**** BEGIN LOGGING AT Tue Oct 23 03:00:00 2012
Oct 23 03:00:06 <sampullman> the scheme wiki is entirely scheme, i think
Oct 23 03:00:10 <grothendieck> neat
Oct 23 03:00:14 <muhoo> grothendieck: are you using jdee mode?
Oct 23 03:00:25 <muhoo> for android in emacs, i mean
Oct 23 03:00:28 <grothendieck> No, just regular Java-mode
Oct 23 03:00:46 <muhoo> the only thing i wish i had in emacs, was auto-imports
Oct 23 03:00:47 <grothendieck> I don't write much Java though for Android
Oct 23 03:00:48 <sampullman> same
Oct 23 03:01:01 <grothendieck> NDK?
Oct 23 03:01:03 <muhoo> what language are you writing android apps in?
Oct 23 03:01:03 <sampullman> muhoo: i recall seeing a plugin for doing that in the past
Oct 23 03:01:11 <grothendieck> muhoo: mostly C
Oct 23 03:01:15 <grothendieck> with some Java glue
Oct 23 03:01:17 <muhoo> :-0
Oct 23 03:01:18 <grothendieck> via JNI
Oct 23 03:01:31 <grothendieck> Math libraries, etc
Oct 23 03:01:36 <muhoo> oic
Oct 23 03:01:48 <grothendieck> Java is just for the interface
Oct 23 03:01:58 <grothendieck> And unfortunately, I suck at interfaces way more than I suck at Java
Oct 23 03:02:01 <muhoo> lucky you
Oct 23 03:02:15 <grothendieck> haha
Oct 23 03:02:17 <grothendieck> why?
Oct 23 03:02:57 <grothendieck> When I spend months trying to debug some JNI functions
Oct 23 03:03:02 <grothendieck> does not feel so lucky :P
Oct 23 03:03:08 <muhoo> well, true
Oct 23 03:03:09 <grothendieck> such a freakin mess
Oct 23 03:03:42 <sampullman> do you have any experience with openCV on android?
Oct 23 03:03:53 <grothendieck> No
Oct 23 03:04:02 <sampullman> nuts
Oct 23 03:04:07 <grothendieck> ooh but I did look at it once
Oct 23 03:04:12 <grothendieck> had an idea for a project but never got around to it
Oct 23 03:04:20 <grothendieck> seems interesting
Oct 23 03:04:24 <grothendieck> why? what are you up to?
Oct 23 03:04:45 <sampullman> they want you to ask your users to download their manager/some package in order to use it
Oct 23 03:05:02 <sampullman> but i got static linking to work, and was wondering if there was a downside to that
Oct 23 03:05:09 <grothendieck> No
Oct 23 03:05:11 <grothendieck> lol
Oct 23 03:05:15 <grothendieck> Absolutely no downside at all
Oct 23 03:05:20 <grothendieck> That's actually fantastic
Oct 23 03:05:23 <sampullman> besides apk size, i guess
Oct 23 03:05:27 <grothendieck> Well
Oct 23 03:05:37 <grothendieck> That and now you have to update it
Oct 23 03:05:40 <sampullman> it's not bad, ~4mb
Oct 23 03:05:52 <grothendieck> That's reasonable
Oct 23 03:05:58 <grothendieck> for a freaking CV library
Oct 23 03:05:58 <grothendieck> :P
Oct 23 03:06:12 <grothendieck> performance will be better
Oct 23 03:06:15 <sampullman> yeah, opencv is one of the best, too
Oct 23 03:06:16 <grothendieck> with static linking
Oct 23 03:07:13 <sampullman> for starters im trying to write an app that detects new features in an image, and automatically separates them into new layers
Oct 23 03:07:14 <grothendieck> I've been working on a tool to help NDK hackers port libraries to android easier
Oct 23 03:07:20 <grothendieck> oooh
Oct 23 03:07:21 <grothendieck> neat
Oct 23 03:07:26 <grothendieck> fourier transform
Oct 23 03:07:54 <sampullman> openCV has builtin feature recognition, i probably won't even have to think at that level
Oct 23 03:08:00 <grothendieck> ah:)
Oct 23 03:08:16 * grothendieck is writing an image comparison library in Perl at work
Oct 23 03:08:19 <grothendieck> :O
Oct 23 03:08:35 <grothendieck> it's basically just a toy though
Oct 23 03:08:46 <sampullman> that sounds awful, wouldn't c be better in every way?
Oct 23 03:09:01 <grothendieck> It uses XS
Oct 23 03:09:12 <grothendieck> But to answer your question, yes absolutely
Oct 23 03:09:12 <grothendieck> :)
Oct 23 03:09:17 <sampullman> ah
Oct 23 03:10:37 <grothendieck> I've never really understood if there is a problem with using a native binary executable on android
Oct 23 03:10:48 <grothendieck> provided it can run within the applications sandbox
Oct 23 03:11:47 <dragorn> grothendieck: A problem in what sense?
Oct 23 03:11:51 <grothendieck> I got some C program that runs as a daemon and connects to a client over a Unix domain socket
Oct 23 03:12:09 <grothendieck> client sends commands to the server, the server pipes them into this math library and returns the result
Oct 23 03:12:10 <dragorn> grothendieck: You're much better off making it a .so and loading it inside of JNI if you can - let it live in a java thread on its own
Oct 23 03:12:15 <grothendieck> right
Oct 23 03:12:24 <grothendieck> but damn
Oct 23 03:12:53 <dragorn> grothendieck: the java exec() function, i forget what it is, has issues - dalvik doesn't really like the fork/delay/exec it does, tho I've used it with success, i've read that it's got a good chance to crash
Oct 23 03:13:03 <grothendieck> yeah
Oct 23 03:13:06 <dragorn> For some stuff, i ended up writing my own fork/exec with a JNI hook
Oct 23 03:13:06 <grothendieck> I've played with that too
Oct 23 03:13:24 <dragorn> and then using JNI to make a FileDescriptor and populate it with dup'd fd's for stdin/out
Oct 23 03:13:45 <KaiKai> anyone know of a way i can convert the pub key in developer console which is a base64 encoded string to .pem?
Oct 23 03:13:46 <mrjeff> Hi All.  I'm using loadermanager with a custom simplecursoradapter.  It is not being updated automatically when db changes occur.  any idea what I might be missing?
Oct 23 03:13:46 <dragorn> so I could get popen functionality w/ java file streams that way
Oct 23 03:13:50 <grothendieck> I wrote some Java method that will start a shell script, and had this shell script to start/stop a daemon...it works but yeah, no idea why it is unsupported
Oct 23 03:14:02 <grothendieck> cool
Oct 23 03:14:07 <grothendieck> thanks dragorn that is food for thought :)
Oct 23 03:14:10 <ron_frow_> dragorn - honestly its getting better... but it seems like android was kinda like one of those things where
Oct 23 03:14:26 <ron_frow_> if the stars align in a very specific way things will work
Oct 23 03:14:30 <grothendieck> yes
Oct 23 03:14:32 <grothendieck> exactly
Oct 23 03:14:50 <dragorn> grothendieck: you can also do some tricks where you compile your exe as a .so then spoof argc/argv and call main directly in jni
Oct 23 03:14:57 <grothendieck> oh
Oct 23 03:14:59 <grothendieck> damn that is clever
Oct 23 03:15:00 <grothendieck> heh
Oct 23 03:15:07 <ron_frow_> I dont have a partnership with google so I cant tell you 3rd parties etc have more support/access
Oct 23 03:15:08 <grothendieck> what I did a while back (nto sure if this still works)
Oct 23 03:15:09 <ron_frow_> I assume they do
Oct 23 03:15:10 <profligacy> hey
Oct 23 03:15:23 <ron_frow_> but from what I've seen, its a miracle any 3rd party releases a stable android build =)
Oct 23 03:15:23 <dragorn> grothendieck: which I've also used to great success, but with some caveats - for a one-shot it's pretty good, but for multiple runs of the same binary, global variables have a tendency to stay initialized
Oct 23 03:15:25 <grothendieck> was bundle the executable in /assets, stuck on a .jpg extension
Oct 23 03:15:37 <grothendieck> dragorn: use volatile
Oct 23 03:15:47 <grothendieck> it should solve problems
Oct 23 03:15:51 <dragorn> grothendieck: yes, the problem in this case was that I wasn't interested in rewriting the native code
Oct 23 03:15:55 <grothendieck> ah
Oct 23 03:16:02 <dragorn> grothendieck: it was a work project to integrate existing tools
Oct 23 03:16:02 <grothendieck> right
Oct 23 03:16:22 <dragorn> I'd jump through a lot of hoops to avoid modifying major parts of the native code, because that would have long-lasting maintenance impacts
Oct 23 03:16:39 <grothendieck> of course
Oct 23 03:17:03 <grothendieck> nobody wants a mess of #ifdef ANDROID
Oct 23 03:17:04 <grothendieck> :P
Oct 23 03:17:10 <Orbitrix> #ifdefinitely
Oct 23 03:17:16 <dragorn> grothendieck: in this case we didn't own the native code either
Oct 23 03:17:21 <dragorn> grothendieck: so major sync issues
Oct 23 03:17:23 <grothendieck> :)
Oct 23 03:17:24 <grothendieck> right
Oct 23 03:17:30 <grothendieck> makes sense
Oct 23 03:17:57 <grothendieck> for future reference, if on some other project you have this problem
Oct 23 03:18:14 <grothendieck> volatile is useful
Oct 23 03:18:15 <dragorn> that said, I've wedged a 120k line c++ program doing a ton of stuff into a jni lib and spoofed main
Oct 23 03:18:19 <dragorn> and it runs pretty happily
Oct 23 03:18:22 <grothendieck> nice
Oct 23 03:18:25 <dragorn> yea I'll keep it in mind, thanks for reminding me
Oct 23 03:18:30 <grothendieck> I'll have to give that a try
Oct 23 03:18:35 <grothendieck> np :)
Oct 23 03:18:53 * grothendieck going for a walk, bbl
Oct 23 03:19:27 <scant> i want to inflate views by their ID (since they aren't each in their own file, and this do not each have their own layout number in the R file). however, the LayoutInflater only seems to accept layout numbers, not ID numbers. how can i get around this?
Oct 23 03:19:36 <scant> *thus
Oct 23 03:19:51 <dragorn> you might have to split them into their own files
Oct 23 03:20:01 <ron_frow_> dragorn just ndk stuff or platform dev
Oct 23 03:20:06 <dragorn> ron_frow_: ndk
Oct 23 03:20:19 <scant> are you sure? that'd be unpleasant
Oct 23 03:20:54 <dragorn> ron_frow_: For the most part I'm not too interested in platform dev - I don't work for a phone manufacturer, so if the answer is "roll your own rom", it means "no-one can use this"
Oct 23 03:21:09 <ron_frow_> understandable
Oct 23 03:21:15 <ron_frow_> just curious more than anything else
Oct 23 03:21:28 <ron_frow_> seems to be outside the realm of normal for the avg android dev
Oct 23 03:21:38 <dragorn> ron_frow_: similarly, if the answer is "requires root" it means "i probably won't bother".  I've jumped through some ridiculous hoops to avoid needing root.
Oct 23 03:21:52 <ron_frow_> dragorn any fun stuff thats public
Oct 23 03:21:57 <dragorn> ron_frow_: In this case, it was getting usb host + userspace re-implementation of realtek wifi drivers + kismet running on it
Oct 23 03:22:06 <ron_frow_> ahh
Oct 23 03:22:11 <ron_frow_> interesting
Oct 23 03:22:15 <dragorn> which was a huge hack, working on a re-implementation of it now
Oct 23 03:22:25 <dragorn> Of course, the new problem with THAT is, every manuf pooches USB host in some new way
Oct 23 03:22:37 <dragorn> which means I had to go out and spin some USB hardware to fix that
Oct 23 03:22:47 <dragorn> now i'm working on finalizing that stuff and getting a manufacturer lined up
Oct 23 03:22:50 <ron_frow_> what is the hardware
Oct 23 03:23:09 <ron_frow_> wireless hardware
Oct 23 03:23:11 <ron_frow_> ?
Oct 23 03:23:33 <dragorn> Not much.  It's a 1-port usb2 hub, with the ability to either a) power off the phone, b) split the power so an external battery pack powers usb devices w/out draining the phone, or c) backfeed +5v into the phone USB host chip to power it
Oct 23 03:23:54 <dragorn> because while the galnex/n7 can do usb host w/ power just fine, their batteries drop dead pretty quick if you throw a big usb radio on them
Oct 23 03:24:02 <dragorn> moto phones won't do usb host w/out 5v incoming
Oct 23 03:24:02 <ron_frow_> ahh
Oct 23 03:24:24 <dragorn> and a bunch of phones, moto included, won't let something as high power as a radio operate w/out a powered hub in the chain
Oct 23 03:24:35 <dragorn> so this sort of covers all the points in one board
Oct 23 03:24:52 <ron_frow_> cool
Oct 23 03:25:12 <dragorn> ron_frow_: https://lh4.googleusercontent.com/-akjkPU1z2NM/UGDHU_8I1nI/AAAAAAAACjk/wNcmbPu84F4/s720/1348519736633.jpg
Oct 23 03:25:17 <ron_frow_> in my exp with android... moto phones generally are the phones that are a pita and you have to work around stuff for
Oct 23 03:25:30 <dragorn> yeah they don't make life easy
Oct 23 03:26:07 <dragorn> I'm talking w/ either $newjob or some other folks about making a custom wifi usb radio w/ an OTG host cable integrated, plus the hub and power, and a lipo
Oct 23 03:26:21 <dragorn> so you plug the radio in to charge, then when you want to use it w/ the phone, turn it on and plug it in and it "just works"
Oct 23 03:27:08 <p_l> dragorn: regarding global variables staying initialized - it should be possible to explicitly unload the binary
Oct 23 03:27:19 <ron_frow_> I am gonna have some stuff I am gonna need to dig into some new jni... I've done a lot in the platform internals but generally hadnt needed to go creating a ton of jni integration
Oct 23 03:27:31 <dragorn> p_l: should be; isn't as far as I could fine
Oct 23 03:27:32 <dragorn> find
Oct 23 03:27:59 <dragorn> ron_frow_: other than being a bit clunky it's not bad
Oct 23 03:28:16 <dragorn> ron_frow_: if you've done reflection it's not so different
Oct 23 03:28:25 <ron_frow_> I just think its a pita... because it feels like a bunch of magic
Oct 23 03:28:34 <ron_frow_> yeah done tons of reflection
Oct 23 03:29:07 <KaiKai> how would i go about converting androids public key from base64encoded string to a format acceptable by openssl_verify?
Oct 23 03:29:46 <dragorn> what format is it before they base64 it?
Oct 23 03:30:49 <dragorn> (most ascii key formats are base64 encoded)
Oct 23 03:34:43 <ron_frow_> jni just feels unnatural
Oct 23 03:35:01 <ron_frow_> c++ fine, java fine... the integration point of the both... just odd
Oct 23 03:35:27 <ron_frow_> dragorn did you find every security restriction in android sdk is there with the ndk?
Oct 23 03:35:39 <ron_frow_> eg, still have to request network perms if you want to use network
Oct 23 03:35:45 <dragorn> yes
Oct 23 03:35:54 <dragorn> barring a few bugs
Oct 23 03:36:17 <dragorn> only one comes to mind, I reported it to google a few years ago and they said "will be fixed in 3.0, will not be fixed in 2"
Oct 23 03:36:36 <dragorn> namely, 'ping' used l2 sockets to inject pings, which bypassed the network restriction
Oct 23 03:37:23 <ron_frow_> I assume you informed them to help them fix rather than its actually something you need
Oct 23 03:37:54 <ashes> i'm looking for a kernel patch to increase audio volume for htc desire hd. can any of you point me in the right direction?
Oct 23 03:38:02 <dragorn> ron_frow_: yes
Oct 23 03:38:13 <dragorn> ron_frow_: it's just an interesting artifact; the attack surface is relatively limited
Oct 23 03:38:18 <cyclick2> To call an activity from a widget's button I can use the following intent: Intent intent = new Intent(context, DK.class);, but what do I need to do if I only want to call one of that Activity's method from the widget's button?
Oct 23 03:38:33 <ron_frow_> yeah so they implemented the restrictions in bionic?
Oct 23 03:38:46 <dragorn> ron_frow_: tho if I really wanted, i could write something that had external_storage only, and parsed your photos for gps and leaked your location over ping
Oct 23 03:39:16 <dragorn> cyclick2: unless it's a static method you can't call it w/out creating an activity
Oct 23 03:39:50 <dragorn> cyclick2: if you want to go directly to some functionality in another activity, you'd have to make an intent that took you there (in the other activity - have it inspect it's creation intent and jump to that mode of operation)
Oct 23 03:40:01 <dragorn> ron_frow_: well the network stuff is, afaik, pid+iptables
Oct 23 03:40:19 <ron_frow_> oh thats right
Oct 23 03:40:29 <ron_frow_> actually a LOT of the permission shit works exactly the same way
Oct 23 03:40:32 <ron_frow_> I forgot about that
Oct 23 03:40:52 <dragorn> ron_frow_: tho I haven't looked at it in a while.  The other restrictions - yeah probably, whatever they jailed each user with.  Since anything you exec lives under that wrappered jail
Oct 23 03:41:07 <dragorn> ron_frow_: I haven't looked into it heavily
Oct 23 03:41:37 <ron_frow_> I only found that out because of some stuff I fixed, but never realized because package manager hadnt reset-up the permissions
Oct 23 03:44:14 <cyclick2> thank you dragorn
Oct 23 03:49:26 <mrjeff> Anyone have time help me with LoaderManager?  Mines not watching for changes like it should
Oct 23 04:06:42 <cyclick2> Would using extras be the easiest way to determine where the intent was called from?
Oct 23 04:08:58 <sampullman> probably
Oct 23 04:13:51 <KaiKai> sorry dragon are you still here?
Oct 23 04:13:57 <KaiKai> im taking it straight from the developer panel
Oct 23 04:14:16 <KaiKai> Base64-encoded RSA public key
Oct 23 04:18:44 <invisiblek> so, if i go to repo upload, and things have been merged since my last rebase, it shows two commits. mine and something that must have been merged recently
Oct 23 04:18:56 <invisiblek> bleh wrong chan, sry
Oct 23 04:24:30 <scant> why won't this code to draw on a view on a surfaceview work?
