that, i just am confused on where you are telling me to use Context.getSharedPreferences(). use that instead of PreferenceData.getLoginUID(this) ??? Jan 25 05:29:18 str8uplinux: I'm telling you to use that to get your shared preferences... Jan 25 05:29:31 AHHHHHHHHH Jan 25 05:29:37 its just shared preferences for christs sakeS! Jan 25 05:29:49 just get defaultsharedprefereences and getString / pustring Jan 25 05:29:58 pass in default value as second parameter Jan 25 05:30:00 not hard Jan 25 05:32:44 ron_frown: if you could only afford one simple, low cost item for android dev, what would it be? preferrably around $200. i'm guessing a n7 Jan 25 05:33:13 honestly Jan 25 05:33:15 oh wow, oreilly came out with a book on Enyo Jan 25 05:33:19 * g00s has ADD Jan 25 05:33:34 I dont do much tablet shit... I've found my panda es's to be invaluable Jan 25 05:33:37 and my odroidx Jan 25 05:33:54 i remember looking for panels and it was a pain Jan 25 05:34:04 I like dev boards better as they have basically huge expansion options Jan 25 05:34:05 what has official support in the tree? panda or beagle ? Jan 25 05:34:11 panda Jan 25 05:34:31 * g00s is a poor panda Jan 25 05:34:31 though I've fixed countless bugs in the panda support Jan 25 05:34:32 haha Jan 25 05:34:49 oh man, so i'm going to be using your code? thats a showstopper Jan 25 05:34:59 I never submitted to google Jan 25 05:35:00 so Jan 25 05:35:01 haha Jan 25 05:35:02 :) Jan 25 05:35:24 they say you can only build android stuff from head Jan 25 05:35:33 but the head is their dev branch and is almost ALWAYS unstable Jan 25 05:35:33 lol Jan 25 05:36:16 yeah, right now, i dont want to compile my own roms and stuff; i dont have time for that. Jan 25 05:36:48 well if you dont ever plan on doing perhiferal stuff or if a arduino would suffice Jan 25 05:36:51 just get a n7 Jan 25 05:36:57 oh shit, so if i get a panda, will it have Play services ? Jan 25 05:37:11 I guess not Jan 25 05:37:17 fsck Jan 25 05:37:29 so, that won't work if i decide to do any mapping stuff … uuuh Jan 25 05:37:30 well thats an easy fx =) Jan 25 05:37:59 I've got i2c stuff, usb and serial perhiferals etc Jan 25 05:38:19 scanners, scales, ultrasound, hand punches etc all hooked up Jan 25 05:38:27 yeah, i'll just get a n7 for now, dont care if its going to be updated in a month or whatever - seems like, it s good to dev on things lots of people have Jan 25 05:38:50 i haven't done i2c in at least 10 years Jan 25 05:39:39 I only got a few chips that even need it Jan 25 05:42:13 a lot of what I do could easily be solved with arduino Jan 25 05:43:02 fun stuff; i like sensors & measuring the environment Jan 25 05:43:10 I got that Jan 25 05:44:13 i like stuff along the lines of the air quality egg Jan 25 05:44:24 too bad you dont live out here Jan 25 05:44:32 I got some ideas to make some bucks along those lines Jan 25 05:44:36 not air quality Jan 25 05:44:41 but stuff like that Jan 25 05:45:03 i dont think arizona is telling the truth about the air; time for citizens to do it themselves Jan 25 05:45:58 for example, they only measure particulates for a small part of the valley, and everything outside of that area is 'green' Jan 25 05:46:03 which is, of course, total BS Jan 25 05:48:50 would be cool to do aerosol spectroscopy Jan 25 05:49:20 nice Jan 25 05:50:30 but, once you start building stuff - you need a lot of space Jan 25 05:50:38 haha Jan 25 05:50:41 for test equipment - etc, and space i do not have Jan 25 05:50:58 one of my current idaes could be multimillions Jan 25 05:51:16 in fact, i was going to try living out of my backpack for a year, or maybe at most my truck to do some photography Jan 25 05:51:35 hahah Jan 25 05:51:36 nice Jan 25 05:51:48 I'm far far far too obsessed with technolog Jan 25 05:52:32 well the nice thing, is that - and thats why i really wanted a Surface Pro - is that i could have in a small package my ereader, Lightroom, and regular apps … along with my camera Jan 25 05:52:46 and maybe a lilliputian fuel cell system Jan 25 05:54:39 all that stuff could fit in my pack; no laptop, etc Jan 25 05:54:57 your photography is your art Jan 25 05:55:02 my software is my art Jan 25 05:55:09 well that may be the case with you as well Jan 25 05:55:10 but yeah Jan 25 05:55:12 oh well Jan 25 05:55:19 :) Jan 25 06:04:49 damn it ron left; panda as BLE … wonder if there is any way to get it to work under android Jan 25 06:04:58 * has BLE Jan 25 06:26:22 in my eclipse IDE, the type head for android stuff does not seem to be working. anything I can do about that? Jan 25 06:48:17 hello all Jan 25 06:48:37 in ABS, how do you add a header/title to an action item that expands out a sub-menu? Jan 25 06:51:39 what do you mean header? Jan 25 06:52:20 Hey guys, I'm pretty new to java, if I want to read the bytes from an InputStream into a byte[] array how could I do that? This example I'm trying to use calls getBytesFromFile but that method does not seem to exist. Jan 25 06:52:32 so I have a + icon that expands a sub-menu saying Add Task, Add Destination Task, Add Proximity Task... I want on top of that to say Quick Action Modes Jan 25 06:52:41 should I just add that as an item that is unclickable or something? Jan 25 06:53:05 plantian: you should go to #java for that Jan 25 06:54:17 QubeZ: Okay, its an android app but I'll try them. Jan 25 06:54:34 plantian: its more java specific than Android -- thats why Jan 25 06:54:59 JakeWharton: I'm implementing your ABS into To Do Mapr Jan 25 06:55:04 plantian, you want the entire inputstream in one byte array? Jan 25 06:55:18 probably you want to read chunks of bytes at a time Jan 25 06:55:35 look at InputStream.read() Jan 25 06:55:41 RustyShackleford: ha yeah that is what I thought, I know its a bad idea, I'm just making a prototype though and following someone else's example for better or worse Jan 25 06:55:58 RustyShackleford: In that case though, would I read it into an ArrayList? Jan 25 06:55:59 QubeZ: you can only do what the native API allows Jan 25 06:56:02 refer to it Jan 25 06:56:16 JakeWharton: ok Jan 25 06:56:19 plantian, I don't think read() can return an arraylist Jan 25 06:56:56 RustyShackleford: Sorry, I mean I have to make the array list and then accumulate the chunks into it. Jan 25 06:57:00 plantian: i happen to have that code open Jan 25 06:57:03 on the note of proximity alerts then... I have implemented the proximity alert and it shows up when I run a 'dumpsys location' but it wont fire. I have an Apple store entered in which is 1 mile away and I have radius set to 5 miles but still not firing. Jan 25 06:57:12 registered receiver in Manifest Jan 25 06:57:16 plantian, perhaps that will work. Are you retrieving a file over the network? Jan 25 06:57:27 plantian: http://pastebin.com/KBsKknxm Jan 25 06:57:56 plantian, you are better off reading 1024 (or whatever) bytes at a time, then writing them to an OutputStream Jan 25 06:58:12 BUFFER_SIZE in mine is 0x1000 Jan 25 06:59:24 JakeWharton, does your buffer size make a huge difference? Jan 25 06:59:34 i suppose it depends what you are reading Jan 25 06:59:46 i'm reading HTTP response bodies Jan 25 07:00:22 here's my code for the proximity alert: http://pastebin.com/HRpymNRj Jan 25 07:00:32 RustyShackleford: Yeah I will have to rewrite this part of the example, I just wanted to get something working for proof of concept. Thanks for the help. Here is the blog post I'm trying to replicate: http://eggie5.com/8-hook-share-picture-via-menu-android Jan 25 07:00:43 I am using the Fluffy Location library Jan 25 07:00:57 JakeWharton: I think that will do it, I can clobber my code into that I think. Jan 25 07:16:18 Leeds: have you tried using the rpi for nas? Jan 25 07:25:52 looks like fluffy lib is not good to use to implement proximity alerts Jan 25 07:27:02 im not even sure how people implement proximity alerts properly because if you set your own listener in the activity then its gone when you leave the activity... how to track the user for use of proximity alerting then outside the app? Jan 25 07:29:03 proximity alerts are broadcasts, not sure with fluffy Jan 25 07:29:49 fluffy broadcasts too for use of user location within an app but I'm curious how proximty alerts work in android once you register one Jan 25 07:30:26 oh, i would think the radio would store the coordinates in a register and wake up hw once the fence is crossed Jan 25 07:30:28 For example, I know the user location in my app -- I have added a proximity alert and registered the reeiver in my manifest. I can see the proximity alert when I 'dumpsys location' with radius 5 miles. But it never fires. Jan 25 07:30:46 so I'm a little baffled why it will not fire Jan 25 07:30:55 try another phone ? Jan 25 07:31:05 I have a broadcast receiver with onReceive implemented to catch the intent Jan 25 07:38:49 do you have to implement your own location listener in the activity to create the proper proximity alert? Jan 25 07:44:05 Hello! I am trying to draw in surface view using own thread (draw ring and rotate. at first i draw ring on bitmap create bitmap drawable and draw it on surface canvas always with different angle). But actually rotation is not smooth when ring makes 1 turn per second (if it faster - it is not so visible). Movement is jerking and it is visible. I had tried to sleep thread on 20ms and move ring... Jan 25 07:44:06 ...on 7.2 angle or another ratios but result always bad. After i use time current and previous time compressions instead of TimeUnit sleep but no results. Movement is jerking. May be i should put ring in view an use animation framefork? Jan 25 07:46:14 or use opengl es 20? Jan 25 07:48:01 ;( Jan 25 07:48:34 http://www.vgmemes.com/wp-content/uploads/2012/01/sad-frog.jpg Jan 25 07:48:42 can anyone give me the basic rundown on what is needed to port ICS to a gingerbread phone? Jan 25 07:49:07 MunKy: step #1 : go to #android-root Jan 25 07:49:15 … profit Jan 25 07:49:17 lol, ok Jan 25 07:58:56 hey all Jan 25 07:59:05 hello, anybody can suggest an id3 tag reading library for android? I have tried jaudiotagger and jid3 both are showing errors in logcat. seems like they are not supposed to be used in android. or maybe somebody know how to use one of them? Jan 25 08:02:16 hey all Jan 25 08:02:23 I wanna ask a question Jan 25 08:02:38 I have renamed the folders and recompile my project Jan 25 08:02:50 now im not able to press button Jan 25 08:02:54 i dont know why Jan 25 08:03:20 xyz87: does it give an error in logcat? Jan 25 08:03:50 I have a bitmap that I load through decodeResource of a drawable. I immediately call bitmap.createscaledbitmap and set the bitmap to the result. However, the bitmap does not scale. In debugger, height and width of bitmap are -1, and I can scale the bitmap if I call the function later. What part of the lifecycle do I need to wait for to be able to scale the texture? Jan 25 08:03:52 http://pastebin.com/MiVX4d96 Jan 25 08:05:37 unexpected value from nativeGetEnabledTags Jan 25 08:05:59 xyz87: did you clean the project after renaming everything? Jan 25 08:08:02 how ??? Jan 25 08:08:44 Project -> Clean (Project in top tabs) Jan 25 08:10:43 did it now Jan 25 08:10:47 and get same error Jan 25 08:12:23 xyy87: sorry, unfamiliar with that error (a lot of times resource tags/ids don't get recompiled correctly until you clean, but not the case here I guess) Jan 25 08:12:38 mm Jan 25 08:23:32 nvm, bitmap problem solved; user error Jan 25 08:30:03 PEBCAK Jan 25 08:58:34 what is the corrent size for action bar icon http://developer.android.com/guide/practices/ui_guidelines/icon_design_action_bar.html or the one in action bar icon template pack Jan 25 08:59:08 site xhdpi 48x48 ---> icon pack xhdpi 64x64 Jan 25 08:59:10 ?? Jan 25 09:15:07 zipoboy, http://developer.android.com/design/style/metrics-grids.html Jan 25 09:15:12 this explains why 48 Jan 25 09:18:39 hi Jan 25 09:19:16 hey, is it possible to pass arguments through adb shell [argv1[, argv2 [, arv3 [, ...]]]] ? Jan 25 09:19:26 yes Jan 25 09:20:06 squ, yes for me? Jan 25 09:20:09 for example, adb -s emulator-5556 shell sqlite /data/data/your.app.id Jan 25 09:20:31 squ, and how to get those args inside the appl Jan 25 09:25:07 squ? Jan 25 09:27:12 you want to start application with parameters from command line? Jan 25 09:27:15 :) Jan 25 09:29:33 squ, i really do :) Jan 25 09:29:48 I do not know how to do it Jan 25 09:30:11 uh Jan 25 09:30:29 if you are writing a c command line app, its like any other c command line app :) Jan 25 09:30:49 if its a typical android java app, i don't think there is any way Jan 25 09:30:53 g00s, but I need the app to run specifically on an android OS Jan 25 09:30:53 it might be interesting to fire a device action from, say, web-page event Jan 25 09:31:12 like sms or a call maybe Jan 25 09:31:21 is that what you are going to do? Jan 25 09:33:02 squ, not exacly, this is part of an automation process. This app should be triggered by a remote app. I found it a good idea to run it calling "adb -s IP:PORT shell " Jan 25 09:33:26 if I cannot pass arguments I'll have to stick with default paths for config files with those arguments Jan 25 09:34:30 this is not an ordinary app for a common user, this is for hardware testing purposes Jan 25 09:35:47 why not use the am command ? Jan 25 09:36:00 g00s, never heard of it, lemme google it Jan 25 09:36:14 adb shell am Jan 25 09:36:32 here http://stackoverflow.com/questions/3228245/how-can-i-deliver-parameters-to-a-test-function-that-launched-using-adb-shell-a Jan 25 09:36:56 cheers Jan 25 09:49:00 Hi, would someone comment on my idea? I want to draw on a view multithreaded, drawing a bitmap in onDraw, drawing that bitmap from a Threadpool. The problem is how to get the threads to draw in the right order... Jan 25 09:50:21 flan3002: use the thead mechanisms, like notify and wait, to wait for all previous threads to draw. But as far as I understand, you need a specific order in which your threads draw? So you anyway want to draw one thread after another, do you need threads at all? Jan 25 09:50:53 I want the threads to calculate something, and then draw it. Jan 25 09:51:12 So... I guess I need them, for speeding up my app Jan 25 09:53:40 This is the first time I need to use Threadpools, would you specify how to use notify and wait here? Do I have to wait the correct thread hasn't drawn yet, and after each thread, notify all Threads, for them to check, if its their turn? Jan 25 09:57:05 timroes: Is that the right way? Jan 25 10:02:06 how can i encode date format to strings.xml %1$te.%1$tm.