**** BEGIN LOGGING AT Wed Jan 07 02:59:57 2009 Jan 07 03:40:11 savid: that's the point of recycled views Jan 07 03:41:06 romainguy, ok, so if I understand correctly, the same identical instance of a view object can be used on multiple rows? Jan 07 03:41:11 no no Jan 07 03:41:24 ok, because that's kindof what it looks like in my debug log :-P Jan 07 03:41:26 but you cannot rely on the id for that Jan 07 03:41:29 oh ok Jan 07 03:41:33 don't trust your debug log, it works :p Jan 07 03:41:38 hehe ok Jan 07 03:43:31 I'm just having issues w/ my viewWrapper pattern... for some reason the first row gets the last object's properties, and the rest of 'em are blank :-P I just need to do some more debugging. I'm sure I messed up somewhere Jan 07 03:44:17 (fyi, I was following thus tutorial: http://androidguys.com/?p=636) Jan 07 03:45:29 Is there any way to actually get a unique id on the view object to make sure it actually _is_ a unique instance? Jan 07 03:45:54 (not that I don't think it is, I want to make sure something else I'm doing isn't overwriting the object somehow) Jan 07 03:46:43 It's wierd though, when the second item loads, convertView is definitely not null... Jan 07 05:13:46 savid: you can always add your own "serial number" type field to the class and include it in a custom toString() Jan 07 05:14:19 also, gnight Jan 07 05:15:46 danfuzz, thought of that, but not sure how I can do that to the base view class... Jan 07 05:17:21 not sure if I could extend the class either... I'm using inflate to create the view Jan 07 07:30:54 hey jeffrey Jan 07 14:26:00 hi guys quick question theoretically android could be ported to none arm architecture? Jan 07 14:32:11 it could be ported to any architecture supported by linux (which is a subset of those supported by gcc) Jan 07 14:32:20 chris_99: yes Jan 07 14:32:26 i'd already been ported to x86 Jan 07 14:32:30 (Asus EEE) Jan 07 14:32:59 in theory, there should be nothing to do if the kernel already has the right device drivers Jan 07 14:33:04 just cross-compile for the right architecture Jan 07 14:33:40 some bits of the android userspace may freak out ("omg no radio interface, uh what the heck is this screen", etc) Jan 07 14:33:56 digitalspagetti - awsome didn't realise that. danderson - cheers yeah that makes sense Jan 07 15:21:39 Can you create and access a sqlite db on the flash drive, accessible as a db by an android app? Jan 07 15:22:01 no Jan 07 15:22:23 michaelnovakjr: okay. thanks. dang it. Jan 07 15:22:30 no problem :) Jan 07 15:30:40 BlindOracle: yes Jan 07 15:30:43 you can Jan 07 15:31:12 might not be a great idea, but it is(was in M5 at least) possible Jan 07 15:32:06 how far along is cupcake now days??? Jan 07 15:35:05 BlindOracle: see SQLiteDatabase.openDatabase()...you can pass a String path or a File object and like to /sdcard Jan 07 15:37:09 zhobbs_ i'm not sure that works on a device Jan 07 15:37:20 i know it was attempted but it didn't work for some reason Jan 07 15:37:32 i can't remember exactly :) Jan 07 15:38:05 michaelnovakjr: hmm, interesting... Jan 07 15:38:25 I did it on the emulator, but never on /sdcard, but in my app's files folder Jan 07 15:38:26 its a pretty simple test actually Jan 07 15:42:34 Hmm.. I'm trying to run "adb devices" to see my connected device but it's not showing up. my lsusb shows that it's connected though. Any ideas on why this might be happening? Jan 07 15:42:46 on linux? Jan 07 15:42:55 adb kill-server; sleep 1; sudo adb start-server Jan 07 15:43:11 savid: http://code.google.com/android/intro/develop-and-debug.html#developingondevicehardware Jan 07 15:43:57 zhobbs_, ah, I think that's what I'm looking for. thanks :) Jan 07 15:44:38 I'm running ubuntu, so I probably need the udev rule Jan 07 15:52:07 Is there a way to generate an Ant buiild.xml from a project that was built using eclipse? Jan 07 15:52:44 (ie, so I can just build from the cmdline) Jan 07 15:53:35 savid: use the activitycreator to create a project in a different dir. Take that build.xml file and move it to your directory. Jan 07 15:54:09 BlindOracle, thanks, I'll try that. Jan 07 15:54:40 michaelnovakjr: does the version sqlite used in android support IF NOT EXIST for table creation? Jan 07 15:57:06 BlindOracle: I seem to remember that it does, have you looked at SQLiteOpenHelper though? Helps you with creation and versioning of a sqlite database Jan 07 15:57:36 marklar: I'm using it. It seems to always try to create my tables....so I'm trying to use IF NOT EXIST Jan 07 15:57:59 marklar: I'm using the notepad as an example Jan 07 15:58:10 marklar: difference is, I have three tables to create Jan 07 15:58:29 hmm, strange. I use one as well in my program and it only calls onCreate() it if the database doesn't exist Jan 07 15:58:51 marklar: okay..I'll double check...maybe I'm not doing something right Jan 07 15:59:15 if you'd like, in a few minutes I could post up the code that I am using for my database Jan 07 16:01:06 well now that I look at it, I'm unsure