**** BEGIN LOGGING AT Wed Nov 21 02:59:56 2007 Nov 21 03:32:56 has anyone experimented with loading raw resources from the apk? Nov 21 04:38:00 where in the API does it list the layout_width and stuff for a RelativeLayout Nov 21 05:10:37 I was wondering the same thing - there is a bunch of "general stuff" in the /docs/devel/ui/layout.html document, and if you click on the relativelayout link it also shows all the constants like POSITION_TO_LEFT etc too Nov 21 05:11:23 oh wait... here it is: /docs/reference/android/widget/RelativeLayout.LayoutParams.html Nov 21 05:12:50 hmm Nov 21 05:15:08 "hmmm" doesn't sound good Nov 21 05:22:59 http://pastebin.ca/792154 Nov 21 05:23:21 can somebody have a look at that and tell me why the only thing to appear on screen is the "chat" text view Nov 21 06:54:18 d03boy: Are you still there? Nov 21 06:55:25 yes Nov 21 06:56:04 From reading your layout, I am having trouble understand what you want it to look like. Can you describe what you are trying for? Nov 21 06:57:36 do you use mirc? Nov 21 06:57:40 it should look similar to that.. Nov 21 06:57:59 the chat window where all teh text shows up. the userlist on the right. the input box below everything Nov 21 06:59:07 Yea, I use mirc :) --- ok give me a sec, cuz what you have wont work :P Nov 21 06:59:12 ok Nov 21 06:59:18 i was just messing around. wasnt really sure what to do Nov 21 06:59:45 yea, I wish someone would create a designer... or at least release the schema Nov 21 07:00:39 ive found 1 or 2 tutorials that werent terrible.. but they really didnt say how to make complex designs Nov 21 07:05:57 ugh... I'm confused and I'm lazy... which is the layout that you can add weights to? Nov 21 07:06:33 the gravity? Nov 21 07:06:44 probably linear Nov 21 07:07:16 ya its linearlayout Nov 21 07:09:28 lol - cool thanks, I have everything on the screen... I am just working on the spacing now :) Nov 21 07:09:35 :) Nov 21 08:01:01 Nescio, give up? :P Nov 21 08:01:51 nope... 1 sec Nov 21 08:02:08 tesing it on a diff view, then I'll post back - k :) Nov 21 08:02:17 no problem, just curious Nov 21 08:06:27 boy do I miss drag and drop... Nov 21 08:07:17 and anchors... and docking... Nov 21 08:07:47 *sigh*... alright it's up... GL :) Nov 21 08:08:18 up where? Nov 21 08:08:29 your pastebin Nov 21 08:08:33 oh ok, thanks Nov 21 08:08:35 i'll check it out Nov 21 08:08:44 np - that was fun.. Nov 21 08:09:50 oh wow, thats great. thanks, I owe you one :) Nov 21 08:10:54 no problem... I dont think it's the best wway to do it, (relative would probably be best) but I'm still newb... so, yw Nov 21 08:11:36 you'll also probaly want to add scrolling Nov 21 08:13:25 ah yes, i'll figure it out though Nov 21 08:13:38 getting this started was difficult because there arent many good examples out there yet Nov 21 08:13:59 aye Nov 21 08:20:07 jasta: Are you here? Nov 21 08:25:27 morning Nov 21 08:25:32 hola Nov 21 08:27:44 Anyone know regex? Nov 21 08:28:27 Is there a way i can skip a line if it starts with a * ? Nov 21 08:30:21 if you have each line as a string you can just check with if(!line.startsWith("*")) Nov 21 08:30:42 if ypu dunno regex Nov 21 08:30:46 if you dunno regex Nov 21 08:35:13 if you really want to use regex... "^\**" Nov 21 08:35:17 i think that would match it Nov 21 08:35:33 ^ means the beginning of a line Nov 21 08:35:43 the \ should escape the * Nov 21 08:35:53 and then the following * is a wildcard for "anything" Nov 21 08:38:56 hmm, I want everything but the lines that start with * Nov 21 08:43:31 just use what cutmasta said :P Nov 21 08:45:02 LoL, I was trying to parse my IRC logs... I forgot to turn off Joins/Quits Nov 21 08:45:05 (?<=[<])(.*) works... Nov 21 08:45:29 thanks tho Nov 21 08:56:00 So, what is everyone up to? Nov 21 08:59:31 realizing its 3am and deciding i need sleep Nov 21 08:59:39 ..after committing changes to svn ofcourse Nov 21 08:59:40 hehe Nov 21 08:59:45 hehe Nov 21 08:59:53 cya later :) Nov 21 08:59:59 later skater! Nov 21 09:03:32 'morning Nov 21 09:05:02 mornin :) Nov 21 09:08:04 Does anyone else have a problem using the arrow keys on the Android search bar? Nov 21 09:08:21 http://code.google.com/android/ Nov 21 09:21:51 well...shit Nov 21 09:22:05 sup? Nov 21 09:23:29 http://groups.google.com/group/android-developers/browse_thread/thread/5565fcb18cd17199/3c8115aedb7c8767#3c8115aedb7c8767 Nov 21 09:23:45 I fear I need either an answer from Google or access to the source code:-( Nov 21 09:24:47 davidw: ah, so another 8-hour wait then Nov 21 09:24:56 hrmph Nov 21 09:25:19 I asked Dan about it yesterday, but no dice... and it was sleep time for me Nov 21 09:28:24 yea... out of my league, but good luck :) Nov 21 10:48:48 Hi everyone Nov 21 10:48:56 why do I need BRIT for the Android plugin? Nov 21 10:53:25 BIRT I mean Nov 21 12:32:28 bidi bidi bidi - what's up, Buck? Nov 21 13:53:35 <[AD]Turbo> hi there Nov 21 14:19:49 Ok, added the java chat for this channel to androidwiki.com ... Nov 21 14:35:01 hrm Nov 21 14:35:27 all View type elements need a layout_width and layout_height, right? Nov 21 14:38:24 davidw: not necessarily Nov 21 14:39:22 there are default values (IIRC), so if you don't set layout_width and layout_height, it'll go to those Nov 21 14:39:56 or at least that worked for me last week when testing a few things Nov 21 15:15:56 hehe, straz' domain name ought to have latitude and longitude just for good measure Nov 21 15:16:16 hey, it's rcn Nov 21 15:22:41 jeez, that is verbose. eww Nov 21 15:24:13 folks fyi - twitter client sample code - http://tinyurl.com/3y8w5k Nov 21 15:26:00 Nice work :-) Nov 21 15:28:45 hi dims Nov 21 15:30:14 hi davidw Nov 21 15:35:27 hello -- other than the resources posted at google, does anyone know of any android-oriented tutorials floating around out there? Nov 21 15:45:11 tis a little early for anything comprehensive Nov 21 15:46:25 dims: nice. btw, Handler.postAtTime() would save you the MyRunnable thread Nov 21 15:57:39 There are so many things wrong with the cellphones i've owned i can't wait to change that Nov 21 16:13:16 blootu1: thanks!!!! Nov 21 16:17:31 np Nov 21 16:17:54 mrphrhrargh... xml attributes are a lot clearer and simpler than most of the java methods...would be nice to be able to get/set them directly Nov 21 16:21:37 mm... I think that ViewInflate.inflate() has a bug. Inflated views loose parent context... :( Nov 21 16:37:36 hrm hrm hrm hrm hrm Nov 21 16:53:22 davidw: still waiting? Nov 21 17:26:16 interesting commentary, http://www.theregister.co.uk/2007/11/21/dalvik/ Nov 21 17:31:37 Stephmw: thanks :) Nov 21 17:32:00 Stephmw: did you submit it to slashdot? Nov 21 17:32:25 Stephenw: So Android can run j2me apps? Nov 21 17:32:33 i would like to see more people comment on it :) Nov 21 17:32:40 All my questions in