**** BEGIN LOGGING AT Fri Jan 23 02:59:57 2009 Jan 23 03:10:07 I'm having trouble getting an application to issue a ping to a server. InetAddress.isReachable() doesn't qualify, as it only actually pings when run as root (falling back to trying to do a TCP handshake with port 7.) It seems the only way is to try forking "ping -w 3 -c 1 hostname" and checking the return code, is there a better way? Jan 23 03:15:30 jargonjustin: why do you need an ICMP echo request? Jan 23 03:16:27 I'm having a little trouble with my layout http://pastebin.com/m35253033 Jan 23 03:16:42 I'm trying to setup an application to automatically detect if the application is behind a captive portal. Interestingly, the portal lets TCP handshakes go through but not ICMP echos. Jan 23 03:16:52 I'm trying to get the cancel button to align bottom left Jan 23 03:17:05 and the delete & save buttons to align bottom right Jan 23 03:17:05 It seemed like a low overhead way of checking if the application is behind the portal or not. Jan 23 03:17:41 jargonjustin: it seems like a flimsy way to me. Jan 23 03:18:30 jasta: any suggestions for a better way? I could use a TCP or UDP socket, but it seemed even a little more hacky (since I'd have to code more closely to the behavior of the router.) Jan 23 03:18:30 still, root privileges are required to create ICMP echo requests on Linux. even the ping binary is suid root. Jan 23 03:18:53 on android, i believe it is not suid root, and so you'd need root to fork off and use it anyway Jan 23 03:18:57 jasta: yeah, and forking the process is pretty high overhead so I'd love a better way. Jan 23 03:19:10 it isn't even a workable way on Android Jan 23 03:19:34 jasta: "/system/bin/ping -c 1 -w 3 www.google.com" seems to work fine. Jan 23 03:19:50 detecting a captive portal should be very easy anyway. just hit a web page and see that you get the expected results Jan 23 03:20:10 you could eitehr host a server specifically for this purpose, or you could use something very predictable like google.com Jan 23 03:20:31 jasta: I think that's what I'm going to end up doing. Jan 23 03:20:35 the latter is ill advised as i think it violates google's tos Jan 23 03:20:46 but this approach is substantially better than using ICMP echo anyway Jan 23 04:07:17 I have an intent-filter for an activity in my app. If an intent is created that resolves to this activity, and the intent has "extras", how do I actually get those extras? Jan 23 04:08:17 <_avatar> intent.getExtras() :P Jan 23 04:08:51 _avatar, how do I get the intent object from within the activity? Jan 23 04:09:02 ie, from within onCreate()? Jan 23 04:09:02 getIntent() Jan 23 04:09:06 ahh Jan 23 04:09:11 savid: read examples and read the javadoc Jan 23 04:13:26 <_avatar> *sigh* i wish i could write desktop apps with android, i'm really not looking forward to going back to win32 to write a UI for my server. Jan 23 04:14:18 don't use win32 then Jan 23 04:16:00 <_avatar> well, i don't know what else to use. QT is huge. wxWidgets is garbage. MFC is worse. Jan 23 04:17:08 .NET? Jan 23 04:17:09 Swing? Jan 23 04:17:22 (and yes, I agree, wxWidgets suck) Jan 23 04:17:52 <_avatar> yeah, problem with those is redistribuiting huge runtimes for a really small application, seems excessive. i'd probably make an exception for android, but definitely not .NET. Jan 23 04:18:03 <_avatar> lots of people absolutely refuse to install .NET for various crackpot reasons Jan 23 04:21:15 hehe, distribution is such a stupid problem for developers to be concerned with Jan 23 04:21:19 thank god i use a sane OS ;) Jan 23 04:21:52 jasta: because really, installing programs on Linux is never an issue :) Jan 23 04:21:59 <_avatar> hahah Jan 23 04:22:37 only when the upstream vendor has poor or non-existent support. is