**** BEGIN LOGGING AT Tue Nov 20 02:59:57 2007 Nov 20 03:15:19 teehee, this just really tickles me. Nov 20 03:16:00 I just have to share... Nov 20 03:19:09 lol Nov 20 03:45:37 http://jasta.dyndns.org/android-rss.jpg Nov 20 03:45:38 hehe Nov 20 03:46:14 zomg i know ur codes Nov 20 03:46:35 for some reason, my window manager borders got cut off when i took a screenshot, oh wel. Nov 20 03:46:38 well* Nov 20 03:46:59 duey: uhh, you'll know them much better when i open access to the cvs repository at the end of the week. Nov 20 03:47:19 lol Nov 20 03:47:29 jasta: can u give me that link to the symbian article? Nov 20 03:47:48 from the preview it sounds like symbian is crying Nov 20 03:47:49 i'm just amused by the irony of testing my Android RSS reader and i randomly found that :) Nov 20 03:48:05 everyone is crying. the news is all negative against Android from all the people who desperately need it to fail :) Nov 20 03:48:11 indeed Nov 20 03:48:15 i think its great Nov 20 03:48:31 me too. the response was so sudden that it just makes you chuckle. Nov 20 03:48:52 days after google makes an announcement, all of the competitors come out of the woodwork saying "no, no, ignore them, ignore them please!!!!!" :) Nov 20 03:49:07 anyway, it's on smartphonethoughts.com, you should be able to find it. Nov 20 03:50:08 duey: i'm planning to release this RSS reader app as a kind of tutorial. Nov 20 03:50:33 for what a complete, functional app looks like. Nov 20 03:50:47 gotta start working on notifications and stuff now. Nov 20 03:50:56 and a background service that downloads new posts every hour Nov 20 03:51:01 instead of making the user click "refresh" like they do now Nov 20 03:51:30 you could probably get webkit to do all the work for you Nov 20 03:51:44 all what work? Nov 20 03:51:52 gui display Nov 20 03:51:59 for the entire app? that's ridiculous. Nov 20 03:52:04 what is this, an iphone? Nov 20 03:52:05 is it really? Nov 20 03:52:09 yes, it's totally ridiculous. Nov 20 03:52:26 and yes Nov 20 03:52:29 its an iphone clone Nov 20 03:52:30 from top to bottom Nov 20 03:52:36 webkit? Nov 20 03:52:39 iphone uses webkit Nov 20 03:52:43 it even has iphone icons! Nov 20 03:52:53 linux? the iphone uses linux. open source? the iphone is open source. oh yeah, i see your point now. Nov 20 03:53:26 iphone is darwin/bsd Nov 20 03:54:02 i refuse to be bated by your obvious trolling, sir. Nov 20 03:54:11 lol Nov 20 03:54:12 back to writing software on this wonderful, open platform. Nov 20 03:55:05 it isn't true, it can't be! I CANT HEAR YOU!! Nov 20 04:02:17 hehe Nov 20 04:02:23 duey, biut you cannot run native code on android Nov 20 04:03:22 duey, what is the relationship between webkit and java UI? Nov 20 04:30:13 I set up eclipse and android at home at it's running fine. I've tried to set it up at work, and when I run a project, it runs the emulator - but doesnt push the application to it. While the emulator is running I tried "adb shell" and it says - *daemon not running* error:no device. I tried doing an "adb kill-server". I've also tried following suggestions from this thread: http://groups.google.com/g Nov 20 04:30:15 roup/android-developers/browse_thread/thread/4f460845bb7d5747/94d01ab9e0d252bf?lnk=gst& but no go! Does anyone know what else i might check? Nov 20 04:31:29 (ive also done "reset adb" from the ddms menu in eclipse) Nov 20 04:40:00 and done a little jig. Nov 20 04:40:06 for good luck. Nov 20 04:49:30 mrspeaker: which eclipse version? Nov 20 04:51:31 eclipse version 3.3.1.1 - with the m3-rc22a version of the sdk. When i set this up the first time it was with the orignial version of the sdk, which I later updated. now I have just used the new version of the sdk. If that makes any difference! Nov 20 04:53:21 also, my work machine is "locked down" more than my home machine... is there anything that might be blocking the eclipse talking to the emulator? Nov 20 04:53:55 good question Nov 20 04:59:09 i think i read about issues with certain launch configurations, but i cant find it anymore Nov 20 04:59:19 ever tried the DDMS view? Nov 20 05:05:34 yeah - thats where I did the "reset