[android-developers] Digest for android-developers@googlegroups.com - 21 Messages in 8 Topics

Comments: (0)

Group: http://groups.google.com/group/android-developers/topics

    Marty Rehder <martygeek@gmail.com> Mar 31 07:58PM -0700  

    Just curious as to how many people actually develop Android codes on an Android device.
    Is is feasible to try and do any serious development on a tablet or phone ? interested in hearing others thoughts

     

    Kristopher Micinski <krismicinski@gmail.com> Mar 31 11:35PM -0400  

    Versus the emulator?
     
    I have used devices *exclusively* for development for the past few
    years (though I don't often write actual apps, per se..): the emulator
    has always sucked and hasn't made any fundamental improvements wrt
    speed. This is pretty inherent in the fact that it's based on qemu,
    I've used the x86 image to some success, but not enough to outweigh my
    workflow.
     
    Kris
     
     

     

    miketra <miketra24@gmail.com> Mar 31 07:15PM -0700  

    My app has been successfully downloaded by tablet users over the last year
    without problems. I made an update a week ago, but the only thing changed
    in the Manifest file was the app version number and version string. After
    making that update, users are no longer able to access the app in Google
    Play. I have an Asus Transformer tablet. I am able to load the app onto my
    tablet with no problem. On the tablet, the app won't even appear in Google
    Play unless I search for it on the internet, then link directly to the app
    in Google Play. When I do this, Google Play says "your device isn't
    compatible with this version" even though I loaded the app directly onto
    the tablet myself. Note that in the Android Developer Console, it says that
    2600+ devices are compatible, and it says 0 devices are excluded. I looked
    at the individual list of devices, and my tablet is listed there as being
    compatible. What's the problem, and how do I fix it? To see the Manifest
    file, go
    here: http://stackoverflow.com/questions/15733070/why-is-app-not-compatible-with-tablets-after-update
     
    Thanks,
     
    -Michael

     

    RichardC <richard.critten@googlemail.com> Mar 31 07:43PM -0700  

    There has been an odd post or two in the adt-dev group regarding the
    ordering of elements in the manifest (although I can't find anything in the
    official docs).
     
    So with that said ... I think you should try putting
    <uses-sdk android:minSdkVersion="4" android:targetSdkVersion="8" />
    before the <application> node.
     
    I see that someone has already mentioned the <supports-screens> element on
    SO.
     
    On Monday, April 1, 2013 3:15:05 AM UTC+1, miketra wrote:

     

    plnelson <pnartg@gmail.com> Mar 31 03:56PM -0700  

    On Friday, March 29, 2013 11:16:26 PM UTC-4, Kristopher Micinski wrote:
     
    > No.
     
    > But you should see something in the logcat that shows an intent was
    > fired. Do you not see that?
     
     
    When the Activity (called DGraphActivity (xxxx-out proprietary namespace)
    is set to a "standard" launchMode and I do a search of all the logCat
    output I see two references to it
     
     
    *03-29 19:21:04.357: I/ActivityManager(119): Force finishing activity
    HistoryRecord{40698f18 com.xxxx.remote/.*
    *DGraphActivity}
     
    03-29 19:23:56.454: I/ActivityManager(119): Starting: Intent {
    flg=0x4000000 cmp=com.xxxx.remote/.**DGraphActivity (has extras) } from pid
    13727*
     
    What was interesting about this is that my program has lots of Activities
    and in all the other ones the "Starting: Intent" was followed immediately
    by a "trying to launch..." the Activity in question. But DGraphActivity
    had no "trying to launch", it just stops at "starting: intent" - that must
    be some kind of clue but I don't know what.
     
    *But*, when DGraphActivity's launchmode is set to singleInstance the logcat
    shows this . . .
     
    *03-31 18:49:12.224: I/ActivityManager(119): Starting: Intent {
    flg=0x4000000 cmp=com.assemblyguide.remote/.DGraphActivity (has extras) }
    from pid 19024
     
    03-31 18:49:12.298: W/ActivityManager(119): Trying to launch
    com.assemblyguide.remote/.DGraphActivity
     
    03-31 18:49:12.318: D/DGraphActivity(19024): created*
     
     
    For my purposes I need it to be standard but I don't know how to interpret
    this to glean why Android can start the Activity in singleInstance, but not
    in standard mode
     

     

    plnelson <pnartg@gmail.com> Mar 31 03:57PM -0700  

    On Friday, March 29, 2013 11:16:26 PM UTC-4, Kristopher Micinski wrote:
     
    > No.
     
    > But you should see something in the logcat that shows an intent was
    > fired. Do you not see that?
     
     
    When the Activity (called DGraphActivity (xxxx-out proprietary namespace)
    is set to a "standard" launchMode and I do a search of all the logCat
    output I see two references to it
     
     
    *03-29 19:21:04.357: I/ActivityManager(119): Force finishing activity
    HistoryRecord{40698f18 com.xxxx.remote/.*
    *DGraphActivity}
     
    03-29 19:23:56.454: I/ActivityManager(119): Starting: Intent {
    flg=0x4000000 cmp=com.xxxx.remote/.**DGraphActivity (has extras) } from pid
    13727*
     
    What was interesting about this is that my program has lots of Activities
    and in all the other ones the "Starting: Intent" was followed immediately
    by a "trying to launch..." the Activity in question. But DGraphActivity
    had no "trying to launch", it just stops at "starting: intent" - that must
    be some kind of clue but I don't know what.
     
    *But*, when DGraphActivity's launchmode is set to singleInstance the logcat
    shows this . . .
     
    *03-31 18:49:12.224: I/ActivityManager(119): Starting: Intent {
    flg=0x4000000 cmp=com.xxxx.remote/.DGraphActivity (has extras) } from pid
    19024
     
    03-31 18:49:12.298: W/ActivityManager(119): Trying to launch
    com.xxxx.remote/.DGraphActivity
     
    03-31 18:49:12.318: D/DGraphActivity(19024): created*
     
     
    For my purposes I need it to be standard but I don't know how to interpret
    this to glean why Android can start the Activity in singleInstance, but not
    in standard mode
     

     

    plnelson <pnartg@gmail.com> Mar 31 04:44PM -0700  

    On Friday, March 29, 2013 11:16:52 PM UTC-4, Kristopher Micinski wrote:
     
    > You might try asking the package manager if the intent you're sending
    > will have any receivers, also..
     
    I'm not sure how do that. Using the intent I use to launch my activity, I
    tried . . .
     
