**** BEGIN LOGGING AT Tue Feb 26 10:59:56 2008 Feb 26 11:37:23 Good morning :) Feb 26 11:37:26 anyone knows if DataStateIntentReceiver is working properly ? Feb 26 11:41:51 good evening :) Feb 26 11:42:57 ;) Feb 26 11:44:13 I did the following and never a message is received: Feb 26 11:44:22 this.dsir = new DataStateIntentReceiver(this, new Handler(){ Feb 26 11:44:28 @Override Feb 26 11:44:31 public void handleMessage(Message msg) { Feb 26 11:44:35 Log.d(Constants.DEBUGTAG, "DataStateIntentReceiver Msg received."); Feb 26 11:44:39 super.handleMessage(msg); Feb 26 11:44:44 } Feb 26 11:44:47 }); Feb 26 11:44:51 this.dsir.notifyConnectionState(CONNECTIONSTATE_WHAT_ID); Feb 26 11:46:13 I can only see the system-logging a change: Feb 26 11:46:13 DEBUG/Sync(508): received data connection state changed: DISCONNECTED Feb 26 12:17:16 anyone ? Feb 26 14:35:39 anyone knows a source where to find the cycles needed to perform specific arithmetic operations on ARM-cpus ??? Feb 26 14:36:32 in a manual maybe Feb 26 14:36:45 !google arm manual clock cycle Feb 26 14:36:53 no google bot :) Feb 26 14:45:10 thx, I found: Feb 26 14:45:28 QADD, QDADD: 1 Cycle Feb 26 14:45:39 Q is ? QuadWord ? 64 bit ? Feb 26 16:14:14 why is it that in the submission page (,http://code.google.com/android/adc-submit/index.html#individual) I can only "choose" to submit an M3-rc37 sdk version project?? Feb 26 16:15:15 weird Feb 26 16:15:46 Lede, does the same thing happen to u? Feb 26 16:15:52 yes Feb 26 16:16:00 but i havnt submited anything yet Feb 26 16:16:15 * Lede pokes morrildl Feb 26 16:16:36 mike1o: no worry you can submit an M5 app Feb 26 16:16:41 mike1o: morrildl is working on fixing the submission app Feb 26 16:18:50 romainguy_: no, actually they can't right Feb 26 16:18:53 right now Feb 26 16:19:11 if they submit it as M3, it will get judged on M3, fail to run on that emulator, and be all around bad :) Feb 26 16:19:54 morrildl: I meant we won't forbid M5 entries Feb 26 16:20:05 * romainguy_ should not speak in the morning :)) Feb 26 16:22:39 jeffb: you here yet? Feb 26 16:22:44 ... ww Feb 26 16:23:22 is it normal that contacts in content://im/contacts is not updated? Feb 26 16:25:46 s/is/are Feb 26 16:34:44 hey, have you guys had any trouble with the manifest editing pane in the new sdk? Feb 26 16:35:32 by 'sdk' i mean 'eclipse plugin' of course. i get an 'editor failed to initialize' Feb 26 16:35:35 whoops Feb 26 16:35:48 'editor could not be initialized' in Editor could not be initialized. Feb 26 16:36:02 fucking a, why didn't it paste the path Feb 26 16:36:18 in com.android.eclipse.editors.manifest.ManifestSourceViewerConfig Feb 26 16:36:38 and then later if I change attributes in this pane and save, i get Save Failed:null Feb 26 16:36:45 anyone run into this? Feb 26 16:45:48 I haven't figured out how to use the manifest wizards Feb 26 16:54:23 thedaniel`: Feb 26 16:55:13 do you have the WST bits for Eclipse installed? Feb 26 16:59:28 jasonchen: I do Feb 26 16:59:50 jasonchen: let me fire it back up - i had switched back to the cli tools for the time being Feb 26 17:01:45 perhaps also relevant to this issue is that I can't run projects from Eclipse either (on ubuntu 7.10 btw) - i am told that it can't find /bin/foo.apk Feb 26 17:02:03 hmmm, that's odd Feb 26 17:02:19 no build.xml is generated, but i can shell out and run activityCreator on the sample apps which will generate a build.xml that leads to a properly compiled app Feb 26 17:02:21 are you sure you set the location of the SDK in the preferences? Feb 26 17:03:15 also, what JDK do you have installed? Feb 26 17:03:34 jasonchen: yup, the current value of that field is /home/daniel/src/android-sdk_m5-rc14_linux-x86/ Feb 26 17:03:41 which is the top-level directory with android.jar etc Feb 26 17:03:59 i have sun jdk 1.6 Feb 26 17:04:20 thedaniel`: what is JAVA_HOME set to? Feb 26 17:04:38 the first ilne of /etc/eclipse/java_home is /usr/lib/jvm/java-6-sun/bin/java Feb 26 17:05:04 there is no $JAVA_HOME env var tho Feb 26 17:05:25 i'll try setting that in .bashrc to the same value i pasted from /etc and relaunch eclipse? Feb 26 17:05:46 well, check out the last item on: http://code.google.com/android/RELEASENOTES.html Feb 26 17:06:40 jasonchen: ah ha - i figured it gcj somehow.. but was sure /etc/eclipse/java_home was the palce for it Feb 26 17:07:09 jasonchen: thanks for the link, hope that works, and if it does, i'll make sure the next question i ask in here isn't so obviously rtfm Feb 26 17:07:24 thedaniel`: no worries :-) Feb 26 17:08:12 jasonchen: ah crap, i see what i did Feb 26 17:08:24 that line i pasted from java_home ends in /bin/java Feb 26 17:08:31 but what that var wants to be is the top level directory Feb 26 17:08:37 that contains /bin/java :) Feb 26 17:09:03 there you go Feb 26 17:12:12 thedaniel`: which distro are you using? Feb 26 17:43:36 yawwwnn Feb 26 17:44:24 jasta, do you think is it normal that contacts in content://im/contacts are not updated? Feb 26 17:47:40 mike1o: are you just messing with me? :) Feb 26 17:48:11 jasta, no :-) why should I ? Feb 26 17:48:24 i went on this big tirade last night about how you can't expect undocumented portions of Android to work intuitively, well, or at all. Feb 26 17:48:31 and the im provider is undocumented ;) Feb 26 17:48:56 yup... I guess it's not something we should know then Feb 26 17:55:22 re Feb 26 18:04:08 romainguy: think there is a workaround for this: http://code.google.com/p/android/issues/detail?id=357 ? Feb 26 18:05:45 zhobbs: I have no idea Feb 26 18:05:52 ok Feb 26 18:06:56 hey guys Feb 26 18:07:25 Would be great if anyone could give some feeback on my project "AndNav!": Feb 26 18:07:34 Flash-Video: http://www.anddev.org/images/tut/map/andnav/screencast.htm Feb 26 18:08:20 i can't watch a video right now, but what is AndNav? Feb 26 18:08:27 something you wrote? Feb 26 18:09:25 plusminus: looks pretty nice Feb 26 18:10:31 we need a speech synthesizer and the ability to rotate the map to really