installing software not supported on OS X any easier? Jan 23 04:23:20 and by the way, this isn't a Linux issue at all. There are plenty of pathetic Linux distros that don't handle distribution well :) Jan 23 04:23:34 quite frankly, only Debian and Ubuntu do handle it well :) Jan 23 04:23:47 well on OS X I never had the issue of a missing dependency or a versioning problem Jan 23 04:23:55 I'm not saying it's perfect or sane or whatever Jan 23 04:23:59 and quite frankly, Microsoft and Apple would be doing their users a great service to adopt those strategies. Jan 23 04:24:19 I'm just saying that in *my* experience, I found installing software on Linux distros (including Ubuntu) a pain in the ass sometimes Jan 23 04:24:38 <_avatar> apt-get is quite good, but the learning curve for average users is too high. i've had the best luck with things "just working" in OSX. Jan 23 04:25:09 there's also the case of software you want to install that's not available for the deployment system of your distro (I included Debian/Ubuntu again :) Jan 23 04:25:09 <_avatar> not sure i'd attribute that to the platform, though. Jan 23 04:25:31 im not advocating Linux you see, i'm advocating moving distribution and dependency management away from developers. Jan 23 04:25:55 some platforms have that, and the burden on developers and maintainers is reduced substantially. Jan 23 04:26:17 with ideally no impact on users. that's a win-win situation that is unpopular on most platforms for some reason. Jan 23 04:26:44 huge amounts of energy have gone into solving this problem in a localized space (every major application on Windows for instance does this in a custom way) Jan 23 04:26:55 Windows Update is one enormous example of this. Jan 23 04:27:41 any of them individually aren't even as simple or graceful as APT. it's pathetic ;) Jan 23 04:28:55 if, for instance, more operating systems had this, _avatar's initial argument about .NET would be -- as it seems to be -- absurd. Jan 23 04:29:24 well the .NET runtime is distributed and updated through the windows update mechanism Jan 23 04:29:35 if the administrators elect to reject the update... Jan 23 04:29:37 yes, poorly. Jan 23 04:29:57 .NET is not a required update, nor is there a hook for any developer to specify that they require it short of simply bundling it. Jan 23 04:30:08 <_avatar> right, ideally the "Windows Update" mechanism would be extended to act more like apt, where my installer could use it to install dependencies. if the user declined, the app wouldn't be installed. Jan 23 04:30:29 _avatar: I've definitely seen installers do that with the .NET runtime Jan 23 04:30:30 this is an utter failure of the distribution system. also what i meant by saying that any of the individual systems fail to achieve as much functionality as APT while being even larger and more complex. Jan 23 04:30:49 _avatar: and the runtime wasn't part of the installer itself Jan 23 04:30:53 <_avatar> romainguy: yeah, that's true. but usually the installers have to do most the work manually... Jan 23 04:31:03 <_avatar> more steps for the user, more potential points of failure Jan 23 04:31:10 and even more work for the developer. Jan 23 04:31:31 good distribution management is, as i said, a win-win for both users and developers. Jan 23 04:32:07 nobody said the contrary Jan 23 04:32:15 if Debian/Ubuntu had as much commercial support from upstream vendors as OS X or Windows had, it would seem even more obvious how APT is the right solution. Jan 23 04:32:19 but unlike you I don't find apt to be this miracle solution :) Jan 23 04:32:34 <_avatar> i don't think it's a miracle solution either, but i think it's a better model Jan 23 04:32:51 romainguy: the failure of APT is a result of administration and maintenance, not