**** BEGIN LOGGING AT Wed Feb 13 02:59:57 2008 Feb 13 03:05:19 hmmm. Feb 13 03:05:21 # ls Feb 13 03:05:21 -rw-rw---- app_6 app_6 26624 2008-02-13 03:05 five.db Feb 13 03:05:21 -rw-rw---- app_6 app_6 26624 2008-02-13 03:05 five.db Feb 13 03:05:26 * jasta raises an eyebrow Feb 13 03:06:52 duplicates? Feb 13 03:15:46 yeah Feb 13 03:16:12 i wonder if android creates temp files using trailing space characters or something? Feb 13 03:24:34 yay, my program finally has a successful sync. Feb 13 03:25:10 took like 5 minutes simulating 3G speeds, though. that's not good ;) Feb 13 03:25:35 but there's a lot of room to improve. I'm using XML right now, even though the protocol I chose supports WBXML. Just didn't want to turn it on yet because it's harder to debug. Feb 13 03:26:13 congrats Feb 13 03:31:53 grats Feb 13 03:32:04 I need to get back to my project so I can finish it by march Feb 13 03:33:53 i'm planning to nearly rewrite every part of it after the initial deadline :) Feb 13 03:34:02 i'm really cowboy coding on this thing Feb 13 03:34:12 Me too :-X Feb 13 03:34:47 anyway, goin home Feb 13 03:34:48 later folks Feb 13 03:36:33 yeah, I can't say I'm planning my code out all that well either Feb 13 04:09:35 you think I can create an SQL lite DB one time, then save the file into my /res/raw folder and copy it to my /data/data//databases/ the first time the app boots up? Feb 13 04:10:14 or is there some kind of security/id numbers/ etc that would not allow the db file to be valid Feb 13 04:40:35 lumme Feb 13 04:40:43 strike a light Feb 13 04:58:49 zhobbs: i think that it would work, but it might not be the best approach. Feb 13 04:59:18 specifically, upgrades might not work very gracefully in that arrangement, and you'll waste a lot of space for a marginal performance gain at first startup. Feb 13 05:01:54 as for security, i doubt it, however there does seem to be some extra magic that version controls your database if you use the database helper. Feb 13 05:08:22 hmmm Feb 13 05:08:51 have you guys read the beginners newsgroup recently? Feb 13 05:09:01 i haven't ever read it, why? Feb 13 05:09:22 there are some interesting posts sometimes Feb 13 05:09:46 I read all the forums fairly often Feb 13 05:09:47 I had assumed android-beginners would just be enthusiasts, no real developers. Feb 13 05:10:00 romainguy___: what specifically, though? Feb 13 05:12:00 recent posts Feb 13 05:12:57 using an eeepc for android development doesnt sound too fun Feb 13 05:13:06 i just glanced and didn't see anything of interest? Feb 13 05:16:03 may be just the level of activity is exciting to romain :) Feb 13 05:16:09 maybe :) Feb 13 05:16:25 just look harder :) Feb 13 05:16:39 no. Feb 13 05:17:12 maybe romain is excited about 2gb of free storage Feb 13 05:17:21 that's what i was wondering about. Feb 13 05:17:28 ah well, too bad for you guys Feb 13 05:17:43 we're going to miss out on the limited offer. Feb 13 05:18:10 seriously,w hat are you talking about? i went back quite a few pages and didn't see much threading at all Feb 13 05:18:22 mostly just 1 or 2 message topics that are relatively short and uninformative Feb 13 05:18:36 hey Jason Feb 13 05:18:49 what's up, romain? Feb 13 05:18:56 romainguy: I hope you guys figure out some kind of api or way to allow applications to display screens or "widgets" when the phone is ringing (incoming call), from what i understand, its not possible yet Feb 13 05:18:59 Idling in the shuttle ) Feb 13 05:19:29 jtoy: What would you like to do with such an API? Feb 13 05:20:25 display some relavent info to the user regarding the incoming call Feb 13 05:21:48 so for a more concrete example: the number is checked against a database of spam numbers, and if its a spam number, then display the message, "this number may be a telemarketer according to spamdb.com" Feb 13 05:22:27 Oh I see Feb 13 05:22:55 well, file a feature request: http://code.google.com/p/android/issues/list :) Feb 13 05:23:38 but there is a lot more cool stuff you could do, I built some applications for a startup I used to work at that had some pretty tight integration with asterisk/webservices Feb 13 05:23:42 ok, thanks Feb 13 05:32:15 almost home Feb 13 05:32:17 see ya later Feb 13 05:45:15 I made a Google maps stitcher thingy, makes big posters from google: http://www.jportsmouth.com/code/Stitch/stitch.html Feb 13 05:45:42 that's cool :) Feb 13 05:46:30 not enormously useful but fun Feb 13 05:47:17 re Feb 13 05:47:54 snoktruix: neat Feb 13 05:48:07 :) Feb 13 05:50:32 snoktruix: that's pretty cool Feb 13 05:51:12 cheers Feb 13 05:54:29 jasta: I think romainguy_ was refering to the DB topic which was talking about exactly what I was talking about :) Feb 13 05:54:39 zhobbs: I doubt it :) Feb 13 05:55:09 oh, well, that wasn't interesting all the same :) Feb 13 05:58:53 romainguy_: how might i have a list which is allowed to grow as much vertically as it would like, but not necessarily take up the entire screen? Feb 13 05:59:14 i want to basically position a view at the end of the list, without it being part of the list. Feb 13 05:59:51 height=0 weight=1 with the other view height=wrap_content Feb 13 06:00:20 ahh, height=0 Feb 13 06:01:00 for the listview? Feb 13 06:01:53 yes Feb 13 06:02:16 jasta: another height would work but height=0 goes through as faster code path Feb 13 06:02:24 hmm. i set the listview layout_height="0px" layout_weight="1", and a textview immediately below it with layout_height="wrap_content", no weight, and it still shows at the very bottom of the screen. Feb 13 06:02:37 aah sorry Feb 13 06:02:41 I thought that's what you wanted Feb 13 06:02:55 no, i want it to show at the bottom of the list, as though it was the last entry in the list. Feb 13 06:03:16 then the best thing to do is to set it as a footer in the list Feb 13 06:03:28 lists have footers, eh? Feb 13 06:03:35 they support headers and footers Feb 13 06:03:41 if not in your SDK, in the next one :) Feb 13 06:04:03 no, i see it. just never realized this existed. you guys really have to expand the API demos ;) Feb 13 06:04:21 Read the APIs :)) Feb 13 06:04:47 there's still lots of api secrets out there Feb 13 06:07:11 romainguy_: how would i specify a view in XML and still use setFooterView()? Feb 13 06:07:28 put the view in another XML file Feb 13 06:07:40 inflate it in code, then call setFooterView() Feb 13 06:07:50 although we might have a way to do this in XML Feb 13 06:07:55 but I'm not sure Feb 13 06:08:03 ahh, well, inflating views from XML is a mystery to us lowly mortals. Feb 13 06:08:15 it's pretty easy though Feb 13 06:08:52 yeah? how do you? :) Feb 13 06:12:00 well, getViewInflate().inflate(R.layout.mylayout, aLayout, true, null); Feb 13 06:12:23 except in your case you'll want to use null, false, null Feb 13 06:12:27 Nice...I'll have to remember that Feb 13 06:12:30 (for the last 3 parameters) Feb 13 06:13:12 excellent. Feb 13 06:13:18 you can also obtain a ViewInflate object by calling getSystemService() on Activity Feb 13 06:13:25 or, if I remember, ViewInflate.from(myContext) Feb 13 06:15:00 that's actually quite helpful. thanks. Feb 13 06:16:05 yeah, ViewInflate is a pretty sweet tool Feb 13 06:16:13 especially from within Adapters Feb 13 06:16:34 exactly..I was struggling with doing a layout via code in an adapter last night Feb 