**** BEGIN LOGGING AT Mon Nov 10 02:59:57 2008 Nov 10 03:47:01 Can anyone offer resources for someone new to android development? Nov 10 03:49:52 Are you trying to write an App? Nov 10 03:50:03 And are you new to Android, or new to mobile devices, or new to development in general? Nov 10 03:50:11 Trying to write an app Nov 10 03:50:22 I've got some of the basics down, the file structure, etc Nov 10 03:50:33 but getting into the docs and moving forward has been frustrating Nov 10 03:50:39 Basic Java knowledge? Nov 10 03:50:52 Somewhat, not as strong as I'd like Nov 10 03:51:19 I'd start with http://code.google.com/android/intro/develop-and-debug.html Nov 10 03:51:34 Which talks about getting your dev environment set up and launching and running your first app. Nov 10 03:51:39 (Or it did last time I looked) Nov 10 03:52:04 I've got Eclipse all set up and written a few simple apps, just can't get much further than that Nov 10 03:52:07 There's a tutorial as well on how to hack on Notepad, which might be a good start. You can also look at the other sample code in the SDK package. Nov 10 03:52:17 for instance: onClick handlers for buttons and such Nov 10 03:52:18 Ah -- do you know what kind of app you're going to write? Nov 10 03:52:45 Working my way towards games, I understand I'm not very good at Java yet, heh Nov 10 03:53:03 Generally, you use the layout xml resources to define the structure and location of widgets, and then hook up the handlers for them. Nov 10 03:53:33 I've gotten layouts made/working fine and such, it's more of the Java itself that's giving me trouble Nov 10 03:53:34 Although with games, it sometimes depends: if you're writing something that does all its own drawing, you don't deal with layout much (like LunarLander or Snake). Nov 10 03:53:47 switching views, things like that Nov 10 04:03:53 So you'd recommend taking a look at the LunarLander/Snake source? Nov 10 04:04:35 Sounds like that might be a place to start. Nov 10 04:04:43 They both basically set up a view and an event/update loop. Nov 10 04:05:05 The LunarLander one is "draw as fast as possible" while the Snake one is "update on a clock" Nov 10 04:05:43 Thanks for the help, I appreciate it Nov 10 04:06:10 No problem -- if you're looking for an idea, I really want someone to write a "Battleship" game. :) Nov 10 04:06:27 Use networking or SMS or instant messaging to have two devices communicate or something. Nov 10 04:06:28 haha no way? That's the game I was going to make Nov 10 04:06:46 That, and a remake of the old tank game from atari Nov 10 04:06:51 Combat! Nov 10 04:07:00 Yeah, loved that Nov 10 04:07:09 Just moving from C to Java has been a painful transition Nov 10 04:07:40 Yeah, it's a little tough. If you've done some C++ or other OOP languages, it's easier. Nov 10 04:08:03 I know OOP, Java just seems very belt and suspenders Nov 10 04:08:14 And it helps to look at a lot of examples (at least for me) because there's a lot of things you can do in Java that're crazy in C. Nov 10 04:08:28 A good IDE helps a lot with Java too, since there's so much strong typing wandering around. Nov 10 04:08:38 Yeah, Eclipse has been a life saver so far Nov 10 04:09:08 anyone is familiar with the ranking-algorithm from the market ? Nov 10 04:09:45 how is the averaged-ranking determined? Just the average over all rates ever? Nov 10 04:10:06 Thanks again for the help, gotta just bear down and get it done :) Nov 10 04:10:14 Yeah, that's always my problem, too. Nov 10 04:11:47 or sth smarter, because when I had a bug in my app and some rated it bad, I can never get up again :7 Nov 10 04:11:49 :/ Nov 10 04:12:38 plusminus_ how do they handle updates for the market apps? Nov 10 04:24:41 I just uploaded a new apk Nov 10 04:24:47 :/ Nov 10 04:25:16 plusminus_: which app? Nov 10 04:26:26 AndNav!!!!!!!!!!!!! Nov 10 05:00:10 michaelnovakjr f00f- right Nov 10 05:09:00 shouldn't the android source (source.android.com) include the source for things like EditText ? Nov 10 