%1$tY %1$tH:%1$tM:%1$tS so it's valid...? Jan 25 10:05:07 does any application ever actually use AudioManager.AUDIOFOCUS_GAIN_TRANSIENT_MAY_DUCK? Jan 25 10:05:38 miha: i think you need %1 %2 %3 etc Jan 25 10:05:38 hi! is here someone who is involved in the development of the "jobb" tool which seems to be part of the android SDK? it seems this tools uses the "fat32-lib" I maintain, and people are having trouble with that. Jan 25 10:05:46 because I implemented reacting to AudioManager.AUDIOFOCUS_LOSS_TRANSIENT_CAN_DUCK but I can't find any application to trigger it for me :p Jan 25 10:06:02 Zharf: use am :) Jan 25 10:06:21 am? Jan 25 10:06:27 activity manager Jan 25 10:06:31 (I'm getting bug reports which seem to be related how the fat32-lib is used by the jobb tool) Jan 25 10:06:32 for sending an intent Jan 25 10:07:28 g00s, it's not an intent though Jan 25 10:07:34 oh, sorry Jan 25 10:08:08 g00s: thx! Jan 25 10:09:56 for android unit testing, is there some kind of @Ignore or @Test annotation? Currently it seems like a method that is a test HAS to be prefixed with test, and the typical JUnit Annotations do not work at all Jan 25 10:10:16 thats because its junit 3.8 Jan 25 10:10:23 annotations came in junit 4, i think Jan 25 10:10:46 i have no idea why we're still using such an old version of junit, i think from 5 years ago at least Jan 25 10:11:07 flan3002: sorry was shortly away, i made something similiar, that i can show you Jan 25 10:12:06 timroes: Wow, thanks Jan 25 10:12:37 ooh, that's too bad Jan 25 10:13:10 flan3002: one minute :) Jan 25 10:13:11 so I actually have to make tests "ignored" by commenting out or renaming? Jan 25 10:14:03 Hi Jan 25 10:15:42 I have a problem with JSONException: Value null at uTraf JSONObject$1 cannot be converted to JSONArray Jan 25 10:17:29 When I do this Jan 25 10:17:30 JSONObject json = jsonFunctions.getJSONfromURL("http://url.de); Jan 25 10:17:30 An it return null Jan 25 10:17:30 I do this: Jan 25 10:17:30 if (json != null){ Jan 25 10:17:30 But this does not work. Why? Jan 25 10:18:21 flan3002: so now i am back :D Jan 25 10:18:40 flan3002: i give you my usecase, so you can understand what I tried to do. Jan 25 10:24:59 well, I wrote a simple test program for the audiofocus stuff, but it's kind of annoying that nothing seems to actually use the transient may duck thing Jan 25 10:29:30 I've declared a service in app 1 which is also usable from app 1, however when I try to access it from app 2 it says "unable to start service intent $servicename not found" Jan 25 10:29:42 Any clue what could cause it or how to debug that? Jan 25 10:30:39 missing from manifest in app 2? Jan 25 10:31:19 pleed: have you exported it from app 1? Jan 25 10:32:31 is my question stupid? Jan 25 10:35:16 dot8, show jsonFunctions.getJSONfromURL Jan 25 10:37:42 timroes: Will you explain it? Are you still there? Jan 25 10:38:13 dot8, maybe this actually not initializes JSONObject json = jsonFunctions.getJSONfromURL? try split to JSONObject json = null; and son = jsonFunctions.getJSONfromURL; and watch what happens in this case Jan 25 10:39:28 Udmurt, I will try it Jan 25 10:39:39 flan3002: sorry one moment :D Jan 25 10:39:46 office wards going on here :D Jan 25 10:40:17 :D Jan 25 10:45:40 Meh. that WebView.loadData() is so b0rked. works on android 4.x, uses wrong encoding on cm7, and displays android's "web site is not available" error on samsung galaxy s2 original 2.x rom... Jan 25 10:45:43 Leeds: in the manifest, yes Jan 25 10:48:07 I tried again using the componentname and now the service can be bound. Ran into another error saying "binder invocation to an incorrect interface" Jan 25 10:49:22 Is it possible to use Apache Mahout on Android? Jan 25 10:51:06 whoops. http://code.google.com/p/android/issues/detail?id=4401 Jan 25 10:52:39 webview is a pos Jan 25 10:58:21 ok, the loadDataWithBaseURL() workaround seems to work Jan 25 10:58:41 flan3002: dude, war is over now :D Jan 25 10:58:45 still there? Jan 25 10:59:19 the war is over ? Jan 25 10:59:22 \o/ Jan 25 10:59:29 g00s: webview in 2.3 .. massive pos :( Jan 25 10:59:30 office war :D Jan 25 10:59:54 timroes: still there :D Jan 25 11:02:06 so my use case was: i have objects in a list and want to convert each object in that list, so i have afterwards a new list with the converted objects. Now i want to convert each object parallel, so it can do the conversaion all besides each other Jan 25 11:02:14 but i still want to have the same order in the new list Jan 25 11:02:43 thats pretty much what you want to do. do all calculations beside each other, but afterwards assure the right order, when in my case i add to list you want to draw Jan 25 11:03:23 exactly, not much difference at all... Jan 25 11:04:15 https://github.com/inovex/jMOM/blob/dev/src/de/inovex/jmom/CollectionConverter.java#L95 Jan 25 11:05:03 i create a list of threads in line 95, afterwards i iterate over all objects in dblist (the original obejcts) and create for each a new thread, and start it (if multithreading is enabled). Afterwards I add the thread to the thread list. Jan 25 11:05:58 in line 112 then i iterate obver all the thread in the thread list, wait for each to be finished, and then add the result to the new list Jan 25 11:06:37 Couldn't you pass a "holder" with what you want processed and a "slot number" to tell the thread where to put the result? Jan 25 11:07:03 and where does it get started (where do you supply the executorservice with you threads)? Jan 25 11:07:10 cantido: I could, but since I anyway need to wait for all, to be finished, i cal also just insert them in that order :) Jan 25 11:09:29 * cantido is glad java has decent concurrency stuff Jan 25 11:10:10 Anyone know any good neural network lib to use on Android? Jan 25 11:12:29 valgrind: what are you training your NN for ? Jan 25 11:12:51 timroes: Thank you, I understood, how you did it. This is going to help me alot... Jan 25 11:13:17 I wanna use it instead of a decision making tree, for game AI Jan 25 11:13:43 flan3002: okay, there are ofc other ways, perhaps some more android specific one, but that lib must run also on JavaSE, so I used the regular threading stuff Jan 25 11:17:47 timroes: Sorry for another question, but I'm using threadpools for the first time. How do I wait for one of them? Jan 25 11:18:16 There is no join(). Jan 25 11:19:22 No but there's a wait() function... Jan 25 11:25:47 valgrind: But then, how do I coordinate the order? timroes approach was to wait for each thread in order, and then insert the computed result. Jan 25 11:33:42 aren't threads scheduled in an arbitrary way by the machine itself? you might want to check tjee oracle website for a way to enforce an order Jan 25 11:39:04 valgrind: Yeah, they are, but I want them to draw their results in a specified order, Jan 25 11:39:16 Thats why I need join(). Jan 25 11:40:34 fine... a join calls allows you to tell the machine i want my main thrad to wait for a specific thread (given its pid) to finish... Jan 25 11:40:53 the wait function if i remmeber correctly pauses your main thread until the worker calls notify Jan 25 11:41:14 so somehow you could notify the main thread in order Jan 25 11:41:19 valgrind: have you looked at Handler's? Jan 25 11:41:19 But I need at least 4 workers Jan 25 11:41:54 flan3002: you may want to look at either a Handler or Activity.runOnUiThread Jan 25 11:42:05 clever: Handlers? They do allow me do control the order in which Threads draw? Jan 25 11:42:26 flan3002: they let you run code on a certain thread Jan 25 11:42:38 clever: Thats now what I want... Jan 25 11:42:42 *not Jan 25 11:43:27 flan3002: if the order matters, why cant you just put all the actions in the same thread? Jan 25 11:43:33 The problem is, that I want them to draw onto a bitmap in MY order. Jan 25 11:43:50 each draw action must then follow the last Jan 25 11:44:21 But because I do calculation in between, thats slower Jan 25 11:44:43 what if one thread did nothing but calculations, and a second did nothing but drawing? Jan 25 11:44:49 It already works, but now I want it to be multithreaded Jan 25 11:45:38 And how do I get the results from the four workers to the drawing thread? How does the drawing thread know when to draw? Jan 25 11:46:02 I want NO static fields! Jan 25 11:46:30 handlers Jan 25 11:46:45 when your finished doing work, use a handler to run the drawing code on the drawing thread Jan 25 11:47:20 But handlers do it in the order of posting. And the order of posting is more or less random when having multiple threads Jan 25 11:47:39 So I get flickering because of z-order randomness Jan 25 11:47:49 what if you just had 1 function that draws everything in order, every time its called? Jan 25 11:47:58 Thats too slow :D Jan 25 11:48:09 what about 5 bitmaps, one for each z layer? Jan 25 11:48:12 with alpha Jan 25 11:48:23 so you only redraw the layer that changed, then lay them ontop of eachother Jan 25 11:48:46 No, that'd be a bad thing, big memory consumption, it must be doable like this. Jan 25 11:49:36 you will always need to redraw the things at the top, if you did things at the bottom Jan 25 11:50:07 so your only choices are to either wait for it to finish that work (causing the flicker you dont want) Jan 25 11:50:13 ot redraw the top things with old data Jan 25 11:50:31 flan3002: habven't read anything, but as far as i know, threadgroup won't help you with that Jan 25 11:50:47 or blow your memory budget and store every object to its own bitmap Jan 25 11:50:50 its more like a grouping class, for easier passing around threads Jan 25 11:51:44 But, cant I just let each thread wait, except for the thread whos drawing, and when he's finished, wake up all others, for them to check if its their turn? Jan 25 11:51:59 yeah you can Jan 25 11:52:28 when you want to wait, call wait in the thread, Jan 25 11:52:41 and when you drawing thread is done, call notify on all threads Jan 25 11:52:43 I know that much ::D Jan 25 11:53:08 so? Jan 25 11:53:33 threadgroup is afaik really just a container for several threads, without much functionallity Jan 25 11:54:36 Eliminating the overhead of starting 4 threads (4 cores in Nexus 7) per frame... Jan 25 11:54:56 you anyway shouldn't start threads each frame :D Jan 25 11:55:11 Thats why I use a Threadpool Jan 25 11:55:28 ah we are talking about a thread pool and not a thread group :D Jan 25 11:55:32 sorry must have missed that part Jan 25 11:55:54 It keeps them started, and only assigns tasks to them. Threadpool Jan 25 12:04:04 clever: any Idea? Jan 25 12:04:14 already listed several Jan 25 12:04:39 flan3002: what are you using to draw? Jan 25 12:04:47 Do you know if I can use ONE canvas from multiple threads? Jan 25 12:05:11 timroes: Canvas :D Jan 25 12:05:28 flan3002: then you deinfately don't want to create bitmaps for each thread with alpha and draw them over each other Jan 25 12:05:35 like suggested above. It will get slow like.. Jan 25 12:05:43 there is no comparision how slow it will get :D Jan 25 12:06:04 hmm I cannot see Chrome Beta for Android on my Galaxy Note 10.1 (Android 4.0.4) Jan 25 12:06:10 anyone know why that is? Jan 25 12:06:17 Sorry, what does deinfately mean :) Jan 25 12:06:24 (in the Play Store) Jan 25 12:06:26 definately* Jan 25 12:06:40 perhaps still written wrong, i am no native speaker, sorry :D Jan 25 12:06:43 Is it possible to bind, call and unbind a service while blocking the UI thread? Jan 25 12:06:51 absolutely Jan 25 12:06:54 Me not too... Jan 25 12:07:34 I never wanted to create multiple bitmaps. In fact I didn't create ONE until I wanted to use the full speed... Jan 25 12:07:38 How would one do that? I've got a working example but after binding to the service I have to return to the UI thread loop and wait for the ServiceConnection.onConnected event Jan 25 12:07:41 flan3002: since you anyway are stuck with the right order, why would you need to draw in any non ui thread on the canvas? Jan 25 12:07:41 definitely don't == don't ever Jan 25 12:08:52 Because I want to do calculations, then draw onto a bitmap, and in onDraw paint that bitmap to my view. And that's faster with multiple cores Jan 25 12:09:36 yeah but you need to draw in the right oder? wasn't that your problem in the beginning? Jan 25 12:10:12 so you anyway need to make in one thread make sure the order is right (e.g. the way I showed you in my code), and if you just do all the painting in that thread then (not the calculations) you will be fine? Jan 25 12:10:15 odd, when I browse to Play Store through the browser, I _can_ download Chrome For Android Beta Jan 25 12:10:21 is that intentional ? Jan 25 12:12:06 I should, but 1. I already have the code to draw it while calculating, and 2. the data organization is difficult, how do I pass the calculation and how do I notify the thread that they are finished? Jan 25 12:13:00 flan3002: if you don't want to separate drawing and calculation, but you want the calculation to be done parallel, but the drawing to be done in the right sequential order, you are a bit limited :D Jan 25 12:13:14 Yeah Jan 25 12:13:29 because you have mainly ONE block that you want to be executed parallel but sequentiell at the same time Jan 25 12:13:34 what just won't work : Jan 25 12:14:17 then the only possibility I see is what clever suggested, using one bitmap for each thread and painting them oever each other in a last thread, what will be a real performance killer Jan 25 12:15:06 Oh god, I hate multithreading :D Jan 25 12:15:06 at least tried it some weeks ago on a galaxy nexus and an acer iconia a501, and by just painting one fullscreen bitmap over another in a canvas, you already had a real drop in the framerate Jan 25 12:15:32 I'm not going to do the layering. Jan 25 12:16:04 so I guess usual approach (from game design) is to have the calculation and drawing seperated, you can multithread the calculations, but leave the drawing in the right z order Jan 25 12:17:41 Well, I have several objects (count unknown), and I would'nt know how to pass the data. Jan 25 12:19:26 i don'tknow exactly wich data you mean :) Jan 25 12:20:39 i would make an interface containing one onUpdate and one onDraw method, so each object must implement that interface, and you can happily write a thread, that calls onUpdate asynchronously and call onDraw synchronosously (and also make sure onUpdate has already finished or so) Jan 25 12:21:10 Thats the problem, the objects draw themselves without storing what they drew... Jan 25 12:21:24 no problem, the onDraw method gets passed in the canvas Jan 25 12:21:31 Thats why I cant pass the drawing to another Thread. Jan 25 12:21:34 and each object draws itself on that Jan 25 12:22:03 So I would have to store the data of these objects? Jan 25 12:22:18 just because two methods are in the same class doesn't mean you need to call them from the same thread? Jan 25 12:22:46 you only need to store your calculation stuff that you made in onUpdate inside the object itself Jan 25 12:22:49 No, but when I call onUpdate, the calculation would be nonsense if not stored somewhere... Jan 25 12:23:14 that method is called from another thread, that calls onUpdate on all updateas parallel, and onoDraw just draws on the given canvas taking the object state into account Jan 25 12:23:36 yeah you would ofc need to store your calculation (state of objects) in side the objects Jan 25 12:24:03 this is pretty awesome, Android running on Windows 7 seamlessly using the Windows kernel http://www.socketeq.com/ Jan 25 12:24:24 I don't know eexactly your use case, but e.g. from a gaming points of view. I have a player class, in onupdate it calculates it movements, make collision detection, whatsoever, and updates its x and y coordiante respectively. OnDraw, the just draws the player sprite on the canvas Jan 25 12:24:31 at its x and y position Jan 25 12:24:32 Well thats at least a way to go, although I wanted to avoid this. Jan 25 12:24:48 mr0wl: ive always wondered how hard it would be to port the android framework to a linux or windows kernel Jan 25 12:25:17 mr0wl: it feels like it would be as simple as compiling the dalvikvm to x86 and writing a 'video driver' that just draws to an x11 window Jan 25 12:26:20 clever: android uses a linux kernel? Jan 25 12:26:31 timroes: yes Jan 25 12:26:42 timroes: yes, but technically it's not part of spec Jan 25 12:26:42 timroes: yeah, but with its own libs Jan 25 12:27:01 clever: isn't that's how at least two different ports work? Jan 25 12:27:28 p_l: which ones? Jan 25 12:27:34 so if you find linux drivers for the hardware used in the device, you can just compile it for any device? Jan 25 12:27:52 timroes: Thank you a lot for your time, I guess you cant help any further. Jan 25 12:27:56 timroes: thats about half the problem Jan 25 12:27:58 Good bye Jan 25 12:28:02 http://www.bluestacks.com/ Jan 25 12:28:06 clever: whats the other half? Jan 25 12:28:07 timroes: nope Jan 25 12:28:32 timroes: embedded devices don't work the same as PCs, there's for example no way to autodetect that many devices are actually tehre Jan 25 12:28:38 and the locations differ Jan 25 12:28:53 yeah, thats more of a driver issue, what io port and memory address to poke at Jan 25 12:29:05 applies to running any flavor of linux on the device Jan 25 12:29:17 yeah thats what i would have guessed? Jan 25 12:29:39 timroes: except on PCs, various workstations etc. you have autodetection