13 06:39:02 sorry for all the disconnects/reconnects Feb 13 06:39:06 playing with the router Feb 13 06:42:28 ever had one of those odd philosophical dialogues with youreslf about why you constantly refer to actions performed by your code as "we"? Feb 13 06:42:48 i sometimes wonder if programmers identify with their code like they are cooperating ;) Feb 13 06:59:19 good lord, the horizontal progress bar drawable is uuuggly ;) Feb 13 07:10:08 there is so much interesting stuff in res/drawable in the android.jar :) Feb 13 07:10:24 yep Feb 13 07:10:38 lots of goofy looking white icons too Feb 13 07:10:54 probably statubar's icons Feb 13 07:11:03 i think they are too big for that. i see those as well. Feb 13 07:11:20 what does the white goofy icons look like? Feb 13 07:11:52 like, res/drawable/contact_fax.png Feb 13 07:12:01 ah yes Feb 13 07:12:03 they seem to all be contact icons Feb 13 07:12:18 contact_pager.png hehe. i hope you guys don't ship with a pager field :) Feb 13 07:29:01 hehe, what the hell. you can click to select a row not selectable. Feb 13 07:30:43 Does the OutputStream of a Socket block? or will it return right away when you add something to it? even if it's a very large write? Feb 13 07:33:49 Yes, it will. Feb 13 07:34:44 The associated SocketChannel is capable of supporting non-blocking operations, but the Socket is not. Feb 13 07:35:22 Err, sorry, I meant Yes, it will block. Feb 13 07:36:33 SocketChannel is not yet implemented in the SDK is it? Feb 13 07:36:50 That I don't know. I presume not because it is part of the nio package. Feb 13 07:37:07 aye Feb 13 07:42:22 hrm Feb 13 07:42:31 there's no way for me to tell how much i can feed into it without it blocking? Feb 13 07:49:08 what advantage is there pretending a blocking socket is non-blocking? Feb 13 07:50:09 i dont have to create a seperate thread to manage traffic Feb 13 07:50:50 you must create a separate thread if you using a blocking interface. that is the fundamental premise and Java will not expose sufficient information for you to implement a non-blocking layer on top of that. Feb 13 07:51:01 though that information does exist, in the kernel. Feb 13 07:51:48 ive given up on waiting for nio to be implemented Feb 13 07:52:00 im just using the standard java.net classes now Feb 13 07:52:14 hopefully that package is actually implemented Feb 13 07:52:38 great, but that layer lacks sufficient information to implement non-blocking sockets on top of it. Feb 13 09:41:33 hi all Feb 13 09:42:35 does any body how to install org.eclipse.equinox.app ? Feb 13 10:27:13 has anyone layered the way they access sqlite3 on android? Feb 13 10:27:45 instead of just calling the android provided sqlite3 classes directly? Feb 13 14:32:14 sdk? Feb 13 14:32:26 new ui? Feb 13 16:34:56 couldn't we just do it like all those who are logged to the channel win the prize? :) Feb 13 16:43:44 except we represent the laziest of all the participants. Feb 13 16:43:57 therefore, however, the best programmers. :) Feb 13 16:48:51 right Feb 13 16:48:59 the ultimate in code resuse Feb 13 16:49:05 err.. reuse Feb 13 16:50:11 * jasta scratches his head Feb 13 16:51:55 jasta: we don't write any code as that would just be duplicating someone else's efforts Feb 13 16:52:04 jasta: so clearly, we should get the prizes Feb 13 16:55:51 yay, made my service thing work. Feb 13 16:56:00 so the activity can just bind to it and figure out what the service is doing at any time Feb 13 16:56:03 without disrupting it Feb 13 16:56:18 how? Feb 13 16:56:56 well startService and bindService, but my arrangement also has an observer interface on the service itself. Feb 13 16:57:28 i just call startService, then bindService in onResume() and in onPause I unregister my observer, then unbindService. Feb 13 16:57:42 ok.. anyways the service needs to be disrupted Feb 13 16:57:51 no it doesn't. Feb 13 16:58:30 it gets a second call to onStart(), which it safely ignores, and no unwarranted calls to onStop. Feb 13 16:58:49 ok Feb 13 17:01:22 i really enjoy the service architecture in android. i just wish some of the stupid smaller bugs would hurry up and get fixed. Feb 13 17:01:53 asynchronous services are so easy to implement correctly, too. Feb 13 17:02:54 hope you do it right, otherwise it becomes a feature :) Feb 13 17:03:12 i am certain i have done it correctly. Feb 13 17:04:26 brb, shower, then work. Feb 13 17:40:00 re Feb 13 17:57:55 * stifffff slaps morrildl around a bit with a large trout Feb 13 18:01:26 alert, altert: http://feeds.feedburner.com/~r/blogspot/hsDu/~3/234479957/android-sdk-m5-rc14-now-available.html Feb 13 18:01:32 ^^ Feb 13 18:02:20 jeez, zhobbs Feb 13 18:02:34 you are on the ball Feb 13 18:02:45 steal our freakin' IRC thunder, will you Feb 13 18:02:47 ;) Feb 13 18:02:51 hehe Feb 13 18:03:16 osnap! Feb 13 18:03:21 and if I may join in Feb 13 18:03:29 please, read the release notes :) Feb 13 18:06:10 * morrildl loves that "osnap" flowchart Feb 13 18:07:37 wut ! Feb 13 18:08:10 morrildl: the docs diff is really cool Feb 13 18:09:09 woohoo! Feb 13 18:09:16 what a treat! Feb 13 18:09:36 i knew it was coming soon. Feb 13 18:10:03 romainguy: yeah that was crazybob Feb 13 18:10:06 that diff is sweet Feb 13 18:10:18 jasta: Yeah, cause we told you it was :) Feb 13 18:10:21 morrildl: it's quite impressive the number of changes... Feb 13 18:10:30 It's funny how sometimes we say things but people don't believe us :) Feb 13 18:10:34 ;) Feb 13 18:10:36 :)) Feb 13 18:10:51 romainguy: yeah I know, unfortunately b/c we skipped M4 Feb 13 18:10:58 morrildl: and also yesterday people on android-beginners noticed the plugin was asking for M5 :) Feb 13 18:11:07 romainguy: that too Feb 13 18:11:07 ohh, very nice! support for Ogg now, this will be a great help to me. Feb 13 18:11:25 morrildl: you didn't tell us this week :) Feb 13 18:11:45 romainguy: Is that what you were referring to yesterday? Feb 13 18:11:46 jasta: tru dat Feb 13 18:11:52 jasta: maybe :) Feb 13 18:11:55 UI looks different than the ones at the show Feb 13 18:12:11 zhobbs: it's the same Feb 13 18:12:16 zhobbs: they are not necessarily running this latest build Feb 13 18:12:20 it's somewhat older Feb 13 18:12:20 except you have a fresher version Feb 13 18:12:28 i can't believe you guys waited for me to get into work to release this though :) Feb 13 18:13:33 jasonchen: see??? I TOLD you we should have waited for jasta Feb 13 18:13:48 now he's mad at us. NICE GOING Feb 13 18:13:55 morrildl: doh! all my fault Feb 13 18:13:57 ;) Feb 13 18:14:05 finally, the respect i deserve. Feb 13 18:14:21 jasta: next time we'll work around _your_ schedule :-) Feb 13 18:15:19 jasonchen: can we get some sleep first? :) Feb 13 18:15:44 * jasonchen checks schedule to see if it allows for sleep Feb 13 18:15:52 nice, you guys released the traceview tool Feb 13 18:15:55 i was really hoping for something like this Feb 13 18:16:02 traceview is really nice Feb 13 18:16:10 I spent some much time using it for M5... Feb 13 18:16:20 yeah, i had been looking at all the screenshots wondering why we missed out Feb 13 18:20:16 nice job on the SimpleCursorAdapter improvements. Feb 13 18:20:25 thank you very much ;-) Feb 13 18:20:54 * jasta raises an eyebrow Feb 13 18:21:08 TabWidget is no longer deprecated? I had been told that you guys were getting rid