exactly how it is being done because my table creation strings do have "IF NOT EXIST" in them Jan 07 16:06:47 marklar: something very odd is going on. I'm testing on emulator first. my app is not running. The database file does not exist. I instantiate my helper and call open. Open tries to create the db and tables. Every table creation fails stating the table already exists when it does not exist. table creation fails. using sqlite3 in emulator shows tables do NOT exist. Jan 07 16:07:11 marklar: if I use "IF NOT EXIST", it pikes saying I can't use "IF NOT EXIST". Jan 07 16:07:15 wtf?! Jan 07 16:07:21 lol wow Jan 07 16:07:29 here let me post the code that I am using Jan 07 16:08:33 BlindOracle: is this on the sdcard? Jan 07 16:08:47 michaelnovakjr: no, on phone Jan 07 16:09:04 and SQL works using sqlite3 (and other tools) Jan 07 16:09:09 SQL = DDL Jan 07 16:09:18 BlindOracle: you are sure the database file exists at /data/data//databases/ ? Jan 07 16:09:30 marklar: yes Jan 07 16:09:38 it does create the db...just tables fail Jan 07 16:09:51 that's how I use sqlite3 on phone to see if tables exist Jan 07 16:09:56 pastebin your db class Jan 07 16:11:54 * BlindOracle testing some stuff first....and restarting emulator in case it got in bad state Jan 07 16:15:22 BlindOracle: http://www.pastebin.ca/1302813 this is a portion of my database class Jan 07 16:15:42 Its pretty long though so I took out the stuff that I don't think would relate to your problem Jan 07 16:16:35 marklar: looking Jan 07 16:18:59 anyone know (jasta maybe :) ) how to diagnose the error numbers in LogCat from MediaPlayer? I'm seeing things like: ERROR/MediaPlayer(652): Error (-17,0) Jan 07 16:20:25 what kind of pattern does SimpleDateFormat expect? Jan 07 16:20:28 mediaplayer stuff is included in the source, find . -iname 'MediaPlayer*' | xargs grep -i "Error (" I guess? Jan 07 16:20:34 i'm not really getting the doc Jan 07 16:20:44 eldenz: there's numerous patterns Jan 07 16:21:36 http://java.sun.com/j2se/1.5.0/docs/api/ Jan 07 16:22:27 vol: good idea, thanks Jan 07 16:22:29 ah i'm stupid... didn't realize that's a java native class Jan 07 16:22:30 thanks Jan 07 16:22:38 eldenz: SimpleDateFormat formatter = new SimpleDateFormat("MM-dd-yyyy"); Jan 07 16:23:42 great, thanks michaelnovakjr Jan 07 16:24:22 no problem Jan 07 16:34:43 trying to compile cupcakeand I get an error like this guy: http://groups.google.com/group/android-platform/msg/db57c7f1c506214b Jan 07 16:34:45 any ideas? Jan 07 16:35:09 javac is 1.6.0_10 Jan 07 16:36:13 famast: that looks like C errors Jan 07 16:36:25 BlindOracle, hmm, it seems that ant can't find my sdk path. I'm getting this error: build.xml:129: Execute failed: java.io.IOException: Cannot run program "${sdk-folder}/tools/aapt": java.io.IOException: error=2, No such file or directory Jan 07 16:36:34 actually its C++ look at the filenames Jan 07 16:37:13 famast: looks like you are missing some header files for C++ dev Jan 07 16:37:22 or it can't find them Jan 07 16:37:29 famast: what compiler? Jan 07 16:37:30 michaelnovakjr: Yea apparently it is. Maybe my checkout of cupcake was incorrect? Jan 07 16:37:44 famast: they would be stored on your system Jan 07 16:37:52 not in cupcake Jan 07 16:37:59 zinx: dunno? gcc? Jan 07 16:38:12 famast: what os? Jan 07 16:38:46 michaelnovakjr: ubuntu 8.10 Jan 07 16:38:56 ah, looks like I need the default.properties file :-p Jan 07 16:39:00 famast: install build-essential Jan 07 16:39:02 famast: install the libstdc++ and its -dev packages Jan 07 16:39:32 michaelnovakjr & zinx: ok thanks let me try that Jan 07 16:39:47 its weird because I've compiled an older branch before Jan 07 16:39:57 but that was a while back Jan 07 16:42:55 hm, shouldn't View#getMeasuredHeight() and View#getHeight() be different if there is a scroller? Jan 07 16:43:29 or how do i get the height the whole content is taking (not the visible area) Jan 07 16:44:42 eldenz: I would assume that property does give you the entire visible area. I don't know the answer to your question but you could try using the hiearachyViewer and see if any of the view's properties reflect the answer your looking for Jan 07 16:44:48 It looks like my build.xml is built for windows?? it's pointing ${apk-builder} to a .bat file, and all the os_family properties are set to "windows". I'm on linux :-P Jan 07 16:45:34 savid: Wrong sdk? Jan 07 16:45:39 savid, you can (re)create the build.xml on your specific platform i think Jan 07 16:46:09 eldenz, I created the build.xml in linux Jan 07 16:46:19 famast, possibly, yeah :-p I guess I'll re-download Jan 07 16:46:38 famast, ah cool, thanks, didn't know the hierarchyviewer :) Jan 07 16:46:58 yea the hiearchyviewer has saved me so many times Jan 07 16:52:28 i don't see anything i need though ;< Jan 07 16:52:44 Were you able to see a list of properties your view has? Jan 07 16:52:52 yep Jan 07 16:52:54 hmm Jan 07 16:53:01 but it has always the visible height Jan 07 16:54:17 michaelnovakjr: marklar: Thanks guys, I figured out. I had two bugs. One of which was I broke the DDL when I had gone back and forth trying to find the first bug. Jan 07 16:54:40 eldenz: this isn't a listview right? Jan 07 16:55:05 famast, no, it's a TextView that has ScrollingMovementMethod Jan 07 16:55:15 i probably need to get hand on the scroller to retrieve its height Jan 07 16:55:18 no luck with the cupcake compile even after verifying that I had those libraries Jan 07 16:55:24 (i'm trying to add autoscroll) Jan 07 16:56:10 eldenz: Yea I'm out of ideas. If the hieararchyViewer doesn't know, maybe no one does! Jan 07 16:56:19 : ) Jan 07 17:03:23 ok i think an answer to my compiling question can be found here (incase anyone is interested) http://groups.google.com/group/android-platform/browse_thread/thread/41877abca5b2d66a Jan 07 17:03:35 still boggles my mind how these things slip through Jan 07 17:04:55 BlindOracle: good to hear Jan 07 17:22:10 how is that menu called that pops up in many apps if you hit 'More'? Jan 07 17:56:25 Oh hey cool Jan 07 17:56:53 W 32 AudioTrack *** SERIOUS WARNING *** obtainBuffer() timed out but didn't need to be locked. We recovered, but this shouldn't happen (user=00000100, server=00000002) Jan 07 17:57:01 shit just got real? Jan 07 18:35:43 I'm trying to catch when a user pressed the cancel button on a DatePickerDialog, does anyone know how I would do that? Jan 07 18:36:09 the closest I've come is using an onDismissListener, but that catches everything, not just the cancel button Jan 07 18:36:42 OnCancelListener apparently only listens for the BACK button? Jan 07 18:37:02 it implements onClickListener() doesn't it? Jan 07 18:37:14 err, without the () Jan 07 18:37:49 I tried overriding the onclick(), unfortunately I can't remember why it didn't work and now I'm not at home Jan 07 18:38:06 Would that be the recommended way though? Jan 07 18:38:25 so its onClick() method will be called telling it which button was clicked Jan 07 18:38:27 yes Jan 07 18:38:36 ok, I'll look at that again, thanks Jan 07 18:38:55 what you do is instantiate a subclass of DatePickerDialog that overrides onClick() to do what you need Jan 07 18:39:41 yes, I tried that, and was tried to listen for all three buttons... I wish I had the code infront of me Jan 07 18:39:58 Thanks though, I'll take a look at it again Jan 07 18:40:29 see DateWidgets1.java in ApiDemos for examples Jan 07 18:40:48 k Jan 07 18:41:17 oh hm, that doesn't look specifically for cancel :) Jan 07 18:42:05 (why do you need to know whether it was cancelled, btw?) Jan 07 18:42:32 Trying to get a range of dates from users, so there are two of them in a row Jan 07 18:42:52 unfortunately, if they cancel the second, next time they try to enter a range it takes the first entered as the second date Jan 07 18:42:58 if that makes sense :) Jan 07 18:43:21 ah, i see Jan 07 18:43:28 you really need your own widget :) Jan 07 18:43:49 haha yeah, trying to keep it easy Jan 07 18:44:02 you mean view right? I thought widgets couldn't be made atm Jan 07 18:44:21 i actually mean your own TwoDatePickerDialog Jan 07 18:44:26 not 'widget' in the technical sense Jan 07 18:44:32 yeah, thats what I figured Jan 07 18:44:40 it actually works pretty well though Jan 07 18:44:46 anyway, the basic thing you probably want to do looks like this: Jan 07 18:45:56 mMyDatePicker = new DatePickerDialog([args here]) { public void onClick(DialogInterface dialog, int which) { super.onClick(dialog, which); /* now put your code to deal with 'cancel' being hit here */ } } Jan 07 18:46:04 instead of just instantiating one straight Jan 07 18:46:42 yup, thats what I was trying to do, since I was overriding the onDateSet() as well... it was probably some small error on my part Jan 07 18:46:52 you have to call the base DatePickerDialog.onClick() because that's where onDateSet() is dispatched Jan 07 18:48:14 ah, that may have been it Jan 07 18:49:18 yeah, now that I think about it I don't think I was calling super.onClick() Jan 07 18:49:19 Thanks! Jan 07 18:49:26 but of course Jan 07 20:04:05 hello ppl Jan 07 20:04:18 hello Jan 07 20:04:44 i just installed the sdk and i am trying to get started on dev a app but i dont know java i need some help Jan 07 20:04:59 jasta, whats up Jan 07 20:05:21 jasta, can you help mme Jan 07 20:05:24 me* Jan 07 20:05:35 actually, the tutorials and materials online can help you. we are not here to teach anyone Java. Jan 07 20:06:02 :) Jan 07 20:11:22 jasta, I know thats how i got as far as i got Jan 07 20:12:00 but i need some one to guid me through somme thing ill even give credit on the app Jan 07 20:12:25 you do not need anyone to guide you. just read materials online. start with the java tutorial on sun's page. Jan 07 20:12:54 im stuck on hello world Jan 07 20:13:12 *ugh*. Jan 07 20:13:29 i know i know Jan 07 20:13:30 if it's an android-specific question, please just ask. Jan 07 20:14:41 what is the script to show a site with downloading? Jan 07 20:14:43 :) Jan 07 20:16:04 