**** BEGIN LOGGING AT Sun Nov 04 03:00:02 2012
Nov 04 03:00:23 <Striker121> I'm trying to get a timer to work in a program but I seem to be getting an odd error
Nov 04 03:00:29 <Striker121> Timer timer = new Timer();
Nov 04 03:00:29 <Striker121> 		TimerTask updateToast = new UpdateToasterTask();
Nov 04 03:00:29 <Striker121> 		timer.scheduleAtFixedRate(updateToast, 0, 3000);
Nov 04 03:01:30 <Striker121> Getting a syntax error on the final ")" and a "VariableDeclarator expected instead error
Nov 04 03:01:31 <Striker121> Any ideas?
Nov 04 03:01:41 <Striker121> (On the timer.schedule line)
Nov 04 03:02:06 <tdignan> did you declare your 'timer' and initialize it?
Nov 04 03:02:12 <tdignan> hm, yes
Nov 04 03:02:33 <Striker121> Oh, by the way, delays on in milliseconds, right?
Nov 04 03:02:38 <tdignan> rtfd
Nov 04 03:03:00 <Striker121> Didn't see it indicated
Nov 04 03:03:09 <Striker121> Simply says it's "Long delay"
Nov 04 03:03:36 <Striker121> Nevermind, I'm dumb
Nov 04 03:03:43 <Striker121> But still, any idea with the error?
Nov 04 03:04:03 <tdignan> make a clean class and try to repro
Nov 04 03:04:09 <tdignan> paste that if you can
Nov 04 03:04:34 <tdignan> it's probably something on a previous line
Nov 04 03:04:37 <tdignan> like a missing semicolon
Nov 04 03:04:47 <tdignan> or a : where there should be a semicolon
Nov 04 03:09:29 <QubeZ> hello all
Nov 04 03:15:11 <tdignan> hey QubeZ
Nov 04 03:15:40 <QubeZ> yo tom
Nov 04 03:18:40 <tdignan> still working on the ios version?
Nov 04 03:19:09 <QubeZ> was but stopped working on it to focus on being a new father... then work (real work) also started getting hectic
Nov 04 03:19:23 <QubeZ> just getting back to releasing the new version for Android (in a few days)
Nov 04 03:21:37 <tdignan> oh yeah
Nov 04 03:22:33 <tdignan> hows that going?
Nov 04 03:24:29 <QubeZ> being a dad?
Nov 04 03:28:17 <g00s> QubeZ: todomarp right ?
Nov 04 03:28:25 <g00s> *todomapr
Nov 04 03:28:37 <jessicajames> I'm trying to work with threading, but my GUI won't update like I want it to: https://gist.github.com/dff1f4a1244bca2f03cd
Nov 04 03:28:39 <jessicajames> Any ideas?
Nov 04 03:28:49 <QubeZ> g00s: ya
Nov 04 03:29:06 <g00s> QubeZ: have you done a tablet optimized version yet?
Nov 04 03:29:45 <QubeZ> g00s: not yet, after I get this release out in a few days... I am going to focus on the Tablet ver
Nov 04 03:29:59 <g00s> i think my struggle with that , has been i can see my program as entirely different beasts on phone/tablet
Nov 04 03:30:18 <QubeZ> g00s: right, as they should be
Nov 04 03:30:27 <g00s> oh i mean, /very/
Nov 04 03:30:51 <QubeZ> I'll be moving to fragments which I haven't used in TDM but will be easier transition to tablet support
Nov 04 03:31:10 <QubeZ> i have a lot of work to do :\
Nov 04 03:31:14 <g00s> :)
Nov 04 03:31:20 <g00s> hopefully its worth it
Nov 04 03:31:31 <QubeZ> but people seem to like it based on the feedback.. of course there's always stupid 1-star comments too
Nov 04 03:31:35 <QubeZ> but comes with the territory
Nov 04 03:32:15 <tdignan> QubeZ: yeh i heard being a dad can be tiring
Nov 04 03:32:18 <jessicajames> It's very short code, guys :) I would appreciate a quick look if anyone has time
Nov 04 03:32:31 <jessicajames> and if anyone has familiarity with multithreading
Nov 04 03:32:33 <QubeZ> tdignan: it is, he's very good kid though but still exhausting
Nov 04 03:32:45 <QubeZ> jessicajames: why are you not using ASyncTask?
Nov 04 03:33:09 <QubeZ> native threading in Java is probably not a good idea when you have other methods like ASyncTask that will make things easier for you
Nov 04 03:33:57 <tdignan> jessicajames: you are posting a runnable to the UI thread and sleeping in it. That's going to block the UI.
Nov 04 03:34:18 <tdignan> you should research how to do what it is your goal is instead of guessin
Nov 04 03:34:57 <tdignan> sure, you have a thread. all it does is post a runnable back to the UI thread that will block it.
Nov 04 03:35:10 <jessicajames> AsyncTask seemed more advanced/difficult so I thought this approach would be okay. But if you guys think otherwise, I will go ahead and read up on AsyncTask
Nov 04 03:35:30 <tdignan> yes
Nov 04 03:35:50 <jessicajames> thank you
Nov 04 03:37:08 <QubeZ> jessicajames: its actually super easy
Nov 04 03:37:27 <QubeZ> not to mention less error-prone i.e. unlike what you just did in that code :P
Nov 04 03:39:11 <ESphynx> So on a termWindow, all GL textures should be invalidated?
Nov 04 03:43:01 <jessicajames> QubeZ, do you have any tutorials? I'm thinking of starting with this: http://www.youtube.com/watch?v=JVaGZwuYmck&list=PL5A73AACC3037C6F7&index=101&feature=plpp_video
Nov 04 03:49:20 <ESphynx> guys?
Nov 04 03:49:25 <ESphynx> terminate window -- GL textures gone?
Nov 04 03:55:12 <ESphynx> Apparently clicking the 'Back' arrow causes my app to be destroyed?? Why's that?
Nov 04 03:55:42 <tdignan> ESphynx: i can't say for sure for OpenGL, but once the window is gone you can kiss all your Views goodbye. They will need to be recreated. Retaining them is wrong, because you'll leak an activity instance.
Nov 04 03:55:56 <tdignan> same for orientation change
Nov 04 03:56:22 <tdignan> i'm dealing with a similar problem right now, just not using opengl. writing code to persist the view state separate from the views
Nov 04 03:56:22 <tdignan> then going to profile them to make sure they're created fast
Nov 04 03:57:15 <ESphynx> hmm
Nov 04 03:57:49 <tdignan> about to smoke test my code
Nov 04 03:58:40 <tdignan> heh
Nov 04 04:09:58 <regedit> hai
Nov 04 04:10:09 <regedit> anyone here at this hour of weekend
Nov 04 04:10:22 <regedit> echo
Nov 04 04:11:07 <maxlamer> there has been some activity here
Nov 04 04:11:21 <regedit> cool
Nov 04 04:12:19 <regedit> is there any way for my android app to send a direct email without bothering with external email-sending app (with app chooser & Intent etc)
Nov 04 04:12:42 <regedit> much like one would do on *nix machine simply "mailx bla bla bla"
Nov 04 04:13:26 <maxlamer> there is, but there isn't any standard lib to do it
Nov 04 04:13:46 <maxlamer> i saw some example code a few days ago
Nov 04 04:14:22 <regedit> what am i dealing with - low-level android APIs? raw java libraries? other?
Nov 04 04:14:44 <regedit> NDK even?
Nov 04 04:15:57 <Leeds> you should be able to lob a Java SMTP library in, if you really want to
Nov 04 04:15:58 <SpeedEvil> open port 25 to the ISP...
Nov 04 04:16:27 <regedit> perhaps there is a more important question: how do email apps send emails; directly from device (by a real MTA), or the device just sends an HHTP request or somesuch to a server which actually sends the mail
Nov 04 04:16:44 <Leeds> and you might find a lot of mobile networks block direct external port 25 connections
Nov 04 04:17:18 <regedit> *http
Nov 04 04:19:03 <regedit> haz i not clear enough question
Nov 04 04:19:05 <Leeds> that's sounding a lot like an "if you need to ask..." question
Nov 04 04:20:27 <regedit> if an android device generally doesn't really directly send emails but rather requests a server to do so, then i wont bother much and ill do the same. if however there is a way to get the device to mail somethin' out directly
Nov 04 04:20:33 <regedit> i might prefer to try that
Nov 04 04:21:08 <regedit> again - without the app chooser & Intent process
Nov 04 04:21:38 <Leeds> presumably you don't mean sending email on behalf of the user?
Nov 04 04:22:17 <regedit> whadyou mean, yes the user does want this message to be sent
Nov 04 04:22:49 <regedit> not necessarily From the user's email address, if thats what your pointing out
Nov 04 04:23:48 <regedit> it can be from lolrofl@myappmailer.com
Nov 04 04:24:04 <regedit> or even lolrofl@12.34.56.78
Nov 04 04:24:54 <regedit> the user wants the message to be sent, but doesnt care for it to be sent as originating from the user's own email address
Nov 04 04:27:32 <Leeds> that sounds smappy
Nov 04 04:27:34 <Leeds> er, spammy :)
Nov 04 04:27:39 <maxlamer> yea
Nov 04 04:27:59 <maxlamer> that last address will get caught by spam assassin
Nov 04 04:28:31 <regedit> fine then myappmailer.com
Nov 04 04:28:46 <Leeds> yeah, still spammy
Nov 04 04:29:01 <maxlamer> where is all the email going to be going to anyway?
Nov 04 04:29:10 <regedit> point is, does the device send out mailings itself? or does it pass all the info to a server to do so
Nov 04 04:29:20 <Leeds> that's not the point
Nov 04 04:29:26 <regedit> to the destination of the user's choosing
Nov 04 04:29:29 <tdignan> someone has to write apps for the nigerian market
Nov 04 04:29:35 <regedit> verified, of course
Nov 04 04:29:47 <regedit> lol
Nov 04 04:30:07 <regedit> am i really asking about something that shady
Nov 04 04:30:15 <maxlamer> if you give users access to sending mail by your domain it will quickly get flagged
Nov 04 04:30:31 <maxlamer> if i was a spammer i would abuse it
Nov 04 04:30:45 <tdignan> yea, that's called an open relay.
Nov 04 04:31:21 <regedit> hmm..
Nov 04 04:31:21 <tdignan> just use GCM or something
Nov 04 04:31:27 <maxlamer> if users are going to be able to send mail through your app, then you will need them to connect to their own MTA
Nov 04 04:32:15 <Leeds> not to mention, again, that a lot of ISPs will block outgoing port 25
Nov 04 04:32:18 <regedit> ok basically what i'm trying to achieve
Nov 04 04:32:40 <kbs> In theory, you could always roll your own (e)smtp client, but the problem I suspect will be the amount of code to reliably deliver email. The gmail client (for instance) relays email via smtp.google.com:465 rather than directly connect to the mx host for the target.
Nov 04 04:33:15 <g00s> anyone have any links to empirical data showing that desktop & mobile users are different based on the class and nature of tasks they perform ?
Nov 04 04:33:42 <kbs> so if you're going through that much trouble, may as well just use http and do the heavy lifting of dealing with smtp at the server.
Nov 04 04:34:10 <Leeds> kbs: I suspect regedit has neither the capability nor the inclination to run his own proxy
Nov 04 04:34:17 <Leeds> regedit: tell us about the app...
Nov 04 04:36:17 <edris_> i have a question or two for android developers.
Nov 04 04:36:33 <regedit> i want the app to send some desired message to a desired destination email. the particular use case i'm thinking of ATM involves sending a shopping list as a SMS, in USA you can email to 1234567890@messageing.carrier.com and it comes as SMS
Nov 04 04:36:36 <edris_> i currently work on an embedded linux project that targets the arm926v5
Nov 04 04:36:55 <edris_> the build system uses OpenEmbedded.  And I really don't like OE very much.
Nov 04 04:37:17 <regedit> Leeds: ^ does that help
Nov 04 04:37:45 <edris_> at first glance android appears to have a more organized build system that is better documented
Nov 04 04:38:03 <edris_> our product would be "headless".
Nov 04 04:38:04 <Leeds> edris_: note that this channel is for app development - not many people here have experience of the system build... system
Nov 04 04:38:29 <Leeds> regedit: a little...
Nov 04 04:39:07 <edris_> fair enough.  but let me ask my first question:  our system is C++ based, talks to an fpga that talks to other devices.  I'm curious have non-java apps will do in the android world.
Nov 04 04:40:25 <edris_> our current solution runs apache2, so we'd need to have that running as well.  short of that, busybox leverages most of our commandline processing.
Nov 04 04:41:12 <edris_> so most apps that are being developed by folks in this channel.... are they java centric?
Nov 04 04:44:12 <kbs> edris_: not that I know the answer to your question; but to ask a dumb question -- is there any reason just compile/installing apache on your target architecture wouldn't work?
Nov 04 04:44:39 <regedit> Leeds: so, what would be the best way of going about that
Nov 04 04:46:24 <gutano> edris_, this is not a #java channel if that is what you are asking..  but yes, java as it relates to traditional android dev
Nov 04 04:46:44 <regedit> Leeds: also, do note that my app is trying as mush as possible to be usable in a "hands-free" manner (takes SpeechRecognizer voice commands) so im trying to send this message out without needing to choose email app, click "Send", etc
Nov 04 04:47:07 <Leeds> regedit: probably doing it through your own server... meh, SMS, is old and boring
Nov 04 04:47:33 <gutano> edris_, ppl often discuss things like phonegap, ndk or deeper aspects
Nov 04 04:47:37 <g00s> edris_: headless … running apache … are you sure android is the right thing :P
Nov 04 04:47:51 <g00s> sounds more like embedded linux project
Nov 04 04:48:32 <g00s> oh heh, you already said that
Nov 04 04:48:36 <edris_> i sort of said it was an existing embedded project targeting and arm926v5 using embedded linux.
Nov 04 04:48:38 <g00s> hehe its late
Nov 04 04:49:12 <edris_> i want to get a better idea of the build system for constructing the android solid image, and the contracts with executing applications
Nov 04 04:50:00 <edris_> apache was just 1 example.  we run multiple other common services (sshd, and others)
Nov 04 04:50:44 <tdignan> weirdness happening here. I'm saving a bundle in Fragment.onSaveInstanceState but when it's recreated I get null back in onCreateView's Bundle. I'm positive onSaveInstanceState is being called, and that I'm calling super. Also, I've checked the logs for exceptions and none are occuring.
Nov 04 04:50:49 <tdignan> any ideas
Nov 04 04:51:18 <edris_> ok, so regarding ndk...  i've already got a gcc build for the arm926, so i would probably just use it to construct the libc/whatever choice of libs
Nov 04 04:54:12 <g00s> uh, he left.
Nov 04 04:54:36 <tdignan> :| i have to write a test case.
Nov 04 04:54:37 <tdignan> sigh
Nov 04 04:54:38 <g00s> oh well, was going to say, this looked kinda cool http://www.amazon.com/Embedded-Android-Porting-Extending-Customizing/dp/1449308295
Nov 04 04:55:20 <tdignan> looks like an interesting book
Nov 04 04:55:32 <tdignan> i don't know enough about modding stuff
Nov 04 04:55:40 <g00s> due out very soon, so he could have probably gotten an early edition or something
Nov 04 04:55:57 * g00s just realized apress has a fuckton of android books
Nov 04 04:56:06 <g00s> i remember when it was like 2,3
Nov 04 04:56:32 <g00s> heh, Big Nerd Ranch Android book isn't by Murphy, even though he taught at BNR
Nov 04 04:59:53 <regedit> how is online data syncing done (like twitter or http://www.rememberthemilk.com/ etc) is everyone using GCM for that?
Nov 04 05:08:14 <tdignan> rest apis, push, polling, syncadapters
Nov 04 05:08:38 <tdignan> although I advise staying away from the last one unless you actually have a good use case.
Nov 04 05:09:51 <regedit> so GCM isn't the only way of keeping a device in sync with online stuff (and vice versa) and for example spawning notifications about such syncs
Nov 04 05:11:16 <regedit> tdignan: what's so interesting about GCM, or is it just a convenience instead of rolling one's own
Nov 04 05:11:34 <gutano> regedit, push instead of pull
Nov 04 05:12:02 <g00s> i havent used GCM, but i thought it was more of a signal that something changed … not the actual payload; so you still need whatever other machinery to do the actual syncing
Nov 04 05:12:05 <regedit> nothing bu GCM can push to an android device?
Nov 04 05:12:22 <regedit> *but
Nov 04 05:12:23 <SuperRoach> Evening :) Any Australians in here?
Nov 04 05:15:51 <maxlamer> SuperRoach yes
Nov 04 05:16:45 <SuperRoach> maxlamer, curious if you've seen about the 10k prize on the sensis hackathon?
Nov 04 05:17:49 <maxlamer> GCM is really only got notifying apps to collect payload data...  and there is no guaranteed delivery
Nov 04 05:18:19 <maxlamer> RuperRoach no i havn't seen it
Nov 04 05:18:20 <tdignan> regedit: read the docs and find out
Nov 04 05:19:09 <SuperRoach> See what you think: http://yorkbutterfactory.com
Nov 04 05:19:24 <SuperRoach> Oops thats the place, i meant http://www.androidaustralia.org/
Nov 04 05:21:12 <g00s> doest say much
Nov 04 05:22:06 * regedit smells advertisement
Nov 04 05:22:47 <regedit> topic violation?
Nov 04 05:23:52 <SuperRoach> disclaimer: I have nothing to do with it, but i apologise. I was interested in the venue
Nov 04 05:26:12 <gutano> regedit, push instead of pull
Nov 04 05:26:21 <gutano> whoops.. mean to repeat my build in another temr.
Nov 04 05:26:47 <regedit> lol
Nov 04 05:41:05 <regedit> is it possible for an app to download additional functionality (code) as 'downloadable content" ?
Nov 04 05:41:55 <regedit> in other words an app that has options of what "functionality modules" you'd like to download and enable in the app
Nov 04 05:44:01 <arthurdent> i'm no expert, but i've definitely seen apps do that
Nov 04 05:44:10 <arthurdent> i'm pretty sure you have to supply the download server yourself though.
Nov 04 05:44:41 <regedit> are you sure it wasn't just a data download only, and the app already had the code/logic to handle that data from before
Nov 04 05:44:55 <regedit> *beforehand
Nov 04 05:45:26 <gutano> I'm not sure where the line is, but many things need to be present in the Manifest (like Activities, Services)
Nov 04 05:45:36 <arthurdent> yeah i think it was just a data download
Nov 04 05:45:46 <arthurdent> i didn't realize you meant like... custom addons
Nov 04 05:45:48 <gutano> I would guess, as a matter of security, that you can't "extend" an app too much
Nov 04 05:46:05 <regedit> i see...
Nov 04 05:46:19 <gutano> webview is one way to "extend" an app
Nov 04 05:46:19 <regedit> then i have a dillemma
Nov 04 05:46:42 <gutano> and apps can expose functionality to eachother, so you could have 2 apps that have "synergy"
Nov 04 05:46:47 <g00s> http://android-developers.blogspot.com/2011/07/custom-class-loading-in-dalvik.html
Nov 04 05:47:28 <regedit> my dillemma is as follows;
Nov 04 05:47:39 <regedit> i fell in love with SpeechRecognizer and im trying to make an app that can take hands-free voice commands to do all sorts of cool stuff
Nov 04 05:48:14 <regedit> like send messages, shopping lists, todo lists, play music, show wheather, check mail, whatnot
Nov 04 05:48:27 <regedit> i.e. Voice Actions on steroids
Nov 04 05:49:17 <regedit> at first i was thinking of implementing all this stuff in the app itself - ambitious i know
Nov 04 05:49:47 <regedit> but come to think of it, each of those above-mentioned actions are really a separate app in their own right
Nov 04 05:50:00 <gutano> that sounds like it fits right in with the "apps can expose functionality to eachother" pattern, I beleive that is a matter of (public) intent filters (handlers)
Nov 04 05:50:24 <regedit> yes i'm seeing that now
Nov 04 05:50:33 <regedit> which brings me to my dillemma
Nov 04 05:51:08 <regedit> i want this app to be a "center of operations" where the user comes back to after any actions, in a totally hands free manner
Nov 04 05:51:17 <regedit> as if the user never switched away to the other apps
Nov 04 05:51:27 <regedit> and can continue speaking more voice commands
Nov 04 05:52:47 <regedit> so anything like a Chooser screen, or even a "send email" button in the email app the user switched to, breaks the flow of my app
Nov 04 05:53:32 <regedit> makes me once again want to implement all the actions in my own app itself - ambitious i know
Nov 04 05:54:02 <gutano> if "Voice Actions on steriods" wasn't ambitous, it would probably already exist ; )
Nov 04 05:54:59 <regedit> so you're basically telling me that yes, the only choice is to implement everything into the one app in order to keep the experience flow?
Nov 04 05:56:55 <gutano> I try not to speak in such certainties..  I can only say that I don't know how to do it, I would expect the receiving apps to respond to a "no-interaction" flag
Nov 04 05:57:30 <gutano> s/would/would need to/
Nov 04 05:57:53 <gutano> it's late, you get my pt
Nov 04 05:58:06 <regedit> what if...... i make a general center-of-operations app, and its SpeechRecognizer framework which developers can build into their app. everyone who builds this framework into their app can be used seamlessly with my Voice Actions on steroids
Nov 04 05:58:28 <regedit> and this would basically depend heavily on developer-community acceptance / jumping on my bandwagon
Nov 04 05:58:35 <tdignan> i'd say the key part of the sentence is the 'what if'
Nov 04 05:58:44 <tdignan> maybe start coding instead of speculating and it will actually happen
Nov 04 05:58:54 <regedit> i haz teh codez
Nov 04 05:58:59 <regedit> it worx
Nov 04 05:59:11 <tdignan> invisible codes?