the technology. Jan 23 04:33:08 Ubuntu and Debian have extremely limited resources and hardly any upstream support, even from open source projects. Jan 23 04:33:21 Microsoft or Apple could not say the same. Jan 23 04:33:36 and if they impelmented such a system, they would see even greater success. Jan 23 04:34:31 <_avatar> it's such a departure from the way things work in Windows right now, though. I doubt we'll ever see anything like that. but it'd be nice :) Jan 23 04:35:04 _avatar: right, i can appreciate that. but this is why i am talking to you now as someone genuinely impressed with what Debian and Ubuntu have accomplished :) Jan 23 04:36:10 Debian more so, of course. Ubuntu deserves very little credit here, they just have more people working for them and a more proactive mindset. Jan 23 04:36:14 Debian built everything :) Jan 23 04:36:16 and still continues to do most of the work Jan 23 04:36:37 one model that I like even better (but it would need to be improved on the developer's side) is what Java WebStart and Microsoft's OneClick Jan 23 04:36:43 (if you look at most Ubuntu patches, updates, etc you will see that they all came from the Debian team hehe) Jan 23 04:36:43 -what Jan 23 04:39:46 Jan 23 04:39:57 I'm always amazed by how stupid people who leave comments on blog can be Jan 23 04:40:08 you fool :) Jan 23 04:40:46 jasta: look at this one I just got http://www.curious-creature.org/2009/01/22/in-the-crowd/#comments :)) Jan 23 04:41:35 you know what the worst part is romain? these people really exist and say stupid things to co-workers, family, and -- if they have any -- friends. Jan 23 04:41:56 that's what scares me the most indeed Jan 23 04:42:58 my onery side is tempted to post a comment. one with the depth of stupidity you could only expect from an anonymous internet poster. and you'll never know it was me :) Jan 23 04:43:07 <_avatar> romainguy: ugh. lovely comment Jan 23 04:43:19 and every time you get a comment like "you smell like poop!", you'll wonder if it could have been me. Jan 23 04:43:22 <_avatar> btw, you're a great photographer Jan 23 04:43:32 jasta: feel free to :) Jan 23 04:43:37 _avatar: thanks Jan 23 04:43:46 _avatar is right. your work is amazing. Jan 23 04:43:50 jasta: at least thats still referring to the original person Jan 23 04:43:56 jasta: just remember that more than 50% of the population is below average intelligence ... Jan 23 04:44:00 i have told many people about your wallpapers on the G1 actually Jan 23 04:44:01 some of those market comments are just people discussing cupcake Jan 23 04:44:17 <_avatar> jsharkey: first! Jan 23 04:44:22 <_avatar> *sigh* Jan 23 04:44:22 lol Jan 23 04:44:26 ahha Jan 23 04:44:37 I'm actually surprised that so many people on market even heard about cupcake ^^ Jan 23 04:44:40 jasta: thanks :) Jan 23 04:45:02 romainguy: one of my coworkers even printed one of your photos and has it in a small frame on her desk Jan 23 04:45:27 cool ^^ Jan 23 04:45:41 now if I could only convince Google to pay me to take photos all day long... Jan 23 04:46:13 jasta: btw, what's the status on Five? Jan 23 04:46:21 awesome. Jan 23 04:46:45 still not on Market :P Jan 23 04:47:30 is there a windows server for five yet? Jan 23 04:48:12 romainguy: well, it's coming. Jan 23 04:48:21 jsharkey: no, but thats only because i havent built a biunary yet Jan 23 04:48:26 i know that it does build on windows :) Jan 23 04:48:48 oh cool, does windows provide something inotify-like? Jan 23 04:48:51 romainguy: sorry, ireally am feeling onery today. Jan 23 04:49:02 jsharkey: yes, GLib generalizes it as GFileMonitor Jan 23 04:49:13 nice :) Jan 23 04:49:28 jasta: sorry for...? Jan 23 04:49:29 romainguy: more seriously, the status is that it's base requirements are implemented and