adb" command from (I have to do that sometimes on my other machine when it doesnt push the app). i think its a sign anyway - the your're-really-not-supposed-to-be-doing-this-at-work sign. Nov 20 05:06:21 nothing in the log? Nov 20 05:19:25 hmmm... no not a thing? how do I get it logging?! Nov 20 05:20:27 actually - when I hit run, a console window pops up briefly and says "Ping not..." and that's all I can read before the window closes ;) Nov 20 05:22:37 you want to open the logcat eclipse view Nov 20 05:31:26 http://www.blinkerfluid.org/logcat.jpg Nov 20 05:31:37 that's what it should look like Nov 20 05:33:54 hmmm my DDMS view has no entries - no "devices", nothing in processes, file expplorer, logcat... it's just blank. Nov 20 05:36:09 but the emulator launches? Nov 20 05:39:02 yeah - emulator launches to the home screen, but the app isn't pushed to it. Nov 20 05:40:36 strange Nov 20 05:44:50 i think i must have a blocked port or some-such... time to search the docs! Nov 20 05:44:59 good morning Nov 20 05:45:42 mrspeaker: good luck :) Nov 20 05:45:48 Zoolooc: good morning Nov 20 06:04:10 mrspeaker: should be tcp port 5555 the emu is listening on Nov 20 06:44:23 eh, going home where it works real-good. Thanks for all your help though gang! Nov 20 08:36:28 morning Nov 20 09:00:04 ehlo Nov 20 11:25:26 anyone know what these inflateParams are? Nov 20 11:28:42 dammit, I want source code:-/ Nov 20 11:32:58 you *can't* handle source code. Nov 20 11:34:59 hahah Nov 20 11:35:25 "You want the source code? You can't HANDLE the source code!" Nov 20 11:51:07 Morning all :-) Nov 20 11:51:35 morning Dralspire Nov 20 11:51:48 morning Nov 20 13:23:45 Dralspire, i just updated the board i mentioned some days ago Nov 20 13:24:00 yay :-) Nov 20 13:24:03 wanna register? :) Nov 20 13:24:10 let's take a look :-) Nov 20 13:24:24 and share the link with the channel :-) Nov 20 13:24:37 http://forum.android-portal.com/ Nov 20 13:24:39 :) Nov 20 13:24:56 yay orange :-) (I like orange :D) Nov 20 13:24:58 i have to link it to the main page Nov 20 13:25:02 me too Dralspire Nov 20 13:29:33 It's good you got it set up :-) Nov 20 13:50:25 sorry for the double post to all who read the developers mailing list Nov 20 14:13:47 is jythonc working with Dalvik? Nov 20 14:13:56 jython you mean Nov 20 14:14:18 it should work, but it wouldn't support all android's api Nov 20 14:14:20 gpolo nop, the bytecode compiler of jython Nov 20 14:14:50 to use the bytecode compiler you used something like.. jython, right ? Nov 20 14:15:11 the command is jythonc, yes Nov 20 14:15:16 i know what is the command Nov 20 14:15:30 but what provides it ? Nov 20 14:15:32 it is jython Nov 20 14:15:36 ok, im not continuing with this Nov 20 14:15:41 as i understand the bytcode of dalvik is not compatible with suns, right? Nov 20 14:15:51 android has a tool called dx Nov 20 14:15:57 that converts java bytecode to dalvik Nov 20 14:16:06 gpolo ah okay, thanks! Nov 20 15:20:06 Hello everyone :) Nov 20 15:20:13 Hi hi :-) Nov 20 15:20:51 I wish I could stay home all day.. Between the groups and IRC, I just had like an hours worth of reading to do >.< Nov 20 15:21:07 now I'm late for werks... Nov 20 15:21:10 :-) Nov 20 15:21:15 heh, cya Nov 20 15:31:14 hi all, any google/android devs in the room? Nov 20 15:37:17 dims: I can see at least one Nov 20 15:51:10 dims! Nov 20 15:51:21 you still working on web service security? Nov 20 16:02:07 Essington: sometimes :) Nov 20 16:03:51 Essington: So, how are you doing? What have u been upto :) Nov 20 16:04:12 dims: I've been working with GWT Nov 20 16:04:39 and a bit with bouncy castle openpgp stuff Nov 20 16:04:55 separate projects Nov 20 17:16:14 Hi, does someone know the number of people working at Google for Android (i.e developers) ? Nov 20 17:17:05 That would be interesting to know. Nov 20 17:17:15 is it for a study ? Nov 20 17:17:27 Heard that they're around 50 at Trolltech working on Qtopia Nov 20 17:17:42 woah Nov 20 17:17:56 50?? Nov 20 17:20:16 Lede: http://lists.openmoko.org/pipermail/openmoko-devel/2007-November/001704.html Nov 20 17:20:40 I remembered wrong, it wasn't 50 at Trolltech, 