Nov 04 05:59:14 <regedit> 102 lines of code
Nov 04 05:59:23 <tdignan> oh, you mean you've got the speech api working
Nov 04 05:59:29 <tdignan> so, what else are you going to do with it?
Nov 04 06:00:00 <regedit> simple
Nov 04 06:00:05 <tdignan> pick something small and targeted and get it working
Nov 04 06:00:26 <regedit> if("lalala".equals(text)) executeThis();
Nov 04 06:00:27 <gutano> better yet, have you reviewed apples patents to see if you aren't going to get sued into oblivion... I'm betting this will be hard to do without rounded corners
Nov 04 06:00:35 <regedit> if("lololo".equals(text)) executeThat();
Nov 04 06:01:12 <gutano> apples got rounded corners on LOCK DOWN
Nov 04 06:01:28 <tdignan> regedit: so what's stopping you then?
Nov 04 06:02:11 <regedit> i was just musing aloud, as i started off saying "what if....."
Nov 04 06:02:44 <tdignan> here's a potentially scientific fact I read about a while back
Nov 04 06:02:56 <regedit> frankly i'm terribly surprised no-one's done anything like this yet
Nov 04 06:03:04 <tdignan> the more you talk about your project idea, the more people will validate it, and the less you will feel like you need to do it.
Nov 04 06:03:11 <tdignan> oh, I've seen apps that do similar things.
Nov 04 06:03:13 <tdignan> They just all suck.
Nov 04 06:03:20 <tdignan> see if you can do better
Nov 04 06:03:37 <gutano> this sounds like a ploy to get me booted from #android... like if I say "siri"
Nov 04 06:03:48 <tdignan> getting booted from here. haha
Nov 04 06:03:52 <tdignan> you'd have to be very juvenile.
Nov 04 06:04:04 <regedit> what's so hard of doing if("we need moar onions".equals(text)) shoppingList(text);
Nov 04 06:04:17 <tdignan> the speech recognitions accuracy
Nov 04 06:04:27 <speakingcode> i wanted to make a joke about hot boy cash money millionaires
Nov 04 06:04:32 <tdignan> things are of course going to be a lot better now that it works locally
Nov 04 06:04:35 <speakingcode> but i don't think most of anyone would get it
Nov 04 06:04:37 <g00s> what if the user meant to open more onion articles ?
Nov 04 06:05:06 <regedit> the app wold have clear instructions about limited vocab
Nov 04 06:05:13 <regedit> think of it as *nix commands
Nov 04 06:05:30 <regedit> "echo" is the one and only command for that purpose
Nov 04 06:05:39 <g00s> regedit: read a book on voice UIs :) there are a few
Nov 04 06:05:49 * tdignan writes down idea for voice-to-unix terminal app
Nov 04 06:05:51 <tdignan> thanks.
Nov 04 06:05:53 <speakingcode> i probably missed something but errr... wtf u talkin about?
Nov 04 06:05:55 <tdignan> free ideas are always appreciated.
Nov 04 06:06:04 <regedit> heh good luck saying ls
Nov 04 06:06:07 <regedit> fsck
Nov 04 06:06:11 <tdignan> "list"
Nov 04 06:06:18 <speakingcode> "el es"
Nov 04 06:07:19 <speakingcode> good luck saying cat mkinitcpio.conf | grep -i ([8?azAZ}**)5
Nov 04 06:07:26 <regedit> hey honestly i'm just in this for the fun atm. looking for ways to make life easier around the house. that's why i called my app FridgeBot
Nov 04 06:07:32 <tdignan> speakingcode: cat | grep? please
Nov 04 06:07:32 <regedit> nexus 7 on the fridge
Nov 04 06:07:35 <g00s> regedit: are you trying to do something like this ? http://www.amazon.com/Practical-Speech-Interface-Factors-Ergonomics/dp/1439815844
Nov 04 06:07:40 <tdignan> grep foo mkinitcpio.conf
Nov 04 06:08:43 <speakingcode> just an example, the point being piping, regexes, etc
Nov 04 06:09:04 <regedit> g00s: i dunno about no fancy textbooks, i speak into SpeechRecognizer and i get very nicely accurate-enough results in text. what else can i ask for
Nov 04 06:09:34 <g00s> so you arent working on the voice recognition part, you are working on the VUI then
Nov 04 06:09:43 <g00s> ok, great … you still can't be a dumbass. get educated
Nov 04 06:09:48 <g00s> :)
Nov 04 06:09:58 <regedit> yup, figured out the speech part within about 3-4 days
Nov 04 06:10:31 <regedit> never having developed with android before, mind you
Nov 04 06:10:36 <regedit> although do have Java experience
Nov 04 06:10:56 <tdignan> see, this was my point. seeking validation from us isn't going to help you write code
Nov 04 06:10:59 <tdignan> just code!
Nov 04 06:11:20 <regedit> might even rig my kitchen with additional mics for convenience/clarity, with some Arduino hack or whatnot
Nov 04 06:11:29 <speakingcode> majke a frt soundboard app thjat lets u record ur own farts
Nov 04 06:11:47 <regedit> you seem to have a fart keyboard
Nov 04 06:12:01 <speakingcode> eating a sub
Nov 04 06:13:14 <regedit> im still pondering about the apps part, whether to implement EVERYTHING into one app (shopping lists, too, email, music, whather, news, etc etc) or whether to make it a framework - will developers care and will they make the apps i'm thinking of.....
Nov 04 06:13:29 <regedit> s/too/todo
Nov 04 06:13:56 <speakingcode> err what?
Nov 04 06:14:15 <speakingcode> b4 making a framework ask what it achieves
Nov 04 06:14:21 <tdignan> speakingcode++
Nov 04 06:14:23 <regedit> tdignan: may i please have your validation about that ^ :D
Nov 04 06:14:33 <regedit> i said
Nov 04 06:14:38 <tdignan> I think you should make a fart app, put admob ads
Nov 04 06:14:53 <regedit> it achieves a streamlined uninterrupted hands free voice commanding of your android
Nov 04 06:14:58 <tdignan> hell, the director of apple's app store did
Nov 04 06:15:12 <tdignan> regedit: i am silently jotting down your ideas and plottng to steal them
Nov 04 06:15:24 <speakingcode> didntv google already do that regedit?
Nov 04 06:15:31 <regedit> tdignan: please just give me the google play link when its done
Nov 04 06:15:53 <tdignan> it's unlikely, i'm in the shit working on my own project already
Nov 04 06:15:53 <regedit> like i said, ATM i really just want the convenience of such a technology, not so much teh moneyz
Nov 04 06:16:03 <tdignan> but maybe i'll take an hour to hammer something funny out.
Nov 04 06:16:42 <regedit> speakingcode: if you're referring to Voice Actions, they should turn that into a framework
Nov 04 06:16:58 <regedit> Voice Actions should be something you can plug into your app
Nov 04 06:17:06 <speakingcode> ou can
Nov 04 06:17:10 <speakingcode> you can*
Nov 04 06:17:17 <regedit> and not only that - but in a totally hands-free manner
Nov 04 06:17:32 <ESphynx> Why am I getting these error:undefined reference to `__fprintf_chk when building with the x86 compiler?
Nov 04 06:17:32 <regedit> like in my SpeechRecognizer loop
Nov 04 06:17:34 <speakingcode> that kills battery
Nov 04 06:17:45 <speakingcode> it would have to always listen
Nov 04 06:17:55 <regedit> i dont care 'cuz its plugged in & stuck on my fridge
Nov 04 06:18:19 <regedit> "hey fridgie, text my wife to STFU"
Nov 04 06:18:23 <speakingcode> so ur making a battery draining framework
Nov 04 06:18:25 <ESphynx> m that's fixed :|
Nov 04 06:18:53 <regedit> you can turn off continuous listening if youd like
Nov 04 06:19:10 <speakingcode> that what the little microphone button does
Nov 04 06:19:18 <speakingcode> ya know, th eone that says "speak now" ;-)
Nov 04 06:19:38 <regedit> did you say Voice Actions is already framework'ized? how so, how can i make my custom voice commands that do things i want
Nov 04 06:20:05 <speakingcode> aside from the active listening
Nov 04 06:20:21 <speakingcode> you basically call the proper intent
Nov 04 06:20:39 <speakingcode> the system will do the recording, make the guess, you can return back a set of results or the "most likely"
Nov 04 06:20:45 <speakingcode> as a string basically
Nov 04 06:20:59 <regedit> link to API pages about this please
Nov 04 06:21:09 <ESphynx> apk/lib/armeabi  X86--> ??
Nov 04 06:22:52 <regedit> speakingcode: really, where do the docs explain this please, i want to read
Nov 04 06:24:28 <speakingcode> looking, one sec
Nov 04 06:25:53 <regedit> why am i remembering a stackoverflow responsa saying Voice Actions is only available for the few things that google hardcoded it for, no custom hacking possible
Nov 04 06:27:40 <speakingcode> http://developer.android.com/reference/android/speech/RecognizerIntent.html
Nov 04 06:28:19 <regedit> fail
Nov 04 06:28:20 <regedit> try again
Nov 04 06:28:51 <ESphynx> How to properly respond to a 'destroy' ?
Nov 04 06:28:51 <regedit> that has nothing to do with voice actions
Nov 04 06:29:02 <ESphynx> I'm getting a destroy time out...
Nov 04 06:29:06 <ESphynx> thought my app was all done
Nov 04 06:30:16 <speakingcode> oh, it doesnt? cause i think i t does
Nov 04 06:30:31 <speakingcode> i think it has everything to do with what i just told you
Nov 04 06:31:21 <speakingcode> ACTION_VOICE_SEARCH_HANDS_FREEStarts an activity that will prompt the user for speech without requiring the user's visual attention or touch input.
Nov 04 06:31:33 <speakingcode> hmmm, sounds familiar...
Nov 04 06:32:31 <regedit> yes RecognizerIntent is for speech recognition, i use it in my code
Nov 04 06:32:39 <regedit> its for recognizing speech
Nov 04 06:32:44 <tdignan> If you get sued by apple, you win
Nov 04 06:32:50 <tdignan> it will be good publicity
Nov 04 06:32:56 <regedit> no mention of lanuching arbitrary apps based on spoken text
Nov 04 06:33:18 <tdignan> g00s would probably read about it on his blogs then come paste a link to it here
Nov 04 06:33:50 <regedit> show me how to lanuch _any_arbitrary_app_ by its name or some-such using the Voice Actions facility
Nov 04 06:34:21 <speakingcode> regediot...
Nov 04 06:34:30 <speakingcode> like i just said
Nov 04 06:34:36 <speakingcode> you run the recognizer - that's step 1
Nov 04 06:34:39 <speakingcode> you get the result
Nov 04 06:34:50 <speakingcode> you match it to your set of available commands, then take the appropriate action
Nov 04 06:35:02 <regedit> lolrofllolrofllolrofllolrofllolrofllolrofllolrofllolrofllolrofllolrofllolrofllolrofllolrofllolrofllolrofllolrofl
Nov 04 06:35:33 <regedit> thank you, i didn't know that. this totally means that android's Voice Actions facility is an open framework
Nov 04 06:35:56 <regedit> not
Nov 04 06:36:15 <speakingcode> what are you talking about?
Nov 04 06:36:33 <speakingcode> how do you normally launch an app?
Nov 04 06:36:42 <speakingcode> how owuld yuo build a launcher?
Nov 04 06:36:59 <speakingcode> without voice?
Nov 04 06:37:01 <regedit> perhaps you confuse android's Speech Recognition facilities with android's Voice Actions feature
Nov 04 06:37:09 <speakingcode> well, do the same thing, and use the speech recongixer to voicify, durrr
Nov 04 06:38:09 <regedit> anyway, nothing to see here, moving along
Nov 04 06:39:38 <ESphynx> so guys, why do I get this destroy timeout now? For all I know my native app is done :\ the thread terminated...
Nov 04 06:40:31 <speakingcode> make ab utton. onclick, speech recognizer. take the result, which is a string. match that string against context.getPackageManager().queryIntentActivities(new Intent(Intent.ACTION_MAIN, null)).addCategory(intent.CATEGORY_LAUNCHER)))
Nov 04 06:40:52 <speakingcode> i mean seriously i think i could do this in one ugly line of code
Nov 04 06:41:33 <regedit> http://xkcd.com/386/
Nov 04 06:42:56 <speakingcode> you try to help someone
Nov 04 06:43:05 <tdignan> ESphynx: you should be stopping everything in onPause
Nov 04 06:43:27 <tdignan> saving any data that needs to be retained
Nov 04 06:43:35 <speakingcode> ESphynx paste some code on pastebin n we mite be able to tell you specifically in your case
Nov 04 06:43:37 <tdignan> using Fragments, though.
Nov 04 06:44:02 <ESphynx> tdignan: at this point all I care is to not get this timeout, so my app can restart cleanly :)
Nov 04 06:44:11 <ESphynx> I hit the 'back' button on the emulator, and that sends me a 'destroy'
Nov 04 06:44:14 <tdignan> so in your onPause() callback stop everything
Nov 04 06:44:19 <regedit> tdignan: btw i dont even know how worth it is to make this kinda app/service/framework, i anticipate google making Voice Actions an open framework anytime soon
Nov 04 06:44:21 <ESphynx> in response to that, my thread exits
Nov 04 06:44:39 <tdignan> i just got done making my app work nice on config changes
Nov 04 06:44:50 <tdignan> completely retains all the state for my custom view inside my fragment
Nov 04 06:44:58 <tdignan> picks up right where it left off :)
Nov 04 06:45:06 <speakingcode> as far as active listening, regedit, all you need to do is keep the mic on and run the input through a pretty basic classifier to decide the user is talking
Nov 04 06:45:13 <tdignan> are you using fragments ESphynx ?
Nov 04 06:45:19 <tdignan> if not, you should be
Nov 04 06:45:35 <ESphynx> V/ecere-app( 1989): THE END.
Nov 04 06:45:35 <ESphynx> W/ActivityManager( 1013): Activity destroy timeout for ActivityRecord{b48c3a08 com.ecere.Hello/android.app.NativeActivity}
Nov 04 06:45:43 <speakingcode> once it classifies a positive you can then pass the input throug the speech recognizer
Nov 04 06:45:45 <ESphynx> THE END is right before my thread terminated
Nov 04 06:45:51 <speakingcode> as described above
Nov 04 06:45:58 <regedit> speakingcode: not so simple actually, finding intentional speech among random chatter
Nov 04 06:46:00 <tdignan> ESphynx: oh, native activity eh
Nov 04 06:46:04 <tdignan> does it have onPause callback?'
Nov 04 06:46:19 <speakingcode> regedit that problem has largely been well solved
Nov 04 06:47:15 <ESphynx> tdignan: there's an onpause, but my issues is with onDestroy, not onPause :P
Nov 04 06:47:50 <regedit> "...massive catastrophic proportions of unpredictable scope hey fridgie we need more beer so anyway as i was saying cosmological apocalypse..."
Nov 04 06:47:56 <speakingcode> i'm sure some basic features o nthe audio signal are suitable for binary classification
Nov 04 06:48:11 <speakingcode> well, make a reasonable init command
Nov 04 06:48:20 <speakingcode> "fridgie! we need more beer..."
Nov 04 06:48:27 <speakingcode> like xbox does
Nov 04 06:48:51 <speakingcode> "xbox! open netflix... xbox! find the lion king
Nov 04 06:48:55 <ESphynx> speakingcode: https://github.com/ecere/sdk/blob/master/ecere/src/gui/drivers/AndroidInterface.ec
Nov 04 06:50:42 <speakingcode> obviously relying on a fixed init signal is gonna be easier to implement than distinguising among chatter, pick ur battles
Nov 04 06:50:46 <regedit> speakingcode: i do anticipate this kind of functionality being implemented either into the android OS or at least google apps (voice actions/search etc)
Nov 04 06:51:14 <regedit> how wrong might i be about that
Nov 04 06:51:42 <regedit> and then i'd have wasted my weekends for no reason
Nov 04 06:51:45 <speakingcode> everythingi s there fr it
Nov 04 06:52:10 <speakingcode> go do it n make a rough cut within n app, put it out there
Nov 04 06:52:31 <regedit> except google didnt make it an official API kind of thing, to make your app available from a continuous listening mic, or from the voice actions button
Nov 04 06:52:45 <regedit> but i anticipate them doing this soon
Nov 04 06:52:48 <regedit> unless im very wrong
Nov 04 06:52:53 <speakingcode> not sure
Nov 04 06:52:58 <speakingcode> i don't know the mechanics of the mic on the android
Nov 04 06:53:06 <speakingcode> may vary by device too
Nov 04 06:53:16 <speakingcode> but mics are usually low power
Nov 04 06:53:20 <regedit> lots of things vary by device
Nov 04 06:53:33 <ESphynx> well that's interesting. my onDestroy() is not returning... it's not my main thread that's the problem :|
Nov 04 06:53:36 <regedit> even if i hack this app together myself, itll still vary by device
Nov 04 06:53:43 <speakingcode> so having it doing passive listening is reasonable
Nov 04 06:54:17 <regedit> my target is pretty much tablet devices that are plugged in and stuck to a fridge
Nov 04 06:54:21 <regedit> or toilet seat
Nov 04 06:54:33 <regedit> not so much phones
Nov 04 06:55:06 <speakingcode> the passive listener just needs to take samples of data, and classify that a command is happening - that will be exponentially lighter than the actual speech recognition
Nov 04 06:55:21 <regedit> phones is something you hold in your hand. if you're holding it anyway, smash your thumb on it and talk
Nov 04 06:55:56 <regedit> and while you're at it, browse to the app you need and stfu
Nov 04 06:55:57 <speakingcode> the problem is that dependingo n your classifier slicing the input stream into samples can result in slicing the signal for the command, and cause an incorrect negative
Nov 04 06:56:14 <regedit> but what im trying to make is for the car, or the fridge
Nov 04 06:56:30 <regedit> or desk
Nov 04 06:56:40 <speakingcode> but that problem is solved by a sliding window approach, i know that solution exists becasue i worked with people that developed it in the past
Nov 04 06:57:34 <regedit> i dont need to slice the audio data, ill just slice the resulting text itself no?
Nov 04 06:57:51 <speakingcode> i mean before you get to speech recognition
Nov 04 06:57:57 <speakingcode> so like the mic is always on
Nov 04 06:58:00 <speakingcode> passively listening
Nov 04 06:58:00 <regedit> why bother
Nov 04 06:58:14 <speakingcode> and it waits to determine "yes, the person is barking a command at me"
Nov 04 06:58:17 <regedit> send everything to speech recognition, slice the resulting text
Nov 04 06:58:22 <speakingcode> THEN it does speech recognition.
Nov 04 06:58:26 <speakingcode> well you could
Nov 04 06:58:36 <regedit> hog googles servers to the bone
Nov 04 06:58:54 <speakingcode> but then you're constnatly sendnig to speech recognition - you still have to slice and rejoin the segments you send, and that will perform crappy
Nov 04 06:59:48 <speakingcode> the load of loading the speech recognizer, doing the process, returning the result, constantly, then building some text stream, parsing it out, looking for commands... awfully heavy especially if yuo want a general framework
Nov 04 07:00:02 <regedit> commands arent that long
Nov 04 07:00:19 <speakingcode> right but how do you know when a command starts and stops if it's always listening and there's chatter?
Nov 04 07:00:45 <regedit> speechRecognizer.startListening(intent); is all it takes every round
Nov 04 07:01:03 <speakingcode> that's why i'm saying first "passively listen" and take the mic data (always on right) through a classifier that determins "bam! a command is happening now start recording for speech recognizer!"
Nov 04 07:01:06 <regedit> whatever, i have it all figured out about that
Nov 04 07:01:28 <regedit> too complicated playing with audio data
Nov 04 07:01:36 <speakingcode> nah, pretty simple
Nov 04 07:01:44 <regedit> much easier to hog googles servers
Nov 04 07:01:49 <speakingcode> there's plentyo f literature and code out there for this
Nov 04 07:01:51 <speakingcode> nah
Nov 04 07:02:17 <speakingcode> the input comes ni faster than you can constantly slice, recognize, result, rejoin, find command
Nov 04 07:02:51 <regedit> i have to do my own speech recognition then, to find the init utterance
Nov 04 07:02:54 <speakingcode> test it - see how fast you can take one segment of speech and get a result back - probably not near as fast as you can continuously talk in a conersation
Nov 04 07:02:58 <speakingcode> nope
Nov 04 07:03:01 <speakingcode> you only need to classify it
Nov 04 07:03:04 <speakingcode> not recognize it
Nov 04 07:03:38 <speakingcode> you can use features of the singal like amplitude over duration, for instance, or other basic statistis related to this sort of thing - there's lots of literature and code out there for this
Nov 04 07:04:04 <speakingcode> 2 steps - determine IF a command, then record and recognize it if so
Nov 04 07:04:06 <regedit> this is too complex for noob like me, ill just hog googles servers
Nov 04 07:04:29 <speakingcode> certainly feel free to try, i just fear its not near instaneous enough
Nov 04 07:04:40 <regedit> doesnt need to be
Nov 04 07:04:57 <regedit> for example, siri isnt either, but its acceptable enough
Nov 04 07:05:10 <speakingcode> what will happen is you talk for 15 seconds of chatter, then say a command, and by the time you say the command the first 15 seconds (even as multiple segments) is still processing
Nov 04 07:05:12 <tdignan> ESphynx: any luck with onPause & shutting your stuff down before it leaks?