gibberish, try again. Jan 07 20:26:01 neat, the MediaPlayer documentation seems to have received significant attention in cupcake Jan 07 20:26:11 thank god Jan 07 20:26:48 are you running cupcake on the g1? Jan 07 20:26:51 I need help. Launching my emulator with a build i did this morning on mainline is giving me a the window but no keyboard or buttons Jan 07 20:28:00 it seems to run OK. I can see the desktop and everything, but the rest of the emulator is missing Jan 07 20:29:17 jasta, how do i embed a website for the android Jan 07 20:30:18 or anybody Jan 07 20:30:32 akuma55: Use a webview? Jan 07 20:30:39 ask the SpeedyMarks author^^ Jan 07 20:42:38 famast, do i download it Jan 07 20:42:41 ? Jan 07 20:44:21 android for this device please :) http://ces.cnet.com/8301-19167_1-10134418-100.html Jan 07 20:44:41 hah Jan 07 20:45:05 Android: Dick Tracey Edition :) Jan 07 20:45:13 one icon per menu ;p Jan 07 20:45:22 http://www.engadget.com/2009/01/06/pharos-rolls-out-traveler-137-coming-this-quarter-for-599-95/ <- also needs some android luvin Jan 07 20:47:02 Anyone know when the next version of the SDK will be out? Jan 07 20:47:22 probab;y when cupcake is released Jan 07 20:47:46 Any idea when that will be? :D Jan 07 20:49:16 how long is a piece of string? Jan 07 20:57:41 what's the difference between a chicken? Jan 07 20:57:58 the size of its cage Jan 07 20:58:48 anyone know if schema creation in sqlite honors transactional boundaries? Jan 07 20:59:43 now this is the killer app for the G1: http://www.thestandard.com/news/2009/01/07/wireless-electricity-coming-ces-dont-expect-much-mobility Jan 07 20:59:54 "If your cellphone is recharging on your desk all day, you won't be thinking about it." Jan 07 21:00:44 "Also, you'll have such a large tumor in your brain you won't be thinking about much of anything." Jan 07 21:01:24 * jasta curses his slow work computer Jan 07 21:01:32 building android is taking forever :) Jan 07 21:01:38 LOS power.. really inefficient too unless they have some newfangled infrared Jan 07 21:01:45 erh, newfangled infrared solar cells Jan 07 21:05:26 jasta, 64-bit and a lot of memory helps :) Jan 07 21:05:37 quadcore too :) Jan 07 21:07:21 marklar: Are you creating multiple tables in your db helper's onCreate? Jan 07 21:08:07 michaelnovakjr: anyone know if schema creation in sqlite honors transactional boundaries? Jan 07 21:09:49 BlindOracle: in what sense? Jan 07 21:10:45 what's your issue? Jan 07 21:18:53 Why does activitycreator put when I'm on linux? Jan 07 21:19:09 I'm using the linux sdk Jan 07 21:19:18 it makes a build.xml which supports both windows and linux Jan 07 21:19:36 help Jan 07 21:19:40 Description Resource Path Location Type Jan 07 21:19:40 ERROR Error: String types not allowed (at 'description' with value 'akuma55 zone'). AndroidManifest.xml akuma55G1 line 14 aapt Problem Jan 07 21:19:48 what does this mean? Jan 07 21:19:59 it means description="akuma55 zone" is gibberish. Jan 07 21:20:41 jasta, oh ok. I think I'm still running into an issue though. When building I get this error: BUILD FAILED: build.xml:230: exec returned: 1 Jan 07 21:20:58 can you show me build.xml:230? Jan 07 21:21:13 jasta, I looked on line 230 : Jan 07 21:21:24 wait Jan 07 21:21:28 that's not 230 Jan 07 21:21:28 one sec Jan 07 21:21:39 Jan 07 21:21:44 ^ 230 Jan 07 21:21:52 what's ${apk-builder}? Jan 07 21:22:07 line 111: Jan 07 21:23:02 I looked and "apkbuilder" does exist in the tools dir Jan 07 21:23:10 i'd never heard of it... Jan 07 21:23:22 is that a windows thing? Jan 07 21:23:22 but it does exist (i just checked) Jan 07 21:23:34 if you are missing apkbuilder, then thats the issue :) Jan 07 21:23:44 nope, it's definitely there Jan 07 21:23:50 in tools/ Jan 07 21:23:55 oh i misread that as "doesn't exist" Jan 07 21:24:08 run ant with debug so you can see what exactly its executing Jan 07 21:29:49 hmm.. I ran the cmd it says it's executing, and all it does is return 1 Jan 07 21:32:13 but what does ant -d say about it? Jan 07 21:32:21 it will tell you the full command line so you can investigate Jan 07 21:33:43 Yeah, It showed the full command and I executed it myself, and all it does is say: "Using keystore: /home/ben/.android/debug.keystore", then "null" Jan 07 21:33:52 $? is set to 1 Jan 07 21:34:03 apkbuilder is a shell script Jan 07 21:38:04 then run sh -x Jan 07 21:38:11 see where it fails :) Jan 07 21:39:02 ah, good idea :) Jan 07 21:42:44 michaelnovakjr: not really an issue, I just want to better understand how things work. I know not all SQL engines honor transaction when it comes to DDL execution. This got me wondering, mostly because of this morning's bugs. Specifically, if sqlite does not honor transactions with regard toward schema generation then the db helper class provided with Android is fundamentally broken. I'm not saying it is at this point, but I'd like to Jan 07 21:42:44 know for sure. Jan 07 21:44:48 jasta, I think it's because I have a space in my project name :-p Jan 07 21:45:25 