50 developers Nov 20 17:20:56 So don't quote me on that first thing ;) Nov 20 17:23:42 hm Nov 20 17:25:27 why arnt there more frequent updates on the official android blog anyway, like adressing problems and what is being worked on Nov 20 17:26:03 because then they wouldnt get any real work done Nov 20 17:26:42 hehe Nov 20 17:27:15 :) Nov 20 17:27:53 they cant even ship a updated version of adb in a new package Nov 20 17:27:56 >_> Nov 20 17:28:09 "-version now shows version number" uh yeah right Nov 20 17:28:32 And someone know when Android project has started? Nov 20 17:29:06 edu, 12.11.2007 Nov 20 17:29:22 google acquired the company in 2005 Nov 20 17:31:13 no no, the date when the team Android group started... does someone know? Nov 20 17:31:14 I mean, I'd like to know more info about Android Team... Nov 20 17:31:34 Lede: thanks 2005 Nov 20 17:32:09 edu: i check the wikipedia article, so you might find more useful information there Nov 20 17:32:16 s/check/checked Nov 20 17:32:31 Lede: thks Nov 20 17:38:14 hmmm. Nov 20 18:50:09 * morrildl is back Nov 20 18:50:40 morrildl: hello Nov 20 18:52:12 morrildl: not a major problem, but the adb shipped with android_sdk_linux_m3-rc22a is the same one as in android_sdk_linux_m3-rc20a, according to the release notes of the newer it should have a minor change (adb -version should return version number) Nov 20 18:53:01 morrildl: i have one more question, are there plans for supporting .svg files? Nov 20 18:53:07 :> Nov 20 18:55:29 version doesn't seem to work does it... oh well Nov 20 18:56:54 Lede: interesting, thanks for letting me know Nov 20 18:57:51 Lede: not currently. It should be possible to build an SVG engine using the Android/SGL primitives, but there are no current plans to do so (we don't have any use cases for it yet) Nov 20 18:58:36 i hope you find use cases for it :) Nov 20 18:58:37 morrildl, I have one I've been thinking about some... you have all these XML attribute names that parallel a bunch of Java methods. Wouldn't it be cool to simply use the xml attributes? setAttr(String name, value) where value could be various types Nov 20 18:58:50 it would be handy for me at least:-) Nov 20 19:00:19 davidw: I'm not sure I follow, can you give an example? Nov 20 19:00:28 hi Nov 20 19:01:08 anybody knows how can i get the current user home path? Nov 20 19:01:33 morrildl, ok, for instance, TextView: android:capitalize setInputMethod(InputMethod) or android:height setHeight(int) Nov 20 19:01:59 morrildl, I was contemplating the possibility of utilizing the xml attribute names from Java Nov 20 19:03:24 nobody? Nov 20 19:03:38 are "layouts" similar to "content panes" or "panels" Nov 20 19:04:37 yeap Nov 20 19:04:43 or form, containers Nov 20 19:04:46 :P Nov 20 19:05:50 hmm.. i wish there was a simple guide explaining how the hierarchy should be set up (with pictures) Nov 20 19:06:14 im very visual. reading through the code isnt all that exciting to me Nov 20 19:06:51 ViaToR_SG: why do you need it? Nov 20 19:07:45 davidw: I guess I still don't understand your use case. It sounds like you're already using Android's XML-based layout... How would you use SVG in that scenario? Nov 20 19:08:09 morrildl, svg is someone else's question Nov 20 19:08:23 davidw: ah, that would explain my confusion :) Nov 20 19:08:51 morrildl: i'm making a open final dialog and I would add a home button Nov 20 19:08:55 then... Nov 20 19:09:06 I need user home path Nov 20 19:09:11 I try System.getProperty("user.dir") Nov 20 19:09:14 but not working Nov 20 19:09:14 ViaToR_SG: there is no concept of a user home path Nov 20 19:09:22 each application has its own user ID Nov 20 19:09:36 e Nov 20 19:09:53 Also, in general applications will not be able to access the entire filesystem but instead only a tree they own Nov 20 19:10:18 For instance, eventually you probably won't be able to access files in / Nov 20 19:10:28 As I recall, those access restrictions aren't all implemented yet though Nov 20 19:10:46 the path /sdcard will be an exception, of course Nov 20 19:11:44 the final access restrictions policy hasn't yet been decided. I think it may come eventually