Nov 04 07:05:18 <regedit> btw ops kick me now for mentioning siri
Nov 04 07:05:19 <speakingcode> and then the command is missed
Nov 04 07:05:30 <ESphynx> tdignan: onDestroy you mean? :P
Nov 04 07:05:36 <tdignan> no no!
Nov 04 07:05:46 <ESphynx> why do you keep talking about onpause :P
Nov 04 07:05:50 <ESphynx> i don't case about onpause :P
Nov 04 07:05:52 <hoangtg> when open send mail form (via android.content.Intent.ACTION_SEND), the cursor is at the bottom of content textview (my preset text is long), is there a way I can move to the top of the form (either the cursor or just scroll the form up to top)?
Nov 04 07:05:52 <tdignan> onDestroy is an unreliable call only meant for freeing up resources. It should never be used to like, stop a thread.
Nov 04 07:05:55 <regedit> speakingcode: ill just throw a "whuut? fridgie dont grok what you said. come again please"
Nov 04 07:05:56 <speakingcode> ESphynx don't do anything in ondestroy
Nov 04 07:05:57 <tdignan> ESphynx: that's what you dont' get
Nov 04 07:05:59 <tdignan> onDestroy SUCKS
Nov 04 07:06:04 <tdignan> onPause rules
Nov 04 07:06:06 <tdignan> :P
Nov 04 07:06:10 <ESphynx> I need to use onDestroy here man :P
Nov 04 07:06:14 <speakingcode> regedit yu can, i guess
Nov 04 07:06:19 <ESphynx> this is the native app stuff :P
Nov 04 07:06:30 <speakingcode> but it will constantly be saying that over and over
Nov 04 07:06:32 <tdignan> let me take a look at the native activity
Nov 04 07:06:44 <speakingcode> while you and your drinkng buddy argue over who has the loudest fart app
Nov 04 07:06:55 <speakingcode> and constantly missing your commands
Nov 04 07:07:12 <regedit> speakingcode: ill make initial beep like siri
Nov 04 07:07:13 <tdignan> ESphynx: the thing about onDestroy is that it's often not going to be called.
Nov 04 07:07:24 <regedit> where appliccable
Nov 04 07:07:30 <speakingcode> regedit, when do you trigger the beep, if it's constantly listening?
Nov 04 07:08:03 <speakingcode> i think there is a circular logic going on you're missing, maybe. and i don't say this to be a dick i'm just seeing an issue and i'm not sure you do
Nov 04 07:08:11 <tdignan> you might clean up your threads and stuff in onDestroy, but don't use it as a place for stopping them.
Nov 04 07:08:17 <regedit> speakingcode: when you prompt it to (like when you say just "fridgie!" and wait a bit, or when a "fridgie dont grok" gets thrown)
Nov 04 07:08:26 <lasserix> Anyone know why my action bar menu icons wouldn't be centered in their spaces?
Nov 04 07:08:28 <speakingcode> right
Nov 04 07:08:32 <speakingcode> but how does it know you prompted it?
Nov 04 07:08:43 <speakingcode> by doing speech recognition you say right
Nov 04 07:08:47 <speakingcode> so it keeps doing it over and over
Nov 04 07:09:14 <speakingcode> and it doesn't recognize the "init" command.. so it does what? says it doesn't recognize, over and over?
Nov 04 07:09:33 <ESphynx> tdignan I realize that, I'm handling the case when it 'does' get called
Nov 04 07:09:41 <regedit> if recognition obtained just the word "fridgie!" then it goes "yes? <MEEEP>"
Nov 04 07:10:25 <regedit> if recognition obtained "bla bla bla fridgie bla bla bla bla" then it goes "whuut? <meeeep>"
Nov 04 07:10:34 <ESphynx> can't we standardize on just popping beers?
Nov 04 07:10:38 <speakingcode> ok so what happens when you're chattering and it gets behind constantly processing looking for "fridgie"
Nov 04 07:11:06 <regedit> how so
Nov 04 07:11:17 <speakingcode> cause the processing isn't instant
Nov 04 07:11:28 <speakingcode> thats why i'm saying check the timing of it.. it may work but i fear it isn't fast enough
Nov 04 07:11:29 <regedit> its a simple text.contains("fridgie")
Nov 04 07:11:41 <speakingcode> right
Nov 04 07:11:44 <ESphynx> I was missing an incref app =)
Nov 04 07:11:50 <speakingcode> but the input stream (your voice) is sconstantly coming in
Nov 04 07:11:56 <speakingcode> it's producer consumer problem see
Nov 04 07:12:02 <speakingcode> you produce input faster than it can consume
Nov 04 07:12:50 <regedit> i think its pretty improbable that this missing-init-word issue happens more than say, twice. the user will definitely get the meep within 2 tries
Nov 04 07:13:09 <g00s> man you guys ar estill talking about this :P
Nov 04 07:13:21 <ESphynx> I'm liking this =) I think I've fixed it =)
Nov 04 07:13:24 <speakingcode> maybe so.. put it to test i suppose
Nov 04 07:13:32 <speakingcode> if simpler works, then hey it works
Nov 04 07:13:34 <ESphynx> So guys, you can now code Android apps in eC =)
Nov 04 07:14:10 <speakingcode> its jsut a loto f resources, which is generally "bad practice"
Nov 04 07:14:11 <regedit> wtf is eC and why should i care. i already dont
Nov 04 07:14:18 <speakingcode> lol
Nov 04 07:14:42 <ESphynx> It's a native/compiled/OO/modern/reflection/properties/C-like language I designed and you should care because it's awesome.
Nov 04 07:15:11 <ESphynx> and because java sucks ass.
Nov 04 07:15:12 <speakingcode> oh another one! :p
Nov 04 07:15:21 <speakingcode> why does java suck?
Nov 04 07:15:21 <ESphynx> no, it's the only one :P
Nov 04 07:15:25 <ESphynx> 'cuz it's java =)
Nov 04 07:15:30 <regedit> ESphynx: what was missing so much that you made eC
Nov 04 07:15:40 <tdignan> g00s: ikr
Nov 04 07:15:42 <speakingcode> is ec dynamic or static typed?
Nov 04 07:15:45 <ESphynx> regedit: a native/compiled/C-like OO language that did not suck like C++
Nov 04 07:16:13 <ESphynx> speakingcode: kindof both. I don't care about dynamic or statically type classifications :P
Nov 04 07:16:23 <speakingcode> and you mae this langauge
Nov 04 07:16:23 <regedit> ESphynx: RAII built in?
Nov 04 07:16:32 <speakingcode> and yuo don't care about one of the must fundamental aspects of a language
Nov 04 07:16:33 <speakingcode> great
Nov 04 07:16:35 <ESphynx> What's RAII?
Nov 04 07:16:36 <speakingcode> sounds like a winer
Nov 04 07:16:45 <regedit> and you ask what RAII is
Nov 04 07:17:39 <ESphynx> yeah none of that nonsense is needed :P
Nov 04 07:17:42 <speakingcode> what does it compile to?
Nov 04 07:17:51 <ESphynx> speakingcode to native code...
Nov 04 07:17:57 <speakingcode> ok good day
Nov 04 07:18:00 <ESphynx> through intermediate C right now =)
Nov 04 07:18:11 <regedit> ESphynx: garbage collection? reference counting?
Nov 04 07:18:24 <ESphynx> it's got a bit of reference counting, it's gonna get a bit more...
Nov 04 07:18:26 <tdignan> it's fucking C, why would it have garbage collection
Nov 04 07:18:41 <speakingcode> can't find anything about ex
Nov 04 07:18:43 <tdignan> oh, surprise
Nov 04 07:18:43 <speakingcode> ec
Nov 04 07:18:51 <tdignan> http://www.ecere.com/
Nov 04 07:19:08 <ESphynx> here's what my android sample look like: https://github.com/ecere/sdk/blob/master/samples/android/hello/helloAndroid.ec
Nov 04 07:20:48 <regedit> ESphynx: so for your last chance, show me 3 quick code snippets where eC makes my life easier than c/c++/java
Nov 04 07:20:49 <speakingcode> you say no concern of typing but the first thing i seei nt he doucmentation is data types
Nov 04 07:20:59 <ESphynx> regedit : well that was 1
Nov 04 07:21:02 <speakingcode> so are they statico r dynamic? iot has to be one or the other, i'm pretty damn sure of that
Nov 04 07:21:24 <ESphynx> That's 2  https://github.com/ecere/sdk/blob/master/samples/3D/Test3D/Test3D.ec
Nov 04 07:22:00 <ESphynx> https://github.com/ecere/sdk/blob/master/samples/db/MovieCollection/movieCollection.ec  -- I guess that's 3 =)
Nov 04 07:22:15 <ESphynx> test3D == http://ecere.com/images/test3D.png
Nov 04 07:22:28 <regedit> ESphynx: this is meant for android only or general computing platforms (desktop etc)
Nov 04 07:22:45 <ESphynx> speakingcode: http://rosettacode.org/wiki/Category:EC  -- see type checking there =)
Nov 04 07:23:05 <ESphynx> regedit: desktop etc. been working on the SDK since 1996... initially as a game SDK =)
Nov 04 07:23:27 <regedit> 1996? i feel so bad....
Nov 04 07:23:36 <regedit> :D
Nov 04 07:23:38 <ESphynx> :P
Nov 04 07:23:41 <speakingcode> what kind of libraries are there, or does it support c libs?
Nov 04 07:24:07 <ESphynx> speakingcode: it supports C libs... and Ecere is a cross platform GUI toolkit , also network libs , system libs
Nov 04 07:24:10 <ESphynx> DB libs
Nov 04 07:24:22 <ESphynx> 3D lib
Nov 04 07:25:12 <speakingcode> hmm
Nov 04 07:25:17 <speakingcode> who uses it?
Nov 04 07:25:20 <ESphynx> Me =)
Nov 04 07:25:31 <speakingcode> looks intriguing but i fear using it for anyhing thats non-experimental
Nov 04 07:26:15 <ESphynx> haven't reached the momentum yet to be above the air friction
Nov 04 07:26:29 <ESphynx> but that's coming :P I'm hoping Android/iOS support will help =)
Nov 04 07:26:47 <speakingcode> it might
Nov 04 07:26:57 <speakingcode> website needs a full rewrite
Nov 04 07:27:15 <ESphynx> I'm actually rather happy with the website we have right now :P
Nov 04 07:27:17 <speakingcode> modernize that puppy with some twitter bootstrap or smoething n the adoption will skyrocket
Nov 04 07:27:26 <ESphynx> It's certainly better than in the past. it can benefit from improvements.
Nov 04 07:27:27 <speakingcode> it looks like garbage, no offense
Nov 04 07:27:34 <ESphynx> That Sea Otter mascott? ;)
Nov 04 07:27:41 <ESphynx> Hmm, thanks, I guess :P
Nov 04 07:27:43 <speakingcode> it doesn't shout "good developers want this"
Nov 04 07:28:03 <ESphynx> depends what websites you compare too I guess :P
Nov 04 07:28:09 <speakingcode> more like "im old and unsupported" i hate to be so harsh
Nov 04 07:28:15 <speakingcode> check out, knockoutjs website
Nov 04 07:28:21 <ESphynx> i just don't have money to give away to a design firm to get a design that in the end I like less that the current site.
Nov 04 07:28:37 <speakingcode> http://knockoutjs.com/
Nov 04 07:28:49 <ESphynx> see
Nov 04 07:28:56 <speakingcode> see there, alignment, typography, whitespace...
Nov 04 07:28:57 <ESphynx> I prefer 100 times http://www.ecere.com/ over that garbage.
Nov 04 07:28:58 <g00s> why would anyone want to use eC ?
Nov 04 07:29:17 <speakingcode> well swallow your pride or don't get adopted, thats on you
Nov 04 07:29:41 <ESphynx> speakingcode that wasn't about pride, it was about design tastes.
Nov 04 07:29:53 <speakingcode> but i'm telling you your site failsm iserable at serving a purpose. it should focues on one thing and call attention to why i want eC and how to get it
Nov 04 07:29:56 <ESphynx> g00s: to get C performance
Nov 04 07:30:07 <infernus> Hi
Nov 04 07:30:14 <regedit> windows95 Netscape Navigator tastes
Nov 04 07:30:18 <ESphynx> speakingcode I agree it could benefit from improvements
Nov 04 07:30:18 <speakingcode> just my opinon, of course. i tend to speak it matter-of-factly
Nov 04 07:30:44 <speakingcode> i like the otter
Nov 04 07:30:46 <g00s> the website needs rounded corners :P
Nov 04 07:30:49 <ESphynx> thanks
Nov 04 07:30:55 <ESphynx> it's got round buttons!
Nov 04 07:31:04 <infernus> I need to know about callback url for foursquare
Nov 04 07:31:20 <regedit> you can keep the otter, thats not the main prob
Nov 04 07:31:43 <ESphynx> well, thanks you guys for not calling it a cat.
Nov 04 07:31:50 <regedit> main issue here is to upgrade to web 2.0 from uh,
Nov 04 07:31:54 <speakingcode> the thing is, from my perspectiv,e is my attentino goes right to the otter and i don't fee lcompelled to look anywhere else
Nov 04 07:31:55 <ESphynx> it's Oliver, the Sea Otter!
Nov 04 07:32:18 <ESphynx> LOL
Nov 04 07:32:20 <speakingcode> i'm worried this is gonna be a 404 or a virus or something
Nov 04 07:32:24 <infernus> I need aauthorization process that returns back to the application
Nov 04 07:32:31 <ESphynx> that's interesting
Nov 04 07:32:31 <speakingcode> not really, but that's the first impression
Nov 04 07:32:36 <ESphynx> I do agree Oliver attracts your attention.
Nov 04 07:32:54 <ESphynx> I should add sound that says ' psst.. try out eC man, it's awesome.
Nov 04 07:33:00 <ESphynx> with visual effects on his lips
Nov 04 07:33:01 <speakingcode> i'll give you another example since you don't like knockout
Nov 04 07:33:05 <speakingcode> nooo you shouldn't
Nov 04 07:33:07 <ESphynx> sure
Nov 04 07:33:08 <ESphynx> lol
Nov 04 07:33:17 <regedit> graphically, the LAST thing a visitor will ever ever do is the right-most button "Contribute"
Nov 04 07:33:55 <ESphynx> regedit: well I don't expect a newcomer to click that button =)
Nov 04 07:33:56 <regedit> it's sooooooo out of the way from center of attention
Nov 04 07:34:02 <infernus> Guys.... callback url... anyone??
Nov 04 07:34:04 <ESphynx> I would only want longtime users to contribute code :P
Nov 04 07:34:17 <ESphynx> only after they visit this page 200 times can they even think about contributing :P
Nov 04 07:34:32 <speakingcode> http://www.codecademy.com https://workforpie.com/http://offliberty.com/
Nov 04 07:34:42 <ESphynx> Download, Learn more, Get support... is where you're supposed to click :)
Nov 04 07:34:47 <regedit> so much imbalance too, the hueg logo vs the unevenly layed out buttons
Nov 04 07:35:02 <speakingcode> esphynx if you don't expect them to click that button, get it the f off the top of the fold
Nov 04 07:35:55 <speakingcode> check those three sites ESphynx and notice a running them - white space and alignment and typography-  easy on the eyes and draws the user right to the focal point, no questino about what is going on
Nov 04 07:36:42 <g00s> ok, ok, the website needs work … that seems minor
Nov 04 07:36:47 <infernus> Has anyone here integrated an android app with foursquare????
Nov 04 07:36:57 <g00s> reading through the pages, i still dont understand why anyone would use this for anything
Nov 04 07:37:00 <speakingcode> nope sorry infernus
Nov 04 07:37:34 <speakingcode> g00s its not IMO... with all the myriad competing frameworks and such in that space he won't get adoption without a slick site, point blank
Nov 04 07:37:48 <ESphynx> speakingcode: I do want them to understand it's open source software and they can eventually contribute =)
Nov 04 07:38:07 <ESphynx> g00s: why would anyone NOT use it?
Nov 04 07:38:16 <speakingcode> sure, and i wouldn't exclude that, but it's not necessarily the focua
Nov 04 07:38:35 <infernus> Dude...  em so hopeless now... searching since last two days about what's a callback url..
Nov 04 07:38:49 <speakingcode> nfernus is it jsonp?
Nov 04 07:39:09 <regedit> speakingcode: codecademy has really awesome immeadiate user captivation with that interactive prompt tutorial, i actually played it through
Nov 04 07:39:17 <speakingcode> ESphynx i'll just be honest. if i weren't talking to you in here, if i just happened across this, i would immediately have clicked back
Nov 04 07:39:40 <infernus> No... oauth authorization process for foursquare
Nov 04 07:40:02 <regedit> ESphynx: looks like an old irrelevant website
Nov 04 07:40:07 <g00s> speakingcode: i guess what i;m saying is, no matter how nice the website is, i don't see the value proposition for using this
Nov 04 07:40:16 <ESphynx> well well. website needs improvements.
Nov 04 07:40:17 <tdignan> infernus: it's the URL that the server responds with your authorization code/token to
Nov 04 07:40:19 <infernus> After authorization, the control should get back to the app
Nov 04 07:40:26 <tdignan> infernus: for client side oauth, often you use a local http server.
Nov 04 07:40:42 <tdignan> infernus: search oauth on d.android.com, since they have an api that abstracts this for you already.
Nov 04 07:40:44 <speakingcode> nothing about the page makes me feel like this is a framework langauge or platform i want anything to do with, and being that it says "cross platform" it should look damn good on the web, ya know?
Nov 04 07:40:45 <ESphynx> g00s: A lot of people don't see it.
Nov 04 07:40:49 <ESphynx> I don't understand though.
Nov 04 07:40:52 <infernus> Owkk... thanks tdignan..
Nov 04 07:41:02 <regedit> g00s: http://www.ecere.com/wiki/index.php?title=Main_Page#Yet_a_New_Language..._Why_eC_over_....3F
Nov 04 07:41:04 <speakingcode> i mean, it's unfair to rate this by its site and not by what it really is, but i'm just seaking n what most people are likely to do when they see this
Nov 04 07:41:10 <infernus> But what is it for an android app
Nov 04 07:41:19 <ESphynx> eC is the ONLY programming language that is Native and Compiled, and has modern features such as reflection, properties and import mechanism.
Nov 04 07:41:30 <tdignan> infernus: read on d.android.com. there is a whole guide on oauth with android.
Nov 04 07:41:45 <speakingcode> so eC can interact with the native android sdk components?
Nov 04 07:41:49 <ESphynx> And it comes with a cross platform IDE and cross platform toolkit.
Nov 04 07:41:54 <g00s> regedit: yes, saw that. nothing is convincing :/
Nov 04 07:42:09 <ESphynx> speakingcode: yeah I got it hooked with the NDK and everything
Nov 04 07:42:15 <tdignan> http://developer.android.com/training/id-auth/authenticate.html
Nov 04 07:42:16 <ESphynx> see that Android sample I put above =)
Nov 04 07:42:21 <speakingcode> doesi t compile to machine code or to java byte code?
Nov 04 07:42:23 <speakingcode> oh
Nov 04 07:42:40 <ESphynx> g00s: what's not convincing about that? if you don't care about native, compiled or modern features, I understand how this has no proposition for you. But *I do*
Nov 04 07:42:47 <infernus> For Twitter it was 'twitter-oauthflow-something'
Nov 04 07:43:41 <speakingcode> ESphynx i'm _curious_ about your langauge and i'm not trying to knock you, i just thik you would see a surge in success with a better presentation
Nov 04 07:43:53 <speakingcode> remember selling is all in the presentation
Nov 04 07:44:28 <g00s> no
Nov 04 07:44:39 <infernus> Thanks tdignan.... I will go through the link you sent
Nov 04 07:44:53 <infernus> I guess it might help
Nov 04 07:45:07 <ESphynx> speakingcode I do agree there.
Nov 04 07:45:29 <g00s> ESphynx: post to HN
Nov 04 07:45:36 <ESphynx> hacker news?
Nov 04 07:45:39 <g00s> yeah
Nov 04 07:45:48 <speakingcode> i would think about shrinking the logo 10x and putting a couple of app screenshots, maybe one android, one ios, one desktop? be creative there. then have some large bold calls to action - download, learn more, contribute
Nov 04 07:45:52 <ESphynx> I don't know . I'll let others post when people care enough =)
Nov 04 07:46:03 <ESphynx> /. never ran an article on it :P
Nov 04 07:46:10 <ESphynx> and it easy dissappear from programming.reddit
Nov 04 07:47:27 <regedit> website doesnt look like its been designed within 5 years of reddit launch
Nov 04 07:47:51 <asterisk-Tester> Hi has anyone compiled successfully ndk sample razzlegames : http://code.google.com/p/razzlegames-android-ndk-tutorial/source/checkout ? or have good experiance with stlport for ndk ?