    PackageManager p = ctx.getPackageManager();
    final List<ResolveInfo> info=p.queryIntentActivities(intent,
    PackageManager.GET_ACTIVITIES);
     
    ...but I had no idea how to interpret the output. Here it is, first the
    successful (singleInstance) case, and second the failing (standard)
    case. I don't see a difference but I may just be too blurry-eyed to see
    it (it's 8PM here).
     
     
    info ArrayList (id=830008681048)
    array Object[1] (id=830008671280)
    [0] ResolveInfo (id=830008643864)
    activityInfo ActivityInfo (id=830008586728)
    applicationInfo ApplicationInfo (id=830008677824)
    configChanges 0 [0x0] [^@ (NUL)]
    descriptionRes 0 [0x0] [^@ (NUL)]
    enabled true
    exported false
    flags 0 [0x0] [^@ (NUL)]
    icon 0 [0x0] [^@ (NUL)]
    labelRes 0 [0x0] [^@ (NUL)]
    launchMode 3 [0x3] [^C]
    logo 0 [0x0] [^@ (NUL)]
    metaData null
    name "com.xxxx.remote.DGraphActivity"
    (id=830008007480)
    nonLocalizedLabel null
    packageName "com.xxxx.remote" (id=830007988376)
    permission null
    processName "com.xxxx.remote" (id=830007818216)
    screenOrientation 1 [0x1] [^A]
    softInputMode 0 [0x0] [^@ (NUL)]
    targetActivity null
    taskAffinity "com.xxxx.remote" (id=830007817688)
    theme 0 [0x0] [^@ (NUL)]
    filter null
    icon 0 [0x0] [^@ (NUL)]
    isDefault false
    labelRes 0 [0x0] [^@ (NUL)]
    match 0 [0x0] [^@ (NUL)]
    nonLocalizedLabel null
    preferredOrder 0 [0x0] [^@ (NUL)]
    priority 0 [0x0] [^@ (NUL)]
    resolvePackageName null
    serviceInfo null
    specificIndex -1 [0xffffffff]
    modCount 1 [0x1] [^A]
    size 1 [0x1] [^A]
     
     
     
     
    info ArrayList (id=830008607104)
    array Object[1] (id=830008590088)
    [0] ResolveInfo (id=830008265248)
    activityInfo ActivityInfo (id=830008525064)
    applicationInfo ApplicationInfo (id=830008598288)
    configChanges 0 [0x0] [^@ (NUL)]
    descriptionRes 0 [0x0] [^@ (NUL)]
    enabled true
    exported false
    flags 0 [0x0] [^@ (NUL)]
    icon 0 [0x0] [^@ (NUL)]
    labelRes 0 [0x0] [^@ (NUL)]
    launchMode 0 [0x0] [^@ (NUL)]
    logo 0 [0x0] [^@ (NUL)]
    metaData null
    name "com.xxxx.remote.DGraphActivity"
    (id=830008010320)
    nonLocalizedLabel null
    packageName "com.xxxx.remote" (id=830008010248)
    permission null
    processName "com.xxxx.remote" (id=830007812448)
    screenOrientation 1 [0x1] [^A]
    softInputMode 0 [0x0] [^@ (NUL)]
    targetActivity null
    taskAffinity "com.xxxx.remote" (id=830007796672)
    theme 0 [0x0] [^@ (NUL)]
    filter null
    icon 0 [0x0] [^@ (NUL)]
    isDefault false
    labelRes 0 [0x0] [^@ (NUL)]
    match 0 [0x0] [^@ (NUL)]
    nonLocalizedLabel null
    preferredOrder 0 [0x0] [^@ (NUL)]
    priority 0 [0x0] [^@ (NUL)]
    resolvePackageName null
    serviceInfo null
    specificIndex -1 [0xffffffff]
    modCount 1 [0x1] [^A]
    size 1 [0x1] [^A]
     
    Thanks for your help on this!!

     

    plnelson <pnartg@gmail.com> Mar 31 03:38PM -0700  

    On Friday, March 29, 2013 7:06:59 PM UTC-4, Streets Of Boston wrote:
     
    > Log.*e*("Commands", "failed to start DGraphActivity"*, e*);
     
    >> OK, I've added that although, since it's not hitting that (or any)
    exceptions it's not really telling me anything.
     
    So we're still kinda back on the original question - if an activity fails
    to start is there anyway of finding out why?

     

    Simon Giddings <mr.s.giddings@gmail.com> Mar 31 06:59AM -0700  

    I am writing a music notation app for android tablets.
    As I draw everything (ie. this is not a "simple" image (jpg, pdf, etc)
    which needs displaying), I need to ensure that the choice of drawing method
    is the fastest possible.
     
    My options, as I see them at present, are
     
    1. Classic on-screen drawing.
    2. Draw on a memory based canvas and bitblt the resulting bitmap onto
    the screen
     
    Will either option give better rendering time results ?
     
    I have chosen to use classic drawing methods, rather than an OpenGL type
    surface.
    Is this a mistake ?
     
    Any help will be appreciated.
    Thank you

     

    Jason <jason.polites@gmail.com> Mar 31 12:12PM -0700  

    My first reaction is that if you really want render speed then you need to
    be leveraging the GPU which is going to be most reliable with OpenGL (as I
    understanding it more recent versions of Android will actually hardware
    accelerate for you, but there are many variables that affect whether this
    works, not least of which is the Android version running on the user's
    device).
     
    However, the requirement for "fast rendering" may be misleading. Render
    speeds will vary wildly from device to device regardless of what mechanism
    you use, and you don't have any control over this. Game development faces
    this issue perhaps more than most but the key thing is to effectively
    "synchronize and interpolate".
     
    This means two things:
     
    1. Changes in pixel color on screen (any change in the screen image is
    effectively a pixel color change) need to be synchronized with changes in
    the objects they represent. That is, if there are several objects in the
    scene that are changing position it is important that all the changes for a
    single frame are done without it affecting what's rendered to screen until
    ALL changes are done. Failure to do this will often result in "tearing" of
    visuals or jittery movement.
     
    2. If there is a separation between changes in "world coordinate" objects
    and "screen coordinate" objects (i.e. the actual objects and their
    on-screen representations) there can sometimes be a case where changes in
    world coordinates are out of sync with screen refreshes. This is common in
    games where the "game thread" is using a fixed time step to calculate
    changes in world coordinates but the screen refresh rate is variable. In
    these cases it's important to keep both coordinate spaces in sync and one
    way to do this is to interpolate the world coordinates where there is an
    uneven number of "steps" required to synchronize the amount of time that
    has passed between screen render operations. For example if the fixed time
    step is 16ms, but the last render took 24ms you need 1.5 steps to
    accomodate this. In a fixed time step environment there is no such thing
    as a "half a step" so you need to perform 2 full steps then "estimate" the
    position at 1.5 steps by interpolating back from 2 to 1.5.
     