modes etc Jan 25 12:30:00 but isn't that a problem of the drivers and the bios handover? Jan 25 12:30:05 now imagine manual setting of io ports, addresses etc. of IBM PC/XT, except you need it for everything, including knowing where the RAM chips are Jan 25 12:30:17 timroes: there's no BIOS handover Jan 25 12:30:27 yeah so the driver has to get it? Jan 25 12:30:27 and no autodetection support Jan 25 12:30:44 for that hardware, whatsover Jan 25 12:30:51 timroes: so you have compiled-in memory maps, device config lists etc. that form "board config" Jan 25 12:31:04 p_l: arm config tree's do something similar Jan 25 12:31:12 hmm, it was called device tree Jan 25 12:31:14 for example telling you that GPIO[1-4] is bit-banged i2c Jan 25 12:31:34 okay so I still don't see how that is not a problem, that the driver has to get its address Jan 25 12:31:58 and as such more or less a driver problem, that it need to work on that specific hardware Jan 25 12:32:03 I've just read the following sentence about android.app.Application: "There is normally no need to subclass Application. In most situation, static singletons can provide the same functionality in a more modular way. If your singleton needs a global context (for example to register broadcast receivers), the function to retrieve it can be given a Context which internally uses Context.getApplicationContext() when first constructing the si Jan 25 12:32:22 timroes: the only devices that have proper board-independent interface for autoconfiguration are the WinPhone/Win8 ones Jan 25 12:32:40 p_l: why can't you detect stuff on ARM machines? Jan 25 12:32:57 cantido: because what is the blinking led on one device can be power button on another Jan 25 12:33:08 cantido: its the same as old ISA based computers, the interface to auto-detect doesnt exist on the hardware Jan 25 12:33:34 p_l: You might have to tell the kernel where certain things are connected.. but that's pretty far away from no detection at all Jan 25 12:33:34 except IBM PC with ISA had a bunch of standard addresses, so you could get floppy, display and serial port running Jan 25 12:33:40 p_l: sorry i might be a bit slow today, but I still only see: no autodetection, driver must to get its memory ports, ram config whatsoever withotu autodetection Jan 25 12:33:43 cantido: they expect the ROM being loaded in, to be made by somebody who has all the specs, so there is no need to waste resources on autoconfig Jan 25 12:33:46 I want a button to allow a user to dial or text a predefined number, when the user presses it I want a box like the contact box in the phone to show up showing the number and allowing the user to dial or text just as they are used to. I'm pretty sure I can do this with an intent, but among the intents I've found there's ACTION_VIEW, ACTION_CALL, ACTION_CALL_BUTTON and probably more, what should I use? Jan 25 12:33:49 p_l: and recent kernels have device tree Jan 25 12:33:57 and us such need to be specific to that hardware or now where to get that information from that hardware Jan 25 12:34:05 clever: you can probe stuff on arm just the same, stop talking out of your arser Jan 25 12:34:11 s/arser/arse/ Jan 25 12:34:25 cantido: yeah, but there is a chance of hitting a device at a non-standard address and doing who knows what Jan 25 12:34:43 and you can only do that kind of probing once basic stuff like ram and atleast serial console are up Jan 25 12:34:52 clever: you can probe i2c etc too Jan 25 12:34:55 cantido: a lot of the devices on embedded stuff are on non-probe-capable buses (SPI, for example) Jan 25 12:34:55 I have a suspicion it's ACTION_VIEW, but I haven't seen something that answers it for sure Jan 25 12:35:18 p_l: in which case you tell linux where they are, exactly as you need to do on x86 Jan 25 12:35:18 cantido: but if you cant turn on the backlight, how do you give any output to say what the probing found? Jan 25 12:35:29 clever: this is in the kernel Jan 25 12:35:57 clever: hint the i2c stuff is clever enough to detect all but really stupid i2c devices Jan 25 12:35:57 cantido: yes. And you can't enable half of the device because you don't have the pin-mapping of GPIO to various busses Jan 25 12:36:19 p_l: you would have to do the same on x86 Jan 25 12:36:21 cantido: first you need to know how to access the i2c bus Jan 25 12:36:36 yeah, if i remember correctly, the GPIO directly on the cpu, is wired to several things like the status LED's and vib motor, on the treo 650 Jan 25 12:36:36 p_l: the kernel knows about the i2c buses on a SoC Jan 25 12:36:45 and the exact wiring, depends on who laid out the PCB Jan 25 12:36:48 p_l: have you done a kernel port? Jan 25 12:37:00 cantido: it knows because of something called "board config" which has it hardwired for the specific device Jan 25 12:37:03 p_l: or written a driver for i2c or spi? Jan 25 12:37:08 p_l: or device tree Jan 25 12:37:27 p_l: that is the same for gpio on x86 Jan 25 12:37:30 cantido: I didn't do full port, but I was involved a bit in upgrading kernels on some devices Jan 25 12:38:19 p_l: what changes did you need to make? Jan 25 12:38:44 cantido: FYI, some of the stuff involved hardcoded values like what physical memory pages were owned by GPU and were "do not touch" Jan 25 12:38:55 hint: ofcourse shit like gpio configuration that can't be detected can't be detected, duh Jan 25 12:38:57 or similarly hardcoded flash config Jan 25 12:39:08 cantido: yes, that's what we have been talking about Jan 25 12:39:16 and every device out there got it *custom* Jan 25 12:39:20 p_l: thats the same on x86 Jan 25 12:39:32 p_l: that's not a problem that only affects SoCs Jan 25 12:39:57 you might want to look at how the beaglebone handles it's pinmux/gpio configuration for so hints.. Jan 25 12:40:01 s/so/some/ Jan 25 12:40:19 cantido: IBM PC-compatible != x86. PC is also the set of specs that tells you how to get certain shit up, with all the later-coming firmware interfaces Jan 25 12:40:35 p_l: how is gpio magically configured on a PC? Jan 25 12:41:10 how does a PC magically configure stuff connected via SMBUS? Jan 25 12:41:24 so on PC-compatible laptop, I don't have to know which GPIOs on EC will talk to keyboard, I just use the standard PS/2 address. Jan 25 12:41:49 I have a standard way of telling the EC to shut down power Jan 25 12:42:00 I don't have to know how to drive the battery controller directly Jan 25 12:42:11 My app runs multiple instances of "void startCountdown(final int uniqueId)" that includes "countDown = new CountDownTimer(...)". How can I cancel specific instances of that countdown? Jan 25 12:42:17 p_l: you mean standards like ACPI? Jan 25 12:42:19 on a mobile phone, the kernel needs that *hardcoded* Jan 25 12:42:26 p_l: device tree etc Jan 25 12:43:15 cantido: device tree... assuming the boot firmware tells you anything instead of dumping you into memory and jumping in with bare minimum (fortunately, it tends to give working serial port) Jan 25 12:43:16 device tree has made it into the ARM ports exactly for this reason, one kernel image working on multiple SoC configurations Jan 25 12:43:47 p_l: you can't compare a PC BIOS with a dumb loader Jan 25 12:44:07 cantido: can I ask you to step rehashing what I said trying to make it sound like I'm saying the opposite? Jan 25 12:44:25 because what was my point is that a lot of embedded devices have *dumb loader* Jan 25 12:44:55 so you either need an intermediate step, or hardcoded board config Jan 25 12:45:00 p_l: what embedded devices are we talking about here.. lots of people use uboot Jan 25 12:45:19 Udmurt, I try it. But now I see that json is not null, because the result is like this: {"uTraf":null} Jan 25 12:45:19 this super-h machine does hardware detection pretty well too :p Jan 25 12:45:55 cantido: lots of devices use custom vendor forks of uboot with custom protocols and configurations which don't necessarily leave you with enough data to get everything running Jan 25 12:46:55 Hey guys, I have problem with bluetooth baud rate using rfComm serial protocol. I need to set the baud rate to a specific value but i realized there is no way to do it. Current baud rate is default 9600bps and when i try send data to our custom designed hardware it's taking way too long. When hardware's baud rate is set to 115200 it's again slower then expected Jan 25 12:47:12 p_l: so, that's an issue with the bootloader opposed to some SoC issue? Jan 25 12:47:22 could somebody help in *unlocking* Samsung Galaxy Note 2 (GT-N7108) which is locked with a specific carrier? Jan 25 12:47:30 cantido: can you point where I said it was SoC that were responsible? Jan 25 12:48:00 we referenced "ARM" because typical person thinks "x86 = PC, ARM = their device" Jan 25 12:48:04 To speed it up we set the baud rate on hardware to 2.76Mbps and tried again, result was about 90kbps. failed Jan 25 12:48:13 21:28 < p_l> timroes: embedded devices don't work the same as PCs, there's for example no way to autodetect that many devices are actually tehre Jan 25 12:49:01 cantido: it doesn't say shit about SoC. It refers to whole device, which yes, includes the fw loaded by vendor Jan 25 12:49:38 your PC does just as much detection as an "embedded system" does Jan 25 12:50:57 cantido: at the point of handover from fw to system, I have pretty much most stuff autodetected or provided. I don't have to directly fiddle with low level details to get keyboard working Jan 25 12:50:59 Is there a way to fix it or atleast figure where the bottleneck is? Jan 25 12:51:14 p_l: thats what the bios does Jan 25 12:51:26 clever: "firmware" Jan 25 12:51:34 p_l: it provides a standard api, to call 'drivers' in the bios to access the keyboard, disk drive, and display Jan 25 12:51:44 yes, I know. Jan 25 12:51:47 And I fucking mentioned it Jan 25 12:51:48 which are only ever used to load the real drivers, and by old stuff like dos Jan 25 12:51:57 not paying close attention here Jan 25 12:52:38 p_l: if you attached a keyboard made of bunch of buttons to a PC you would have to do the same stuff Jan 25 12:52:47 p_l: you are comparing apples to oranges Jan 25 12:53:36 and you still have to mess with keymaps and shit on a PC Jan 25 12:54:14 cantido: thats a different layer, the keyboard controller converts each key to a number, which the pc converts back to ascii codes and key codes Jan 25 12:54:27 seperate from just figuring out what button was hit in the first place Jan 25 12:54:28 cantido: you know what? For some reason you decided to flame me about telling someone to not expect bare drivers to get initialized because "this is not a PC and no one needs to give a shit about interoperability unless it's worth something for their own projects", trying to educate me on the same poitn Jan 25 12:54:59 saying that nothing can be detected on "embedded systems" is just tripe Jan 25 12:56:10 clever: The bios keyboard handling stuff only actually works for very limited stuff Jan 25 12:56:14 cantido: next time, I suggest talking Navier-Stokes equations when a kid asks you "why airplanes fly" and expect them to understand Jan 25 12:56:27 p_l: you said, that it won't work if you find linux drivers for that specific device, or at least negated me, when I say, finding working drivers for a specific hardware, won't work on that Jan 25 12:56:42 can I please ask my question again: Jan 25 12:56:43 I have a problem with JSONException: Value null at uTraf JSONObject$1 cannot be converted to JSONArray Jan 25 12:56:43 When I do this Jan 25 12:56:43 JSONObject json = jsonFunctions.getJSONfromURL("http://url.de); Jan 25 12:56:43 An it return null Jan 25 12:56:43 I do this: Jan 25 12:56:44 if (json != null){ Jan 25 12:56:44 But this does not work, cause it returns {"uTraf":null} How can I catch this? Jan 25 12:56:52 but i don't anyhow want to addfuel to the fire here :D Jan 25 12:56:57 timroes: "device driver" in Linux kernel, and the configuration to run it, are separate things Jan 25 12:57:10 no matter how you gain the configuration data Jan 25 12:57:24 timroes: supplying configuration data for stuff that can't be probed is fairly simple though Jan 25 12:57:33 with exceptions for things that are completely hardcoded Jan 25 12:57:41 timroes: and for a lot of things can be done via sysfs at run time Jan 25 12:57:41 okay, sorry then for beeing unclear. With drivers for a specific device I mean, driver with the configuration that is needed to run on that device Jan 25 12:58:21 dot8: can you paste (to pastebin) the whole stacktrace from logcat? Jan 25 12:59:26 and the code snippet with line numbers where the error occurs Jan 25 13:00:43 dot8: use the debugger and step through getJSONfromURL Jan 25 13:01:07 timroes: http://pastebin.com/ygXjQrT1 Jan 25 13:02:22 dot8: that is what you already wrote above, if there is no exception , do as cantido said and debug into getJSONfromURL, where it causes the error Jan 25 13:02:49 or anyone why i s the function trying to convert a null value to an array Jan 25 13:03:04 dot8: I would suggest using GSON over the bare JSON implementation that is Android too Jan 25 13:03:29 could somebody help me with my galaxy note 2. the language is all in chinese and i can't read chinese. i would like to change it to english. if somebody could help me with the menu please Jan 25 13:03:57 expert: there are icons? :) Jan 25 13:04:15 expert: don't know about chinese, but in Japanese language is 言語 Jan 25 13:04:17 ok, thx I try it Jan 25 13:04:32 one of those symbols should match or look similar to something in the menu Jan 25 13:04:55 cantido: i usualy just go by position, i think it was 3rd from the bottom then 2nd from the top in my kindle Jan 25 13:05:05 but that often depends on what the OEM changed Jan 25 13:05:21 is it 'Language' in English menu? I can do a Google translate but just want to make sure it is actually 'Language' text for that icone in Settings Jan 25 13:05:23 best to compare to another matching device Jan 25 13:05:32 expert: don't you have any icons in your menu? Jan 25 13:06:06 expert: on AOSP it's Language & input Jan 25 13:06:10 but that's 4.2 Jan 25 13:06:24 expert: One sec, I have a samsung phone with chinese, I can tell you what it looks like Jan 25 13:06:29 thank you! Jan 25 13:06:38 does samsung remove the icons from the menus? Jan 25 13:06:40 cantido: great! :) Jan 25 13:06:50 timroes there are icons! Jan 25 13:07:03 timroes thanks for your input Jan 25 13:07:03 how can i remove an activity's title in code? Jan 25 13:07:19 i would go for the one with the A in there :) Jan 25 13:07:48 and then the topmost setting Jan 25 13:08:01 expert: it has the 言 symbol in there, but yeah, look for a key with an a on it Jan 25 13:08:12 timroes: wicked! got it! Jan 25 13:08:15 :D Jan 25 13:08:20 cantido, thanks :) i got it! Jan 25 13:08:36 how the hell did you came to the idea to translate the stuff to chinese, if ther eare icons in front of it :D Jan 25 13:08:38 * p_l for a moment, for lulz, switched to jpn Jan 25 13:08:48 now, does anybody know how to unlock this phone? it's locked with a chinese carrier Jan 25 13:09:00 I can just about read enough Chinese to change a phone to English :) Jan 25 13:09:15 p_l: to get the real japanese phone experience you need a vendor theme that is all really shitty icons and lots a purple :( Jan 25 13:09:17 expert: root it, then use some unlocker? Jan 25 13:10:13 or try the link i sent you Jan 25 13:10:16 believe it or not, but this phone comes without Google Play Store Jan 25 13:10:36 expert: can you not flash a local rom? Jan 25 13:10:36 expert: how are you so sure it's not fake? Jan 25 13:10:41 timroes, i tried! thanks, but the menu that guide shows is different to the menu i am getting on this phone! Jan 25 13:11:01 cantido, i would like to.. it's my first time on any android device.. can you help me *how to* ? Jan 25 13:11:06 anyway, carrier locks are a) almost entirely separate from the OS, and b) nothing to do with app development Jan 25 13:11:15 cantido: or THIS: http://www.aniplogs.com/2012/07/install-evangelion-sh-06d-theme-on-your.html Jan 25 13:11:37 Leeds, if you don't like to help, let others! thank you Jan 25 13:11:59 expert: you are explicitly off-topic Jan 25 13:12:12 p_l: that bets the disney phones you can get here by a long way ;) Jan 25 13:12:32 i'm working on something that is going to revolutionize your mom! Jan 25 13:12:40 https://play.google.com/store/apps/details?id=yau.QNBT.NERV&hl=pl <--- a bit better solution, imo Jan 25 13:12:53 mikedg: what? you can make her undead? Jan 25 13:13:00 is it possible to start a timer e.g counting 30 second and then invoke an event even if the user closes the app? I want to make sure that in this example the game can't be tampered. A round is only lasting 30 seconds Jan 25 13:13:07 mikedg: sounds like pr0n ad Jan 25 13:13:17 anyone work for a company doing Android stuff and have a job opening where I could just sit and post crap on google+ all day Jan 25 13:13:32 Daani: alarms can fire after the user moves away from the app Jan 25 13:13:44 p_l: ahhh damn you link with language component in :D Jan 25 13:14:10 do complete ui themes like that work without root? Jan 25 13:14:18 anyone going to be in NY feb 1,2? Jan 25 13:14:19 cantido: ah, alarms.. will look into it. Thanks Jan 25 13:14:27 timroes: sorry, just hit 'y' in my browser :P Jan 25 13:16:55 cantido: does the alarm manager open up the app that created the alarm? Jan 25 13:17:06 Daani: you give it a PendingIntent Jan 25 13:17:35 So you could start up an IntentService or something to do your checks in the background Jan 25 13:18:00 I that sounds nice Jan 25 13:18:42 Daani: could you not work it into the lifecycle of your app though? I.e. when the activity gets stopped it does all the magic you need Jan 25 13:20:56 cantido: I thought of that, but, when the game starts you have 30 seconds to achieve the round, what if you accidently closes the app and starts it right away e.g you have 20 seconds left Jan 25 13:21:00 My app runs multiple instances of "void startCountdown(final int uniqueId)" that includes "countDown = new CountDownTimer(...)". How can I cancel specific instances of that countdown? Jan 25 13:21:25 Daani: recheck the time when the activity resumes Jan 25 13:22:03 cantido: you should not be able to pause the round, that will lead to cheating. It is a quiz kind of app Jan 25 13:22:30 aah, I might check if the user has been "away" for more than the time he had left.. Jan 25 13:22:51 Daani: yeah, save the time when the round starts, and if the activity is paused/resumed recheck the time Jan 25 13:22:56 when he closes the app I store the timestamp and check against that when he opens the app. Jan 25 13:23:04 if the difference is too big end the round Jan 25 13:23:16 thanks, that is an great idea! Jan 25 13:33:18 If I use a ViewPager and I have a fragment with a ListView inside it that contains several ImageViews, why might i get OutOfMemoryException when flinging between fragments? Jan 25 13:33:43 brx_: lots of Bitmaps not getting recycled maybe Jan 25 13:35:45 cantido, do I recycle them in the getView method of my ListAdapter? Jan 25 13:36:16 brx_: I'm not sure what the accepted way of recycling Bitmaps from ImageViews is Jan 25 13:36:29 it seems to work differently on 2.x and 4.x too Jan 25 13:36:46 I am targeting >3.0 here anyway Jan 25 13:37:16 <3.0 version of my app needs a rewrite since there are some api calls not available which i need in the <3.0 version Jan 25 13:38:55 21:28:28 < g00s> lov said their code was like sausage Jan 25 13:39:04 fwiw I think that's every OEM. Jan 25 13:41:55 brx_: you have a separate version for older releases? Jan 25 13:43:06 cantido, in this particular app yes, I couldnt find a better workaround for the methods which are needed Jan 25 13:43:26 I need getPrefferedViewPreview(), it isnt available is there any way to change the name my device reports in adb? Now it's "012fdsf4asdf" and i want it to say like "nexus7" Jan 25 13:45:17 That's the device serial number, so no. Jan 25 13:45:25 The problem is that you may have several devices of the same type plugged in Jan 25 13:46:36 but if i could specify an alias or stomething Jan 25 13:47:53 cantido, here is my ListAdapter http://pastebin.com/5UjspzhZ Jan 25 13:49:50 brx_: you're recycling view correctly.. I would use the heap dump tool to work out what is using all of your memory Jan 25 13:50:30 hmm tbh ive never used that before (yes im still a bit of a noob) Jan 25 13:50:49 something new to learn I guess :) Jan 25 14:10:34 hi Jan 25 14:11:29 is it possible to do an app that reads nfc tags sent by another? (from the same device) Jan 25 14:12:44 probably not Jan 25 14:13:00 not sure what the use case is there Jan 25 14:13:28 a background service called when a nfc tag arrives Jan 25 14:13:34 and foreground app sending tags Jan 25 14:13:44 in the air the tags return to the phone Jan 25 14:13:49 right, but why would you want to do that Jan 25 14:13:49 isn't it? Jan 25 14:14:02 is a proof of concept Jan 25 14:14:11 sha0coder: no, the NFC can either transmit or receive, it can't do both at the same time. Jan 25 14:14:13 about security on cards Jan 25 14:14:29 sha0coder: bidirectional transfer is handled by negotiating which is doing which Jan 25 14:15:57 oks thanks Jan 25 14:24:10 hello.. I'm still looking for a solution how to combine sliding-menu, and viewpageindicator with fragments. http://stackoverflow.com/questions/14259170/no-activity-exception-when-using-viewpager-and-fragments-inside-fragments Jan 25 14:24:31 someone suggested me to use activities and transitions for sliding menu work. but latest support library supports fragments inside fragments so i would relay like to implement them this way Jan 25 14:24:58 My app runs multiple instances of "void startCountdown(final int uniqueId)" that includes "countDown = new CountDownTimer(...)". How can I cancel specific instances of that countdown? Jan 25 14:26:11 Are the any docs on what android's libc supports? I want to know if it can do RPC Jan 25 14:26:13 I'm not sure how to cancel the right countDown Jan 25 14:26:36 hey zipoboy Jan 25 14:26:44 hi Jan 25 14:26:52 im the person that replied to that stackoverflow question Jan 25 14:26:58 ql:D Jan 25 14:27:17 cantido: no IPC routines in Bionic, afaik Jan 25 14:27:24 NDK download has details Jan 25 14:28:02 i check and I'm using ChildFragmentManager as you can see on git Jan 25 14:28:17 p_l: I want Sun RPC ;) Jan 25 14:31:04 cantido: that is too missing, I think Jan 25 14:31:17 cantido: thanks for your ideas and for pointing me towards right direction.. time to sleep some!! Jan 25 14:31:19 bye Jan 25 14:31:27 Daani: no problem :) Jan 25 14:31:34 I have a question regarding using root commands, if anyone is up for it Jan 25 14:31:44 notbryant: don't ask to ask, just ask. Jan 25 14:31:46 p_l: Yes, I guessed as much, the right headers are in the source but I doubt it actually works Jan 25 14:31:50 And if no one answers here, #android-root Jan 25 14:31:56 (but #android-root is kind of a cesspool sometimes) Jan 25 14:32:13 it's all the root going to their heads Jan 25 14:32:48 #android-dev is kind of a cesspool sometimes Jan 25 14:33:02 welcome to the interwebs Jan 25 14:33:04 ok so is there a way to reliably launch a Process with su and then execute individual commands and wait for them to finish? Jan 25 14:33:27 notbryant: start a shell and talk to it by pipe Jan 25 14:34:11 I know how to do that on Linux, not sure how to do it on Android Jan 25 14:34:58 notbryant: equivalent of popen("su -c /system/bin/sh") Jan 25 14:35:45 except using Java apis for that Jan 25 14:36:19 I tried that, but the problem is that there's no way of waiting for the command to complete and retrieve the data Jan 25 14:36:52 in my eclipse IDE, the inteli typing stuff for android stuff does not seem to be working. anything I can do about that? Jan 25 14:39:44 Hi! Is it possible to guess why my Application descendant instantiated? (I mean to launch main activity or for broadcast receiver) Jan 25 14:39:48 irishgringo: not really sure what you mean. Ctrl-space? Jan 25 14:39:56 Udmurt: huh? Jan 25 14:40:42 Udmurt: you'll have to wait until the actual code (activity/receiver) is called Jan 25 14:40:56 oh Jan 25 14:40:57 Heh, found the way to do ) Jan 25 14:41:01 yeah, what Ge0rG said. Jan 25 14:41:09 oh? I'd love to hear about it. Jan 25 14:41:35 not with application… in other classes Jan 25 14:45:41 hello, does any1 know a channel for discussing android kernel hacking? Jan 25 14:46:04 root Jan 25 14:46:18 there's #android-platform but it's mostly dead. You can try #android-root. Technically, here works, but you won't get much response. Jan 25 14:46:35 Root? Modded ROMs? Head to #android-root Jan 25 14:46:43 topic at #android Jan 25 14:47:51 And another question. I have an Authenticator and service for it and want to show Toast in my Authenticator, Toast is not showing (from Authenticator service it works fine). How to solve it? Jan 25 14:48:23 yeah thx, im already at android-root but i thought there were (another) for discussing specifically kernel development Jan 25 14:52:48 iriki: what stuff do you want to do in the kernel? Jan 25 14:54:00 cantido thx. well, first im troubleshooting why wifi wont work with the kernel sources i have. and secondly, im trying to make this RK3066 tablet charge when in OTG-USB mode. Jan 25 14:54:24 the second issue i have some ideas to try first. but the wifi problem is more urgent, i guess Jan 25 14:54:29 iriki: I would guess the wifi is some "binary blob" issue Jan 25 14:54:50 i wanna pop kernels Jan 25 14:55:10 perhaps. check this out: 255|root@android:/system/lib/modules # ll Jan 25 14:55:23 (...) -rwxrwxrwx root root 535790 2012-12-01 07:07 rkwifi.ko (..) Jan 25 14:55:28 iriki: for the charger.. you'll need to find out what the charger chip etc are and if you can fiddle any registers I guess Jan 25 14:55:35 but, when i try to insmod... root@android:/system/lib/modules # insmod rkwifi.ko Jan 25 14:55:47 insmod: init_module 'rkwifi.ko' failed (No such file or directory) Jan 25 14:56:27 cantido: yes, ive done some searching, tho i need to solve this wifi first before fiddling and studying the otg charge thereafter Jan 25 14:56:53 iriki: I think thats going to be firmware loading.. do you have dmesg? Jan 25 14:56:59 how do I find what's the power chip? Jan 25 14:58:10 iriki: you should take a look through the board file for that phone to see if you can find out what stuff is wired up Jan 25 14:58:42 I guess there is some usb charger chip in there that you can talk to over i2c Jan 25 14:59:27 I just added a PreferenceActivity for my app, but what is the "correct way" to have the settings button launch that activity? Jan 25 15:00:43 iriki: sysfs and debugfs will help too Jan 25 15:00:50 here is pastebin: pastebin.com/B9yLwrJM Jan 25 15:03:55 I mean, since it's a built-in Settings button in the Menu, I can't use onclick like I normally would Jan 25 15:04:07 will study how to use then, then. thanks for the pointers Jan 25 15:06:05 Quacked, PreferenceActivity is ordinary activity, so launch it as normal activity Jan 25 15:06:15 iriki: the otg charging thing might be in hardware though, so you might not actually be able to change that Jan 25 15:06:49 Quacked, what is built-in Settings button? Jan 25 15:06:58 cantido: yes i know, i've been reading that. but still I have to try Jan 25 15:10:01 Any of you guys develop with Sublime Text 2 and tried this plugin http://gitorious.org/sublime-android ? Jan 25 15:11:06 sublime is sublime Jan 25 15:11:49 Udmurt: I mean the option menu that has the Settings-button option Jan 25 15:12:34 mikedg: you using it? Jan 25 15:12:38 Quacked, in your activity/fragment override method public boolean onOptionsItemSelected(MenuItem item) Jan 25 15:12:44 i do use it for heavy text editing Jan 25 15:12:49 but mostly as a temporary work space Jan 25 15:12:51 also file searching Jan 25 15:12:56 thanks, Udmurt Jan 25 15:13:16 mikedg: regular expression building with it is pretty awesom Jan 25 15:13:19 aweoms Jan 25 15:13:33 or however you spell that word! Jan 25 15:13:48 try awesome =) Jan 25 15:14:12 I'd like to try and get away from Eclipse Jan 25 15:14:16 it's so heavy and bulky Jan 25 15:14:29 i wouldn't care so much about that Jan 25 15:14:37 if it didn't lose all its settings and installed software every month Jan 25 15:14:41 (because of a new version..) Jan 25 15:15:28 When I use getApplicationContext(), can I use this context object safely allover the lifetime of my application? Jan 25 15:15:48 I know that getApplicationContext() should be avoided if possible but I have a special use case Jan 25 15:16:18 well Jan 25 15:16:21 "yes" Jan 25 15:16:42 but caching the application context is something you don't want to get into the habit of doing Jan 25 15:17:06 what is wrong with getApplicationContext()? Jan 25 15:17:48 this is why i don't install most of the apps on play :P Jan 25 15:18:18 fusion27: once your application grows larger than 100 classes, you need the enterprisy abilities of eclipse. Jan 25 15:19:30 fusion27: the refactoring capabilities of eclipse are really worth all the "trouble" Jan 25 15:19:36 or IntelliJ idea, community edition work great for me, including maven Jan 25 15:22:27 * cantido wishes the CDT stuff worked as well as the Java stuff Jan 25 15:27:33 does anyone by chance know if there is a limit in the length an url can have in an HttpUrlConnection? Jan 25 15:29:17 If you have to find out, you should reconsider your webservice. Jan 25 15:30:09 lov: yeah i reconsidered it already several times Jan 25 15:30:32 retrieving objects from a server, the request can contain a query Jan 25 15:30:41 there is no limit in the length the query could have Jan 25 15:30:42 ... ugh Jan 25 15:31:06 timroes: consider using POSTs in this case. Jan 25 15:31:12 well Jan 25 15:31:14 could put it in the get body, isn't really in the http specs and i didn't found any JAX-RS implementation, that could handle that Jan 25 15:31:15 I guess it depends on what you're doing Jan 25 15:31:49 POST could handle it, but would break the REST paradigm, and i would need to additionally code what i am trying to do (editing/ getting) in the message, since its not coded by the http method anymore Jan 25 15:32:39 inventing an own http method (e.g. SEARCH) would be able in jersey, but you cannot add cusotm methods to HttpUrlConnection easily. HttpClient would do, but isn't recommended anymroe since android 2.3. Also there were deep package inspection issues, regarding custom methods Jan 25 15:33:12 yeah don't invent your own method Jan 25 15:33:21 ok, lets step back Jan 25 15:33:26 what, exactly, are you doing? Jan 25 15:33:31 there were some other possibilities, but we eventually we decided do code the query in the url, what get is actually made for :) Jan 25 15:33:47 the only problem in that case is, that there might (or might not) be a limit to url length Jan 25 15:34:07 (but in fact we could split up queries so it's not that bad) Jan 25 15:34:27 timroes: I'm getting a little concerned that you're writing GET requests about "queries" that are this length. Jan 25 15:34:27 transparent synchronisation of java object on a server Jan 25 15:34:34 ... Jan 25 15:34:52 use POST. You're changing states. Jan 25 15:34:56 there is also only one "use-case" that could get in danger to the length :D Jan 25 15:35:14 don't use SQL for the query either. Jan 25 15:35:15 sounds like mobile code :> Jan 25 15:35:34 and that is, since i collect multiple requests, to not need to transfer data so often, and I might collect a high numbers of object ids, I want to get from the server Jan 25 15:35:42 (you can do all sorts of funny things to the server even with only read access) Jan 25 15:35:49 zinx: ofc i dont use sql :) Jan 25 15:36:01 timroes: don't do that. Jan 25 15:36:03 well Jan 25 15:36:04 lov: where am i changing state? Jan 25 15:36:10 I guess do that but sanity check your requests. Jan 25 15:36:18 break