Nov 04 07:47:55 <speakingcode> i would do away with those platform logos all together and just use text under the main "top call to action" section somewhere , shrug
Nov 04 07:49:04 <ESphynx> I kinda like my front page :|
Nov 04 07:49:13 <ESphynx> Open space zen design'ish :|
Nov 04 07:49:32 <ESphynx> you guys are all web 2.0 brainwashed
Nov 04 07:49:52 <regedit> i think ESphynx is pretty happy with his userbase
Nov 04 07:49:56 <ESphynx> lol
Nov 04 07:50:11 <ESphynx> I generate enough bug reports by myself as a uer.
Nov 04 07:50:12 <ESphynx> user*
Nov 04 07:50:55 <regedit> sounds like i'm right
Nov 04 07:51:22 <regedit> it's a fun lifetime solo project
Nov 04 07:51:54 <regedit> meant only for people of "his type"
Nov 04 07:54:59 <tdignan> guys, take it to #amateurish-criticism.
Nov 04 07:55:44 <ESphynx> yeah, and I won't join you there :P
Nov 04 07:55:58 <ESphynx> regedit -- I guess that is partially true =)
Nov 04 07:56:04 <_deXter_> He lies, there's no #amateurish-criticism !
Nov 04 07:56:04 <regedit> nobody like nowadays tastes/style/marketing techniques, but what sells sells
Nov 04 07:56:17 <lasserix> Anyone know any good design for listview rows with block text on linear gradient background?
Nov 04 07:56:30 <ESphynx> regedit -- at some point I will hire a marketing expert :P
Nov 04 07:56:41 <regedit> ESphynx: i doubt it
Nov 04 07:56:47 <ESphynx> but all that was beyond the point
Nov 04 07:56:48 <regedit> its fun asi is!
Nov 04 07:56:55 <regedit> *as is
Nov 04 07:57:16 <ESphynx> I just wanted to share with you the fact that you can now use eC for writing Android apps, for those who care (if any) :) just pushed a commit that fixed my onDestroy handling...
Nov 04 07:57:49 <regedit> let celebrate then
Nov 04 07:57:55 <regedit> *let's
Nov 04 07:58:19 <ESphynx> damn right
Nov 04 07:58:28 <ESphynx> I'm going to take out the Wyobora Wodka and Ungava Gin
Nov 04 07:58:34 <ESphynx> with a splash of Kina Lillet...
Nov 04 07:58:50 <ESphynx> if I can find ice cubes, that is
Nov 04 07:58:53 <ESphynx> I like it shaken, not stirred!
Nov 04 08:00:45 <speakingcode>  that's what she said
Nov 04 08:01:30 <ESphynx> lol
Nov 04 08:04:59 <infernus> I'm still in limbo...  :(
Nov 04 08:05:22 <infernus>  Foursquare android integration
Nov 04 08:05:32 <infernus> Callback url
Nov 04 08:05:54 <regedit> infernus: You're waiting for a train, a train that will take you far away. You know where you hope this train will take you, but you don't know for sure
Nov 04 08:06:11 <infernus> Lol..
Nov 04 08:07:09 <infernus> But I'm quite serious... it's being days em brainstorming over this glitch...
Nov 04 08:07:46 <SuperRoach> what's it doing?
Nov 04 08:08:03 <regedit> i know one day ill be stuck in your situation and ill regret not having tried to solve heh
Nov 04 08:08:28 <lasserix> What's recomended square size for background ninepatch?
Nov 04 08:08:32 <infernus> Correct callback url for foursquare in order to get back to app after oauth authorization
Nov 04 08:08:57 <infernus> ???
Nov 04 08:10:04 <sampullman> lasserix: what do you mean recommended?
Nov 04 08:10:22 <sampullman> as small as possible, I would say...
Nov 04 08:10:35 <Leeds> infernus: https://github.com/foursquare/android-oauth-example doesn't work?
Nov 04 08:11:40 <lasserix> Yeah... hmm maybe I can just use more drawables
Nov 04 08:12:22 <infernus> Yes Leeds... it works
Nov 04 08:12:38 <infernus> But it's not using callback url
Nov 04 08:12:46 <Leeds> correct, it's not
Nov 04 08:13:19 <infernus> It's doing manual handling by intercepting client methods
Nov 04 08:13:24 <Leeds> yes, it is
Nov 04 08:13:44 <infernus> I need to do it through native browser
Nov 04 08:13:44 <Leeds> you have correctly understood the 40 lines of example code
Nov 04 08:13:46 <Leeds> why?
Nov 04 08:14:29 <infernus> There are cases where a user doesn't trust third party apps while providing credentials
Nov 04 08:14:45 <infernus> They feel Luke getting phished
Nov 04 08:14:52 <infernus> Like*
Nov 04 08:15:06 <Leeds> Luke... I am your identity thief
Nov 04 08:15:38 <infernus> So I need to provide authentication process using browser as well
Nov 04 08:16:23 <infernus> Leeds do you have any idea about this implementation??
Nov 04 08:16:25 <Leeds> well... firstly I think you're wrong
Nov 04 08:16:50 <infernus> Ummm...? Wrong with what?
Nov 04 08:16:51 <Leeds> I don't think there's any issue doing it internally - it's not like you're asking the user to provide their bank details
Nov 04 08:17:35 <infernus> Yes.. not that grave.. but people often don't trust third-party apps
Nov 04 08:18:12 <infernus> You would never like anyone getting into your FB or Twitter account
Nov 04 08:18:14 <Leeds> then they might not use your app's foursquare integration... is that a deal-breaker?
Nov 04 08:18:34 <Leeds> alternatively, it's time to get down and intimate with intent-filters
Nov 04 08:18:51 <infernus> Ummm. Actually em bent to do it... client's requirement
Nov 04 08:19:15 <infernus> Otherwise I don't give a crap about foursquare.. :P
Nov 04 08:19:40 <Leeds> the client requires authentication outside your app?
Nov 04 08:19:44 <infernus> Intent filter... yea
Nov 04 08:19:58 <infernus> Yes... through browser
Nov 04 08:20:08 <Leeds> time to smack the dumb client around then
Nov 04 08:20:41 <infernus> Control goes to the browser, user authenticates, control should get back to the app with access token
Nov 04 08:21:05 <infernus> :P... the dumb client is paying me a million..
Nov 04 08:21:46 <Leeds> then I'm going to shut up until I get cut in
Nov 04 08:22:20 <infernus> The last step in the required flow needs a callback url that makes browser to identify the activity that should catch back the control
Nov 04 08:22:45 <Leeds> sure
Nov 04 08:22:47 <infernus> Leeds.....duuuuude!!!  Cmonnn!! :P
Nov 04 08:23:27 <infernus> Please do help if you can..
Nov 04 08:23:30 <infernus> Thanks
Nov 04 08:24:22 <Leeds> pastebin your existing intent-filter
Nov 04 08:24:32 <Leeds> the one which doesn't work
Nov 04 08:26:10 <infernus> Oh Shit... email doing irc through cellphone... will login through system and will tell you about my current intent-filter
Nov 04 08:26:39 <infernus> Actually it's normal , just with additional scheme and hot attributes
Nov 04 08:27:08 <infernus> Host*
Nov 04 08:27:33 <speakingcode> i've decided i'm not voting in this stupid election
Nov 04 08:28:34 <infernus> Speakingnode... you probably need an android app to vote
Nov 04 08:28:42 <infernus> :p
Nov 04 08:28:53 <infernus> Don't ya!!??!
Nov 04 08:28:57 <Leeds> keep the politics out
Nov 04 08:28:57 <speakingcode> i'd like to start a national don't vote campaign
Nov 04 08:29:04 <speakingcode> k
Nov 04 08:29:13 <lasserix> its called third party
Nov 04 08:29:29 <infernus> Yes actually... it corrupts the room.... android is so pure..
Nov 04 08:29:49 <speakingcode> pure as pure itself
Nov 04 08:29:57 <Leeds> I'm serious - no politics here
Nov 04 08:30:02 <infernus> Leeds... do you often keep inside this room?
Nov 04 08:30:25 <Leeds> infernus: I'm essentially always in this channel when my PC is on
Nov 04 08:30:41 <infernus> If I might get back here after sometime with my intent filter
Nov 04 08:30:46 <Leeds> (this isn't AOL, they're channels not rooms)
Nov 04 08:30:58 <speakingcode> lol
Nov 04 08:31:17 <infernus> Thanks dude... will ping you then..
Nov 04 08:31:58 <infernus> Yeaa.. channels... I'm new to irc... will get even with terms soon :)
Nov 04 08:43:12 <regedit> back to my speech recognizer; i'd like to have a omni-running service (until stopped) that listens to voice and (based on computed speech recognition results) it can 1) open/switch-to different apps 2) manipulate / cause things to happen inside those apps
Nov 04 08:43:31 <regedit> is this possible?
Nov 04 08:43:47 <g00s> doubt it
Nov 04 08:44:08 <tdignan> coding is hard, just give up and watch tv
Nov 04 08:44:36 <ESphynx> you americans better vote for Obama.
Nov 04 08:44:58 <regedit> i should add that the apps in question are initially programmed/rigged to work with this voice service - were it feasibly programmable
Nov 04 08:45:29 <regedit> i.e. i create the omni service, and i create the apps that work with it
Nov 04 08:45:42 <Leeds> NO POLITICS
Nov 04 08:45:52 <tdignan> http://www.youtube.com/watch?v=I9zpnLBtwwg
Nov 04 08:46:04 <tdignan> not a political youtube, an awesome classic pop hit.
Nov 04 08:46:17 <ESphynx> why is this damn thing killing an object i'm currently working with?
Nov 04 08:46:32 <g00s> Leeds: we're a handful tonight :P
Nov 04 08:46:43 <regedit> g00s: ^ ?
Nov 04 08:47:03 <Leeds> NO POLITICS... except http://www.youtube.com/watch?v=fWEm4p3Dopw
Nov 04 08:47:21 <tdignan> i'll vote for whoever can help me expatriate... any takers?
Nov 04 08:47:23 <g00s> fuck man, needs flash
Nov 04 08:47:41 <regedit> google "fWEm4p3Dopw"
Nov 04 08:47:56 <tdignan> Leeds: "the uploader has not made this available in your country"
Nov 04 08:48:03 <tdignan> no back to this earlier point about expatriating
Nov 04 08:48:08 <tdignan> now*
Nov 04 08:48:17 <Leeds> wow, sorry... anyway, enough
Nov 04 08:48:32 <tdignan> good tip regedit
Nov 04 08:48:44 <tdignan> wait no, still not available
Nov 04 08:49:00 <regedit> so can i haz service that continuously listens to voice and makes stuff happen in apps
Nov 04 08:49:29 <regedit> i know i can haz app that does that, i already made it
Nov 04 08:50:07 <tdignan> cat picture project
Nov 04 08:50:09 <tdignan> g00s:
Nov 04 08:50:43 <tdignan> make speech-to-lolcat
Nov 04 08:50:47 <tdignan> I bet it would be a hit
Nov 04 08:51:04 <regedit> hah
Nov 04 08:51:17 <regedit> le'me break this down;
Nov 04 08:51:28 <regedit> can a long-running service start an app
Nov 04 08:51:58 <Leeds> yes, of course a service can start an app
Nov 04 08:52:09 <regedit> good good, we're making headway
Nov 04 08:52:14 <regedit> can a long-running service switch from one app to a different app
Nov 04 08:52:36 <Leeds> only in the sense that... of course a service can start an app
Nov 04 08:53:07 <kaerf> has anyone used app inventor?
Nov 04 08:53:28 <regedit> so if a user was playing Where's My Water and suddenly an obnoxious service decides "HEY ITS TIEMZ TO USE THIS OTHER APP" it can just switch away like that?
Nov 04 08:55:29 <Leeds> in principle yes, AFAIK
Nov 04 08:55:40 <regedit> ok good
Nov 04 08:57:16 <regedit> next; can a service manipulate an app and cause things to happen, for example; service runs SomeApp and makes X Action/Intent/wtvr happen, and then proceeds to make Y Action/Intent/wtvr happen
Nov 04 08:57:17 <lasserix> If i am doing a shape drawable with a linear gradient with start/center/stop colors, is there a way to give these relative size weights like you can do with a linear layout?
Nov 04 09:02:22 <regedit> Leeds: ^ plausible?
Nov 04 09:02:49 <Leeds> regedit: not particularly sensibly, unless the app is co-operating
Nov 04 09:03:42 <regedit> Leeds: yes let's say the app is rigged to cooperate. the idea here is hands-free automation, no user intervention needed
Nov 04 09:04:16 <regedit> _can_ such a cooperation be rigged?
Nov 04 09:04:34 <Leeds> an app and a service can co-operate with each other, yes
Nov 04 09:05:01 <Leeds> service sends a message to app saying "do X", app does X and sends a message back saying "finished doing X"
Nov 04 09:05:05 <marcnesium> Hello everyone! i have here a little tricky problem with my nexus s. i encrypted my phone with build-in functions (stock rom), but now my "usb storage" can't be mounted any more. seems corrupted. my first idea was to get an image using dd and try to mount that with my laptop. anyone here with some expiriences regarding this?
Nov 04 09:05:47 <Leeds> marcnesium: hi! this channel is for app development, not user support, rooting, etc.
Nov 04 09:06:40 <regedit> Leeds: i that covers everything then. to reiterate original question;
Nov 04 09:06:45 <regedit> Leeds: "i'd like to have a omni-running service (until stopped) that listens to voice and (based on computed speech recognition results) it can 1) open/switch-to different apps 2) manipulate / cause things to happen inside those apps. i should add that the apps in question are initially programmed/rigged to work with this voice service - were it feasibly programmable. i.e. i create the omni service, and i create the apps that wor
Nov 04 09:06:57 <regedit> Leeds: all in all - plausibly feasible?
Nov 04 09:07:04 <Leeds> I'd say the listening to speech from the background is the bit which would worry m
Nov 04 09:07:34 <marcnesium> oh, ok, sorry. can you recommend an other channel for this?
Nov 04 09:07:57 <Leeds> marcnesium: #android or #android-root, depending
Nov 04 09:08:04 <regedit> Leeds: and aside from that?
Nov 04 09:08:40 <marcnesium> Leeds: Thanks!
Nov 04 09:08:41 <regedit> Leeds: think red dot in notification area that denotes "listening in progress"
Nov 04 09:09:02 <regedit> perhaps flashing red dot even
Nov 04 09:09:26 <Leeds> no, I'm thinking who the hell would want a service sitting with the mic permanently open listening to everything I say and do... and would Android allow it
Nov 04 09:09:39 <Leeds> the voice bit is the hard bit, trust me
Nov 04 09:10:07 <lasserix> and i imagine would drain the battery relatively fast if you plan on letting it run constantly
Nov 04 09:10:12 <Leeds> that too
Nov 04 09:10:17 <regedit> Leeds: 1) turn it off if you dont want it 2) this is mainly targeted at stationary usage, such as in a car, at desk, or on kitchen fridge
Nov 04 09:10:36 <regedit> as such #2, it is assumed to be plugged in most of the time
Nov 04 09:10:44 <Leeds> honestly, I don't know if you can do that sort of voice recognition from a service, and I'd sort of hope not
Nov 04 09:11:11 <regedit> sounds like its possible from here http://stackoverflow.com/questions/4559930/speechrecognizer-causes-anr-i-need-help-with-android-speech-api
Nov 04 09:11:23 <lasserix> regedit are you expecting commands to be issued, or are you looking for a continous flow of speech?
Nov 04 09:11:33 <regedit> commands to be issued
Nov 04 09:11:40 <regedit> totally hands free
Nov 04 09:12:32 <regedit> ofcourse id never want to betray user trust so i make it very clear that audio capture is happening, like i said flashing red dot in notification area, or some-such
Nov 04 09:13:18 <Leeds> that seems to be the record of someone not getting recognising to work in a Service :)
Nov 04 09:13:35 <regedit> Leeds: and an answer of how to do it right
Nov 04 09:13:41 <Leeds> does it work?
Nov 04 09:13:50 <regedit> does it not
Nov 04 09:14:04 <Leeds> um, no - you're the one trying to do it, not me
Nov 04 09:14:12 <Leeds> so does that answer work?
Nov 04 09:14:28 <regedit> im going to assume it can then, i was just asking about the other bits of functionality
Nov 04 09:14:52 <regedit> i.e. if i can make this voice-processing service to perform all sorts of interesting actions
Nov 04 09:15:30 <regedit> open apps, perform things inside apps, etc
Nov 04 09:16:00 <Leeds> I suggest you spend some time to see if your basic USP - background speech recognition - works
Nov 04 09:16:47 <regedit> Leeds: ya that's my first step at this point, just wanted to check if its even worth bothering - meaning whether i can indeed achieve the functionality effects im trying
Nov 04 09:17:08 <Leeds> yes, Services and Activities can communicate with each other... now go play with speech stuff
Nov 04 09:17:37 <regedit> yes dad *runs off like a good little puppy*
Nov 04 09:18:40 <Leeds> if you'd just asked the straightforward question in the first place - simple question, simple answer, nothing to do with voice stuff
Nov 04 09:20:14 <regedit> guess so, 'ts not always so clear in my own head
Nov 04 09:20:22 <regedit> im only a <5 days old android developer
Nov 04 09:20:59 <adq> don't worry, be hacky
Nov 04 09:23:04 <lasserix> Drawing ninepatch, left and top control padding for content (content will not go where black pixels are) and bottom and right control stretching (stretching occurs where black pixels are) or do I have this backward?
Nov 04 09:23:10 <regedit> adq: oh thanx, that's nice
Nov 04 09:23:29 <regedit> adq: also nice tagline/mantra
Nov 04 09:27:55 <Ge0rG> anybody has got reliably working code to record record audio from a bluetooth SCO headset?
Nov 04 09:42:31 <cagedwisdom> what are the reasons my phone would reboot frequently for no good reason?
Nov 04 09:42:57 <cagedwisdom> w/o otas etc
Nov 04 09:43:55 <Leeds> cagedwisdom: while you're working on an app?
Nov 04 09:45:11 <regedit> Leeds: would you care to explain about that stackoverflow http://stackoverflow.com/questions/4559930/speechrecognizer-causes-anr-i-need-help-with-android-speech-api how come doing it in the service is (apparently) not in the main thread?
Nov 04 09:45:12 <cagedwisdom> Leeds, no. I am testing my engineers signe builds, but not coding myself. and haven't updated that app in a week or so but had probably at least 2-3 reboots in last ~48 hrs.
Nov 04 09:46:02 <Leeds> regedit: not really, no
Nov 04 09:46:26 <Leeds> cagedwisdom: then you shouldn't see regular reboots (on a stock OS, I assume)
Nov 04 09:47:08 <cagedwisdom> hoping for a galaxy nexus update to jb after xmas. my wife's had it on her nexus s for a few months.
Nov 04 09:47:56 <Leeds> CDMA, presumably
Nov 04 10:43:56 <tdignan> cagedwisdom: using bluetooth?
Nov 04 10:59:27 <tdignan> every now and then eclipse likes to surprise me by segfaulting the jvm
Nov 04 10:59:31 <tdignan> :D
Nov 04 10:59:56 <shmooz> total eclipse of the heart !
Nov 04 11:00:26 <Zider> twilight: eclipse
Nov 04 11:01:47 <Mavrik> tdignan: using OpenJDK by any chance? :P
Nov 04 11:04:03 <fenicks> hi
Nov 04 11:06:08 <tdignan> Mavrik: nope
Nov 04 11:06:32 <tdignan> ~$ which java
Nov 04 11:06:33 <tdignan> /home/tom/opt/jdk1.6.0_37/bin/java
Nov 04 11:07:00 <tdignan> jdk 1.7 is significantly worse on crashes, i have that installed too
Nov 04 11:07:24 <tdignan> on the latest eclispe juno. 32 bit linux.
Nov 04 11:19:41 <Freezingwind> Hello all
Nov 04 11:26:46 <Freezingwind> anyone available?
Nov 04 11:27:43 <Leeds> if you've got a question, ask it
Nov 04 11:29:17 <Freezingwind> :D
Nov 04 11:29:32 <Freezingwind> will ODIN really factory restore your phone?
Nov 04 11:30:14 <Leeds> ah, the wrong question... read channel topic, go to #android-root, ask question there
Nov 04 11:30:39 <Freezingwind> but im not trying to root my phone :/
Nov 04 11:31:15 <Freezingwind> after installing aokp and etc of other roms, they've all left unused folders and files, just wanna wipe clean so i can fresh install aokp
Nov 04 11:31:16 <Freezingwind> o-o
Nov 04 11:31:56 <Freezingwind> If android-root is the designated channel for that, then im sorry for asking it here.
Nov 04 11:32:09 <shmooz> isn't there a channel for android-roms ?
Nov 04 11:33:20 <Freezingwind> is this really a rom question?
Nov 04 11:33:38 <Freezingwind> I just wanna erase all data on the phone.
Nov 04 11:33:40 <adq> for sure this is not a "dev" question
Nov 04 11:35:04 <Freezingwind> ...
Nov 04 11:35:09 <Freezingwind> My mistakes.
Nov 04 11:35:19 <Freezingwind> I thought i was in the aokp channel.