At the end of the APK builder script, it runs a java -jar command and doesn't enclose all the path arguments in quotes :-p Jan 07 21:46:32 actually, it's whatever is calling apkbuilder that's not quoting the arguments Jan 07 21:48:17 BlindOracle: all operations other than select are run through transactions Jan 07 21:49:07 michaelnovakjr: in other words, if you create multiple tables and table creation fails on any table other than the first one, and all created tables are not rolled back, then db helper is broken because it will not create the tables as it now thinks they already exist despite the fact they don't. Same logic can be applied to the onUpgrade method too. Jan 07 21:49:20 michaelnovakjr: ok...good to know. thanks Jan 07 21:50:20 savid: well you should of course be murdered for putting spaces in your project name Jan 07 21:50:26 :) Jan 07 21:50:35 get a rope... Jan 07 21:50:39 ;) Jan 07 21:50:43 * savid runs and hides Jan 07 21:50:58 wait what? spaces? Jan 07 21:51:04 why would anyone do that? Jan 07 21:51:26 sheesh Jan 07 21:51:40 :) Jan 07 21:51:41 hehe... Jan 07 21:52:55 michaelnovakjr: so if I need to create multiple tables in db helper's onCreate, it should be done with a single call using multiple statements? Jan 07 21:53:20 should be Jan 07 21:54:37 michaelnovakjr: that's multiple SQL statements via single call to execSQL...just to clear/sure Jan 07 21:55:08 actually, if _any_ part of the path to my project had spaces it would fail :-p Jan 07 21:56:20 spaces in dir/file names is bad... Jan 07 21:58:04 sure, but not disallowed... Jan 07 21:58:33 right, so there's a bug in apkbuilder, but you still should change your project name :) Jan 07 21:59:59 savid: not disallowed only because you've not seen the type of knot we're tying in this here rope... ;) Jan 07 22:00:17 * savid runs and hides again Jan 07 22:03:54 BlindOracle: it seems android uses transactions and it the transaction returns unsuccessful then it runs a rollback Jan 07 22:04:52 so if a table is created successfully and then the next table fails it should also rollback the successful table Jan 07 22:05:27 but that depends on your code also Jan 07 22:07:03 if that makes any sense :) Jan 07 22:07:12 michaelnovakjr: on my code also? How so? execSQL is executed within a transaction boundary. What else is required? Failed transactions should automatically be rolled back? No? Jan 07 22:07:34 BlindOracle: if you span your table creation over multiple execSQL calls Jan 07 22:07:58 in that case each would be their own transactions Jan 07 22:08:40 michaelnovakjr: right. That's the point I was making. If coder spans multiple execSQL then they are potentially breaking the helper class. I'm creating an enhancement request to make sure this gets documented/captured else someone be surprised in future. Jan 07 22:09:47 michaelnovakjr: and for what it's worth, it does not appear sqlite allows embedded transactions. so failure means manually dropping tables...not really a roll back. Jan 07 22:10:07 it rolls back the transaction Jan 07 22:10:59 Ok, so I did away w/ the spaces in my path, and the java cmd that apkbuilder is running is outputting "null" and exiting w/ 1. http://pastebin.com/m33e47588 Jan 07 22:11:11 michaelnovakjr: but if you have multiple execs, each executing DDL, the coder must then capture the exception and manually drop the tables which did succeed. the previously created tables will not be rolled back as the transaction containing said DDL completed. Jan 07 22:11:26 yes Jan 07 22:12:01 ok...we're on the same page. You're 100% DDL honors transactions in sqlite?! I looked but I've not found anything which assures that. Jan 07 22:13:24 interesting.... when you call getReadbleDatabase() it actually attempts to return a getWritableDatabase first :) Jan 07 22:13:49 wow...that's odd. Jan 07 22:16:45 michaelnovakjr: I take it you're browsing source to confirm transaction support for DDL - which is how you noticed it returns a writable DB? Jan 07 22:16:52 yes Jan 07 22:17:07 ah nm, works now. I just nuked the bin/ dir and re-ran ant Jan 07 22:18:11 Where does ant store the .apk file? I don't see an apk in my project dir anywhere Jan 07 22:18:22 bin/ Jan 07 22:18:35 I reported http://code.google.com/p/android/issues/detail?id=1732 Jan 07 22:18:36 hmm. I see "MyProject.ap_" Jan 07 22:23:32 I NEED HELP I WILL PAY Jan 07 22:23:41 you need to go away Jan 07 22:24:03 haha Jan 07 22:28:40 michaelnovakjr: if you find DDL is not honored for transactions, please let me know. Jan 07 22:31:47 I can trade knowledge if thats not goood enough Jan 07 22:32:32 akuma5, did you not see my priv msg? Just ask your question and be patient Jan 07 22:33:30 akuma5: get lost Jan 07 22:34:00 will you trade sex? Jan 07 22:34:11 sorry savid Jan 07 22:34:17 I will fill you with knowledge the only way I know how Jan 07 22:34:26 * savid runs and hides Jan 07 22:34:47 thank for the generous off michaelnovakjr_ Jan 07 22:34:55 offer8 Jan 07 22:35:01 offer* Jan 07 22:35:08 