Oct 23 04:24:31 <scant> https://gist.github.com/bd52ae94f112ee254ec5
Oct 23 04:26:22 <tophyr> any particular reason why an emulator wouldn't have internet access? the host does.
Oct 23 04:26:29 <tophyr> it's not in airplane mode
Oct 23 04:28:41 <tophyr> ah, it's probably trying to connect via my disconnected ethernet, instead of my connected wifi
Oct 23 04:31:28 <g00s> ron_frow_: http://www.phoronix.com/scan.php?page=news_item&px=MTIxMzA
Oct 23 04:31:47 <ron_frow_> kick ass!
Oct 23 04:31:54 <ron_frow_> I usually read phoronix every day
Oct 23 04:32:32 <brickhead> hey guys
Oct 23 04:33:35 <brickhead> I'm looking for a way to have a object tree, but disk based
Oct 23 04:34:12 <brickhead> for Jake Wharton's LRU Cache looked like what I needed, but it supports string values, I would need to persist objects
Oct 23 04:34:25 <JakeWharton> it supports anything
Oct 23 04:34:32 <ron_frow_> BAM
Oct 23 04:34:36 <ron_frow_> the man in person.
Oct 23 04:34:41 <JakeWharton> the API is a stream
Oct 23 04:35:07 <brickhead> Hey JakeWharton, I was looking at the test code as reference
Oct 23 04:35:47 <JakeWharton> the tests use strings for convenience
Oct 23 04:36:58 <JakeWharton> you should use the input stream methods to write whatever you want
Oct 23 04:37:20 <brickhead> ok thanks I'll look at it
Oct 23 04:45:39 <vanson2012> hey did anyone knows an equivalent of iOS's UINotificationCenter in Android??
Oct 23 04:46:07 <vanson2012> I wish to update my Fragment from a singleton manager async download
Oct 23 04:46:44 <ron_frow_> otto
Oct 23 04:46:54 <bitplane> Hi all, where do Android devs hang out and discuss new features / bugs etc? the bug tracker has been dead for ages, the google groups seem really quiet
Oct 23 04:47:05 <ron_frow_> but thats not part of android frameowrk out of box
Oct 23 04:48:17 <vanson2012> thanks!
Oct 23 05:08:06 <g00s> cool, jdbm3 is now an apache project. i used jdbm2, that was the bees knees
Oct 23 05:08:20 <g00s> works great with android
Oct 23 05:49:12 <Udmurt> Hi! I'm using android-maven-plugin. Created project with release archetype. And after added dependency to actionbarsherlock I unable to build entire project. On integration tests it fails to add actionbarsherlock's classes because whey already added. I do not add dependency to integration tests subproject. How to fix it?
Oct 23 06:18:51 <mrjeff> is there an easy way to set the background color of a checkbox?
Oct 23 06:19:00 <mrjeff> mines grey on white
Oct 23 06:23:30 <soa2ii> mrjeff: I don't know how but did you have a look at custom styles?
Oct 23 06:24:05 <soa2ii> I don't know how familiar you are with that … I'm not at all but it'd be my first guess.
Oct 23 06:24:46 <mrjeff> I'm not too familiar with styles at all
Oct 23 06:25:52 <mrjeff> i'm hoping I can just point to the right drawable?  i get the right image when using simple_list_item_multiple_choice
Oct 23 06:53:40 <tophyr> gat damn do i love me some emulator action
Oct 23 07:06:17 <gaz`> damn abs still dies on 4.0+
Oct 23 07:22:32 <f2prateek> i'm using this to return the device height and widht http://pastebin.com/JX0tDgCY. it works fine on my htc desire z, but on my tab 10.1 the height shows up as 1232*800, when i'm looking for 1280*800. i'm guessing it returns the total height minus the ics navigation bar, while i want the height for both combined
Oct 23 07:22:50 <f2prateek> desire z was on gingebread,m tab 10.1 on jellybean
Oct 23 07:26:26 <Ge0rG> f2prateek: you can't change the nav bar anyway, so why bother?
Oct 23 07:26:37 <Ge0rG> f2prateek: you could hard-code common values and detect your orientation as well ;)
Oct 23 07:27:00 <Ge0rG> ,oO( also, totally off-topic: is there any device available in europe that can replace the good old desire z? )
Oct 23 07:28:58 <f2prateek> f2prateek: i dont actually need it it to draw a view or anything, its so that the app automatically selects a device with matching dimensions from the available list for the user (http://bit.ly/ORuWAv)
Oct 23 07:29:27 <f2prateek> Ge0rG: what do you mean hardcode common values? like automatically interpret 1232 as 1280?
Oct 23 07:30:12 <Ge0rG> f2prateek: exactly
Oct 23 07:31:21 <f2prateek> might work, but if user is on custom rom then it wont work right? i know in AOKP you can change the hiehgt of the navbar so it could really be anything
Oct 23 07:32:29 <Ge0rG> f2prateek: nice app :D
Oct 23 07:32:52 <Ge0rG> f2prateek: but don't screenshots have the full screen res anyway? just look at the screenshot resolution...
Oct 23 07:33:03 <f2prateek> thanks, still lots to be done though :)
Oct 23 07:36:10 <f2prateek> yeah, but then that takes away option from the user, since i have to check it every time user pick a screenshot and there can be multiple device with the same size (eg: htc one x, galaxy nexus, galaxy s3), so either a) i have to prompt them for every picture, or b) automatically choose one for them, whichc means it will could end up using a deivce user didnt want
Oct 23 07:37:02 <f2prateek> all i'm trying to do is get the device size of the current deivce, find a device with matching size and set it as default the first time user runs the app
Oct 23 07:37:20 <f2prateek> after that if its not the one that they wanted, they can simply change it
Oct 23 07:37:38 <f2prateek> rather than do it for every pic
Oct 23 07:41:25 <evelyette> hi
Oct 23 07:41:47 <evelyette> is there an option in galaxy s3 to not lock screen while usb connected ?
Oct 23 07:41:51 <evelyette> and where is it
Oct 23 07:50:24 <pacior> maybe you lock screen through services or in android manifest
Oct 23 07:51:43 <f2prateek> evelyette: maybe look at the keep screen awake setting
Oct 23 07:51:48 <f2prateek> its under developer options
Oct 23 07:57:38 <pacior> have You had problems with asyncTask during invoking doInBackground? i am trying to download a file, but do this in a loop, waiting for tasks to be completed. seems i close the streams, but during reading data from last file it stops
Oct 23 07:58:26 <pacior> its rather network sending data question, i connect through http
Oct 23 07:59:24 <pacior> ok thanks it works now
Oct 23 08:03:41 <pacior> i have file in 10.1.1.7 on server. i try to download a file and it hungs during a connection. why?sometimes works sometimes not.
Oct 23 08:16:17 <siyb> is there a simple way to spot open cursors in android? i cannot find the cursors that cause this warning: Cursor finalized without prior close()
Oct 23 08:24:02 <mr6> siyb: just log everywhere you obtain one and where you close them
Oct 23 08:26:53 <siyb> mr6: that seems like a very inconvenient and painful task, isn't there a better solution? i cannot understand that there is no stacktrace indicating where the cursor was opened. i suspect changeCursor(...) to be defective (v4 supportlib)
Oct 23 08:27:54 <anthony_dev> guys, I have just started to learn how to develop for android. so I was wondering, is there anyway to write game without using this 9pixel thing?
Oct 23 08:29:02 <siyb> do you mean 9patch?
Oct 23 08:29:21 <anthony_dev> siyb: yes.
Oct 23 08:29:26 <siyb> and yes, you can write a game without using 9patches
Oct 23 08:29:35 <siyb> your question doesn't make any sense to me btw
Oct 23 08:30:07 <siyb> it's like saying can i eat a breakfast egg without drinking orange juice
Oct 23 08:31:13 <anthony_dev> I thought there will be some stretching mechanism so if I use something like drawImage(). and my game runs mainly for 1024x600. it will be scaled\streched to work on 1024x768 or even 800x480 etc.
Oct 23 08:32:17 <timroes> anthony_dev: the stretching can also be done with bitmaps, but then it is really stretching. You use 9path images, if you don't want to stretch your image, but want to extend it "clever", like enlarge only the body of it, but leave the corners the same
Oct 23 08:32:27 <mr6> siyb: my other suggestion is to code better
Oct 23 08:32:51 <anthony_dev> hmm
Oct 23 08:34:39 <gregtom6> hy all
Oct 23 08:34:54 <anthony_dev> timroes: thank you. I'll read more about this subject.
Oct 23 08:34:59 <mr6> siyb: seriosuly though, i have seen stacktraces that indicate where the cursor was opened
Oct 23 08:35:29 <siyb> mr6: well i only get warnings and i don't see where the cursor could still be open
Oct 23 08:35:39 <siyb> unless changeCursor doesn't do its job properly
Oct 23 08:35:48 <timroes> anthony_dev: do you want to make some real-time game or some round based/puzzle/what ever not needing 30fps game?
Oct 23 08:36:04 <gregtom6> how can I set the background color of the application?
Oct 23 08:36:26 <mr6> siyb: verify by closing it before calling change cursor
Oct 23 08:36:30 <mr6> that is valid anyway
Oct 23 08:37:44 <siyb> mr6: hmm i could do that, but the app is pretty big, will take a long time ;)
Oct 23 08:38:12 <siyb> but thanks for the suggestion
Oct 23 08:39:02 <mr6> heh well it's what i'd try.  I assuming you're not using ms notepad to write code so you probably can use search and replace functionality.
Oct 23 08:39:34 <anthony_dev> timroes: I already have in some way simple and not finished engine for (whoa!) j2me devices. so I thought it will be nice to convert that thing for android, because j2me has issues with music and sound effects. yes. I need fps and its kind of realtime engine (actually you can switch between rpg like mode and arcade)
Oct 23 08:40:10 <mr6> gregtom6: use a magic marker on your screen. good luck.
Oct 23 08:40:26 <timroes> okay, if you are coding the engine you, will need to read a lot about android :) otherwise I would have recommended, taking a look at andengine (an android gaming engine)
Oct 23 08:41:00 <timroes> and for real time games under android I would highly recommend you, to use OpenGL ES, instead of using the android Canvas stuff
Oct 23 08:41:16 <pacior> i recommend andengine
Oct 23 08:41:47 <siyb> mr6: that's a foreground color though :p
Oct 23 08:43:41 <mr6> siyb: it's fg and bg imo ;>
Oct 23 09:00:35 <anthony_dev> timroes: got any suggestions "which books to read" or something like that? because I dont like how they develop a game in Practical Android 4 Games Development (from Apress) book but it is seems the best what I found on my local market.
Oct 23 09:01:58 <timroes> i neither have found any really good book, but the developer documentation of android is in my opinion good enough on all topics, to make the last step yourself, (especially, if you allready developed some game framework)
Oct 23 09:02:09 <gregtom6> can I use android:background in the manifest?
Oct 23 09:02:19 <gregtom6> if yes, how?
Oct 23 09:02:25 <gregtom6> can't find any example
Oct 23 09:02:40 <Tee_Pee> not in the manifest, but in the layout
Oct 23 09:02:50 <anthony_dev> timroes: well, thanks :}
Oct 23 09:03:13 <Tee_Pee> You can apply a layout to an activity, like such
Oct 23 09:03:23 <Tee_Pee> android:theme="@style/MyStyle" >
Oct 23 09:03:23 <gregtom6> where is the layout? in every activity?
Oct 23 09:03:32 <Tee_Pee> You can apply that in the manifest to the whole application even
Oct 23 09:03:39 <gregtom6> okay, but I don't want a style
Oct 23 09:03:45 <Tee_Pee> in your values folder, you have a styles.xml
Oct 23 09:03:49 <gregtom6> I could set theme to the application
Oct 23 09:04:07 <gregtom6> should I define a style?
Oct 23 09:04:10 <Tee_Pee> yup
Oct 23 09:04:16 <Tee_Pee> define a style in styles.xml with
Oct 23 09:04:18 <gregtom6> I want only a green background
Oct 23 09:04:23 <gregtom6> okay
Oct 23 09:04:36 <Tee_Pee> <item name="android:windowBackground">"#00FF00"</item>
Oct 23 09:05:09 <Tee_Pee> styles are a tad tricky if you want backwards compatibility
Oct 23 09:07:03 <Tee_Pee> What's your min SDK version?
Oct 23 09:07:07 <Tee_Pee> and what version are you compiling with?
Oct 23 09:13:06 <gregtom6> Tee_Pee: I don't know, 2.2 I think
Oct 23 09:13:25 <gregtom6> I want a solution what will be okay in every other Android version
Oct 23 09:14:05 <Tee_Pee> here's what you want then
Oct 23 09:14:29 <Tee_Pee> in your res folder, you should have a values subfolder, as well as values-11 and values-14, correct?
Oct 23 09:15:20 <gregtom6> res has values, menu, layout folders, and drawable folders
Oct 23 09:16:12 <vadi2> How can I check if I'm doing a debug or a release build with ant? build.mode.debug does not seem to be working anymore
Oct 23 09:16:45 <Tee_Pee> alright greg, in that case, create a XML file called colors.xml in it
Oct 23 09:16:58 <Tee_Pee> (in values)
Oct 23 09:17:02 <gregtom6> Tee_Pee: okay
Oct 23 09:17:41 <Tee_Pee> and in it, you write <resources>   <color name="my_green_color_name">#00FF00</color>       </resources>
Oct 23 09:18:00 <Tee_Pee> just substitute the values with your real ones obviously
Oct 23 09:18:38 <Tee_Pee> Next up - create an xml called styles.xml (if it doesn't exist already)
Oct 23 09:18:56 <Tee_Pee> and in it, you write this:
Oct 23 09:20:07 <Tee_Pee> http://pastebin.com/y1hnE97W
Oct 23 09:20:53 <gregtom6> okay
Oct 23 09:21:04 <Tee_Pee> lastly, in your manifest, you apply the android:theme="@style/MyTheme" attribute to any activity (or the entire application)
Oct 23 09:22:02 <gregtom6> I couldn't write hex number values in the styles.xml, so this seems to be the only way to do this I think :)
Oct 23 09:22:32 <pacior> couldn't you just set android:background in activities layouts?
Oct 23 09:22:44 <pacior> how many activities you have?
Oct 23 09:23:10 <Tee_Pee> You could, but this is a better practice overall.
Oct 23 09:23:30 <Tee_Pee> It's a few extra steps, but it's neater :p
Oct 23 09:23:36 <pacior> yes it is, you set once for all:)
Oct 23 09:23:37 <gregtom6> okay
Oct 23 09:24:13 <pacior> is it through inheritance? all options in styles go to all layouts?
Oct 23 09:24:47 <Tee_Pee> nope, but if you say
Oct 23 09:25:33 <Tee_Pee> android:theme="@style/MyTheme" in the application section of the manifest file (Right below the android:icon, before the first <activity>)
Oct 23 09:25:34 <gregtom6> Tee_Pee: before you said this colors.xml thing, I tried this in styles.xml: http://pastebin.com/vASbvSaQ and it said "error: string types not allowed bla bla bla"
Oct 23 09:25:38 <Tee_Pee> it will apply to the entire application
Oct 23 09:26:12 <Tee_Pee> yeah I know, you can't set string values there... to tell you the truth I don't even know how you'd hard-code a hex value into there, but it's irrelevant as you should always separate it :p
Oct 23 09:26:25 <gregtom6> so I don't know how would that previous solution be good
Oct 23 09:26:42 <gregtom6> :D
Oct 23 09:26:45 <gregtom6> okay
Oct 23 09:28:17 <Tee_Pee> apparently, you just code it like #000000 without the quotation marks.
Oct 23 09:29:27 <pacior> thats great idea with styles, like css styles for entire document lol
Oct 23 09:30:25 <Tee_Pee> that's the point :p
Oct 23 09:30:29 <Tee_Pee> However,
Oct 23 09:30:35 <gregtom6> Tee_Pee: error: unable to start activity componentinfo    nullpointerexception
Oct 23 09:30:36 <gregtom6> wow
Oct 23 09:30:42 <Tee_Pee> If you want to make sure your app looks correct on API levels 11+ and 14+
Oct 23 09:31:03 <Tee_Pee> pastebin the stacktrace.
Oct 23 09:31:25 <Cook13> hey guys I got a strange issue. I have one main activity and a 2nd activity wich is used to scan NFC-Tags. So when I read a new tag I call finish in the 2nd activity and pass the uID via putExtra() to the mainActivity. The mainActivity calls onActivityResult(). The strange thing is that onActivityResult() is never called. It seems like calling finish in the 2nd Activity doesn't work for me.
Oct 23 09:31:55 <gaz`> grr trying to port some obj-c to android
Oct 23 09:32:03 <gaz`> but i suck at obj syntax
Oct 23 09:32:10 <pacior> couldn
Oct 23 09:32:38 <Tee_Pee> did you actually call setResult() in the 2nd activity?
Oct 23 09:32:41 <pacior> couldn't you just finish second activity, put exstra and read intent in first? you would have first working and second would be finished
Oct 23 09:32:49 <Cook13> yes I did Tee_Pee
Oct 23 09:32:55 <gregtom6> Tee_Pee: http://pastebin.com/CAKaNWgP
Oct 23 09:33:08 <Tee_Pee> and did you call the 2nd activity via startActivityForResult() rather than startActivity()
Oct 23 09:33:55 <Cook13> yes Tee_Pee it seems like it doesn't even finish my NFC-Activity
Oct 23 09:34:06 <pacior> gregTom6, what is in line RoboLOGO.java:63
Oct 23 09:35:50 <gregtom6> pacior: btn_connect.setOnClickListener(new View.OnClickListener() {
Oct 23 09:36:00 <Tee_Pee> Cook13 - Afraid I'm not exactly sure what would cause that... You tried logging the codepath I take it?
Oct 23 09:36:33 <Tee_Pee> greg - btn_connect isn't initialized.
Oct 23 09:36:41 <CVirus> I want to display multiple buttons on the screen and associate each one of them with an object in a HashMap .. How can I do so ?
Oct 23 09:36:49 <Tee_Pee> hence the null pointer exception :p
Oct 23 09:36:54 <Cook13> Tee_Pee , is it possible that pendingIntent could cause it?