Nov 04 11:35:22 <shmooz> yes I guess just ask it in #android
Nov 04 11:52:51 <cagedwisdom> tdignan, no to bt
Nov 04 12:03:45 <tdignan> there is a bug in 2.2 that causes a device reboot after you connect to bluetooth 512 times
Nov 04 12:14:22 <Ge0rG> tdignan: how did you find out? :D
Nov 04 12:14:28 <Freezingwind> well they didn't help me out
Nov 04 12:16:14 * Ge0rG still needs to figure out how to record audio from a SCO headset... the usual howtos fail with strange error messages :(
Nov 04 12:20:48 <shankey_androdev> hii
Nov 04 12:24:37 <shankey_androdev> does android support mullti touch
Nov 04 12:24:53 <shankey_androdev> bcos initially i was told it doesnt
Nov 04 12:24:59 <ESphynx> it does
Nov 04 12:25:01 <shankey_androdev> ny1
Nov 04 12:25:05 <shankey_androdev> ok
Nov 04 12:25:20 <shankey_androdev> how many touch
Nov 04 12:26:09 <SimonVT> As many as your touch screen supports
Nov 04 12:28:32 <shankey_androdev> ok then it is hardware dependent.?
Nov 04 12:30:41 <shankey_androdev> m creating a game which is multi touch but then how would i know on which device it is compatible or how many touch a device can have
Nov 04 12:35:44 <SimonVT> Yes, not all devices support multitouch
Nov 04 12:37:33 <SpeedEvil> not all devices support touch
Nov 04 12:39:53 <Hodapp> booo, the SoundCloud app is "not compatible" with my MK802 but all I wanna do is stream music :| wonder if it's bitching about the lack of a microphone
Nov 04 12:42:42 <Freezingwind> does skyrocket support multitouch?
Nov 04 12:43:16 <Freezingwind> actually nevermind, I already know the answer.
Nov 04 12:43:43 <Freezingwind> I remember me holding an icon with one finger and swiping through homes on the other finger
Nov 04 12:45:45 <adq> Freezingwind, for recording audio from BT, maybe you forgot permissions in manifest?
Nov 04 12:46:01 <adq> otherwise take a look on stackoverflow, I'm sure you will find relevant answers
Nov 04 12:52:20 <tdignan> Ge0rG: someone asked me to fix it...
Nov 04 12:52:41 <tdignan> i killed the process and used the alarm manager to restart it :)
Nov 04 13:25:20 <elf-> hey
Nov 04 13:39:01 <sciepy> hey, anyone has experience with cross compiling c programs for android?
Nov 04 14:01:29 <napster> I use setFillAfter(true) after the animation. But the click areas stays unmoved. How can I move it as well?
Nov 04 14:26:01 <SimonVT> Same way you put it in its initial position.. layoutparams
Nov 04 14:28:52 <Harrold> hi
Nov 04 14:49:15 <dioHere> hey guys. Am building a game with andengine and I want to show a dialog. So the dialog must be created inside runOnUiThread. Through the dialog I am showing an xml view that has buttons. I want to set click listeners to the buttons, and I am doing that inside the runOnUiThread because this is where dialog created. But they never fire. Can you help me please?
Nov 04 15:00:32 <regedit> is it feasible to develop nothing but a long-running service with no actual app associated with it, just an on/off switch to start/stop it and some settings to configure it
Nov 04 15:00:38 <regedit> i'm asking because i'd need to call startService() since no app is gonna initiate it with no bind functions, but if i dont have an app/action how can i call startService
Nov 04 15:01:23 <tdignan> you can start it off a broadcast, like boot completed.
Nov 04 15:01:30 <tdignan> you don't need to have any activities
Nov 04 15:01:42 <tdignan> but, I would write one, because it's convenient.
Nov 04 15:02:14 <regedit> tdignan: feels like just clogging user's app screens. and what happens when you click on it - nothing?
Nov 04 15:03:34 <tdignan> remove the launcher intent-filter if you don't want an icon.
Nov 04 15:05:45 <regedit> tdignan: so if it's not triggered by say boot completed, rather it runs on its own little convenience action, how/when does action get executed
Nov 04 15:06:35 <tdignan> what action?
Nov 04 15:06:48 <regedit> i meant activity
Nov 04 15:07:00 <tdignan> when you click the icon?
Nov 04 15:07:01 <regedit> [10:01] <tdignan> you don't need to have any activities [10:01] <tdignan> but, I would write one, because it's convenient.
Nov 04 15:07:19 <tdignan> does that comment really require elaboration?
Nov 04 15:08:37 <regedit> tdignan: first you recommended to have a convenience activity, then you advised me about removing the launch icon if i wish, so i was just wondering what use would the convenience activity be of
Nov 04 15:09:27 <tdignan> Those two statements were not related.
Nov 04 15:10:20 <regedit> so it's either "on boot completed" triggered with no activity whatsoever, or the app has a launcher icon - that's what you meant
Nov 04 15:11:59 <regedit> tdignan: and when the user launches your convenience activity, what happens (visually) - nothing?
Nov 04 15:14:15 <tdignan> if you have a background app, the ideal experience is probably: starts on boot, shows an ongoing notification. touch the ongoing notification to open an activity with some information.
Nov 04 15:14:37 <regedit> i see
Nov 04 15:19:50 <dioHere> hey guys. Am building a game with andengine and I want to show a dialog. So the dialog must be created inside runOnUiThread. Through the dialog I am showing an xml view that has buttons. I want to set click listeners to the buttons, and I am doing that inside the runOnUiThread because this is where dialog created. But they never fire. Can you help me please?
Nov 04 15:29:14 <regedit> tdignan: is an idle service (not currently executing any codepaths or responding to any events/broadcasts etc) as good as a nonexistent service, performance-wise?
Nov 04 15:31:17 <regedit> oh, he's gone
Nov 04 15:31:25 <vandenoever> is it possible to get a blob from an android database via a uri in a webview?
Nov 04 15:36:15 <regedit> is it possible to make a service sometimes execute some work and sometimes remain idly standing by, perhaps controlled by a notification-area on/off switch
Nov 04 15:37:16 <vydd> hi. could someone here help me with deciding on whether I should use java+opengl or ndk for a particular problem? of course, it would be best if I could write it both ways and do benchmarks, but since I'd be learning in the process, that's not an option at the moment
Nov 04 15:37:37 <vandenoever> vydd: should it run on multiple platforms?
Nov 04 15:38:18 <vydd> vandenoever: that doesn't matter much; platforms supported by ndk are enough
Nov 04 15:38:43 <vandenoever> vydd: what's the particular problem?
Nov 04 15:39:19 <vydd> vandenoever: image deformation using moving least squares
Nov 04 15:39:36 <vandenoever> vydd: so most code will be a shader?
Nov 04 15:43:02 <vydd> vandenoever: well, I guess. I haven't really done much opengl before (only some examples)...basically, there's a function that's done for all pixels which takes matrices and vectors as its arguments, in part calculated from other pixels
Nov 04 15:43:48 <vandenoever> vydd: it sounds like doing that with a shader is feasible and then ndk vs java is not import
Nov 04 15:44:07 <vandenoever> so starting from a simple java with shader example is easiest
Nov 04 15:45:00 <vydd> vandenoever: any good, advanced books that you could recommend?
Nov 04 15:47:05 <vandenoever> vydd: OpenGL Shading Language
Nov 04 15:47:10 <vandenoever> rost et al
Nov 04 15:48:35 <vydd> vandenoever: thanks!
Nov 04 15:48:44 <vandenoever> vydd: have fun! :-)
Nov 04 15:48:54 <vydd> :)
Nov 04 16:03:12 <regedit> how do i make my service with a quick-settings on/off switch in the expanded notification bar
Nov 04 16:03:28 <regedit> much like the WiFi on/off switch
Nov 04 16:07:23 <kowalej> regedit: so this app is meant to run in the background? and you want to be able to pull down the notification bar to toggle it?
Nov 04 16:07:49 <regedit> kowalej: precisely
Nov 04 16:08:01 <Strangebrew> is it even possible?
Nov 04 16:08:18 <regedit> i was also meaning to ask about that
Nov 04 16:08:49 <kowalej> regedit:http://developer.android.com/guide/topics/ui/notifiers/notifications.html#Actions
Nov 04 16:09:13 <kowalej> you can at least start and stop activities
Nov 04 16:09:16 <regedit> when the switch is ON, my service does XYZ. when OFF, uh i dunno service hangs around idly waiting to be turned on?
Nov 04 16:09:33 <speakingcode> when switch is off, stop the service ;-P
Nov 04 16:09:34 <regedit> is that possible?
Nov 04 16:09:47 <regedit> no but then the on/off switch isnt there anymore
Nov 04 16:09:58 <speakingcode> o, right
Nov 04 16:10:16 <speakingcode> make it a widget instead of a notification bar
Nov 04 16:10:54 <regedit> i definitely do want to show a notification icon though, my service must make it well aware that it is operational - when it is
Nov 04 16:11:14 <regedit> *make the user well aware
Nov 04 16:11:35 <speakingcode> could do a combination
Nov 04 16:12:18 <kowalej> why can't the service just run on its own activity and the app has a seperate activity for idle state
Nov 04 16:12:22 <regedit> is there no such concept as a service idling around with a switch button to activate it, or something close enough to that?
Nov 04 16:12:45 <speakingcode> sure
Nov 04 16:12:49 <kowalej> when you toggle on it launches that activity, when you toggle off it launches the idle activity which closes the other one
Nov 04 16:12:53 <speakingcode> but i dunno about through the notification bar
Nov 04 16:13:16 <speakingcode> just keep a service running and use bound service to rebind to it
Nov 04 16:13:51 <speakingcode> thjink about it from a UX perspective tho - how do you start it to begin with? you can't start it from the notification bar if it isn't started
Nov 04 16:14:54 <speakingcode> thats why a toggle widget seems to make sense to me, like a gps/blutooth/wifi toggle widget; fairly common
Nov 04 16:16:08 <Strangebrew> is there a way to start android gallery to display an image (using Intent.ACTION_VIEW) but disable the delete functionality?
Nov 04 16:17:12 <Strangebrew> oh, and where can i find the options that can be used with each intent action?
Nov 04 16:17:20 <regedit> i start it on boot complete
Nov 04 16:18:26 <regedit> speakingcode: these activities you speak of (actual execution and idle), they will execute in the background?
Nov 04 16:18:51 <Strangebrew> regedit: it seems that it's possible  http://forum.xda-developers.com/showthread.php?t=1037817..  i thought it's only possible with system modifications (custom roms)
Nov 04 16:19:04 <speakingcode> trying to do anything in background with an activity is fail
Nov 04 16:19:36 <speakingcode> system will say "oh this activity in the background is diong work but it's not in the foreground, kill it"
Nov 04 16:20:14 <speakingcode> so you must use service. and  i believe someone else suggested using two activities cause i didn't
Nov 04 16:20:28 <regedit> right i was getting confused about that. this is a service im talking about here, not an activity
Nov 04 16:20:39 <speakingcode> so you could use one service
Nov 04 16:20:48 <speakingcode> yuo start it on boot with a broadcast receiver, sure
Nov 04 16:21:10 <speakingcode> it posts a notification. it starts work, a button in the notifcation bar allows it to stop work, btu the service itself is still running, just not doing the workload
Nov 04 16:21:45 <regedit> how is that part coded
Nov 04 16:22:00 <regedit> the part where its still running just not doing anything
Nov 04 16:22:13 <speakingcode> you could do it a few ways
Nov 04 16:22:17 <regedit> and is that a performance hog in the slightest
Nov 04 16:22:24 <speakingcode> no
Nov 04 16:22:41 <speakingcode> a service in the background diong nothing is less drain than one in the background doing something
Nov 04 16:22:54 <regedit> is it (almost) as good as a nonexistent service
Nov 04 16:23:16 <speakingcode> thats a sort of loaded question
Nov 04 16:23:35 <regedit> performance / system resource-wise
Nov 04 16:23:42 <regedit> cpu-wise
Nov 04 16:24:04 <speakingcode> what does almost mean? i mean from a human pointo f view i doubt you would notice any difference
Nov 04 16:24:19 <Leeds> regedit: can you do speech recognition from your Service yet?
Nov 04 16:24:44 <regedit> Leeds: not yet, working on it
Nov 04 16:24:55 <speakingcode> in your service,use a sep thread to do work.. if the user hits the notificaton bar, the service can get the thread to stop (a few ways to do that really)
Nov 04 16:25:04 <Leeds> then you're wasting you time with this on/off stuff
Nov 04 16:25:23 <regedit> Leeds: i found it interesting to learn about in general
Nov 04 16:25:52 <regedit> speakingcode: the work my service is doing must run in main thread....
Nov 04 16:26:03 <Leeds> okay, as long as you know it's nothing to do with the app you're trying to write
Nov 04 16:26:04 <regedit> SpeechRecognizer is like that
Nov 04 16:26:37 <regedit> see overview http://developer.android.com/reference/android/speech/SpeechRecognizer.html
Nov 04 16:26:56 <speakingcode> o
Nov 04 16:28:57 <regedit> can a service not be in a state of "not executing anything ATM, just waiting around for an event to be called on me" or some-such?
Nov 04 16:41:46 <regedit> what happens with a Service that was started by say startService() and then its onStartCommand() method returns, and stopService() or stopSelf() is not called (yet)
Nov 04 16:42:14 <regedit> will it effectively be in a state of "idling" until something causes it to enter some execution path again?
Nov 04 16:42:56 <SimonVT> Same thing that happens to an activity that is not finished
Nov 04 16:43:57 <kowalej> regedit: is your service bound?
Nov 04 16:45:01 <regedit> kowalej: i'm thinking of starting it on boot complete - therefore not bound - and then the service gets bind requests here 'n there throughout its existence to execute stuff
Nov 04 16:45:15 <regedit> if that makes any sense
Nov 04 16:46:03 <kowalej> well, as long as it's bound it shouldn't get killed
Nov 04 16:46:30 <regedit> sounds like if its started - even if not bound - it shouldnt get killed either
Nov 04 16:46:31 <regedit> http://developer.android.com/reference/android/app/Service.html#ServiceLifecycle
Nov 04 16:46:49 <regedit> see last paragraph of that section
Nov 04 16:47:19 <SimonVT> It'll get killed if the memory is needed elsewhere
Nov 04 16:47:45 <regedit> SimonVT: would that not happen if it were bound?
Nov 04 16:48:22 <SimonVT> It might happen in either case
Nov 04 16:48:37 <regedit> ok so nothing to lose there
Nov 04 16:49:16 <SimonVT> It seems wasteful to start your service at boot if it's not gonna get used
Nov 04 16:49:47 <regedit> in conclusion it seems like it's feasible to have a background service that is "activated/deactivated" by a notification-bar on/off switch (like the wifi switch)
Nov 04 16:50:57 <regedit> when in on state (ever since boot), a code path will begin to be executed. when in off state, all functions return and service kinda does nothing but wait for on switch
Nov 04 16:54:01 <speakingcode> what happens when it is started is entirely up to the programmer
Nov 04 16:54:34 <regedit> speakingcode: what are you pointing out
Nov 04 16:54:56 <speakingcode> <regedit> what happens with a Service that was started by say startService() and then its onStartCommand() method returns, and stopService() or stopSelf() is not called (yet)
Nov 04 16:55:58 <regedit> speakingcode: so if all functions return and the service was not shutdown yet (with stopService/Self()) it is effectively in a kind of "idle" state?
Nov 04 16:56:04 <speakingcode> that lapse betwen onStartCommand and stopService is what you would call the lifetime of the service, and what happens in its lifetimei s entirely up to the programmer. a service is always "running" but what it is actually diong is up to you
Nov 04 16:56:33 <regedit> or if its running anything at all - is what im asking
Nov 04 16:56:40 <speakingcode> err always "running" until it is stopped**
Nov 04 16:56:46 <speakingcode> sure
Nov 04 16:56:57 <speakingcode> you can start a service that does absolutely nothing, if you want
Nov 04 16:57:05 <regedit> cool
Nov 04 16:57:16 <regedit> thats all i was trying to find out
Nov 04 16:57:45 <speakingcode> there's somewhat of an asynchronous model going on tho you might need to think about
Nov 04 16:57:56 <regedit> so now all i need to figure out is how to make the service's notification on/off switch effectively execute stuff or return all methods appropriately
Nov 04 16:58:30 <speakingcode> there's no way to force a method to return really, think about it from an async perspective
Nov 04 16:59:16 <regedit> well how does a service's notification controls (like this on/off switch) report back to the service
Nov 04 16:59:50 <speakingcode> you're gonnah ave to read the docs,  i can't succicntly explain it any better
Nov 04 16:59:51 <regedit> it's probably a settings thing, no?
Nov 04 17:00:20 <speakingcode> you probably need to go with a bound service (you do, heh)
Nov 04 17:00:50 <regedit> i understand the async ideas going on here, the service might be running along its methods when the switch is turned off, how do the methods know to return
Nov 04 17:01:28 <regedit> i'm imagining that when the switch changes state, some event is posted to the service?
Nov 04 17:01:55 <speakingcode> you can give the notification a pending intent, so that when it's clicked the intent can effectively 'notify' your service
Nov 04 17:02:25 <regedit> ah excellent
Nov 04 17:02:41 <shmooz> is there anyone pro enough here to compile ffmpeg4android (from sourceforge) using Windows 7 instead of ubuntu ?
Nov 04 17:02:52 <regedit> and then my service just sets a variable or whatnot which the running methods check periodically, and return themselves if necessary
Nov 04 17:03:09 <speakingcode> well your service is basically an infinite loop right constnalty doing speech recongition. so put a conditon on that loop
Nov 04 17:03:12 <speakingcode> yup exactly
Nov 04 17:03:21 <shmooz> see someone has done it with older version of ubuntu, but if its possible with Win7 I'd rather use Win7
Nov 04 17:03:54 <speakingcode> don't think anyone is pro enough here fort that
Nov 04 17:04:35 <regedit> speakingcode: then when the switch goes back on, is it easy enough for the pending intent to cause the methods to start running again?
Nov 04 17:04:54 <kowalej> why are you trying to have it idle then, why can't you stop and restart?
Nov 04 17:05:14 <regedit> kowalej: because i need the on/off switch in the notification bar
Nov 04 17:05:38 <speakingcode> don't understand why your so set on it benig in the notif. bar, but shrug
Nov 04 17:06:04 <kbs> sorry to flog a possibly comatose horse :) but the android design style says 48dp is (on average) 9mm. I'm not seeing how this math works? On a 160ppi device for example, 48dp == 48pixels, which works out to 48/160 inch ~ 7.6mm
Nov 04 17:06:08 <regedit> speakingcode: feels more natural, like thats where it really should be
Nov 04 17:07:24 <speakingcode> kbs because it's avg?
Nov 04 17:07:42 <shmooz> argh I guess I have to install loonix
Nov 04 17:07:45 <speakingcode> perhaps 160ppi isn't the avg density
Nov 04 17:07:51 <shmooz> just to get ffmpeg working
Nov 04 17:08:15 <regedit> speakingcode: why should it be in some homescreen widget? feels out of place there... my service is kinda like the wifi service, available to be switched on/off system-wide. not a piece of homescreen widget imho
Nov 04 17:08:17 <shmooz> I wonder how fast eclipse on ubuntu would run with a 2.4 Ghz Celeron
Nov 04 17:08:34 <kbs> speakingcode: I thought so, but that doesn't jive either, because the docs claim the range is 7mm-10mm. on a 192 ppi device, it gets quantized into a 160ppi bucket -- which works out to .635 mm
Nov 04 17:09:16 <speakingcode> shrug
Nov 04 17:09:40 * regedit wonders who that shrug was for
Nov 04 17:10:04 <speakingcode> more at the ppi/dp question but i guess at yours too :-) go with what you feel is best
Nov 04 17:10:12 <regedit> :D
Nov 04 17:10:19 <regedit> k thanks for help
Nov 04 17:10:20 <kbs> :-) the problem here is that this falls under the recommended minimum touchable area size (7mm), meaning that using 48dp isn't actually correct
Nov 04 17:10:57 <speakingcode> tbh i haven't done much with notifications and services, so i can't really tell you whats best here
Nov 04 17:11:18 <speakingcode> usually when i have a service that posts an otification, that notification leads back to an activity that binds to the service
Nov 04 17:11:55 <speakingcode> and the user can take action from that ativity. jellybean (maybe ICS too?) introduced some more features with notifications but i haven't played with those, so i dunno
Nov 04 17:12:14 <speakingcode> using 48dp for what?
Nov 04 17:12:54 <kbs> the recommended metric on android: http://developer.android.com/design/style/metrics-grids.html
Nov 04 17:13:16 <kbs> the exact quote: "If you design your elements to be at least 48dp high and wide you can guarantee that your targets will never be smaller than the minimum recommended target size of 7mm regardless of what screen they are displayed on."
Nov 04 17:13:58 <gutano> seems like a silly thing to get hung up on, I'm sure they picked 48 because it is a very standard computer number (32 + 16).. 49, 50 would have been odd
Nov 04 17:13:59 <speakingcode> you say 48 dp is > 7 mm. so in terms fo area thats over 79 square units
Nov 04 17:14:21 <kbs> no, I said 48dp on a 192ppi device is 6.35mm
Nov 04 17:14:39 <speakingcode> mmm
Nov 04 17:14:41 <kbs> I have a color nook, and the standard buttons are too small -- this is why it's relevant :-)
Nov 04 17:14:58 <Vlad__> Hi. I am trying to get intent filters to open my own app in gmail with a custom extension, but nothing I have found seems to work right. They all do at least one of: only work on 4.1, prompt to open other files too, prompt to open my app when tapping an email in the notification bar. Does anyone have any experience with this?
Nov 04 17:15:23 <Vlad__> I have also asked on stackoverflow if anyone is interested: http://stackoverflow.com/questions/13156212/intent-filter-for-custom-file-type but no answers yet.