this forum about j2me have been ignored. Nov 21 17:32:56 I didn't check back to see if they were answered in google groups. Nov 21 17:32:59 Interesting register article Nov 21 17:33:48 I'm getting the feeling that google employees are avoiding the issue due to legal reasons. Nov 21 17:35:25 Dralspire: the author just needs to run sed with s/OS/OSS on the article ;) Nov 21 17:35:37 lmao Nov 21 17:36:36 Lede: nah, I haven't given it to /. Nov 21 17:36:57 ozarka: the answer is pretty simple really Nov 21 17:37:11 ozarka: no J2ME APIs, so straight no Nov 21 17:37:27 ozarka: but if they did... still no, not without translating your j2me bytecode to dalvik bytecode Nov 21 17:37:35 Yeah. That's kind of what I thought. Nov 21 17:37:36 ozarka: so that means no OTA delivery Nov 21 17:37:47 ozarka: not unless someone else adds a proper java vm Nov 21 17:38:00 But I asked some more... thoughtful questions about j2me which were ingnore as well. Nov 21 17:38:32 I probably didn't see em, can you summarise? Nov 21 17:38:44 note: I'm not google-related Nov 21 17:38:45 So the register article is wrong. Nov 21 17:38:59 which bit is wrong? Nov 21 17:39:11 Register says that Dalvik can run j2me Nov 21 17:39:57 ozarka: it said the platform could Nov 21 17:40:03 Stephmw: I have to run, but I asked about porting strategies for j2me. Nov 21 17:40:05 ozarka: with appropriate vm Nov 21 18:17:44 re Nov 21 18:17:51 Can Android SDK answer phones? Nov 21 18:18:51 HMMmmmmm Nov 21 18:20:42 Hmmm Nov 21 18:20:57 mele-: Not real phone calls, but simulated ones yes. Nov 21 18:21:24 Android phone home! Nov 21 18:23:12 Simulated ones? Nov 21 18:23:34 would be fun if you coudl start two emulator and make them talk to each other virtually Nov 21 18:23:46 (not that i need it) Nov 21 18:27:07 damn... ((Button) view).setLayoutParams(new LinearLayout.LayoutParams(LayoutParams.FILL_PARENT, LayoutParams.WRAP_CONTENT)); Nov 21 18:27:37 vs android:layout_width="fill_parent" android:layout_height="fill_parent" Nov 21 18:28:17 * davidw really wants to be able to do something like myview.setAttr(attrname, attrvalue); Nov 21 18:28:55 um ... simulated or stimulated calls? :P Nov 21 18:30:01 :) Nov 21 18:30:15 what does that mean? A program in a real phone can't do it? Nov 21 18:31:07 i hope the new phones that come with android have quad core intels in them Nov 21 18:31:26 i hope the new phones that come with android have fans in them Nov 21 18:31:33 lmao Nov 21 18:32:04 at least 512 ram and i hope 8 gig storage Nov 21 18:32:13 fans ... Nov 21 18:32:16 ;] Nov 21 18:32:16 with 800mhz fsb Nov 21 18:32:34 and tv, mirror, fridge ... Nov 21 18:32:34 other wise it wont be able to play crysis Nov 21 18:32:35 :D Nov 21 18:32:38 I'd like a dvd player Nov 21 18:32:48 i bet they would show the tiny model of the phone on tv, and when you buy it there will be this power brick you need to carry around Nov 21 18:32:52 if it cant play crysis then its as bad as every other phone Nov 21 18:33:11 brick phones are cool Nov 21 18:33:21 they go with my mullet Nov 21 18:33:26 haha Nov 21 18:35:02 Google phone: http://news.bbc.co.uk/olmedia/1775000/images/_1778067_fonecash-bbc300.jpg Nov 21 18:35:23 lmao funny (that was my first cellphone) Nov 21 18:35:24 Lede: yeah Nov 21 18:36:15 brick phones are awesome they match your blue suit and tie Nov 21 18:36:37 But you look cool using it everyone sees it Nov 21 18:36:38 http://www.grooveking.com/blog/uploaded_images/g-738997.jpg Nov 21 18:36:57 Weeee!!! I like that movie Nov 21 18:38:00 it looks like you need massive hands for those things Nov 21 18:38:18 lol there was a time when I was driving this thing: http://www.handy-sammler.de/Handys/Bosch_se_of_71.htm Nov 21 18:39:05 (box goes into the trunk, handset to the in-car console :D) Nov 21 18:39:48 uhhhm Nov 21 18:39:57 i have a 1980's mobile "car phone". Nov 21 18:40:00 like, it's downstairs. Nov 21 18:40:20 it comes in it's own little briefcase with a handset, giant battery, and a pen/paper holder. Nov 21 18:40:33 Dralspire: that thing looks like a defibrillation ;p Nov 21 18:40:46 i thought about taking an old cell phone i have and soldering it inside, replacing the car phone electronics and actually making it work. Nov 21 18:40:49 Lede: good idea Nov 21 18:40:52 lol and an anti-carjacking device in one unit ... 12 pounds (5kg :P) Nov 21 18:41:20 uhm i meant to say defibrillator Nov 21 18:43:46 jasta: that would be awesome Nov 21 18:44:00 jasta: make it the first android phone! Nov 21 18:44:17 lol oh the display clarity Nov 21 18:44:26 so canada should have an android phone by say 2009 Nov 21 18:44:33 xbit: uhm, the LCD on it is a 1-line character display screen. Nov 21 18:44:59 osearth: and the US will have one by, say end of 2008? Nov 21 19:05:14 Hi, someone knows if there will be unlocked handsets with Android? opensource? Nov 21 19:06:11 someone probably has an idea, but it's not us Nov 21 19:06:18 if it's going anywhere else than us it will definitely be unlocked Nov 21 19:12:05 I doubt it Nov 21 19:12:10 oh, hand on Nov 21 19:12:12 *hang on... Nov 21 19:12:13 misread Nov 21 19:12:31 unlocked, very likely worldwide, except US.... hackable, probably not Nov 21 19:12:53 thomasd: problems with your connection? Nov 21 19:14:02 as far as I know the handset can be locked in hardware (the GSM chip),... By lock I mean, the handset doesn't allow all SIM Nov 21 19:15:35 yeah Nov 21 19:17:06 edu: right, it's just not all that common outside the US Nov 21 19:17:28 edu: dirt-cheap PAYG handsets are one market segment where it's seen, for instance Nov 21 19:17:28 Well, most prepaid plan phones are locked, aren't they Nov 21 19:17:40 haha happy icecream :-) Nov 21 19:17:47 ;) Nov 21 19:17:56 * Stephmw should type faster Nov 21 19:18:18 * Lede has motorla f3 Nov 21 19:18:27 Stephmw: all countries has locked phones, it's the way they can offer good plans to people Nov 21 19:18:32 most basic phone ever :) Nov 21 19:18:46 edu: sure, but you have to keep segmentation in mind Nov 21 19:19:06 edu: android won't even run on the dirt-cheap hardware where locking is common Nov 21 19:19:16 lmao enter the IKEA Android phone ... kit + solder iron :D Nov 21 19:19:32 heh Nov 21 19:19:33 stephmw: so there will be unlocked handset? very expensive? Nov 21 19:20:00 edu: that completely depends on who makes them - Google hasn't really been that clear on this point Nov 21 19:20:09 it's illegal to not offer unlocked options somewhere, look at the iphone in germany Nov 21 19:20:13 edu: just figure your usual smartphone prices Nov 21 19:20:16 you can get that for just $1500 ;) Nov 21 19:20:40 well, it's not the lock part that was petitioned to the court, it was the exclusive offer on one network Nov 21 19:21:43 okay, but that's just a step to lock applications uploading also... I