Oct 23 09:37:02 <Cook13> I use it in the NFC Activity
Oct 23 09:37:29 <pacior> greg: try extracting it from layout through casting to Button class
Oct 23 09:37:57 <Tee_Pee> Cook13 - Honestly, beats the shit out of me. :p
Oct 23 09:38:13 <pacior> CVirus: put buttons in gridview using adapter
Oct 23 09:38:26 <gregtom6> Tee_Pee: It works with another theme
Oct 23 09:38:45 <Tee_Pee> what android version are you running your app with?
Oct 23 09:38:56 <gregtom6> Tee_Pee: if I change the style the problem will occurs, with the old style it works
Oct 23 09:39:08 <CVirus> pacior: I don't want to use a gridview .. the buttons will be shown on a camera preview on different locations
Oct 23 09:39:15 <CVirus> pacior: accroding to the sensor data
Oct 23 09:39:16 <gregtom6> Tee_Pee: 2.3.3
Oct 23 09:39:50 <gregtom6> pacior: what do you mean?
Oct 23 09:39:57 <pacior> cvirus: do you have them in layout
Oct 23 09:40:14 <CVirus> pacior: a relativelayout yeah
Oct 23 09:40:36 <pacior> gregTom6: Button btn_connect = findViewById(R.layout.id.button2);
Oct 23 09:40:52 <pacior> gregTom6:and then senonclicklistener
Oct 23 09:41:29 <gregtom6> so just but the "layout" word in that?
Oct 23 09:42:15 <pacior> for each activity you have layout assigned which is stored in file res/layouts regular xml file.
Oct 23 09:42:32 <pacior> read about creating activity in adroid developers website
Oct 23 09:43:21 <gregtom6> pacior: but it worked for me with a default style
Oct 23 09:43:40 <gregtom6> why it doesn't work now?
Oct 23 09:45:25 <pacior> read about observer pattern in java
Oct 23 09:46:38 <gregtom6> pacior: I don't want to read a lot because of changing a background color :) it seems to me like creating an elephant from a mouse
Oct 23 09:48:16 <gregtom6> at least you wrote that for me? or for another person?
Oct 23 09:49:33 <gaz`> damn
Oct 23 09:49:33 <pacior> paste entire code of robologo method containing line 63
Oct 23 09:49:39 <gaz`> i need someone who know obj-c and android
Oct 23 09:49:46 <gaz`> properly stuck ;/
Oct 23 09:51:48 <pacior> gregtom: then i will write it for you and others, ist not a secret
Oct 23 09:54:08 <gregtom6> pacior: I spoke with my friend, I remembered it could solve a very easy way
Oct 23 09:54:30 <gregtom6> inside LinearLayout tags:  android:background="#FFFFFF"
Oct 23 09:54:36 <gregtom6> it's the simplest way
Oct 23 09:55:44 <gregtom6> styles and colors.xml are unusable only for changing a background color, if I have a lot of problem caused by this
Oct 23 10:06:50 <ali__> whats the best way to communicate between service and activity. The service makes network calls
Oct 23 10:07:05 <ali__> should I offload it to a database
Oct 23 10:07:17 <Ge0rG> ali__: depends on the type of state you want to communicate
Oct 23 10:07:52 <Ge0rG> ali__: for things like connection status you can bind the service; if you want to "ping" in either direction without an answer, you can send broadcasts
Oct 23 10:08:50 <ali__> well my case the service makes calls to a remote webservice and that data changes frequently
Oct 23 10:09:11 <ali__> I need a way to send the parsed data back to the activity
Oct 23 10:11:46 <ali__> Ge0rG: I think I'm going to stick with the database approach. That way I can support offline mode as well.Thanks!
Oct 23 10:12:18 <Ge0rG> ali__: yeah, offline mode is a great idea
Oct 23 10:12:35 <appel1> sounds awful for batterylife to query a server frequently :)
Oct 23 10:12:35 <Ge0rG> ali__: also, please do not create a "downloading stuff" progressdialog which blocks everything :D
Oct 23 10:13:03 <ali__> Ge0rG: I hear you! I hate that stuff as well
Oct 23 10:24:08 <Mavrik> this might be a stupidly obvious question but anyway:
Oct 23 10:24:30 <Mavrik> if I register listener for GPS location in service, do I have to keep a CPU wakelock on to get updates while phone is locked?
Oct 23 10:25:09 <Ge0rG> Mavrik: no
Oct 23 10:25:21 <Ge0rG> Mavrik: you'll get proper updates without a wakelock
Oct 23 10:25:38 <Ge0rG> Mavrik: I'm using the periodic GPS updates in my app to wake up the app every half an hour or so
Oct 23 10:26:13 <Mavrik> Ge0rG: thanks :)
Oct 23 10:26:47 <pragma-> that reminds me, is there an app that will report which apps have wakelocks, so I can find these retarded apps and uninstall them?
Oct 23 10:27:57 <Ge0rG> pragma-: there is a cryptic pm command to see the active wakelocks, iirc
Oct 23 10:28:06 <Mavrik> pragma-: on newer Android (like 4.0+) you'll get those apps shown in battery usage screen
Oct 23 10:28:16 <Mavrik> if they lock CPU for noticable amount of time
Oct 23 10:30:35 <bubbleguuum> we enter "I want my device to do millions things without ever consuming any CPU" territory here
Oct 23 10:31:43 <pragma-> no, it's "i want my device to do certain things while i'm actively using it instead of when it's in my pocket and i don't care about it anymore"
Oct 23 10:35:34 <pacior> save state in database, so user can switch battery ( e.g. during walking in mountains )
Oct 23 10:38:26 <clever> Mavrik: when the battery app says the cell radio has been running x hours, is that just idle time, or actual time it spent transmiting something (a ping, data, anything)
Oct 23 10:39:46 <Mavrik> clever: IIRC it's the actual active time of radio but it could vary between android versions :\
Oct 23 10:40:16 <clever> what exactly is considered 'active' though?
Oct 23 10:42:53 <Mavrik> clever: sending data, receiving data, transmitting control data (switching cells, SMSes, establishing data connections)
Oct 23 10:43:16 <clever> ah, so it should be usefull, if its fully working
Oct 23 10:43:59 <clever> ive wondered if it would be feasible to put an actual current measurement sensor on each module, but that would depend on the hardware design and drive up costs too
Oct 23 11:22:09 <pacior> how about progressDialog? i show it on my listener using show() but it doesn't show. it is in method of asynctask which is invoked.
Oct 23 11:22:14 <gaz`> whats the best way of making a type liek tyepdef in android? i just wanna do a wrapper i guess aorund int somehow, inherit from Iteneger?
Oct 23 11:22:44 <pacior> when i invoke on OnCreate it works, otherwise ( later ) no.
Oct 23 11:23:09 <pacior> int doesn't inherit
Oct 23 11:23:15 <gaz`> bugger
Oct 23 11:23:16 <pacior> try Integer.valueOf
Oct 23 11:23:39 <gaz`> im trying to make a type
Oct 23 11:23:42 <gaz`> based on thi obj-c
Oct 23 11:23:43 <gaz`> typedef unsigned int TSPathVertexPosition;
Oct 23 11:23:56 <gaz`> so i want a type TSPathVertexPosition that is basiclaly an int
Oct 23 11:26:52 <pacior> so this is a wrapper. you can inherit from some numeric interface maybe?
Oct 23 11:27:16 <Cook13> hey guys I have Activity A which calls Activity B (ActivityForResult). Activity B starts a PendingIntent for the NFC-Sensor. When I read a NFC-Tag in Activity B the result should be passed to Activity B by the method onActivityResult. Is it possible that the callback doesn't work because of PendingIntent?
Oct 23 11:29:52 <Leeds> Cook13: erm... what does "starts a PendingIntent" mean?  do you mean the result should be passed back to Activity A?
Oct 23 11:32:26 <Cook13> Leeds I do this in activity B: pendingIntent = PendingIntent.getActivity(this, 0, new Intent(this, getClass()).addFlags(Intent.FLAG_ACTIVITY_SINGLE_TOP), 0); I'm not sure what it exactly does since the android doc doesn't explain that in the NFC-tutorial. And yeah the result should be passed to Activity A
Oct 23 11:32:40 <Cook13> I'm pretty sure that pendingIntent is the reason for not finishing activity B
Oct 23 11:32:52 <Cook13> (I call finisch() in activity B)
Oct 23 11:32:58 <Cook13> *finish
Oct 23 11:33:01 <Leeds> that line you quoted doesn't do anything at all
Oct 23 11:34:13 <Cook13> I use this in onResume in Activity B mAdapter.enableForegroundDispatch(this, pendingIntent, intentFiltersArray, techListsArray);
Oct 23 11:34:49 <Leeds> no point quoting random lines from your code when you don't really know what they do
Oct 23 11:36:08 <Cook13> Leeds, I worked through the NFC Android tutorial by google and didn't understand everything. That's why I'm asking =)
Oct 23 11:38:23 <RichardS> heya all
Oct 23 11:41:55 <RichardS> I have a ClassNotFoundException which I cant track down,  its telling me that my Activity_Start. but it is marked in the manifest as android:name=".Activity_Start" and when I control+click it it takes me to the class,  I'm compleatly lost
Oct 23 11:43:15 <pacior> try to debug
Oct 23 11:43:26 <RichardS> it does not start though
Oct 23 11:43:35 <RichardS> when i click debug it crashes instantly
Oct 23 11:43:40 <pacior> write numbers line by line
Oct 23 11:43:48 <pacior> no, write to console
Oct 23 11:44:02 <pacior> in onCreate first line, and second and so on
Oct 23 11:44:24 <RichardS> k will try
Oct 23 11:44:48 <torniker> Hello,  I have listView over VideoView, the problem is the focus. after app is loaded the focus goes to videoview, but i need to have focus on listview any suggestions?
Oct 23 11:44:50 <RichardS> it doesnt even get to onCreate
Oct 23 11:45:21 <RichardS> it just crashes with this
Oct 23 11:45:22 <RichardS> 10-23 13:44:55.115: E/AndroidRuntime(9791): java.lang.RuntimeException: Unable to instantiate activity ComponentInfo{com.richard.aau.rebound.tracker/com.richard.aau.rebound.tracker.Activity_Start}: java.lang.ClassNotFoundException: com.richard.aau.rebound.tracker.Activity_Start
Oct 23 11:45:25 <pacior> clean project
Oct 23 11:45:27 <RichardS> tried
Oct 23 11:45:30 <RichardS> no change
Oct 23 11:45:52 <lasserix> Is it possible to add views in code to the row view in the getView method of an arrayadapter?
Oct 23 11:46:36 <superlinux-hp> how can I set the general application screen size?
Oct 23 11:46:36 <pacior> RichardS:check android manifest?
Oct 23 11:46:45 <RichardS> thats where i thought the problem was
Oct 23 11:46:51 <superlinux-hp> i want a unified size for all layouts?
Oct 23 11:46:53 <RichardS> but i cant see an error there,  let me pastebin it
Oct 23 11:46:56 <torniker> here is xml of my layout http://pastebin.com/bXr3x6g8
Oct 23 11:47:20 <pacior> lasserix: i did so to the dridview in adapter getView method in class having arraylist
Oct 23 11:47:32 <lasserix> pacior cool
Oct 23 11:47:34 <lasserix> thanks
Oct 23 11:47:44 <RichardS> heres my manifest
Oct 23 11:47:45 <RichardS> http://pastebin.com/LSSmk9F5
Oct 23 11:47:49 <RichardS> i cant see any errors?
Oct 23 11:48:21 <RichardS> and Activity_Start is the class which it says is missing,  but it is not
Oct 23 11:48:23 <pacior> superlinux-hp: you have four different sizes
Oct 23 11:48:48 <RichardS> when i control click it, it takes me to the Activity_Start class
Oct 23 11:49:00 <superlinux-hp> pacior, i mean from the manifest can i do that?
Oct 23 11:49:20 <pacior> richards: could you remove         <activity android:name="com.richard.aau.rebound.Activity_TeamManagement" >
Oct 23 11:49:29 <RichardS> k, checking
Oct 23 11:49:37 <pacior> richards: othervise describe it properly
Oct 23 11:50:12 <pacior> superlinux-hp just set width and height in dp to minimum expected
Oct 23 11:50:32 <RichardS> same error
Oct 23 11:53:52 <Leeds> superlinux-hp: erm, you know your app is going to run on different sized (physical and resolution) screens, right?
Oct 23 11:53:56 <pacior> richards: android:name="Activity_start" instead of long package name in manifest
Oct 23 11:54:05 <RichardS> already tried,  same error
Oct 23 11:54:27 <superlinux-hp> Leeds, that's what's making my boss crazy!
Oct 23 11:54:56 <superlinux-hp> on Samsung Geo everything works fine
Oct 23 11:55:06 <superlinux-hp> on S3 no
Oct 23 11:55:24 <superlinux-hp> i get heaps of crashes
Oct 23 11:55:52 <Leeds> then debug the crashes
Oct 23 11:56:00 <superlinux-hp> and on Geo : i can see the grid view
Oct 23 11:56:05 <superlinux-hp> on S3: i don't
Oct 23 11:56:12 <superlinux-hp> Leeds, i did
Oct 23 11:56:32 <superlinux-hp> but that's weird about the gridviews of images
Oct 23 11:58:15 <pacior> could you getmetrics and get width and height of device?
Oct 23 11:59:28 <Leeds> superlinux-hp: you're not making much sense...
Oct 23 11:59:55 <superlinux-hp> i am telling you what's happening to me
Oct 23 11:59:58 <pepigno75> hi, i have a problem with background layout. I have this layout http://pastebin.com/UUVPRnnf and this 9png background https://www.dropbox.com/s/spnizexd5munppe/ballondiary.9.png
Oct 23 12:00:06 <superlinux-hp> Leeds, got me?
Oct 23 12:00:15 <Leeds> no, absolutely not
Oct 23 12:00:33 <pepigno75> Result is here https://www.dropbox.com/s/u64cbiy40vf1qqb/risultato.png  TextView have a top margin
Oct 23 12:01:04 <pepigno75> i don't understand because layout have this topmargin
Oct 23 12:01:05 <superlinux-hp> but i press where they are supposed to be in the grid and i get the event of clicking working
Oct 23 12:03:47 <lasserix> you want padding
Oct 23 12:13:41 <pacior> how to enrypt entire code? can i put private key in java file to decrypt files needed by my application?
Oct 23 12:14:01 <pacior> how simple is decompiling apk file?
Oct 23 12:15:20 <lasserix> you want to use progaurd
Oct 23 12:15:51 <lasserix> but there is no way to completly "safeguard" your code or any data
Oct 23 12:15:56 <lasserix> you can make it very difficult
Oct 23 12:16:19 <pacior> thanks
Oct 23 12:16:36 <lasserix> *You can make it difficult, this is called obfuscation
Oct 23 12:18:58 <pacior> lasserix: yeah i found this. do you know if it changes strings names too?
Oct 23 12:19:13 <pacior> sorry. strings values
Oct 23 12:20:26 <evident> Can I access logcat messages on my device somehow that happened in the past? I really rarely notice a bug in an app and would like to find out, what exactly crashes. All apps I found that log the logcat don't allow me to go back a few minutes to see earlier messages...
Oct 23 12:21:09 <pacior> evident: ctrl+a, save as and open in notepad
Oct 23 12:21:53 <monsti> hi
Oct 23 12:22:25 <monsti> is there a getDimensionPixelSize where i can give a namespace string?
Oct 23 12:22:26 <Leeds> evident: AFAIK it's a buffer - it only goes back so far
Oct 23 12:24:05 <evident> sure saving logcat out of eclipse works fine, I don't need to go back very far... but I don't want it when I am connected via adb...
Oct 23 12:24:27 <evident> I've found some apps taht save the logcat, but only when these apps are running and visible...
Oct 23 12:24:39 <evident> But I guess I will have to check some more... :)
Oct 23 12:26:47 <brian_de> did anybody experience the datepicker/timepicker on a sony ericsson with white text on white background?
Oct 23 12:27:20 <brian_de> ... and if yes, is there a non-hacky way to get this look acceptable like on all other devices? ;)
Oct 23 12:51:33 <monsti> is there a way to get the id of a view from a library?
Oct 23 12:51:41 <monsti> e.g. by a string
Oct 23 12:52:32 <alankila> the IDs are not exactly strings, but you can pass an instance of R.id and maybe use introspection to look up the ID against it
Oct 23 12:52:39 <alankila> sorry, not pass an instance but pass the reflection of R.id
Oct 23 12:52:41 <monsti> int resID = getResources().getIdentifier("button_%i",  "id", getPackageName());
Oct 23 12:52:49 <monsti> got it ;
Oct 23 12:57:36 <pepigno75> anyone can help me
Oct 23 13:00:38 <wei2912> pepigno75, your question?
Oct 23 13:01:51 <pepigno75> i have this layout http://pastebin.com/UUVPRnnf
Oct 23 13:01:55 <pepigno75> simple RelativeLayout
Oct 23 13:02:09 <lov> and?
Oct 23 13:02:10 <pepigno75> background is this https://www.dropbox.com/s/spnizexd5munppe/ballondiary.9.png a 9png
Oct 23 13:02:26 <pepigno75> i have result: https://www.dropbox.com/s/u64cbiy40vf1qqb/risultato.png
Oct 23 13:02:34 <superlinux-hp> can I set using code the wallpaper without making it stretched on multiple screens?
Oct 23 13:02:35 <pepigno75> there is a topmargin
Oct 23 13:03:11 <lov> superlinux-hp: huh?
Oct 23 13:03:23 <superlinux-hp> lov, yes
Oct 23 13:03:37 <lov> superlinux-hp: your question doesn't make any sense. Please try again.
Oct 23 13:03:48 <lov> pepigno75: get to the point.
Oct 23 13:04:19 <wei2912> lov, he means that he wants to set a wallpaper without it being stretched on multiple screens
Oct 23 13:04:20 <lov> pepigno75: if you're not sure what's affecting your layouts, try using hierarchyviewer
Oct 23 13:04:26 <wei2912> but stretched from one end to another?
Oct 23 13:04:26 <pepigno75> if i use background ballondiary.9.png content of RelativeLayout have a topmargin
Oct 23 13:04:40 <superlinux-hp> lov, what i want is to set the wallpaper programatically . but i want the wallpaper not scrolling after setting it.
Oct 23 13:04:54 <lov> hmm.