functional, but some stability and performance improvements are in the pipeline before an initial release can be amde Jan 23 04:50:09 it's not far from that release though. Jan 23 04:50:16 reminds me I need to finish my own damn app Jan 23 04:50:23 i got a neat letter from tmobiles legal team today actually Jan 23 04:50:35 your contract? Jan 23 04:50:51 <_avatar> i need to finish my app too. wish there were more hours in the day Jan 23 04:50:55 i have been working with them on IP issues and such, and they "compromised" with me by providing a full release of my IP in certain areas Jan 23 04:51:18 one of which was Five, of course. so i have this letter legally defining Five and that any work i contribute to it on my own personal time is owned by me Jan 23 04:51:28 regardless of any products or services offered now or in the future by tmobile Jan 23 04:51:34 good call Jan 23 04:51:36 <_avatar> jasta: make sure you get it in writing that the contract also applies to future versions of the software Jan 23 04:51:45 nor can i be terminated for it, even if it competes with something they do Jan 23 04:51:51 <_avatar> my company got me on a technicality, i'm no longer allowed to work on v2 of my software, because it's technically a different product Jan 23 04:52:04 _avatar: just name it something else ;) Jan 23 04:52:05 _avatar: of course. thats why this is so contentious because they know its ongoing Jan 23 04:52:17 the legal team kept asking me if i would be able to finish the work before i begin employment and i told them hel lno :) Jan 23 04:52:30 <_avatar> jsharkey: yeah, thought about that, actually. Jan 23 04:52:44 _avatar: was your app opensourced? Jan 23 04:52:56 <_avatar> jsharkey: yeah, with a new BSD license. Jan 23 04:53:02 <_avatar> er, new-style BSD license Jan 23 04:53:31 I'm glad that Google is really cool on stuff you do on your own time (at least that's my experience so far) Jan 23 04:53:43 jasta: a little worrisome that their legal team is doing more than their due dilligence. Jan 23 04:53:57 but otherwise really good to hear continued development of five hasnt been impeded. Jan 23 04:54:09 romainguy: yea ive been really impressed with that too, but havent had time to do anything yet Jan 23 04:54:41 <_avatar> speaking of which i should probably tell the company i've been writing a few android apps in my free time. i'm worried they'll take 'em from me, though. Jan 23 04:54:47 romainguy: i cant say either way whether tmobile planned to exercise their right of refusal, but i didnt want to take any chances signing a document giving them the legal authority to Jan 23 04:54:57 jasta: you did well Jan 23 04:55:10 _avatar: just write/release them under a different online handle Jan 23 04:55:45 <_avatar> jsharkey: yeah, thought about that too :) Jan 23 04:56:04 although that is clearly *not* legal advice ;) Jan 23 04:56:09 <_avatar> hehe Jan 23 04:56:18 especially because this conversation is publicly logged Jan 23 04:56:18 hehe Jan 23 04:56:48 <_avatar> yeah, plus i already have a website that outlines how the app works. oh well. Jan 23 04:57:20 btw, i'd love to see an IRC chat log enter into court as evidence some day. that would be sweet having a jury look at a bunch of crap like [20:56] i'm gonna kill my wife! Jan 23 04:57:26 and the channel clamoring "do it dude!" Jan 23 04:57:27 hehe Jan 23 04:58:15 jasta: but you don't have a wife Jan 23 04:58:26 i should probably not have used my own nick in that example Jan 23 04:58:39 err, thats not quite how i meant that to sound Jan 23 04:59:54 gah, im gonna go get work done Jan 23 05:00:03 I have tons of stuff to do Jan 23 05:00:07 adn the will to do none of them Jan 23 05:00:37 especially the huge backlog of photos to sort and develop ^^ Jan 23 05:01:43 <_avatar> i wish i had a huge backlog of photos to sort, i haven't been shooting in months. Jan 23 05:01:49 im excited to start my new job on monday. but i shouldnt be, the first week is gonna be lame :) Jan 23 05:02:03 _avatar: my new camera helps reducing the amount of photos I shoot Jan 23 05:02:12 _avatar: 25 to 30 MB per RAW file @!