mean, providers will develop there application and not users... may be? Nov 21 19:22:32 It's really all speculation at this point :-) Nov 21 19:23:04 aye, and there's all kinds of different 'locking' aspects, hardware is just one Nov 21 19:26:09 yeah... so my last question: android will be better than openmoko? Nov 21 19:27:57 edu: yes/no Nov 21 19:28:00 :) Nov 21 19:42:08 salad Nov 21 20:07:14 Hehe I'm so pleased with myself getting that chat plugin running :D Nov 21 20:12:04 hi Nov 21 20:12:50 Hi there :-) Nov 21 20:13:41 Dralspire: a chat plugin? Nov 21 20:13:53 for eclipse or what Nov 21 20:14:09 Yeah, on androidwiki.com ... had to make a new extension because the existing ones don't work ... to this fine channel here Nov 21 20:14:28 That's where AWiki001 etc. users come from Nov 21 20:33:06 question: Anyone tried Translucent View demo? Nov 21 20:35:32 hey there, anyone having a phone running already a linux kernel (for OpenMoko, or Opie, GPE, etc.) thought or tried already of running android on it? Nov 21 21:18:11 is andtriod a good way to start learnign programming ? Nov 21 21:18:23 no Nov 21 21:18:27 it is not a book or anything like that Nov 21 21:18:46 is it this like javascript tho Nov 21 21:18:47 huh? Nov 21 21:18:53 no it is not Nov 21 21:18:55 android is an operating system Nov 21 21:19:00 i mean if i wanna go get a book Nov 21 21:19:07 there is no book for android, yet Nov 21 21:19:08 what book should i get Nov 21 21:19:10 you will find for java Nov 21 21:19:15 there is android Nov 21 21:19:18 and there is dalvik Nov 21 21:19:29 if you want to know about android os, get a book for linux kernel Nov 21 21:19:49 as-if Nov 21 21:21:19 anybody has tried InflateView.inflate()? Nov 21 21:21:37 why? Nov 21 21:23:23 passing parent param always throw stackoverflow exception and returned view has diferent context Nov 21 21:23:25 :S Nov 21 21:24:42 well if it's throwing an exception it's guaranteed its internal state is buggered ;) Nov 21 21:26:23 yeap... I can use null for parent param... but I cant change view context Nov 21 21:37:19 I get an error message on startup: The application com.google.android.providers.media is not responding. Nov 21 21:40:18 It is when I start the demo on youtube tutorial Nov 21 22:23:25 Is there a way to shut down the emulator with a button instaed of alt+f4? Nov 21 22:25:42 lnxnt_, the one on the front of your computer ought to do it Nov 21 22:26:05 otherwise, the one in the upper left hand side Nov 21 22:26:15 it's sort of hard to see Nov 21 22:26:28 hrm Nov 21 22:26:35 doesn't seem to actually work though, either Nov 21 22:26:58 davidw: hehe Nov 21 22:27:20 davidw: yes, the left one in the upper corner seems to be the one but it doesn't work. That's poor Nov 21 22:27:52 Also i have no red "running" button in my eclipse console, maybe that's an eclipse issue Nov 21 22:29:23 * davidw just uses the command line Nov 21 22:30:13 davidw: you don't work with eclipse? Nov 21 22:30:59 * davidw uses emacs Nov 21 22:33:49 the power button works fine for me? Nov 21 22:36:06 haavi: how does it work for you, does the application close? Nov 21 22:36:19 I press it and get a dialog with three buttons Nov 21 22:36:31 haavi: yes. Nov 21 22:36:36 Turn Off Screen, Turn Of Radio and Power Off Nov 21 22:36:47 haavi: yes, that's correct. Nov 21 22:37:03 yea Nov 21 22:37:17 haavi: that's not all to it. Power Off should close the app Nov 21 22:37:44 should it Nov 21 22:37:57 I don't get it. Two projects created, with two different emulator configurations and on both I have my contact list I created with the first emualtion? Nov 21 22:38:11 only works one time though Nov 21 22:38:20 lnxnt_: yes Nov 21 22:38:27 anyone knows the git command to sync (update) an already cloned dir? Nov 21 22:38:31 lnxnt_: the emulator uses ~/.android/userdata.img Nov 21 22:38:51 that is where the "on-device" data lives. Removing that file is equivalent to wiping the device to restore factory settings. Nov 21 22:39:18 because it's a file, it needs to be locked. The 2nd emulator will have read-only access to that file Nov 21 22:39:26 morrildl: so there's no use in creating two different configurations besides the emulator screen size and some other things.. Nov 21 22:39:38 lnxnt_: currently, that's correct Nov 21 22:39:49 ah ok :) Nov 21 22:43:34 I had a weird thing happening earlier today where I couldn't create new databases, only solution was to delete userdata.img :/ Nov 21 22:43:52 and that userdata.img.lock folder Nov 21 22:44:32 hey where's the proper place to define new custom action strings for Intents? just constants in any class or is there some convention on where? Nov 21 22:45:04 cybereagle: wherever you like Nov 21 22:45:17 ok thanks Nov 21 22:45:43 nevermind, found it, just had to run insinde that dir Nov 21 22:45:55 swetland: ping Nov 21 22:46:37 heehe, that's the proper way to get someone's attention in here? Nov 21 22:47:01 it's just common IRC channel practice Nov 21 22:47:11 I had no idea :) Nov 21 22:47:33 usually, the "pinged" one responds at some point, with "pong" Nov 21 22:50:03 does that make any sound? Nov 21 22:51:57 only on the really cool IRC clients Nov 21 22:52:54 To clean an application from the emulator, I have to delete that userdata.img then? Nov 21 22:52:56 lnxnt_: don't know if it does in some particular client, but someones name spelled exactly usually blinks the item in the taskbar,, highlights, and so on Nov 21 22:53:43 I don't want to delete that because I have contacts there ;) Nov 21 22:53:50 lnxnt_: some clients beep when you address someone directly ... Nov 21 22:53:56 no I don't think so, I delete apps from the shell Nov 21 22:54:12 haavi: how? Nov 21 22:54:15 therefor the ping is just a noop that you place behind someones nick to get their attention Nov 21 22:54:24 run adb shell Nov 21 22:54:39 and then go to the apps folder and delete the apk file for your app Nov 21 22:54:53 Essington: I see, I just thougth maybe that ping text makes noise ;) Nov 21 22:55:06 adb is in the tools folder btw.. Nov 21 22:56:29 it could if you set "ping" as a trigger word in a client that supports it heh Nov 21 22:57:38 haavi: wow that's cool Nov 21 22:57:44 haavi: it's a small linux then? Nov 21 22:58:14 yeah Nov 21 22:58:43 your app should be in data/app/ I think Nov 21 22:59:08 why do you guys think I asked then, if somebody already having a phone booting linux, tried android on it Nov 21 22:59:35 but it has to be run on a specific chipset I think Nov 21 22:59:45 I don't think it would compile Nov 21 23:00:01 haavi: the application file (.apk) will be in /data/app Nov 21 23:00:25 haavi: the data files created by the application when it runs are generally in /data/data/ Nov 21 23:00:51 yeah that's right sorry Nov 21 23:03:26 what CPU speed is the