your request into multiple requests Jan 25 15:36:27 yeah thats what i said, :) Jan 25 15:36:42 why this isn't actually a too big problem, i can split them up into several if these kind of requests get to long Jan 25 15:36:53 which frankly you ought to be doing anyway Jan 25 15:37:04 You've got not guarantee that you won't make a request, say, 1M chars long Jan 25 15:37:13 that will, likely, cause problems locally Jan 25 15:37:18 even if it doesn't cause them remotely Jan 25 15:37:37 thats why i was interested in the maximum length of urls android can handle with httpurlconnection Jan 25 15:37:50 ive a question about deviceadmin.. currently i start the deviceadmin out of an activity to force lock of the device. Is it possible to set up a deviceadmin on STARTUP of a SERVICE which is available over all other Intents/Classes? Jan 25 15:43:33 in order to write apps for Android do I have to learn Java or HTML5 ? Jan 25 15:43:43 b0ss_: yes :) Jan 25 15:43:56 there are also some other possibilities like C# via Mono for Android Jan 25 15:44:23 or writing parts of your app in C/C++, but java is the main language for android, and you can create webapps with html5 and js Jan 25 15:45:30 A question regarding service IPC and blocking: When binding to a service, is the onConnected callback always called inside the main thread? I'm wondering how I could prevent an ANR if a service RPC call blocks my thread for let's say 20 seconds. How to execute the RPC then in another thread than the main thread? Jan 25 15:46:14 Or is the only way then using asynchronous communication? Jan 25 15:46:31 What if I want the IPC to happen synchronously? Jan 25 15:49:58 timroes: hmm. I'm aware HTML5 has limitations concerning filesystem resources... Am I right ? Jan 25 15:50:06 not only that Jan 25 15:50:16 with html5 you can make mainly very basic apps Jan 25 15:50:30 but if it come to advanced topics, you can hardly reach it with html Jan 25 15:50:42 i mean most you get out of frameworks like phonegab is camera support or such Jan 25 15:51:24 Right. I'm not aware of phonegab Jan 25 15:51:44 "phonegap" Jan 25 15:51:57 isn't it anyway renamed completely? :) Jan 25 15:52:09 I didn't even know there were frameworks for HTML5. Are they MVC ? Jan 25 15:52:09 or should it? Jan 25 15:53:17 So Android is the JDK running on Linux based devices, right ? Jan 25 15:53:21 ah no i guess i am mixing it up with something Jan 25 15:53:27 b0ss_: no :D Jan 25 15:53:38 android is the operating system/framework running on several devices Jan 25 15:53:53 ah ok Jan 25 15:53:54 dalvikvm is the JVM used inside android to run the java stuff Jan 25 15:54:06 not a JVM... Jan 25 15:54:14 -j Jan 25 15:54:32 not a JVM? What then ? Jan 25 15:54:37 a VM Jan 25 15:54:54 a jvm is running bytecode, dalvik vm runs dexcode Jan 25 15:54:54 not all VMs are JVMs Jan 25 15:55:03 Leeds: is it related to that Oracle-Google court case ? Jan 25 15:55:22 nope, dalvik wasn't a JVM before that Jan 25 15:58:21 hey guys, i got problem with bluetooth baud rate does anyone know what maximum baud rate is ? using samsung galaxy w i8150 - bluetooth v3.0 with A2DP Jan 25 15:58:36 potentially flame generating question: what's more powerfull and expressive Android or JME ? Jan 25 15:58:46 Android. Jan 25 15:58:58 If by JME you mean stuff like J2ME then it's a lollerific comparison Jan 25 15:59:05 JME is AWFUL Jan 25 15:59:21 i wouldn't even come to the idea to compare android to jme? Jan 25 15:59:23 its like.. Jan 25 15:59:25 it had an absolutely tiny subset of the classes available on desktop java, and did everything poorly Jan 25 15:59:38 what is more powerfull, a horse or a car Jan 25 15:59:56 Well JME was also big when phones were phones, not dual-core 1ghz computers Jan 25 16:00:09 horses were also big in middeavel ages? Jan 25 16:00:18 noone used cars that time, believe me Jan 25 16:00:22 timroes: so were JME phones Jan 25 16:00:27 Yes, I'm seconding your point. Jan 25 16:01:02 JME made sense when you had an 8088 in your phone with 4KB memory. Jan 25 16:01:10 i mean i don't have anything against JME, but its just not a comparission that makes anyhow really sense Jan 25 16:01:13 We have moved on in the last decade and it no longer makes sense. Jan 25 16:04:05 yeah.. poor old horses .. and jme devices Jan 25 16:05:25 plus #jme is a bit empty :) Jan 25 16:06:34 question: Android seems to overcome iOS in midterm. Right ? Jan 25 16:06:47 b0ss_: can't say that easily Jan 25 16:07:11 depends if you are meaning absolute numbres, and even there I guess noone can predict the future in here. Jan 25 16:09:10 How can I get those 3 settings dots in top right corner even if my phone has a settings button? Jan 25 16:09:29 my personal opinion 'bout that: they aim at a compleltey different audience. you will android phones from 100$ on. That are other ppl buying these, that buying iPhones. But thats something also apple recognized, why they plan to build cheap leightweight iPhones. Jan 25 16:09:45 Quacked: www.actionbarsherlock.com Jan 25 16:11:11 doesnt look like its working here Jan 25 16:11:26 You can't Jan 25 16:11:28 Quacked: What version of Android is your phone running? Jan 25 16:11:30 for me neither Jan 25 16:11:55 Quacked: then head over here https://github.com/JakeWharton/ActionBarSherlock Jan 25 16:11:59 well, my phone is 4.2 Jan 25 16:12:02 is it a bug (of my app) or is it an android feature that camera.takePicture() is not availalbe while the screenshot is locked? Jan 25 16:12:03 Actionbar Sherlock uses fallback to the native implementation if one is available Jan 25 16:12:10 So that might be why Jan 25 16:12:11 I see Jan 25 16:12:35 oh you are using already an post 3.0 phone, then your manifest is just wrong :) Jan 25 16:12:38 Won't bother with that Jan 25 16:12:53 but what should your minimum supported android version be? Jan 25 16:13:05 me? I'm thinking 2.3 Jan 25 16:13:15 then you will need that suport lib to use the actionbar Jan 25 16:13:23 I see Jan 25 16:13:25 because it was introduced with android 3.0 Jan 25 16:13:27 I'll skip it :) Jan 25 16:13:29 misterli: I think that requires root access Jan 25 16:13:43 Quacked: you don't want to skip actionbar in your app :) Jan 25 16:13:57 my app consists of 16 large buttons Jan 25 16:14:21 and basically 2 activities :P Jan 25 16:14:27 Mhm, okay. But there are several anti thief apps which shoots a pic when the keylock was not successfully disabled. (I guess they use DeviceAdmin and onPasswordFailed and shoot a picture this way. But shooting does not work -_- Jan 25 16:15:43 misterli: Depending on how used you are to application decompiling, you could either look at their implementation or just choose the easy way out and e-mail the developer about it :) Jan 25 16:17:09 okay. i dont want to use the DeviceManager to unlock -> take picture -> lock again. It's a securitypurpose :| Jan 25 16:17:09 My MainActivity is running a bunch of CountDownTimer objects (countDown = new CountDownTimer(numberOfMinutesInSec *1000, 1000)). I know countDown.cancel() will stop the countDown, but how can I specify which of the countDowns? Jan 25 16:19:18 hey, can't access the back byte array of a ByteBuffer I have in my code, I get an UnsupportedOperationException when I call .array() method, and .hasArray() return false. How can I get that byte array still? Jan 25 16:19:46 Can I use CountDownTimer like that or must I go over to AlarmManager Jan 25 16:23:22 Quacked: I think you're confused about how java works./ Jan 25 16:23:40 Quacked: if you have a CountDownTimer in countDown, and then assign countDown to a new CountDownTimer, you don't have a reference to the first timer instance Jan 25 16:23:46 (unless you stored a reference somewhere else) Jan 25 16:23:59 If you're going to have multiple timers, I'd strongly recommend using something like an ArrayList. Jan 25 16:24:04 so who is countdowntimer? Jan 25 16:24:21 use a Vector, you don't know what you are doing, so it's safer ot be threadsafe Jan 25 16:24:42 lov, good, I was thinking I might need something like ArrayList Jan 25 16:24:43 thanks Jan 25 16:24:44 I've reading about Android vs JME and people agree the updates of JME are not that frequent. Otherwise Android is frequently updated. Right ? Jan 25 16:24:46 * Ge0rG would vote for AlarmManager nevertheless Jan 25 16:24:49 when in doubt, throw it out Jan 25 16:24:52 you never know when the cpu falls asleep Jan 25 16:25:25 b0ss_: I've been reading some random blogs which say that JME is due for a resurgence this year as part of BB11 Jan 25 16:25:46 I like my CountDownTimer, hehe Jan 25 16:25:46 BB11? Jan 25 16:26:03 one better than BB10 Jan 25 16:26:25 lov: fyi, app crashed if you try to send urls larger than around 1MB, and from around 70kb you notice that the garbage colelctor has to run several times during one request send :) Jan 25 16:26:32 ha Jan 25 16:26:36 yeah, there's your answer :> Jan 25 16:26:43 http://www.dailytech.com/Android+App+Take+Pictures+of+Phone+Thieves+From+Unlock+Screen/article23122.htm hmm... it seems that it must be possible without root Jan 25 16:26:48 Leeds: it's BB10 but with the 10 crossed out and 11 written in Jan 25 16:26:51 but i bet it has smthg todo with the preview Jan 25 16:27:04 because you're jamming out with your cellphone, but then you want to go 1 higher Jan 25 16:27:05 tested on android 2.2 :) (because I would assume the limit would increase with increasing android version) Jan 25 16:27:07 so you turn it up to 11 Jan 25 16:27:56 misterli: sure it is. There are broadcast events for the screen turning on. Jan 25 16:28:16 lov: the BBC's online TV service has a volume slider which goes to 11 :) Jan 25 16:28:46 i know. but i dont want the screen to turned on (Anti-Thief!). it better must be shot while the keyguard is activated Jan 25 16:32:15 can a CountDownTimer set at 60 minutes be interrupted if I lock my phone? Jan 25 16:32:22 Quacked: yes Jan 25 16:32:29 not interrupted but delayed Jan 25 16:34:15 * Quacked facepalms Jan 25 16:34:36 What's the updating model of Android? Is it open source and there is a JCP for it like JME or is it all about the Google's engineers in a close room ? Jan 25 16:34:40 Quacked: Not if you're *cough* lucky to have an installed app which keeps the cpu awake Jan 25 16:34:51 * Quacked looks up AlarmManager again Jan 25 16:35:22 b0ss_: are you writing an article, doing homework, or building an app? Jan 25 16:36:39 Leeds: none of them Jan 25 16:36:45 just getting to know Android Jan 25 16:36:50 b0ss_: it is open source project. anyone can commit. there is a regular review process that any commit has to go through http://source.android.com/source/life-of-a-patch.html Jan 25 16:36:57 anyway of internal or external commit Jan 25 16:37:21 there are some google apps, that are not open source. Jan 25 16:37:42 timroes: except that realistically, the vast majority of work is done behind closed doors at Google Jan 25 16:37:44 and also afaik new android versions are developed inside google, so device manufacturar get early preview to it, before developers get Jan 25 16:37:54 Leeds: I would assume so Jan 25 16:38:01 b0ss_: this is an app development channel, not general chat - #android might be a better place for your questions Jan 25 16:38:20 and i get the feeling if you work at google your chances are way higher, someone will review your code within a view days :) Jan 25 16:38:38 at least still waiting for some reviews for my commits :D Jan 25 16:38:42 nah, you just have to get reddit to spam the bug report tool and gerrit Jan 25 16:38:50 b0ss_: Note that builds on devices aren't necessarily based on AOSP, nor is there a guarantee that they're going to get a particular update. That's based on the manufacturer and the carrier. Jan 25 16:38:55 then you get a good chance of being reviewed Jan 25 16:39:20 mikedg: its luckily nothing really important atm, just some documentation and some bug fi in the build process Jan 25 16:39:25 so I won't spam anyone for that :D Jan 25 16:40:23 ok Jan 25 16:40:49 where does android store the version number of app's SQLite databases? Jan 25 16:40:56 so Google plays very much the same role as Oracle in the evolution of Java platform ? Jan 25 16:41:00 in the database Jan 25 16:41:01 Ge0rG: I believe within the database itself Jan 25 16:41:10 b0ss_: in a sense. Jan 25 16:41:28 google is the architect not the oracle Jan 25 16:41:31 get your shit straight Jan 25 16:41:44 and mikedg is neo Jan 25 16:41:51 lov: ie, open to everyone commit but the final word belongs to Google. Jan 25 16:41:54 actually you're more like cypher... Jan 25 16:41:59 yeah Jan 25 16:42:01 lov: but there is no sqlite3 command to fetch it? Jan 25 16:42:12 Ge0rG: it's probably stored in a metadata table Jan 25 16:42:12 it is a pain every time I up/downgrade the app while dev'ing Jan 25 16:42:20 'kay :) Jan 25 16:42:53 er Jan 25 16:43:01 It's almost certainly in some metadata table Jan 25 16:43:08 b0ss_: that's how AOSP works. Jan 25 16:43:25 AOSP stands for Android .... ? Jan 25 16:43:30 open source project Jan 25 16:43:30 open source project Jan 25 16:43:35 head to source.android.com Jan 25 16:43:38 it might be more helpful Jan 25 16:43:40 also, d.android.com Jan 25 16:43:54 except it's actually more than that... because the tree is developed internally at Google and only published when there's a new release, you can't know if your patch will apply to the current internal tree Jan 25 16:43:58 AOSP YOLO Jan 25 16:44:10 ... Jan 25 16:44:27 why there should be always a pragmatic purpouse on learning something? Curiosity and prospection aren't just good enough reasons ? Just saying... Jan 25 16:44:30 or if you're trying to fix something which is already fixed Jan 25 16:45:06 b0ss_: again, this is not a channel for general chat about Android Jan 25 16:45:27 well, people at #android sent me here Jan 25 16:45:59 for app dev questions, yes - not for generic questions about the platform Jan 25 16:48:48 b0ss_: head to #android-root and see where they send you :) Jan 25 16:49:03 then its 1 vs 2 and the losing channel must help you :) Jan 25 16:49:16 timroes: straight to hell Jan 25 16:49:25 * Pattehh feels sorry for the losing channel Jan 25 16:49:30 hmm that would be bad, then we still have a tie :D Jan 25 16:55:55 hmm since there are already several ppl here in common with the review process. If i commited a bugfix for an open bug, should i somehow comment on that in the bugtracker? Jan 25 16:57:16 close it Jan 25 16:57:32 i think i don't have the right to close bugs in the bug tracker :D Jan 25 16:57:55 i think googlers can o that Jan 25 16:57:58 you should try to get a job Jan 25 16:58:02 i have a job :D Jan 25 16:58:05 but not at google Jan 25 16:58:17 bit too far away from here :) Jan 25 16:58:18 timroes: My guess is that you reference the bug in your commit and then the Google guys do the job for you Jan 25 16:59:04 cantido, I fixed the outofmemoryerror i was getting earlier (with images in the listview within the view pager) Jan 25 16:59:16 Pattehh: just thought, noone is going through the bug tracker and think oh a nice bug, i take this, and then it has already been fixed Jan 25 16:59:18 even when i try to unlock the screen with KeyGuard and shot the picture immediately after that it wont work.. just if the screen is completly unlocked it works -_- annoying Jan 25 16:59:40 timroes: mostly, noone goes through the bug tracker... except canadiancow when he's bored Jan 25 16:59:56 misterli: Blank image or an Exception? Jan 25 17:00:00 i followed this link http://developer.android.com/training/displaying-bitmaps/load-bitmap.html but I hope it isnt a hack and the problem really lies somewhere in my adapters/releasing of objects Jan 25 17:00:04 is he often bored ;) Jan 25 17:00:22 Nothing of Both. It just dont use the Callback of takePicture() Jan 25 17:00:24 well, the wide open fields of Canadia don't provide much stimulation, even for a bovine Jan 25 17:01:00 timroes: Good point Jan 25 17:01:17 misterli: Ah, so