get it going Feb 26 18:10:49 but nice job plusminus, is the code available? Feb 26 18:10:57 or will it be Feb 26 18:11:00 rotation is not possible, right? Feb 26 18:11:02 plusminus: looks great! Feb 26 18:11:13 I'm plannign to submit to ADC Feb 26 18:11:38 I'll have to have a look on what I'm allowed to release or not.. Feb 26 18:11:53 plusminus: don't think you can rotate the map Feb 26 18:11:59 jasta: Yes I wrote it Feb 26 18:12:19 plusminus: I like the start address you use in the video... this is *exactly* where I used to live last year :) Feb 26 18:12:56 =D I think its where the MockLocation Data starts Feb 26 18:13:07 aah Feb 26 18:13:19 romainguy: the world is small ;) Feb 26 18:14:19 there will be that "arrow" you all know from other navigation system in the lower left corner Feb 26 18:15:34 it runs even >pretty< smooth on my emulator without any explicit optimization. Feb 26 18:15:49 very nice so far, plusminus Feb 26 18:16:00 50-150 ms / full invalidation on HVGA Feb 26 18:16:03 good job Feb 26 18:16:16 thanks =) Feb 26 18:17:52 I'd like to make a bluetooth car dock to interface with some of the high end phones so you can hook into the stereo for apps like this and media player, etc Feb 26 18:18:34 btw its pretty funny to see that the mock-location-data is "played backwards" once it has finished ^^ Feb 26 18:18:44 are we not supposed to write google talk apps? Feb 26 18:18:51 has anyone gotten GmmGeocoder to display business names? Right now when I run GmmGeocoder.query() it's returning an array of Address objects, but the business name is not stored in that object Feb 26 18:18:57 was wondering if someone hacked around the problem. Feb 26 18:19:06 romainguy: turns out the guy who wrote Shorewall not only used to live in the house I live in now, but he currently lives in the cul-de-sac in north seattle that my buddy lives. Feb 26 18:19:26 really weird :) Feb 26 18:19:34 I would be very impressed if I knew what Shorewall is :) Feb 26 18:20:14 It's a powerful intermediate tool to manage Linux iptables. Feb 26 18:20:16 Google says its a firewall under GPL Feb 26 18:20:32 I guess silence answers my question :) Feb 26 18:20:49 i suppose silence answers mine too, mike1o :p Feb 26 18:21:23 oh,not GPL but "GNU Free Documentation License". Feb 26 18:27:19 g2g, bye Feb 26 18:40:41 * jasta could never understand why Java did not specify a binary tree collection. Feb 26 19:17:11 good evening Feb 26 19:20:04 hmm, the iPhone has a much nicer window title than android :) Feb 26 19:20:07 i wonder if that is being looked at Feb 26 19:20:19 it's better called out, and can have useful buttons in it Feb 26 20:08:25 romainguy: How does Google recommend developers to publish the interface for services and providers for other applications to use? Feb 26 20:08:48 Just distribute a jar file with the interface classes? Feb 26 20:10:14 jasta: aren't different applications only meant to interact via ContentProviders? Feb 26 20:21:01 dragor43: No; The Android platform heavily depends on the service architecture to let applications interact with other packages. Feb 26 20:32:44 hrm hrm hrm hrm... Feb 26 20:46:31 it's annoying not being able to change the apk on the fly Feb 26 20:47:18 davidw: it's a very specific usage that you have Feb 26 20:49:05 in some ways... Feb 26 20:49:34 unix uses it: think of programs that switch behaviors on argv[0] Feb 26 20:56:33 romainguy: do you have any thoughts on exposing my service to other packages? should i just make a jar file and distribute it for developers? Feb 26 20:56:42 davidw: you mean generate bytecode on the fly? Feb 26 20:57:06 jasta: I have no idea Feb 26 20:57:45 That's essentially what Google has done, so I guess I can take their lead. You guys just distribute one giant jar file instead ;) Feb 26 20:58:05 Distribution of software in Java is such a mess. Feb 26 20:58:36 how so? Feb 26 20:59:43 dims, nah Feb 26 21:00:11 dims, with j2me Hecl, I have a java program that twiddles the .jar and .jad a bit, so that you can generate new ones without any compilation Feb 26 21:00:18 There's no convention anywhere. There's no common build targets, no common installation mechanism, no recommendations for versioning... Feb 26 21:00:30 it replaces the Hecl script contained within, and changes the name Feb 26 21:00:41 No prescribed way to either provide or use third party functionality Feb 26 21:02:44 romainguy: It's bad ;) Feb 26 21:03:08 jasta: sure, a lot worse that in other languages Feb 26 21:03:19 Try sitting in the shoes of a package maintainer for a major distribution. Feb 26 21:03:54 romainguy: Yeah, but for as broken and stupid as autoconf/automake is, it certainly does solve this problem well. Feb 26 21:04:00 they're probably sandals Feb 26 21:04:07 ^^ davidw Feb 26 21:04:29 jasta: use Maven :) Feb 26 21:04:59 I probably should, actually. Feb 26 21:05:35 It's frustrating that ant doesn't have strong guidelines on what functionality is appropriate and what it should be called, how it should be used, etc. Feb 26 21:05:49 While any project can deviate, it's very important to beat people over the head with convention ;) Feb 26 21:06:24 * jasta reads up on Maven, perhaps this will solve my problem nicely Feb 26 21:07:00 jasta: we've been struggling with Maven2 for a year or so now...it's been tough to say the least Feb 26 21:07:50 i have never worked with it, but i am encouraged by the sensibility of its description :) Feb 26 21:08:03 Plus I am a total apache whore Feb 26 21:08:04 never looked at maven - the vague impression I get is a 'do everything our way and it will be great' project, but I could be completely wrong, as is often the case :-> Feb 26 21:08:51 davidw: i could not have summarized better Feb 26 21:08:58 davidw: That sounds perfect. Feb 26 21:10:33 davidw: That's the point of convention. Deviate only when you must, not just when you can. Feb 26 21:10:59 But I will learn more before I will speak to Maven specifically :) Feb 26 21:12:54 Maven's not in Debian? I'm