emulator running at? Nov 21 23:05:35 haavi: once the chipset is supported by a linux kernel (boots), that should abstract the hardware enough Nov 21 23:06:44 morrildl: do you have one of these Gphones that the emulator emulates and that we can see in the presentation video? Nov 21 23:07:19 I still geht errors on startup like The application com.google.android.providers.media is not responding. The application com.google.android.providers.contacts is not responding. Nov 21 23:07:40 haavi: you mean in hardware? Nov 21 23:08:06 yeah Nov 21 23:08:10 no Nov 21 23:08:30 ok, but he's a Google employee right? Nov 21 23:08:56 the presentation guy? Nov 21 23:09:05 yeah, I think so Nov 21 23:10:58 I supose they have to hold on to them pretty tight at Google Nov 21 23:11:23 haavi: there is no "gPhone" Nov 21 23:12:00 haavi: There are indeed some development devices we have internally, and one of them was shown in the video, and that is indeed the one that the emulator is modeled to look like Nov 21 23:12:01 but "that phone" we can see in the presentation Nov 21 23:12:35 ok, so do YOU have one? Nov 21 23:12:52 haavi: if I need one, I can get my hands on one, yeah Nov 21 23:13:12 don't you want to check it out? Nov 21 23:13:16 and when you do, how can you resist the urge to constantly play quake on it? :P Nov 21 23:13:24 morrildl: does that one has anything in common with the halibut evaluation board? Nov 21 23:13:26 cybereagle: lol Nov 21 23:13:35 cybereagle: ever tried playing quake on a d-pad? Nov 21 23:13:40 Zoolooc: I don't know the term "halibut" Nov 21 23:13:41 cybereagle: it's bloody hard! Nov 21 23:13:57 Zoolooc: we have various names for our development boards and devices, but that's not one I've heard Nov 21 23:14:19 * Stephmw readies a dictionary attack on morrildl :) Nov 21 23:14:24 what about the Qualcomm's msm7200A chipset? Nov 21 23:14:29 Zoolooc: (which is to say, I'm not one of hte hardware guys on this project :)) Nov 21 23:14:47 * jasta dropkicks Android Nov 21 23:14:53 the one for which the kernel was published in the git repository Nov 21 23:15:08 BitmapFactory.decodeStream apparently cannot decode .ico files, as in favicon.ico. Nov 21 23:15:18 morrildl: ok, I get it Nov 21 23:15:21 Zoolooc: what about it? That is indeed what we're using, etc. Nov 21 23:15:29 Zoolooc: doesn't you have to compile the kernel on a device that has that specific chipset? Nov 21 23:15:30 Zoolooc: was there a specific question...? Nov 21 23:15:51 haavi: not if you have an appropriately-built cross-compiler Nov 21 23:15:57 if that phone the guys here where asking about, use this chipset Nov 21 23:16:06 * morrildl 's favorite cross-compiler was the one that built binaries for the Lego Mindstorms RCX Nov 21 23:16:11 ahh, those were the days Nov 21 23:16:23 I wish I still had enough free time to spend hours and hours building cross-compilers for toys :) Nov 21 23:16:29 morrildl: what are the supported image formats for the BitmapFactory? Nov 21 23:16:37 :) Nov 21 23:16:48 jasta: should be PNG, JPEG, GIF. Apparently not ICO. :) Nov 21 23:16:49 (and can you add to the supported formats?) Nov 21 23:16:56 No, apparently not :) Nov 21 23:17:24 also, what Stephmw said. Nov 21 23:17:27 Zoolooc: good question, I don't really know the answer. swetland might Nov 21 23:17:43 guess he does, but thanks :) Nov 21 23:18:08 Stephmw, jasta: at runtime, probably not. It will definitely be possible to add support at the system level Nov 21 23:18:23 morrildl: ok Nov 21 23:18:25 * Stephmw makes a note Nov 21 23:19:04 morrildl: Well, hmm. I loaded a favicon.ico to my res/drawable folder and called iconview.setImageResource(R.drawable.favicon); and it worked. Nov 21 23:19:05 i.e. the usual sense of "find an implementation of a .ico codec that can be ported to ARM, integrate it with the libraries and framework, etc." Nov 21 23:19:18 jasta: oh? hmm. interesting Nov 21 23:19:23 morrildl: I know this isn't your problem, but have thoughts been given to ensuring one of the shipped devices next year isn't locked to hell and back by the manufacturers? Nov 21 23:20:07 morrildl: So riddle me that. I am simply using BitmapFactory.decodeStream(new URL("http://www.smartphonethoughts.com/favicon.ico").openStream()); Nov 21 23:20:11 Stephmw: I actually don't know a whole lot about what the shipping devices will look like Nov 21 23:20:42 In fact, it looks like no excetpion is being thrown. It just returns null. Nov 21 23:20:51 Stephmw: when we say that right now the SDK is about getting feedback about the APIs, we really do mean it :) Nov 21 23:21:15 As time goes on, additional announcements will be made that will clear up questions like that Nov 21 23:21:28 morrildl: it's a shame that for most of the people I speak to in R&D currently evaluating the platform, that's currently a detail Nov 21 23:21:48 morrildl: the worry is about investing time & money now in something that they can't really use later Nov 21 23:22:14 Stephmw: yeah I understand Nov 21 23:22:32 We can talk and swear up and down and promise and talk and say that it'll be open-source and all that Nov 21 23:22:41 But until it is, it's a matter of trust Nov 21 23:22:44 * Stephmw nods sadly Nov 21 23:22:59 We understand that Nov 21 23:23:24 it's not actually a Google thing, the mobile industry is just scarred and battered... and really doesn't trust much Nov 21 23:23:31 So while I know it probably won't make much difference, I'll say it again: this WILL be open-sourced Nov 21 23:23:33 all of it Nov 21 23:23:42 * Stephmw tickles morrildl Nov 21 23:23:46 Dalvik, SGL, the libc implementation, the application framework, etc. Nov 21 23:24:06 morrildl: that's a wonderful part of the problem, and I honestly believe you Nov 21 23:24:12 heh heh :) Nov 21 23:24:21 There's believing, and then there's betting real money on it :) Nov 21 23:24:25 it's the other half that's a cause for concern Nov 21 23:24:25 the emulator takes time to boot... Nov 21 23:24:38 Stephmw: the other half of what? Nov 21 23:24:50 morrildl: the other half being manufacturers and operators Nov 21 23:24:53 ahh Nov 21 23:24:55 Right Nov 21 23:25:23 morrildl: historically, you must have seen what happens to other open platforms once an operator or manufacturer gets its hands on it Nov 21 23:25:45 Stephmw: oh yes, definitely Nov 21 23:25:49 morrildl: so when the PR talk about replaceable call control or other main apps... well Nov 21 23:26:00 morrildl: it's great for the OEMs, ODMs and Ops Nov 21 23:26:00 Stephmw: eh well, I've done that Nov 21 23:26:09 replaced the dialer I mean Nov 21 23:26:18 * Stephmw nods Nov 21 23:26:21 just for the hell of it, really :) Nov 21 23:26:30 morrildl: I did the same thing on SavaJe Nov 21 23:26:40 morrildl: in fact, I removed the whole User Experience Nov 21 23:26:45 ha