    Wow.. that was probably way more complicated than it needed to be :/
     
    On Sunday, March 31, 2013 6:59:44 AM UTC-7, Simon Giddings wrote:

     

    Romain Guy <romainguy@android.com> Mar 31 03:19PM -0700  


    > My first reaction is that if you really want render speed then you need
    to be leveraging the GPU which is going to be most reliable with OpenGL (as
    I understanding it more recent versions of Android will actually hardware
    accelerate for you, but there are many variables that affect whether this
    works, not least of which is the Android version running on the user's
    device).
     
    There are only two variables: are you running Android 3.0+ and are you
    using supported operations.
     
    In this particular case I doubt using OpenGL directly is necessary. Android
    tablets run on Android 3.0+ so you can simply rely on built-in hardware
    acceleration. Your first step should be to test your app with hardware
    acceleration on and see whether you are running into

     

    Jason <jason.polites@gmail.com> Mar 31 12:07AM -0700  

    Hey folks,
     
    So I'm in the midst of creating a game and I'm hitting some occasional slow
    down which "seems" to be a result of GC calls triggered by other processes
    on the device. I have profiled *my* app and am 100% certain there are no
    allocations going on while the game is running, but in certain
    circumstances (i.e. most of the time) logcat is spewing out GC_CONCURRENT
    freed...blah blah entries. Here's a typical output from the console:
     
    03-30 23:58:58.868: DEBUG/dalvikvm(2590): GC_CONCURRENT freed 407K, 6% free
    8758K/9224K, paused 5ms+3ms, total 29ms
    03-30 23:59:05.970: DEBUG/dalvikvm(2590): GC_CONCURRENT freed 408K, 6% free
    8757K/9228K, paused 4ms+2ms, total 30ms
    03-30 23:59:13.063: DEBUG/dalvikvm(2590): GC_CONCURRENT freed 407K, 6% free
    8757K/9224K, paused 4ms+4ms, total 33ms
    03-30 23:59:20.142: DEBUG/dalvikvm(2590): GC_CONCURRENT freed 407K, 6% free
    8757K/9228K, paused 5ms+3ms, total 30ms
    03-30 23:59:27.227: DEBUG/dalvikvm(2590): GC_CONCURRENT freed 407K, 6% free
    8757K/9224K, paused 6ms+2ms, total 31ms
    03-30 23:59:34.306: DEBUG/dalvikvm(2590): GC_CONCURRENT freed 407K, 6% free
    8757K/9232K, paused 4ms+2ms, total 23ms
    03-30 23:59:41.392: DEBUG/dalvikvm(2590): GC_CONCURRENT freed 407K, 6% free
    8757K/9224K, paused 4ms+2ms, total 27ms
    03-30 23:59:48.470: DEBUG/dalvikvm(2590): GC_CONCURRENT freed 407K, 6% free
    8757K/9232K, paused 5ms+2ms, total 30ms
     
    Now I have no idea which processes are to blame, but I did notice that when
    "other" events on the device occur it can lead to a spike in (I assume) CPU
    usage which causes a visible drop in frame rate in the game.
     
    For example my wifi connection dropped out one time and the logs went crazy
    with other processes complaining about a lost connection and trying
    desperately to tell me about it. This caused a fairly severe jolt in the
    game dropping the frame rate from 60fps to around 20fps for about a minute
    while everything settled down and got on with things.
     
    I had noted on a previous thread<https://groups.google.com/forum/?fromgroups=#!searchin/android-developers/Garbage$20collector$20suspend/android-developers/85yj9CZEqH4/YIiElemqiuoJ>a statement from Dianne Hackborn:
     
    *"as of 1.6 such GCs in a background process can not together take more
    than ~10% of the CPU from any foreground process that wants to run."*
     
    This sounds good, but I'm wondering if this is still the case as of 4.2.2?
    I am seeing behavior that appears to indicate this may not the case.
     
    Is there anything I can do to ensure other processes are not using more
    than 10% of CPU (or any number really)? Or some other
    restrictions/priorities that might help alleviate this external resource
    usage?
     
    Thanks,
     
    Jason.

     

    RichardC <richard.critten@googlemail.com> Mar 31 02:46AM -0700  

    Is/was 2590 your process?
     
    On Sunday, March 31, 2013 8:07:50 AM UTC+1, Jason wrote:

     

    Jason Polites <jason.polites@gmail.com> Mar 31 07:38AM -0700  

    I'm 90% sure it wasn't, but I make this claim based on profiling heap
    allocations and I don't think I confirmed that the process ID was actually
    different. I'll confirm (or otherwise) this and post back.

     

    Anthony Prieur <anthony.prieur@gmail.com> Mar 31 06:42PM +0200  

    The frame drop looks huge as since Gingerbread the GC is clamied to work
    concurrently.
     
     

     

    Jason Polites <jason.polites@gmail.com> Mar 31 11:56AM -0700  

    OK.. so I've confirmed it's not my process. Here's an example (apologies
    for the length of the log)
     