end up just being permissions-based: that is, /sdcard will be world-readable and probably world-writable so all apps will be able to access it. Otherwise applications will probably only have permission to see their own files Nov 20 19:12:24 mmm... and if i would save my pictures or mp3 in native memory??? Nov 20 19:12:32 I cant? Nov 20 19:12:52 ViaToR_SG: you can, but they will be stored in a particular application's tree Nov 20 19:13:31 If you connect an Android device to a PC, it will appear as a USB mass storage device, but that will only allow you to access the contents of the external storage (e.g. SD card) Nov 20 19:13:39 ... ok then... the home path of this case... I suposse that its app path Nov 20 19:13:55 Users won't be able to drag/drop files onto the built-in memory, in other words Nov 20 19:14:07 right Nov 20 19:14:23 thx :) Nov 20 19:14:43 ViaToR_SG: you probably want /data/app/ Nov 20 19:14:56 specifically, /data/app/ Nov 20 19:15:13 oops, sorry Nov 20 19:15:25 I meant /data/data/ Nov 20 19:15:34 /data/app is for .apk files that are installed Nov 20 19:15:47 /data/data/* is for application data. For instance if you create a SQLite database the files get stored there Nov 20 19:17:22 i need to store mp3 files... Nov 20 19:17:52 but the "." application path is "/" xD Nov 20 19:20:11 hi Nov 20 19:20:22 ViaToR_SG: you should be able to access /data/data/ Nov 20 19:22:50 yes i know... but i would do some more raltive... I want not scpecificate "/data/app/myapp/"... I would something like ".", init... Nov 20 19:23:12 but ".", nows, link to root Nov 20 19:23:21 ViaToR_SG: okay -- I'll mention that to the folks working on the filesystem stuff Nov 20 19:26:33 how do I view the "title_editor" layout in the NotePad application? I dont see any place to edit a title Nov 20 19:26:42 thx morrildl :) Nov 20 19:28:07 d03boy: the title_editor.xml file contains an EditText field? Nov 20 19:28:33 ya, and a button.. Nov 20 19:30:37 I like how the adb package install thing kicks you out of your running app - that's handy to eliminate one extra step when debugging/reloading Nov 20 19:30:57 davidw: yeah :) Nov 20 19:31:20 I like how overwriting a package doesn't require a lengthy reboot, unlike some platforms! :) Nov 20 19:31:28 * orrc glowers at the BlackBerry Nov 20 19:31:39 BB is poopy Nov 20 19:32:33 it is indeed. if only I could work write android applications instead! Nov 20 19:35:21 er, work *writing* android apps Nov 20 19:36:24 * Dralspire giggles Nov 20 19:37:02 morrildl: can you say how simulating SMS is going to work? Nov 20 19:44:40 ViaToR_SG: see the method Nov 20 19:44:49 ViaToR_SG: see the method Context.getFileStreamPath(String) Nov 20 19:45:09 That will return you a File object which you should be able to use to figure out the directory name Nov 20 19:45:38 It's a little bit of a kludge, because it seems like there should be a method to just request that directory name straight up, but for now that should get you running in a reasonably "standard" way Nov 20 19:46:30 orrc: what do you mean by "simulating"? Nov 20 19:47:36 morrildl: simulating SMS messages to/from the device, since the emulator doesn't have GSM access Nov 20 19:48:47 orrc: let me check into that Nov 20 19:49:10 thanks! Nov 20 19:49:44 well... I'm using /data/app + AppName now... works tmp Nov 20 19:57:10 evening Nov 20 19:57:52 guardian: evenin' Nov 20 20:07:02 orrc: currently there's no way to do that, but we are working on it now Nov 20 20:09:21 morrildl: ah ok, I didn't think it was possible just now; I was just curious as to how it's planned to work. thanks for checking into it. Nov 20 20:14:29 morrildl what is your function at google if i may ask, are you directly involved with android? Nov 20 20:16:49 oh wait, woops Nov 20 20:16:52 you are Dan Morrill Nov 20 20:16:53 ._O Nov 20 20:17:11 * Lede bows Nov 20 20:18:35 Lede: right :) Nov 20 20:19:18 * Dralspire giggles Nov 20 20:26:50 wow, 108 users Nov 20 20:26:59 Does anyone how I can style the text I pass to a ListAdapter? Nov 20 20:27:47 I've got to make the text spannable somehow, but in the faq it says you do that by using setText