ha :) Nov 21 23:27:01 I don't get the stuff related to chipset...is it not possible to run android on every phone then? Nov 21 23:27:33 lnxnt: it will be possible to run Android on any phone to which it's been ported, and which has a route to physically replace the operating system Nov 21 23:27:58 lnxnt: it is an unfortunate truth that many devices sold to consumers don't have, for example, a USB mechanism to re-flash the OS Nov 21 23:28:08 ah Nov 21 23:28:46 That is what Stephmw is referring to -- the OS can be as open-source as we want it to be, but if a given carrier modifies the OS and locks it down, and then doesn't give you a way to update your device's OS, then you are still out of luck Nov 21 23:28:50 i wonder how long before someone tries to get it running on an iphone :P Nov 21 23:29:28 morrildl: We simply don't think that will happen. Nov 21 23:29:38 morrildl: incidentally, how would we find out if Google is planning to come to FOSDEM in february? Nov 21 23:29:39 if a carrier wanted to do that i cant see them wanting to run android in the first place, openness is pretty much its point from what i can see Nov 21 23:29:46 * morrildl is talking to himself apparently Nov 21 23:29:49 haha Nov 21 23:29:56 Stephmw: you would ask me, actually :) Nov 21 23:30:00 what is FOSDEM? Nov 21 23:30:12 http://www.fosdem.org/2008/ Nov 21 23:30:19 open source conference in brussels Nov 21 23:30:20 Hello Nov 21 23:30:26 cybereagle: Hah. Nov 21 23:30:34 maybe it takes a lot to boot the emulator because of these errors I get? Nov 21 23:30:36 morrildl: I really think something might be broken here... Nov 21 23:30:53 cybereagle: I can with complete conviction say that "Open OS" means something completely different to operators and manufacturers than to the rest of the world Nov 21 23:31:01 lnxnt: well, you're booting a whole OS, on a procesor that is emulated. depending on your workstation it could simply take a while Nov 21 23:31:03 cybereagle: to them, WinCE and Symbian are 'Open' Nov 21 23:31:10 Hey jasta, how's your tutorial coming? Nov 21 23:31:19 Well, I'm adding bells and whistles. Nov 21 23:31:24 jasta: are you referring still to the ICO issue? Nov 21 23:31:24 :) cool Nov 21 23:31:32 jasta: I totally agree, and forgot to mention that I am entering a bug :) Nov 21 23:31:32 Stephmw, yeah i guess thats true. i'm purely a user, and i'm just happy to have a platform thats gnna let me easily program my own phone Nov 21 23:31:45 morrildl: Yes. decodeStream() returns null, but I don't think there is anything wrong w/ the stream. Nov 21 23:31:54 cybereagle: and that's my worry, that there won't be an unblocked device when this ships Nov 21 23:32:09 jasta: does the same decodeStream() call succeed if you feed it a bitmap? (or other supported file type?) Nov 21 23:32:28 cybereagle: the only way I'd be convinced would be if either a manufacturer or Google actually made such a declaration Nov 21 23:32:43 morrildl: Let's try... Nov 21 23:33:41 morrildl: it's a fair size free-entry conference, over 2 days Nov 21 23:33:47 it works if you supply a png file. Nov 21 23:33:49 morrildl: Sun started attending last year Nov 21 23:33:53 Dralspire: ping Nov 21 23:34:03 well it only takes one vulnerable library to get a foothold if needs be... not gonna help froma business perspective but it'll do me lol Nov 21 23:34:21 i may be able to try a different ico file Nov 21 23:34:22 cybereagle: :D so you've got an iphone already... Nov 21 23:34:32 morning Nov 21 23:34:37 hiya rwhitby Nov 21 23:34:38 Stephmw: I've been to small-ish conferences (like GNOME) where Google attended, so I think FOSDEM sounds quite likely Nov 21 23:35:02 orrc: cool Nov 21 23:35:13 morrildl: It definitely fails on any .ico file. Nov 21 23:35:22 the bug is simple to confirm. Nov 21 23:35:39 Did anyone try M3-RC22a on the 3.4 line of Eclipse? Nov 21 23:35:43 run BitmapFactory.decodeStream(new URL(anywhere).openStream()); Nov 21 23:35:58 jasta: what about local .ico files? Nov 21 23:36:02 replace anywhere with a URL to an ico, it will return null. replace it with a png path, it will succeed. Nov 21 23:36:04 rcjsuen: that is known to not work right now, I believe Nov 21 23:36:23 Stephmw: you really think it is likely to implement the InputStream parsing different for local files? Nov 21 23:36:33 jasta: bug has been entered, thanks Nov 21 23:36:45 morrildl: but they claimed it works ;p Nov 21 23:36:46 jasta: I'm just wondering if it's similar to bug I've seen elsewhere Nov 21 23:36:53 "ADT 0.3.1 is now compatible with Eclipse 3.4." Nov 21 23:36:56 Lies? Nov 21 23:36:58 jasta: where the buffering of external streams is at fault Nov 21 23:37:18 i am quoting http://code.google.com/android/RELEASENOTES.html Nov 21 23:37:22 I suppose I could try, but I'd prefer to just call it Google's bug and move on. Nov 21 23:37:22 well i must say thats the quickest vendor response toa bug i've ever seen lol Nov 21 23:37:30 Stephmw: re: FOSDEM -- cool, we are planning for 2008 conferences now. I'll add FOSDEM to our list of conferences to consider. (This would be specifically about Android -- there may or may not be a larger Google presence.) Nov 21 23:38:01 rcjsuen: ha ha, looks like they fixed it then :) Nov 21 23:38:09 rcjsuen: are you having trouble with it? Nov 21 23:38:18 morrildl: No, I haven't tried, jus twondering if anyone did. Nov 21 23:38:21 morrildl: ok, glad to know - be warned, they're already starting their calls for speakers/devrooms, etc Nov 21 23:38:23 cybereagle: we aim to please Nov 21 23:38:40 Stephmw: okay, thanks. We will definitely be doing SOMETHING in Europe next year, regardless Nov 21 23:38:54 yay :) Nov 21 23:39:58 So I've been approached by the owners of androidwiki.com to purchase the android-internals.org domain from me. How much should I charge? ;-) Nov 21 23:40:44 * Stephmw does the pinky-to-lips move.... One Million Dollars! Nov 21 23:41:19 rwhitby: haha Nov 21 23:42:14 But seriously, since I created android-internals.org as a community resource intended to be the antithesis of the plethora of wanna-be android portals whose only purpose is to draw in clicks for advertisiing revenue, I doubt that my ethics would allow me to sell it. Nov 21 23:42:40 why would they want some .org? Nov 21 23:42:50 and who are behind that wiki? Nov 21 23:46:10 parti: I am behind android-internals.org (along with most of the other people who have posted *new* information about android internals, as opposed to just rehashing existing information already published by Google). Was that the wiki you are referring to? Nov 21 23:46:10 I have a problem...the hello world example doesn't display anything anymore and the app is not deployed on the emulator Nov 21 23:46:37 rwhitby, ye i was curious who owns the wiki Nov 21 23:47:22 parti: I just