# Jan 23 05:02:15 <_avatar> what kind of camera? Jan 23 05:02:19 5D Mk II Jan 23 05:02:23 <_avatar> awesome :) Jan 23 05:02:36 <_avatar> i'm still using my first generation D70, it's time to upgrade. Jan 23 05:02:37 raw? isn't there some sort of lossless compression for images available yet? Jan 23 05:02:56 RAW can be compressed Jan 23 05:03:05 but it's not the raw you might be thinking about Jan 23 05:03:25 probably not Jan 23 05:03:28 _avatar: http://www.curious-creature.org/2009/01/13/one-wheel-with-new-camera/ << look at the full res version on this one, the 5D is amazing Jan 23 05:03:43 jasta: the "RAW" format is different for each manufacturer/camera Jan 23 05:03:51 jasta: it's actually the raw data from the sensor Jan 23 05:03:57 it's not even pixels Jan 23 05:04:02 <_avatar> romainguy: holy crap that's awesome. Jan 23 05:04:07 <_avatar> i'm jealous Jan 23 05:04:19 _avatar: the drawback is that the camera is quite large and heavy Jan 23 05:04:59 <_avatar> how many continuous shots can you get with the the 5D Mk II without a write delay? Jan 23 05:05:48 in RAW, about 13 frames Jan 23 05:05:56 in JPEG, something like 300 Jan 23 05:05:58 <_avatar> that's pretty impressive Jan 23 05:06:11 <_avatar> i can get, like, 3 with my old D70 :P Jan 23 05:06:35 unfortunately it's only 3.9 frames per second Jan 23 05:21:56 how would I go about getting all the numbers for a given contact? I don't see a CONTENT_URI for People.Phones Jan 23 05:53:10 wow... Jan 23 05:53:24 http://news.nationalgeographic.com/news/2008/05/photogalleries/volcano-photos/ Jan 23 05:53:33 this is an old story that i somehow missed... Jan 23 05:55:49 yeah I saw that picture a while back Jan 23 05:55:50 so sick Jan 23 06:07:19 That 1 volcano could balance out human attempts to reduce carbon emissions for the next 100 years **** BEGIN LOGGING AT Fri Jan 23 08:55:23 2009 Jan 23 14:15:55 Hi, guys Jan 23 14:16:17 I have a little problem with listactivity layout Jan 23 14:16:24 I have a listview that i bind a cursor to using a simple cursor adapter, how can i insert a row to the top of the listview? Jan 23 14:17:15 very simple layout (only a listview and a buton on the bottom), but probebly I made something wrong into the layout xml file Jan 23 14:18:01 esses: paste your xml into pastebin Jan 23 14:18:30 I got an error : your content must have a listview whose id attribute is android.R.id.list Jan 23 14:18:35 oki Jan 23 14:20:16 here is the link : http://pastebin.com/d2deee3df Jan 23 14:21:37 uhmm connection problem Jan 23 14:21:41 any suggestions? Jan 23 14:22:02 i copied the example Jan 23 14:22:03 should be android:id="@+id/YOUR_ID" Jan 23 14:22:32 if it is the problem... LOL Jan 23 14:22:35 :-P Jan 23 14:26:20 don't work Jan 23 14:26:27 doesn't work Jan 23 14:26:29 .-) Jan 23 14:32:11 the solution is if someone else has the some problem : android:id="@id/android:list" Jan 23 14:32:16 thanks google Jan 23 14:32:44 only this name for the listview into a listactivity is accepted. Jan 23 14:33:06 i go out, bye Jan 23 14:33:27 how can I add a string item to a bound listview? Jan 23 14:37:58 or better yet how can i add an item to a list adapter that has already been constructed via a simple cursor adapter? Jan 23 14:54:09 can i cast a SimpleCursorAdapter as an ArrayAdapter so that I can insert an object to the top of the ListView? Jan 23 15:01:33 anyone got an easy to install web based bug tracking system? Jan 23 15:02:09 :) Jan 23 15:02:19 i