of that system? Feb 13 18:21:17 No Feb 13 18:21:23 We told you that it might go away Feb 13 18:21:38 interesting. Feb 13 18:21:42 And since it was not ready to be used, it was marked deprecated Feb 13 18:21:45 Now it's ready to be used Feb 13 18:21:55 You can actually see it in action in the Contacts app Feb 13 18:22:16 i can't wait to tear into this Feb 13 18:22:34 the menu button is a lot different Feb 13 18:23:49 hmm, the view threading improvements seem kind of clumsy. Feb 13 18:24:40 how so? Feb 13 18:25:06 gah. well, i'm glad I already implemented a expandablelistivew. :-/ Feb 13 18:25:14 It just seems like it's promoting magic, hiding away the Handler concept which is much cleaner and more general. Feb 13 18:25:48 jasta: we've added these methods because very often you just need to post an action to the UI thread and creating a Handler just for that purpose is cumbersome Feb 13 18:25:49 has the documentation been improved? Feb 13 18:25:57 especially since the View has access to a Handler Feb 13 18:25:58 posting a runnable to a View also seems like an unusual relationship. Feb 13 18:26:17 romainguy: I can understand the utility, it just seems magical. Feb 13 18:26:25 I disagree Feb 13 18:26:30 Threading should not have to be hard Feb 13 18:26:42 And considering how difficult it's been for some developers to understand and use Handlers... Feb 13 18:26:43 private Handler mHandler = new Handler(); doesn't make anything harder. Feb 13 18:27:37 But it does have the advantage of exposing the framework in a manner more consistent with how it actually works. Feb 13 18:28:01 You can keep using a Handler if you prefer Feb 13 18:28:10 but anyway, I won't be the one confused by it so... Feb 13 18:28:13 lol "Toasts" Feb 13 18:28:31 yay new sdk :) Feb 13 18:28:38 romainguy: I understand that; suffice it to say I simply disagree that such a utility was necessary. Feb 13 18:29:45 romainguy: What is this Touch mode bit here? I'm intrigued, but don't understand what specific purpose you guys had in mind creating this? Feb 13 18:30:01 Which one? Feb 13 18:30:31 The documentation refers to something called "touch mode". I'm curious why that was introduced? Feb 13 18:30:44 skicson: the Toasts concept reminds me of the "Toastie" guy that you could make pop-up in Mortal Kombat Feb 13 18:31:00 jasta: it just means that we're claiming that touch-screen support is usable now Feb 13 18:31:07 In M3 it really wasn't Feb 13 18:31:10 morrildl: Ahh. Feb 13 18:31:27 It also means that when you touch the screen, selection and focus disappear Feb 13 18:31:37 because they don't really have any strong meaning in touch mode Feb 13 18:31:37 oh they actually stop drawing? Feb 13 18:31:56 yes, but they completely disappear Feb 13 18:32:01 interesting. Feb 13 18:32:12 in touch mode, getSelectedItem() on a ListView will return null for instance Feb 13 18:32:26 However, a View can be focusable in touch mode to receive key events directly Feb 13 18:32:33 right, that is absolutely a smart way to do this. Feb 13 18:32:34 for instance an EditText Feb 13 18:32:55 note that pressing a direction will exit touch mode and refocus something Feb 13 18:34:41 Nice, tree view ;) Feb 13 18:35:18 jasta: Not quite, it supports only 2 levels Feb 13 18:35:30 Oh, yeah, I see that now. Feb 13 18:35:51 Oh well, trees aren't very good interfaces on mobile devices anyway. Feb 13 18:35:57 That's the reasoning Feb 13 18:38:11 whoa, ContentURI was renamed. Feb 13 18:38:15 that's a lot of code to update now ;) Feb 13 18:38:51 oh boy, more than just renamed. yiikkes :) Feb 13 18:38:52 jasonchen: so does the Toast have a pic of you in there somewheres? :-) Feb 13 18:39:16 jasta: Good thing you got a deadline extension... :) Feb 13 18:39:24 hehe, yes ;) Feb 13 18:39:25 skicson: haha. thankfully, it does not. Feb 13 18:40:44 ohhh, very nice on the openFile method in ContentProvider. Feb 13 18:40:51 that was really a chore previously. Feb 13 18:43:32 hmm Feb 13 18:43:58 so now mediaplayer supports mp4 http streaming, but youtube mobile still only offers rtsp :/ Feb 13 18:44:13 i was hoping for a youtube breakthrough with the new sdk Feb 13 18:44:17 but oh well Feb 13 18:44:22 umdk1d3: you can rest assured that android will offer a way to access youtube. Feb 13 18:44:26 google owns youtube, remember. Feb 13 18:44:30 lol yea Feb 13 18:44:39 im thinking this is the last sdk before the contest deadline Feb 13 18:44:48 they claim not. Feb 13 18:44:48 so if we want youtube, we need to hack over an rtsp client ;) Feb 13 18:45:05 I thought rtsp worked? Feb 13 18:45:37 orly? Feb 13 18:45:51 don't really know, just thought someone said that Feb 13 18:45:57 hehe, just got to the toast part of the changelog. this does sound like the Mortal Kombat thing ;) Feb 13 18:46:01 ive hacked together a way to get http streaming working Feb 13 18:46:07 but havent touched rtsp yet Feb 13 18:46:11 its not too hard to code up Feb 13 18:49:22 Hmm, new Android Manifest Editor for eclipse Feb 13 18:49:29 yeah, im anxious to see that. Feb 13 18:54:50 howdy Feb 13 18:55:13 hello, new SDK. Feb 13 18:55:28 ooh. pretty new bg on the emulator Feb 13 18:56:01 glad you like it :) Feb 13 18:56:32 so which mobile will use android ? Feb 13 18:57:15 i like the transition animations too.nifty. Feb 13 18:57:34 any screenshots of the new UI floating around yet? Feb 13 18:57:51 http://www.helloandroid.com/node/258 Feb 13 18:58:29 sorry for the annoying flash gallery, but was easy Feb 13 18:58:42 romainguy: why is the bug droid not the default background, anyway? Feb 13 18:58:54 morrildl: it didn't work well with this layout Feb 13 18:59:27 ah Feb 13 18:59:50 can I change the background? Feb 13 18:59:58 it says feature not supported or something Feb 13 19:02:35 are those lists in dev tools, api demos, the default way a ListView looks? Feb 13 19:02:41 zhobbs: yes Feb 13 19:03:34 ooh now to interface android with http://blogs.computerworld.com/here_come_the_teeny_tiny_mini_projectors :) Feb 13 19:05:23 Nice, if I run ddms while eclipse is running I get infinate popup warnings Feb 13 19:07:02 YaY new sdk! Downloading now. Feb 13 19:07:04 * zhobbs feels like I spent a lot of time on stuff I'm going to have to rework Feb 13 19:07:25 zhobbs: aren't you glad about the deadline extension? :) Feb 13 19:07:36 hmm, Ubuntu is all screwy and broken and won't let me install WST. Feb 13 19:07:57 zhobbs: there is a DDMS perspective in Eclipse. Feb 13 19:08:23 jasta: yeah I know Feb 13 19:08:58 romainguy: yeah, no way you could release this and then expect us to figure it out for march 3rd Feb 13 19:14:44 zhobbs: you didn't show the new status bar notifications in your screenshots ;-) Feb 13 19:15:07 ok, I'll get that updated Feb 13 19:17:49 * jasta doesn't want to see it yet Feb 13 19:17:55 i'm installing the SDK right now ;) Feb 13 19:19:57 whoa, the android manifest editor is really elaborate. Feb 13 19:20:20 holy cow Feb 13 19:21:01 aww, Errors (100 of 147 items) :( Feb 13 19:21:21 jasta: lot of work ahead? Feb 13 19:21:26 hehe, yeah :) Feb 13 19:21:55 the best part, though, is that my ADC project just reached a major milestone so is in a good state to go tearing through it with API updates. Feb 13 19:21:55 wow lots of code breakage just by looking at the changelog Feb 13 19:22:14 lol romainguy took the picture