Oct 23 13:05:11 <superlinux-hp> i can set it using code. but i cannot hold the scroll
Oct 23 13:05:17 <lov> sorry, I don't know how that's handled. I'm assuming that you'll have to take the device display height/width and scale appropriately to your crop size.
Oct 23 13:05:52 <superlinux-hp> lov all the images are set on a default size.
Oct 23 13:06:08 <superlinux-hp> on a chosen well known size
Oct 23 13:06:22 <lov> superlinux-hp: ok. Sorry, I don't know enough about this to help you out :(
Oct 23 13:06:36 <superlinux-hp> ok.. thanks mate
Oct 23 13:06:41 <superlinux-hp> i understand
Oct 23 13:06:47 <pepigno75> this is strange, because if png is background why content of RelativeLayout have margins?
Oct 23 13:07:07 <wei2912> ah damn
Oct 23 13:07:20 <wei2912> wrong channel... thought this was a linux help channel so i gave that answer, sry
Oct 23 13:07:22 <pacior> can you set margintop to -10dp?
Oct 23 13:07:26 <pacior> od padding?
Oct 23 13:07:39 <pepigno75> pacior is  a bad solution
Oct 23 13:08:57 <pacior> no i am a human beeing, thanks for advertisement:)
Oct 23 13:09:50 <pepigno75> eheh
Oct 23 13:16:05 <Tee_Pee> I have a textview with a compound drawable, which right now appears rather small... is there any way to control the size?
Oct 23 13:16:23 <Tee_Pee> I set it programatically via textView.setCompoundDrawablesWithIntrinsicBounds(new BitmapDrawable(mContext.getResources(), avatar), null, null, null);
Oct 23 13:26:45 <Tee_Pee> God... Perhaps someone remember what method you call for this, I knew it the other day but can't recall where I read about it
Oct 23 13:27:02 <Tee_Pee> normally the viewpager keeps only the 2 adjacent fragments in memory... but there's a way to increase that
Oct 23 13:28:00 <Tee_Pee> setOffscreenPageLimit
Oct 23 13:28:04 <Leeds> setoffscreenpa... yeah :)
Oct 23 13:29:10 <lov> View.AccessibilityDelegate cannot be resolved to type
Oct 23 13:29:13 <lov> compiling against API level 15
Oct 23 13:29:16 <lov> fuck you ADT.
Oct 23 13:33:15 <lasserix> I can set this in styles, but is there a way to set android:panelFullBackground in code?
Oct 23 13:53:33 <snooplsm> I have a popup window that is showing a webview.  If I call loadUrl on the webview, sometimes the data doesn't load.  Know what the deal is with that?
Oct 23 13:54:48 <Tee_Pee> So... I get a funny crash - I have an activity, in it I execute an asynctask and implement an onSuccess and onFailure method. In the onFailure method I want to show a dialog showing that the operation failed. However, here's the scenario
Oct 23 13:55:27 <Tee_Pee> Open activity, asynctask starts, finishes, error pops up <---this is fine                        Open activity, asynctask starts, close activity, asynctask finishes, application crashes
Oct 23 13:55:52 <snooplsm> you need to check if the activity is null
Oct 23 13:56:00 <snooplsm> are you in a fragment?
Oct 23 13:56:17 <Tee_Pee> nope
Oct 23 13:57:09 <snooplsm> well what exception you getting?  NPE?
Oct 23 13:57:38 <Tee_Pee> I honestly can't recall, I just remembered that it happened 2 days ago, haven't had time to recreate it today
Oct 23 13:57:43 <Tee_Pee> but I imagine it's a NPE
Oct 23 13:58:12 <Tee_Pee> I know that a view has method isActive(), but that requires api level 11, and I'm in activity anyhow
Oct 23 14:00:51 <sdaq> hello
Oct 23 14:08:28 <evancharlton> Tee_Pee: cancel the asynctask in onPause, check for canceled before showing the dialog
Oct 23 14:08:58 <lasserix> tee_pee why not use an asynctask loader
Oct 23 14:09:03 <lasserix> it'll handle life cycle stuff for you
Oct 23 14:10:12 <lasserix> you know? cause its always so much fun to refactor
Oct 23 14:10:29 <mrenouf> +1
Oct 23 14:10:40 <Tee_Pee> lasserix - I probably ought to implement a loader one of these days...
Oct 23 14:10:57 <Tee_Pee> Gotta finish up all the actual content of the application, then I'll work on optimization and debugging.
Oct 23 14:12:29 <pacior> in asynctask i have public variable finished and checking in in a while loop
Oct 23 14:12:53 <pacior> additionally running all class in a looped handler
Oct 23 14:13:00 <pacior> pretty nice huh?
Oct 23 14:17:09 <pacior> no its not nice. do you know how to solve synchronization with dialog?
Oct 23 14:21:41 <polomolo777> hello, where can I find the source for adbd?
Oct 23 14:21:58 <evancharlton> system/core/, I believe
Oct 23 14:23:51 <polomolo777> That seems about right, thank you.
Oct 23 14:25:57 <sdaq> polomolo777: shalom
Oct 23 14:27:16 <Leeds> oy vey
Oct 23 14:36:30 <tophyr> polomolo777: what are you working on with adb?
Oct 23 14:41:33 <polomolo777> I worked for nokia, we are developing ndb and i needed some reference.
Oct 23 14:42:27 <wei2912> polomolo777, why didn't nokia choose android?
Oct 23 14:42:46 <wei2912> polomolo777, wait sry there are nokia android phones
Oct 23 14:43:56 <tophyr> polomolo777: ah, cool
Oct 23 14:44:01 <timroes> wei2912: there are nokia android phones?
Oct 23 14:44:08 <Leeds> there are no Nokia Android phones
Oct 23 14:44:24 <fakingfantastic> Anyone here ever use SlidingMenu, I'm having issues getting it working
Oct 23 14:44:39 <wei2912> timroes, last time i believe, but not now
Oct 23 14:44:41 <fakingfantastic> it's all running fine, just not actually doing the slide
Oct 23 14:45:11 <wei2912> after i read an article... so now to rephrase my question - why is nokia abandoning android?
Oct 23 14:45:26 <timroes> because they want to ruin their company?
Oct 23 14:45:36 <wei2912> timroes, ah sry it is a hacked version lol
Oct 23 14:45:40 <wei2912> http://www.ubergizmo.com/2012/08/nokia-n9-runs-on-android-4-1-jelly-bean/
Oct 23 14:45:58 <timroes> i mean look at their sale figures, and stocks
Oct 23 14:49:01 <polomolo777> Nokia has a very vast legacy, "going under" with android is not they way to survive as a company that only develops cellular phones(look at HTC's figures).
Oct 23 14:51:23 <dot8> I have a question: constructor
Oct 23 14:51:23 <dot8> I do this:
Oct 23 14:51:23 <dot8> String mlat = String.valueOf(latitude);
Oct 23 14:51:23 <dot8> start a AsyncTask:
Oct 23 14:51:23 <dot8> new PostData().execute(mlat);
Oct 23 14:51:23 <dot8> but wenn I put it here:
Oct 23 14:51:24 <dot8> nameValuePairs.add(new BasicNameValuePair("lat", mlat));
Oct 23 14:51:24 <dot8> it is Null
Oct 23 14:51:42 <dot8> what is wrong?
Oct 23 14:52:41 <matt_j> Hey all :) newbie android dev here
Oct 23 14:53:00 <matt_j> does anyone know of an eclipse tool to convert a string literal in my code to a reference and a new string in strings.xml?
Oct 23 14:53:10 <matt_j> thereby saving me some typing :3
Oct 23 14:54:02 <dragorn> if there is, let me know ;P
Oct 23 14:54:03 <matt_j> lol wait...
Oct 23 14:54:06 <matt_j> I thin it's built in
Oct 23 14:54:49 <matt_j> well kinda, but it doesn't replace it in your code
Oct 23 14:54:58 <matt_j> better than nothing though
Oct 23 14:58:31 <tophyr> matt_j: Source -> Externalize Strings... ?
Oct 23 14:58:51 <tophyr> dot8: gist.github.com, paste full code
Oct 23 15:00:46 <Newb101> Hi, I have my imei number can someone recommend a free site to unlock my phone?
Oct 23 15:01:30 <Leeds> Newb101: not here, please leave
Oct 23 15:02:17 <dot8> tophyr: https://gist.github.com/3939259
Oct 23 15:02:24 <matt_j> tophyr: yeah, that's close to what I want. I was hoping it'd convert my existing highlighted code to a getString(...) for me ;)
Oct 23 15:02:42 <Newb101> Leeds, your every where cant get away from you
Oct 23 15:02:47 <Leeds> correct
Oct 23 15:03:04 <tophyr> matt_j: Ah :) sounds like a decent project for a contribution to AOSP/ADT ;)
Oct 23 15:03:11 <matt_j> haha
Oct 23 15:03:11 <Newb101> like nightmare on elm street
Oct 23 15:03:13 <matt_j> I'm a newb
Oct 23 15:03:14 <tophyr> Leeds is like Jesus
Oct 23 15:03:18 <matt_j> converted to the dark side from iOS
Oct 23 15:03:19 <tophyr> everywhere and nowhere
Oct 23 15:03:24 <Newb101> freddie krugger
Oct 23 15:03:28 <matt_j> at work I will be the only guy here who knows both :D
Oct 23 15:03:31 <Leeds> tophyr: I help those who help themselves?
Oct 23 15:03:36 <tophyr> Leeds: that too
Oct 23 15:03:42 <tophyr> matt_j: makes you very valuable :)
Oct 23 15:03:46 <matt_j> yeah :D
Oct 23 15:03:51 <matt_j> got 5 weeks to make an enormous project
Oct 23 15:03:59 <matt_j> so I will basically become a guru in 5 weeks :D
Oct 23 15:04:00 <matt_j> woo
Oct 23 15:04:07 <MDijkstra> heh
Oct 23 15:04:15 <Frigolit> i wish i had time to put on an enormous project :(
Oct 23 15:04:28 <Frigolit> i really want to make a game
Oct 23 15:04:40 <frankbro|work> I would've though that setting a view's visibility to GONE and notifying an adapter would remove it from being displayed but it doesn't. Is there a better way?
Oct 23 15:05:01 <MDijkstra> UIViewController!=Activity is the only thing I can say
Oct 23 15:05:33 <MDijkstra> and design your app with fragments from the start, don't think you can easily add them later
Oct 23 15:05:37 <tophyr> matt_j: GL with it :) loads of fun. biggest tip i have for the iOS->Android transition is do not underestimate the testing workload. the fragmentation of the android devices provides for a LOT of "wtf?!" moments compared to iOS
Oct 23 15:06:20 <tophyr> not only device size differences, but the various versions of Android and especially the varying preloaded apps, can cause you a load of headaches if your app does any inter-app comm via intents
Oct 23 15:06:46 <Leeds> matt_j: does that mean we're going to see you around a lot
Oct 23 15:07:09 <dragorn> frankbro|work: gone has always removed it from the view for me
Oct 23 15:07:43 <frankbro|work> dragorn: even adapter?
Oct 23 15:07:58 <frankbro|work> I did the notify thing too
Oct 23 15:08:31 <Leeds> what are you trying to remove it from?
Oct 23 15:08:34 <dragorn> frankbro|work: i've never had a view not disappear when i set it to view.gone, but maybe not in the same situation you're in
Oct 23 15:08:56 <frankbro|work> Leeds: a custom ArrayAdapter
Oct 23 15:09:16 <Leeds> which is displaying in?
Oct 23 15:09:21 <tophyr> dot8: your PostDataActivity.mlat variable is not the same as your mlat variable inside onClick()
Oct 23 15:09:37 <frankbro|work> Leeds: a relativelayout
Oct 23 15:09:42 <tophyr> you need to set it from the 'params' variable. bone up some more on Java and how variables and scope work
Oct 23 15:09:56 <Leeds> why are you using an adapter for a relativelayout?
Oct 23 15:10:10 <tophyr> ... and, *how* are you using an adapter for a RL
Oct 23 15:10:19 <Leeds> yeah, that too
Oct 23 15:10:41 <frankbro|work> Leeds: to place several things around it
Oct 23 15:11:08 <frankbro|work> wait its an adapterView
Oct 23 15:11:08 <Leeds> so you're just using it as an array?
Oct 23 15:11:15 <Leeds> okay, bored now
Oct 23 15:11:39 <frankbro|work> relativelayout holds the AdapterView<ListAdapter>, sorry
Oct 23 15:12:32 <Leeds> still doesn't make any sense, but good luck with that
Oct 23 15:12:38 <tophyr> AdapterView is abstract i thought. what actual *view* are you using?
Oct 23 15:13:43 <dot8> tophyr: thx, I will have a look on it
Oct 23 15:14:21 <frankbro|work> tophyr: https://github.com/vieux/Android-Horizontal-ListView
Oct 23 15:14:44 <matt_j> Leeds: possibly lol
Oct 23 15:14:55 <matt_j> maybe not now since I work with a guru, but later perhaps :)
Oct 23 15:15:06 <Leeds> ah... oh well :)
Oct 23 15:19:42 <matt_j> whoa actually extract string did do a getString
Oct 23 15:19:43 <matt_j> woo
Oct 23 15:35:55 <sdaq> OpenGLES is easier to program than canvas... agree?
Oct 23 15:40:48 <genio> I'm sure this is a FAQ, but Googling around hasn't yielded much in the way of a concrete answer.  Is there some sort of accepted design pattern framework (MVC/MVP) and testing platform?
Oct 23 15:44:34 <xlinearx> Hey, I have a quick question.  I keep forgetting the workaround for this issue: if I have two elements in a layout XML that reference each other, what special syntax do I use?  For instance, a Button has layout_above="@id/the_text_view", and TextView has nextFocusDown="@+id/the_button".  No matter which one I define first, lint whines that it can't find the other one (an issue that was solved years ago with respectable parsers).  Do
Oct 23 15:46:50 <ubert_mac> hi, Im a android newb, yaaaaaay :)
Oct 23 15:47:16 <evancharlton> xlinearx: @+id/someId defined someId if it doesn't otherwise exist. @id/someId refers to an already-existing ID
Oct 23 15:48:18 <ubert_mac> I m working on an existing project, trying to understand how pngs are pulled into the project, I don't see the png file referencd ANYWHERE, its weird
Oct 23 15:49:14 <Leeds> ubert_mac: are they not in the res directory, probably res/drawables(-something)?
Oct 23 15:49:29 <ubert_mac> AHHHHHHH it references them without .png extension, so i cant find with search
Oct 23 15:49:53 <ubert_mac> they are, but Im trying to see how they get pulled into the code
Oct 23 15:50:13 <Leeds> are you asking how Android code references images?
Oct 23 15:51:08 <ubert_mac> yes, starting to understand though
Oct 23 15:51:22 <ubert_mac> apparently they are turned into reference able resources via R.java or seomthing
Oct 23 15:51:25 <ubert_mac> odd, but Ill take it! :)
Oct 23 15:51:29 <Leeds> correct
Oct 23 15:51:58 <ubert_mac> i dropped a png in the folder and will now turn that into a tppable button on the bottom toolbar
Oct 23 15:55:11 <xlinearx> I asked this again a little while ago but it might have gotten buried already.. does anyone know how to make two views in the XML reference each other? Since one has to be defined above another with @+id, and it's impossible for both of them to be above each other, how do you do it?
Oct 23 15:55:30 <evancharlton> you can also just ignore the linter
Oct 23 15:55:36 <jonp> really, really, silly question:
Oct 23 15:55:41 <jonp> why does adb not like shell scripting?
Oct 23 15:56:00 <evancharlton> jonp: ?
Oct 23 15:56:24 <tophyr> xlinearx: you don't have to define the id (@+) in the view whose id it is
Oct 23 15:56:36 <jonp> evancharlton: https://gist.github.com/3939634
Oct 23 15:56:50 <jonp> evancharlton: run `adb` in a bash loop: Failure. Run the ~same command "by hand": Success
Oct 23 15:56:55 <xlinearx> evancharlton: perhaps, but if I turn of lint, then I miss a bunch of other important things too.
Oct 23 15:56:56 <jonp> it's _really_ annoying
Oct 23 15:57:00 <tophyr> you can do <Button android:id="@+id/foo" android:layout_toLeftOf="@+id/this_view_is_below_me_in_xml"/>
Oct 23 15:57:30 <tophyr> then the below view, <FooBar android:id="@id/this_view_is_below_me_in_xml"/>
Oct 23 15:57:36 <tophyr> which, really, would be a very confusing view id
Oct 23 15:57:59 <xlinearx> tophyr: that's not what I mean. I have two views, and they reference each other.  The top one throws an error because the bottom one is defined below it, so it can't reference it.
Oct 23 15:58:01 <evancharlton> jonp: try adb wait-for-device uninstall maybe? not sure
Oct 23 15:58:20 <tophyr> jonp: lemme guess, it hangs?
Oct 23 15:58:26 <jonp> oh, it's because adb prints \r in the package name: http://stackoverflow.com/questions/7426070/adb-uninstall-command-doesnt-work-in-bash-shell-for-statement
Oct 23 15:58:37 <jonp> so i need to kill it: for p in `adb shell pm list packages Scratch. | tr -d "\r" | cut -c 9-` ; do adb uninstall $p ; done
Oct 23 15:58:42 <xlinearx> tophyr: Ah, ok.  So I use +'s in the top view, and don't use them in the bottom view
Oct 23 15:58:45 <jonp> for the love of $deity, why is adb using \r\n?
Oct 23 15:58:59 <evancharlton> jonp: windows compat? dunno
Oct 23 15:59:06 <jonp> yes, but i'm on OSX
Oct 23 15:59:07 <tophyr> xlinearx: correct, that should shut lint up
Oct 23 15:59:16 <tophyr> that's what i do and it doesn't yell at me, at least.
Oct 23 15:59:17 <jonp> it's presumably adbd on the device that's doing it
Oct 23 15:59:50 <xlinearx> tophyr: Ok, that's the trick then.  Thanks.
Oct 23 16:04:48 <monsti> how is "20dp" translated?
Oct 23 16:04:58 <monsti> getWidth() is an integer
Oct 23 16:05:09 <monsti> any idea how android makes the translation?
Oct 23 16:05:30 <ubert_mac> wouldn't it be dpi?