Nov 04 17:16:31 <speakingcode> how are you getting 6.35? i'm not following your math
Nov 04 17:18:00 <kbs> a 192ppi device gets treated as a 160ppi device (that's its closest bucket)
Nov 04 17:18:14 <kbs> so a 48dp line is rendered as 48 pixels -- so 48/192 inches long
Nov 04 17:19:19 <speakingcode> wouldn't it be 48/160 if its treated as such?
Nov 04 17:19:52 <kbs> no -- it always draws 48 pixels if you specify 48dp (if the 'normalized' ppi is 160)
Nov 04 17:20:23 <kbs> the problem is basically that the system thinks there are fewer pixels per inch than there really are. This means everything si smaller by about 20%, roughly
Nov 04 17:20:35 <speakingcode> where is this stated?
Nov 04 17:22:28 <SimonVT> That's just how android works
Nov 04 17:22:40 <speakingcode> oh ok i see the math now. 48 * 1in/192 = 6.35mm
Nov 04 17:22:54 <kbs> http://stackoverflow.com/questions/6656651/difference-between-android-dimension-pt-and-dp/6656800#6656800 has some more concrete examples
Nov 04 17:24:21 <speakingcode> shoulda got a kindle fire i guess
Nov 04 17:43:08 <voso> I think 9 out of 10 people who for some reason think they want to use View.onMeasure() do it, because they want to set up dimensions based on total available pixels. Yet it is impossible to get hold of the available total real estate (width x height) a that time of the app reliably.
Nov 04 17:43:17 <voso> Do you all agree, or am I missing something?
Nov 04 17:44:02 <shmooz> you can get screen dimensions no ?
Nov 04 17:44:02 <kbs> voso: as I'm clueless enough to do just that :-) could you tell me what the gotcha is?
Nov 04 17:44:57 <alankila> I have worked with the onMeasure successfully to make a custom widget that has height which is 1/12th of its width
Nov 04 17:45:06 <voso> there is no way as far as I know to get your apps W x H pixels (taking into account the statusbar and systembar across all devices - if/when they exist)
Nov 04 17:45:23 <kbs> oh, you do mean the total device screen pixels -- misunderstood.
Nov 04 17:45:42 <shmooz> using DisplayMetrics
Nov 04 17:46:28 <shmooz> getWindowManager().getDefaultDisplay().getMetrics(display);
Nov 04 17:46:55 <shmooz> screenHeight = display.heightPixels;
Nov 04 17:47:01 <voso> I've been all over SO and the docs.. DisplayMetrics does not account properly for sys/staatus bars
Nov 04 17:47:09 <shmooz> screenWidth = display.widthPixels;
Nov 04 17:47:32 <shmooz> it does for me, but I do hide the title bar
Nov 04 17:47:45 <voso> ..very funny. ;)
Nov 04 17:47:51 <shmooz> you can hide the status and title bars at first and get metrics and then show them again
Nov 04 17:47:53 <voso> ..and you are not thinking about tablet compat. ;)
Nov 04 17:48:20 <shmooz> what would be the problem ?
Nov 04 17:48:25 <voso> systembar?
Nov 04 17:48:31 <voso> or whatever its called..
Nov 04 17:48:40 <voso> you cant hide that
Nov 04 17:48:45 <shmooz> set FULLSCREEN and get metrics and then unset FULLSCREEN
Nov 04 17:49:04 <Vlad__> i'm using a viewtreeobserver with vto.addOnGlobalLayoutListener and then getHeight on the view
Nov 04 17:49:07 <voso> I understand your workaround.
Nov 04 17:49:15 <Vlad__> full screen is not full on tablets
Nov 04 17:49:25 <voso> I was looking for a 'clean' solution
Nov 04 17:49:28 <Vlad__> the status bar that shows the clock and stuff will always be there
Nov 04 17:49:38 <voso> found none, came here for confirmation! ;)
Nov 04 17:50:00 <shmooz> well it seems to work for me but I only test on phone
Nov 04 17:50:02 <Vlad__> it's rather clean, only a few lines, and you can save it in shared preferences
Nov 04 17:50:21 <Vlad__> i save it in shared prefs in my first activity then get it from there
Nov 04 17:50:28 <kbs> that's a nice idea, thanks
Nov 04 17:51:22 <Vlad__> on > 3.1 it works with the usual method IIRC
Nov 04 17:51:28 <Vlad__> even on tablets
Nov 04 17:52:02 <voso> Vlad__ that sounds interesting. I'll look into it. Thanks.
Nov 04 17:52:04 <Chainfire> any of you guys using ACRA with a non-Google-Docs backend, and if so, which one ?
Nov 04 17:52:35 <Vlad__> let me know if you need code
Nov 04 17:52:51 <voso> ok, thanks.
Nov 04 17:52:59 <Vlad__> np
Nov 04 17:59:49 <kbs> Vlad__: does gmail actually provide a standardized intent across all its versions to handle attachments? (or is that essentially what you're finding doesn't exist :-)
Nov 04 18:01:26 <Vlad__> yup, i think that's what doesn't really exist :/
Nov 04 18:02:05 <Vlad__> but there has to be a way to at least support most devices, right?
Nov 04 18:06:49 <kbs> Heh. Well, one silly thought is just to filter on the mime-type rather than the path pattern [assuming you're able to set this when sending the email.]
Nov 04 18:07:45 <kbs> another thought is to see what logcat dumps when the app actually successfully calls another intent, to see what things it is putting into the intent
Nov 04 18:07:53 <Vlad__> how do i set the mime type?
Nov 04 18:07:56 <kbs> er, "calls another app"
Nov 04 18:08:20 <kbs> that should be something feasible when sending the email -- do you have control over this?
Nov 04 18:08:40 <Vlad__> yes i do, but i didn't know i could set the mime type or an attachment
Nov 04 18:08:47 <Vlad__> of an*
Nov 04 18:09:09 <kbs> it should be -- you can specify the mime-type within a multi-part mime, of the attachment itself.
Nov 04 18:09:44 <kbs> but i tthink it might be best to see if you can pull out a successfully launched intent out of logcat, so you can see what gmail actually uses
Nov 04 18:10:16 <kbs> I think i can look in my nook here, let's see...
Nov 04 18:10:45 <Vlad__> oh there is a setDataAndType method, that looks to be it
Nov 04 18:11:01 <Vlad__> not sure what i should pass for the Uri parameter though
Nov 04 18:11:22 <kbs> wait, now you've confused me. I thought the issue was to get gmail to launch you, not the other way around?
Nov 04 18:11:59 <Vlad__> yes, but weren't you suggesting to set my own mime type explicitly and filter on that?
Nov 04 18:12:11 <Vlad__> i do that when attaching to gmail with an intent, right?
Nov 04 18:12:18 <Vlad__> right now it's                     share.setType("application/octet-stream");  actually
Nov 04 18:12:50 <Vlad__> filtering on that would be too general i think
Nov 04 18:13:30 <kbs> do you send an email from within your app first? Or does it arrive from somewhere else?
Nov 04 18:13:40 <Vlad__> i send it from within my app
Nov 04 18:13:56 <Vlad__> by launching gmail
Nov 04 18:14:10 <kbs> ah, I see. So the sequence is a) create intent to trigger gmail b) user taps on attachment in gmail app to call you back
Nov 04 18:14:23 <Vlad__> yes
Nov 04 18:15:07 <kbs> gotcha. Well, I don't know about the (a) part, as I was assuming the email was sent from somewhere else
Nov 04 18:16:12 <Vlad__> i wonder if i can put anything i want with setType and filter on that
Nov 04 18:17:14 <kbs> I think you're now in the realm of needing to work with two undocumented sets of APIs :-) one to send it info, and next to get it to send you info
Nov 04 18:17:33 <Vlad__> lucky me :)
Nov 04 18:17:52 <kbs> but I can check here if I can get a logcat (2.3 device) intent for (b) at least
Nov 04 18:18:14 <Vlad__> that would be great, thanks
Nov 04 18:20:21 <Vlad__> oh well so far this works
Nov 04 18:20:56 <Vlad__> i put application/umsgsave in setType and filtered with <data android:mimeType="application/umsgsave" /> and it works on 4.1 cyanogen mod
Nov 04 18:21:06 <cube> is it possible to get the current status of Android's 100sms/hour limit?
Nov 04 18:21:44 <Vlad__> fingers crossed that it works on others too but i cant test that until tomorrow
Nov 04 18:23:06 <adq> cube, I thought sms is limited only by your carrier
Nov 04 18:23:07 <Vlad__> thanks for giving me this idea kbs :)
Nov 04 18:24:11 <cube> adq: stock android has that limit, if you try to send more than 100sms/hour it will show you a popup window to confirm sending
Nov 04 18:24:13 <kbs> { dat=content://gmail-ls/messages/kbsriram%40xxxxx.xxx/110/attachments/0.1/SIMPLE/false cmp=com.google.android.gm/.ViewAttachmentActivity }
Nov 04 18:24:53 <kbs> Vlad__: so on 2.3, it appears that it's only putting that data and nothing else
Nov 04 18:25:06 <kbs> oh no, I'm sorry
Nov 04 18:25:11 <adq> cube, it's probably to avoid automated spamming :D looks like this counter and its limit is stored in settings.db
Nov 04 18:25:41 <kbs> Starting: Intent { act=android.intent.action.VIEW dat=content://gmail-ls/messages/kbsriram%40xxxxx.xxx/110/attachments/0.1/SIMPLE/false typ=image/jpeg flg=0x80001 cmp=android/com.android.internal.app.ResolverActivity }
Nov 04 18:26:05 <Vlad__> i think that's what it's putting on 4.1 too, i remember seeing something similar when printing the path of a file opened when you click view first instead of save
Nov 04 18:26:31 <kbs> somewhere in there should be enough -- I guess it's only matching on the mime type
Nov 04 18:26:44 <cube> adq: yes it is, can i query that db from my app>
Nov 04 18:26:44 <Vlad__> yup, thanks
Nov 04 18:26:55 <Vlad__> i will provide more feedback tomorrow about how it works on 2.3
Nov 04 18:27:01 <adq> cube, only if you are root I think
Nov 04 18:27:20 <cube> adq: that's what I thought :|
Nov 04 18:27:43 <adq> hopefully, imagine any app can query other app data...
Nov 04 18:28:46 <cube> adq: well, systems settings could be read safely, I think
Nov 04 18:43:00 <voso> Vlad__: your ViewTreeObserver trick works like a charm! I suppose the result/price is two layout passes instead of one during startup, but I guess I can live with that. Thank you again for the tip.
Nov 04 18:44:27 <Vlad__> voso: glad it works, np
Nov 04 18:44:30 <shmooz> oh , then I should use that too, because I'm making a fullscreen video when device tilted into landscape
Nov 04 18:44:51 <shmooz> and I never tested on tablet but it looks fine on phone
Nov 04 18:46:05 <shmooz> so that will work on tablets and phones to give the real height and width of full screen in pixels
Nov 04 18:46:33 <Vlad__> it's only needed for 3.0 and 3.1 devices in my experience, but yes, that will work in all cases
Nov 04 18:47:45 <voso> shmooz: register a getViewTreeObserver().addOnGlobalLayoutListener on your root View and call getHeight() in the callback.
Nov 04 18:48:00 <shmooz> you guys can show me some working code snippet of it ?
Nov 04 18:48:16 <shmooz> ok
Nov 04 18:50:05 <Vlad__> also you might want to do logo.getViewTreeObserver().removeGlobalOnLayoutListener(this); where logo is the view you're measuring (declaring it with FILL_PARENT will mean you'll be measuring the entire usable screen space)
Nov 04 19:05:10 <opariti> hi! anyone knows how can I use a tethered mobile to send sensor signals to eclipse emulator (under ubuntu)?
Nov 04 19:05:17 <saturation> how to increase resolution of avd, now if I edit the something the virtual keyboards covers the whole screen
Nov 04 19:05:28 <saturation> I am using it like this ./emulator -avd Default -no-skin -gpu off -scale 1 -no-boot-anim -dpi-device 320
Nov 04 19:12:56 <jmfurlott> Can someone please walk me through how to draw 3D objects in android? (all primitive objects). How would something like jogl work?
Nov 04 19:19:10 <sampullman> jmfurlott: what's jogl? java openGL?
Nov 04 19:19:28 <jasta> doclava is so rad :)  thank you google
Nov 04 19:19:28 <jmfurlott> sampullman, a java binding to use openGL
Nov 04 19:19:49 <sampullman> drawing 3d stuff in openGL on android is no different than opengl anywhere else
Nov 04 19:20:39 <sampullman> you just render a bunch of triangles
Nov 04 19:21:07 * vandenoever is looking to make a versioned filesystem via a contentprovider
Nov 04 19:21:26 <jmfurlott> sampullman, okay I will learn how to use openGL then
Nov 04 19:21:31 <jmfurlott> sampullman, thank you sir
Nov 04 19:21:45 <sampullman> openGL ES, which is mostly the same but not quite
Nov 04 19:22:43 <viran> hey, is it ok to use a static variable in an activity to save values when the activity is recreated? lets say when moving from landscape to portrait?
Nov 04 19:23:36 <dragorn> viran: you should use the save state feature
Nov 04 19:24:08 <jasta> dragorn: err, its not that simple.
Nov 04 19:24:19 <jasta> viran: typically you're best served by using fragments and retaining the fragment.
Nov 04 19:24:32 <jasta> there are ApiDemos examples that show this
Nov 04 19:24:32 <dragorn> jasta: if he's storing a boolean?  should be relatively simple
Nov 04 19:24:53 <dragorn> jasta: unless I've been using them wrong all along...
Nov 04 19:25:02 <jasta> dragorn: you very likely have been.
Nov 04 19:25:58 <jasta> there are an unfortunately large number of gotchas and nuiances between the various methods for saving state.  saving instance state, for example, is designed to save _UI_ state that can be recovered for example if your process dies while in the background
Nov 04 19:26:09 <jasta> things like the position in a list is saved that way automatically for you by ListView
Nov 04 19:26:46 <dragorn> jasta: I suppose I should have clarified if his boolean was related to his UI display, yes.
Nov 04 19:26:47 <jasta> retaining complex objects, for example an adapter, some downloaded data, etc to quickly recover from orientation change is saved using "retained" data mechanisms.  the easiest of which to use is offered through Fragment's setRetainInstance
Nov 04 19:27:12 <jasta> but of course other legacy apis do exist
Nov 04 19:27:21 <jasta> they're just kludgey and hard to use
Nov 04 19:27:54 <jasta> dragorn: the distinction is very important.  typically when someone is talking about saving state through orientation they are referring to the retain set of APIs.  most UI instance state is saved for you automatically and there are very, very few exceptions out there.
Nov 04 19:28:07 <jasta> basically unless you've built custom controls, you're probably doing it wrong to use the save instance state method
Nov 04 19:28:24 <dragorn> since we've now discussed his problem about 10x what he asked, I'll wait for him to actually clarify what he needs.
Nov 04 19:28:34 <jasta> and there are very subtle ways that test/demonstrate that you're doing it wrong.  it can be very hard to notice if you don't know your edge cases...
Nov 04 19:28:53 <jasta> dragorn: it's not just about helping him.  i'm explaining this for your benefit too :)
Nov 04 19:28:57 <dragorn> all my state is stored in services except some specific UI tweaks
Nov 04 19:29:18 <dragorn> I simply interpreted his question as pertaining to UI stuff
Nov 04 19:29:52 <jasta> why would a service save state?  the primary function of service is to escape process death when your application is backgrounded.
Nov 04 19:30:14 <jasta> everything on top of that is just a nicety to support situations where that is true (that you wish to escape process death)
Nov 04 19:30:24 <dragorn> jasta: because all the work I do has to occur in services, because it's independent of UI state or has to preserve connections when the UI isn't active
Nov 04 19:30:38 <jasta> ah well, carry on then ;)
Nov 04 19:30:48 <dragorn> therefore all of the state of my application is in the service.  The UI is a bit of glitter on top to give the user something to look at, and, arguably, a way to kill the service
Nov 04 19:30:57 <dragorn> otherwise I practically don't need a UI
Nov 04 19:31:40 <voso> jasta, what is 'legacy api' on a static field? if the data is not UI related, i dont see a problem with a static field either. if there is one, please enlighten me. btw I tend to keep this type of data in my App, not an Activity. but still..
Nov 04 19:31:59 <dragorn> pretty much everything I do is centered around data gathering - either via USB or rfcomm, and logging.  Which need to be independent of the UI display.  Pretty much all my UI interaction is read only database access to the backend, plus some state messages on the service handlers
Nov 04 19:32:04 <jasta> voso: well, you now can't have more than one instance of your activity for example.
Nov 04 19:32:09 <jasta> at least not with differing state
Nov 04 19:32:43 <voso> thats true.
Nov 04 19:32:45 <jasta> also, i think the argument is best reversed: what advantage is there to using a static field over fragment's retain instance capability?  both are extremely simple.  the latter, however, has few drawbacks.
Nov 04 19:32:49 <jasta> if any
Nov 04 19:33:15 <jasta> bbl, lunch
Nov 04 19:56:30 <FuriousScissors> Hi all
Nov 04 19:57:16 <FuriousScissors> I've been attempting to configure Ortho for a Nexus 7 device but I've been having some troubles, any one have any experience with OpenGL ES 2?
Nov 04 19:57:50 <dioHere> can someone tell me how can I know that a code that is run with runOnUIThread is finished, in order to continue from another point of my code? I want to make something like wait and signal but I don't know how to stop my code from continuing before the UI thread finishes
Nov 04 19:58:42 <FuriousScissors> I get "Called unimplemented OpenGL ES API" every time i call gl.glOrtho or GLU.gluOrtho2d
Nov 04 19:58:57 <dragorn> dioHere: you could pass a message with handlers.  Also since it's all the same activity (I presume) you could use traditional thread sharing, like locking an object while it's running and waiting for it to unlock
Nov 04 20:00:01 <dioHere> dragorn I have the runonuithread method to run in another class. can I do that with synchronized block?
Nov 04 20:00:09 <dragorn> dioHere: there's also asynctask, depending on what you're doing
Nov 04 20:00:48 <dindinx> FuriousScissors: ES 2 doesn't have fixed-pipeline matrices.
Nov 04 20:00:54 <dragorn> dioHere: ah.  So you'd have to have an instance - you could pass your instance in your function on your UI class you're running on the UI thread and have it call your function as a callback
Nov 04 20:02:45 <dioHere> dragorn I will try everything you've said. thank you very much
Nov 04 20:03:34 <dragorn> dioHere: ie, public class whatever { uiDone(); }  ... class myactivity { runningOnThread(whatever cbwhatever); } ...  runOnUi() { someActivity.runningOnThread(this); }
Nov 04 20:04:02 <dragorn> dioHere: similar to an onClick handler calling back
Nov 04 20:04:58 <FuriousScissors> dindinx: I see, is there any recommended way of drawing 2D within ES 2 or is the solution to stick with ES 1?
Nov 04 20:07:51 <dioHere> dragorn: yes I think I understand. I am a little confused about these kind of things but I hope I will manage to do it
Nov 04 20:08:03 <dindinx> FuriousScissors: juste use your own matrix library.
Nov 04 20:09:08 <dragorn> dioHere: I suspect doing a callback is going to be what you want
Nov 04 20:09:30 <FuriousScissors> dindinx: I'll do some research on that, thank-you.
Nov 04 20:16:49 <dindinx> rakshith wants to ask something, but prefer to anoy people in private
Nov 04 20:18:17 <rakshith> sorry ?
Nov 04 20:18:48 <dindinx> rakshith: ask your question here, not in private.
Nov 04 20:18:57 <rakshith> dindinx: got it
Nov 04 20:19:25 <rakshith> dindinx: i asked you because,i saw you active
Nov 04 20:19:32 <rakshith> dindinx: sorry man
Nov 04 20:19:53 <gutano> rakshith, I doubt he's offended... but you never know who's got info that could help you
Nov 04 20:20:20 <gutano> plus, it's nice to learn from Q/A (even if they aren't your own)
Nov 04 20:20:28 <rakshith> gutano: the problem is,i have signed
Nov 04 20:21:05 <gutano> if your going to violate, might as well do it publicly, lol
Nov 04 20:21:24 <rakshith> gutano: a apk,and when i try to install it in emulator,it says "file xx.apk has missing AndroidManifest.xml "
Nov 04 20:21:53 <Vlad__> is it missing that file?
Nov 04 20:22:23 <gutano> rakshith, does the non-signed ver have this prob?
Nov 04 20:22:35 <rakshith> Vlad__: i did zip the apk and extracted the file,AndroidManifest.xml was there
Nov 04 20:22:42 <rakshith> gutano: no
Nov 04 20:23:34 <rakshith> gutano: i dont think i will be able to upload it to the market as well
Nov 04 20:24:19 <gutano> well if it doesn't work, you certainly don't want to do that nayway
Nov 04 20:24:46 <rakshith> gutano: how do i check it then ?
Nov 04 20:25:09 <gutano> rakshith, http://developer.android.com/tools/publishing/app-signing.html has several methods of signing
Nov 04 20:25:15 <rakshith> gutano: like if my apk is ready to be released to everybody
Nov 04 20:25:23 <gutano> it's not
Nov 04 20:25:27 <gutano> or it wouldn't error when you ran it
Nov 04 20:25:34 <gutano> you have to fix your signing issue
Nov 04 20:25:47 <rakshith> gutano: signing is working cool
Nov 04 20:25:59 <rakshith> gutano: it says jar verified
Nov 04 20:26:02 <dragorn> pretty sure when the market attempts to read the manifest and blows up it won't let you publish anyhow :P
Nov 04 20:26:06 <gutano> ok
Nov 04 20:26:12 <gutano> nothing to discuss then
Nov 04 20:26:21 <rakshith> sorry ?