own the domain - the community owns the wiki. Nov 21 23:47:59 sorry i ment who's behind androidwiki.com Nov 21 23:48:28 parti: I believe Dralspire owns it Nov 21 23:55:07 http://castfortwo.blogspot.com/2007/11/impression-of-googles-android-ui-on.html <-- some people run android on devices Nov 21 23:55:33 ;) Nov 21 23:56:29 on an ipod touch? niiceee Nov 21 23:58:38 still I like the iphone menu more than the android menu navigation :( it's stack way of navigating has a feeling of being digged up from the attic. Nov 21 23:58:47 :( Nov 21 23:59:05 make a better one ;P Nov 21 23:59:16 at least you have the option Nov 22 00:05:56 lnxnt: anything but iStuff Nov 22 00:06:17 lnxnt: the current android UI is a placeholder Nov 22 00:06:23 It is far from the final UI Nov 22 00:07:13 lnxnt: ...although the "backstack"/history model will remain. What do you dislike about it? Nov 22 00:07:40 morrildl: Is there a task manager in the works? - I do agree that the navigation leaves a bit to be desired... I like being able to quickly switch between tasks Nov 22 00:08:16 morrildl: can you get someone to add the "or exploring its use on various hardware." from http://code.google.com/android/groups.html to the charter of the android-internals group on http://groups.google.com/group/android-internals/about ? Nov 22 00:08:26 The inconsistency is causing questions on the discussion group. Nov 22 00:08:44 rwhitby: can you clarify? Nov 22 00:08:54 with a bit of work someone could make a kind of pseudo window manager/task bar as an app perhaps Nov 22 00:08:56 e.g. can you show me a specific thread? Nov 22 00:09:44 on the group summary page, it says the android-internals group is "For "hackers" interested in discussing the innards of Android or exploring its use on various hardware.". On the charter page for the group itself, the last part of that sentence is left off. Nov 22 00:09:58 cybereagle2: /Nod, justs seems like it should be built into the system... Nov 22 00:10:26 morrildl: the thread starting at http://groups.google.com/group/android-internals/msg/676bbc06b4720bbb Nov 22 00:11:45 Nescio|Away: but the way threds and activities come and go when they aren't in the foreground would make it harder to do in a consistent way, it'd have to have a nice dose of "illusion" Nov 22 00:12:11 yep Nov 22 00:12:42 pictured - just a way of keeping track of bundles... Nov 22 00:13:30 well apps should resume themselves intelligently, so you could just have a list of tasks that the user would consider "open" for ease of switching Nov 22 00:13:57 but, how could my app. know what else is running? Nov 22 00:14:06 it wouldnt have to Nov 22 00:14:19 So, you mean like a quick launch bar then? Nov 22 00:14:29 processes can just randonly be terminated, but when the user "switches" to that task, it would just be started again Nov 22 00:14:56 well, under the hood it wuld be more like a quick launch, but from the users perspective it could still "look" like a window manager lol Nov 22 00:15:14 hehe Nov 22 00:15:20 cybereagle: there is indeed to reason that wouldn't be possible Nov 22 00:15:28 But what is it you guys don't like about hte current model? Nov 22 00:15:34 Just that it's too hard to start/restart apps? Nov 22 00:15:51 oh i do like it, the idea of making it act in a way it doesnt really actw as just too tempting ;) Nov 22 00:16:16 and as a dev it took a while to get my head around the app lifecycle Nov 22 00:16:19 * morrildl wants to hear from lnxnt who started the discussion :) Nov 22 00:16:25 but once you figure it out its good Nov 22 00:16:25 Right now when I accidently hit [Home] inside a textview I go back to the home screen, then i have to navigate back to the app I was running Nov 22 00:17:19 * Nescio|Away Agree - didn't mean to silence lnxnt, sorry bud! Nov 22 00:17:50 yeah, it would be nice to be able to go "back" form home, maybe some kind of circular stack? i dunno i'm thinking aloud Nov 22 00:18:22 Nescio|Away: heh, previously the "back button" DID work "across" Home, and that was also confusing :) Nov 22 00:18:42 cant please everyone :) Nov 22 00:18:47 You would have Activities from different apps shuffled together like a deck of cards Nov 22 00:19:31 the way you can browse the history in the browser in 3D would be great in other contexts Nov 22 00:19:37 This idea was missing on WM5 as well, HTC finally released a small app that would let you switch/close apps Nov 22 00:19:43 g'night! Nov 22 00:19:49 g'night Nov 22 00:20:33 yeah but apps dont really "close" on android. it took me a while ot get my head around that Nov 22 00:21:30 cybereagle: as a user, once you get used to it you'll love it :) Nov 22 00:21:36 Just a thought... I'm sure I'll get used to the concept, I'm just too used to my current phone... Nov 22 00:22:00 cybereagle: I also want the carousel-style history representation :) Nov 22 00:22:24 Nescio|Away: yeah I hated that about WM5 -- I spent a LOT of time in the Memory Manager app killing processes :) Nov 22 00:22:26 oh definately, i dont think its confusing as a user, its when you think like a developer/geek it gets confusing. its easy ot navigate and use, but when you think in terms of processes and threads it gets a little more complicated Nov 22 00:22:33 hehe Nov 22 00:22:42 cybereagle: the trick is to let yourself forget about the processes :) Nov 22 00:22:45 but geeks dont minda leraning curve if there's a payoff :) Nov 22 00:22:53 yeah, its very zen Nov 22 00:23:00 cybereagle: we've tried really hard to make it so that while you have to know there ARE processes, you don't ever have to WORRY about them Nov 22 00:23:09 i only just got into the proper mindset yesterday Nov 22 00:23:13 cool Nov 22 00:23:57 What are the battery life implications of this model? - Will the scheduler take into account amp ussage, as well as memory? Nov 22 00:24:01 its like is my app runing or not, and then it hits you, neither! :P Nov 22 00:25:08 morrildl: I assume you've been using it :) -- I know devices vary, but what kind of performance have you observed? Nov 22 00:36:34 Nescio|Away: the development hardware we have are not really everyday devices. I wouldn't really want to make any conclusions based on their battery life Nov 22 00:37:00 morrildl: sorry, I missed the discussion. What I wanted to say is that I had a bad feeling the first time I operated this emulator. When I was a kid and if I would get an android phone the first time in my hand in that time and when I would know about the iphone way of navigation, I would be pretty dissapointed about the stack navigation, because it's not "cool" enought ;) Besides I get immediately the feeling I know what the whole a Nov 22 00:37:03 alright, I was just curious :) Nov 22 00:39:25 the 3D view of browser history would be a "cool" addition to the