Oct 23 16:05:34 <ubert_mac> dots per inch
Oct 23 16:06:11 <monsti> you can set 20px, 20dp, 20dpi
Oct 23 16:06:24 <monsti> getWidth() is some translation of that
Oct 23 16:06:42 <monsti> how does this exactly work? there must me some device specific formula
Oct 23 16:08:38 <brickhead> Hello Guys
Oct 23 16:08:40 <evancharlton> getWidth() is pixels
Oct 23 16:09:10 <evancharlton> 20dp is (20 * (density / 160))px
Oct 23 16:09:31 <monsti> thx
Oct 23 16:10:21 <SimonVT> (density bucket, not actual device density)
Oct 23 16:10:47 <evancharlton> SimonVT: actual device density (as defined in build.prop)
Oct 23 16:10:59 <evancharlton> which is usually rounded to the bucket, but doesn't have to be
Oct 23 16:15:56 <monsti> evancharlton: is there a utility class that can do this translation?
Oct 23 16:16:23 <monsti> evancharlton: i am reading the attributes of a layout file in a library, because i want to ship the library without any ressources
Oct 23 16:16:37 <SimonVT> Doesn't have to be, but pretty much always is (on production builds anyway.. root users do strange stuff)
Oct 23 16:16:43 <evancharlton> monsti: I don't know how you'd do this without resources
Oct 23 16:16:44 <monsti> e.g. i have myFancyControl:layout_width="20dp"
Oct 23 16:17:01 <monsti> i know how to do this - but in this case i only get the string "20dp"
Oct 23 16:17:10 <monsti> which is quite odd
Oct 23 16:17:57 <evancharlton> generally speaking, you shouldn't be doing this conversion yourself
Oct 23 16:18:17 <evancharlton> you should use Resources#getDimensionPixelSize(int) or Resources#getDimensionPixelOffset(int)
Oct 23 16:18:32 <monsti> yeah that would require a resource file
Oct 23 16:18:47 <evancharlton> just make your library take in a context and the <dimen> resource ID which points to the desired dimension
Oct 23 16:18:49 <monsti> which i am trying to avoid - because i'd like to ship a jar and no android library
Oct 23 16:19:05 <SimonVT> So the user passes in a layout resources?
Oct 23 16:19:14 <monsti> in custom attributes
Oct 23 16:19:22 <monsti> e.g. i have myFancyControl:layout_width="20dp"
Oct 23 16:19:22 <evancharlton> SimonVT: no, pass in a <dimen> value
Oct 23 16:19:26 <Napalm> is it me or does ViewPager have a major leak problem. Yes when the view is destroyed it removes its views. But the removal of a fragments view doesn't destroy the fragment and the destoryItem of the adapter is never called.
Oct 23 16:19:38 <monsti> earlybirdd: yea a dimension value
Oct 23 16:19:43 <monsti> but there is no paser for that :(
Oct 23 16:19:58 <evancharlton> monsti: er, can't you define that attribute to be a dimension and not a string?
Oct 23 16:20:06 <evancharlton> and the platform will parse it automatically
Oct 23 16:20:34 <monsti> evancharlton: i can - but there is only attrs.getAttributeResourceValue(namespace, attribute, 0)
Oct 23 16:20:48 <monsti> or  attrs.getAttribute(namespace, attribute, 0) .. or int/bool
Oct 23 16:21:49 <monsti> evancharlton: check this http://developer.android.com/reference/android/util/AttributeSet.html - maybe i didn't find it
Oct 23 16:21:59 <evancharlton> monsti: alright, so what your clients should be doing is <dimen name="fancy_layout_width">20dp</dimen> and myFancyControl:layout_width="@dimen/fancy_layout_width"
Oct 23 16:22:08 <romainguy> monsti: first you need to obtain a TypedArray
Oct 23 16:22:24 <evancharlton> listen to romainguy :)
Oct 23 16:22:29 <romainguy> use context.obtainStyledAttributes(attrs, ...)
Oct 23 16:22:34 <monsti> romainguy: yeah but that's my point - i don't have the resource in the library
Oct 23 16:22:51 <monsti> romainguy: well can you define a typedarray by code?
Oct 23 16:23:06 <romainguy> you want to use attributes without using resources?
Oct 23 16:23:11 <monsti> romainguy: yes
Oct 23 16:23:32 <romainguy> that doesn't make sense
Oct 23 16:23:33 <monsti> romainguy: i'd like to avoid linking to a library - i'd like to have a jar file
Oct 23 16:24:12 <romainguy> you cannot use the resources APIs without resources
Oct 23 16:24:59 <SimonVT> If you don't want resources at all, you need to put your xml in res/raw/ or something, and parse it manually..
Oct 23 16:25:15 <monsti> SimonVT: that's what i am asking for
Oct 23 16:25:35 <monsti> SimonVT: we have a parser for int, string, bool, resources - but none for dimensions :(
Oct 23 16:25:53 <monsti> is this true?
Oct 23 16:25:53 <SimonVT> So make one
Oct 23 16:26:09 <evancharlton> monsti: I'm curious how you're planning on obtaining the display density without android APIs
Oct 23 16:26:09 <Napalm> romainguy: have you noticed that ViewPager doesnt call its adapters destroyItem. So for a instance of FragmentPagerAdapter the fragments remain in the FragmentManager
Oct 23 16:26:24 <monsti> evancharlton: my lib is an android library
Oct 23 16:26:32 <evancharlton> jackiechanwtf.jpg
Oct 23 16:26:43 <monsti> evancharlton: but i don't want to force the user to include it as source
Oct 23 16:27:11 <evancharlton> what does using android.content.res.* apis have to do with source?
Oct 23 16:27:53 <monsti> i don't know how this is related to my question - i alredy can get the strings in my custom attributes
Oct 23 16:28:23 <SimonVT> Napalm: destroyItem doesn't destroy/remove the Fragment
Oct 23 16:28:27 <SimonVT> It just detaches it
Oct 23 16:28:33 <SimonVT> The fragment is still managed by the fragmentmanager
Oct 23 16:28:52 <monsti> ok thank you guys i will investigate this more or create my own parser
Oct 23 16:31:01 <Escherial> hmm, i've included jodatime as an external library in my android project, yet i'm getting a NoClassDefFound exception when i try to use ISODateTimeFormat
Oct 23 16:31:10 <Escherial> (using eclipse, by the way)
Oct 23 16:31:21 <Escherial> is there some chance that jodatime's not getting compiled into the apk or something?
Oct 23 16:31:32 <SimonVT> Probably, depending on how you added it
Oct 23 16:31:56 <SimonVT> if you did it the right way (just put the jar in libs/, adt handles the rest), then it should be included.. if you didn't do that, you probably did it wrong
Oct 23 16:37:53 <fakingfantastic> Gang, I have an activity that loads a fragment. The fragment has a button. The button click needs to fire a method on the Activity. How do I access the activity from the button?
Oct 23 16:38:09 <Napalm> SimonVT: sorry i was talking about FragmentStatePagerAdapter which does call remove
Oct 23 16:38:21 <DysonReturns> I'll be rebelious one in the gang who dies first.
Oct 23 16:38:35 <fakingfantastic> Should the button on click be defined in the activity?
Oct 23 16:39:03 <SimonVT> The fragments article should be read, as it explains how to call back to the activity
Oct 23 16:39:35 <Napalm> SimonVT: not following? I know how to call back to the activity. Thats not the issue here.
Oct 23 16:39:57 <SimonVT> That's fakingfantastic's issue
Oct 23 16:40:11 <Napalm> oh sorry
Oct 23 16:40:15 <Napalm> didnt see that
Oct 23 16:40:17 <fakingfantastic> haha sorry for the confusion
Oct 23 16:41:09 <arturaz> Hey. Anyone tried building android apps with scala?
Oct 23 16:41:43 <Napalm> but you get my point, destroyItem gets the reference to the object which is the fragment, so i cant call destroyItem directly and the fragments arraylist in the FragmentStatePagerAdapter is private so i cannot access it
Oct 23 16:41:44 <Escherial> SimonVT: ah, i did it through eclipse's project settings dialog...i guess that's the wrong way, then :\
Oct 23 16:42:06 <Escherial> SimonVT: thanks, i'll try putting it in libs/ as you suggested and seeing what happens
Oct 23 16:42:18 <SimonVT> destroyItem should be called when the fragment is not one of the cached views
Oct 23 16:47:40 <kwstasl> Hey. Although the API doc explicitly states that "a fragment must always be embedded in an activity" , has anyone tried nested fragments?
Oct 23 16:48:15 <SimonVT> It's not supported
Oct 23 16:48:25 <SimonVT> Might work in some cases, but likely to break
Oct 23 16:49:15 <mobile8> for mobile site development, i need a good simulator of the mobile phone browser
Oct 23 16:49:19 <mobile8> what can i best use for this?
Oct 23 16:50:39 <kwstasl> Thanks SimonVT! I see. any ideas on the cases it will break?
Oct 23 16:51:38 <kwstasl> in fact, I've tried creating nested fragments using a <fragment> tag in a fragment's layout, and the nested fragment is attached as defined in the layout.
Oct 23 16:52:58 <theTerry> I have dynamically nested fragments before, it works, but they were only simple views with no interaction other than a click
Oct 23 16:53:24 <CharlesWHChan> I am trying to figure out how to get LTE cell information from Android 4.x API. From the SDK, I can only find GsmCellLocation and CdmaCellLocation ... any suggestions? Thanks!
Oct 23 16:55:44 <SimonVT> I don't have a test case at hand, but people usually have issues when their activity is restored from saved state
Oct 23 16:56:19 <theTerry> ^makes sense
Oct 23 16:57:18 <kwstasl> theTerry, did you test it with runtime orientation changes?
Oct 23 16:58:00 <SimonVT> Something to do with the order in which fragments are restored
Oct 23 17:00:46 <kwstasl> yes, I read on stackoverflow that even the nested fragments are still related to the underlying Activity , something like that
Oct 23 17:00:51 <mobile8> for mobile site development, i need a good simulator of the mobile phone browser
Oct 23 17:00:54 <mobile8> what can i best use for this?
Oct 23 17:01:07 <SimonVT> Yes, fragments are attached to the activity
Oct 23 17:02:23 <kwstasl> so , maybe, one would have to keep the state in the 'parent fragment' and also remove the nested fragments when it is detached, in order to make this work.
Oct 23 17:02:44 <SimonVT> So if fragment 2 is in fragment 1, and fragment 2 is restored first, it'll complain that the layout it was added to doesn't exist
Oct 23 17:02:51 <Escherial> hmm, my cursorloader's not being notified of updates despite passing it the CONTENT_URI of the collection that's being modified...pretty sure that the insert is going through and that the contentprovider is notifying on the same URI, too
Oct 23 17:03:23 <kwstasl> yes, exactly
Oct 23 17:03:44 <Escherial> actually, wait, no, i'm not notifying on it :\ apologies
Oct 23 17:04:40 <SimonVT> You'll be fine in a simple app, probably
Oct 23 17:04:50 <SimonVT> But once it gets more complex, so does your fragment handling
Oct 23 17:05:22 <kwstasl> do you think that one could manage these issues with nested fragments? Or have you heard of any undefined behaviors and/or errors? Of course, there is always the risk of the app breaking in a next Android version.
Oct 23 17:05:24 <SimonVT> When you should really just keep it simple and not nest fragments
Oct 23 17:05:38 <Escherial> er, terribly sorry, but i actually am notifying on the URI and i've verified that the one i'm passing to CursorLoader's constructor is the same as the one on which i'm notifying
Oct 23 17:05:43 <Escherial> totally mystified at this point :\
Oct 23 17:05:54 <SimonVT> Nesting fragments is basically relying on undefined behavior
Oct 23 17:06:41 <monsti> that's the reason why i decided against fragments :(
Oct 23 17:06:55 <SimonVT> That's a pretty bad reason
Oct 23 17:07:36 <monsti> i don't judge on that :)
Oct 23 17:07:38 <kwstasl> true simon, literally speaking, it is not defined on the API :)
Oct 23 17:07:42 <monsti> it's a decission
Oct 23 17:07:47 <tnzr> anyone know how to get rid of the "Shape round rect too large to be rendered into a texture"
Oct 23 17:08:56 <kwstasl> monsti, they are a very important part of the framework from tablets. Just use them as the API says. I play on the edge cases for reasons that are really irrelevant now.
Oct 23 17:09:15 <kwstasl> *for tablets
Oct 23 17:09:27 <meetoman> hey guys
Oct 23 17:09:31 <meetoman> http://pastebin.com/XpC5c2BQ
Oct 23 17:09:38 <monsti> kwstasl: yeah a fragment is a FrameLayout ;)
Oct 23 17:09:46 <meetoman> I am keep getting this error, which I dont really know how to fix
Oct 23 17:10:38 <meetoman> nothing on google
Oct 23 17:10:52 <meetoman> would lack of memory cause this type of problem?
Oct 23 17:10:55 <meetoman> android.content.res.Resources$NotFoundException
Oct 23 17:11:12 <Escherial> i'm supposed to call setNotificationUri() on incoming Cursors in my ContentProvider's query() method, correct?
Oct 23 17:11:56 <Escherial> at least if i want them to requery when i notifyChange() from a ContentResolver, right? also, ContentResolver is a singleton, right?
Oct 23 17:12:06 <Escherial> there's no chance of notifying on the "wrong" CR?
Oct 23 17:14:02 <SimonVT> Your cursors should get the uri you used to query your contentprovider with
Oct 23 17:14:06 <SimonVT> As notification uri
Oct 23 17:15:10 <SimonVT> That isn't necessarily what uri you want to listen to updates to, but manage that where you implement your cursor, not in the contentprovider
Oct 23 17:15:58 <Escherial> hmm, then i shouldn't be calling setNotificationUri() in ContentProvider.query()?
Oct 23 17:17:15 <Escherial> i hate to post big blocks of code, but does this look roughly right for a listfragment that should be updating when its target changes?: http://paste.awesom.eu/tIU&ln
Oct 23 17:17:34 <Escherial> (you can ignore the package-level class; it's just for displaying the list items)
Oct 23 17:18:14 <Escherial> i'm not sure if there's something i should be doing in onCreateLoader() aside from just calling CursorLoader...i thought i'd followed the example to a tee (including even the comments :P)
Oct 23 17:18:19 <kwstasl> thanks guys! bb
Oct 23 17:18:49 <meetoman> anyone for my problem?
Oct 23 17:18:53 <meetoman> I can't seem to recreate it
Oct 23 17:19:31 <docmur> I'm editing or trying to edit this code, on line 15 what is edit doing?  There is no edit definded in the program anywhere, I'm assume edit is attached to bundle http://pastebin.com/NEvfbxMN
Oct 23 17:21:13 <Escherial> docmur: i'm guessing that it's a preferences editor, which was probably a member in the class from which you took that code
Oct 23 17:22:12 <Escherial> as a general rule of thumb, writing your own code rather than copying and pasting it (unless you're sure that there aren't dependencies in that code) is the way to go
Oct 23 17:22:39 <docmur> in this case I was told to take this code and work with it
Oct 23 17:22:42 <Escherial> case in point, your mysterious "edit" variable
Oct 23 17:22:44 <docmur> I actually have to merge other code into it
Oct 23 17:23:10 <docmur> so the first thing I have to do is go in and comment it
Oct 23 17:23:11 <Escherial> i hope for your sake that you're familiar with java
Oct 23 17:23:20 <mobile8> for mobile site development, i need a good simulator of the mobile phone browser
Oct 23 17:23:21 <docmur> I know Java but I've never done android development
Oct 23 17:23:23 <mobile8> what can i best use for this?
Oct 23 17:23:36 <SimonVT> edit is from SharedPreferences.edit()
Oct 23 17:24:21 <docmur> so basically a shared preference is a way to build up a set of actions and commit them all at once?
Oct 23 17:24:39 <Escherial> a set of key-value pairs, yes
Oct 23 17:25:38 <Escherial> you can then retrieve them later on...generally people store preferences that are shared by different components of the application there, hence the name
Oct 23 17:27:13 <docmur> So it would be the same as making a config file and just setting all the values in it, in effect
Oct 23 17:27:47 <drknzz> docmur: Essentially, but it's on your app's protected storage so other apps can only read/write to it with root permissions
Oct 23 17:28:01 <drknzz> AFAIK
Oct 23 17:28:09 <docmur> ya thats fine
Oct 23 17:28:14 <docmur> okay
Oct 23 17:28:34 <drknzz> Can anyone check my question on Parcelables? http://stackoverflow.com/questions/13035618/how-to-parcel-custom-list-on-android I'm getting a really weird behaviour
Oct 23 17:29:07 <Escherial> in order to change the collection, you generally first get a handle to the shared preferences object for your app via SharedPreferences prefs = context.getSharedPreferences(name, mode), then get a handle to the editor by calling Editor edit = prefs.edit()
Oct 23 17:31:33 <Sicp> I think you call the editor through SharedPreferences.Editor editor = prefs.edit()
Oct 23 17:32:00 <Escherial> right, that's more direct :)
Oct 23 17:32:21 <SimonVT> parcel.writeList((ArrayList<Song>)this);
Oct 23 17:32:25 <SimonVT> I'm surprised this works
Oct 23 17:32:36 <Escherial> if you just want to read preferences, you can do so from that prefs handle, e.g. prefs.getBoolean("firstRun", <default value if not found>)
Oct 23 17:33:02 <Sicp> Excess flood?
Oct 23 17:33:10 <Sicp> that's ...so 1996!
Oct 23 17:33:56 <drknzz> SimonVT: Well, i want my PlayList to behave like an ArrayList without having to wrap a ton of methods for a private ArrayList instance
Oct 23 17:34:46 <drknzz> That's the closest thing to it, without the cast it would make a recursive call to writeToParcel as this would implement Parcelable
Oct 23 17:34:53 <drknzz> Ot that's how i think it should of behave
Oct 23 17:34:57 <pacior> drknzz: logged in, what is the problem? Collections have many interfaces you can implement
Oct 23 17:35:57 <pacior> for example comparator
Oct 23 17:36:18 <drknzz> pacior: Basically, i am trying to pass a PlayList object around activities, it writes itself to a parcel fine, but on deserialization, a NullPointerException/BadParcelableException (Caused by a ClassLoader Exception) is raised
Oct 23 17:36:37 <drknzz> It can't find a ClassLoader for a string member of the Song class
Oct 23 17:37:18 <pacior> can't you use sqlite and map PlayList to database?