Nov 04 20:26:36 <rakshith> dragorn: any way we can fix it ?
Nov 04 20:26:43 <dragorn> no clue
Nov 04 20:26:59 <dragorn> I don't even know how you'd get it into such a state to be honest
Nov 04 20:27:05 <Vlad__> just blindly guessing here, but are you using the code obfuscator too?
Nov 04 20:27:35 <Vlad__> if you are, try making a signed apk without it
Nov 04 20:27:45 <Vlad__> see if that works
Nov 04 20:27:57 <rakshith> dragorn: i am signing it manually,without using eclipse
Nov 04 20:28:24 <rakshith> Vlad__: no,i am not using code obfuscator
Nov 04 20:28:51 <rakshith> dragorn: thru terminal using jarsigner and keytool
Nov 04 20:28:57 <dragorn> rakshith: I'd guess you're signing the wrong thing, or somehow otherwise missing a step.  Like I said, I have no idea how you'd get such a failure
Nov 04 20:29:11 <rakshith> ok,thanks
Nov 04 20:29:11 <dragorn> I've never encountered it.
Nov 04 20:29:17 <Vlad__> try signing it with eclipse then
Nov 04 20:30:09 <gandhijee_> hi, i am trying to view the pdk after building it from the android sources...
Nov 04 20:30:11 <rakshith> Vlad__: yes,will give a  try
Nov 04 20:30:22 <gandhijee_> but the directions from the readme don't seem to work.... any suggestions?
Nov 04 20:30:57 <rakshith> pdk ??
Nov 04 20:36:26 <mdkess> I have a ListView where I want items to be swipeable left and right to scroll through pictures (similar to the pictures in the Verge's cool app). Should I have a FragmentPagerAdapter per row of the list, or should that be attached to my ViewHolder?
Nov 04 20:36:58 <lasserix> Is there a way to change the default path nine patch tool editor open/save dialog points too?
Nov 04 20:38:45 <lasserix> mdkess it sounds like you are using both a pager and listview which defeats the function of both since, one is supposed to scroll vertically and the other horizontally
Nov 04 20:44:36 <mdkess> not only that, but several viewpagers :)
Nov 04 20:45:05 <mdkess> So I have a viewpager that is a set of ListView based fragments. These scroll horizontally.
Nov 04 20:45:31 <mdkess> Then each ListView scrolls vertically, naturally. But some of the ListView items will scroll horizontally too (the ones with their own pagers). The Verge does this, it's not as complicated as it sounds.
Nov 04 20:45:49 <mdkess> I mean, complicated from a user's perspective.
Nov 04 20:46:03 <lasserix> Yeah I just downloaded it, I think I see what you're getting at
Nov 04 20:46:13 <mdkess> Pretty sweet app too, right?
Nov 04 20:46:52 <lasserix> It's okay: to swip left and right but also have the sidebar thing that comes out on a swipe left to right is a little convuloted
Nov 04 20:47:35 <mdkess> I like the idea of side drawers (I think Dolphin was the first app that I saw that really used that pattern), but I find it's hard to communicate to the user.
Nov 04 20:49:47 <gandhijee_> rakshith: Product Development Kit
Nov 04 21:00:13 <rakshith> hello
Nov 04 21:00:16 <rakshith> ??
Nov 04 21:36:26 <mrtncx> Hi devs. Just reading the API guide and noticed this "Each process has its own virtual machine (VM), so an application's code runs in isolation from other applications.". Does it just mean virtual memory with its own address space? Or is it a full blown virtual machine? Virtual machine implies hypervisor, hardware abstraction etc. Is that the case here?
Nov 04 21:37:44 <bankai_> it's talking about jvm
Nov 04 21:37:57 <jonrafkind> I am trying to use gdbserver to debug my ndk app but how do I run gdbserver on the device? if I push it over adb its not executable and chmod can't seem to make it executable
Nov 04 21:39:14 <mrtncx> You know what, I read that back just now and realised that it was obvious. Thanks.
Nov 04 21:47:58 <Syzygy> Is there a guide somehow on how to create animated splash screens?
Nov 04 21:55:25 <Syzygy> Every guide on it I find says that for a splash screen I should create it's own activity... but wouldn't it make more sense to simply put it into the onCreate() method?
Nov 04 21:55:56 <adq> more sense not sure, but more easy maybe
Nov 04 21:56:26 <kowalej> it's better to seperate it
Nov 04 21:56:51 <kowalej> such as if you wantet to launch your app and bypass the splashscreen
Nov 04 21:57:21 <adq> fact is when the splash screen has finished its animation, there is no need to keep it (in memory per example), so if the splash is its own activity, ondestroy is called when the splash is finished (end of animation or by user) and the resource of the splash are gone
Nov 04 21:58:03 <Syzygy> I can see why I would like to do that as a dev, but I don't see why I would give users the ability to bypass my splash screen
Nov 04 21:58:11 <Syzygy> that makes sense adq, thanks
Nov 04 21:58:13 <kowalej> ya and it's also a bit cleaner if you want to reuse the splashscreen
Nov 04 21:58:32 <adq> users don't like to wait 10 seconds (if there is no purpose except the displaying of the splash)
Nov 04 21:58:48 <kowalej> ya definitely make a click through option
Nov 04 21:59:22 <AlexZan> is app engine a good web service solution for native android apps? or are there better options? REST, RPC, etc?
Nov 04 21:59:40 <Syzygy> yeah, it will be only 1 second and (obviously) only when starting the app (a game) and not when bringing it back from the background
Nov 04 21:59:52 <adq> Syzygy, if you really don't want to bother with a splash in a separate activity, you can use a relative layout, put the splash on top of everything, apply an animation (per example a fade out), and then remove the view
Nov 04 21:59:55 <adq> but it's not the way to go
Nov 04 21:59:57 <Belacan> hey
Nov 04 22:00:14 <Syzygy> actually, not a second but depending on how long everything will take to load
Nov 04 22:00:18 <Belacan> does anyone know the sample rate for the gps?
Nov 04 22:00:22 <Belacan> the maximum
Nov 04 22:00:26 <monsti> Syzygy: i delayed my splash
Nov 04 22:00:30 <Syzygy> (and it's really just a requirement from my uni course)
Nov 04 22:00:35 <monsti> Syzygy: a min time of 3 seconds
Nov 04 22:00:54 <Syzygy> monsti, so you have to wait at least 3 seconds no matter what?
Nov 04 22:00:56 <adq> Syzygy, you can short-circuit an animation too, via overriding its methods
Nov 04 22:01:18 <Syzygy> that's a bit to advanced for me right now i guess
Nov 04 22:01:25 <monsti> Syzygy: yeah
Nov 04 22:01:40 <monsti> Syzygy: i have an Appllication class that does some initalization
Nov 04 22:01:49 <Syzygy> if i put my splash screen in a seperate app, can I still tie it to the time it takes my app to initialize?
Nov 04 22:01:49 <monsti> Syzygy: the splash polls if this is finished
Nov 04 22:02:09 <monsti> Syzygy: if it's not yet finished it waits until done - else wait minimum of 3 seconds
Nov 04 22:02:19 <Syzygy> ok, that makes sense
Nov 04 22:02:30 <Syzygy> would you be kind enough to share your code?
Nov 04 22:02:43 <monsti> Syzygy: it's just a AsyncTask
Nov 04 22:02:56 <Syzygy> I'll look it up then
Nov 04 22:03:02 <monsti> a busy wait while(1) ... sleep(50ms) ...
Nov 04 22:03:14 <monsti> if(app.done) break;
Nov 04 22:03:41 <Syzygy> and is it it's own activity?
Nov 04 22:03:47 <monsti> yeah
Nov 04 22:03:54 <monsti> it's the start activity
Nov 04 22:04:02 <Syzygy> ok
Nov 04 22:04:13 <monsti> but - of course - you can do this in any other activity
Nov 04 22:04:21 <Syzygy> yeah
Nov 04 22:05:05 <AlexZan> an android app with GWT does not use the native UI correct?
Nov 04 22:05:40 <Syzygy> so my final question on this is... can someone recommend a guide for how to create animated splash screens? all i found till now is how to fade them.
Nov 04 22:06:04 <monsti> Syzygy: android 3.0+ ?
Nov 04 22:06:14 <monsti> Syzygy: read the heavy chapter about animations :)
Nov 04 22:06:25 <Syzygy> 2.3 is our minimum required platform iirc
Nov 04 22:07:17 <ivdorelian> or you can always make your frames as png files in photoshop and animate them manually in onDraw()
Nov 04 22:07:21 <kowalej> or use a video
Nov 04 22:07:33 <adq> so many possibilities :)
Nov 04 22:07:38 <monsti> Syzygy: i started by looking at the drawables for the progress cursor
Nov 04 22:08:19 <Syzygy> I guess making a video would be the easiest for me
Nov 04 22:09:02 <kowalej> it's the best in terms of consistency
Nov 04 22:10:37 <Syzygy> I mean the animation wouldn't be that hard to do i guess
Nov 04 22:10:41 <Syzygy> https://dl.dropbox.com/u/31690810/nan.png
Nov 04 22:10:48 <Syzygy> just the not sliding in from the left
Nov 04 22:10:52 <adq> yeah but think of loading the video & co, you will face more problems :)
Nov 04 22:10:56 <adq> go simple first
Nov 04 22:11:10 <adq> (because from the complexity you cannot go back to simplicity too ;))
Nov 04 22:11:32 <Syzygy> But loading videos is a requirements of the course as well
Nov 04 22:11:35 <adq> Syzygy, nice logo
Nov 04 22:11:42 <Syzygy> so 2 flies with one stone
Nov 04 22:11:44 <Syzygy> thanks
Nov 04 22:11:47 <adq> (looks like for an irc client :))
Nov 04 22:11:51 <kowalej> ya but when you need your splash animation to be consistent over all hardware especially older devices video is the way to go
Nov 04 22:12:03 <kowalej> and if it's a small animation the video should be small
Nov 04 22:12:05 <adq> agreed
Nov 04 22:12:28 <kowalej> of course if you don't know what you're doing with rendering the video it could end up way too big and cause problems :P
Nov 04 22:12:36 <Syzygy> well, preferably I would time the animation so that it finishes when loading is done
Nov 04 22:12:41 <Syzygy> I've rendered videos before
Nov 04 22:12:53 <Syzygy> any recommended resolution?
Nov 04 22:16:35 <kowalej> I used 800x480
Nov 04 22:16:53 <kowalej> it's 4 seconds and is 121kb
Nov 04 22:19:19 <Syzygy> so no real conventions there apperently, in that case I'll stick to the highest resolution the phones we use for developing have.
Nov 04 22:21:23 <kowalej> well there is a guide to screen support: http://developer.android.com/guide/practices/screens_support.html
Nov 04 22:22:15 <ivdorelian> i suggest you do the video at whatever resolution looks good and is small enough for your liking and then resize it to a percentage of the screen at runtime
Nov 04 22:26:58 <Syzygy> in the manifest, if i do this: android:screenOrientation="landscape" in the activity tag, I can enforce landscape mode. can i enforce it on all activities if i just do it in the application tag?
Nov 04 22:27:21 <AlexZan> hey guys, is there a good RPC solution for App engine that works on android native, not web app? and has rpc code generators? backend is in java so that I can use tdd both front and backend?
Nov 04 22:27:57 <ivdorelian> Syzygy why don't you try it out :P
Nov 04 22:28:07 <ivdorelian> afaik you can't do that at application level
Nov 04 22:28:16 <ivdorelian> it must be done for each activity
Nov 04 22:28:52 <Syzygy> I don't know where all my cables are D:
Nov 04 22:29:28 <ivdorelian> you don't need cable, your IDE should tell you if you can use android:screenOrientation in the application tag or not
Nov 04 22:29:45 <ivdorelian> type it and see if autocomplete fires up
Nov 04 22:30:40 <ivdorelian> and you should get an error if you try to compile and it doesn't belong there, or at least a warning
Nov 04 22:32:14 <Syzygy> don't get anything and it doesn't work
Nov 04 22:33:21 <marsilainen> I've spent the whole weekend, on & off, trying to figure out how to do image caching... I feel like I have a better understanding of what I need to do know, though I'm still some way off having a working implementation... anyone else had similar frustrations with caching?
Nov 04 22:33:46 <marsilainen> s/know/now
Nov 04 22:34:07 <jasta> uhh.
Nov 04 22:34:07 <ivdorelian> what do you mean by image caching?
Nov 04 22:34:08 <jasta> no?
Nov 04 22:34:28 <Syzygy> I have similar frustrations with my whole life D:
Nov 04 22:34:30 <marsilainen> I mean generating bitmaps and putting them into memory & disk caches
Nov 04 22:34:35 <marsilainen> and then managing those
Nov 04 22:34:39 <jasta> that's really a trivial problem
Nov 04 22:34:44 <jasta> 100 line utility class tops
Nov 04 22:34:48 <marsilainen> images generated in the background and so on
Nov 04 22:35:23 <marsilainen> well my main frustrations have been that the example of how to do it from the android docs contains over 10 classes for a start
Nov 04 22:35:28 <jasta> generating them (why are you generating images?) is perhaps somewhat trickier, but still pretty simple.
Nov 04 22:35:39 <jasta> marsilainen: where are you looking at android's docs?
Nov 04 22:36:03 <marsilainen> http://developer.android.com/training/displaying-bitmaps/cache-bitmap.html
Nov 04 22:36:16 <marsilainen> I was looking at that doc and the example they provide
Nov 04 22:37:36 <marsilainen> I'm generating images because my app does custom drawing (of a board game) and I want to provide some thumbnails for browsing through saved game positions
Nov 04 22:37:43 <jasta> just scanned it, sounds perfectly reasonable for the case its outlining
Nov 04 22:37:59 <jasta> is this somehow excessively complicated to you?
Nov 04 22:38:22 <jasta> they've basically given you the code plainly.  the only frustration i could think of is if you can't grok it.
Nov 04 22:38:48 <marsilainen> well I think the example they provide is over-complicated really
Nov 04 22:39:05 <jasta> you think that based on what?
Nov 04 22:39:26 <marsilainen> they have three classes - one extended from the other from the other - when I think they could have done that as one class for simplicity and ease of understanding
Nov 04 22:39:35 <marsilainen> for example.
Nov 04 22:39:55 <jasta> ok, i don't agree, but whatever.
Nov 04 22:40:06 <marsilainen> well clearly you're smarter than me
Nov 04 22:40:30 <jasta> can you please just narrow in on your actual concern so that i could address something tangible
Nov 04 22:41:17 <ivdorelian> it might help to start an empty project, cnp the code and then run through it with debug statements (Log, alertdialogs etc.) until you understand it
Nov 04 22:41:23 <marsilainen> I was only asking if other people had taken as long as I had to get their head around implementing caching for the first time given the docs...
Nov 04 22:41:43 <jasta> marsilainen: i implemented this stuff myself prior to google doing it and, no, i didn't even find that difficult or frustrating.
Nov 04 22:41:58 <jasta> you now have well thought through and tested code to just copy and paste into your project.  *confused*
Nov 04 22:42:24 <marsilainen> ivdorelian: yeah, that's the sort of thing I ended up doing, just found it took me longer to get my head round it than other android things
Nov 04 22:42:45 <marsilainen> jasta: sure - it's probably at least in part because I'm fairly new to android dev
Nov 04 22:43:02 <jasta> so you're frustrated that you're learning new things?  you're really losing me here...
Nov 04 22:43:45 <jasta> when you encounter material you don't know, it takes time to understand it.  when you encounter material you understand, you quickly move through it.  with experience, most things will be easy to move through.
Nov 04 22:43:49 <jasta> moving on now...
Nov 04 22:44:21 <marsilainen> I love learning new things - I found this particular aspect took a bit longer to get my head round than I would have liked... but I'm perfectly willing to accept that my lack of knowledge/skill is to blame... just wondered if other people have had the same issue when implementing caching the first time
Nov 04 22:44:56 <jasta> well i will tell you that way back in the early days of android running on a G1 it was a nightmarish problem to try to achieve decent performance scrolling through a list of images
Nov 04 22:45:00 <jasta> google themselves struggled with it
Nov 04 22:45:23 <jasta> and through all the pain and torture, some very strong and simple best practices emerged such as the one you found.
Nov 04 22:46:42 <jasta> consider yourself lucky to have the opportunity to learn from those headaches and not repeat them.  and i assure you, if you find the results confusing or overcomplicated, you're just lacking experience in this space.
Nov 04 22:46:51 <marsilainen> sure. And that's fine, I can understand that. I would recommend to Google that for beginners such as myself that the docs on that probably could do with being made a little easier to understand. People like you who obviously have experience probably don't need such things
Nov 04 22:47:05 <jasta> marsilainen: patches welcome.
Nov 04 22:47:16 <marsilainen> where do I submit?
Nov 04 22:47:25 <jasta> source.android.com
Nov 04 22:47:37 <marsilainen> ok
Nov 04 22:47:40 <jasta> those guides are in the aosp packages
Nov 04 22:48:25 <marsilainen> as I say, I'm not complaining about the solution - just that for beginners it needs a bit better documentation IMHO - speaking as someone doing his best to learn
Nov 04 22:48:53 <marsilainen> I have to say that in general I've found the android docs pretty good
Nov 04 22:48:55 <jasta> that's probably a very accurate observation that you're in a unique position to correct.
Nov 04 22:48:59 <jasta> so, please do :)
Nov 04 22:49:33 <marsilainen> the docs available on vogella.com have been invaluable to me too
Nov 04 22:49:49 <marsilainen> the docs there have got me through learning to use loaders, content providers etc
Nov 04 22:51:01 <kazimierz> test
Nov 04 22:51:07 <kazimierz> ha!
Nov 04 22:51:11 <kazimierz> hi :)
Nov 04 22:51:22 <kazimierz> sorry for that
Nov 04 23:17:01 <neure> hi
Nov 04 23:17:13 <neure> can i use functions in Android.mk ?
Nov 04 23:26:47 <vydd> neure: I haven't tried that, but you could test it quickly; here's the syntax for gnumake: http://www.gnu.org/software/make/manual/html_node/Functions.html
Nov 04 23:26:52 <Seus> question: I'm working on a simple app, and am trying to utilize the Apache EntityUtils class. I've imported 'org.apache.http.util.EntityUtils', yet Eclipse says 'EntityUtils.toString cannot be resolved to a type'…I've tried cleaning the project, but no luck. Just updated to the latest version of Eclipse as well…thoughts?
Nov 04 23:28:19 <Belacan> yo
Nov 04 23:28:30 <Belacan> who knows what a blinking GPS icon at the top of the phone means
Nov 04 23:28:38 <Belacan> does it mean that a location is not fixed?
Nov 04 23:29:13 <Seus> i would think it just means gps is on, no?
Nov 04 23:31:52 <vydd> neure: tested. you can.
Nov 04 23:35:30 <neure> how do i get verbose output from ndk-build ?
Nov 04 23:35:40 <neure> i need to see which command it is executing
Nov 04 23:35:46 <neure> as my includes are not workinbg
Nov 04 23:38:40 <RustyShackleford> i need to make 143 requests to a server to populate my database
Nov 04 23:39:02 <RustyShackleford> I've done it a few times rapid fire but I need to be kinder to their server
Nov 04 23:39:26 <RustyShackleford> how can I space them out a bit?
Nov 04 23:43:22 <neure> ndk complains string header is not found
Nov 04 23:49:36 <vydd> neure: --debug ?
Nov 04 23:49:54 <lasserix> RustyShackleford use a handler/runnable with a post delayed
Nov 04 23:50:41 <neure> fatal error: string: No such file or directory
Nov 04 23:50:47 <neure> what could be causing that?
Nov 04 23:51:02 <neure> #include <string> should work, shouldn't it?
Nov 04 23:51:48 <neure> i also have -std=c++0x
Nov 05 00:08:05 <mrenouf> setBackgroundColor(0) causes nothing to be painted at all, right?
Nov 05 00:08:25 <mrenouf> due to opacity = 0 ?
Nov 05 00:17:27 <bliynd> can anyone help me with this? http://stackoverflow.com/questions/13221970/ontouch-with-multitouch-and-soundpool-motion-events
Nov 05 00:24:06 <pragma-> stackoverflow can help you with it
Nov 05 00:26:22 <pierpark> Hey guys. Have a quick question (I hope). I have this "GPSTracker" that extends Service and Implements locationListener... I use it from another activity like so: GPSTracker gps; and then do stuff like gps.getLatitude; Well, I have a "public void onLocationChanged(Location location) {" line that works and lets me know when their location has changed... is there any way to listen for this
Nov 05 00:26:22 <pierpark> in my other activity? I have a map view that I use in there, and it'd be nice to get that "change" data inside my activity.
Nov 05 00:26:42 <pierpark> Or do I have to set up a timer and see "if location changed" by calling it
Nov 05 00:27:43 <diki> Can using a few buttons,textviews and edittext fields cause my app to use 10 megabytes for just that?
Nov 05 00:27:58 <diki> Cause my avg usage is over 10 megabytes...
Nov 05 00:28:13 <pierpark> diki: Doesn't sound right to me.
Nov 05 00:28:48 <diki> I did perform an analysis with MAT, but..