on the background of the new SDK. Feb 13 19:22:20 i recognized that right away Feb 13 19:22:27 :)) Feb 13 19:22:41 i am guessing the reason for the extension... Feb 13 19:23:02 ok i must admit i like the new UI a lot more than i thought from the videos i saw. Feb 13 19:23:12 the videos didn't do it justice. there is a lot of subtle design here that looks very nice Feb 13 19:23:31 i agree Feb 13 19:25:36 the bubbly icons on the top, however, i don't like. Feb 13 19:25:49 the font used in the api demos is great :) Feb 13 19:25:53 and all that padding, finally. Feb 13 19:26:04 romainguy: I don't see any changes to the notification section of apidemos? Feb 13 19:26:47 tap the status bar Feb 13 19:30:40 all this new stuff is really great guys Feb 13 19:32:06 ohh, i like the buttons with arrows on them. Feb 13 19:32:08 that's very nice. Feb 13 19:32:39 romainguy: I don't get what to do with the status bar? Feb 13 19:32:47 zhobbs: drag it down Feb 13 19:33:05 Oh, I wasn't dragging it down far enough... Feb 13 19:34:18 we need a task switcher Feb 13 19:36:46 niiicee, the horizontal scroll bar in the title is great. Feb 13 19:37:17 i also like the new circular scroll bar. i love how it kind of pauses each rotation. Feb 13 19:37:46 nice, primary and secondary progress built in Feb 13 19:37:49 fucking awesome Feb 13 19:38:07 horizontal progress dialog looks nice Feb 13 19:39:04 i still wonder about the soft keyboard.... Feb 13 19:41:13 I like the progressbar in status bar Feb 13 19:41:22 or title bar I mean Feb 13 19:44:24 * jasta likes. Feb 13 19:50:28 romainguy: you got your status bar screenshot :) Feb 13 19:51:19 is there tabs in m5? Feb 13 19:51:42 zhobbs: yes Feb 13 19:52:06 hmm, just no api demo Feb 13 19:56:04 wow... that's quite an update Feb 13 19:57:57 zhobbs: look at the contacts app. Feb 13 19:58:03 the tabs actually look nice. Feb 13 19:58:18 i'm pleased they finally amped up the padding of everything and utilize all the screen real-estate. Feb 13 19:58:32 it will make touch-screen users very happy :) Feb 13 20:01:41 jasta: try it with QVGA-L, everything is too big Feb 13 20:07:08 fortunately, that problem is solveable with the SDK. Android was obviously designed to have scalable UIs. Feb 13 20:07:58 but i do think that's a job for Google. the open source community is not going to screw with UI stuff that much ;) Feb 13 20:08:39 that's what you think... Feb 13 20:10:10 screwing up the UI is perhaps not what he meant :D Feb 13 20:13:35 jasta is precisely right Feb 13 20:13:57 the first device that will be available will have a touch screen, so the first UI optimization is for touch-screens Feb 13 20:14:29 Are the judges going to even fire up any emulators other than HVGA-P? Feb 13 20:14:31 when someone gets ready to launch a non-touchscreen device, somebody (perhaps TAT, perhaps someone else) will do a non-touch-optmized UI Feb 13 20:14:39 zhobbs: I'd be surprised if they do :) Feb 13 20:15:36 Just finished flicking through all the demos Feb 13 20:15:54 Nice to see a lot more nested table layou demos Feb 13 20:16:01 layout even Feb 13 20:17:34 They UI, although I'm not awe inspired by it, shows more clarity in the design and lends well to help visualise the final designs. Feb 13 20:17:47 I very much like the buttons Feb 13 20:18:11 The small additions of the arrow on the buttons are an excellent touch. Feb 13 20:18:12 ui seem clunky to me Feb 13 20:18:22 Yeah Feb 13 20:18:23 I guess it's good for big fingers though Feb 13 20:18:32 I don't mind big and well pronounced Feb 13 20:18:54 I'd understand big and well pronounced for accessibility reasons Feb 13 20:19:50 ooh this is really breaking a lot of stuff Feb 13 20:20:07 hehe, no kidding Feb 13 20:20:19 get ready to break out grep/sed Feb 13 20:21:00 it's worse than that. Feb 13 20:21:08 infact, looking at it, it may just be because only some parts of the UI are rendered. The Home screen just has a background behind the icons which looks a bit tattered. But demonstrates the functionality well enough. Feb 13 20:21:27 Very much like the layout animations, they are well swish Feb 13 20:22:44 There is one fundamental point about the UI from the last release I'd like to point out, and I hope is addressed for the final release. Feb 13 20:23:10 The scrolling feels somewhat unintuitive. Feb 13 20:23:33 I was thinking that too, but might be a mouse vs touchscreen issue Feb 13 20:23:33 I don't know why but there is a mild feeling of disorientation when you scroll down a large area of text Feb 13 20:25:11 User agents such as web browsers and text programs have got in the habit of displaying your position by the use of a highlighted bar at the side of the application. This at one time displays your current position and an approxomation of the complete text available Feb 13 20:25:46 Taking that away from a user is mildly disorientating. Even though the IPhone manages it. Feb 13 20:27:10 It might be a good idea to experiment with a small arrow that would appear at the bottom right and top left corners when there is extended content available. Feb 13 20:29:10 chomchom: The scrolling feels somewhat unintuitive. << we are still working on that and we made huge improvements on it this week Feb 13 20:29:26 cool Feb 13 20:30:25 Be interesting to see how this pans out. User interfaces are having a bit of a rethink with this mobile market. Feb 13 20:38:28 anything about battery life? Feb 13 20:39:04 battery usage is crucial in my applicatino Feb 13 20:47:11 michaelnovakjr: is there a way to test battery consumption? Feb 13 20:47:36 zhobbs: that is what I am wondering... Feb 13 20:47:54 that would be handy Feb 13 20:49:27 definitely Feb 13 20:49:35 how about video and android... Feb 13 20:49:46 codecs... do they play nice if at all? Feb 13 20:52:45 http://android-developers.blogspot.com/2008/02/android-sdk-m5-rc14-now-available.html Feb 13 20:53:01 "New media codecs - The MediaPlayer class has added support for the OGG Vorbis, MIDI, XMF, iMelody, RTTL/RTX, and OTA audio file formats." Feb 13 20:53:29 are there formats that it supported before? Feb 13 20:57:38 wow, helloandroid got linked on engadget...server melting Feb 13 20:57:58 zhobbs...shared hosting?? Feb 13 20:58:11 VPS account...actually doing ok now Feb 13 20:58:21 zhobbs: congrats? condolences? Feb 13 20:58:23 hehehe Feb 13 20:58:27 maxClients was set to 150 Feb 13 20:59:16 that's what you get when you scoop everyone Feb 13 20:59:32 you get popular :) Feb 13 21:00:13 vps melted :( Feb 13 21:00:25 that's no good Feb 13 21:00:34 which VPS did you move to? Feb 13 21:00:49 No, was my fault, it's doing ok now Feb 13 21:00:58 I love ServerAxis Feb 13 21:04:02 so it looks like we can't just send text messages from the emulator to somebody else with a gtalk account Feb 13 21:05:22 that is - running in a browser for example? Feb 13 21:05:35 i can't wait to actually test it on hardware Feb 13 21:05:47 that is when i will know if my application sucks or not :) Feb 13 21:07:24 i wonder if android will still accept IM inputs from browser-based gtalk users? Feb 13 21:09:22 cool, 200 guests on site and the load it pretty low Feb 13 21:09:26 is Feb 13 21:11:00 Is anyone having a problem with the android:normalTextColor attribute? Feb 13 21:11:25 yeah, it doesn't look like it exists anymore Feb 13 21:11:44 Whats the