Oct 23 17:38:33 <drknzz> pacior: I'd rather not, not right now at least
Oct 23 17:38:45 <drknzz> I still have to think out a sane database schema for such a structure
Oct 23 17:39:41 <drknzz> Do you think making PlayList a linked list would alleviate the issue?
Oct 23 17:39:51 <pacior> drknzz: have you get correct class loader?
Oct 23 17:40:06 <drknzz> I call Song.class.getClassLoader()
Oct 23 17:40:16 <pacior> PlayList.class.getClassLoader()
Oct 23 17:40:17 <drknzz> passing in null would not change things
Oct 23 17:41:18 <drknzz> Hmmm, i believe a Song classLoader would be more appropiate, as i am tring to deserialize a Song arrayList
Oct 23 17:41:41 * drknzz secretly curses Java on his dreams
Oct 23 17:41:48 <Komak57> Does anyone have a socket handler class made, or do I need to design it myself?
Oct 23 17:42:02 <pacior> maybe try as with hibernate, deserializing collections separately?
Oct 23 17:43:14 <drknzz> pacior: As in serializing each song separately with a loop?
Oct 23 17:45:18 <pacior> drknzz: yes maybe, and with keeping order of writing then reading
Oct 23 17:45:32 <drknzz> i will try that then
Oct 23 17:54:27 <Looney> a design/software engineering question => I have a listview which I populated from xml file. the file tells me the next logical listview items which I need to render if I click one of the item in listview. whats the best approach
Oct 23 17:54:58 <Looney> I mean, I can think of intent to start new activity and stuff
Oct 23 17:56:27 <tech1> maybe you are looking for arrayadapter ? depending on the data source
Oct 23 17:56:47 <Looney> I have displayed data already in list view
Oct 23 17:57:43 <Looney> once I click on an item in listview, there is corresponding image or some  stuff which I need to show next
Oct 23 17:58:03 <tech1> on a new screen?
Oct 23 17:58:35 <Looney> yeah, new activity won't hurt
Oct 23 17:59:53 <tech1> activity, intent would work. i dont know of a better way
Oct 23 18:00:41 <ubert_mac> whats the difference between running as ANdroid Applicaton and Native Android Applicaton?
Oct 23 18:01:37 <lov> ubert_mac: whether your code is written in C or Java, mostly :P
Oct 23 18:01:51 <ubert_mac> ahh ok, C it is
Oct 23 18:01:52 <ubert_mac> I love C
Oct 23 18:01:57 * ubert_mac ducks.
Oct 23 18:02:06 <lov> are you talking about stuff like http://developer.android.com/reference/android/app/NativeActivity.html ?
Oct 23 18:02:18 <lov> It's a non-standard but totally cromulant path to write your entire app in C as of Gingerbread
Oct 23 18:02:23 <lov> that said, I don't envy the poor sod who does it.
Oct 23 18:02:29 <ubert_mac> in Eclipse you can chose between those 2 options
Oct 23 18:02:47 <ubert_mac> k gotcha
Oct 23 18:02:49 <lov> where?
Oct 23 18:02:58 <ubert_mac> Im a senior java eng, but I come from the server side, Apache
Oct 23 18:03:08 <ubert_mac> new to android, but have extensive os experience
Oct 23 18:03:12 <ubert_mac> iOS
Oct 23 18:03:20 * ubert_mac ducks from all the flying shoes.
Oct 23 18:03:20 <Escherial> for the record, i'm an idiot :|
Oct 23 18:03:26 <lov> well we knew that.
Oct 23 18:03:27 * Escherial chucks a few shoes
Oct 23 18:03:44 <Escherial> <3
Oct 23 18:04:09 <ubert_mac> get in line :P
Oct 23 18:04:10 <Escherial> in short, my issues with my loader were more related to my switch statement falling through to "cursor = null"
Oct 23 18:04:28 <ubert_mac> watch get for not using debugger hehehehheheehhehe
Oct 23 18:04:33 <ubert_mac> break points are your friend :P
Oct 23 18:04:56 <ubert_mac> i tent o just run the app in run and not debug and then wonder why nothing is breaking lol
Oct 23 18:07:50 <Looney> how to change an image of imageview while one is uses trackpad to browse through listview items
Oct 23 18:08:33 <ubert_mac> [myimageview setImage:xxxxxx];
Oct 23 18:10:23 <Looney> like we have onsetclicklistener for clicks, any listener for highlight
Oct 23 18:11:19 <ubert_mac> too busy here sorry, maybe post to paste bin.com
Oct 23 18:11:30 <ubert_mac> oh crap wrong channel lol
Oct 23 18:11:49 <ubert_mac> my answer was for iOS sorry, need to change the labels on these rooms heh *blush
Oct 23 18:23:43 <nbadal> Is it possible to have a button map to CTRL and fallback on HOME? I know you can do this with normal keys.
Oct 23 18:26:22 <docmur> Is it possible to create a static list view object from a string rather then populating from a database?
Oct 23 18:27:21 <sampullman> docmur: you can populate a listview with any sort of list adapter
Oct 23 18:27:47 <docmur> Oh so I have to write hte row I want into the list adapter then it will populate into the list view?
Oct 23 18:27:53 <docmur> *the
Oct 23 18:28:36 <sampullman> for example, you could create an arrayadapter from a String array, and call listView.setAdapter(myArrayAdapter)
Oct 23 18:31:14 <sampullman> does anyone know how to catch the event (that i assume) edittexts throw when the ticker-thing is used to change the selection index?
Oct 23 18:31:18 <sampullman> if that makes sense..
Oct 23 18:44:52 <mrenouf|work> ticker-thing?
Oct 23 18:44:58 <mrenouf|work> is that the technical term? ;-)
Oct 23 18:45:16 <mrenouf|work> screenshot might be helpful
Oct 23 18:46:43 <SrPx> For checking a database periodically to display certain notifications, should I use postDelayed? Will it work even when the app is on the background?
Oct 23 18:46:47 <sampullman> ah, i just figuered it out
Oct 23 18:47:02 <sampullman> somehow i missed onSelectionChanged in the docs
Oct 23 18:50:00 <evancharlton> sampullman: were you talking about d-pads or other directional controllers, btw?
Oct 23 18:50:15 <sampullman> hold on, ill grab a screenshot
Oct 23 18:52:32 <sampullman> http://imgur.com/kGe0g
Oct 23 18:52:51 <sampullman> evancharlton: the blue thing under the edittext
Oct 23 18:53:31 <evancharlton> oh, whoa. I completely mis-read your question and thought you were talking about listviews
Oct 23 18:53:34 <sampullman> when you tap any edittext (on my sgs 3) it shows, and lets you drag to change the current selection index
Oct 23 18:53:34 <evancharlton> derp.
Oct 23 18:53:38 <evancharlton> yeah
Oct 23 18:53:47 <sampullman> oh
Oct 23 18:54:01 <RustyShackleford> I have a .svg (I will convert it to png) I would like to use for icons. What size should I make the .png?
Oct 23 18:54:18 <Sicp> 92, 72, 48, 32 RustyShackleford
Oct 23 18:54:28 <Sicp> for XHDPI, HDPI, MDPI and LDPI
Oct 23 18:54:29 <sampullman> Rusty: http://developer.android.com/guide/practices/ui_guidelines/icon_design.html
Oct 23 18:54:34 <Sicp> or that
Oct 23 19:05:25 <RustyShackleford> Sicp: perhaps this is a dumb question, but all icons are squares?
Oct 23 19:07:33 <tech1> i think so
Oct 23 19:08:11 <tech1> http://developer.android.com/guide/practices/ui_guidelines/icon_design_launcher.html#size
Oct 23 19:08:50 <tech1> 36 x 36 px 	48 x 48 px 	72 x 72 px 	96 x 96 px
Oct 23 19:11:27 <Sicp> squares in what sense?
Oct 23 19:11:32 <belgianguy> is there a way to prevent Google Docs from opening links to Excel files?
Oct 23 19:11:37 <Sicp> evenly sized? or the shape that the icon is going to take on the launcher?
Oct 23 19:11:47 <belgianguy> eg I just want to download the spreadsheet, not open it in the browser
Oct 23 19:11:57 <Sicp> the latter is not true, the former is; they are all evenly sized (i.e. 32 means 32 x 32 px/dpi)
Oct 23 19:13:00 <Sicp> ^^ afaik
Oct 23 19:16:37 <RustyShackleford> ugh I don't want to learn image editing software. Is there an easy program that shrinks a picture to a specific size?
Oct 23 19:17:25 <tech1> tons
Oct 23 19:18:04 <tech1> though it would take just as long to resize with a program that does only that
Oct 23 19:18:28 <Sicp> it's usually the job of the GD in your team, but..we've all had to do it at some point
Oct 23 19:18:34 <Sicp> learnin 9patch also
Oct 23 19:18:38 <Sicp> +g
Oct 23 19:19:14 <tech1> how can you build apps and not know how to resize an image
Oct 23 19:19:21 <Sicp> +1
Oct 23 19:21:40 <fakingfantastic> Hey gang, Im trying to theme a sidenav like this (http://4.bp.blogspot.com/-pm6gUKGPF6U/T9O1zOlTMfI/AAAAAAAAHe0/-jPLw5WDOv4/s1600/7.png). Im using a Text field and a ListView. How do I make the whole panel scroll, and disable the ListView scroll? When I do ScrollView > ListView, i get a warning that I have two scroll items
Oct 23 19:21:55 <fakingfantastic> Note: I'm not asking how to pull of the sidenav pattern, that's all working fine
Oct 23 19:22:39 <RustyShackleford> it seems that the ADT plugin creates icons for you from a picture
Oct 23 19:22:41 <RustyShackleford> sweet
Oct 23 19:23:07 <freeone3000> How do I cause a service to die when all bound activites are unbound?
Oct 23 19:23:29 <evancharlton> that should happen automatically, no?
Oct 23 19:23:46 <evancharlton> just make everyone connect with bindService(..) and it'll terminate on its own, afaik
Oct 23 19:25:05 <freeone3000> evancharlton: That's what I've got, but I've still got the service alive in the background waking the phone from sleep.
Oct 23 19:25:16 <Belacan> Hi all
Oct 23 19:25:30 <Belacan> has anyone here used kaufman filters for GPS data?
Oct 23 19:25:51 <freeone3000> evancharlton: http://ideone.com/L1QXQ4 Line 92 and Line 99.
Oct 23 19:27:01 <freeone3000> Belacan: Kalman filter, you mean?
Oct 23 19:28:05 <Belacan> haha yeah
Oct 23 19:28:07 <Belacan> kalman filter
Oct 23 19:28:08 <Belacan> sorry
Oct 23 19:28:26 <Belacan> Wait maybe theres something better
Oct 23 19:28:33 <freeone3000> Seems redundant. GPS is extraordinarily accurate.
Oct 23 19:28:42 <Belacan> really?
Oct 23 19:28:49 <freeone3000> To within a few feet, yeah.
Oct 23 19:28:57 <Belacan> hmm
Oct 23 19:28:58 <Belacan> well
Oct 23 19:28:59 <kevinb> you actually need EKF for gps data
Oct 23 19:29:05 <kevinb> and it can be helpful if you have temporary outages
Oct 23 19:29:36 <Belacan> I dont need 'real time' filtering
Oct 23 19:29:47 <Belacan> Say i recorded a number of GPS fixes over some time
Oct 23 19:30:16 <Belacan> Then afterwards i need to find the most probably locations
Oct 23 19:30:20 <g00s> i think if you dont need real time, you can do least squares fit
Oct 23 19:30:22 <Belacan> would you still reccomend a kalman filter?
Oct 23 19:31:20 <freeone3000> Belacan: What're your other inputs?
Oct 23 19:31:32 <Belacan> g00s does that still work for a non linear relationship?
Oct 23 19:32:02 <Belacan> freeone3000: i need to accurately measure distance from a starting point and time
Oct 23 19:32:17 <Belacan> to get the deceleration profile of a car
Oct 23 19:32:25 <Belacan> or speed
Oct 23 19:32:30 <Belacan> whatever is more accurate
Oct 23 19:32:40 <freeone3000> Belacan: Your *only* input is GPS?
Oct 23 19:33:05 <Belacan> umm
Oct 23 19:33:12 <Belacan> I think so?
Oct 23 19:33:19 <Belacan> yeah
Oct 23 19:33:22 <g00s> Belacan: kalman filter would also use inertia
Oct 23 19:33:28 <freeone3000> Then the most accurate measure is what the GPS is saying.
Oct 23 19:33:49 <Belacan> freeone3000: I know but i also need to filter it somehow because the GPS is surely a bit noisy
Oct 23 19:33:52 <g00s> so a nonlinear lesast squares fit is your huckleberry, most likelt. btw, my gps is very noise, like 20ft
Oct 23 19:34:21 <freeone3000> Nonlinear least-squares assumes you already know the function-type you're mapping to.
Oct 23 19:34:31 <g00s> if i want to be sure i get the same 'point' for a real spot, i often have to come back to that area a few hours later when satellite positions have moved
Oct 23 19:34:38 <Belacan> so taking all that noise into acount the app needs to guess the probable distance
Oct 23 19:35:02 <freeone3000> g00s: ...The satellites don't move.
Oct 23 19:35:28 <Belacan> freeone3000: true dat
Oct 23 19:35:48 <Belacan> they are geosynchronous
Oct 23 19:36:01 <kevinb> weather changes though, there's DGPS to try to reduce that noise, but it sounds like Belacan doesn't care about absolute position, just relative during the testing
Oct 23 19:36:17 <freeone3000> Actually, I'm completely wrong. They totally move.
Oct 23 19:36:35 <tech1> they just move at the same parallel orbit?
Oct 23 19:37:06 <Belacan> kevinb: correct
Oct 23 19:37:14 <freeone3000> tech1: Yes, they orbit. No, they're not on the same orbit. Yes, at least 3 are always visible. It's really cool, wikipedia has a diagram.
Oct 23 19:37:24 <freeone3000> One day I'll fact-check *before* I say something, though.
Oct 23 19:38:26 <Belacan> so is kalman filtering still the way to go? even though real-time data is not neccecary?
Oct 23 19:38:33 <tech1> i see
Oct 23 19:38:37 <kevinb> what's the rate of GPS on Android? sadly i'm guessing it's either manufacture specific or some manufactures cheat and repeat the same readings multiple times?
Oct 23 19:39:15 <freeone3000> Belacan: Kalman filtering assumes that you have another way of determining the state of the vehicle...
Oct 23 19:39:41 <freeone3000> Belacan: If you have a few measurements of the vehicle at rest, for instance, you can use Kalman filtering to determine its position with greater certainty.
Oct 23 19:39:45 <Belacan> freeone3000: I beleive it uses an estimate of your velocity to work it out
Oct 23 19:39:53 <freeone3000> Belacan: Yes. Which means you need velocity.
Oct 23 19:40:04 <Belacan> which is from the previous GPS data
Oct 23 19:40:16 <g00s> velicity and acceleration
Oct 23 19:40:16 <Belacan> its itterative
Oct 23 19:40:36 <kevinb> if you know the car is driving in a straight line and it's current speed, you do have a prediction step so kalman is doable, however since you don't need real time i imagine some other algorithm would be more accurate. maybe monte carlo or something? I don't have experience there so i don't know.
Oct 23 19:40:53 <freeone3000> Belacan: You need a prediction step independent of GPS.
Oct 23 19:41:20 <Belacan> are you sure freeone3000 ?
Oct 23 19:41:23 <kevinb> just GPS is doable
Oct 23 19:41:32 <freeone3000> Belacan: If your position is uncertain, dx/dt is uncertain, so your velocity cannot be mroe certain than any position. Therefore, you can't use velocity to refine position, because velocity is just as unreliable.
Oct 23 19:42:39 <Belacan> freeone3000: but surely velocity can be more certain when you consider the history of your position
Oct 23 19:42:54 <Belacan> from the previous fix no, but from many fixes i'd argue yes
Oct 23 19:43:16 <freeone3000> Belacan: Also: Static analysis on variables: You're looking at deceleration, so that's probably a function of x^3 (rate of change of acceleration is constant), which can be least-squared pretty easily.
Oct 23 19:44:39 <RustyShackleford> is it easy to build an application that behaves differently for different API levels?
Oct 23 19:44:51 <freeone3000> RustyShackleford: Generally not what you want to do.
Oct 23 19:45:11 <freeone3000> RustyShackleford: You can use the support libraries to bring selected API features to previous API levels.
Oct 23 19:45:37 <RustyShackleford> and then run the support library on newer platmorms as well?
Oct 23 19:45:44 <freeone3000> RustyShackleford: Yes.
Oct 23 19:45:57 <tech1> i think you are meant to do the opposite (make sure the app works the same on all APIs)
Oct 23 19:46:11 <g00s> "Zynga just laid off 100+ employees in its Austin office, closes down its The Ville game" wut a bunch of sleezeballs
Oct 23 19:46:20 <g00s> during the apple announcements, heh
Oct 23 19:46:42 <g00s> i'm glad bovines have moved to greener pastures ;)
Oct 23 19:46:46 <pragma-> what great news
Oct 23 19:46:52 <RustyShackleford> i have a tendency to try things that are bleeding edge
Oct 23 19:47:09 <RustyShackleford> i could probably target this for an old API version and it would look fine
Oct 23 19:47:21 <lov> RustyShackleford: are you asking about backwards compat in general?
Oct 23 19:47:24 <pragma-> canadiancow: did you hear g00s's wonderful news?
Oct 23 19:47:26 <lov> Yes, it's totally possible and expected to do this.
Oct 23 19:47:45 <lov> RustyShackleford: http://developer.android.com/design/patterns/compatibility.html
Oct 23 19:49:33 <RustyShackleford> I'll just use the support package. I was picturing a setup where I have one build for API's that need the support package, and another for targets that have those features natively
Oct 23 19:49:39 <RustyShackleford> but that sounds like a lot of work
Oct 23 19:49:45 <lov> RustyShackleford: It is.
Oct 23 19:49:48 <lov> Just use the compat libraries.