just ask your question akuma Jan 07 22:35:11 don't be a dong Jan 07 22:35:33 BlindOracle: ping Jan 07 22:36:58 i need to make a app for my webiste that allows ppl to listen to the audio stream and have some thing like webview so i can update stuff for ppl to see Jan 07 22:37:21 and you would like us to design it for you? Jan 07 22:37:57 akuma5: i would suggest starting with a few smaller projects first just to learn your way around java/android Jan 07 22:38:15 or go to a freelance site and not here :) Jan 07 22:38:16 i was look for help so i can get started but it look like noboody wonts to help so thats my onley choice Jan 07 22:38:17 The MediaPlayer has support for audio streams. It wouldn't be hard to whip something up. Jan 07 22:38:39 well, maybe if you asked for help instead of screaming "NED HLP PAY MONEYS PM ME $$$$4141q3rt50weu" we'd be friendlier Jan 07 22:38:57 i did about a hour or two ago Jan 07 22:40:00 actually, you didn't. you've acted helpless and whiny since you got here :) Jan 07 22:40:32 oh sorry Jan 07 22:41:04 you didn't ask a specific question Jan 07 22:41:04 you said "help me" Jan 07 22:41:04 17:23 < akuma5> I NEED HELP I WILL PAY Jan 07 22:41:05 akuma5: have you looked at the ApiDemos? and the apps-for-android sources? Jan 07 22:41:05 they are great places to start Jan 07 22:41:13 now i see what it feel like for newbies on the offsec channel Jan 07 22:41:22 akuma5. For your view use a webview. For your music use MediaPlayer. Jan 07 22:41:56 but the thing about that is i dont know java Jan 07 22:42:06 akuma5: and that's the thing we find helpless and whiny. Jan 07 22:42:11 im playing with the sdk and im stuck Jan 07 22:42:19 dont know what to do next Jan 07 22:42:22 why are you all arguing? ;> Jan 07 22:42:25 what can you expect us to do for you? teach you our many years of experience on IRC? Jan 07 22:42:30 michaelnovakjr: I'm here Jan 07 22:42:31 akuma5, IRC != Support Jan 07 22:42:42 i know Jan 07 22:42:46 akuma5: Then your first step is to learn some basic java Jan 07 22:43:12 is there any good video tut Jan 07 22:43:18 for the android Jan 07 22:43:21 ? Jan 07 22:43:23 akuma5: you won't learn java by watching video tutorials. Jan 07 22:43:32 akuma5: No Jan 07 22:43:35 nor would you learn anything significant about android this way. Jan 07 22:43:35 BlindOracle: one sec Jan 07 22:43:43 reading is confuseing for programming Jan 07 22:43:52 michaelnovakjr: I've run into a problem. I thought it was another bug but it doesn't seem to be the case. It appears only the first statement is executed inside of execSQL calls. This means of the three tables I required, only the first gets created. Jan 07 22:43:52 wow. Jan 07 22:43:58 akuma5: The best thing to do is learn Java first. Jan 07 22:44:03 akuma5: expect that learning will be a significant investment of time, and no time spent here initially will be helpful. so leave, and go try something else. Jan 07 22:44:12 michaelnovakjr: should I be creating tables via a different call? Jan 07 22:44:30 akuma5, I believe some literacy is a prerequisite to programming Jan 07 22:44:43 akuma5: Learn java first. Jan 07 22:45:00 If you're completely new to programming, and you prefer visual feedback, you might want to try the Processing programming language Jan 07 22:45:01 BlindOracle: that is correct it wouldn't work Jan 07 22:45:03 :) that was my purp vol Jan 07 22:45:08 execSQL is doing only the first one Jan 07 22:45:09 it's built off of java, but is meant for artists and non-technical types Jan 07 22:45:19 so it is very easy to just make things appear on the screen Jan 07 22:45:22 akuma5: you are off topic and distracting.... stop Jan 07 22:45:25 akuma5: start here http://eclipsetutorial.sourceforge.net/totalbeginner.html Jan 07 22:45:47 michaelnovakjr: yes. Jan 07 22:45:49 vol: Cool my sister might like that. She keeps trying to read my java book but gets bored Jan 07 22:46:00 alternately, http://processing.org/ Jan 07 22:46:17 bear in mind that it's something of a limited language, and you're basically programming java in a limited environment Jan 07 22:46:21 thanks morgan Jan 07 22:46:29 once this whets your appetite for java, you might want to move on to real java tutorials Jan 07 22:46:49 im in to offsec thats easy for me Jan 07 22:46:56 and web dev Jan 07 22:46:59 michaelnovakjr: I'm using the following; create table if not exists "c" ( pk integer primary key autoincrement not null );create table if not exists "a" ( pk integer primary key autoincrement not null );create table if not exists "b" ( pk integer primary key autoincrement not null ); Jan 07 22:47:31 michaelnovakjr: it works from sqlite3. But if I execSQL it, only table "c" gets created. I can rotate the statements within and only the first table is created. Jan 07 22:49:47 michaelnovakjr: obviously I'm using different quoting. I've tried both with and without quotes - with quotes escaped - pragmatically that is. Jan 07 22:50:16 michaelnovakjr: This also gets the same result; "create