out. :) Feb 26 21:33:48 romainguy: I really rather like this thing, thanks :) Feb 26 21:34:12 it's so imperialistic, gives me goose bumps. Feb 26 21:34:14 you're welcome Feb 26 21:34:40 it seems very heavy, though Feb 26 21:34:47 but it distributes itself well Feb 26 21:35:56 hrm Feb 26 21:38:25 there isn't an INTENT to handle random gtalk packets is there? Feb 26 21:48:54 jasta: do parcelables work with services? Feb 26 22:38:34 zhobbs: Yes, of course. Feb 26 22:40:38 zhobbs: That's principally how services work. Feb 26 23:13:06 jasta: so I create the parcelable class and an aidl file for that parcelable Feb 26 23:13:13 but how do I as the docs say "Last but not least, add an aidl file for your parcelable class so the AIDL tool can find it, but don't add it to your build. This file is used like a header file in C. You don't compile the aidl file for a parcelable just like you wouldn't normally compile a .h file." Feb 26 23:14:40 eclipse is telling me my java file is in the way Feb 26 23:23:05 Huh? Feb 26 23:23:24 You're trying to create create a new parcelable type for your interface? Feb 26 23:23:54 err, create* :) Feb 26 23:24:46 yeah Feb 26 23:25:09 So I defined that class in Class.java, and the docs say I need a Class.aidl Feb 26 23:29:03 Yeah, you do, however I have not worked with this yet :) Feb 26 23:29:25 Ahh, somehow I have to tell eclipse to leave me alone about that aidl file Feb 26 23:29:37 Oh, that's easy Feb 26 23:29:42 just say exclude from build Feb 26 23:29:45 somewhere in the right click contxt menu Feb 26 23:31:47 sweet, i found a maven android plugin :) Feb 26 23:39:41 damn, can't figure out how to exclude the file Feb 26 23:43:21 its in there, trust me. Feb 26 23:43:51 right click the file you want and it's somewhere in there to say exclude from build Feb 26 23:44:46 there is a greyed out "Exclude from context" Feb 26 23:45:19 "Remove from context" Feb 26 23:46:28 no its not that, here let me find it Feb 26 23:46:49 :) Feb 26 23:51:00 oh interesting ;) Feb 26 23:51:06 Stupid eclipse only has that option for Java files Feb 26 23:51:10 hehe Feb 26 23:51:21 so this is a deficiency in the android plugin then. Feb 26 23:51:43 bummer, you might be hosed :) Feb 26 23:52:08 http://groups.google.com/group/android-developers/browse_thread/thread/a98df6ea4c93146f/212c7f015314eebf?lnk=gst&q=aidl+parcelable#212c7f015314eebf Feb 26 23:52:11 perhaps you could just inline the contents of that file in your other AIDL files as a workaround Feb 27 00:04:12 hmm im wondering if i can use an ADC open source media player for my app? Feb 27 00:05:32 if its open source i should be able to modify it for my needs, but what about the challenge? i can't pretend i wrote the player, but its going to be a part of my app Feb 27 00:06:09 kurzum: are you adding some kind of extension to it or something? Feb 27 00:06:58 well part of my app will have to play media files, so ill need a media player, either self written or "borrowed" Feb 27 00:08:35 so im wondering if there are clear rules for such a case Feb 27 00:09:16 what music player is it? Feb 27 00:09:47 if it's MusicDroid you have my permissions :) Feb 27 00:10:04 oh i dont have any specific player in mind, but i remember seeing a couple of media players on code.google some weeks ago Feb 27 00:11:34 I think a lot of those players might wait until after the challenge deadline to release their stuff Feb 27 00:11:41 I haven't seen any release anything yet Feb 27 00:11:58 cool thanks! :) i need a video player tough... Feb 27 00:12:16 oh ok Feb 27 00:12:49 yeah i guess it makes sense they arent releasing their source just yet Feb 27 00:24:07 Hooray! The ADC submission app now has an entry for M5. Feb 27 00:27:33 hehe Feb 27 00:27:53 morrildl: some major technical hurdles with adding that option? :) Feb 27 00:34:07 kurzum: keep in mind that the challenge specifically states that your application must not be distributed illegaly. meaning, if you have tainted yourself with the GPL, then you must comply with the GPL. Feb 27 00:34:21 And if you do not, your application would be considered illegaly distributed, and would be ineligible to win the contest. Feb 27 00:35:05 so be very aware of the licensing terms of any software you use. Feb 27 00:57:05 Holy Smokes... Feb 27 00:57:29 Did anybody here go to Code Day? Feb 27 01:01:26 home early today :) Feb 27 01:02:11 Hey Jasta Feb 27 01:02:16 Whats up? Feb 27 01:02:53 nothing? Feb 27 01:03:04 I'm in Baltimore... not a developer but definitely Android obsessive: I run http://www.phandroid.com and hoping to interview some developers who attend the Code Day. Anyone interested? Feb 27 01:03:12 Wish I could have gone... guh Feb 27 01:03:14 ugh Feb 27 01:05:52 i did not go, sorry. Feb 27 01:06:11 have you checked the forums for folks impressions? Feb 27 01:09:10 what forums? Feb 27 01:09:29 the google groups topics. Feb 27 01:10:15 I've got it in my netvibes but didn't see anything specifically Code Day related Feb 27 01:10:25 are you sure, because i did. Feb 27 01:10:40 Hmmm... okay I'll take a closer look. Thanks for the idea Feb 27 01:27:42 trigatch, was that the boston code day? i'm asking because i'm also in the baltimore area Feb 27 01:28:23 some guy posted about the boston code day on android groups and wasn't too happy about it Feb 27 01:35:38 meh. i got that guy beat. i did 3000 miles to his pathetic 1200 :-P (for the campfire) Feb 27 01:54:54 skicson: what did you really expect from that thing? Feb 27 01:55:29 i mean, sure i'd like to go to one if it's nearby, but it's not a huge deal :) Feb 27 01:55:38 it's not like they're gonna reveal anything they haven't already told the web. Feb 27 01:56:19 it'd be cool to just checkout the Googleplex Feb 27 01:56:32 yeah, but the googleplex only is in mountain view Feb 27 01:56:59 yeah, campfire was in mtn view Feb 27 01:57:06 oh Feb 27 01:57:12 still, not that cool Feb 27 02:25:41 sweet, got my Android build environment in Maven now. this has simplified everything :) Feb 27 02:25:50 i had 4 separate projects before. Feb 27 02:26:18 well i still do, but now they're not hardcoding paths ;) Feb 27 02:34:52 never understood the point of maven when there is ant Feb 27 02:34:59 then again I never really looked at maven Feb 27 02:35:03 Marv|LG: Maven does much more than Ant Feb 27 02:35:15 Especially when it comes to managing your dependencies Feb 27 02:36:45 Ant fails very critically at trying to understand its environment and how the building project fits into it. Feb 27 02:37:16 Ant is analogous to make in that sense; make is totally dumb and unaware of the host. Feb 27 02:38:02 Which is why almost no one writes straight make files and expects their project to build anywhere but their own host computer Feb 27 02:39:12 Maven makes an effort to fix that, and from what I have learned today does so with mixed success. Ultimately, it's still better than Ant :) Feb 27 02:40:01 sounds like a linux vs windows arguement to me ;) Feb 27 02:40:10 The dependency management seems like it was narrowly implemented, adhering to the specific policies found in Apache projects and not necessarily everywhere else. Feb 27 02:40:22 Marv|LG: If it does, then you have misunderstood me completely. Feb 27 02:40:56 the basis of that argument is that each has it own set of plus/minuses Feb 27 02:41:41 Ant encourages developers to design monolithic applications, not stopping to separate into smaller, re-usable (as in, separately distributed) components. Feb 27 02:42:19 If you want to do so, for example, you have to considerably increase the complexity of your ant environment, certainly falling back to creating something very custom, very unconventional, and very crappy :) Feb 27 02:42:29 just like people did with Make :) Feb 27 02:43:49 jasta: it's not that hard to do with ant Feb 27 02:44:35 romainguy: Oh? How might you do it then without requiring that the packages are either delivered or built together? Feb 27 02:46:51 Try to imagine that the packages are truly separate, but both written by the same authors. The user may have one without the other already. How would you know that? The only way to deal with that case is to ignore it, which will cause distribution managers to shoot you in the face. Feb 27 02:48:11 Java developers just plain don't think like this, and that's because they're happy to just "make it build", and rarely consider more advanced systems of distribution. Probably because Windows doesn't even have one :) Feb 27 02:48:41 show server Feb 27 02:48:45 oops sorry Feb 27 02:49:08 jasta: Java WebStart works that way actually Feb 27 02:49:21 jasta: and you are only thinking with a Linux mind right now :) Feb 27 02:49:50 It's not my fault that Debian owns you all :) Feb 27 02:50:06 I personally hate Linux packages based distributions Feb 27 02:50:15 I much rather prefer the monolithic .app of Mac OS X for instance Feb 27 02:50:19 all Linux distributions are package-based. Feb 27 02:50:26 I know Feb 27 02:50:28 even Slackware and Gentoo Feb 27 02:50:29 and I hate them all Feb 27 02:50:54 I would imagine that you do. The monolithic approach is great for users. Feb 27 02:51:00 But it's terrible for everyone else. Feb 27 02:51:07 That was my earlier point. Feb 27 02:51:09 Which is pretty much nobody :) Feb 27 02:51:15 Gah! Feb 27 02:51:29 There is harmony that exists, and it is called APT. Respect it :) Feb 27 02:51:51 Yeah yeah I know Feb 27 02:52:12 if only I got a penny every time a Debian user mentioned apt... :) Feb 27 02:52:48 The reality is that it stands out as a such a great concept because it is hard to imagine why it is so rare. Feb 27 02:53:21 It is truly difficult to understand why Windows, for example, has no package management. Every application just throws together it's own piece of crap updater and the user is continually lost in the mess. Feb 27 02:53:41 Then, it pretends like it has such a thing with the 'Add/Remove programs' list which just heightens the user's confusion. Feb 27 02:55:33 hmm, my raid node is so loud i can hear it in my bedroom now. Feb 27 02:59:49 romainguy: You gotta think like Reagan here :) Feb 27 03:00:17 trickle-down economics. if developers are happy, more and better software will get written and users will be happy :) Feb 27 03:00:29 (disclaimer: no, i don't seriously believe that) Feb 27 03:23:37 romainguy: you ever been to ireland? Feb 27 03:24:00 i'm going in september, but i'm not sure what my exact itinerary is yet Feb 27 03:59:39 jasta: your animation bug has been fixed since M5 alrady Feb 27 03:59:41 already Feb 27 04:00:49 excellent, but didn't you already tell me that? :) Feb 27 04:01:15 no I told you I was pretty certain it was fixede Feb 27 04:01:19 but I had to check Feb 27 04:01:20 hehe Feb 27 04:01:22 oh Feb 27 04:01:24 I did check and it is fixed :) Feb 27 04:01:34 thanks for following up then :) Feb 27 04:01:55 you're welcome Feb 27 04:03:55 my stupid custom cursor has all sorts of data change bugs i think hehe Feb 27 04:05:16 i assume all i do is just call registerDataSetObserver() on the wrapper cursor and pass along onChanged and onInvalidated to super.notifyDataSetChanged and notifyDataSetInvalided, right? Feb 27 04:05:26 wrapped* Feb 27 04:05:50 should work yes Feb 27 04:06:08 * jasta tries that Feb 27 04:32:01 romainguy_: hmm, another bug :\ Feb 27 04:34:14 just reproduced it in the apidemos too Feb 27 04:34:41 scrolling very quickly through listviews that have rows of differing sizes seems to let you "offset" the listview from its original layout. Feb 27 04:35:02 such that it appears to have a top padding relative to its parent until you click on it and it is then reset Feb 27 04:35:16 known? Feb 27 04:35:39 scrolling with the mousewheel makes it easy to show. Feb 27 05:14:20 jasta: yeah we fixed that already Feb 27 05:28:10 excellent, got my adapter to stop throwing exceptions :) Feb 27 05:28:38 and i implemented one of the most lacking features in nearly every digital music player in the world Feb 27 