    03-31 11:46:12.587: DEBUG/dalvikvm(16714): GC_CONCURRENT freed 325K, 5%
    free 8827K/9208K, paused 10ms+4ms, total 34ms
    03-31 11:46:12.594: DEBUG/656565(16714): Table:account Count:7 Migration
    Time:48
    03-31 11:46:12.610: DEBUG/656565(16714): Table:advice Count:3 Migration
    Time:15
    03-31 11:46:12.758: DEBUG/656565(16714): Table:category Count:128 Migration
    Time:153
    03-31 11:46:12.868: DEBUG/dalvikvm(16714): GC_CONCURRENT freed 403K, 5%
    free 8840K/9272K, paused 14ms+6ms, total 43ms
    03-31 11:46:12.962: DEBUG/656565(16714): Table:transaction_bankcc Count:116
    Migration Time:202
    03-31 11:46:12.969: DEBUG/656565(16714): Table:budget_overall Count:1
    Migration Time:6
    03-31 11:46:12.985: DEBUG/656565(16714): Table:budget Count:14 Migration
    Time:13
    03-31 11:46:12.993: DEBUG/656565(16714): Table:fi_login Count:3 Migration
    Time:9
    03-31 11:46:13.001: DEBUG/656565(16714): Table:spending Count:1 Migration
    Time:13
    03-31 11:46:13.008: DEBUG/656565(16714): Table:alert Count:3 Migration
    Time:5
    03-31 11:46:13.040: DEBUG/dalvikvm(16714): GC_CONCURRENT freed 395K, 5%
    free 8842K/9272K, paused 4ms+19ms, total 51ms
    03-31 11:46:13.040: DEBUG/1010(16714): Throwable
    t:android.database.CursorIndexOutOfBoundsException: Index 0 requested, with
    a size of 0
    03-31 11:46:13.040: INFO/MintDBAdaptor(16714): Upgrading database from
    version 6 to 12
    03-31 11:46:13.040: INFO/MintDBAdaptor(16714): running migration7.sql
    03-31 11:46:13.071: INFO/MintDBAdaptor(16714): running migration8.sql
    03-31 11:46:13.071: INFO/MintDBAdaptor(16714): running migration9.sql
    03-31 11:46:13.094: INFO/MintDBAdaptor(16714): running migration10.sql
    03-31 11:46:13.094: INFO/MintDBAdaptor(16714): running migration11.sql
    03-31 11:46:13.102: INFO/MintDBAdaptor(16714): running migration12.sql
    03-31 11:46:13.587: DEBUG/dalvikvm(16758): GC_CONCURRENT freed 406K, 6%
    free 8774K/9260K, paused 2ms+2ms, total 69ms
    03-31 11:46:13.610: DEBUG/dalvikvm(16714): GC_CONCURRENT freed 360K, 5%
    free 8929K/9316K, paused 4ms+9ms, total 98ms
    03-31 11:46:14.040: DEBUG/dalvikvm(16714): GC_CONCURRENT freed 294K, 4%
    free 9067K/9388K, paused 5ms+6ms, total 42ms
    03-31 11:46:14.508: DEBUG/dalvikvm(16714): GC_FOR_ALLOC freed 137K, 5% free
    9136K/9568K, paused 23ms, total 23ms
    03-31 11:46:14.532: DEBUG/dalvikvm(16714): GC_FOR_ALLOC freed 148K, 6% free
    9137K/9632K, paused 22ms, total 22ms
    03-31 11:46:14.555: DEBUG/dalvikvm(16714): GC_FOR_ALLOC freed 61K, 5% free
    9229K/9632K, paused 24ms, total 24ms
    03-31 11:46:14.688: DEBUG/dalvikvm(16714): GC_CONCURRENT freed 377K, 5%
    free 9348K/9756K, paused 3ms+4ms, total 38ms
    03-31 11:46:14.969: INFO/Database(16714): sqlite returned: error code = 17,
    msg = statement aborts at 26: [INSERT INTO budget_overall(id,
    performanceStatus) VALUES(?, ?);] database schema has changed
    03-31 11:46:15.790: DEBUG/dalvikvm(16714): GC_CONCURRENT freed 635K, 7%
    free 9181K/9844K, paused 3ms+2ms, total 33ms
    03-31 11:46:15.962: DEBUG/123(16714): Modified From::Mar:31 07:15 2013
    :1364739342001
    03-31 11:46:16.102: DEBUG/dalvikvm(16714): GC_CONCURRENT freed 433K, 7%
    free 9161K/9844K, paused 2ms+2ms, total 26ms
    03-31 11:46:16.829: DEBUG/dalvikvm(16714): GC_FOR_ALLOC freed 159K, 7% free
    9250K/9844K, paused 22ms, total 22ms
    03-31 11:46:16.852: DEBUG/dalvikvm(16714): GC_FOR_ALLOC freed 40K, 6% free
    9269K/9844K, paused 19ms, total 19ms
    03-31 11:46:17.227: DEBUG/dalvikvm(16714): GC_FOR_ALLOC freed 59K, 6% free
    9299K/9844K, paused 26ms, total 26ms
    03-31 11:46:17.227: INFO/dalvikvm-heap(16714): Grow heap (frag case) to
    9.236MB for 136844-byte allocation
    03-31 11:46:17.258: DEBUG/dalvikvm(16714): GC_FOR_ALLOC freed 89K, 7% free
    9344K/9980K, paused 29ms, total 29ms
    03-31 11:46:17.407: DEBUG/dalvikvm(16714): GC_FOR_ALLOC freed <1K, 7% free
    9344K/9980K, paused 23ms, total 23ms
    03-31 11:46:17.415: INFO/dalvikvm-heap(16714): Grow heap (frag case) to
    9.258MB for 115094-byte allocation
    03-31 11:46:17.430: DEBUG/dalvikvm(16714): GC_FOR_ALLOC freed 0K, 7% free
    9456K/10096K, paused 22ms, total 22ms
    03-31 11:46:17.454: DEBUG/dalvikvm(16714): GC_FOR_ALLOC freed 133K, 8% free
    9322K/10096K, paused 21ms, total 21ms
    03-31 11:46:17.469: DEBUG/dalvikvm(16714): GC_FOR_ALLOC freed <1K, 7% free
    9435K/10096K, paused 14ms, total 15ms
    03-31 11:46:17.469: INFO/dalvikvm-heap(16714): Grow heap (frag case) to
    9.403MB for 172636-byte allocation
    03-31 11:46:17.485: DEBUG/dalvikvm(16714): GC_FOR_ALLOC freed 112K, 8% free
    9491K/10268K, paused 14ms, total 14ms
    03-31 11:46:17.501: DEBUG/dalvikvm(16714): GC_FOR_ALLOC freed 168K, 9% free
    9435K/10268K, paused 14ms, total 14ms
    03-31 11:46:17.555: DEBUG/dalvikvm(16714): GC_CONCURRENT freed 383K, 7%
    free 9564K/10268K, paused 3ms+2ms, total 34ms
    03-31 11:46:17.555: DEBUG/dalvikvm(16714): WAIT_FOR_CONCURRENT_GC blocked
    10ms
    03-31 11:46:17.688: DEBUG/dalvikvm(16714): GC_CONCURRENT freed 237K, 6%
    free 9734K/10268K, paused 3ms+2ms, total 26ms
    03-31 11:46:18.024: INFO/dalvikvm(16714): Total arena pages for JIT: 11
    03-31 11:46:18.024: INFO/dalvikvm(16714): Total arena pages for JIT: 12
    03-31 11:46:18.024: INFO/dalvikvm(16714): Total arena pages for JIT: 13
    03-31 11:46:18.024: INFO/dalvikvm(16714): Total arena pages for JIT: 14
    03-31 11:46:18.024: INFO/dalvikvm(16714): Total arena pages for JIT: 15
    03-31 11:46:18.024: INFO/dalvikvm(16714): Total arena pages for JIT: 16
    03-31 11:46:18.024: INFO/dalvikvm(16714): Total arena pages for JIT: 17
    03-31 11:46:18.071: DEBUG/dalvikvm(16714): GC_CONCURRENT freed 940K, 11%
    free 9232K/10268K, paused 2ms+3ms, total 33ms
    03-31 11:46:18.493: INFO/dalvikvm(16714): Total arena pages for JIT: 18
    03-31 11:46:20.657: DEBUG/dalvikvm(16758): GC_CONCURRENT freed 432K, 6%
    free 8749K/9268K, paused 4ms+7ms, total 33ms
    03-31 11:46:27.704: DEBUG/dalvikvm(16758): GC_CONCURRENT freed 383K, 6%
    free 8774K/9236K, paused 2ms+1ms, total 23ms
    03-31 11:46:29.008: DEBUG/dalvikvm(16714): GC_CONCURRENT freed 354K, 8%
    free 9286K/10088K, paused 3ms+3ms, total 31ms
     