backstack navigation Nov 22 00:40:04 cybereagle: haha, kinda like the browser history? (probably a bit resoucre intensive tho Nov 22 00:42:27 yeah, but it seems to handle it ok in other situations Nov 22 00:43:11 it would definitely add to the "cool" factor :) Nov 22 00:43:23 i need to make some kind of screen saver in opengl with a nice evil looking rotating logo Nov 22 00:44:40 ok, Happy Turkey day everyone... cya later :) Nov 22 00:48:05 Hmm. How can I get the row view from a ListAdapter? There is a .getView() method but looking at view/List4.java from the ApiDemos, I don't see how it is relevant. Nov 22 00:49:31 I have extended CursorAdapter, and essentially I want to get back, by row position, the View I created in newView(). Nov 22 00:50:27 set up a onListItemClick listener? Nov 22 00:50:35 or what do you mean get the row view? Nov 22 00:51:30 I mean get the View that the list row is using. I created the View by overriding newView on my CursorAdapter class. Nov 22 00:51:57 I just want to get it back in another part of the program. Let's say I had an icon and in the row and I want to change the icon now as a result of user interaction. Nov 22 00:52:17 in a way that does not persist in the database, so updating the cursor is not relevant. Nov 22 00:52:43 I think you pass the view you want to use to the constructor of your specific ListAdapter Nov 22 00:53:08 What? Nov 22 00:53:45 don't you pass the view you want for each row to the constructor of the ListAdapter you want to use? Nov 22 00:54:00 no, you do not. Nov 22 00:54:11 cybereagle: an OpenGL screensaver? Speaking of battery life........ Nov 22 00:54:13 yuuh that's what I do Nov 22 00:54:25 the Id of the view anyhow Nov 22 00:54:29 haavi: It can't be. Nov 22 00:54:52 wow its amazing what is going on ... hard to catch up with all this activity for it being just one week announced... Nov 22 00:55:06 jasta: getView() should be what you want Nov 22 00:55:16 haavi: See view/AutoComplete4.java, you'll see what I mean by newView(). Nov 22 00:55:28 jasta: when the onListItemClick is called, the int argument (position) should contain the row that was clicked Nov 22 00:55:32 okay I'll check, I don't think I understand you correctly Nov 22 00:55:33 morrildl: convertView should be null, then? Nov 22 00:55:53 jasta: what are you trying to do? Nov 22 00:56:11 are you trying to fuss with a View in response to a user click? Nov 22 00:56:36 Sort of, in response to a user menu item selection. Nov 22 00:57:02 I want to take the list row that is currently selected when the user performs this action, and add an animating progressbar icon until some action (network activity) is finished. Nov 22 00:57:05 but not blocking the UI. Nov 22 00:58:17 in this specific case, the progressbar will indicate that the RSS channel is being refreshed. the progress will be the amount of data received out of how much is expected (Content-Length) Nov 22 00:59:22 but it is important that i not just block the whole UI and show a progress dialog. the user should still be able to click around and view feeds while certain channels are refreshing. Nov 22 00:59:29 but i also want a progress indicator. Nov 22 00:59:42 so it seemed fitting to put said indicator in the list row. Nov 22 01:01:29 I think you have to make your own adapter to be able to do that Nov 22 01:01:35 i have made my own adapter. Nov 22 01:01:52 ok Nov 22 01:02:05 but it is still not clear how i could implement the functionality that i want. Nov 22 01:02:11 but then you can return the view you want for each row with the getView method Nov 22 01:02:26 unless, of course, i do something ridiculous like manage my own list of constructed View's and query them separately. Nov 22 01:02:56 haavi: getView does not do what I want, I am certain of it. Look at view/List4.java's overriden getView method. it either creates a brand new view, or it just binds the one you pass it to data. Nov 22 01:03:12 have getView return a view that has that icon in it but only make it visible when the list item is clicked Nov 22 01:03:24 isn't that possible? Nov 22 01:03:28 haavi: You're more confused than I am, I think :) Nov 22 01:03:34 ok =) Nov 22 01:04:12 I just started messing with Lists a few days ago myself Nov 22 01:04:39 I see a path to solve my problem myself, by adding a simple mapping in my custom CursorAdapter class, and then an extra method like View getViewByRowID(long id), but that seems really ridiculous. Nov 22 01:05:23 why wouldn't my idea work then? Nov 22 01:05:36 it sounds reasonable to me :) Nov 22 01:05:43 because the getView() method has nothing to do with yielding a view by row position. Nov 22 01:05:49 It seems to have some other purpose, and is poorly named. Nov 22 01:06:45 I thought getView is the method that gets called one time for every row to set your view when you call setListAdapter Nov 22 01:07:15 haavi: It may well be, but it does not yield an existing view by row position. Nov 22 01:07:37 morrildl: Are you there? Do you have any thoughts? Nov 22 01:11:54 if you set up your own view that has your progress icon for each row with getView and then make that icon visible when the onListItemClick listener gets called, why wouldn't that work? Nov 22 01:11:54 jasta: sorry, back Nov 22 01:12:11 you don't have to yield a completely new view on the click just show the icon Nov 22 01:12:26 haavi: in onListItemClick, how would you access that particular row's ProgressBar? Nov 22 01:12:28 jasta: ahhh I see Nov 22 01:12:43 jasta: Personally I would create a custom View for the indicator, and then create a layout file that includes it Nov 22 01:13:04 morrildl: But still, how do you manipulate a ListAdapter's row? Nov 22 01:13:22 Nov 22 01:13:31 jasta: generally, you aren't expected to Nov 22 01:13:41 That is silly. Nov 22 01:14:02 the general idea is that your ListActivity/ListView only "need" to interact with a given row once it's clicked Nov 22 01:14:15 and all the guts of ListView are aimed at that Nov 22 01:15:13 morrildl: There are fundamental problems with that assumption that will be exposed quickly when applications begin to appear for this platform. Nov 22 01:15:32 yeah I don't like that either, I want to be able to style the text for each list row Nov 22 01:15:34 It may still be possible to fool around with the Views in a ListView that way (particularly if you subclass ListView) but we would suggest that the "right" way to do what you're trying to do is to put your progress indicator in the XML layout Nov 22 01:15:39 and I still haven't figured out how to do it Nov 22 01:15:44 My specific application will not be the only one that needs to interact more generically with it's own layout. Nov 22 01:16:04 jasta: I'm not saying your use case isn't legitimate :) Nov 22 01:16:07 morrildl: How will it help to put it into the XML layout? Nov 22 01:16:46 well, I'm assuming that what you are talking about is a progress indicator at the top of the screen, and the list of RSS entries below Nov 22 01:16:49 Is that inaccurate? Nov 22 01:17:24 he wants it to show up on the list row that he clicks Nov 22 01:17:27 morrildl: No, incorrect. I am talking about putting a progressbar icon in the right-hand column of a list. The icon indicates the progress of each row, independently. Nov 22 01:17:46 yeah that'd be nice indeed Nov 22 01:17:48 jasta: ahhhhhhhhhhh Nov 22 01:17:50 haavi: Not that I click; this feature is engaged by the user opening a menu or even by a background service one day. Nov 22 01:18:00 jasta: then in that case, I'd write a custom View Nov 22 01:18:01 oh Nov 22 01:18:14 I have, in a sense, written a custom View. Nov 22 01:18:24 Alternatively, you can set IDs on the individual elements in the row's XML layout Nov 22 01:18:29 I have extended CursorAdapter, and have done special magic in newView() and bindView(). Nov 22 01:18:35 and then get a handle to the progress indicator via getView(), and set updates on it that way Nov 22 01:18:56 you mean findViewById()? getView() is magic, it does not do what I want. Nov 22 01:19:06 what does getview() not do? Nov 22 01:19:18 jasta: actually -- I have to run now Nov 22 01:19:23 jasta: however, I don't want to drop this :) Nov 22 01:19:31 Can you write me an email about this? Nov 22 01:19:38 I'll check it when I get home Nov 22 01:19:41 It is an undocumented method, to which I do not have the source, but judging from List4.java, it recycles views. Nov 22 01:19:53 jasta: ahh, okay Nov 22 01:19:53 yep Nov 22 01:21:08 List4.java overrides getView(), and it most certainly does NOT yield a row View by position. Nov 22 01:21:49 morrildl: I can send you an e-mail; what is your address? Nov 22 01:22:00 jasta: morrildl@google.com Nov 22 01:22:43 Ok, I will probably start a Google Groups thread and e-mail you the reference to it. Nov 22 01:22:52 jasta: the ListAdapter stuff is a little odd, but powerful :) I bet there's a way to do this, and if not I'll get the feature request in Nov 22 01:23:03 jasta: sounds great! Thanks Nov 22 01:24:51 jasta: but whatever view you return from getView is the view that gets populated on each row of your list right? Nov 22 01:25:08 haavi: getView() is not correct, trust me. Nov 22 01:25:40 okey :/ Nov 22 01:25:45 if i was going to hack my way around the problem I would use my overridden newView() and bindView() methods to track, by underlying cursor id, the View I manage. Nov 22 01:26:07 then add my own ListAdapter method to access that list I maintain, and yield the row View by row position. Nov 22 01:26:08 and here I was subclassing a whole class in the belief that that was what it did Nov 22 01:26:54 seriously, read view/List4.java's implementation of getView(). It is painfully clear that it does not yield an existing view. Nov 22 01:32:15 makes SpeachViews for each row Nov 22 01:32:25 that's what I thought Nov 22 01:33:24 Now imagine that there is a menu item that when you click on it, changes the dialogue of the current SpeechView row to "Foobarbaz!" instead of whatever it was before. Nov 22 01:34:26 yeah Nov 22 01:34:40 it would be like calling setDialogue("Foobarbaz!"); on the currently highlighted row. But how do you get the currently highlighted row? Nov 22 01:34:43 Do you see the problem now? Nov 22 01:35:14 yeah that's a problem Nov 22 01:36:09 The only way, then, is to manage the mapping yourself. Absurd. Nov 22 01:36:59 Obviously some layer somewhere is already managing this. It just might not expose itself because Google for wahtever did not think this would come up. Nov 22 01:37:33 they're gonna release the source for the library sometime next year I think Nov 22 01:37:38 that's gonna help Nov 22 01:38:15 that class's method is very poorly documented imo Nov 22 01:38:21 methods Nov 22 01:44:02 i guess i'll move on to the other part of this functionality. Nov 22 01:44:11 creating a service to do all the work :) Nov 22 01:44:19 it's easy to get stuck with the details ;) Nov 22 01:44:34 it needs to be in a service since the point is that the user can happily keep navigating through the program while this function is being performed, whether they are visualizing the progress or not. Nov 22 01:48:21 I would like to check out this RSS reader of yours when it's done Nov 22 01:48:28 if you're going to release it that is Nov 22 01:49:11 i'm planning to release it very soon, as a demonstration of a "complete" app Nov 22 01:49:28 cool, with source code too? Nov 22 01:49:29 i will, of course, create a sourceforge or google code project to host the code. Nov 22 01:49:36 nicee Nov 22 01:50:15 i hope that Android is a platform where users are not surprised to find source available for all their favorite 3rd party apps. Nov 22 01:50:38 it should be the norm. Nov 22 01:50:46 I always release the source to the stuff I write Nov 22 01:53:05 if it's asked for Nov 22 01:54:37 btw... Nov 22 01:54:45 the project is currently up to almost 2k loc :) Nov 22 01:55:22 bigger and more thorough than even the NotePad sample Google created. Nov 22 01:55:42 nice :) Nov 22 01:55:57 before I release it, though, I want to clean up all the things I know aren't "the android way" Nov 22 01:56:07 since I want it to serve as a tutorial for how to do things correctly on Android. Nov 22 01:57:07 that makes sense Nov 22 01:57:45 I like that Google released so many examples Nov 22 01:57:58 that's always the best way to learn I think Nov 22 01:58:05 for me it is anyway Nov 22 01:58:22 my big milestone goal is making the feed refresh process happen on an interval, and in a background Service that the main Activity can interact with when it needs to. Nov 22 01:58:27 for example, schedule a refresh. Nov 22 01:58:35 and show the progress of active refresh commands. Nov 22 01:59:41 haavi: examples and good documentation must be wed together, either without the other is inefficient communication. Nov 22 02:00:03 it's like a sexless marriage, or loveless sex :) Nov 22 02:00:17 true :) Nov 22 02:00:30 except that documentation is like sex, because it still works without examples :) Nov 22 02:00:52 anyway, time to write the service Nov 22 02:00:59 ok, gl! Nov 22 02:17:44 rockstars Nov 22 02:44:36 jasta: if you send the position parameter in the getView method to your View constructor you can assign unique ids for every rows views by using that parameter Nov 22 02:45:13 which enables you to reference to the the clicked view by that unique id Nov 22 02:50:41 yeah that works, I just tested it by changing the text for the rows I click on Nov 22 02:56:10 you mean like View.setId()? **** ENDING LOGGING AT Thu Nov 22 02:59:56 2007