cant get bugzilla installed on my stupid server grrr Jan 23 15:05:16 i think we use bugnet here Jan 23 15:05:54 i didn't install so I don't know how easy it is... can't image it's too complicated though Jan 23 15:08:58 bugnet windows only? Jan 23 15:09:17 ah, right... might be Jan 23 15:15:10 maybe ill just install it on my own pc at home :/ Jan 23 15:34:43 just installed Mantis Jan 23 15:34:45 that was easy Jan 23 15:34:53 I'd rather use bugzilla, but i cant win everything Jan 23 15:51:28 or google code offers bug tracking and source control wrapped into one ;) Jan 23 15:52:37 How can I insert a row into a SimpleCursorAdapter? Jan 23 16:03:00 you cant? can you? Jan 23 16:04:08 not sure Jan 23 16:04:40 but look at the messages app... there's a list of recent messages/contacts and at the top there's "New Message" Jan 23 16:05:16 I'm totally guessing, but maybe using a MergeCursor? Jan 23 16:05:25 http://code.google.com/android/reference/android/database/MergeCursor.html Jan 23 16:06:00 herriojr: didn't you write something similar to that object? Jan 23 16:06:32 hmm... looks about right, but can you construct a cursor row??? Jan 23 16:07:08 yup, using a MatrixCursor Jan 23 16:07:26 ah ok Jan 23 16:07:43 thats the only way I can think of to do it, there might be something more elegant Jan 23 16:08:01 Well... that will work for now. Thanks Jan 23 16:08:13 np, good luck, I'm not even sure if it will work :) Jan 23 16:22:56 marklar: i did what you suggested and it worked partially. The original cursor binds weird though Jan 23 16:23:09 the id is correct, but the "Name" field is 1 for every row Jan 23 16:23:10 look at the source for the messages app Jan 23 16:23:20 i haven't downloaded the source yet Jan 23 16:23:31 you can just download the messaging from the git web Jan 23 16:23:33 i don't want to install all that junk just to see the source Jan 23 16:23:38 link? Jan 23 16:24:33 try http://android.git.kernel.org/?p=platform/packages/apps/Mms.git;a=snapshot;h=refs/heads/master;sf=tgz Jan 23 16:24:39 if not poke around http://android.git.kernel.org/?p=platform/packages/apps/Mms.git;a=summary Jan 23 16:24:43 i gotta run out Jan 23 16:24:47 k thanks Jan 23 18:20:53 zhobbs: MergeCursor is different Jan 23 18:21:53 zhobbs: MergeCursor doesn't sort, it just treats them all as a single list, such as, when approaching the end of one Cursor, it starts on the next Cursor Jan 23 18:27:00 it doesn't do any type of comparison Jan 23 18:33:58 grrr ant get this dumb bug tracker integrated with cvs Jan 23 19:00:40 ?? Jan 23 19:01:17 im trying to integrate Mantis with CVS Jan 23 19:01:32 and I think i have it set up correctly but it's not working Jan 23 19:01:44 so when I check something in regarding certains issues, it adds a note to the bug report Jan 23 19:04:08 duh think i figured it out Jan 23 19:04:43 err nope :( Jan 23 19:04:47 maybe the documentation is bad Jan 23 19:05:07 time to go through cvs docs Jan 23 19:15:03 think im missing hte format string Jan 23 19:18:29 ugh maybe not Jan 23 19:18:32 i dont get this Jan 23 19:18:34 ughhhhhhh Jan 23 19:23:07 youre using cvs? hehe Jan 23 19:25:39 blau1, it was pretty simple to do with SVN Jan 23 20:46:21 back Jan 23 20:46:34 of course i leave just as people start answering about my gripe Jan 23 20:46:38 sigh Jan 23 20:47:16 i dont remember why i started using cvs, maybe cause eclipse has a built in plugin for it Jan 23 20:47:16 shrug Jan 23 21:03:35 stupid cvs doesnt appear to be givin ginfo when it commits Jan 23 21:06:01 don't use cvs :) Jan 23 21:06:19 there is a good SVN plugin for eclipse blau-MikeDG Jan 23 21:07:46 yeah but i already have everyhting set up for cvs Jan 23 21:21:31 sucks to be me :( Jan 23 21:28:32 is it possible for a content provider to make it so anything that