and the flag TextView.BufferType.SPANNABLE Nov 20 20:27:57 morrildl: 107 ;) Nov 20 20:28:15 univac: eh, it was less than a 1% change, so it doesn't matter ;) Nov 20 20:28:29 ;-) Nov 20 20:28:41 haavi: are you referring to the per-row text? Nov 20 20:28:48 yes Nov 20 20:29:02 what are you trying to span? Nov 20 20:29:21 the text I pass to my ListAdapter Nov 20 20:29:28 the ArrayList Nov 20 20:29:41 right, but span across what? Nov 20 20:29:46 Are you using a TableLayout...? Nov 20 20:30:28 ListView Nov 20 20:31:26 ListAdapter adapter = new ArrayAdapter(this, Nov 20 20:31:26 R.layout.project_row, Nov 20 20:31:26 projectNames); Nov 20 20:31:26 setListAdapter(adapter); Nov 20 20:31:43 I need to style the text in projectNames somehow Nov 20 20:32:24 haavi: right, that's generally what you do in your project_row.xml file Nov 20 20:32:32 project_row is a TextView Nov 20 20:32:43 then there's nothing to span Nov 20 20:32:56 yeah but I have to do it programatically because what I have to style depends on certain criterias Nov 20 20:33:04 haavi: ahh Nov 20 20:33:13 Then you'll need to subclass ListAdapter and build in your own logic Nov 20 20:33:24 ok.. Nov 20 20:33:35 The ListAdapters generally just map names from data you provide to XML id-s in the layout Nov 20 20:34:01 it's a simple one-to-one mapping, so if you need more sophisticated styling as you describe, you'll need to handle it yourself Nov 20 20:34:29 yeah so I have to extend ListAdapter and then override getView? Nov 20 20:34:41 haavi: that sounds right, yes Nov 20 20:34:42 or how does it set up the views for each row? Nov 20 20:34:54 getView should be called once for each row Nov 20 20:34:55 let me check Nov 20 20:35:43 haavi: yes, that looks right Nov 20 20:35:52 ok I'll try Nov 20 20:35:56 thanks alot Nov 20 20:36:12 np! Nov 20 21:43:22 this faq wasn't in the faq I was reading: where are the phones? Nov 20 21:43:59 nowhere Nov 20 21:44:12 non-existant Nov 20 21:44:13 Nada Nov 20 21:44:16 yet =) Nov 20 21:44:23 ok Nov 20 21:44:48 release early release often software first Nov 20 21:46:37 agile Nov 20 21:46:49 heh Nov 20 21:47:10 wow that must suck having an alias with such a common programming word Nov 20 21:48:22 /nick android Nov 20 21:48:24 it wasn't quite as common when I began using it but it's always nice to hear none-the-less Nov 20 21:48:36 lol Nov 20 21:48:38 :D Nov 20 21:55:20 does anybody know if there's a way to control transitions between Activities? Nov 20 21:55:42 for example, when you first start your main Activity, it sort of fades and zooms in Nov 20 21:56:21 is there a way to control that yourself? so you could create an iPod-esque slide transition between screens etc. Nov 20 22:07:48 morrildl, do you have any idea what sort of Map one should pass to a widget's constructor? new Button(Context, AttributeSet, Map) Nov 20 22:08:09 or is that just something that's impossible to hijack for one's own purposes... Nov 20 22:08:43 is that a java.util.Map or is that a GPS map? Nov 20 22:08:50 java.util.Map Nov 20 22:09:31 surely it doesn't matter, otherwise the constructor would require a specific Map implementation? Nov 20 22:09:56 I think he's asking what the map should contain, not the implmeemntatino Nov 20 22:11:02 yeah.... it's not documented much, other than to say that it's basically used to create stuff from XML Nov 20 22:11:05 ah, ok Nov 20 22:12:40 public Button(Context context, AttributeSet attrs, Map inflateParams) Nov 20 22:12:54 i think its for the icon? Nov 20 22:12:56 maybe Nov 20 22:14:17 "Additional run-time inflation parameters." says the docs. not so handy :) Nov 20 22:15:18 the WidgetInflate docs expand on that a bit saying, "Currently, this parameter is only intended to be used internally by Android." Nov 20 22:15:28 inflateParams Optional map of additional attribute name/value pairs to apply to the root element. Can be null. Currently, this parameter is only intended to be used internally by Android. Nov 20 22:15:49 http://code.google.com/android/reference/android/view/ViewInflate.html Nov 20 22:16:48 davidw: the process of building an in-memory representation of a layout defined