alternative? Feb 13 21:11:58 i just used textColor until i find something more relevant Feb 13 21:12:20 I've seen a few changes to the stylable stuff in naming conventions Feb 13 21:12:34 foregroundColor is now colorForeground :-P Feb 13 21:13:13 makes more sense from a package structure point of view Feb 13 21:13:20 sure Feb 13 21:26:13 where can i find org.eclipse.wst.sse.ui Feb 13 21:52:08 morrildl: is a major bullet-point of GTalk and "always-on networking" to replace SMS as a proprietary mechanism for alerting devices? Feb 13 21:52:28 i am curious about that since my program deals with synchronization of data. should we assume that the server can wake up the client when changes come in using GTalk alone? Feb 13 21:52:36 heh... I doubt they're going to wave that flag too much before they get carriers on board Feb 13 21:52:51 that second question's a good one though and I'm curious too Feb 13 21:53:55 that would be huge for any application performing data synchronization. absolutely ground-breaking. Feb 13 22:01:01 michaelnovakjr: it's WST, you can install it using Eclipse's Callisto repository. Feb 13 22:01:09 it's under web tools or something. Feb 13 22:01:10 thanks Feb 13 22:01:18 but you have to install a bunch of dependencies (i just ran into this) Feb 13 22:01:30 like, 6 or 7. Feb 13 22:01:47 and i had to do them manually for the most part (find the one it wants and install it as well) Feb 13 22:02:45 thanks Feb 13 22:03:30 the drop-down notification area is strange, i think. Feb 13 22:03:40 i feel like i should be able to just tap it, but that makes it appear and disappear rapidly Feb 13 22:03:44 i have to tap and drag Feb 13 22:03:45 jasta: I need to be able to wake the phone up from a server...and I really wish I didn't have to use gtalk Feb 13 22:04:08 well, the current de-facto standard with mobile devices is SMS. Feb 13 22:04:28 which is just so brutally locked down Feb 13 22:06:16 and expensive sometimes Feb 13 22:06:24 thats what i meant. Feb 13 22:06:35 so if GTalk was our option, that would be a relief. Feb 13 22:06:42 I think it is Feb 13 22:06:56 You can setup a Listener for it right? Feb 13 22:07:34 the new UI in google maps is very nice ;) Feb 13 22:07:42 yeah, for sure Feb 13 22:07:47 zhobbs: yes, but i'm just not sure if that's how they're positioning it. Feb 13 22:08:12 that is, is it really going to automatically sign on when the device powers on and just always be reachable via XMPP/GTalk? Feb 13 22:09:10 how do the messengers have open connections like they do? how are they not intensive? Feb 13 22:09:10 i dont see how to zoom in, tho Feb 13 22:09:26 jasta: i Feb 13 22:09:44 right, but that's not very convenient for this form factor Feb 13 22:09:51 should be able to tap twice or tap and hold Feb 13 22:09:56 oh, you can ;) Feb 13 22:09:58 tap and hold hehe Feb 13 22:10:39 michaelnovakjr: well, there was talk of optimizing GTalk substantially for mobile operation. Feb 13 22:10:42 that may be what theyre talking about. Feb 13 22:10:48 (in addition to the obvious) Feb 13 22:11:03 my application would benefit from being able to monitor my web services frequently Feb 13 22:13:21 michaelnovakjr: that's not a problem using an AlarmManager Feb 13 22:13:30 Is there somewhere we can see a vague list of the milestone expected modifications to the SDK? Feb 13 22:15:49 zhobbs: yea, i want to check back with the web services every minute or so... so yea kind of frequent... that is why i am worried about performance... its not a monster query by any means but it is frequent Feb 13 22:17:08 michaelnovakjr: yeah, would drain battery Feb 13 22:17:45 I have an app in mind where it's very important that a client is alerted of certain updates immediately Feb 13 22:17:59 that is pretty much what i am doing zhobbs Feb 13 22:18:36 so I need an always on service using sockets or something, but would be nice if there was a way to just wake the app up Feb 13 22:18:58 zhobbs, some kind of listener? Feb 13 22:19:02 yeah Feb 13 22:19:19 I noticed that there is a new attribute "android:port" for intent filters Feb 13 22:19:31 docs on it? Feb 13 22:19:39 haven't got that far Feb 13 22:19:46 just looking for the M5 Changes Feb 13 22:20:14 zhobbs, your project is chat based or messaging system? Feb 13 22:20:16 /s/for/over Feb 13 22:21:11 michaelnovakjr: neither...the app needs to be alert in case of certain emergencies Feb 13 22:21:11 sweet, just crashed the maps app. Feb 13 22:21:23 emergencies? Feb 13 22:21:26 I can't type today... Feb 13 22:22:07 yeah, if the server feels there is an emergency it needs to go out and wake the phone up to tell the user Feb 13 22:22:24 like a technical issue or something? Feb 13 22:22:52 yeah, i dont follow either. Feb 13 22:23:16 emergency is a weird word to use her e:) Feb 13 22:23:29 for example, a kid feels in harm, can press button to alert parents Feb 13 22:23:47 he should call 911 for that. Feb 13 22:24:13 for sure, maybe he pressed button, it calls 911 and alerts parents Feb 13 22:24:38 i am working on a warning system for mobile devices Feb 13 22:24:48 zhobbs: wel, regardless, no need to define it to that level. Feb 13 22:24:48 michaelnovakjr: what type? Feb 13 22:24:58 weather related Feb 13 22:25:01 why don't we just call it an alert. the specific meaning of that data can be interpreted by the receiver. Feb 13 22:25:26 jasta: exactly Feb 13 22:25:52 michaelnovakjr: oh yeah, me too :) Feb 13 22:26:00 what do you mean? Feb 13 22:26:10 michaelnovakjr, you should call it "Danger, Will Robinson!" Feb 13 22:26:45 haha Feb 13 22:27:29 so we can keep using handlers if we want to right? Feb 13 22:27:37 zhobbs, how's yours weather related? Feb 13 22:28:35 the app I'm about to finish is weather related Feb 13 22:28:49 a warning system zhobbs? Feb 13 22:29:02 i'm gonna port my news reader to the new SDK tonight Feb 13 22:29:29 The notification manager has changed how you add a visual notification to a running service, the upgrade docs are a bit thin on the ground as to how one would now apply an image on the status bar. Any ideas anyone? Feb 13 22:29:50 there's an API demo that does it. Feb 13 22:30:10 oh is there? Feb 13 22:30:10 thanks missed that Feb 13 22:31:09 oh yeah showNotification() Feb 13 22:38:49 does android support chinese input? Feb 13 22:40:44 mnf. I guess it does still support chat via a IChatSession Feb 13 22:47:37 zhobbs: not yet Feb 13 22:56:18 god I'm glad I kept the warnings clean. This upgrading is a task. Its going to have to be a concerted community effort to help revamp all of plusminus's tutorials Feb 13 22:59:12 chomchom: yes :( Feb 13 22:59:41 We realize that dropping a change of this magnitude on developers all at once is harsh Feb 13 23:00:01 We don't expect future releases to be this big a change, but then again we can't promise anything either :) Feb 13 23:00:28 I've got a few checked out that I'll try and contribute back, it would be a shame if new starts were put off by the amount of broken tutorials! Feb 13 23:00:34 fortunately for the Challenge (at least) you don't have to use M5, you can stick with M3 Feb 13 23:00:40 Its been a strength so far Feb 13 23:01:35 It would also be a shame if plusminus and other avid contributers to the tutorial forums got swamped with requests to update the old ones rather than making new ones! Feb 