table if not exists c ( pk integer primary key autoincrement not null );create table if not exists a ( pk integer primary key autoincrement not null );create table if not exists b ( pk integer primary key autoincrement not null );" Jan 07 22:50:47 interesting Jan 07 22:54:13 michaelnovakjr: I think that means I'm back to using multiple execSQLs to get 'er done Jan 07 22:54:27 yuck Jan 07 22:54:36 execSQL is a transaction, but it only execs one statement Jan 07 22:55:14 michaelnovakjr: ya, that's what I'm finding. any work arounds? Jan 07 22:56:11 checking Jan 07 22:58:14 what ever happen to dev with notpad Jan 07 22:58:32 or am i old school Jan 07 22:59:01 uh yea..... Jan 07 23:00:29 are you guys windows base Jan 07 23:00:33 michaelnovakjr: so how would you do multiple inserts as a single transaction? Jan 07 23:00:33 ? Jan 07 23:00:42 notepad? Jan 07 23:00:44 really? Jan 07 23:00:57 I mean, you can write things on slate with chalk, instead of paper and pencil, or keyboard and screen Jan 07 23:01:02 but the world has moved on Jan 07 23:01:06 I just found where someone was asking about this type thing. ...well, for inserts... Jan 07 23:01:28 that how i learn python Jan 07 23:01:35 basics Jan 07 23:01:39 if you're not using autocomplete or intellisense stuff, you're wasting cycles trying to figure out what you misspelled Jan 07 23:01:49 BlindOracle: hm, the query i ran works on the sqlite3 shell but not in the code Jan 07 23:01:54 BEGIN; CREATE TABLE test ( _id INTEGER not null, name text not null ); CREATE TABLE testme (_id integer not null, name text not null); COMMIT; Jan 07 23:02:40 will try Jan 07 23:02:54 i tried rawQuery too, didn't work Jan 07 23:02:59 execSQL that query will break Jan 07 23:03:03 I think that will fail because a transaction is already open via execSQL Jan 07 23:03:06 yes Jan 07 23:03:14 but rawQuery it doesn't fail.... Jan 07 23:03:15 ok Jan 07 23:03:24 AHhh!!! Will try that! Jan 07 23:03:26 uhm, some dolt put bash-isms in build/core/find-jdk-tools-jar.sh Jan 07 23:03:48 jasta: there can be only one... :P Jan 07 23:03:57 bash rocks :) Jan 07 23:03:58 sh rejects them. Jan 07 23:04:13 meaning it breaks regardless Jan 07 23:04:32 they wrote [ foo && bar ]. test requires foo -a bar Jan 07 23:04:34 unless you use bash as your build shell? Jan 07 23:05:02 also hum, what tree are you looking at? i see it using [[ instead of [ Jan 07 23:06:12 oh...rawQuery takes an array of strings Jan 07 23:10:56 (and yeah, [[ seems to be a bash intrinisic) Jan 07 23:13:56 michaelnovakjr: http://pastebin.com/d453ee3be Jan 07 23:14:45 michaelnovakjr: in both cases the db gets created but no tables do Jan 07 23:18:07 For highlighted version: http://pastebin.com/d70623aa6 Jan 07 23:19:48 michaelnovakjr: I'm thinking rawQuery isn't going to work. Jan 07 23:20:43 ctate: two brackets is just a style thing AFAIK, its not syntactically significant. Jan 07 23:20:49 ctate: anywya, i just committed the fix: Jan 07 23:20:58 http://review.source.android.com/6559 Jan 07 23:21:00 on my machine i have /bin/[ but no corresponding binary for [[ :) Jan 07 23:21:17 ctate: [ is just test, and nested tests won't do anything different functionally. Jan 07 23:21:28 there is a difference though Jan 07 23:21:41 [[ lets you use && and || Jan 07 23:21:48 not with dash it doesn't. Jan 07 23:21:49 whereas [ lets you use only -a etc. Jan 07 23:21:51 romainguy_: beat me to it Jan 07 23:21:54 as you just saw. what are you using as your build shell? Jan 07 23:22:01 aha Jan 07 23:22:16 at least with bash Jan 07 23:22:21 obviously the real answer is to use bash :) Jan 07 23:22:22 I don't know about other shells ^^ Jan 07 23:22:38 dash is Ubuntu's default for sh Jan 07 23:22:39 romainguy_: it's a bash thing. Jan 07 23:22:40 [but thanks for the patch to make it not require bash] Jan 07 23:22:44 and this is a bashism for sure. Jan 07 23:24:12 damnit, i was wrong. [[ is also unsupported by dahs. Jan 07 23:24:14 you need to make /bin/bash point to bash, not dash Jan 07 23:24:18 * jasta revisesammends Jan 07 23:24:35 the scripts that use [[ specify /bin/bash Jan 07 23:24:44 this one does not. Jan 07 23:24:52 * BlindOracle fears he ran michaelnovakjr off Jan 07 23:25:05 if it's called from the build system, the build system forces the shell to bash Jan 07 23:25:10 no it doesn't :) Jan 07 23:25:16 yes it does Jan 07 23:25:26 it doesn't for that script then. Jan 07 23:25:33 how do you think i noticed it? choosecombo fails. Jan 07 23:25:36 the first non-comment line in build/core/main.mk Jan 07 23:25:49 it's also an existing bug in the issue list Jan 07 23:26:10 likely a pretty common problem considering this is Ubuntu's default. Jan 07 23:30:09 wait, ubuntu ships with /bin/bash being a link to a non-bash-compatible shell? LAME Jan 07 23:31:01 no, /bin/sh being linked to dash. Jan 07 23:31:10 which is not bash compatible. sh is not expected to be bash compatible. Jan 07 23:31:16 whew Jan 07 23:31:22 Android's build system somehow makes the assumption that