in XML is known as "inflating" the layout Nov 20 22:16:49 aha, that's a little bit more info at least Nov 20 22:17:15 * d03boy insists someone write that down Nov 20 22:17:30 The various layout parameters from the XML (like width and height and so on) need to get passed to the in-memory object Nov 20 22:17:36 That's basically what that Map is for Nov 20 22:17:41 you could probably use it, but I wouldn't bother Nov 20 22:17:52 you sure that's not what the AttributeSet is for? Nov 20 22:18:00 lol Nov 20 22:18:22 well, they're both used, I think Nov 20 22:18:35 I'm just not sure what each one does Nov 20 22:18:41 So, is there any way to reuse an Activity or improve efficiency of layout and view construction? Nov 20 22:19:03 Imagine an app that shows an activity listing entries, and when you click on an enry, it shows the entry text as a separate Activity;. Nov 20 22:19:19 then, the user presses back, which destroys the activity, and moves down one entry in the list, then picks another. Nov 20 22:19:40 here we go again, creates a new activity instance of the same type with all the same controls, but different data. Nov 20 22:20:18 and your having efficiency problems with that? Nov 20 22:20:27 duey: Yes, try creating a WebView in the Activity. Nov 20 22:20:41 can you reuse a the webview? Nov 20 22:20:56 When you expect that the user will close and open the same Activity class with a different instance 10 or 20 times in a given session, it seems really wasteful. Nov 20 22:21:10 duey: That's what I'm asking: can you reuse the Activity it belongs to? Nov 20 22:21:11 sure but thats what garbage collection is for :o) Nov 20 22:21:24 duey: No, that's not what Garbage Collection is for. Nov 20 22:21:27 heh Nov 20 22:21:42 GC is for the re-use of resources, not data. Nov 20 22:21:54 i was meaning reuse the webkit object in your acitvities Nov 20 22:21:57 in this case, you want to re-use substrate data underneath the views and activity. Nov 20 22:22:24 davidw: okay Nov 20 22:22:34 davidw: yeah I had them backwards -- attr is the XML attributes Nov 20 22:22:35 duey: How would you do that? You can't pass a View to an Activity, and you can't permit an Activity to store data in memory after it is destroyed (onDestroy() is called) Nov 20 22:22:37 I was going to say that the Activity would remain on the history stack, but since you're explicitly closing it, I don't know what happens internally Nov 20 22:22:39 layoutParams is system-level stuff Nov 20 22:22:50 orrc: What happens internally is that it is destroyed, completely. Nov 20 22:22:56 currently the only thing it's used for is telling the layout engine where the status bar is Nov 20 22:23:12 when you press back on an activity at the end of the history stack, onDestroy() gets called right away. Nov 20 22:23:39 so as the user goes down in the list, opening and closing Activity's, they end up wasting a lot of cycles re-doing the same exact layout over and over. Nov 20 22:23:41 jasta: right, ok. which sucks given that it's in the same, active process Nov 20 22:23:49 morrildl, really? I thought that that's where all the xml attributes got fed into the java code in order to set up a widget/view/something... that was my "mental model" at least Nov 20 22:24:02 orrc: Perhaps this is something Google needs to rethink about the Application Lifecycle? Nov 20 22:24:04 davidw: that is the attrs Nov 20 22:24:19 davidw: system-level layout properties are in layoutParams Nov 20 22:24:49 orrc: It seems like perhaps there should be a way to allow you to re-use an Activity and just change out it's data, not it's layout/view. Nov 20 22:24:59 But that would need to be implemented by the Window Manager. The app can't affect that. Nov 20 22:25:01 jasta: I though OnPause() is called when yo go back? Nov 20 22:25:06 thought* Nov 20 22:25:09 duey: No, onPause() is called when you go forward. Nov 20 22:25:16 jasta: it's not that simple, unfortunately Nov 20 22:25:22 oh right Nov 20 22:25:34 morrildl, aha... mmmm so they're sort of an override? Nov 20 22:25:37 morrildl: Trust me, I know. But it is a solvable problem, and an important optimization to be