queries/inserts/deletes to the content provider can only access those rows related to it? Jan 23 21:28:52 so basically allowing row permissions Jan 23 21:29:11 Yes. The download manager does that. Jan 23 21:29:21 ok, I'll look into that then Jan 23 21:29:23 thanks Jan 23 21:29:30 ugh if have to compile one more time.... (up to android compile #32) Jan 23 21:29:33 Binder.getCallingUid() Jan 23 21:31:46 i have like 40 friggen checkings for this file Jan 23 21:37:28 jbq: ohh, thats nice Jan 23 22:00:53 argh, I'm in a losing battle with marketing Jan 23 22:01:43 is there such a thing as winning with them? Jan 23 22:02:53 they want more ponies? Jan 23 22:03:44 i'm making an AlertDialog with a single choice list but when OK is clicked how do i tell which list item is currently selected? Jan 23 22:03:49 http://android.pastebin.com/m229db02d Jan 23 22:05:11 jasta: probably not Jan 23 23:02:47 morgan, see "case DIALOG_LIST" at http://code.google.com/android/samples/ApiDemos/src/com/example/android/apis/app/AlertDialogSamples.html Jan 23 23:05:03 I'm trying to use the "sysinfo" call in my shared lib Jan 23 23:05:12 unfortunately, on the link stage I get undefined reference error Jan 23 23:05:26 is there a link flag I'm forgetting? Jan 23 23:09:51 joakime: thanks. i was using that sample but I was using the DIALOG_SINGLE_CHOICE case Jan 23 23:13:20 agh >:( Jan 23 23:13:34 vol: native call to sysinfo? Jan 23 23:15:57 yes Jan 23 23:16:05 no Jan 23 23:16:07 yes Jan 23 23:16:21 the #include sysinfo Jan 23 23:18:55 ctate: any idea/ Jan 23 23:18:57 ? Jan 23 23:19:11 so it compiles but doesn't link? Jan 23 23:19:15 I'm really just trying to get total memory and avail memory (not to the runtime, to the system) Jan 23 23:19:18 yes Jan 23 23:19:40 I could screen scrape proc/meminfo like java.os.Process does Jan 23 23:19:47 but this frankly feels like it just cuts out the middleman Jan 23 23:20:00 y'know, i don't see a sysinfo implementation in bionic. Jan 23 23:20:03 that's ... curious. Jan 23 23:20:12 i suppose it's supposed to just be a syscall, but still. Jan 23 23:20:29 well that's interesting because the call exists in the header file Jan 23 23:20:56 bionic/libc/include/sys/sysinfo.h Jan 23 23:21:18 right Jan 23 23:21:33 but there's no syscall shim for it in bionic/libc/arch-arm/syscalls Jan 23 23:21:40 oh sorry, misread what you said Jan 23 23:21:53 ugh Jan 23 23:21:58 I guess screenscraping it is Jan 23 23:22:58 yeah. if android.os.Process does that..... Jan 23 23:26:47 I guess I'm stuck. Oh well. Jan 23 23:29:23 meh Jan 23 23:29:34 (what info do you need from sysinfo, if i may ask?) Jan 23 23:30:41 total free memory, total available memory Jan 23 23:31:06 not vm memory, system memory Jan 23 23:33:29 tricky. Jan 23 23:34:13 if you felt REALLY daring you could write your own sysinfo.S shim and see whether the syscall is really implemented :) Jan 23 23:34:21 but no guarantees yadda yadda Jan 23 23:35:01 haha Jan 23 23:35:09 I don't think I'll venture into writing asm, thankyouverymuch Jan 23 23:35:20 scaredycat! ;) Jan 23 23:35:34 anyway, the semi-official answer at this point is read /proc/meminfo Jan 23 23:35:45 http://www.amazon.com/Professional-Assembly-Language-Programmer/dp/B000Q7ZETY/ref=sr_1_12?ie=UTF8&s=books&qid=1232753733&sr=8-12 Jan 23 23:35:50 I picked that book up a few years ago Jan 23 23:36:00 I read up to the point where he starts talking about using 3dnow stuff Jan 23 23:36:06 and I realized "my god, I will never use this for anything" Jan 23 23:36:11 and just put the book down forever Jan 23 23:36:35 ha ha ha Jan 23 23:38:57 anyway, I'm off for the weekend. Ho hum, thanks for your help. Jan 23 23:39:07 luck, and have a good weekend Jan 23 23:39:10 thanks, you too **** ENDING LOGGING AT Sat Jan 24 02:59:57 2009