05:29:25 a mundane unique name comparator. Feb 27 05:29:42 now, "The Arcade Fire", and "Arcade Fire" are the same thing :) Feb 27 05:29:51 "Minus the Bear" and "Minus The Bear". Oh what a novelty! Feb 27 05:30:52 i'd also like to implement a locale-independent comparison as well, but not sure how just yet. Feb 27 05:31:01 that's a tricky area to work in Feb 27 05:45:57 iTunes does it well ^^ Feb 27 05:46:02 (most of the time) Feb 27 05:46:12 yep Feb 27 05:49:02 gah, no it doesn't Feb 27 05:49:13 i hate my girlfriend's ipod because there's shit like that everywhere Feb 27 05:49:19 i also hate that she doesn't organize it better Feb 27 05:49:23 but thats probably itunes' fault too Feb 27 05:50:47 yes it does jasta Feb 27 05:51:35 really? i'm gonna look more closely next time i have that thing Feb 27 05:51:51 John butler trio Feb 27 05:51:56 The John Butler Trio Feb 27 05:52:00 is the ordering Feb 27 05:52:07 er Feb 27 05:52:10 the ordering? Feb 27 05:52:11 John Butler Feb 27 05:52:15 The John Butler Trio Feb 27 05:52:20 John Butler Trio Feb 27 05:52:22 as an example Feb 27 05:52:32 order by artist and thats what you get Feb 27 05:52:40 i'm not talking about ordering Feb 27 05:52:58 ? Feb 27 05:53:07 grouping by artist? Feb 27 05:53:15 i'm actually talking about the comperator that decides if "The John Butler Trio" and "John Butler Trio" are the same. Feb 27 05:53:25 ah Feb 27 05:53:26 soundex Feb 27 05:53:32 no, not soundex, just obvious. Feb 27 05:53:33 metaphones Feb 27 05:54:09 John Butler and John Butler Trio of course will be different, but come on. "The" is optional. Feb 27 05:54:19 right so you remove the Feb 27 05:54:30 soundex would be cool Feb 27 05:54:35 and compare using and and & as similar. Feb 27 05:54:37 but something better than soundex Feb 27 05:54:38 soundex would be too unreliable. Feb 27 05:54:45 plenty of things match soundex when they aren't the same. Feb 27 05:54:55 metaphone is pretty decent Feb 27 05:55:05 also, it's very important to me that locale-sensitive compares are supported, although they are not yet. Feb 27 05:55:18 right Feb 27 05:55:33 such as for example "Bjork" versus the umlauted version. Feb 27 05:55:49 they specify the same artist, and should match in that sense. Feb 27 05:56:08 meta data is so sloppily maintained by most users that the software they use needs to be pragmatic. Feb 27 05:56:37 most software i have used doesn't even do the simple case-insensitive comparison for crying out loud Feb 27 05:57:32 i am struggling with the decision to make artist, album, and song title a unique index as well. many people (thanks to piece of crap itunes) have multiple copies of the same albums in different formats. Feb 27 05:57:54 so showing them multiple versions helps no one, however on the other hand some artists legitimately duplicate song titles ("unknown" in particular) Feb 27 05:58:12 so it would be necessary to show that. but then it gets overly mucky to try to look at the tracknumber meta data bla bla Feb 27 06:07:06 jasta: gah, no it doesn't << my iTunes definitely doesn't take into account stuff like The for sorting Feb 27 06:07:21 If I want to jump to The Killers in my list I just hit the K Feb 27 06:07:34 i'm not referring to sorting. Feb 27 06:07:47 does it think killers = Killers? Feb 27 06:08:05 let me check that Feb 27 06:08:12 (since my tags are all proper and clean... :) Feb 27 06:08:23 or Tom Petty & The Heartbreakers = Tom Petty And The Heartbreakers? Feb 27 06:08:41 Or "The Killers - " = "The Killers"? :) Feb 27 06:08:41 yeah it does show as the same artist Feb 27 06:08:52 I don't know how far they went but it works for The Feb 27 06:09:21 well The is the obvious one that everyone is going to support because it would look absurd to sort it with the T's. Feb 27 06:09:38 but the actual fuzzy matching seems to rarely get supported well or at all. Feb 27 06:09:47 locale-independent matches especially Feb 27 06:10:02 normally, you want locale-sensitive comparisons, but this is one of the rare cases that you very certainly do not. Feb 27 06:10:41 that plus the fact that two names that look alike could legitimately be really different :) Feb 27 06:10:59 * romainguy_ just makes sure the tags are correct, it's much easier Feb 27 06:11:27 yes, and so do i of course Feb 27 06:11:38 but really, most people couldn't care less and assume that their software should handle it for them Feb 27 06:12:32 "I feel that i am a very good Java programmer but learning the language Feb 27 06:12:32 of Android is foreign." Feb 27 06:12:39 interesting post on the newsgroup ^^ Feb 27 06:13:11 They probably didn't literally mean langauge. That phrasing tells me they are not native English speakers. Feb 27 06:13:26 Oh you think? :) Feb 27 06:13:45 Yes ;) Feb 27 06:14:14 Still, a lot of people seem to complain that Android is somehow difficult to learn, which seems odd to me. Especially those that assert they are "very good java programmers" Feb 27 06:14:34 Well he goes on saying that he knows that in Java to paint on screen you have to override the paint() method Feb 27 06:14:47 which is actually untrue if you want to do it the right way Feb 27 06:15:10 I would have to imagine that any seasoned programmer could cut through Android like butter. In any language. Feb 27 06:15:40 I'm actually glad to see we attract all sorts of developers Feb 27 06:15:54 You mean not just Java fanboys? Feb 27 06:15:58 attracting the average developer is vital Feb 27 06:16:01 I meant experience wise Feb 27 06:16:24 romainguy_: Ahh, I would have to agree. I am surprised by the range. Feb 27 06:16:58 ror cannot become mainstream Feb 27 06:17:01 There are a lot of people on the forums that would be better suited for less...restrictive programming :) Feb 27 06:17:28 you have to learn somehow Feb 27 06:17:36 i would like android to be like java.. not ror Feb 27 06:17:52 what is ror? i thought that was a typo at first :) Feb 27 06:18:00 ruby on rails Feb 27 06:18:03 jasta: I believe he must be talking