05:21:55 KNY: it does Nov 10 05:21:55 oh, I found it, nevermind Nov 10 05:22:04 heh, jinx Nov 10 05:22:20 it didn't get imported into Eclipse Nov 10 05:52:50 sweet Nov 10 05:53:45 hehe, i should totally update screenshots on five.googlecode.com Nov 10 05:54:21 * michaelnovakjr_ agrees Nov 10 05:54:38 i have one more set of visual changes i want to accomplish before i do that Nov 10 05:54:53 back to working on it? Nov 10 05:55:00 i'm gonna adopt the music players "main" UI where it shows a few huge buttons to get you started Nov 10 05:55:04 to eliminate the tabs Nov 10 05:55:17 and then i am going to take their "now playing" indicator (the bottom bar that says the name of the current playing track) Nov 10 05:55:43 also, i'm gonna take the YouTube apps bottom bar design with the progress slider that disappears when buffering has finished Nov 10 05:55:50 nice Nov 10 05:56:19 i wonder if i can extract the YouTube apps colors :) Nov 10 05:56:54 its a better translucent gray than five has currently Nov 10 05:56:55 :) Nov 10 05:58:40 i bet my axml2xml script still works ;) Nov 10 06:01:33 :) that would be helpful Nov 10 06:10:24 turns out it does kind of work Nov 10 06:10:29 but there is some silly bug somewhere Nov 10 06:11:10 it thinks everything is the same value Nov 10 06:11:15 android:id="visibility" Nov 10 06:11:15 android:background="visibility" Nov 10 06:11:15 android:layout_width="visibility" Nov 10 06:11:17 :) Nov 10 06:12:01 i might take a look at this :) Nov 10 06:12:12 because i'd really like to know what color they use Nov 10 06:19:18 seems worth it to me Nov 10 06:19:25 would taking a screenshot and using photoshop/gimp color picker not give ypu the correct color? Nov 10 06:19:55 * languish just curious Nov 10 06:21:22 depends on how you took the screenshot Nov 10 06:21:31 still, just use xmag Nov 10 06:22:40 you could find out more about the transparency with the xml stuff though Nov 10 06:23:13 yeah computing the transparency from a screenshot would be really hard Nov 10 06:23:22 I see. thanks Nov 10 06:23:36 you could write a program to do it assuming the color is constnat, but that would be much less work than just examining the xml files in the app Nov 10 06:23:53 nod Nov 10 06:24:21 if the source is there hey why not :) Nov 10 06:24:29 source isnt there for youtube Nov 10 06:24:35 but i have a tool that can still get into it Nov 10 06:24:42 nod Nov 10 06:35:17 so who do I thank for this? http://www.droiddraw.org/ Nov 10 06:43:33 g1er: clicking that paypal donate button might be what you're looking for Nov 10 06:47:13 thx. ;) Nov 10 07:07:21 hmm Nov 10 07:07:32 i'm having a really hard time figuring out the colors used in the youtube app :) Nov 10 07:07:37 for that overlay when you tap the video Nov 10 07:18:04 oh wait, this is in the msuic app Nov 10 07:20:29 this must be part of VideoView... Nov 10 07:21:29 ahh, MediaController, nice Nov 10 07:32:46 well hmm. when i think im copying them, it still looks like hell Nov 10 09:27:05 Any idea why one specific installation of eclipse/sdk on linux wouldn't be able to make .apks? Nov 10 09:29:19 not sure if it matters but do you have the android plug-in installed also? (is it even need on linux/or to create apks) Nov 10 09:29:32 yes Nov 10 09:29:48 I think I may have just found the answer though :) Nov 10 09:30:07 cool :) Nov 10 12:13:59 is it possible to change the background color of Buttons? I mean the default orange gradient? I would just like to change the tint of it.. just replace the orange one with another but so that the border/shadow stays the same.. Nov 10 15:46:00 tauno: sure, use a style. Nov 10 16:28:19 jasta, then I'd have to implement my own style including the 9.png's and stuff etc that are used in the default style? Nov 10 16:29:20 no, inherit from the default style Nov 10 16:29:36 there are plenty of examples of this in the source and sampels Nov 10 16:44:48 ok, thanks Nov 10 16:50:00 jasta, didn't find anything in the API demos neither in "Implementing a User Interface" or "Building Custom Android Components" in the documentation Nov 10 16:50:04 any pointers? Nov 10 16:50:14 sry for being a pain in the ... Nov 10 16:57:14 http://code.google.com/android/reference/android/content/package-descr.html defines "Bag resources" that sounds something like it would fit the bill ("Bags also support inheritance") but this term is not used anywhere (at least searching for it doesn't turn up anything except the same page) Nov 10 16:58:57 ok, I'm blind it seems.. Nov 10 16:59:05 found it on the same page :) Nov 10 17:01:54 wonder where is the default style for buttons? I don't know what exactly do I have to override to get the background color to change (so the borders and shadows and gradient etc stay the same.. only the color changes (I still guess I can't do it that easily..)) Nov 10 17:04:12 tauno: might want to checkout /frameworks/base/core/res/res/values/themes.xml in the git tree Nov 10 17:06:25 thanks Nov 10 17:06:59 I'll try to get it into my little brain :/ Nov 10 17:22:05 ok if I know the width of my textview, is there anyway to tell the maximum number of characters it can show Nov 10 17:22:29 I'm thinking if I know the ems property, that could give me a rough idea Nov 10 17:22:43 famast, only if you're using a fixed-width font Nov 10 17:24:47 hmm I see. This is because a user sets a bigger font size because he has a hard time reading, and thus the width increases? Nov 10 17:25:27 it seems that the button just uses a 9 patch png for the look.. (including borders and shadow) so I can just copy it and edit it Nov 10 17:25:31 thanks Nov 10 17:26:38 my whole problem is I'm trying to append an ellipses if the entire text is not shown Nov 10 17:27:23 famast: you mean as in android:ellipsize="end" ? Nov 10 17:28:25 ah brilliant! thank you Nov 10 17:29:23 I hadn't finished going over the textview attributes, and my search failed on ellipses Nov 10 18:28:05 This is very interesting. File dir = new File(Environment.getExternalStorageDirectory, "mydir"); String[] list = dir.list(); the last element of list is the file I want to open, but trying to open that file still doesn't work. Nov 10 18:28:33 ... sigh. Nov 10 18:28:37 No, it's not the same name, it has a typo Nov 10 18:28:42 durf durf I'm dumb Nov 10 18:31:37 is source.android.com supposed to give you the source to all the apps that are installed by default? Nov 10 18:31:50 I'm trying to find the source to Market Nov 10 18:31:57 KNY: except amazon, Market, YouTube, and Gmail Nov 10 18:32:10 KNY: Market is not open source Nov 10 18:32:18 dang. Nov 10 18:45:12 vol: does "mydir" actually exist? Nov 10 18:45:28 Never mind. Didn't read your next lines :) Nov 10 19:03:07 I know the answer here is probably "no", but is there any way to prevent default handling of the Dial, Home and End buttons? I can see the KeyDown events on these, but I can't prevent the handset from taking the default actions. Is there a flag I can set somewhere to prevent that? Nov 10 19:08:19 I somewhat suspect not Nov 10 19:21:45 I strongly suspect not. But there's always hoping. Clearly there needs to be some way for a user to escape to the desktop, but, those buttons do also make awfully attractive game controls... Nov 10 19:25:33 I think the ability of the user to say "fuck this bullshit" and "ok I want to use my cellphone as a cellphone now" trumps those desires Nov 10 19:27:16 Yeah, that's the frame of the argument. Nonetheless, users would happily play a good game that broke those rules. I was just curious about the API flexibility. Nov 10 19:29:44 "Application is requesting the following priviledges:" Nov 10 19:29:46 "Brick" Nov 10 20:33:23 I'm having a hefty amount of trouble creating my own /data image Nov 10 21:04:11 romainguy: regarding our conversation about the efficiency of persistent connections, it is important to factor in that polling is likely to be even less efficient in the long run Nov 10 21:04:48 if each check is scheduled at, say, 