    When this was recorded, the process ID for my process was 16810 (
    http://cl.ly/image/1X0I3e2W423k). Most (although interestingly not all) of
    the above logs are for process ID 16714 which we can see is actually the
    Mint app. Not only is it hammering the GC but it also seems to be doing
    database migrations :/
     
    This is all fine IF it runs in a resource limited space (i.e. limited CPU,
    RAM & Disk) but it appears to not be the case as either one or all of these
    resources may be blocking causing lag in the game.
     
    Not sure what the 16758 process is.
     
    This is just a small sample (covering about 10 seconds) of an otherwise
    much larger log with several GC calls.
     
    Is anyone able to shed some light on whether this is expected/normal
    behavior, or if there's anything that can be done to mitigate?
     
    Thanks.
     
    P.S. I don't see the same behavior on older 2.3 devices, but this is not a
    very controlled experiment as the older devices tend not to have many apps
    whereas the newer (4.2.2) device (which is a Galaxy Nexus btw) is my every
    day device so has more apps.
     
     
     
    --
    Ozzy's Odyssey! A new game for Android
    https://market.android.com/details?id=com.carboncrystal.odyssey
    http://www.carboncrystal.com/ <http://www.carboncrystal.com/droid-odyssey/>

     

    Sunil Deshpande <deshpande.sunil@gmail.com> Mar 31 03:42AM -0700  

    Hi,
     
    I want to use usb bluetooth dongle with the tab.
     
    From forum I can see you have already tried this. Do you have any
    suggestions?
     
    1) Do I need to hack kernel or tab image?
    2) dongle bt apps are already available on play store?
     
    Thanks and Regards,
    Sunil
     
    On Friday, 19 November 2010 21:09:07 UTC+5:30, Daddym85 wrote:

     

    Brofalad <hihansson@gmail.com> Mar 30 11:42PM -0700  

    Thanks for your answer,

    but I don't know how to use the rotation. The LatLonBox is from parsed KML
    file.

    I'm adding the OverLay like this

    layMap = new GroundOverlay(bmpImage, topGeoPoint, bottomGeoPoint);
    mapOverlays.add(0, layMap );

    But "new GroundOverlay" don't seem to support rotation






     
    Den lördagen den 30:e mars 2013 kl. 23:30:29 UTC+1 skrev gjs:
     

     

    Brofalad <hihansson@gmail.com> Mar 30 11:44PM -0700  

     

    Brofalad <hihansson@gmail.com> Mar 30 11:45PM -0700  

    Den söndagen den 31:e mars 2013 kl. 06:26:37 UTC+2 skrev lbendlin:

     

    Brofalad <hihansson@gmail.com> Mar 30 11:50PM -0700  

    Thanks for your answer,
     
    but I don't know how to use the rotation. The LatLonBox is from parsed KML
    file.
     
    I'm adding the OverLay like this
     
    layMap = new GroundOverlay(bmpImage, topGeoPoint, bottomGeoPoint);
    mapOverlays.add(0, layMap );
    But "new GroundOverlay" don't seem to support rotation
     
    Den lördagen den 30:e mars 2013 kl. 23:30:29 UTC+1 skrev gjs:
     

     

You received this message because you are subscribed to the Google Group android-developers.
You can post via email.
To unsubscribe from this group, send an empty message.
For more options, visit this group.

--
--
You received this message because you are subscribed to the Google
Groups "Android Developers" group.
To post to this group, send email to android-developers@googlegroups.com
To unsubscribe from this group, send email to
android-developers+unsubscribe@googlegroups.com
For more options, visit this group at
http://groups.google.com/group/android-developers?hl=en
---
You received this message because you are subscribed to the Google Groups "Android Developers" group.
To unsubscribe from this group and stop receiving emails from it, send an email to android-developers+unsubscribe@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out.
 
 

[android-developers] Digest for android-developers@googlegroups.com - 9 Messages in 4 Topics

Comments: (0)

Group: http://groups.google.com/group/android-developers/topics

    gjs <garyjamessilva@gmail.com> Mar 30 03:30PM -0700  

    Hi,
     
    You are already using <rotation>0.477325392961741</rotation> in your
    example, just try other rotation values & see what you get.
     
    Regards
     
    On Friday, March 29, 2013 7:04:56 PM UTC+11, Brofalad wrote:

     

    lbendlin <lutz@bendlin.us> Mar 30 09:26PM -0700  

    you could pre-rotate the image for the ground overlay via a matrix
    transformation.
     
    On Friday, March 29, 2013 4:04:56 AM UTC-4, Brofalad wrote:

     

    tom_mai78101 <tom.mai78101@gmail.com> Mar 30 02:16PM -0700  

    I had a bug that occured when Android 4.0 and Android 4.1 both tried to share a Bluetooth connection. I was denied support for those two versions and was asked for trying to test it on Android 4.2.
     
    The main problem for me would be the lack of hardware available to use. I don't have Android 4.2, and I can't upgrade my devices to 4.2.
     
    My manufacturer don't have Android 4.2 patches available for upgrade anyway, so it is hard for me to get Android 4.2 on my devices.
     
    Here, I thought Android 4.X has support for all ICS versions, like how Microsoft supports their Windows products based on their shelf life. But it seems to me, that's not the case.
     
    Do we still have product support for Android 4.0 and 4.1? Or developers are keen on getting the latest versions and drop support for older versions, starting from ICS?
     
     
     
     
     
     
     
     
     
     
    F

     

    Kristopher Micinski <krismicinski@gmail.com> Mar 30 05:50PM -0400  

    This isn't an appropriate message for android-developers, you'll be
    more likely to get a response on android-platform (though I sort of
    doubt the people there would know the answer either).
     