about Ruby on Rails Feb 27 06:18:09 although I have no idea how it relates to Android :) Feb 27 06:18:15 me neither Feb 27 06:18:26 in terms of adaptability Feb 27 06:18:53 at the end of the day you want all kinds of developers flocking android Feb 27 06:18:59 You're stretching muth :) Feb 27 06:19:01 muthu* Feb 27 06:19:58 muthu: Well, you will find that many developers will be unable to work with the constraints Android places on them. Mobile development will be a stretch for the common Java developer. Feb 27 06:20:10 need not be Feb 27 06:20:16 Yes, it need be. :) Feb 27 06:20:31 agreed, current mobile development is difficult Feb 27 06:21:28 Mobile development on Android is no different. You have to be mindful of efficiency at all times. Even benchmarking how well your application performs isn't good enough because it might still be draining the battery very fast. Feb 27 06:21:55 correct.. and i'm no expert Feb 27 06:22:54 i'm just hoping android makes it easy for the end developers Feb 27 06:23:07 muthu: It won't. Feb 27 06:23:16 nothing is easy Feb 27 06:23:25 problem with making something to easy is you get a lot of crap Feb 27 06:23:36 muthu: Android cannot help you rethink your broken assumptions. You'll have to do that all on your own :) Feb 27 06:23:37 duey: how? Feb 27 06:23:42 well for example Feb 27 06:23:53 if you were to create an rss reader as an pi Feb 27 06:23:55 api* Feb 27 06:24:02 you would get 100 rss readers Feb 27 06:24:03 For example, Microsoft Access applications ;) Feb 27 06:24:07 rather than a couple of good ones Feb 27 06:24:08 exactly Feb 27 06:24:49 easy and innovative - too much to ask? Feb 27 06:25:27 i'm sure that's android goal Feb 27 06:25:30 not much innovation in doing all the work and letting users glue it togeather Feb 27 06:25:32 muthu: You're just missing the point. Feb 27 06:25:45 (in terms of getting an innovative development community) Feb 27 06:26:02 you don't want it to become the VB for mobile platforms Feb 27 06:26:10 why not? Feb 27 06:26:15 because VB is crap Feb 27 06:26:21 :) Feb 27 06:26:27 duey: and had and has lots of success :) Feb 27 06:26:33 yes thats true Feb 27 06:26:47 romainguy_: Oh quiet, you know damn well you don't want Android to be like VB :P Feb 27 06:26:51 if one of the methods in my IBinder implementation throws, and i don't catch it, does the service die or does the activity die? Feb 27 06:27:01 jasta: if it could be as successful, why not? Feb 27 06:27:06 in other words, is the exception translated over ipc? Feb 27 06:27:07 as long as they are great applications... Feb 27 06:27:09 it depends on what you define success Feb 27 06:27:09 muthu: Albert Einstein has a great quote that applies to precisely why Android cannot be made truly "easy". Feb 27 06:27:21 "Make everything as simple as possible, but not simpler." Feb 27 06:27:24 if you define it as hugely popular for mainly crap applications sure Feb 27 06:27:54 jasta: agreed with einstein Feb 27 06:27:58 Android has limits placed on it by the devices which will run it. No one can do anything to work around that. Code's gotta be written small and light, network protocol efficient, etc. Feb 27 06:28:17 The common developer just doesn't have the attitude that those things are important. Feb 27 06:28:42 there are 3 levels .. beginners, intermediate and advanced Feb 27 06:29:06 Don't be so narrow. There are no levels ;) Feb 27 06:29:19 basically i don't think you want everyone releasing there hello world apps Feb 27 06:29:19 jasta: case closed ;) Feb 27 06:29:47 duey: what's the problem? why not? Feb 27 06:30:01 because thats what VB is like Feb 27 06:30:24 and that's why they are successful Feb 27 06:30:38 well i personally don't call VB successful Feb 27 06:30:40 popular yes Feb 27 06:31:18 duey: and tremendously useful and used within companies Feb 27 06:31:29 where they do stuff a tad more complicated than hello world Feb 27 06:31:53 are you talking vb or vb.net Feb 27 06:31:57 my organization uses a very elaborate sales instrument with the desktop client coded in VB. Feb 27 06:31:57 either Feb 27 06:32:31 it also is a terrible piece of shit. Feb 27 06:32:40 it's really, really slow :) Feb 27 06:32:54 if companies are under such missmangagement to choose such a product then well Feb 27 06:32:57 they can keep it Feb 27 06:34:00 romainguy_: Let's face it, VB is popular because you can higher developers for cheap. The products typically reflect that. Feb 27 06:34:08 I never said the contrary Feb 27 06:34:10 exactly Feb 27 06:34:12 it's also easy to pick up Feb 27 06:34:15 easy to learn Feb 27 06:34:20 and well, easy to get the job done with it Feb 27 06:34:23 Yeah, and impossible to get right. Feb 27 06:35:22 VB is probably an ok "first language" Feb 27 06:35:26 BTW, you work for a very successful company that has much the opposite strategy. Software done right is very commonly easier to write and cheaper in the long run since it does not stagnate as easily. Feb 27 06:35:34 duey: No, it's poison :) Feb 27 06:35:47 jasta, yeah but lots of schools use it Feb 27 06:35:49 at least here Feb 27 06:35:58 That doesn't make it an acceptable choice. Feb 27 06:36:07 That just makes it a convenient one ;) Feb 27 06:36:08 jasta, better than no choice Feb 27 06:36:11 yes correct Feb 27 06:36:52 at any rate, i want to get back to android development ;P Feb 27 06:37:03 yeah i better start.. Feb 27 06:37:06 lol Feb 27 06:37:12 can't wait to see what you guys can come up with Feb 27 06:37:14 i am, oddly enough, optimizing my terribly inefficient code. Feb 27 06:37:21 romainguy_: Can? Psssh :) Feb 27 06:37:33 just need to figure out how to clean up these images.. Feb 27 06:37:39 what's duey and jasta doing? Feb 27 06:37:42 secret Feb 27 06:37:47 jasta: well so much boasting made me curious :) Feb 27 06:37:48 come on! Feb 27 06:37:58 me too :) Feb 27 06:37:58 I am working on a project to put lasers on various types of animals. Feb 27 06:38:03 waht the heck Feb 27 06:38:08 im doing ocr on photos