it is, however :) Jan 07 23:31:37 it looks like the assumption is that choosecombo scripts dont explicitly set /bin/bash. Or at least, this one doesn't. Jan 07 23:31:38 does the copy of build/envsetup.sh have [[ in it? Jan 07 23:31:44 if i understand joeo above, it sounds like one of the build scripts is just forgetting to refer to /bin/bash explicitly Jan 07 23:32:00 joeo: it's not envsetup.sh that does this. It's build/core/find-jdk-tools-jar.sh Jan 07 23:32:04 ahh ok Jan 07 23:32:08 that could be the real issue. Jan 07 23:32:16 ctate: jasta is right...that's not uncommon. Bash is fairly fat as shells go. Most shells don't need the power (and resulting bulk) provided by bash Jan 07 23:32:27 how about build/core/config.mk, does that start with SHELL := /bin/bash? Jan 07 23:32:31 in your tree? Jan 07 23:32:47 BlindOracle: i got it, np Jan 07 23:32:56 jasta: if the shell starts with /bin/bash, that's the author stating it requires bash Jan 07 23:33:05 i do not have a line like that in my tree. main.mk, however, contains that line. Jan 07 23:33:08 and is not intended to execute under sh Jan 07 23:33:10 ahh ok Jan 07 23:33:18 then it's just that change hasn't been pushed to git yet Jan 07 23:33:23 if you add that, it'll fix it Jan 07 23:33:25 BlindOracle: huh? Jan 07 23:33:46 if the magic bang specifies bash, it's a bash script Jan 07 23:34:07 i know that; this script has no hash bang. Jan 07 23:34:14 as many of them do not. Jan 07 23:34:16 iirc, there are some bashisms in the shell scripts embedded in the build system, which don't have a #!/bin/bash line, although that script should anyway Jan 07 23:34:49 I know that adding that to build/core/config.mk fixes it Jan 07 23:34:52 if it has bashisms, it should have #!/bin/bash...otherwise #!/bin/sh Jan 07 23:35:45 of course, some platforms don't honor magic but that's their broken (window)...*cough* Jan 07 23:38:34 hmm, i feel like such a dolt trying to work with git and repo Jan 07 23:39:04 jasta: I'm still coming up to speed on git my self, but once you learn it, you love it Jan 07 23:39:17 i cant figure out how to ammend the patch set i just committed Jan 07 23:40:39 i didnt change [[ to [ but i should have. no clue how to do this without starting another review. Jan 07 23:40:41 jasta: I don't remember what they call it (I'd have to look) but you can revert your commit by creating a new commit (tool does it for you) which restores to pre-commit state. You then apply what else you need and commit again Jan 07 23:40:57 http://www.kernel.org/pub/software/scm/git/docs/user-manual.html is your friend. ;) Jan 07 23:41:19 http://www.kernel.org/pub/software/scm/git/docs/user-manual.html#fixing-mistakes Jan 07 23:41:26 IIRC, that's where you need to look Jan 07 23:41:26 but i think this is a repo thing, not git. Jan 07 23:42:04 git commit --amend maybe? Jan 07 23:42:15 thats what i did, but repo upload still wants to make a new patch set. Jan 07 23:42:33 oh ya...sorry...forgot you're using repo too Jan 07 23:42:51 amend should only ever be used if you have NOT pushed changes up stream yet Jan 07 23:43:43 I'm calling it a day. sqlite frustration has gotten the better of me today Jan 07 23:44:02 michaelnovakjr: And BTW, thanks so much for your assistance today. It most certainly is appreciated! Jan 07 23:45:13 michaelnovakjr: I'll update the bug report tomorrow to reflect the facts on the ground as we know them now. Depending on what I (we ;) find tomorrow, I may open a bug report against the helper class as it is a support nightmare waiting to happen for multi-table databases. Jan 07 23:51:48 jasta: like joeo said, just change build/core/config.mk as he suggested rather than changing the [[ stuff. Jan 07 23:52:16 the bug you're seeing isn't that it's using [[, it's that it is not correctly ensuring that it's running bash Jan 08 00:23:32 ctate: is that change coming to git? Jan 08 00:23:53 regardless, i still would like to learn more about how to use repo :) Jan 08 00:34:38 isn't it possible to define strings (in strings.xml) in printf() format? like "Blah blah %s" and pass in %s at runtime? Jan 08 00:34:51 it's hard to search for, so I'm asking here :-/ Jan 08 00:35:01 yeah, look at the Email app. it does this. Jan 08 00:35:12 see Resources.getText(int, Object...) Jan 08 00:35:14 the Music app too Jan 08 00:35:19 you use %1, %2, etc. Jan 08 00:35:38 sorry, Resources.getString(int, Object...) Jan 08 00:36:44 there also are some helpers for plurality and such Jan 08 00:36:50 nice Jan 08 00:36:53 thanks, guys Jan 08 00:37:43 using the Resources mechanisms with minimal extra fiddling in your code makes for the best localization experience, too, FYI Jan 08 01:35:02 * jasta wonders if romain has done anything with his app Jan 08 01:35:02 :) Jan 08 02:08:59 jasta: yeah so I almost released it this morning Jan 08 02:09:07 but then I remembered I'm using reflection to access 2 private APIs Jan 08 02:09:12 so I need to fix that first ;) **** ENDING LOGGING AT Thu Jan 08 02:59:57 2009