it is a thread-safe call? Jan 25 17:01:27 Non-blocking** Jan 25 17:02:05 Yes, it is Jan 25 17:02:46 well, im playing around with newKeyguardLock and the return handler. It _must_ work to shot the picture straight after unlock is thrown and enable it back? Jan 25 17:02:47 Any blocking call available? Otherwise I would look into DeviceAdmin or some other call Jan 25 17:03:46 misterli: newKeyguardLock could be something Jan 25 17:04:26 well, its deprecated but i dont know an alternative method for that Jan 25 17:38:01 My client wants a ViewPager of 6 or so images, but we want them to fill the entire activity. How can I know all the different image resolutions / how to decide which to use at runtime? Jan 25 17:39:12 heh, anyone use com.github.USER as a package? Jan 25 17:39:20 chrisbanes does Jan 25 17:39:28 I just imported his com.github.chrisbanes.photoview lib Jan 25 17:39:37 hm, so not entirely unprecedented Jan 25 17:40:14 achuinard: You could try to use some resizing option and providing a multiple of different resources in the correct drawable folder Jan 25 17:40:35 Right ok, so using drawable, drawable-hdpi, etc is the way to go. Jan 25 17:40:52 How could I be sure to cover most cases though? I can pass off the requirements to the designer. Jan 25 17:41:23 But I need to be like 'drawable-hdpi dir with 800x480s, drawable-mdpi dir with 480x320s, drawable-land with 320x480s', etc... Jan 25 17:41:23 I think you need to use the screensize "attribute" instead of the density one Jan 25 17:41:44 Since a tablet at 7" and 1280x720 is hdpi Jan 25 17:42:06 I'm confused between the attribute vs density Jan 25 17:42:14 instead of hdpi Jan 25 17:42:15 xhdpi Jan 25 17:42:17 you mean like Jan 25 17:42:31 small, normal, large, xlarge Jan 25 17:43:16 if I create big notifications with buttons for 4.1, what happens on a 2.3 phone? Jan 25 17:43:42 Quacked: Not sure, but compatibility library lets you specify action buttons for notifs but degrades gracefully Jan 25 17:44:18 hmm Jan 25 17:46:34 achuinard: I would take a look at the table below "How to Test Your Application on Multiple Screens" http://developer.android.com/guide/practices/screens_support.html#testing Jan 25 17:47:29 Quacked: Check the API version before adding those buttons, if it's lower than 4.1 don't add those calls. I'm quite sure you would crash the app on other devices if you choose to not do that Jan 25 17:52:53 Quacked: You can use this to test if you're running on the correct platform: if (android.os.Build.VERSION.SDK_INT >= android.os.Build.VERSION_CODES.JELLY_BEAN) { } Jan 25 17:54:26 isnt it possible to something like if (buildIsHigherThan15) { addBigButtons(); } ? Jan 25 17:55:19 Quacked: That's exactly what that does ;) Jan 25 17:56:11 so I can use that to have big buttons only for newer Android versions? Jan 25 17:57:21 Probably, yes. But if it's some layout thing I would use the folders layout/ and layout-v16/ instead of doing some code solution Jan 25 17:57:39 cool Jan 25 17:57:48 I guess we would have a bunch of different releases on Google Play for different version if this was not possible :P Jan 25 18:00:21 Uh Jan 25 18:00:30 Yeah, generally you want to have one apk when possible Jan 25 18:00:41 the resourcing system and customization is really awesome; strongly consider looking into that Jan 25 18:00:47 use layout-v16 or something like Pattehh said Jan 25 18:03:38 hi... i need to change layout after deleting some data... my problem is that after deleting i manage to put refreshed new data but below(after) old list... how to erase all layout and just put new? Jan 25 18:06:29 nicole24: Supposed you're putting it all in a ViewGroup you can use viewgroup.removeAllViews(); Jan 25 18:07:09 anyone remember who i was talking to about widgets yesterday? Jan 25 18:08:19 Pattehh in these activity i am using -> this.setListAdapter(new ArrayAdapter(this, android.R.layout.simple_list_item_1, results)); Jan 25 18:09:06 and it just keeps adding at bottom current new results... i tried few things but nothing works... Jan 25 18:09:31 Did you try removing the stuff you have in results? Jan 25 18:10:09 yes... and that works... i have only new data.. so logic is good... but this old part of layout remains... Jan 25 18:10:20 nicole24: Oh, well then you should keep a reference to the original array and the adapter, modify this array when data is updated, and then call adapter.notifyDataSetChanged() Jan 25 18:13:59 When you are dealing with various physical screen sizes with similar resolutions, how would you, say, change the text size apporpriately, using dp and sp dont' create the desire effect. Clearly you want a larger text size for the larger screen (inches). Jan 25 18:14:26 Not necessarily Jan 25 18:14:30 Clearly? Doesn't seem so clear to me Jan 25 18:14:45 sp i think is a resolution independent unit Jan 25 18:14:55 I meant "I want", maybe not clear, fair enough. Jan 25 18:14:55 so your text has a consistent physical size Jan 25 18:15:02 as in inches Jan 25 18:15:05 if that's what you want, then just put custom values-large etc... Jan 25 18:15:25 The benefit of a larger screen isn't bigger text, it's more content Jan 25 18:15:40 is values-large the same thing as values-hdpi, values-xhdpi, etc.? Jan 25 18:15:51 except its for screen size Jan 25 18:16:03 but don't do this, it's stupid Jan 25 18:16:06 so they aren't the same thing, large is for larger physical sizes? Jan 25 18:16:19 yes Jan 25 18:16:27 http://developer.android.com/guide/topics/resources/providing-resources.html#AlternativeResources Jan 25 18:17:02 ^ will read. Thanks for the help mikedg and SimonVT Jan 25 18:17:07 no Jan 25 18:17:09 thank you! Jan 25 18:17:12 jdawes: agree with the principle -- I think the text size should change based on expected distance from the viewer Jan 25 18:17:49 jdawes: there isn't any good way to do that from android -- the closest I do, is to base all measurements off a "unit" that I define in values/dimens.xml Jan 25 18:18:14 then, I alter it based on screen width -- which serves as a rough proxy for the screen distance, though not very well Jan 25 18:19:10 also, using pts rather than dp works better to avoid internal inconsistencies that android introduces in actual length, even on devices with identical physical dimensions Jan 25 18:19:30 ugh Jan 25 18:19:32 Understood. I'll give both methods a try I think. Jan 25 18:19:41 i need to stop reading what people write in this channel Jan 25 18:20:14 zinx: pretty much. Jan 25 18:20:29 the truth is hard to read sometimes :-) Jan 25 18:20:49 The wrong is hard to read Jan 25 18:21:01 feel free to enlighten me ;-) Jan 25 18:21:16 The benefit of a larger screen isn't bigger text, it's more content Jan 25 18:22:05 makes sense, if it's held at the same distance as a smaller device; did I say something that says otherwise? Jan 25 18:25:03 For the application that I'm creating I want the same content but with bigger text. It's got nothing to do with displaying more content. I agree with you, though. I just don't have any more content for the particular screen and the text is smaller than what would be comfortable to a user. Jan 25 18:25:36 It pretty much is tho Jan 25 18:25:45 The benefit? Jan 25 18:25:46 Yes. Jan 25 18:26:27 render to a 320x240 bitmap, then put it in an imageview with scale fitxy Jan 25 18:26:37 the viewing distance Jan 25 18:27:14 best of both worlds Jan 25 18:27:31 i'm so myley Jan 25 18:28:42 anyone know how to setup intellij to let me step through the platform source when debugging Jan 25 18:30:21 hey mikey Jan 25 18:30:25 :) Jan 25 18:30:55 Open any framework class, click the "attach sources" button Jan 25 18:31:00 hi Jan 25 18:31:04 Or, I believe it's somewhere in module settings Jan 25 18:31:09 i saw the chat hear recently about intellij.. is it any good, last time i saw anyone speak about it.. it wasnt up to scratch? Jan 25 18:31:26 *here Jan 25 18:31:27 not seeing attach sources, ill check module settings Jan 25 18:31:35 i bought it cause it was cheap Jan 25 18:31:41 it still annoys the shit out of me Jan 25 18:31:51 not sure if its less than eclipse annoys the shit out of me though Jan 25 18:32:12 Napalm: That must have been years ago :p Jan 25 18:32:23 quite possibly Jan 25 18:32:34 Otherwise I'm sure "not up to scratch" meant "no ui designer" Jan 25 18:32:56 (which it has now.. so yeah..) Jan 25 18:33:21 (protip: it can be disabled under plugins) Jan 25 18:33:22 i never use the UI designer anyway Jan 25 18:33:23 lol Jan 25 18:36:25 oh the ui designer exists Jan 25 18:36:31 its terrible from a usability perspective thoguh Jan 25 18:37:26 Having used neither ui designed, I just assume everything it creates is horrible Jan 25 18:37:35 s/designed/designer/ Jan 25 18:39:07 [to flog a dead horse, since I didn't actually see an answer I could see :-) jdawes original question - different physical screen sizes, identical resolution (let's say, 160ppi) is a good question, because using dp results in identical physical sizes. eg: smartphone with 160ppi, versus a wall monitor with 160ppi. a 24dp sized font is not going to be very readable for the latter, as it is always 0.15 inches high on a 160ppi device.] Jan 25 18:40:28 Yup, the values-small/normal/large/xlarge seems to be the solution I'm looking for. Jan 25 18:41:32 kbs I'm using 16dp. There are 4 different files with different values, works like a charm. Jan 25 18:41:57 jdawes: yep - same here. I use values-wnnndp ; but same idea Jan 25 18:42:12 There's also a subtle problem with using dp, that results in inconsistent sizes on the same sized devices... but that's a whole another can of worms Jan 25 18:42:48 You should be using sp for text size Jan 25 18:43:00 That was a typo, you're correct. Jan 25 18:43:03 Just realized that. Jan 25 18:43:08 good luck getting a designer to sign off on using SP Jan 25 18:43:22 "what do you mean the user can change the font size" Jan 25 18:43:29 lol Jan 25 18:43:52 I don't think a designer would sign off on dp either -- even that's not teh same size, in reality :-) Jan 25 18:44:31 kbs, a device with the same parameters and fuck ups will always looks the same as another device with the same parameters and fuck ups Jan 25 18:44:56 it just means a 7" galaxy tab with the same resolution and density as a real tablet usually doesnt look the same Jan 25 18:46:06 mikedg: I dunno all that; but basically two android tablets with the same physical size -- if you use dp (or sp with the same scaling) , there's no guarantee the sizes are the same. Jan 25 18:46:19 pt/in/mm does work correctly, however Jan 25 18:46:25 well, identically at least :-) Jan 25 18:46:41 really? Jan 25 18:46:45 yes Jan 25 18:47:01 i mean barring sub pixel rendering and oled stupidity Jan 25 18:47:01 if one has 140ppi, and the other has 180ppi, they both get treated as 160ppi devices Jan 25 18:47:22 device manufacturers set the scaling used by dp Jan 25 18:47:25 so, a 160dp line on one is shorter than an inch, and the other is longer than an inch Jan 25 18:47:26 oh yeah physical size Jan 25 18:47:34 i thought you were talking about physical and density Jan 25 18:47:58 in general, your layouts should not depend on a particular size Jan 25 18:48:05 see, from a user's point of view, I just don't want to see text too small. If you set it at 9dp, then on the 180ppi device it becomes too small Jan 25 18:48:23 zinx: I'm afraid android forces said manufacturers to quantize it to one of a few pre-defined resolutions Jan 25 18:48:40 if you use dp insted of pt, the physical size of all widgets become different Jan 25 18:49:06 (even though the actual device is the same physical size) Jan 25 18:50:03 o_O Jan 25 18:50:20 kbs: it doesn't force them to Jan 25 18:51:11 most fix it at certain ones so less assets are scaled Jan 25 18:51:24 but there is a separate setting for font scaling. Jan 25 18:52:19 The difference is really minor Jan 25 18:52:31 yes Jan 25 18:53:49 Hey if any body has a moment to help me with UI design ideas, I'd appriciate it. I'm a little bit stuck here, and I want to make this app as a useful tool for our company but also a learning exercise. Jan 25 18:54:39 i bet your company doesnt know shit about good ui's Jan 25 18:54:43 throw some poop at the wall Jan 25 18:54:51 mikedg Jan 25 18:55:11 First this is a very small company, and this is for another programmer, so he knows what sucks Jan 25 18:55:12 Jouva: Jan 25 18:55:17 Second, screw off :) Jan 25 18:55:25 oh it's for a programmer, you can make the ui extra shitty Jan 25 18:55:28 throw a command line in there Jan 25 18:55:31 he'll eat it up Jan 25 18:55:42 ^^ Jan 25 18:55:56 You obviously didn't see the learning exercise part of my statement, so yeah that ain't going to work Jan 25 18:56:13 I'd like to learn to make apps that don't look shitty Jan 25 18:56:43 If I wanted to make an app that looked shitty, I'd not come here and make it the old Android 1.5 way without fragments and screens for each part Jan 25 18:57:44 Read the design guidelines Jan 25 18:57:48 Anyway, I'm trying to figure out the best way to handle the UI for my project. Basically an internal app to handle repair tickets. I'll be using Fragments for the UI. The first screen will list tickets, but I'd like for there to be a choice of which tickets to list (open, closed, all). From there when you choose a ticket, there's going to be 2 possible screens. One for the information Jan 25 18:57:48 about the case, and another that is a "tune-up checklist" of sorts Jan 25 18:58:02 Figure out how you can use the patterns it introduces Jan 25 19:00:19 SimonVT: Well I'm just stuck with how to make this flow right. Partially how to handle switching between open/closed/all tickets, and partially on how to handle switching between the case info and checklist forms, since I want to do this for the various layout orientations for different sizes Jan 25 19:01:39 I guess phones could have a nav spinner on both the list of cases and the switching between forms Jan 25 19:04:22 http://developer.android.com/design/patterns/app-structure.html Jan 25 19:04:22 But what about landscape and portrait tablet? I'm thinking something like the gmail client. Open/Closed/All list, then a ticket list fragment. But what I get stuck on is how they do the landscape look Jan 25 19:04:23 http://developer.android.com/design/patterns/actionbar.html Jan 25 19:04:28 I'm aware of those Jan 25 19:04:33 And have been reading those Jan 25 19:07:00 How they do the landscape look? Jan 25 19:07:47 In the gmail and email app, for tablet landscape, they have folders (Inbox, Sent, etc) then the larger fragment has a list of emails Jan 25 19:08:08 When you tap an email, then the email list becomes the smaller fragment and the contents becomes the larger fragment Jan 25 19:08:44 It's a custom viewgroup Jan 25 19:08:47 And I guess I'm not sure what they do to do this Jan 25 19:08:48 Ok Jan 25 19:08:50 With 3 panes Jan 25 19:09:03 Oh. Ick. Jan 25 19:09:12 You can toggle between content showing, and content not showing Jan 25 19:09:57 So, based on orientation, they just decide whether to display the list along with the content Jan 25 19:12:58 (sorry if someone replied to my dp/pt screen size tirade, network dropped out but would love to hear rebuttals and thoughts :-) Jan 25 19:13:32 We've moved on Jan 25 19:18:09 hopefully, "we've" also moved beyond the knee-jerk dp is always correct theory :-) Jan 25 19:19:29 I don't think anyone believed that Jan 25 19:19:33 The difference is just insignificant Jan 25 19:20:00 Unless your app is a ruler Jan 25 19:21:56 Well. I can see that a 15% swing in font sizes and treating a 24dp font on a smartphone and a wall monitor screen with identical resolutions _might_ insignificant to some :-) Jan 25 19:22:14 *_might be* Jan 25 19:22:38 kbs: the wall monitor likely has a very different setup. Jan 25 19:23:50 zinx: unless you think their pixels per inch are always going to be different. or, compare google-glass type devices and tablets, etc. Jan 