13 23:02:19 I'll be marking mine "depreciated" for now Feb 13 23:05:05 Yeah thats a good start Feb 13 23:08:12 i intend to update a few of my existing projects tonight. Feb 13 23:08:21 actually, it doesn't look like it'll take me more than a few hours Feb 13 23:08:23 if that Feb 13 23:12:54 So far the classes extending the ListAdapters the notifications and the manifests have needed the once over, but I must say the documentation provided for the diff, change over recommendations and emulator installation so far is excellent. It has made it as painless as I could imagine. Feb 13 23:22:01 morrildl: i take it from the rlsnotes posted that nio isn't implemented in this rls either? Feb 13 23:23:05 jerkface03: NIO seems implemented Feb 13 23:23:19 seriously? Feb 13 23:23:23 that's great news Feb 13 23:23:39 take a look to make sure Feb 13 23:23:46 looks like all my bitching actually resulted in something Feb 13 23:23:59 nio? Feb 13 23:24:10 yeah I was thinking the same thing Feb 13 23:24:14 whats nio? Feb 13 23:24:22 it stands for New I/O Feb 13 23:24:28 jerkface03: the diff is for API changes, basically Feb 13 23:24:29 it's non-blocking I/O Feb 13 23:24:45 non blocking? Feb 13 23:25:02 diff? Feb 13 23:25:21 jerkface03: http://code.google.com/android/migrating/m3-to-m5/changes.html Feb 13 23:25:32 looks neat. Feb 13 23:28:09 migrating content providers is a bitch ;P Feb 13 23:30:21 Does anyone know how to select a notification icon from the status bar? Feb 13 23:30:37 Yeah there is a tutorial on it Feb 13 23:30:46 in the provided API demos Feb 13 23:30:52 Dan_U: you mean, with your "finger"? Feb 13 23:31:24 well, whatever way you are supposed to. I spent hours trying to figure it out with the last sdk until someone told me I had to click the home button, then the up key Feb 13 23:31:31 I don't think you can with the emulator Feb 13 23:31:35 I select the icon now and it pops some weird panel up beneath the status bar Feb 13 23:31:37 Dan_U: drag down the status bar Feb 13 23:31:40 yep Feb 13 23:31:41 but I can't do anything with it Feb 13 23:31:47 that panel must be dragged down Feb 13 23:31:48 ah, I see Feb 13 23:32:00 it works beautfully on a touch screen Feb 13 23:32:05 hmm Feb 13 23:32:11 except there's nothing in the panel Feb 13 23:32:13 because when you touch the status bar, the panel appears and your finger also touches this panel Feb 13 23:32:19 because you have no notifications Feb 13 23:32:26 you can fling it to make it expand fully Feb 13 23:32:38 thats a nice idea Feb 13 23:32:38 but I was looking at an api demo for notifications Feb 13 23:32:44 so I do have the notification Feb 13 23:32:52 Dan_U: then drag it to the bottom of the screen Feb 13 23:32:56 it's still a work in progress Feb 13 23:33:00 ah, I see Feb 13 23:33:03 yeah, that needs some help Feb 13 23:33:05 right now the notifications stick to the top Feb 13 23:33:11 not very intuitive. Feb 13 23:33:15 again, it works beautifully on a real touch screen Feb 13 23:33:20 might be good if you didn't have to pull it down all the way Feb 13 23:33:26 yes, that's part of the plan Feb 13 23:33:36 to reveal the notifications early and then stick them to the top Feb 13 23:33:38 or something like that Feb 13 23:33:46 anyway, don't worry, we're working on it Feb 13 23:33:47 romainguy: the notifications should stick to the bottom so you can peek at them Feb 13 23:34:00 beat me to it Feb 13 23:34:01 I still prefer the way it was before, with the exception of having to click home then the dpad up Feb 13 23:34:33 I have one remaining error, and I don't understand it's output message. It results in a "Conversion to Dalvik format failed with error 1" in the problems view. The stack trace gives out a more verbose: UNEXPECTED TOP-LEVEL EXCEPTION: com.google.util.ExceptionWithContext: Couldn't map types in phi insn:v24:[[Z v23:Ljava/lang/Object; v22:Ljava/lang/Class; v19:I". ... ...while processing column_metadata (J)[[Z...while processing Feb 13 23:34:33 org/sqlite/NestedDB.class Feb 13 23:34:41 maybe touching the icon with your finger, then releasing should expand it all the way Feb 13 23:34:50 right now it "blinks" for an instant Feb 13 23:34:51 does anyone understand what its telling me? Feb 13 23:36:55 the diff page doesn't show the nio package as being changed? and doesn't really show any of the nio classes being changed. so it still hasn't been updated? Feb 13 23:37:25 jerkface03: the diff page just shows methods that were added or removed Feb 13 23:37:27 jerkface03: the API hasn't changed Feb 13 23:37:29 it does not say anything about internals Feb 13 23:37:31 it's been implemented Feb 13 23:38:18 chomchom: it's telling you dx exploded converting some Java bytecode to Dalvik .dex bytecode Feb 13 23:38:26 chomchom: you should never see that Feb 13 23:38:44 please file a bug in the issue tracker, and if possible attach the .class file that's failing Feb 13 23:41:44 I have another question about notifications. There's now two intents to use with the constructor. I'm looking at IncomingMessage example and I don't understand the appIntent. Feb 13 23:42:14 What does it do exactly? Feb 13 23:42:35 * jasta scratches his head Feb 13 23:42:52 how on earth can you instantiate a Uri.Builder from an existing Uri? Feb 13 23:43:52 myUri.buildUpon() ? Feb 13 23:44:01 http://code.google.com/android/reference/android/net/Uri.html#buildUpon() Feb 13 23:44:05 Oh. Feb 13 23:44:34 that would make sense ;) Feb 13 23:46:45 I'm not sure which .class is failing as it doesn't appear in the stack trace but I'l have a look. Any pointers in the meantime as to how I could go about getting my app to build? Some pointers? I've tried starting the emulator with -wipe-data and I've tried creating a new workspace and running it from there. Any more ideas? Feb 13 23:50:44 chomchom: are you using external/third-party JAR(s)? Feb 13 23:51:30 Yes, I'm using an odbc implementation of sqlLite for unit testing Feb 13 23:51:40 I'll try removing it Feb 13 23:51:49 yes, do try that Feb 13 23:55:52 Well that seems to have got the emulator started, thank you. There is no need for my app to use third party Jars so its not a really big problem just now. I'll file an issue with the jars I used. Feb 13 23:56:04 chomchom: thanks Feb 13 23:56:08 It's actually a compile-time error Feb 13 23:56:24 the Eclipse builder is determining that your project is out of date, so it's rebuilding prior to starting the emulator Feb 14 00:06:55 uhoh spagettios Feb 14 00:07:12 INFO/DEBUG(494): signal 11 (SIGSEGV), fault addr 00000000 Feb 14 00:07:27 normal? Feb 14 00:10:51 jerkface03: definitely not :) Feb 14 00:11:45 what was the link to bugtrac again? Feb 14 00:11:49 i'll submit the full log file Feb 14 00:12:00 it runs for like 2 seconds Feb 14 00:12:02 then it stops Feb 14 00:12:06 and it shows that :X Feb 14 00:12:20 shows that in the log view i mean Feb 14 00:13:31 The issue with the third party jar crash is "Issue 246", and as a side note. On the Issue submitter, if you create a new attachment form box and then don't enter anything it attaches an extra 0 bytes file. Feb 14 00:14:51 morrildl: the documentation for the changes to AlertDialog seem not to be accurate. Feb 14 00:15:00 you guys added an iconId parameter, but it's not discussed as one of the changes. Feb 14 00:15:53 morrildl: fyi, looks