30 minutes, then instead of running a simple keep-alive every 6 minutes, you will now have to perform a database query, tons of allocations, iterations, and lots of network I/O. Nov 10 21:05:15 so if email is somewhat infrequent, push is actually going to be a big win on battery life Nov 10 21:05:21 jasta: depends on the relative cost of running the CPU vs. running the radio Nov 10 21:06:14 fadden: how do you mean, both approaches have to run the cpu and the radio. the balance is that one ideally runs them frequently but to do very little work, the other does so infrequently but performs lots of work Nov 10 21:07:02 I'm not an expert, but I suspect that turning the radio on 5x and doing 5 units of network traffic (1 unit each time) is more expensive than turning it on once and doing 10 units of network traffic. Nov 10 21:07:23 Especially since the radio doesn't necessarily turn off right after you finish using it. Nov 10 21:07:32 How quickly does a UMTS connection drop when it's idle? Nov 10 21:08:23 fadden: the radio never turns off, actually. Nov 10 21:08:30 (thats why push e-mail works) Nov 10 21:09:01 it might have some type of power saving semi-sleep mode, but it can't truly sleep. the radio can also wake a sleeping cpu. Nov 10 21:09:38 If it turned completely off your phone wouldn't ring. But there's a difference between receiving a transmission and transmitting. Nov 10 21:09:48 for instance, if you connect to a server and the two sides idle, the cpu will eventually fall asleep on the client/phone. if the server says something, the client will wake up and read the data. this is because the radio is on, but the cpu is sleeping, and the radio can wake the cpu to perform work. Nov 10 21:10:53 still, i think that your estimations of units of work are inaccurate, and neither of us know the functions well enough to compare the two sets either way. Nov 10 21:11:10 but i'm guessing that the functions favor the usage case of push e-mail, where you're merely guessing the opposite :) Nov 10 21:11:50 jasta: I yield to our equally inadequate knowledge. Nov 10 21:12:35 We have an "efficient code" document, but no "efficient battery" or "efficient radio" equivalents. Nov 10 21:12:37 but it's kind of a moot point since the feature is worth a loss of battery performance if that is truly what happens. Nov 10 21:13:13 One is led to believe that "don't leave the GPS on at all times" is non-obvious. Nov 10 21:13:16 i've been running this code all day today and actually i havent noticed any difference either way. it could be negligible. Nov 10 21:14:15 fadden: the scheduled checks that the email app does currently are really inefficient though. Nov 10 21:14:45 I would tend toward doing whatever the gmail app does. Nov 10 21:14:59 me too, if only it were public knowledge :) Nov 10 21:15:12 oh, i thought there was just #android :) Nov 10 21:15:17 i suppose thats more the hardware Nov 10 21:15:31 it didn't used to be, but we've been forced out :) Nov 10 21:15:44 jasta: we gave you "dexdump -d". Get busy. Nov 10 21:15:58 fadden: what would be even better is if i could hook into its alarms so that at least we both wake the cpu at exactly the same time :) Nov 10 21:16:48 fadden: hehe, well, i have an implementation that i _think_ is the same. but i may have to do that to confirm at some point. Nov 10 21:16:59 (actually, if we made you promise not to disassembly the code, then I have to discourage you from getting busy) Nov 10 21:17:14 yea, i'm looking for someone that might be interested in a little open source project for android :) I'm starting to do dev with it, but slowly but i have a great idea that leverages on an appengine backend that I can write in Python, and was wondering if anyone may be up to the challenge :) Nov 10 21:17:38 digitalspaghetti: that's usually not how you recruit open source developers, btw. Nov 10 21:17:53 Usually you write something awful and defy anyone to improve it. Nov 10 21:18:02 precisely :) Nov 10 