25 19:24:31 kbs: i think if you're concerned about usability you probably have a different layout for -large / -xlarge or whatever Jan 25 19:25:24 zinx: yes, that's exactly what I indicated earlier, pre your "i need to stop reading what people write in this channel" remark :-) Jan 25 19:25:44 kbs: what you're saying and what i am saying are entirely different things. Jan 25 19:26:17 nobody in this channel ever tells the truth Jan 25 19:26:26 zinx: would you help me understand by clarifying the difference? would much appreciate it Jan 25 19:26:26 mikedg is telling the truth Jan 25 19:26:37 There's a difference between recommending it as best-practice, and recommending it for a specific use-case Jan 25 19:26:38 PARADOX UP IN THIS Jan 25 19:29:47 i'm making a breakthrough in the thermal coupling of widgets to data cross process mr fusion! Jan 25 19:30:17 tl;dr: WIDGETS, WOO! Jan 25 19:30:27 1.21 jiggawatts! Jan 25 19:30:34 SimonVT: well :-) My understanding was exactly that use-case. jdawes specifically asked "When you are dealing with various physical screen sizes with similar resolutions, how would you, say, change the text size apporpriately, using dp and sp dont' create the desire effect." I think that's a very accurate perception of practical reality. Jan 25 19:30:58 kbs: mu Jan 25 19:31:15 that's my answer and i'm sticking to it Jan 25 19:31:16 kbs: And generally, you don't Jan 25 19:31:58 kbs: if i were to buy a new chair, what color would i have to paint it to make the internet go faster? Jan 25 19:32:47 People come in here with a lot of stupid ideas and misconceptions Jan 25 19:33:06 indeed :-) Jan 25 19:33:21 Noone with a clue is going to help those people realise those ideas unless they have a damn good reason Jan 25 19:33:45 stupid people come in here with lots of ideas from conception Jan 25 19:34:34 people come in here with a lot of fast chairs Jan 25 19:34:49 thankfully we have zinx here to help Jan 25 19:35:36 zinx slows everyone down :( Jan 25 19:35:57 dat relativity Jan 25 19:40:08 yawn Jan 25 19:40:20 hi SimonVT Jan 25 19:40:22 and lov Jan 25 19:40:24 and mikedg Jan 25 19:40:29 Hi canadiancow Jan 25 19:40:34 sup dog Jan 25 19:40:47 hey canadiancow Jan 25 19:41:26 hi Napalm Jan 25 19:42:05 canadiancow: have you seen glass? Jan 25 19:42:19 i dont know why you're asking me that Jan 25 19:42:21 you know the answer Jan 25 19:42:29 well you better clean it Jan 25 19:42:35 you ruiend my joke Jan 25 19:42:36 fuck you Jan 25 19:42:39 :D Jan 25 19:42:41 lol Jan 25 19:42:58 anyone going to the AT&T mobile game hackathon thing tonight? Jan 25 19:43:10 in...redwood city Jan 25 19:44:46 i can go if you pick me up Jan 25 19:45:05 yo you can get a flight for like $150 Jan 25 19:45:08 just come down Jan 25 19:45:10 you shouldnt be allowed at a hackathon, i bet you are a zynga spy Jan 25 19:45:16 yeah Jan 25 19:45:17 exactly Jan 25 19:46:38 nah it's cool Jan 25 19:46:51 they'll just kick him out for trying to take pictures before it starts Jan 25 19:46:59 lol Jan 25 19:47:27 mikedg: anything interesting happened since i left the office? Jan 25 19:47:55 i set a chair on fire and pushed it out a window Jan 25 19:48:01 ballmer'd Jan 25 19:48:07 haha Jan 25 19:48:09 Hello Jan 25 19:48:41 I'm using adt/eclipse and when I try to create a new master/detail activity (or any other with navigation for example) it always tells me that minSDK is 1 but it needs to be at least 11 Jan 25 19:48:45 my minsdk in manifest is 15 Jan 25 19:48:58 I need to create a virtual device for samsung galaxy s and sIII. How important is it that I get the skin vs just plugging in the specs like the screen size and OS version? Jan 25 19:48:58 is it a bug? Jan 25 19:49:05 yes Almindor, in your app Jan 25 19:49:09 pastebin your manifest Jan 25 19:49:40 windsurf_: It makes no difference Jan 25 19:49:51 SimonVT: thank you Jan 25 19:50:38 canadiancow: http://pastebin.com/kjwmGThQ Jan 25 19:51:13 so what is telling you that the minsdkversion is 1? Jan 25 19:51:22 yes I thought you jumped conclusion Jan 25 19:51:33 Almindor: clean build? Jan 25 19:51:56 all the 14+ sdk features work, it's just the new activity wizard Jan 25 19:53:20 where's this new activity wizard Jan 25 19:53:24 so if you right click project, new/other/android/master-detail activity on top of that new window it gives me: "This template requires a minimum SDK version of at least 11, and the current min is 1" Jan 25 19:53:25 SimonVT: Would be nice to have the skin – I followed samsung's instructions but it doesn't show up in my list in ave for some reason. I put the unzipped Galaxy_S in //platform/android-10/skins Jan 25 19:53:36 nm Jan 25 19:53:46 windsurf_: I don't know Jan 25 19:53:50 it relaly makes no difference Jan 25 19:53:58 SimonVT: yeah ok thanks :) Jan 25 19:53:58 It doesn't emulate the actual hardware Jan 25 19:54:04 yeah so no point Jan 25 19:54:15 good for marketing screenshots i guess Jan 25 19:54:18 but not what i'm after today Jan 25 19:54:20 It's a standard emulator where the window around it just looks like a samsung device Jan 25 19:54:20 Almindor, it works for me Jan 25 19:54:30 do you have the latest ADT? Jan 25 19:54:39 should be, i updated like 2 days ago Jan 25 19:54:42 let me check Jan 25 19:54:56 There are plenty web apps that lets you make device screenshots just by dragging a screenshot taken on a device Jan 25 19:55:04 well i cant reproduce the issue Jan 25 19:55:15 21.0.1 Jan 25 19:55:30 I mean sdk Jan 25 19:55:34 anyway im on a bus, and my secondary battery is almost dead Jan 25 19:55:37 so i need to leave Jan 25 19:55:38 bye Jan 25 19:55:43 no update found for adt Jan 25 19:55:56 so I'll flag this as a bug in linux/eclipse/adt plugin and go blank activity I guess Jan 25 19:56:54 like a pro Jan 25 20:05:10 i hope google cancels the glass thing next week to punish these fuck bags who keep leaking shit Jan 25 20:05:20 and also stops giving out free shit at io Jan 25 20:05:23 oh cool, got the skin working SimonVT, fyi Jan 25 20:05:37 mikedg: why stop giving out free shit at io? Jan 25 20:05:38 needed to use command line to launch the specific avd i created and point it to the -skindir and -skin Jan 25 20:05:53 i understand the leaking, but free stuff? Jan 25 20:05:54 so fuckbags stop gobbling up tickets like cocks at your moms birthday Jan 25 20:06:19 oi? Jan 25 20:06:51 im with you, but for us legit devs its good to get your hands on the latest hardware Jan 25 20:07:17 we got to get Noah to come this year Jan 25 20:07:28 Legit devs will most likely buy the devices once they're out anyway :p Jan 25 20:08:13 yes, but its difficult to get your hands on them Jan 25 20:08:38 i have pratically all the Nexus products Jan 25 20:08:59 everything apart from the Nexus S.. I have the G1 Jan 25 20:09:07 even though thats not exactly "Nexus" Jan 25 20:24:59 why the heck is samsung still releasing ICS phones Jan 25 20:25:26 at least they stopped releasing gingerbread phones Jan 25 20:25:42 because it takes more than an hour to upgrade to jelly bean and do a full QA? Jan 25 20:25:46 and to develop hardware Jan 25 20:25:53 and to manufacture and ship worldwide Jan 25 20:27:57 hopefully now that oems get early access with the PDK this won't happen much more often Jan 25 20:28:21 it will still happen Jan 25 20:33:39 getWindow().addFlags(WindowManager.LayoutParams.FLAG_DISMISS_KEYGUARD); getWindow().addFlags(WindowManager.LayoutParams.FLAG_TURN_SCREEN_ON); Jan 25 20:33:49 its not working. Any idea why? Permissions are set Jan 25 20:36:17 Window flag: when set the window will cause the keyguard to be dismissed, only if it is not a secure lock keyguard. -> its a Secure Lock Keyguard. I think this may be a problem :S Jan 25 20:37:11 misterli: i believe you can show a window above the secure lock keyguard, just not dismiss the secure lock keyguard Jan 25 20:37:21 does that help? Jan 25 20:38:08 misterli: Any progress on the camera capture? Jan 25 20:38:16 no, does not. Maybe i dont need to unlock. Problem is by camera.takePicture. It does not shoot until keylock.. No, not yet :/ Jan 25 20:38:26 It dont call the Callback aslong as the Keyguard is locked Jan 25 20:38:43 If its unlocked it works.. But even unlocking only works with deprecated features. The Line above even with FLAG_SHOW_WHEN_LOCKED does not work Jan 25 20:39:25 misterli: what are you trying to do? Jan 25 20:46:46 what does lid mean re: keyboard lid support Jan 25 20:46:49 ? Jan 25 20:47:19 Napalm. Anti Thief Feature Jan 25 20:47:46 misterli: Thermite is better Jan 25 20:48:17 What is Termite? Jan 25 20:48:29 misterli: a very good "burning" kind of explosive Jan 25 20:48:39 haha ;> Jan 25 20:48:52 well. I want to offer a Free Anti-Thief Software which survives even full wipes Jan 25 20:49:15 It has already lots of features like GPS-Tracking, Navigation, Camera-Shots, SMS-Read, Calllogs, etc etc.. But im missing the Cam-Shot on wrong Unlock Jan 25 20:50:33 misterli: anyway, "Thermite" was the favored BOFHly-method of disk wiping Jan 25 20:51:01 Haha yes, or just a DeviceAdmin which is easier to handle ;) Jan 25 21:04:58 it's annoying that _noone_ knows a solution.. not even a single site on google :/ Jan 25 21:06:04 Lookout does the cam-shot on wrong unlock Jan 25 21:13:32 misterli: i'm late to the discussion and missed part of it, but DeviceAdminReceive has an onPasswordFailed which might be what you're looking for Jan 25 21:16:34 misterli: If you can link me to a (free) apk that does the camera thing I could get my hands dirty in smali and help you later tonight after I finish my essay :) Jan 25 21:58:38 hi guys Jan 25 21:58:57 anyone ever experienced an EditText using the wrong density drawable when you use an asset for the background? Jan 25 21:59:15 it appears that the n7 is using the xhdpi version of the EditText background and not the mdpi version Jan 25 21:59:24 and it doesn't happen on other mdpi devices Jan 25 21:59:39 n7 isn't mdpi (160), it's tvdpi (213) Jan 25 21:59:52 really Jan 25 21:59:55 wtfdpi? Jan 25 21:59:59 wtfdpi indeed Jan 25 22:00:06 what dpi does it default to then if it has no assets? Jan 25 22:00:24 generally n7 prefers hdpi (240) and scales them down, but i guess if there isn't tvdpi or hdpi it'll go with the higher rather than the lower for quality Jan 25 22:00:51 (well it prefers tvdpi over hdpi, but assets for tvdpi aren't very common) Jan 25 22:05:34 tnzr: or the context used on the view might have some weird configuration Jan 25 22:06:25 thanks kevinb!! Jan 25 22:06:36 Napalm: cool I will check that out too, thanks Jan 25 22:07:59 yep it's using h! Jan 25 22:08:15 Yeah, you don't really have to worry about tvdpi drawables Jan 25 22:08:47 iirc, android on n7 only have a few tvdpi resources.. Rest are hdpi Jan 25 22:08:47 tnzr: if you want your app to be listed in play store for n7 remember to add the 213 dpi tags to your manifest Jan 25 22:13:35 even the RIM developers don't like eclipse :| "I understand why RIM is using a lot of third party tools, such as Eclipse. This is the fast and easy approach, but seriously, I have yet to meet a game developer that would not prefer Visual Studio, or any other IDE in fact.": Jan 25 22:15:53 thanks for the help guys Jan 25 22:16:05 it appears to only be happening with this particular asset - all the other hdpi assets seem to look just fine Jan 25 22:16:26 so I'm thinking we can make a tvdpi asset and have the designer just cut a new one at that density Jan 25 22:16:32 does anyone here know how I could add a ViewPager to every row of my listview? I have a custom array list adapter for my listview. I havn't been able to find an answer for two days :( Jan 25 22:16:35 and the rest of the assets can continue using hdpi? Jan 25 22:17:18 Yes Jan 25 22:18:24 For each resource it goes through your res/ dir and finds the best match.. It doesn't care where the last resource was loaded from Jan 25 22:21:21 hey Jan 25 22:21:27 i need help plz Jan 25 22:22:04 my application is not opening until i open bluetooth Jan 25 22:22:38 it crashesh when askes the user if he acccept to open BT Jan 25 22:22:42 MajorZero: it doesnt sound like a good idea. But if you want to, you would forget ArrayAdapter and create your own.. and return your own custom layout in getView Jan 25 22:22:58 im using this Jan 25 22:23:01 Intent enableBtIntent = new Intent(btAdapter.ACTION_REQUEST_ENABLE); Jan 25 22:23:01 startActivityForResult(enableBtIntent, REQUEST_ENABLE_BT); Jan 25 22:24:32 any help Jan 25 22:26:13 Napalm: So it would be my own adapter (not array adapter) and itl allow me to return a layout in each row? How would I initialize the viewpager though? Just inflate it right there? Jan 25 22:26:34 if you like Jan 25 22:26:35 Thanks Simon Jan 25 22:26:58 is there a way to enable bluetooth without prompt Jan 25 22:27:12 MajorZero: hopefully, your ViewPager doesn't have ListViews also :) Jan 25 22:27:21 lol Jan 25 22:28:22 haha....well funny you mention that....one of the fragments is a listview.....do you think that would be too satanic? Jan 25 22:28:43 It would break in horrible ways Jan 25 22:28:52 You should rethink your layout Jan 25 22:30:10 SimonVT:well idealy, what this layout is, a newfeed like twitter or facebook...so each row has a layout that shows some information...however, I want this layout to have horizontal SLIDING, which is WHY i chose the listview to contain everything...do you have a better recommendation of how to approach this (THank you very much for your input now) Jan 25 22:31:05 >??????????? Jan 25 22:31:53 Sounds confusing Jan 25 22:33:43 MajorZero: you can have nested views that don't overload scroll dimensions; so one scrolls x the other y - technically thats ok Jan 25 22:33:55 but, i haven't seen any apps that do this so ... Jan 25 22:34:13 if you are going to try it, you had better put together a prototype and do some usability testing with 5 or 6 people Jan 25 22:34:49 thanks guys. I really appreciate the feedback im getting. Jan 25 22:34:52 btw, linear paging is considered bad usability in many cases Jan 25 22:35:10 even if the android design guidelines mention it for some cases Jan 25 22:37:51 hi... need hint on how to put spinner "inside" spinner? Jan 25 22:38:55 all the Dawgs are out on friday Jan 25 22:39:36 hello; i have an app that publishes a contentprovider and another one that usually consumes it, but i'm getting the following exception in the consumer now: java.lang.SecurityException: Permission Denial: opening provider edu.ucla.cens.mobility.MobilityContentProvider from ProcessRecord{4187522817710:org.ohmage/u0a10061} (pid=17710, uid=10061) that is not exported from uid 10069 Jan 25 22:39:55 g00s++ Jan 25 22:40:12 diphtherial, is the contentprovider exported? Jan 25 22:40:15 nothing much has changed in the provider; the line that declares the contentprovider in the manifest is still there Jan 25 22:40:16 or does it require a permission? Jan 25 22:40:23 can you pastebin both manifests Jan 25 22:40:35 sure thing, one moment; eclipse is crashing on me repeatedly :x Jan 25 22:41:34 By default, providers are not exported (since API17) Jan 25 22:42:27 You can either set your targetSdk to <=16, or set the exported attribute to true Jan 25 22:42:50 aha, that's very likely it Jan 25 22:42:52 thank you :) Jan 25 22:43:34 Hey I got it all working. ;) Jan 25 22:43:58 Went ahead and installed andchat on my g1 Jan 25 22:47:25 canadiancow: by default they are android:exported="false" now right? Jan 25 22:47:39 seems that way Jan 25 22:47:48 Don't you trust me? :/ Jan 25 22:47:51 perhaps im thinking of something else Jan 25 22:47:57 i trust you SimonVT Jan 25 22:48:03 TRUST FALL Jan 25 22:48:14 SimonVT: i didnt see your response Jan 25 22:48:20 my bad Jan 25 22:48:37 diphtherial: you should add your own custom permission anyway Jan 25 22:50:12 custom permissions for third party apps are buggy, i'd avoid them if possible. if you declare in package A and request in package B it will work as long as A is installed before B. But if B is installed