    Kris
     

     

    Kostya Vasilyev <kmansoft@gmail.com> Mar 31 02:02AM +0400  

    Android is a work in progress. Something's always broken, and then you have
    device manufactures screwing with things on top of it.
     
    What versions you support is a judgement call.
     
    Right now, according to Google's own metrics, Android 2.1 - 2.3 is still
    50% of actual user's devices, and 4.2 is only 1.6% (and many new devices
    come out with 4.1 at most):
     
    http://developer.android.com/about/dashboards/index.html
     
    -- K
     
     
    2013/3/31 tom_mai78101 <tom.mai78101@gmail.com>
     

     

    Boyd Speer <bspeer@shaw.ca> Mar 30 09:44PM -0600  

    What would be a reasonable time from publishing date until an app is available in all countries where GooglePlay is available? If there is any hitch in the publishing process does Google let the developer know? (Of course some apps are likely to have a narrower market and may take more time before the users become aware of the availability...) Thanks for any insights.

     

    sbVB <sbvillasboas@gmail.com> Mar 30 10:38AM -0700  

    When I install and run an application, that has a set of permissions
    granted,
    I feel afraid that something bad might happen.
     
    This is the case of an interesting game, and my personal phone.
     
    If I grant the permissions, it gets dangerous.
    If I don't grant the permissions, I don't run the app.
     
    Here is the feature wish:
    I want to use some sort of "sand box", that is completely external to
    any given application (probably at operation system tier), that
    can configure for each app a set of permission denials.
     
    If I activate the permission denials of the sandbox at a
    game for instance, the game asked for some permission
    and the game believes those permissions were granted,
    so it will be installed.
    But after the installation, the sandbox will deny selected
    permissions, so that the game app will not be able to
    actually do something in the scope of the denied permission.
     
    Google and guru's: implement this and get rich !
    I considered to implement it myself.
    I could do. But as I mentioned, this works at operating
    system tier, and it is quite complex.
    If Google invites me, I would happily join the team implementing
    this. But since I have to work, I get busy, so I won´t
    implement this in near future.
    That's why I'd rather just report this wish.
     
    Yours truly,
     
    ----------------------------------------------------
    +------+ Sergio Barbosa Villas-Boas
    /------/| sbvb@sbvb.com.br
    | sbVB |/ http://www.sbVB.com.br
    +------+ +55-21-7699-1337
    +55-21-2562-8782 (Labotim)

     

    Kristopher Micinski <krismicinski@gmail.com> Mar 30 04:40PM -0400  

    This is really for the android-security-discuss list.
     
    There are implementations of this idea by various research groups:
    enough that you can download and use a few of them. But most of them
    rely on non stock Android ROMs.
     
    FYI, it's impossible to just "deny the permission" to an app. If you
    just flat out "deny" an app the permission, it will (most likely)
    crash.
     
    Why? Because many times a permission gates an API call which returns
    some piece of information an app *has to have* to function. If you
    just "return null" you'll quickly crash the app. So let's consider
    the alternative: return some "sensible null value". To do this, you
    need to do a fairly elaborate form of control flow analysis to detect
    which parts of the app must be "patched" to work with the lack of
    information. This is quite nontrivial, to say the least.
     
    The closest to the solution you want is a slight variation: don't
    grant an app access to a permission until it actually attempts to
    access information guarded by that permission. This is doable, but
    would require similar analysis and transformation (or a non backwards
    compatible change to the API). There are also groups that implement
    variations of those.
     
    Kris
     
     

     

    deb <javibc@esdebian.org> Mar 30 11:23AM -0700  

    Hello, excuse me for the question.
    I am trying, in an app, use this link <a href="tel:897#1">Call me</a>, but
    dial not know "#" and number is blank.
    Thank you.

     

You received this message because you are subscribed to the Google Group android-developers.
You can post via email.
To unsubscribe from this group, send an empty message.
For more options, visit this group.

--
--
You received this message because you are subscribed to the Google
Groups "Android Developers" group.
To post to this group, send email to android-developers@googlegroups.com
To unsubscribe from this group, send email to
android-developers+unsubscribe@googlegroups.com
For more options, visit this group at
http://groups.google.com/group/android-developers?hl=en
---
You received this message because you are subscribed to the Google Groups "Android Developers" group.
To unsubscribe from this group and stop receiving emails from it, send an email to android-developers+unsubscribe@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out.
 
 

[android-developers] Digest for android-developers@googlegroups.com - 10 Messages in 4 Topics

Comments: (0)

Group: http://groups.google.com/group/android-developers/topics

    Kristopher Micinski <krismicinski@gmail.com> Mar 29 11:16PM -0400  

    > Which illustrates my question: which is whether there's a way to
    > systematically find out from Android what happened. Is any of this stuff
    > instrumented so we can see why something did or did not happen?
     
    No.
     
    But you should see something in the logcat that shows an intent was
    fired. Do you not see that? The system does not "drop intents
    randomly". Can you point out concrete cases where it has without
    sufficient error messages, because those would be bugs..
     
    > have worked. Because we're all fumbling in the dark since we don't know
    > what Android is doing behind the scenes. I want to know if there's a way
    > to find out.
     
    There's not logged output from the Android internals that give an
    account of why your intent is "dropped" but I am more confused as to
    why this would be happening in the first place. In the past, whenever
    I've heard this happening, it's because there's always been something
    like a misspelling of package name, or something's been signed
    incorrectly..
     
    Kris

     

    Kristopher Micinski <krismicinski@gmail.com> Mar 29 11:16PM -0400  

    You might try asking the package manager if the intent you're sending
    will have any receivers, also..
     
    Kris
     
    On Fri, Mar 29, 2013 at 11:16 PM, Kristopher Micinski

     

    Jim Graham <spooky130u@gmail.com> Mar 29 06:53PM -0500  

    On Fri, Mar 29, 2013 at 12:36:32PM -0700, tom_mai78101 wrote:
    > I have seen it, but it doesn't answer my question. Why aren't OpenGL
    > ES 2.0 video tutorials available on the net?
     
    Perhaps nobody's made one? Or you just didn't find the right search
    (I've run into this plenty of times, myself). If you can't find one,
    maybe you should pay someone to do one (seriously). If not, I'm sure
    you'll find plenty of books on the subject.
     
    Later,
    --jim
     
    --
    THE SCORE: ME: 2 CANCER: 0
    73 DE N5IAL (/4) MiSTie #49997 < Running Mac OS X Lion >
    spooky130u@gmail.com ICBM/Hurricane: 30.44406N 86.59909W
     
    In light of DHS's recommendation to defend yourself
    against armed intruders with scissors, I'm giving
    my deer rifle a new nickname: Scissors.