like the problem is with attempting to play a midi Feb 14 00:24:57 Is there supposed to be no docs for ScrollIndicator in this sdk? Feb 14 00:29:35 Dan_U: I don't think we documented it Feb 14 00:29:48 and ScrollIndicator is deprecated for now Feb 14 00:32:44 ported my RSS reader to m5. Feb 14 00:33:43 there are a few issues tho. Feb 14 00:34:35 is the default for progressbar's indeterminate now? Feb 14 00:36:21 odd, it was documented before. I actually extended ScrollIndicator for something. Feb 14 00:36:45 Is there supposed to be a replacement for ScrollIndicator? Feb 14 00:37:54 or maybe my question should be, is there some kind of slider control? Feb 14 00:39:47 Dan_U: ProgressBar is planned to become a slider as well Feb 14 00:40:02 but scrollindicator is not something we plan on supporting Feb 14 00:40:21 ok, thanks Feb 14 00:40:56 I believe anddev or another similar site has some code to show you how to use a progressbar as a slider Feb 14 00:41:42 ok, I'll look. It's probably about as simple as overriding some event. Feb 14 00:42:20 exactly Feb 14 00:42:36 btw, now progress dialog and progress bar can be used from non-UI threads Feb 14 00:43:47 cool Feb 14 00:48:50 well, I found it on helloandroid, but it's a bit incompatible with the new sdk Feb 14 00:48:58 since there apparently is no more onMotionEvent Feb 14 00:49:03 onTouchEvent() Feb 14 00:49:23 yeah, I thought it would be that, but I also saw onTrackballEvent and was wondering about that Feb 14 00:49:32 romainguy: really, directly without the need for a handler at all? Feb 14 00:53:21 jasta: that's why UIThreadUtilities was created Feb 14 00:53:41 ProgressDialog detects whether its method is being called on the UI thread Feb 14 00:53:50 if it's not, it spawns off a Runnable itself Feb 14 00:54:10 hmmm, i misunderstood what you meant. Feb 14 00:54:15 that i still don't like, for the same reason :) Feb 14 00:54:24 the show() still has to happen on the UI thread though Feb 14 00:54:57 jasta: you can create your Handler if you want Feb 14 01:15:01 * jasta starts rewriting his _data icon junk into the new scheme... Feb 14 01:17:42 Hmm, i think there may be a big problem here, actually. Feb 14 01:17:49 which was my issue with the original interface as well Feb 14 01:18:25 how can you optionally re-use a raw resource in place of a real file in ContentProvider.openFile? Feb 14 01:44:56 Is it possible to make a TextView clickable? I'm calling setClickable(true) on it and it doesn't help. Feb 14 01:45:53 is your textview singleLine=true? Feb 14 01:46:12 no, but I can try that Feb 14 01:46:17 no,that's the point Feb 14 01:46:20 don't :) Feb 14 01:46:26 ? Feb 14 01:46:49 so if you specify that it's a single line, it can't be clickable? Feb 14 01:46:51 so actually I don't think you can make it clickable with DPAD_CENTER Feb 14 01:47:10 but the touch click should work Feb 14 01:47:26 I'm clicking with the mouse. It just doesn't do anything Feb 14 01:48:05 every touch click has a delay Feb 14 01:48:14 you can't just do a rapid click Feb 14 01:48:37 I'm not sure I understand Feb 14 01:48:47 shouldn't it be just like clicking a button? Feb 14 01:49:07 hmm I'm reading the code and clicking a TextView should work Feb 14 01:49:32 I know it was possible in the last sdk using onMotionEvent Feb 14 01:49:47 but now that doesn't work Feb 14 01:49:59 are you using a OnClickListener? Feb 14 01:50:11 yes Feb 14 01:50:25 then I don't know what to say Feb 14 01:50:29 buttons extend TextView Feb 14 01:50:33 and they most definitely work :) Feb 14 01:50:37 hmm Feb 14 01:50:40 ok, let me experiment a bit Feb 14 01:52:43 what the hell Feb 14 01:52:49 i can't make ContentProvider.openFile work at all Feb 14 01:53:02 oh, and I just noticed with Eclipse that when I type R. that it won't autocomplete with drawable/string/array/etc. Feb 14 01:53:16 small annoyance Feb 14 01:59:23 Even just a simple example of trying to make TextView clickable doesn't work Feb 14 02:00:17 I created a new project and got the default "hello" example, and set the textview to be clickable and gave it an onclicklistener Feb 14 02:00:47 let me try Feb 14 02:03:34 hmm, i see, the file is being created with mode 0 Feb 14 02:06:36 Dan_U: just asked the TextView guy, it's working Feb 14 02:06:47 but it's not! Feb 14 02:06:48 lol Feb 14 02:06:59 TextView.onTouchEvent is returning false apparently Feb 14 02:07:09 I just extended it and returned true, and it works Feb 14 02:07:10 that does not matter Feb 14 02:07:21 the first thing onTouchEvent does is call super.onTouchEvent Feb 14 02:08:00 ok, but I don't think that's where it is returning false Feb 14 02:08:18 that's what I'm saying Feb 14 02:08:25 it should not matter what value you are returning Feb 14 02:08:38 apparently it does though Feb 14 02:09:03 What is the boolean value for? Isn't it to signal that the event has been handled? Feb 14 02:09:13 Dan_U: update from the TextView guy, yes it should return true if super returned true :) Feb 14 02:09:21 So it's bug and it'll be fixed in next release ;-) Feb 14 02:09:33 ok... Feb 14 02:09:48 Probably is a bug since I have TextView decompiled and there are cases where it might return false. Feb 14 02:10:11 "romainguy: So it's bug and it'll be fixed in next release" Feb 14 02:10:30 I think I might see what the problem is. Let me try some stuff Feb 14 02:11:24 yep, got it Feb 14 02:11:30 have to setFocusable(true) Feb 14 02:11:44 I'd think that setClickable(true) would/should imply that Feb 14 02:13:03 we had this discussion internally and decided against it Feb 14 02:13:09 ah Feb 14 02:13:11 because in touch mode the notion of focus disappears Feb 14 02:13:23 so you might want a view that is not focusable but clickable :) Feb 14 02:14:40 Maybe I don't understand what is meant by "touch". Wouldn't that mean that the functionality isn't available unless you have a touchscreen on your phone? Feb 14 02:14:51 Since you wouldn't be able to focus on something, how could you do the onclick? Feb 14 02:15:04 well then you would have to set it focusable Feb 14 02:15:28 I think I see Feb 14 02:15:48 you could have a notion of focus in touch mode if you wanted though Feb 14 02:16:07 although I'm not sure it'd work with some of the other functionality in other views Feb 14 02:16:08 it's called focusable in touch mode in the framework Feb 14 02:16:24 anyway Feb 14 02:16:25 I just tried Feb 14 02:16:38 and setClickable/setOnClickListener works perfectly with EditText Feb 14 02:16:54 must be just TextView then Feb 14 02:16:55 it's just for TextView that you need to work around our bug Feb 14 02:17:02 ok, not a problem, thanks Feb 14 02:17:10 and really, if you want to click a TextView, maybe it's a button then :)) Feb 14 02:17:25 anyway, the bug is logged and understood, thanks for the report Feb 14 02:17:31 well, when I developed my app against the last sdk, I didn't want the button look Feb 14 02:17:42 I suppose I could have overridden ondraw or something though Feb 14 02:17:44 in that case you can do a setBackground(null) Feb 14 02:17:49 that's the easiest way Feb 14 02:17:51 ah Feb 14 02:18:02 but I would argue that something clickable should look clickable Feb 14 02:18:09 and TextView definitely does not look clickable :) Feb 14 02:18:22 Well, this looks clickable by what text is in the textview Feb 