Feb 27 06:38:08 By way of an Android phone taped to their head. Feb 27 06:38:27 and then a) able to take a pic of a business card and insert into address book Feb 27 06:38:35 b) take a pic of a sign and translate it into another language Feb 27 06:38:37 jasta: music album sorter :-D Feb 27 06:39:08 duey: nice Feb 27 06:39:45 yeah its my honours project Feb 27 06:39:45 muthu: I'm writing a media synchronization system that allows very efficient remote access to music, images, video, documents, etc stored on a home PC or some other repository. Feb 27 06:40:16 With a reference implementation of a progressive music player integrating last.fm services. Feb 27 06:40:45 By synchronization, I mean over-the-air. It is uncertain if I will even support tethered synchronization. Feb 27 06:40:50 how would i use it? Feb 27 06:41:02 muthu: Install a server on your computer, and then connect your phone to it. Feb 27 06:41:28 The phone will synchronize meta data (names of songs, albums, etc) and download the content associated on demand for streaming (and caching to your storage card). Feb 27 06:42:10 from my own server? Feb 27 06:42:11 For optimal usage, the user should have a large (1GB or more) storage card and frequent EDGE or 3G data service. Feb 27 06:42:14 muthu: Yes. Feb 27 06:42:39 ok.. so you are testing it with last.fm? Feb 27 06:42:40 im thinking of off loading the ocr onto a webservice rather than doing it embedded Feb 27 06:42:53 at least for initial versions Feb 27 06:42:58 romainguy_: btw, remarking on your previous comment, I am just trying desperately to hold on to my youthful idealism before I slip into my 30's :) Feb 27 06:43:09 lol Feb 27 06:43:28 jasta: get it done before you hit the dreaded 3O's Feb 27 06:43:34 muthu: Not really, last.fm is just a provider of extra value for the software. It's not an essential feature. Feb 27 06:44:15 last.fm is used for artist pictures, bio information, etc and possibly down the road a community aspect of the music player. Feb 27 06:44:30 hmm.. interesting Feb 27 06:44:33 And also playlist suggestions and "auto" mode. Feb 27 06:44:54 So the user could say choose an artist or album and just say "Whatever, pick music for me." Not unlike Amarok's dynamic playlist feature. Feb 27 06:45:03 ok.. so you write the server software too? Feb 27 06:45:08 Yes, already have. Feb 27 06:45:14 great Feb 27 06:45:34 romain: does android send exceptions over ipc? Feb 27 06:45:40 The server is written in C, and is written in such a way that it will be extended to other platforms using a more appropriate shell. Feb 27 06:45:50 So like a systray type implementation for Windows, a daemon for Linux, ... Feb 27 06:46:00 marvinalone: No, I believe that is currently unsupported. Feb 27 06:46:49 jasta: so i need to write my services such that weird calls can't kill them by throwing. Feb 27 06:47:26 marvinalone: I have no idea but it doubt it Feb 27 06:47:33 marvinalone: but again, don't take my word for it Feb 27 06:48:02 i just realized setting a widget disabled causes it to be slightly translucent ;) Feb 27 06:48:27 marvinalone: No, you don't, but for now you should be aware that they are not passed through. Feb 27 06:48:35 jasta: yep Feb 27 06:54:32 romain: what does cedric beust do? Feb 27 06:54:48 muthu: he works on Android Feb 27 06:54:49 always wondered what cedric role is in android Feb 27 06:54:58 romain: exactly what? Feb 27 06:55:02 we actually seat in the same office Feb 27 06:55:06 I cannot tell you :) Feb 27 06:55:11 :( Feb 27 06:55:19 the secrecy of google Feb 27 06:55:31 one get used to it Feb 27 06:55:42 i heard Feb 27 06:55:51 it was quite a shock transitioning from Sun to Google Feb 27 06:56:05 from blogging/writing books & articles/speaking at conferes/etc. to absolute silence :) Feb 27 06:56:20 you like it of course :) Feb 27 06:56:59 working at Google? Yes, I love it Feb 27 06:57:08 but I also miss being involved with the community Feb 27 06:57:16 it's just not my job anymore Feb 27 06:57:35 yeah.. need to be involved in the community Feb 27 06:57:38 that's the real fun Feb 27 07:22:09 i wonder what excactly emulator -qemu -smp 2 does ;) Feb 27 07:23:25 * jasta experiments Feb 27 07:39:51 romainguy_: you're pretty involved with the community here Feb 27 07:40:10 zhobbs: thanks :) Feb 27 07:40:21 it's just a different kind of involvement Feb 27 07:41:53 well, whatever it is you've helped us out a lot :) Feb 27 07:42:40 you probably want Dan's job right? Feb 27 07:42:57 No :) Feb 27 07:43:07 I'm very happy with what I'm doing Feb 27 07:43:15 I just miss doing that kind of stuff, that's all Feb 27 08:13:06 yes, this kind of involvement is lame and dealing with us harping on him for bugs and things he usually has nothing to do with :) Feb 27 08:13:28 and also our perpetual need to misunderstand his grand uniform design :) Feb 27 08:14:01 or more precisely you perpetual need to disagree with the design :) Feb 27 08:14:09 haha Feb 27 08:14:36 i like Android's UI design, personally :) Feb 27 08:15:45 i am a big fan of the APIs in general, actually. especially the IPC stuff, that is just spot on. Feb 27 08:16:00 there are a few things in the UI toolkit I would have done differently Feb 27 08:16:06 but most of it was in place when I joined Feb 27 08:16:22 but yeah, I'm overall really pleased with the APIs Feb 27 08:16:24 me too, specifically the concept of gravity and weight could be more elegantly replaced by packing parameters. Feb 27 08:16:30 especially considering it's for a cell phone Feb 27 08:17:05 well, I would also like to have the notions of preferred/minimum sizes for the views Feb 27 08:17:16 that would make some layouts easier to use/understand I think Feb 27 08:17:31 widgets do have a preferred size, but it doesn't act as a minimum i suppose. Feb 27 08:17:42 they don't really have a preferred size actually Feb 27 08:18:17 well, not so called, but they are offered the chance to set a concrete value when drawing using WRAP_CONTENT. Feb 27 08:18:51 