     

    Lew <lewbloch@gmail.com> Mar 29 06:19PM -0700  

    tom_mai78101 wrote:
     
    > I have seen it, but it doesn't answer my question. Why aren't OpenGL ES
    > 2.0 video tutorials available on the net?
     
     
    Why aren't there tutorials on YouTube regarding any subject not represented
    on YouTube?
     
    It's a silly question.
     
    I think Spooky has the best answer, which I'll paraphrase as "because no
    one has made one
    and uploaded it."
     
    Or maybe he's right and your search wasn't correct.
     
    Or maybe the goons at Censors 'R' Us have had them all deleted.
     
    Or maybe they all violated copyright and were taken down under the DMCA.
     
    Or maybe all the video tutorials are proprietary and you have to pay to see
    them.
     
    Or maybe there's a filter preventing you, personally and specifically, from
    finding them
    while the rest of us enjoy them with white wine and paté de fois gras.
     
    --
    Lew

     

    plnelson <pnartg@gmail.com> Mar 29 03:40PM -0700  

    On Friday, March 29, 2013 5:57:46 PM UTC-4, Streets Of Boston wrote:
     
    > Look at your logcat and see why your activity can't be started. Could be
    > as simple as a typo in your manifest.
     
     
    I only see things in LogCat about DGraphActivity when it *does* start
    (ie., when its launchMode is "singleInstance"). Are you saying there
    should be a message that it failed to start (or failed to be created)? I
    run Logcat with no filters and in verbose mode so I think I'm seeing
    everything there is.
     
    The manifest is exactly as it's shown - the *only* difference between the
    successful case and the failing case is the launchMode, so if there was a
    typo it would have to be unique to that line. Furthermore it only works
    for singleInstance, it fails for standard, singleTop and singleTask, so if
    there was a typo it would have to be in all three of those but not in the
    singleInstance case.
     
     
    RichardC wrote
    Technically it needs a "." here:
     
    <activity android:name=".DGraphActivity"

    I'm skeptical. My program has about 20 Activities and none of them have
    .'s in front of the name and I've written a lot of other Android programs
    without .'s. And this works fine without the "." if I launch it in
    singleInstance. From time to time I have heard people mention "."'s
    in the Manifest so I don't know if that used to be a requirement in the
    early days or what. Just for yucks I tried putting the "." in and it
    didn't help.

     

    RichardC <richard.critten@googlemail.com> Mar 29 03:45PM -0700  

    On Friday, March 29, 2013 10:40:44 PM UTC, plnelson wrote:
    > in the Manifest so I don't know if that used to be a requirement in the
    > early days or what. Just for yucks I tried putting the "." in and it
    > didn't help.
     
    I works for me as well, that's why I wrote "technically". It's just not
    documented to work, see:
    http://developer.android.com/guide/topics/manifest/activity-element.html#nm
     
    and as you were having trouble I though it might help.

     

    Streets Of Boston <flyingdutchie@gmail.com> Mar 29 04:04PM -0700  

    I've never seen this. If it fails to start, i get a logcat message and an
    exception is thrown (and you are catching that).
    Have you tried to set the Intent.FLAG_ACTIVITY_NEW_TASK flag when starting
    the activity?
     
    On Friday, March 29, 2013 6:40:44 PM UTC-4, plnelson wrote:

     

    Streets Of Boston <flyingdutchie@gmail.com> Mar 29 04:06PM -0700  

    Also, change the catching of your try - catch block, so that you see more
    of the exception being thrown when the activity fails to start:
     
    Log.*e*("Commands", "failed to start DGraphActivity"*, e*);
     
    This would allow you to see more info of what may go wrong.
     
    On Friday, March 29, 2013 7:04:39 PM UTC-4, Streets Of Boston wrote:

     

    plnelson <pnartg@gmail.com> Mar 29 04:45PM -0700  

    On Friday, March 29, 2013 7:04:39 PM UTC-4, Streets Of Boston wrote:
     
    > I've never seen this. If it fails to start, i get a logcat message and an
    > exception is thrown (and you are catching that).
     
     
    *My bad* - it turned out it did have a filter on. (thanks for saying you
    never saw that - it made me look twice!) With it off I do get 2 logcat
    messages about this activity (I've "xxxx'd out some proprietary
    content). I'm not sure what these mean . . .
     
     
    03-29 19:21:04.357: I/ActivityManager(119): Force finishing activity
    HistoryRecord{40698f18 com.xxxx.remote/.DGraphActivity}
     
     
    03-29 19:23:56.454: I/ActivityManager(119): Starting: Intent {
    flg=0x4000000 cmp=com.xxxx.remote/.DGraphActivity (has extras) } from pid
    13727
     
    What was interesting about this is, as I mentioned, my program has lots of
    Activities and in all the other ones the "Starting: Intent" was followed
    immediately by a "trying to launch..." the Activity in question. But
    DGraphActivity had no "trying to launch"! So that seems to be a clue but
    I don't know of what.
     
     
    Have you tried to set the Intent.FLAG_ACTIVITY_NEW_TASK flag when starting
    > the activity?
     
    No I haven't - thought that just places it outside the others on the
    history stack - why do I want to do that?
     
    (PS - where on the Streets of Boston are you - I'm on the streets of
    Chelmsford).

     

    Brofalad <hihansson@gmail.com> Mar 29 01:04AM -0700  

    Hi,

    I have a overlay image with the following cooridinates.

    <LatLonBox>
    <north>56.56509012808</north>
    <south>56.5455371985222</south>
    <east>13.249250664387</east>
    <west>13.2017977722382</west>
    <rotation>0.477325392961741</rotation>
    </LatLonBox>

    It is no problem to create the overlay without the rotation. But is it
    possible to rotate the image, or can I only use images with rotation = 0.
    /Brofalad

     

You received this message because you are subscribed to the Google Group android-developers.
You can post via email.
To unsubscribe from this group, send an empty message.
For more options, visit this group.

--
--
You received this message because you are subscribed to the Google
Groups "Android Developers" group.
To post to this group, send email to android-developers@googlegroups.com
To unsubscribe from this group, send email to
android-developers+unsubscribe@googlegroups.com
For more options, visit this group at
http://groups.google.com/group/android-developers?hl=en
---
You received this message because you are subscribed to the Google Groups "Android Developers" group.
To unsubscribe from this group and stop receiving emails from it, send an email to android-developers+unsubscribe@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out.
 