Nov 05 00:29:18 <diki> it showed the culprit as android.content.res.Resource which after searching in the histrogram, turns out to be the UI elements(textviews,edittext whatever).
Nov 05 00:30:05 <diki> Is there some cleanup I am forgetting? All my elements are never hidden, always shown to the user.
Nov 05 00:30:23 <pierpark> diki: Ah, right. That wouldn't be your app though.
Nov 05 00:30:33 <pierpark> Are you having memory overload problems?
Nov 05 00:31:05 <diki> pierpark:Not yet. But I can't have my app using that much memory anyway.
Nov 05 00:32:15 <pierpark> There are times when you might load a ton of data in from the web, or end up with a large cache that is being read in all at once, and you can certainly clear your cache.. but as far as the UI elements, those are handled by Dalvik, which is Android's virtual machine that all apps run in. <--- definition might not be accurate.
Nov 05 00:33:01 <diki> I do not use anything web related. Nor have any cache, nor handle bitmaps(however MAT did show references to bitmaps for whatever reason).
Nov 05 00:33:15 <tdignan> mrenouf: so I tried getRuntime().availableProcessors()... Turns out it only returns CPUs available to the JVM. So for some reason, it always returns 1 on my dual core. However, my /proc/stat reading code ends up detecting the second core when it wakes from sleep
Nov 05 00:33:23 <pierpark> At first, and pretty confident guess... you are in no risk of having an issue from your app. However, other apps with bad programming that are on your device might be causing the high usage.
Nov 05 00:33:33 <mrenouf> hmm
Nov 05 00:33:34 <tdignan> unfortunately this makes me think dalvik is only ever using 1 of my cores
Nov 05 00:33:48 <mrenouf> yeah, wasn't confident about that. that's not necessarily true
Nov 05 00:33:52 <tdignan> its ok though, I add cores dynamically. It would be nice if I knew in advance.
Nov 05 00:33:54 <mrenouf> it's not specified in the spec
Nov 05 00:34:10 <mrenouf> you can do a traceview if you don't beleive it
Nov 05 00:34:18 <mrenouf> or systrace
Nov 05 00:34:19 <tdignan> oh, you mean that dalvik uses both cores?
Nov 05 00:34:26 <tdignan> you're probably right
Nov 05 00:34:29 <mrenouf> don't see any reason it wouldn't
Nov 05 00:34:34 <diki> pierpark:I hope I am not overstepping myself here, but I do have the hprof file. I could send it for someone to take a look at, provided they have free time and all :)
Nov 05 00:34:43 <mrenouf> but it could decide to put a process on a core or set of cores
Nov 05 00:34:48 <tdignan> well I have a fib() function that I spawn threads on to deliberately increase cpu, and it seems to make the other core wake up
Nov 05 00:35:16 <mrenouf> oh right, your trying to burn cpu?
Nov 05 00:35:30 <tdignan> i was earlier. I hid the button. I don't think I want users to have that feature.
Nov 05 00:35:34 <tdignan> I'll just keep it for myself :)
Nov 05 00:35:39 <mrenouf> lol
Nov 05 00:35:39 <pierpark> diki: pop it up somewhere and ask someone to take a look. If nobody responds, try again here in about 2 hours... more people are generally on. In fact, mrenouf might hep you, since they are discussing things along that line.
Nov 05 00:36:17 <mrenouf> sorry, dont have time to analyze hprof traces
Nov 05 00:36:32 <mrenouf> but if you have questions about how to look at them, I might be able to help
Nov 05 00:37:10 <diki> Here it is, http://www35.zippyshare.com/v/95084996/file.html. :)
Nov 05 00:37:23 <pierpark> diki: From what you are describing, there's no way that TextView's and buttons, etc., are going to hog memory. Now if that package you said was the culprit is showing that much, that isn't your app. Follow your package and that should be your only concern ... so look for your com.diki.package .. I'll open that, but I doubt I'll be much help.
Nov 05 00:37:28 <diki> I think it is already converted by hprof-conv/
Nov 05 00:38:10 <diki> pierpark:DDMS also shows my memory usage as 10(even over 12 megabytes).
Nov 05 00:38:16 <diki> for my package, not other.
Nov 05 00:38:30 <pierpark> Odd.
Nov 05 00:39:59 <diki> Ok, the previous hprof I uploaded was not the one from hprof-conv, this one is http://www12.zippyshare.com/v/63901955/file.html
Nov 05 00:40:10 <diki> It will run directly on MAT.
Nov 05 00:40:38 <RustyShackleford> so I finally have a fully populated database on my harddrive
Nov 05 00:40:39 <diki> If this helps, The class "android.content.res.Resources", loaded by "<system class loader>", occupies 6 225 640 (59,31%) bytes. The memory is accumulated in one instance of "java.lang.Object[]" loaded by "<system class loader>".
Nov 05 00:40:47 <RustyShackleford> I want to include it in my app
Nov 05 00:41:04 <RustyShackleford> first of all, where does it go?
Nov 05 00:41:08 <pierpark> I don't think I'm going to be of any use to you, unfortunately.. save your links and hopefully someone that is more 'skilled' in that area can help you. I apologize...
Nov 05 00:41:29 <pierpark> diki ^^
Nov 05 00:41:42 <diki> pierpark:That's ok, I am as mind boggled as the next guy.
Nov 05 00:41:48 <diki> :)
Nov 05 00:42:27 <pierpark> I'm actually trying to work out a massive memory problem. The Maps API really really needs an update.
Nov 05 00:43:10 <RustyShackleford> i've included some text files in res/raw before
Nov 05 00:43:43 <RustyShackleford> I want my database to be stored in /data/data/packagename/Databases/database.db
Nov 05 00:44:18 <pierpark> RustyShackleford: You'll need to take a look into the sqlite ? documentation I believe.
Nov 05 00:44:26 <RustyShackleford> also, do I need to still subclass SQLiteOpenHelper if my database is already populated?
Nov 05 00:44:31 <diki> Just to add more infor for clarity, I am also using JNI, however the C code uses no more than 1 megabyte, and I vouch for even less.
Nov 05 00:44:39 <RustyShackleford> i do not need to change this information, just query it
Nov 05 00:45:35 <diki> The even weirdest part is, that the emulator shows my application using only 2-3 megabytes of RAM, which is multitudes lower than on the phone.
Nov 05 00:45:50 <pierpark> RustyShackleford: http://www.reigndesign.com/blog/using-your-own-sqlite-database-in-android-applications/
Nov 05 00:46:04 <pierpark> RustyShackleford: That should answer your question and provide you an example.
Nov 05 00:47:25 <RustyShackleford> pierpark: perfect, thank you
Nov 05 00:47:34 <lasserix> RustyShakleford, when you include your own sqlite database in the APK note that it will require twice the size
Nov 05 00:47:53 <lasserix> because the included database won't be deleted after you use it to create the database internal to your app
Nov 05 00:48:36 <lasserix> *using the method the link pierpark put up
Nov 05 00:49:02 <RustyShackleford> so it will simply make a copy?
Nov 05 00:49:14 <lasserix> yeah but the copy will be available to your app
Nov 05 00:49:23 <RustyShackleford> 16KB, that shouldn't be too much of a problem
Nov 05 00:49:33 <lasserix> oh yeah no problem :)
Nov 05 00:49:37 <RustyShackleford> can I delete the original copy after installing?
Nov 05 00:49:41 <tdignan> I read a thesis that says right now is when the most people are playing android games
Nov 05 00:49:47 <lasserix> I don't believe so
Nov 05 00:50:39 <lasserix> if it's that small why not just format the text files and populate your database with those
Nov 05 00:50:59 <lasserix> *the text files you formelly put into raw
Nov 05 00:51:19 <lasserix> like in the oncreate method, just go line by line and insert them?
Nov 05 00:52:10 <lasserix> the reigndesign method is usually for large databases (like i used it for a database >2mb) but only because i don't have a dedicated web server i could pull the info from
Nov 05 00:52:31 <pierpark> lasserix is right... after further reading. I'd suggest putting that database file on the web on your server, and that way you could update the database file, and then the app could update the records if you 'flag' a database change. that way, there's no .db file included in your app, and you'll have the power to update the database without making them update the actual app.
Nov 05 00:54:46 <jero> anyone knows what app is that task list from, the one seen when long-pressing the home key?
Nov 05 00:54:51 <pierpark> meaning: app calls up http://site/datebase.db if.. lets say, http://site/datbase_info.txt says.. "update 2" and, the app says "eh, I've got update 1" .. then it'll download the database, update internal database, then you can delete the database file you just downloaded while retaining the data.
Nov 05 00:55:35 <lasserix> pierpark: but he's only using a db that is 16kb in size. easiest way would to be just do read line by line from a text file in raw and populate in oncreate
Nov 05 00:56:12 <pierpark> lasserix: yeah, definitely true.
Nov 05 01:29:54 <xroberx> hi
Nov 05 01:30:49 <xroberx> Does anyone know what's the use of having both onSurfaceCreated() and onSurfaceChanged() if the latter gets called just after the former and both are called when the surface is recreated ?
Nov 05 01:33:00 <xroberx> in other words, is there any case in which onSurfaceCreated() gets called but onSurfaceChanged() isn't ?
Nov 05 01:33:24 <Syzygy> uuhm, is Void (uppercase v!) something java or android specific?
Nov 05 01:33:30 <Syzygy> as I've never encountered it.
Nov 05 01:35:03 <Syzygy> xroberx, I'm not an expert by far, but I belive onSurfaceChanged() is called when the screen orientation changes and judging from the name it might have some connection to fragments as well.
Nov 05 01:35:38 <xroberx> Syzygy: thank you for your interest, but it has nothing to do with fragments
Nov 05 01:36:44 <xroberx> I'm coming to the conclusion that having both functions is just a design flaw like many others
Nov 05 01:39:46 <Syzygy> that's always a possibility as well
Nov 05 01:40:10 <tdignan> anyone have experience with the differences between: free version and paid version, separate apps, and in-app billing, single free version? Which freemium model do you think is better?
Nov 05 01:40:14 <tdignan> I'm thinking of doing both
Nov 05 01:40:29 <tdignan> having a free app with in-app billing, and also cloning it and selling an all-in-one paid version
Nov 05 01:40:36 <tdignan> that has all features enabled by default
Nov 05 01:40:44 <clever> how do i make use of the menu button in my activity?
Nov 05 01:41:04 <tdignan> you create a menu in onCreate/onPrepareOptionsMenu
Nov 05 01:41:18 <clever> tdignan: android.view.Menu?
Nov 05 01:41:31 <tdignan> if the code runs on a device that doesn't have a menu button, it will show up as three dots in the actionbar
Nov 05 01:41:34 <tdignan> yeah
Nov 05 01:41:39 <tdignan> there's a guide on menus in the dev guide section
Nov 05 01:41:47 <clever> nice, i was looking for that and couldnt find it
Nov 05 01:41:54 <tdignan> sec
Nov 05 01:42:08 <tdignan> http://developer.android.com/guide/topics/ui/menus.html
Nov 05 01:42:11 <clever> thanks
Nov 05 01:42:24 <xroberx> tdignan: I really prefer to have two versions, a free and a paid one. Basically because the free version usually has less features and users are not likely to rate it as high as the full version
Nov 05 01:42:25 <tdignan> i don't usually navigate sites. I just google for every page I need
Nov 05 01:42:54 <tdignan> xroberx: that's a good point. I agree free users tend to let their one stars flow freely
Nov 05 01:43:12 <xroberx> tdignan: also if you put ads in the free versions, some users won't like the ads and will rate your app poorly
Nov 05 01:43:23 <tdignan> i've decided on no ads for this already
Nov 05 01:43:35 <tdignan> I'd only do an ad if I intended not to make a paid version
Nov 05 01:43:54 <tdignan> the free version is an ad... for me, and the premium version :D
Nov 05 01:44:06 <xroberx> tdignan: I used to think like you :)
Nov 05 01:44:16 <tdignan> you put ads now?
Nov 05 01:44:17 <tdignan> haha
Nov 05 01:44:35 <xroberx> tdignan: yeah, because most users are cheap and are not willing to pay
Nov 05 01:52:17 <clever> tdignan: testing...
Nov 05 01:53:09 <clever> tdignan: menu works, and i forgot to set the checkbox to the current val
Nov 05 01:56:31 <clever> tdignan: yep, all works in 4.x
Nov 05 01:57:52 <tdignan> yeah, its forward compatible
Nov 05 01:58:00 <tdignan> the actionbar just picks up the menu stuff
Nov 05 02:12:02 <clever> tdignan: is there an intent i can create, to let the user add a google account to the device?
Nov 05 02:12:21 <clever> tdignan: hmmm, actualy, do you need a google account linked to buy an app?
Nov 05 02:13:50 <tdignan> 1. i dont know 2. yes
Nov 05 02:14:02 <tdignan> unless you roll your own app store :)
Nov 05 02:14:10 <tdignan> or use a third party one
Nov 05 02:14:20 <clever> i'm selling the app on the google play store
Nov 05 02:14:31 <clever> so #1 will never happen
Nov 05 02:15:52 <clever> tdignan: do you know about gridview_v7 support?
Nov 05 02:19:00 <tdignan> gridview was added in api level 1?
Nov 05 02:19:05 <tdignan> what is gridview_v6
Nov 05 02:19:09 <tdignan> v7*
Nov 05 02:19:25 <tdignan> I've used a gridview before. That's about the extent of it.
Nov 05 02:19:31 * clever looks
Nov 05 02:19:40 <clever> tdignan: ah gridlayout, not view
Nov 05 02:19:44 <clever> keep getting them mixed up
Nov 05 02:19:45 <tdignan> oh ok, that makes sense
Nov 05 02:19:50 <tdignan> yeah I heard gridlayout is good
Nov 05 02:19:53 <clever> eclipse added it automaticaly
Nov 05 02:19:58 <clever> now i cant build under ant
Nov 05 02:20:05 <clever> and it messed up all the other dev systems
Nov 05 02:20:07 <tdignan> oh, whats your error?
Nov 05 02:20:21 <clever> /home/clever/android-sdk-linux/tools/ant/build.xml:569: Invalid file: /home/clever/workspace-android/gridlayout_v7/build.xml
Nov 05 02:20:39 <tdignan> I'll try creating a project like this,
Nov 05 02:20:59 <clever> it also put the support lib in my workspace, outside of the projects git repo
Nov 05 02:21:10 <clever> so all other dev systems are broken, referencing a dir that doesnt exist
Nov 05 02:21:35 <tdignan> how many dev systems? is this at a company
Nov 05 02:21:42 <clever> just 3, 2 of them mine
Nov 05 02:21:54 <clever> ive been fixing it by manualy copying it to the same path on each
Nov 05 02:22:20 <tdignan> you might consider just doing: android create project -t<#> -p foo --package com.clever.app
Nov 05 02:22:27 <tdignan> then import that instead
Nov 05 02:22:38 <tdignan> see android list targets for -t
Nov 05 02:23:12 <clever> tdignan: in my project dir or the gridlayout_v7 dir?
Nov 05 02:23:32 <tdignan> you can probably fix your existing project if you cd there and do android update project -t<target> -p .
Nov 05 02:23:34 <clever> hmmm, build.xml is missing in the gridlayout dir
Nov 05 02:23:44 <tdignan> is the gridlayout dir a library project?
Nov 05 02:23:48 <clever> yeah
Nov 05 02:23:52 <tdignan> get in there and do android update lib-project -p . -t<target>
Nov 05 02:24:37 <clever> yep, that did it
Nov 05 02:26:42 <clever> tdignan: so i'm guessing, that eclipse bypasses build.xml?
Nov 05 02:31:32 <tdignan> eclipse does not build with ant
Nov 05 02:31:36 <tdignan> it has its own build system
Nov 05 02:31:53 <tdignan> I mean, you can build with ant in eclipse.
Nov 05 02:32:06 <tdignan> But that's not what it's typically doing when you're developing for Android the de-facto way.
Nov 05 02:32:32 <clever> ah
Nov 05 02:32:40 <clever> so that explains why it worked with the build.xml missing
Nov 05 02:34:21 <tdignan> maybe it does use ant internally. I don't think it does, though.
Nov 05 02:35:07 <clever> tdignan: i have had issues changing the build target, eclipse doesnt notice and keeps spewing errors
Nov 05 02:35:12 <clever> have had to restart it
Nov 05 02:35:42 <tdignan> yeah, if you have that issue just change it with android update project
Nov 05 02:35:52 <tdignan> then highlight the project and click f5
Nov 05 02:36:14 <tdignan> also now eclipse has that integrated into the right click menu
Nov 05 02:36:40 <tdignan> right click, android tools, fix project.properties
Nov 05 02:36:52 <clever> ah, nice
Nov 05 02:46:27 <Syzygy> do I need to manage the lifecycle of a splash screen?
Nov 05 02:47:11 <Syzygy> or the main menu
Nov 05 02:48:02 <tdignan> the easiest way to make splash screen is to stick an imageview in an activity and then use view.postDelayed() to fire off your main activity after a 1-2 sec delay
Nov 05 02:48:05 <tdignan> that's all you need to do
Nov 05 02:48:31 <tdignan> you will want to use the no history flag to start the next one
Nov 05 02:48:44 <tdignan> so that when you hit the back button, you don't go back into the splash screen
Nov 05 02:49:08 <canadiancow> dont use a splashscreen
Nov 05 02:49:11 <canadiancow> those are the worst
Nov 05 02:49:35 <Syzygy> canadiancow, i'm doing an university course that requires me to use a splash screen for this game
Nov 05 02:49:49 <canadiancow> the worst
Nov 05 02:50:21 <Syzygy> tdignan, I've created a short video that should act as a splash screen and I've also created an asynchronousTask
Nov 05 02:50:29 <Syzygy> so that I can load things in the background
Nov 05 02:50:36 <Syzygy> although I'm not sure what things
Nov 05 02:50:45 <Leeds> Syzygy: then it's time to tell your teacher that you have spoken to people with professional app development experience who have told you not to do that
Nov 05 02:50:49 <tdignan> a video in a splash screen isn't a good idea
Nov 05 02:50:56 <tdignan> I would neve rely so much on video in android
Nov 05 02:51:01 <tdignan> never*
Nov 05 02:51:11 <Syzygy> I've been told (here) it's a good idea
Nov 05 02:51:14 <tdignan> wrong!
Nov 05 02:51:23 <tdignan> well, have fun appeasing your masters
Nov 05 02:51:29 <tdignan> us free people will keep on livin' free
Nov 05 02:51:38 <Syzygy> but the video is so cool https://dl.dropbox.com/u/31690810/uni/Multimedia/Splashscreen.mp4 D:
Nov 05 02:51:49 <tdignan> video should be optional
Nov 05 02:51:57 <canadiancow> no one intelligent told you a splash screen was a good idea
Nov 05 02:52:23 <Syzygy> I wasn't told the splash screen is a good idea, just that doing it as a video is a good idea
Nov 05 02:52:51 <tdignan> when not dealing in objective facts, you will find diverse opinions
Nov 05 02:52:56 <Syzygy> also, god damn it, i got some chili form my pizza into my open wound on my hand
Nov 05 02:53:17 <tdignan> could be worse. never touch your junk after eating jalapenos
Nov 05 02:53:49 <canadiancow> +1
Nov 05 02:54:04 <Syzygy> so what would you recommend as a splash screen? not ignoring the fact that a splashscreen is reqired
Nov 05 02:54:35 <canadiancow> a textview
Nov 05 02:54:45 <Syzygy> is that animated?
Nov 05 02:54:54 <canadiancow> with the text: "This annoying screen is here because I like annoying users. And because I'm not sure you know what app this is."
Nov 05 02:55:24 <Syzygy> I feel insulted
Nov 05 02:55:52 <Syzygy> i'll stick to videos for now
Nov 05 02:55:58 <tdignan> better to hear it now than to get it in your reviews. i'm pretty sure canadiancow uses google play
Nov 05 02:56:17 <canadiancow> :D
Nov 05 02:56:27 <Leeds> tdignan: note where this is a university project, apparently not something anyone is expected to ever actually run
Nov 05 02:56:33 <tdignan> 'This app is blue. I like green. One-star"
Nov 05 02:56:45 <canadiancow> shouldnt university teach you how to do things intelligently?
Nov 05 02:56:49 <tdignan> Leeds: most apps on the market wouldn't even get a D as a uni project.
Nov 05 02:56:56 <tdignan> he should publish it if it gets anything over C
Nov 05 02:57:09 <Syzygy> Leeds, we plan to refine it after the course is done and then publish it.
Nov 05 02:57:10 <Leeds> tdignan: and most uni projects wouldn't even get 2.5* in the market... goes both ways :)
Nov 05 02:57:29 <Leeds> Syzygy: then start now by refining away the splash
Nov 05 02:57:52 <Syzygy> i do not plan to get bad marks
Nov 05 02:58:00 <Syzygy> plus splash screens for games are somewhat common
Nov 05 02:58:06 <tdignan> yes, they are
Nov 05 02:58:25 <Syzygy> blame that on the industry, but I'm still doing it as well
Nov 05 02:59:28 <canadiancow> usps charges shipping to order stamps... sigh
Nov 05 02:59:42 <canadiancow> of all the places i expect not to get gouged on shipping
Nov 05 02:59:51 <regedit> how can i have an on/off switch in my notification, much like the WiFi on/off switch in the notification bar/drawer
**** ENDING LOGGING AT Mon Nov 05 02:59:59 2012