Oct 23 19:49:53 <lov> they'll handle everything automagically
Oct 23 19:50:19 <lov> Ah, here's what I was looking for
Oct 23 19:50:20 <lov> http://android-developers.blogspot.com/2009/04/backward-compatibility-for-android.html
Oct 23 19:50:20 <freeone3000> By which they mean "they'll always use the compatability libraries, because otherwise stuff will break because the compatability libraries aren't always EXACTLY compatible with the current API".
Oct 23 19:50:30 <lov> Personally, I prefer the factory patterns and interfaces over reflection.
Oct 23 19:50:46 <lov> but reflection is totally valid
Oct 23 19:59:33 <jasonmog> my external storage path is /sdcard0 which is an internal, non-removable sd card and files written to it are private. how do i get the path of the first removable sd card?
Oct 23 20:01:23 <leslie> jasonmog: getExternalStorageDirectory doesn't work?
Oct 23 20:01:39 <jasonmog> no
Oct 23 20:01:44 <jasonmog> that returns /sdcard0
Oct 23 20:01:45 <evancharlton> jasonmog: Android (sadly) doesn't really support multiple storage volumes. If the one you want isn't returned by Environment.getExternalStorageDirectory() then you're stuck with hacks
Oct 23 20:01:55 <jasonmog> that's what i thought
Oct 23 20:02:11 <jasonmog> i couldn't find that explicitly written anywhere in the docs
Oct 23 20:02:29 <jasonmog> would like to show the client something more than an IRC quote or google result
Oct 23 20:03:28 <lov> jasonmog: yeah, you're pretty much stuck.
Oct 23 20:03:33 <lov> You don't have a lot of choices here, sadly
Oct 23 20:03:35 <evancharlton> jasonmog: I don't have a more definitive resource for you, but it's the truth
Oct 23 20:03:38 <evancharlton> point to the source?
Oct 23 20:03:46 <evancharlton> lov: is there a bug on b.android.com, btw?
Oct 23 20:03:50 <lov> oh jesus
Oct 23 20:03:52 <lov> I'm sure there is.
Oct 23 20:04:00 <lov> there's a bug on morrildl I'm sure
Oct 23 20:04:03 <evancharlton> I doubt it'll be fixed since android is moving to single-volume storage anyway, but it can't hurt
Oct 23 20:04:09 <lov> I consider that a more effective bugtraq system
Oct 23 20:04:37 <lov> I actually need to get someone to clandestinely drop insects on him with little notes like "multiple storage volumes" written on them
Oct 23 20:10:18 <Belacan> I think im gonna go with least squares
Oct 23 20:11:59 <MDijkstra> most squares isn't cutting it?
Oct 23 20:13:29 <Belacan> haha
Oct 23 20:13:36 <lov> it's hip to be square
Oct 23 20:18:12 <TheSeven> hm... any wpa_supplicant experts around?
Oct 23 20:19:31 <ron_frown> IGUAY!
Oct 23 20:21:54 * TheSeven is running into weird wifi roaming problems, which are seemingly caused by a race condition (or something similar) in android's wpa_supplicant (it works fine on linux)
Oct 23 20:22:39 <lov> dragorn: ping: ^^
Oct 23 20:22:44 * ron_frown wonders if android platform HEAD should be named devs-playground
Oct 23 20:22:56 <dragorn> hey
Oct 23 20:23:05 <dragorn> lov: I think we've tlaked before
Oct 23 20:23:15 * lov shrugs
Oct 23 20:23:17 <dragorn> TheSeven: I assume you're [7]?
Oct 23 20:23:20 <TheSeven> yes
Oct 23 20:23:20 <lov> unping!
Oct 23 20:23:25 <dragorn> heh
Oct 23 20:23:34 <dragorn> yeah sorry, I don't work at the company anymore w/ the guy working on supplicant changes
Oct 23 20:23:54 <TheSeven> do you have a clue who could be working on this?
Oct 23 20:24:19 <TheSeven> I've been digging deeper during the last days and figured out some more details
Oct 23 20:24:28 <TheSeven> seems like wpa_s->bssid gets updated way too late
Oct 23 20:25:23 <TheSeven> http://pastie.org/5105281 lines 317-344 shows the problem... IIUC the radio is on the new channel starting from line 317
Oct 23 20:25:56 <TheSeven> and thus the packet being sent at line 333 will end up nowhere
Oct 23 20:27:18 <TheSeven> one possible solution might be sending the packet to wpa_s->last_eapol_src instead of wpa_s->bssid, but IIUC that might have some security implications
Oct 23 20:37:04 <tdignan> can anyone recommend a hardware accelerated charting library? For a moving "mountain" chart?
Oct 23 20:37:13 <tdignan> I will be updating it every second, so it has to be kind of fast.
Oct 23 20:39:28 <tdignan> Do you think the canvas would be fast enough?
Oct 23 20:39:44 <tdignan> it can't use a lot of CPU time.
Oct 23 20:39:56 <lov> tdignan: canvas would be fast.
Oct 23 20:40:02 <lov> it wouldn't be game development fast.
Oct 23 20:40:06 <tdignan> I guess I should try it before I knock it
Oct 23 20:40:10 <lov> I don't know of any charting libraries for android
Oct 23 20:40:11 <lov> yes, please do that.
Oct 23 20:40:13 <tdignan> I'll try it and report back
Oct 23 20:40:47 <SrPx> Sorry, I have been disconnected... did anyone answer my question?
Oct 23 20:46:24 <mheld> hey y'all
Oct 23 20:46:58 <mheld> anybody know what happened to canvasGL?
Oct 23 20:47:31 <RobinB> hi
Oct 23 20:54:40 <Belacan> does GPS error follow a normal distribution?
Oct 23 20:54:44 <ron_frown> disappointment abounds
Oct 23 20:55:19 <lov> Belacan: more or less
Oct 23 20:55:28 <lov> Belacan: it's an estimation
Oct 23 20:55:50 <lov> by error you mean accuracy?
Oct 23 20:56:13 <Belacan> relative accuracy I think
Oct 23 20:56:42 <lov> yeah
Oct 23 20:56:43 <Belacan> So what matters is not the absolute position (precision?)
Oct 23 20:56:51 <lov> well
Oct 23 20:56:53 <Hodapp> does Android GPS do dead-reckoning?
Oct 23 20:56:55 <Belacan> so say we get 10 fixes in a second
Oct 23 20:57:05 <lov> Hodapp: it's implementation specific.
Oct 23 20:57:14 <lov> All android knows about is the position fix information coming from the baseband.
Oct 23 20:57:19 <lov> it doesn't do the GPS calculations itself.
Oct 23 20:57:39 <p_l> Hodapp: afaik dead reckoning doesn't really make sense for GPS units themselves, and android simply gets NMEA stream (mostly)
Oct 23 20:57:41 <TheSeven> Belacan: you can't really improve the accuracy by averaging measurements (at least if you have a solid fix)
Oct 23 20:57:57 <p_l> Hodapp: there are, however, devices that have rudimentary INS included
Oct 23 20:58:04 <lov> ^^
Oct 23 20:58:19 <lov> There may be some measure of inertia involved in the calculations, especially to give the appearance of smoothness even when a fix hasn't come in
Oct 23 20:58:30 <lov> p_l: actually, while android gets an NMEA stream, that's only used to pass on to the user.
Oct 23 20:58:33 <Belacan> TheSeven: What i really want to know is say I move 20m away from my starting point, is the GPS accurate in that respect?
Oct 23 20:58:45 <lov> android really gets direct information about what satellite strengths are and what the baseband fix information is.
Oct 23 20:58:54 <lov> Belacan: maybe.
Oct 23 20:59:04 <lov> Belacan: it's supposed to be, but it depends on the quality of your fix.
Oct 23 20:59:13 <lov> if your accuracy is like 50m or something, you may not really get any change
Oct 23 20:59:27 <p_l> lov: aren't those also included as optional data in NMEA? Not all devices have Qualcomm chips with custom GPS protocol :)
Oct 23 20:59:42 <p_l> (iirc, Qualcomm had custom as at least an option)
Oct 23 20:59:52 <lov> p_l: I'm actually the reason that you can request NMEA information, in a roundabout way.
Oct 23 21:00:06 <p_l> hehe
Oct 23 21:00:08 <lov> if you look at how the GPSLocationProvider works, it just gets location information injected into it from the native layer
Oct 23 21:00:19 <CharlesWHChan> I am trying to figure out how to get LTE cell information from Android 4.x API. From the SDK, I can only find GsmCellLocation and CdmaCellLocation  any suggestions? Thanks
Oct 23 21:00:21 <lov> with the NMEA stream just being passed on to any consumers
Oct 23 21:00:34 <lov> the NMEA stream DEFINITELY isn't parsed by android itself.
Oct 23 21:00:39 <lov> at least it wasn't as of like a year ago
Oct 23 21:00:45 <lov> s/a/two/
Oct 23 21:01:04 <KaiKai> hi does anyone here have experience with android in-app-billing through server verification?
Oct 23 21:01:10 <lov> ([ask me about] working at the frameworks layer on Android's GPS!)
Oct 23 21:01:29 <p_l> lov: well, it's parsed (if it's used) by whatever software gizmo implements the location services (which, iirc, is available for customization by vendor)
Oct 23 21:02:03 <lov> p_l: I can only talk about what I've seen personally but my understanding is that said parsing happens on the basebad. I've only worked with QC stuff though.
Oct 23 21:02:29 <lov> with QC you had a nice pretty C interface where you got frequent callbacks of location information all pre-calculated and pre-chewed
Oct 23 21:02:50 <lov> *baseband
Oct 23 21:03:25 <lov> NMEA is really just meant as a communication layer for applications that expect it; I can't imagine passing NMEA anywhere in the baseband as being A Thing.
Oct 23 21:03:41 <lov> You can use NMEA to communicate over bluetooth in an industry standard format
Oct 23 21:04:03 <p_l> lov: I think it's because QC chips have GPS and mobile baseband integrated into same chip and use custom protocol there
Oct 23 21:04:24 * lov shrugs
Oct 23 21:04:36 <p_l> i.e. the chewing happens in OLK4 code running on baseband processor
Oct 23 21:04:44 <lov> right
Oct 23 21:05:02 <p_l> (which runs L4/Linux, btw)
Oct 23 21:05:53 <p_l> had a bit of surprise there back in the old days of Windows Mobile 5.0 and trying to get those phones to update
Oct 23 21:08:35 <Belacan> aight peeps thanks for your input, ;) ;)
Oct 23 21:08:37 <Belacan> im off now
Oct 23 21:16:04 <g00s> i would imagine the visible accuracy of the gps is different from chipset to chipset
Oct 23 21:16:35 <g00s> and error distributions, etc
Oct 23 21:16:51 <g00s> bias error? haven't seen that i guess
Oct 23 21:17:42 <SrPx> What's the correct way to check a site every x seconds for an app to display notifications about it? That is, what is a timer function that will work even when the app is not openned?
Oct 23 21:20:31 <kbs> SrPx: what I do at least, is register a few broadcast receivers: one for boot_received, and another on network changes. The boot_receieved places a service hat's tickled periodically from the alarm manager, while the network change also tickles the service. The service itself checks if it should make a network call, and does so if 'more than x seconds' have passed.
Oct 23 21:21:24 <kbs> one can further tweak the mechanics of when it should be called (eg: don't bother to do so if the phone is not awake, wifi is disabled, etc.)
Oct 23 21:21:26 <Komak57> I want to create an event handler in Java. I know the equivelant coding in C#, but that doesn't seem to be assisting me in any way here. What terms do you suggest?
Oct 23 21:21:57 <SrPx> kbs: let me see. Thanks
Oct 23 21:25:21 <theTerry> are there any Fragment Transaction experts in the house?  Sometimes transaction.replace leaves a fragment behind
Oct 23 21:25:46 <theTerry> esp if the app has been sleeping
Oct 23 21:27:36 <kbs> Komak57: the android event-handling is relatively simplistic. Generally, you write some class that implements a method that's called when an event occurs (eg: View.OnClickListener) and then attach it to a specific control with its setXXXListener (eg: Button.setOnClickListener(your_event_handler_instance))
Oct 23 21:29:34 <Komak57> I want my own event listeners for non-button events?
Oct 23 21:29:56 <Komak57> More specifically, I want to cause an event when a line is read from a socket.
Oct 23 21:30:21 <Komak57> And I want the Activity that uses this class to be able to  implement it's own method of handling it.
Oct 23 21:31:12 <Komak57> C# Example: http://msdn.microsoft.com/en-us/library/db0etb8x.aspx
Oct 23 21:31:37 <kbs> ah. If I'm following what you want correctly, there is no generic event-handling framework -- in essence, the event-handling is coupled very tightly to the actual implementation of each control.
Oct 23 21:32:17 <kbs> What I do myself is simply use a threadpool where I enqueue runnables, and anywhere that I need to pull or post events, I just post the runnable (or other items, depending on need.)
Oct 23 21:33:22 <kbs> eg: I have a queue (or sometimes thread-pool) that subscribers can pull or post events (or runnables)
Oct 23 21:33:50 <kbs> this decouples the event poster from the event handler fairly easily; but this is all a diy event framework
Oct 23 21:34:19 <Komak57> I was thinking of having a read thread asynchronously scan the read buffer and, if it contains a newline char, add it to a list of events of onRead(). My main activity could then create a method like OnRead(MSGEvent e) {} to the event listener.
Oct 23 21:35:03 <Komak57> Meaning this class would pretty much handle all the connectivity stuff between sockets, and leave the activity to do all the coding based on return values.
Oct 23 21:35:51 <Komak57> At the moment, it's simply bufferring and public String Get() would pull the next line, or null. Just trying to cut down on loop resources.
Oct 23 21:36:18 <kbs> Komak57: right. I don't know that there's One Right Way to do anything; but as I mentioned earlier, what I'd do is use a globally accessible queue. In this case, I might create an encapsulated data class that contains each line and have the network code push it into the queue.
Oct 23 21:36:54 <kbs> any class that needs to process it would register itself, and get called when such events occur, etc.
Oct 23 21:37:45 <TheSeven> dragorn: just FYI, I've now filed a report here: https://code.google.com/p/android/issues/detail?id=38830
Oct 23 21:38:25 <kbs> [The one android gotcha with this approach is that the called code doesn't run on the UI thread, so any UI action should use post() or something similar to do any ui work.]
Oct 23 21:38:45 <TheSeven> I wonder why this hasn't been reported yet
Oct 23 21:38:53 <Komak57> think I found something... reading up on it: http://www.exampledepot.com/egs/java.util/custevent.html
Oct 23 21:39:10 <TheSeven> it seems to be affecting all devices I've seen so far and is causing massive wifi problems in enterprise environments
Oct 23 21:40:05 <TheSeven> now how do I get the right people to notice this report?
Oct 23 21:40:36 <TheSeven> any tags which I can add to show that it's related to wpa_supplicant?
Oct 23 21:46:46 <SpeedEvil> TheSeven: people willing to dig into unexplained wierdnesses are regrettably rare.
Oct 23 21:48:51 <TheSeven> yeah, but being the sysadmin who all the users run to swearing about their flaky wifi I'm annoyed enough by it to make me try to nail it down
Oct 23 21:49:56 <SpeedEvil> :-)
Oct 23 21:50:11 <SpeedEvil> I truly hope this sort of shit gets nailed down.
Oct 23 21:50:24 <TheSeven> well I think I'm pretty close now :)
Oct 23 21:51:03 <SpeedEvil> my nexus 7s WiFi connextion stalls randomly. it sits there connected at some level, but IP is broken
Oct 23 21:51:05 <TheSeven> I already have a possible fix, but it might have security implications, so I'm trying to hand this over to the experts
Oct 23 21:51:12 <Komak57> I can't import javax.swing.event. Anyone else have this difficulty?
Oct 23 21:51:23 <romainguy> Komak57: javax.swing doesn't exist on Android
Oct 23 21:51:33 <TheSeven> SpeedEvil: had that with SGS3 before the first few batches of updates
Oct 23 21:51:36 <Komak57> ugh, you're kidding =.=
Oct 23 21:51:45 <TheSeven> seemed to be related to it hopping back and forth between 2.4 and 5 GHz
Oct 23 21:51:46 <SpeedEvil> same with two devices, so not hw, and the 8 other devices I have work fine
Oct 23 21:51:55 <SpeedEvil> 2.4 only her
Oct 23 21:52:07 <tech1> implement the right action listner
Oct 23 21:52:41 <TheSeven> SpeedEvil: so you're observing this one?
Oct 23 21:52:42 <TheSeven> https://code.google.com/p/android/issues/detail?id=35496
Oct 23 21:53:17 <SpeedEvil> not quite
Oct 23 21:53:22 <SpeedEvil> ping stops
Oct 23 21:53:34 <TheSeven> hm, I've seen some reports of this as well
Oct 23 21:54:17 <TheSeven> https://code.google.com/p/android/issues/detail?id=12125
Oct 23 21:54:52 <SpeedEvil> not radius
Oct 23 21:56:00 <TheSeven> well nobody really claims that it only happens with radios
Oct 23 21:56:02 <TheSeven> radius*
Oct 23 21:56:06 <SpeedEvil> ah
Oct 23 21:56:13 <TheSeven> in fact some comments are reporting it for psk as well
Oct 23 21:56:40 <TheSeven> possibly also https://code.google.com/p/android/issues/detail?id=1698
Oct 23 21:57:47 <TheSeven> lots of that kind of reports over there...
Oct 23 21:58:08 <SpeedEvil> yeah
Oct 23 21:58:22 <SpeedEvil> determining exactly which bug applies is fun
Oct 23 21:59:50 <TheSeven> damn, waiting for the sgs3 jb update since more than 30 hours now, it still tells me I'm somewhere in the queue
Oct 23 22:03:31 <Sicp> TheSeven?
Oct 23 22:03:38 <Sicp> there's JB for S3?
Oct 23 22:03:45 <Sicp> when! how!
Oct 23 22:03:56 <TheSeven> Sicp: in most of europe, yes
Oct 23 22:04:11 <TheSeven> they started rolling it out in poland and south korea 2 weeks ago
Oct 23 22:04:24 <TheSeven> now they finally started the rollout in germany yesterday
Oct 23 22:07:15 <kbs> out of curiosity, how functional is the cyanomod version for the s3?