first then B won't have the permission until you uninstall and reinstall it while A is already installed Jan 25 22:50:44 you can declare the permission in both packages, but that doesn't feel right. and potentially some other app could declare your permission Jan 25 23:09:03 help on how to do spinner inside spinner? Jan 25 23:09:34 ack, that sounds like a UI nightmare Jan 25 23:11:48 nicole24: what kind of nonsense is that? Jan 25 23:12:40 nicole24: just have two spinners side-by-side horizontally or vertically and have the second spinner's data depend on the first spinner's selection or something more sane Jan 25 23:16:28 yeah nicole24, what data are you trying to display Jan 25 23:31:15 pragma- thx, thats better idea then i had first... experience :) Jan 25 23:32:38 Hey, odd question, but does anyone know of any good batching tools for the Android Emulator? Jan 25 23:33:02 I'm looking for a way to automate testing mobile websites using the Base Android profile on the Emulator. Jan 25 23:33:11 Err let me clarify that. Jan 25 23:34:10 I mean I'd make a series of AVDs built to vanilla configuration for an API level such as Gingerbread or Ice Cream Sandwich and have a way to simultaneously open all of those AVDs and point the browser to a specific mobile file we'd want hosted locally. Jan 25 23:38:10 well you can send intents through adb Jan 25 23:38:21 an intent can open the browser to a specific url Jan 25 23:38:32 you can fairly trivially do that for all attached devices Jan 25 23:40:01 adb devices | tail -n +2 | cut -sf 1 | xargs -iX adb -s X shell am start -a ACTION_VIEW -d http://www.google.com Jan 25 23:40:26 that will launch google.com (or the activity chooser if you have multiple browsers) on all connected devices Jan 25 23:40:37 canadiancow: ahhh `tail` ... ahhhh :D Jan 25 23:41:02 I just commited a fix some days ago, because there was some false assumptions using `head` in the android build scripts Jan 25 23:41:25 where? Jan 25 23:41:33 https://android-review.googlesource.com/#/c/50390/ Jan 25 23:41:45 oh that one Jan 25 23:41:59 (no covert advertising for my commit intended ;) ) Jan 25 23:42:14 why hasnt it been verified yet Jan 25 23:42:17 i thought that was an automatic process Jan 25 23:42:33 Parent(s) Jan 25 23:42:33 7a184cb70f7fbdf7f9f64b88338dcd155cfe1e3d Merge "Change Email2 reference back to Email" Jan 25 23:42:41 sigh Jan 25 23:42:47 since it was my very first commit to AOSP, i am ferrly new to that, and still am not sure about how everythign works :D Jan 25 23:42:53 Is there a way to animate a ViewGroup width except using a custom IntEvaluator? Jan 25 23:43:09 canadiancow: i guess that was the last commit, when i edited it? Jan 25 23:43:18 yeah yeah Jan 25 23:43:30 but i am very welcoe, if you could explain me, what i should have done :) Jan 25 23:43:38 i dont know what you should have done Jan 25 23:43:39 https://android-review.googlesource.com/#/c/49610/ Jan 25 23:43:42 thats a change i made Jan 25 23:44:04 the autoverifier verified it like 3 hours after i submitted it Jan 25 23:45:28 Oh yeah, I found a great solution http://android.amberfog.com/?p=758 :) Jan 25 23:47:34 Yeah, performance is horrible with that approach Jan 25 23:48:29 SimonVT: Is there another way? Jan 25 23:48:32 Your entire view hierarchy is being re-measured, layout, drawn every ~16 ms, or however often the screen is drawn Jan 25 23:49:44 Well, _basically_ this is the approach, but there's a bit more to it Jan 25 23:50:06 Before starting your animation you want to measure each pane at its new size Jan 25 23:50:41 I believe gmail also does this Jan 25 23:50:56 Then you want to enable stuff like hardware layers to make it smoother Jan 25 23:51:19 And yeah, you don't want to be changing layoutparams every 16 ms Jan 25 23:51:25 Make a custom viewgroup that handles this Jan 25 23:51:36 gmail does what Jan 25 23:52:07 or well, I guess it wouldn't matter Jan 25 23:52:11 Both cases is a requestLayout Jan 25 23:52:43 I'll try with this solution and use traceview to inspect performances to ensure it takes under 16ms to measure child Jan 25 23:53:21 canadiancow: gmail measures the list in the middle pane at its new size before animating Jan 25 23:53:41 middle pane? Jan 25 23:53:49 which list is this Jan 25 23:53:59 Sheesh Jan 25 23:54:02 The list of emails Jan 25 23:54:44 JulienDev: it's not about how long it takes, it's about how much cpu and battery is being wasted on this eye-candy nonsense Jan 25 23:55:44 gotcha, so is it possible to resize only the viewgroup and resize child on animation finished? Jan 25 23:56:24 StingRay_: hey Jan 25 23:57:10 StingRay_: you were right, one can pass List>> through an intent via serializable... but not a vector Jan 25 23:57:19 StingRay_: i finally managed it :) Jan 25 23:57:21 StingRay_: thx Jan 25 23:57:33 t.t Jan 25 23:57:38 cliffer: dafuq. Jan 25 23:57:54 pragma-: why? Jan 25 23:58:02 oh SimonVT, i see what you're talking about lol Jan 25 23:58:26 ^_^ Jan 25 23:58:27 SimonVT: You mean that Gmail measures email list child on animation start and never do it again later during the animation? Jan 25 23:58:42 JulienDev: Yes Jan 25 23:59:09 SimonVT, how do you know how gmail works Jan 25 23:59:28 I know all your secrets Jan 26 00:00:12 I need a slow motion camera :) Jan 26 00:00:34 SimonVT, go run hierarchyviewer on gmail and you'll see soem interesting stuff :P Jan 26 00:00:52 Can't ;O Jan 26 00:00:55 Stock roms Jan 26 00:01:01 ah Jan 26 00:02:53 I'm sure they do some weird shit tho :p Jan 26 00:04:24 I know the row layout is basically just a single view (or it used to be iirc) Jan 26 00:05:33 can anybody confirm correctness and position of line 13 in http://pastebin.com/21C2Y5TB ? the dialog window never shows up. async-task is called from onActivityResult with ListeVerarbeiten(this, webResult).execute(); Jan 26 00:06:19 if it matters: calling activity is a ListActivity Jan 26 00:06:30 I can confirm that is not the correct way to do it Jan 26 00:07:06 SimonVT: thx Jan 26 00:07:13 np Jan 26 00:07:23 SimonVT: really just a single view? I can't hierarchyview, I don't have true root Jan 26 00:07:39 yes it is Jan 26 00:07:48 JulienDev: Yeah Jan 26 00:07:49 and thats what i wanted you to see SimonVT , but sounds like you already have :P Jan 26 00:07:56 SimonVT: so its the postion or the content of line 13? Jan 26 00:08:02 How is it possible? Jan 26 00:08:34 During onMeasure they.. measure.. everything Jan 26 00:08:54 with CompoundDrawable ? Jan 26 00:09:11 No Jan 26 00:09:37 I mean Jan 26 00:09:50 They calculate the size of the checkbox, the title, the date, etc, etc, etc Jan 26 00:10:00 And during onDraw they just draw them at those sizes Jan 26 00:10:26 They might also figure out the position of each element in onMeasure Jan 26 00:10:32 And possibly visibility Jan 26 00:10:59 Sorry, I was talking about "row layout is basically just a single view" Jan 26 00:11:06 Yes Jan 26 00:11:37 Each email in the list is a single view Jan 26 00:11:58 Amazing Jan 26 00:12:51 Not really :p Probably a lot more timeconsuming tho (to code) Jan 26 00:12:58 But they can do a lot of optimizations Jan 26 00:13:01 Making it smoother Jan 26 00:15:06 interesting Jan 26 00:16:01 Framework views have to handle a lot of use-cases. When your ui starts to get more complex, framework views might just be too slow for what you're trying to achieve Jan 26 00:16:24 10-times nested LinearLayouts with weights FTW! Jan 26 00:16:34 e.g. take a look at textview Jan 26 00:16:38 It's just amazing Jan 26 00:16:50 But for a lot of use-cases, Canvas#drawText might be enough Jan 26 00:17:00 And most likely a lot faster Jan 26 00:17:34 i have a ViewPager that pages (duh) between ListViews. scrolling left to right is kind of slow. any tips for making it not choppy ? Jan 26 00:18:58 Yeah, I tried to create a JustifiedTextView but I had to handle to much use-cases I dropped the idea Jan 26 00:23:33 It's a lot easier to customize something for your needs, rather than make a general library Jan 26 00:37:16 Yes sure. Time to sleep now. Thank you very much for all your explanations Jan 26 01:01:31 Napalm: are you currently here? Jan 26 01:01:38 yes Jan 26 01:02:41 Earlier we spoke about implementing a viewpager in each of my listview rows...I finished writing my custom adapters and what not and I can my application running, however I cannot see my viewpager or listview in the beginning...would it be possible if you could check my code to see if maybe a certain view is conflicting with one another? Jan 26 01:05:27 i would guess you could see your listview before you changed the adapter Jan 26 01:05:51 therefore i would assume your new adapter has a bug which is probably returning 0 as its count of items Jan 26 01:06:38 remember you need to call notifyDataSetChanged to get the listview to refresh its views from the adapter Jan 26 01:07:19 I downloaded the master branch from AOSP, and can't find the build target for the nexus 7. Jan 26 01:07:28 Can I just run lunch full-eng to build for the nexus 7? Jan 26 01:07:31 Thank you. Jan 26 01:08:07 napalm: So after I set my listView adapter I should call it? Jan 26 01:09:21 Oh, Is it just the grouper? If so, thank you anyway. Jan 26 01:12:13 for some reason it deosn;t let me call notifyDataSetChanged(); Jan 26 01:13:40 MajorZero maybe your attempting it on the listView rather than the adapter ? Jan 26 01:16:31 Im doing it on my listview and the only methods its letting me call are notify() and notifyAll() Jan 26 01:22:55 http://sphotos-a.xx.fbcdn.net/hphotos-prn1/20413_10200464094507573_636597525_n.jpg this is a copy of my code Jan 26 01:23:04 for the listview Adapter....for some reason my listview isnt showing up Jan 26 01:23:12 any advice would be greatly appreciated Jan 26 01:24:33 MajorZero err...paste bin would be good Jan 26 01:24:40 rather than a screenshot Jan 26 01:25:25 getCount() { return 0; } Jan 26 01:27:13 My apologies, before I was using an ArrayList adapter, since Im extending a base adapter now, how should I get the count? Jan 26 01:27:54 get it from whatever data backs your adapter Jan 26 01:29:09 in this case I assume it would be my pageradapter then Jan 26 01:30:03 guys.. Jan 26 01:30:07 it.. Jan 26 01:30:09 it works Jan 26 01:30:11 :') Jan 26 01:30:13 by the nine divine Jan 26 01:30:15 it works Jan 26 01:30:27 Thank you. Bless you all. This has been a thorn in my side for almost 48 hours Jan 26 01:32:07 StingRay_, my first born will be named after your irc tag. JakeWharton, you will be my second, and Napolm, you will be my third. THank you Jan 26 01:33:13 haha, good luck Jan 26 01:35:23 how come every smartphone screenshot i see, the time is always 4:20? Jan 26 01:35:50 cause thats the only time your allowed to take shots Jan 26 01:35:52 and if the time is always 4:20, how does anyone get any work done? Jan 26 01:35:56 dont work otherwise ! Jan 26 01:41:09 hey JakeWharton can I use your spinner control outside of abs? or is it always part of an action... Jan 26 01:41:16 no, it only works in the action bar Jan 26 01:41:32 that's what i thought, is there a nice compat spinner that you know of? Jan 26 01:43:05 kind of a dumb question i guess, i just want that spinner style across all platforms Jan 26 01:43:12 but if there was ever an easy style to implement... Jan 26 01:43:22 well you can use the style if you don't care about actually having a dropdown Jan 26 01:43:45 but having a dropdown is more complicated than just applying the same style Jan 26 01:44:11 yeah more dumbness, i guess spinner isn't a dropdown Jan 26 01:45:59 in what version of the SDK did spinner become a dropdown? Jan 26 01:46:02 11 Jan 26 01:46:14 we style them the same across all APIs but allow pre-API 11 Spinners to still do a pop-up Jan 26 01:46:50 right, that was what i was hoping for outside of the action bar Jan 26 01:47:33 you should be able to do that, then Jan 26 01:47:41 style="@style/Widget.Sherlock.Spinner.ActionBar or something Jan 26 01:48:11 sry i didn't read your previous statement right... i could live with that Jan 26 01:48:16 Widget.Sherlock.Spinner.DropDown.ActionBar Jan 26 01:48:23 esp. if it's as simple as using one of your styles Jan 26 01:48:44 yep, should be Jan 26 01:49:32 cool, thanks for that option Jan 26 01:49:58 if abs is doing it that way anyhow, it doesn't feel so inconsistent Jan 26 02:11:32 hello all Jan 26 02:13:26 lo Jan 26 02:13:35 anyone know, if i buy a new macbook pro, quad core, i7 – will my emulator (java) run faster  – will it make use of the extra cores? Jan 26 02:13:51 or will the upgrade be wasted i wonder Jan 26 02:14:10 windsurf_: typical emulators will not use the extra cores Jan 26 02:14:33 windsurf_: as to the entirely separate issue, yes, you wasted money, non-Apple hardware is generally far cheaper for the same components. Jan 26 02:15:25 windsurf_: if you're trying to make the android emulator run faster, just use the x86 one. it gets virtualized on most machines - the cpu doesn't have to be emulated as it does for ARM/MIPS Jan 26 02:16:01 or androVM if you want super speed Jan 26 02:16:02 :) Jan 26 02:16:20 the standard SDK using the x86 image is pretty fast apart from EGL Jan 26 02:16:35 i'm not sure they ever implemented acceleration for EGL in qemu Jan 26 02:17:15 zinx: when i run the x86 one it still takes quite some time. Jan 26 02:17:27 windsurf_: first boot, or thereafter? Jan 26 02:17:29 actually, when i boot from snapshot it's not that bad Jan 26 02:17:44 it's a little sluggish when it runs, you're making me remember that it's the compiling that's slow Jan 26 02:17:49 wonder if that would be faster with more cores Jan 26 02:17:53 windsurf_: when you boot for the first time, it converts the .dex files in every apk on the system in to .odex files that can be run by Dalvik Jan 26 02:18:17 yeah ok, i'm not as concerned with first time emulator boot as i don't have to do it often enough Jan 26 02:18:19 windsurf_: compiling /can/ be faster with more cores, if your make and/or compiler support it Jan 26 02:18:39 zinx: ok. Jan 26 02:18:52 windsurf_: the emulator will only get faster with faster ram (or possibly more ram if you are low), and higher single-thread cpu speed (which is not really tied to clock speed so much) Jan 26 02:18:59 sometimes i do native java with the eclipse ide and android sdk and sometimes with ttianium. Jan 26 02:19:24 i assume eclipse is optimized for it but not sure about titanium – both are using the same sdk though so i'm not sure if the question is more centred around the sdk itself. Jan 26 02:19:28 windsurf_: number of cores doesn't matter for the emulator, but you can do other things while using it, with more cores :/ Jan 26 02:19:59 zinx: yeah so you've made me realize my major issue is compiling speed. Jan 26 02:20:27 well, upgrading will likely increase that, depending on what you have now :) Jan 26 02:20:37 Anyone dealt with proximity alerts? I'm having a hell of a time and not sure why. My proximity alert registered and shows the right mRadius value and lat/lng but when I get near the location, nothing fires Jan 26 02:20:53 core 2 duo now zinx Jan 26 02:20:56 2.8 ghz Jan 26 02:21:06 windsurf_: yeah, it'll be faster Jan 26 02:21:07 so, i'd be moving to a 2.6 or 2.7? quad core Jan 26 02:21:19 zinx: gut feeling – 2x faster? less, more? Jan 26 02:21:27 windsurf_: dunno off hand Jan 26 02:21:37 like, right now, it takes me like 30 seconds to run any change i make Jan 26 02:21:39 it's crazy Jan 26 02:24:19 thanks for the input, gotta run Jan 26 02:29:53 MajorZero: sorry was afk, but your welcome. Jan 26 02:29:56 time for sleep Jan 26 02:46:21 Napalm|afk: what about his welcome? Jan 26 02:47:38 hey guys, i'm trying to process a android.intent.action.SEND in my onCreate method to upload an image, my app seems to send the image but my phone is just stuck on my app's empty screen when i really want it to return to the gallery, any direction on how I could get that to happen? **** ENDING LOGGING AT Sat Jan 26 02:59:59 2013