21:18:08 :p Nov 10 21:18:54 well, it's not really open source in that it's a proof of concept where the source would be available Nov 10 21:19:05 digitalspaghetti: you should start your project, host it publicly at google code, and wait to see if anyone thinks your idea is worth a damn Nov 10 21:19:21 i got the idea while writing my jQuery plugin, jMaps Nov 10 21:19:30 if it's a good idea, you write specifications, and a mock up you might find someone Nov 10 21:19:33 and poking about on the phone Nov 10 21:20:11 digitalspaghetti: either way, you're unlikely to recruit anyone serious if you merely have a good idea you want to share... Nov 10 21:20:13 could have that up in about 20 mins, once i re-write some of the stuff and throw it on the wiki Nov 10 21:21:32 there's quite a bit that a non-programmer can do Nov 10 21:21:55 docs, psuedo code, communication protocol, mock up screen shots, etc. Nov 10 21:22:42 if it's described well enough a programmer can bang out something quick Nov 10 21:24:45 spikebike: i am a coder, i do Python, PHP, JavaScript and Actionscript, and a little bit of HaXe Nov 10 21:24:53 it's my Java where I fall down Nov 10 21:25:04 great, that helps alot Nov 10 21:25:38 just be precise, if you specify something cool so that a good programmer can bang it out quick without asking 100 questions they just might Nov 10 21:26:10 basically market your idea (so it sounds good) then lower the barrier to entry by doing the hard work Nov 10 21:26:45 taking a well specified idea and getting the java syntax right is IMO the easy part Nov 10 21:34:45 ok, project page added: http://code.google.com/p/livelocations/ Nov 10 21:36:18 my thoughts are, why make another application for it if it can just be a service, and use the existing contacts manager to display it Nov 10 21:39:49 I haven't looked at your page, but I can assume that it could also be called "StalkerApp" Nov 10 21:40:40 well, that needs to be worked out in the implementation - such as authorising contacts to view your location Nov 10 21:40:53 you could say for example "Only let them see the city i am in" Nov 10 21:45:34 for example, the app could go "Here is a secure hash of all my owners phone numbers/emails, please find them" The app goes, "Ok, But I need to request their permission first". On that persons next sync, it flags a notice to say "This person wants to see your location, do you want to let them? What Granularity?". That then goes back to the server on the next sync and updates the flag Nov 10 21:45:40 to ay what rlevel of detail they get Nov 10 21:47:58 using the contact database as a permission broker seems kinda strange Nov 10 21:48:28 why? it gives you that row-level granulrity for permissions Nov 10 21:49:02 "I don't mind if Bob gets my exact position, but I don't want my mum seeing where I go on the weekends" Nov 10 21:51:37 anyone know what the filter option is for the method createScaledBitmap Nov 10 21:51:53 its a boolean value, but there is nothing in the docs about this method Nov 10 21:52:32 dunno, seems like you should propose a standard format for that kinds of info then offer an google app to do that storage for purpose then if it gets popular enough lobby for it's inclusion in the contact database Nov 10 21:53:03 seems kinda silly to add a bunch of fields to every contact entry for an application that may or may not get popular Nov 10 22:00:30 http://code.google.com/p/livelocations/wiki/Security Nov 10 22:00:57 maybe spikebike Nov 10 22:04:24 it could just be a custom label field added to contacts that are recognised with the service Nov 10 22:05:09 just call it "Current Position" and the app reads teh contact and find that field Nov 10 22:05:22 since the app will run as a service Nov 10 23:27:06 are the icons in the android source free (as in speech) for reuse? Nov 10 23:32:35 The license should be specified; probably Apache2. Nov 10 23:33:36 ah yes, found it (MODULE_LICENSE_APACHE2) **** ENDING LOGGING AT Tue Nov 11 02:59:57 2008