Oct 23 22:07:35 <kbs> i'm in the market for a new phone, and leaning towards the nexus but have the cash to splurge for an s3
Oct 23 22:07:48 <kbs> mostly inhibited because it doesn't run stock android
Oct 23 22:08:00 <Sicp> #android would know much more about that I'd suspect
Oct 23 22:08:14 <Sicp> seriously
Oct 23 22:08:22 <TheSeven> or #android-root
Oct 23 22:08:30 <Sicp> +5
Oct 23 22:08:42 <kbs> okay
Oct 23 22:08:46 <Escherial> hmm, my wakefulintentservice is not getting called, despite me doing this: WakefulIntentService.sendWakefulWork(context, BufferUploadService.class);
Oct 23 22:08:59 <TheSeven> but tbh I like touchwiz better than the gnex UI
Oct 23 22:09:09 <Escherial> (BufferUploadService extends WakefulIntentService, is registered in the manifest, and implements doWakefulWork())
Oct 23 22:09:34 <Escherial> actually, wait, der, it's not registered in the manifest...foot, meet mouth :O
Oct 23 22:10:08 <Sicp> lol
Oct 23 22:20:17 <g00s> interesting blog about html5/mobile debate http://www.ericsink.com/
Oct 23 22:20:22 <g00s> well, probably not too new
Oct 23 22:21:31 <lfaraone> I have a LinearLayout in a ScrollView which itself has a bunch of inner linear layouts. Is there a way to scroll to a specific linearlayout inside the linearlayout inside the scrollview?
Oct 23 22:27:57 <kbs> lfaraone: I haven't tried it myself, but you may have luck with some combination of using the getTop() method on the child linearlayout(s) along with scrollTo(0, child_top)
Oct 23 22:29:29 <kbs> g00s: interesting. I've found the most useful part of using/writing a native app is that it lets me do stuff "disconnected"
Oct 23 22:31:34 <kbs> that said, I have to say that I'm pretty tired of all the web sites that keep popping up that damned floating window asking me to install their native app
Oct 23 22:33:28 <mahamoti> im trying to insert an image into a textview. nothing happens--what am i doing wrong? http://pastebin.com/qdVdE2zX
Oct 23 22:34:39 <tech1> are you meant to try and set it in that method?
Oct 23 22:35:12 <mahamoti> what do you mean?
Oct 23 22:35:54 <mahamoti> editText.getText() returns an Editable object
Oct 23 22:36:05 <tech1> i havnt seen it in getText(), might be ok though
Oct 23 22:36:37 <mahamoti> could be, but there isn't a setText function that accepts anything inheriting from spannable..
Oct 23 22:37:19 <tech1> maybe editTextView.setSpan(is, 5,5+10,0); might work
Oct 23 22:37:42 <mahamoti> nope
Oct 23 22:38:06 <tech1> or set it in xml? android:background="@drawable/image" />
Oct 23 22:39:01 <mahamoti> ah i got it...im just an idiot was passing too large numbers into the index (2nd param)
Oct 23 22:39:24 <tech1> ahh
Oct 23 22:41:02 <DarkGhost`> see this talk about php now being able to develop android apps?
Oct 23 22:41:27 <tech1> where?
Oct 23 22:41:58 <mahamoti> setSpan(is, startIndex, endIndex, flag)
Oct 23 22:45:56 <pfn> huh, why am I intermittently getting TransactionToLargeException on a setView?
Oct 23 22:46:14 <pfn> http://stackoverflow.com/questions/11451393/what-to-do-on-transactiontoolargeexception
Oct 23 22:46:20 <pfn> it looks something like the OP's stack there
Oct 23 22:46:40 <jasta> that's interesting indeed
Oct 23 22:47:22 <jasta> well right away my thoughts turn to user-suppliable pieces of information that would enlarge the binder msg size
Oct 23 22:47:25 <jasta> like tags
Oct 23 22:48:06 <TheFred> database link?
Oct 23 22:48:54 <jasta> but im not actually sure whats logically covered in the window session proxy there
Oct 23 22:49:05 <jasta> im investigating now though.  quite interested.
Oct 23 22:51:59 <jasta> pfn: looking at the code, im absolutely baffled.
Oct 23 22:52:13 <jasta> what version of android?
Oct 23 22:52:19 <socket> hey, any idea if the WebView supports css animations?
Oct 23 22:52:30 <pfn> jasta, in my case, it happened on 4.1.2
Oct 23 22:52:37 <socket> it doesnt seem to be, but maybe my css is wrong
Oct 23 22:52:42 <jasta> it looks quite like google has made a chance that is allowing this issue
Oct 23 22:52:52 <jasta> because the version im looking at, looks impossible
Oct 23 22:58:12 <g00s> interesting, indeed
Oct 23 23:00:07 <pfn> jasta, yeah, it doesn't make sense either
Oct 23 23:00:24 <pfn> http://code.google.com/p/android/issues/detail?id=34828
Oct 23 23:00:34 <Butch128> Having issues with adb/my device... I've never had a problem compiling/installing before... now whenever i try, my device is kicked off adb
Oct 23 23:00:43 <pfn> ^^ that's almost exactly what my exception looks like
Oct 23 23:00:59 <pfn> status: spam?  wtf
Oct 23 23:01:01 <Butch128> adb -d install -r project-debug.apk <---- my device eventually falls off "adb devices"
Oct 23 23:01:37 <Butch128> this was working just last night, no idea what changed...
Oct 23 23:03:56 <pfn> Butch128, unplug and plug it back in
Oct 23 23:09:56 <Butch128> pfn: yep - tried that
Oct 23 23:10:07 <Butch128> dmesg spits out: usbfs: USBDEVFS_CONTROL failed cmd adb rqt 128 rq 6 len 256 ret -71
Oct 23 23:10:18 <Butch128> rebooting pc and phone also doesnt fix it
Oct 23 23:10:39 <pfn> turn on usb debugging on the phone
Oct 23 23:10:43 <Butch128> is
Oct 23 23:10:47 <Butch128> and tried turning it on and off
Oct 23 23:10:57 <Butch128> running the app in an emulator works fine
Oct 23 23:11:11 * pfn shrugs
Oct 23 23:11:12 <pfn> dunno then
Oct 23 23:11:15 <Butch128> but.. not the same as running it on my device
Oct 23 23:12:33 <kbs> I find that my color nook periodically does fall off as well. I usually kill the adb server, and replug the device in, which does the trick for me. It doesn't do it immediately; usually stays up for a few minutes.
Oct 23 23:13:26 <Butch128> yes - my device is listed when i freshly reboot my phone/computer
Oct 23 23:13:31 <Butch128> and stays connected for awhile as well...
Oct 23 23:13:41 <Butch128> until i try and install an app
Oct 23 23:15:36 <kbs> Interesting. So my procedure is to leave a logcat running all the time from a terminal
Oct 23 23:15:56 <kbs> I don't have to go all the way towards rebooting everything, just killing the adb process and re-plugging the device does the trick
Oct 23 23:16:18 <Butch128> yea... whats wierd is this has been working for days and days... and suddenly stopped
Oct 23 23:16:21 <Butch128> maybe i'll try my other phone
Oct 23 23:16:22 <pfn> sounds like the adb process on the phone is dying
Oct 23 23:16:23 <pfn> check logcat
Oct 23 23:16:26 <pfn> restart adb on the device
Oct 23 23:17:21 <Butch128> oh - unrelated question... it is possible to run an emulator with a non-read-only filesystem?  i need to edit the hosts file...
Oct 23 23:18:00 <Butch128> nm, mount -o rw,remount -t yaffs2 /dev/block/mtdblock3 /system works
Oct 23 23:28:28 <mobile8> how can i change the default that my screen turns of on my android phone?
Oct 23 23:28:35 <mobile8> after 10 secs
Oct 23 23:29:43 <vitaminbh> Butch128's issue rings a bell... nothing to do with needing a cert update?  seems like I ran into this before or something similar
Oct 23 23:39:26 <evancharlton> mobile8: #android for questions like that, but it's under settings > display most likely
Oct 23 23:40:12 <Denommus> hi
Oct 23 23:40:24 <cagedwisdom> sometimes in a webview I can select elements, like w/a mouse on desktop. Is there a way to turn that off?
Oct 23 23:40:28 <Denommus> eclipse won't start my activity
Oct 23 23:41:12 <Denommus> ActivityManager: Error: Activity class {package.application/package.application.activity.DoNotStartActivity} does not exist.
Oct 23 23:41:13 <Denommus>  
Oct 23 23:42:08 <Denommus> I think it should be package.application/.activity.DoNotStartActivity
Oct 23 23:57:15 <socket> anyone ever tried using webview with -webkit- ?
Oct 23 23:57:28 <socket> it's acting really funny with some webkit css
Oct 23 23:57:50 <socket> not updating the view after loadData() for example
Oct 24 00:18:19 <birbeck> how long does it take for google to respond to appeals for app removals?
Oct 24 00:18:50 <birbeck> i have no idea why my app was removed, certainly not for copyright infringement which was claimed in the notice
Oct 24 00:21:18 <pfn> I don't think they respond unless you get someone here to look at it or something
Oct 24 00:21:38 <birbeck> it said to appeal, respond to the email
Oct 24 00:22:25 <SpeedEvil> what sort of app was it?
Oct 24 00:22:56 <SpeedEvil> are you given ore than just 'copyright violation' - perhaps who complained
Oct 24 00:25:18 <evancharlton> pfn: no one in here can do anything about it
Oct 24 00:25:35 <evancharlton> as birbeck said, respond to the email
Oct 24 00:26:32 <birbeck> it was cover art downloader (donate). all works are original or public domain
Oct 24 00:27:02 <evancharlton> lols, I just realized I was responding to birbeck and cited birbeck.
Oct 24 00:27:07 <birbeck> its a small utility app to scan your music library for music with no album art, and uses public documented apis to legal download the cover art
Oct 24 00:27:12 <evancharlton> anyway, you were right, birbeck :-) Just respond to the email
Oct 24 00:27:22 <pfn> evancharlton, bummer
Oct 24 00:28:23 <birbeck> and no, i didnt get a dmca notice. ive had two before for different apps. one was legit and the other i settled with starbucks
Oct 24 00:28:35 <pfn> you settled?
Oct 24 00:28:43 <pfn> did you get big bucks out of it?  :)
Oct 24 00:31:01 <birbeck> no, i got to keep my app and them out of my hair for another year
Oct 24 00:31:51 <birbeck> im not excited about the lost income, but i dont want my google account deleted
Oct 24 00:32:14 <birbeck> i was stupid and created my publisher account using my personal account
Oct 24 00:38:59 <birbeck> its been days which is the only reason im asking in case anyone else has had it happen
Oct 24 00:43:57 <SpeedEvil> :-\
Oct 24 01:10:55 <nomi1988> think http://loopj.com/android-async-http/ this library would support oauth? would I have to create the self sign certificate?
Oct 24 01:13:30 <Leeds> g00s: ping?
Oct 24 01:21:33 <tdignan> nomi1988: a quick googl for android oauth points out that it might be a better choice to involve the google api client for java
Oct 24 01:22:01 <tdignan> I use the google api python client for client side oauth in one of my projects, and it turned out after some digging it has a convenience method that does almost the whole process for me.
Oct 24 01:22:17 <tdignan> I found said convenience method after starting to write my own server to receive the token
Oct 24 01:22:28 <tdignan> so it was kind of bittersweet
Oct 24 01:24:16 <tdignan> This library looks useful though.
Oct 24 01:29:04 <Rockmaninoff> how can i have a horizontal LinearLayout with seven children, the first which expands to fill space and the last six which are static (and equal) widths
Oct 24 01:29:23 <Rockmaninoff> setting a weight on just the first one just makes it fill the entire LL
Oct 24 01:35:47 <nomi1988> tdignan were you able to save the cookies?
Oct 24 01:36:10 <nomi1988> I'll be making different server calls throught my app
Oct 24 01:36:25 <nomi1988> I also need remember me functionality.
Oct 24 01:39:31 <edman007> how do I transfer files to the android simulators?
Oct 24 01:39:46 <Leeds> edman007: adb push?
Oct 24 01:39:53 <vadi2> Is it possible to retrieve the view that's powering an item in the list without making a custom adapter?
Oct 24 01:40:35 <edman007> Leeds, hrm, ok...
Oct 24 01:47:21 <tdignan> nomi1988: iirc I was saving stuff locally and reusing it. I forget if cookie is the correct oauth parlance.
Oct 24 01:47:34 <tdignan> nomi1988: my authentication lasted like, weeks
Oct 24 01:48:34 <tdignan> you can use their lib for oauth, since oauth works over http, but it's a pita to implement all the oauth stuff yourself, especially when it's already done. I'd use a third party lib to do the oauth handshake, and use that lib for all your regular http stuff.
Oct 24 01:50:40 <tdignan> nomi1988: https://github.com/tom-dignan/gdrive-cli/blob/master/oauth/simple_cli.py -- see, with the google api client, I only needed to set up a client secrets file and call a couple functions.
Oct 24 01:51:08 <tdignan> the authenticate() method literally does it all, it even starts a webserver for me, and opens a webbrowser to the oauth page!
Oct 24 01:51:41 <tdignan> not java, but I'd bet you'd have similar luck with the api client for java.
Oct 24 01:51:44 <svm_invictvs> I'm having trouble playing back an audio stream using the MediaPlayer class.  It basically will spend a long time preparing, and then never play.  At a bit of a loss as to what actually I need to do.
Oct 24 01:53:53 <tdignan> are you using prepare(), or prepareAsync()
Oct 24 01:54:17 <tdignan> if the latter, I believe you can set an onPreparedListener
Oct 24 01:54:26 <nomi1988> See I used that library in my other project which was heavy server call driven. It made my life really easy, didn't have to worry about saving whole lot of things to the prefrences. This new project is similar except it uses OAuth
Oct 24 01:54:44 <tdignan> nomi1988: you can use that lib
Oct 24 01:54:49 <tdignan> it's just not an oauth lib.
Oct 24 01:55:24 <nomi1988> oauth lib would have things that I would probably need?
Oct 24 01:55:28 <tdignan> once you do the oauth handshake you can pass your oauth data along with your API calls with that library, just use google api java client to do the handshake, and the other lib to do your oter stuff.
Oct 24 01:55:45 <tdignan> nomi1988: yes.. did you see my code? it's literally a couple lines to do an oauth handshake with the google api client.
Oct 24 01:56:51 <tdignan> http://blog.doityourselfandroid.com/2011/04/12/oauth-android-google-apis-client-library-java/
Oct 24 01:57:10 <tdignan> looks harder on android
Oct 24 01:57:17 <nomi1988> I see, so you are suggesting my approach would be to use google api for the oauth handshake and then go on to use that lib for server calls?
Oct 24 01:57:26 <tdignan> thats what I'd try.
Oct 24 01:57:41 <nomi1988> but aren't all the cookies in that handshake?
Oct 24 01:57:46 <doomrobo> I cannot, for the life of me, figure out why this linear layout doesn't scale. I have a linear layout (vertical) that contains many inflations of the same linear layout (horizontal) which is set to weight=1 and fill_parent. Yet, each row wraps to content
Oct 24 01:57:49 <doomrobo> help me
Oct 24 01:58:01 <tdignan> ". One of the big advantages of using OAuth2 is that it should provide alternative ways to let the user authorize access besides using the browser"
Oct 24 01:58:19 <tdignan> Make sure you check out the latest version
Oct 24 01:59:13 <nomi1988> Doesn't it depend what version the the server's already using?
Oct 24 01:59:45 <tdignan> yes
Oct 24 02:00:30 <tdignan> Oh look, what's this!?
Oct 24 02:00:32 <tdignan> http://developer.android.com/training/id-auth/authenticate.html
Oct 24 02:00:39 <tdignan> looks like you can use AccountManager.
Oct 24 02:01:32 <tdignan> https://developers.google.com/google-apps/tasks/oauth-and-tasks-on-android
Oct 24 02:03:09 <svm_invictvs> tdignan: prepareAsync
Oct 24 02:03:33 <svm_invictvs> tdignan: I have a listener added to start playing when it finishes preparing.  Is there something I'm not doinglike adding it to a run loop or something?
Oct 24 02:09:09 <nomi1988> Thx, but if its just like a HTTP call, then I don't see what else I would need to do
Oct 24 02:13:16 <nomi1988> https://github.com/loopj/android-async-http/issues/123 :(
Oct 24 02:13:18 <nomi1988> haha
Oct 24 02:17:48 <opprobrium> anyone here use GCM?
Oct 24 02:30:03 <nomi1988> http://nilvec.com/implementing-client-side-oauth-on-android/ I read this. So the way I understand it is OAuth handshake and servercalls are loosely tied. HttpGet request = new HttpGet(url); consumer.sign(request);
Oct 24 02:30:23 <nomi1988> that HttpGet for me would be different though
Oct 24 02:30:42 <nomi1988> its going to be AsyncHttpClient client = new AsyncHttpClient();
Oct 24 02:31:13 <nomi1988> would I still be able to do this: consumer.sign(client)?
Oct 24 02:33:09 <nomi1988> but I don't need to do that?
Oct 24 02:33:15 <nomi1988> confused
Oct 24 02:33:44 <nomi1988> actually no, I would have to do
Oct 24 02:37:53 <nomi1988> its not going to work is it?
Oct 24 02:38:55 <qkzoo1978> The default checkbox is really small and hard to see. Is there a.recommended.built in style that is.larger?
Oct 24 02:46:43 <nomi1988> sign(HttpRequest request). Signs the given HTTP request by writing an OAuth signature (and other required OAuth parameters) to it. I'm just thinking out loud, could I just cast back and forth? This seems like only one thime thing
Oct 24 02:47:24 <nomi1988> actually there: sign(String url)           "Signs" the given URL by appending all OAuth parameters to it which are required for message signing.
Oct 24 02:48:20 <nomi1988> is that it? is that what I need? then basically use this new string that's been created?
Oct 24 02:48:35 <nomi1988> I'm looking here: www.jarvana.com/jarvana/view/oauth/signpost/signpost-core/1.2/signpost-core-1.2-javadoc.jar!/oauth/signpost/OAuthConsumer.html
Oct 24 02:49:26 <qkzoo1978> Are there different checkbox styles built-in that can be used? Like.themes? I can't seem to find anything online...
Oct 24 02:51:11 <nomi1988> No, android api only provides you with one. You would have to custom make it
Oct 24 02:58:08 <qkzoo1978> Bummer
**** ENDING LOGGING AT Wed Oct 24 03:00:01 2012