14 02:18:33 I doubt the user would be confused Feb 14 02:19:47 Looks like I'm gonna have my work cut out for me trying to fix my apps to look good in the new UI. Feb 14 02:21:35 heh, there's an interesting feature of the edittext password functionality. For a split second, it shows you the character you just typed. Is that on purpose? Feb 14 02:21:43 yes Feb 14 02:21:53 imagine typing your password on a 12-keys phone Feb 14 02:22:00 true Feb 14 02:22:09 it's pretty common amongst phones to do this Feb 14 02:22:16 hopefully nobody is looking over your shoulder though Feb 14 02:22:33 they'd look at the keypad :) Feb 14 02:22:59 yeah, I guess they would Feb 14 02:23:06 so I guess it really wouldn't matter too much Feb 14 02:23:45 Since moving to the new SDK When I call on my Buttons via the (Button)findViewById(R.id.myButtonId) It returns null. I'm really not sure why all of a sudden these are being referenced as null. I switch back to the old SDk and they work fine. I've tried scrapping and regenerating the R file but it still always returns a null reference. Feb 14 02:23:47 unless they were good at inputting their password quickly, then it'd help to have the character show up for a split second Feb 14 02:25:16 chomchom: have you changed your layout to use android:id= instead of id= Feb 14 02:25:51 chomchom: all of your id in the xml files should be android:id= and not id= Feb 14 02:26:01 god damn it, It took it too literally when it said to replace everything with android:name! Feb 14 02:26:06 that was so stupid! Feb 14 02:28:07 that will undoubtebly be it. Feb 14 02:29:16 well, looks like I'm gonna be extending a bunch of UI components to get the look of my app back to where it was. Feb 14 02:29:36 And I don't really want to go back to the previous sdk Feb 14 02:29:58 No way, feel the pain now Feb 14 02:30:03 it wil save you in the long run Feb 14 02:30:08 right Feb 14 02:30:13 some things are ok Feb 14 02:30:20 but ImageButtons really got screwed up Feb 14 02:30:22 :-( Feb 14 02:33:16 Hooray! I've defeated the 800 odd errors. Feb 14 02:33:25 congrats Feb 14 02:33:29 And my application look alien. Feb 14 02:33:35 haha Feb 14 02:33:36 looks even Feb 14 02:33:38 join the club Feb 14 02:34:06 The chunky interface has really messed up some of my stuff. Feb 14 02:34:18 Going to have to have a rethink about the interface Feb 14 02:34:48 personally, I think I'm just gonna try to mimic the components the way they were in the last sdk Feb 14 02:35:31 Dan_U: I think a better idea would be to come up with something that is neutral and not implementation specific Feb 14 02:35:55 yeah Feb 14 02:36:19 really, it would have been nice if these changes were handled like a theme, where we could just specify the old theme versus the new theme Feb 14 02:37:20 That way you are proofing yourself against another fall. God knows how these phone peeps will customise their offerings, you want to give your application an identity unto itself so it becomes just incidental that it is running no the android platform. Feb 14 02:37:34 Yeah a theme is a good idea Feb 14 02:37:42 I think I'll be having one of those Feb 14 02:37:59 funnily enough thats my motto, "You've got tot have a theme". Feb 14 02:38:48 I don't even understand why the UI was changed. Feb 14 02:38:48 One thing that has not been mentioned but I can now see, it the nice improvements to the map UI Feb 14 02:38:56 Thats a really nice zoom tool guys Feb 14 02:39:00 very zippy Feb 14 02:39:05 yeah, I saw that zoom tool. It's nice Feb 14 02:39:37 Dan_U: because the previous UI was too small for touch devices and this is something we want to support Feb 14 02:40:06 I thought about that, but I tried to compare the size of the UI controls to the buttons on the emulator and they seemed fine to me Feb 14 02:40:27 Dan_U: Trust me, we tried Feb 14 02:40:38 and the emulator is way bigger than a real device Feb 14 02:40:42 Dan_U: Chunkiness is a constraint, embrace it. Let it empoer you. Feb 14 02:40:47 empower even Feb 14 02:41:28 I believe they tried. Feb 14 02:41:31 true, but relatively speaking, the buttons (like back, home) on the emulator were the same size or smaller than the 32px square imagebuttons I was using. Feb 14 02:41:53 now they seem huge Feb 14 02:42:00 my imagebuttons I mean Feb 14 02:42:22 I would imagine that google have done plenty of research on UI usability regarding touch screens. Feb 14 02:42:52 It would be awful nice if they could share some of their study findings with the developers for their new platform Feb 14 02:43:10 they could maybe help influence dome design decisions. Feb 14 02:43:18 it seems to me touching a real button on the phone wouldn't be much different than touching the touchscreen Feb 14 02:44:05 Dan_U: HVGA devices usually have a high dpi, something like 160 dpi Feb 14 02:44:20 your computer screen is far from having such a high dpi resolution Feb 14 02:44:25 ah Feb 14 02:44:32 shows how much I know about mobile phones :-) Feb 14 02:46:36 so if that's the case, does that mean what I see in the emulator is going to look different in a real phone in terms of the size of my imagebuttons? Feb 14 02:46:48 Dan_U: *yes* Feb 14 02:46:51 very much so Feb 14 02:47:02 that is very scary Feb 14 02:47:04 the physical size on your monitor is probably easily 50% greater than the handset will be. Feb 14 02:47:10 well, true Feb 14 02:47:22 but I'm talking about relative to say the actual physical buttons on the phone Feb 14 02:49:28 (probably, tho) Feb 14 02:50:22 sorry, got disconnected for a sec Feb 14 02:50:55 * jasta eagerly installs the new SDK at home :) Feb 14 02:51:50 about the screen size, I guess I need to give an example. I have 5 imagebuttons at the bottom of my activity and one is mangled/cut-off because there is not enough space... Feb 14 02:52:04 on actual hardware, is this still the case? Feb 14 02:52:17 why wouldn't it be? Feb 14 02:52:33 Dan_U: yes it will be the case Feb 14 02:53:03 well, with the image size being 32 pixels square and dpi being 160, isn't it different? Feb 14 02:53:13 the image will still be 32 pixels wide Feb 14 02:53:47 I guess since I've never worked with this stuff before, it gets me confused when talking about screen res, when I'm looking at an emulator on my computer screen Feb 14 02:54:06 well Feb 14 02:54:12 the layouts will remain the same Feb 14 02:54:14 but if you take a ruler Feb 14 02:54:26 and measure your buttons on your computer screen, they might be, let's say 1inch long Feb 14 02:54:30 then you do the same on a phone Feb 14 02:54:37 they will be for isntance 0.5inch long Feb 14 02:54:46 but they will still use the same amount of pixels Feb 14 02:54:49 I mean, I'm looking at my app right now and I think about the width of my 5 imagebuttons and approximately what the padding is and I think I need 200 pixels wide Feb 14 02:56:13 actually maybe it's a bit more than that, but not much Feb 14 02:58:39 maybe I'm confusing dpi with pixels Feb 14 02:58:51 they are considered different right? Feb 14 02:58:59 yes Feb 14 02:59:07 dpi means dot per inch Feb 14 02:59:10 ok, then that's probably my problem Feb 14 02:59:17 yeah, I knew that, but I made some bad assumptions Feb 14 02:59:19 roughly, how many pixels a screen can display on one inch of space **** ENDING LOGGING AT Thu Feb 14 02:59:57 2008