yes, most of the time :) Feb 27 08:18:57 most of the time? Feb 27 08:19:18 let's say you have a horizontal LinearLayout A Feb 27 08:19:22 with a height of wrap_content Feb 27 08:19:30 it contains a View B of height 64 pixels Feb 27 08:19:38 and a LinearLayout C of height fill_parent Feb 27 08:19:53 if you run this, A, B and C all have a final height on screen of 64 pixels Feb 27 08:19:55 which make sense Feb 27 08:20:24 but if inside C you put two Views of height=0/weight=1, thinking they will share equally the 64 pixels, you obtain something very different Feb 27 08:20:33 suddenly A and B take up the whole screen Feb 27 08:20:37 oh yes, i see what you're saying. Feb 27 08:20:46 while it's actually the correct behavior it's not intuitive Feb 27 08:21:22 a minimum size actually lends itself to a more elegant packing solution anyway. Feb 27 08:21:36 it also makes writing layout a tad easier Feb 27 08:21:37 since "fill" could be redefined not to fill its parent, but rather to fill available. Feb 27 08:22:04 which is what weight is supposed to do Feb 27 08:22:18 i know, but i'm saying you could blend weight, gravity, and fill/wrap all into one much more elegant concept of packing. Feb 27 08:22:35 a necessary precursor would be what you're referring to as well. Feb 27 08:22:42 yep Feb 27 08:22:49 this would be much more like how GTK does things. Feb 27 08:22:50 mornin Feb 27 08:23:00 there's one layout I really really like in Java: http://www.miglayout.com/ Feb 27 08:23:16 it makes it really easy to describe very complex layouts very cleanly Feb 27 08:23:49 (it was inspired by another great layout called FormLayout) Feb 27 08:24:31 give me a second to look Feb 27 08:25:07 Pretty cool, I was doing some profiling in ListView and to scroll a ListView of 1,000 items, we allocate only 6 TextView Feb 27 08:25:36 yeah, i'm happy to see that android took UI scalability seriosuly. Feb 27 08:26:05 i just went up to my full scale test with my slightly optimized alphabet separator adapter and found very nice performance with i think 850 list items Feb 27 08:26:07 it's a daily source of pain and hard work :) Feb 27 08:26:21 and the next SDK will make it even better Feb 27 08:26:24 yes, but it's very critical. Feb 27 08:26:30 right now we recycle all the views in the same heap Feb 27 08:26:38 but we will use different heaps based on the view type Feb 27 08:26:53 yeah, that allowed me to write an adapter that scrolled through and filter 3000 + items pretty nicely Feb 27 08:26:56 in a listview you mean? Feb 27 08:27:00 yes Feb 27 08:27:15 that is what you were referring to the other night about having to create new views when the type doesn't match? Feb 27 08:27:20 yep Feb 27 08:27:38 just made my cursor only select 50 or so at a time Feb 27 08:27:41 that would be welcome indeed. Feb 27 08:27:45 because you will have more "normal" views than separators, the change to get a recycled separator when you need one are slim Feb 27 08:27:49 it's just an oversight Feb 27 08:30:01 522 rows, not 800. damn, i was more impressed with my adapter's performance when i thought it was larger ;P Feb 27 08:31:42 doh! Feb 27 08:31:46 @!# Feb 27 08:31:55 what? Feb 27 08:32:11 I thought my traces were really small Feb 27 08:32:40 and I just remembered that calling BufferedOutputStream.write(int) actually writes only the first low byte Feb 27 08:33:01 I was just missing 3/4 of my data :) Feb 27 08:33:06 hehe Feb 27 08:34:27 I have this pretty cool "profiler" I wrote Feb 27 08:34:39 it shows you the views tree Feb 27 08:35:03 and replays the measure/layout/draw/invalidate/requestlayout/caching calls Feb 27 08:35:16 by flashing the tree nodes in a different color Feb 27 08:35:20 nice Feb 27 08:35:26 really nice Feb 27 08:35:28 very very handy to optimize this critical code path Feb 27 08:37:07 right now I'm writing another one to show how Views are shuffled in and out of ListView Feb 27 08:37:15 makes debugging a lot easier Feb 27 08:37:29 ui programming is black magic Feb 27 08:37:42 and a pain to debug through a debugger Feb 27 08:37:47 yeah, definitely Feb 27 08:38:51 i had an idea a long time ago for a debugger that ran virtualized like valgrind which would be capable of writing a vectored core file so that you could step backwards through the debugger :) Feb 27 08:39:22 ah here you go Feb 27 08:39:27 I found a screenshot: http://jext.free.fr/tfandroid.png Feb 27 08:39:46 it's an old version Feb 27 08:40:07 seems designed for your 30" screen :P Feb 27 08:40:23 well, this was a worst case scenario Feb 27 08:40:40 this must be for the dialer :) Feb 27 08:40:45 lol Feb 27 08:40:49 that's the Home screen Feb 27 08:41:02 except with something like 42 apps in the grid Feb 27 08:41:03 hehe really? i figured it was the dialer for sure because of all those linearlayouts in a gridview :) Feb 27 08:41:14 didnt count them tho Feb 27 08:43:40 anyway Feb 27 08:43:42 time to go to bed Feb 27 08:43:44 good night Feb 27 08:43:45 me too :) Feb 27 08:44:11 i am so close to having a fully working sample. can't wait to post something online ;) Feb 27 08:52:19 romainguy: good morning Feb 27 08:52:27 you sleeped for 7 minutes! Feb 27 08:52:29 :P Feb 27 08:54:30 hehe :) Feb 27 09:11:11 bed time Feb 27 09:11:14 night folks Feb 27 09:47:48 morning Feb 27 09:52:24 oh no I am out of sync with the rest of us! Feb 27 09:52:31 too much coffee in the morning... Feb 27 10:18:30 think so acsia? Feb 27 10:18:55 it's 5:12am here, i haven't gone to bed yet and i just had a large coffee i scooped up from the gas station... hehe Feb 27 10:18:59 the life of a web developer.. ugh Feb 27 10:19:13 :P Feb 27 10:19:22 yes lovely isn't it? Feb 27 10:19:44 personnally I can only work early in the morning or late at night Feb 27 10:19:52 in between I am non productive Feb 27 10:20:13 probably the reason I have not progressed with my career for a couple of years Feb 27 10:23:05 so what are you working on? Feb 27 10:58:45 Good morning :) **** ENDING LOGGING AT Wed Feb 27 10:59:57 2008