made. Nov 20 22:25:44 an Activity may or may not be a separate process in a given context. More to the point, it has to be prepared to be both in the same process as the caller, as well as its own process Nov 20 22:26:19 morrildl: What would need to be implemented is a way for the Window Manager to allow stale Activity's to persist when they are not in the history stack anywhere, and expose a new set of mechanics to reuse it's data with a new Intent(). Nov 20 22:26:20 jasta: this complicates the prospect of reusing activity instances Nov 20 22:26:35 morrildl: I don't think it's impossible, or even implausible. Nov 20 22:27:07 I didn't say it was impossible, or implausible, just that it's more complicated Nov 20 22:27:12 (aside) this is all rather complex... if carriers ever implement unlimited data plans on a wide scale, javascript web apps are going to clean up, just as they're starting to do with low-hanging desktop fruit:-) Nov 20 22:27:24 What would need to be implemented is an onReuse and onDiscard. onDiscard() would be an intermediate step that happens before onDestroy(). onReuse() would supercede onCreate() in some cases. Nov 20 22:27:57 the complicated part is that the Intent underneath the Activity would change. Nov 20 22:28:14 so any cursors and things would need to be invalidated. but the application developer can and should manage that in the onDiscard() and onReuse() calls. Nov 20 22:28:27 jasta: a hidden performance issue there is that maintaining the cache of Activities itself consumes memory Nov 20 22:28:52 morrildl: Yes, which is why an application would need a mechanism to request this behaviour. Nov 20 22:28:54 jasta: what are you specifically doing that is causing performance issues? Nov 20 22:29:03 to say, startActivity(..., yes_I_want_this_one_to_stick); Nov 20 22:29:15 and a way to say finish(..., nuke_all_my_stickies); Nov 20 22:29:18 jasta: it could also be a flag on the Intent Nov 20 22:29:47 morrildl: yes, the implementation specifics are open to a lot of extra discussion. Nov 20 22:30:04 bye all Nov 20 22:30:39 morrildl: I have an RSS reader app I am writing (to learn, mostly), and the transition from the RSSPostList activity to the RSSPostView activity is a common step. RSSPostView must implemented a WebView since most RSS posts include images and extra HTML gunk. Nov 20 22:30:50 brb, 1s, sorry. Nov 20 22:32:40 back. Nov 20 22:32:46 morrildl: must implement, I mean. Nov 20 22:33:11 jasta: are you constructing your layout from code, or from XML? Nov 20 22:33:20 the one including the WebView I mean Nov 20 22:33:28 morrildl: From XML, in this case. Nov 20 22:33:58 the performance hit I see is that the Activity opens, showing the "rest" of the layout which is not in the WebView, then the WebView appears some time later. Nov 20 22:34:08 so is the expensive operation creating the webview? Nov 20 22:34:19 It certainly seems to be. Nov 20 22:34:24 jasta: what content are you feeding the WebView? Nov 20 22:35:17 this is a bit of a hack, but couldn't you put init the webview in a static class and then add that webview to the layout? Nov 20 22:35:24 morrildl: The text of an RSS feed post, which is usually a couple of paragraphs with an inline'd in some cases, other cases it is like

click here!

at the footer. Nov 20 22:35:25 s/put/ Nov 20 22:36:01 jasta: do you have the content local, or are you issuing a network request for it? Nov 20 22:36:57 morrildl: I can send you my application, if you'd like to see for yourself. Nov 20 22:36:57 i though tRSS feeds were xml Nov 20 22:37:24 morrildl: specifically, i've been testing with SmartphoneThoughts.com and Slashdot.org's RSS feeds. if you want to pull up a lynx -source http://rss.slashdot.org/SlashDot/slashdot you can see some examples of the type of content I'm loading into the WebView. Nov 20 22:37:57 d03boy: the RSS container is XML, but jasta is just displaying the contents of individual posts, which can be anything, but typically HTML Nov 20 22:38:02 jasta: that would be very helpful actually Nov 20 22:38:05 I am also "wrapping" the WebView in an ... Nov 20 22:42:04 welcome back, all :) Nov 20 22:42:12