 

[android-developers] Digest for android-developers@googlegroups.com - 25 Messages in 10 Topics

Comments: (0)

Group: http://groups.google.com/group/android-developers/topics

    plnelson <pnartg@gmail.com> Mar 29 01:16PM -0700  

    Over many projects I've run into various problems with Activities doing
    mysterious things. I've had them not start; I've had onActivityResult not
    get called when I'm trying to send parameters back from an Activity; I've
    had onActivityResult get called *instantly* before the other Activity even
    got created. Googling the web I see that lots of people have these
    problems and often these either stump people on StackOverflow or result in
    random "superstitious" fixes with no obvious basis - even if they sometimes
    seem to work.

    So* My Question*: is there any *systematic* way to debug Activity
    life-cycle problems - are there log files, traces, dumps, or anything that
    programmers can use to figure out why something is happening. If I call
    "startActivity" and the activity is never created or started, is there
    anyplace Android tells us why? If onActivityResult is called prematurely
    (i.e., before the Activity that's supposed to return the result even
    exists) how can we find out why it was called? Are there any hints of
    tricks to interpreting the call stack that can shed light on that?
     
    Thanks in advance for any tips or tricks.

     

    Streets Of Boston <flyingdutchie@gmail.com> Mar 29 02:56PM -0700  

    Look at your logcat. It tells you when an activity hasn't started.
    For the onActivityResult getting called instantly: Your returning activity
    was probably started with a singleTask or singleInstance. This means that
    onActivityResult was called immediately after your activity was started
    (and not upon its return). Use default/standard instead.
     
    On Friday, March 29, 2013 4:16:53 PM UTC-4, plnelson wrote:

     

    plnelson <pnartg@gmail.com> Mar 29 03:23PM -0700  

    On Friday, March 29, 2013 5:56:18 PM UTC-4, Streets Of Boston wrote:
     
    > Look at your logcat. It tells you when an activity hasn't started.
     
     
    When an activity fails to start all I see in LogCat is .. *nothing* related
    to that Activity. If the Activity is created or started (or paused, etc) *
    then* I see it in Logcat. Are you saying that I should see a message
    saying that the Activity failed to start or be created? I run LogCat in
    verbose mode with no filters on.
     

     
    > was probably started with a singleTask or singleInstance. This means that
    > onActivityResult was called immediately after your activity was started
    > (and not upon its return). Use default/standard instead.
     
    In that particular example the returning Activity was standard but the
    Activity *invoking* it was singleInstance. I knew the returning Activity
    had to be standard but I only found out that the caller couldn't be
    singleInstance by time-consuming trial-and-error. I've never seen it
    documented. When I made the caller activity also standard it
    worked. Which illustrates my question: which is whether there's a
    way to *systematically* find out from Android what happened. Is any of
    this stuff instrumented so we can see why something did or did not happen?
     
    I'm currently trying to debug a case where I can ONLY start an activity if
    it's singleInstance. I don't know why and lots of people on StackOverflow
    have made many guesses and suggestions, none of which (as of this writing)
    have worked. Because we're all fumbling in the dark since we don't know
    what Android is doing behind the scenes. I want to know if there's a way
    to find out.

     

    plnelson <pnartg@gmail.com> Mar 29 01:02PM -0700  

    This question has (so far) stumped them on Stack Overflow. . . .
     
    I'm trying to launch an Activity which launches *perfectly fine* when its
    launchMode is set to *singleInstance*. When I change it to *standard*nothing happens; it never gets to onCreate (or onStart or onResume, or
    anyplace in the target Activity - I've overridden all the lifycycle events
    and set breakpoints)
     
    if (DGraphActivity.bitmap != null) {
    Intent intent = new Intent(ctx, DGraphActivity.class);
    intent.putExtra("Buttons", sButtonParam);
    try {
    ctx.startActivity(intent);
    }
    catch (Exception e) {
    Log.d("Commands", "failed to start DGraphActivity");
    }}
     
    in the manifest . . .
     
    <activity android:name="DGraphActivity"
    android:screenOrientation="portrait"
    android:launchMode="standard"></activity>
     
    I need the DGraphActivity launchMode to be *standard* because it will be
    launching another activity and using onActivityResult to collect the
    response and onActivityResult doesn't work with activites launched in
    singleInstance launchMode.
     
    I have lots of other Activities with *standard* launchMode that launch
    perfectly well, I can't figure out what's different about this one. Thanks
    in advance!

     

    RichardC <richard.critten@googlemail.com> Mar 29 02:45PM -0700  

    Technically it needs a "." here:
     
    <activity android:name=".DGraphActivity"
     
    ^
    or the full package name.
     
    On Friday, March 29, 2013 8:02:04 PM UTC, plnelson wrote:

     

    Streets Of Boston <flyingdutchie@gmail.com> Mar 29 02:57PM -0700  

    Look at your logcat and see why your activity can't be started. Could be as
    simple as a typo in your manifest.
     
    On Friday, March 29, 2013 4:02:04 PM UTC-4, plnelson wrote:

     

    xrd <xrdawson@gmail.com> Mar 29 02:53PM -0700  

    I'm trying to get a fragment working such that when the fragment is brought
    into view it refreshes the content in the view.
     
    I thought that I could implement OnPageChangeListener and
    override onPageSelected but my breakpoint is never hit. Is there a better
    way? Or, perhaps this means I am doing things wrong somewhere in my code?
     
    I've also tried using onResume(), but this does not get called each time
    the fragment is displayed, which is what I need.
     
    Any suggestions?

     

    Tobiah <toby@tobiah.org> Mar 29 01:04PM -0700  

    I'm launching another app from my app like this:
     
    Intent LaunchIntent = getPackageManager().getLaunchIntentForPackage("com.other.appname");
    startActivityForResult(LaunchIntent, EXPOSMART_MESSAGE);
     
    Then I set a result from the launched app:
     
    Intent response = new Intent();
    response.putExtra("exit_flag", "y");
    setResult(RESULT_OK, response);
    finish();
     
    I was expecting onActivityResult() to get called in the original
    app, but it never does. Is there something different about launching
    other apps rather than just new activities in the same app?
     
    Thanks!
     
    Toby

     

    Tobiah <toby@tobiah.org> Mar 29 02:06PM -0700  

    On 03/29/2013 01:52 PM, RichardC wrote:
    > Does the other App call setResult() ?
     
    Yeah, It's shown in the code below.
     
    Thanks,
     
    Tobiah
     

     

    tom_mai78101 <tom.mai78101@gmail.com> Mar 29 12:23AM -0700  

    I'm looking around on Youtube for tutorial videos on how to use OpenGL ES
    2.0 for Android, and I see that there aren't any available online. I'm not
    exactly looking at using OpenGL ES 1.X, but I'm okay with it. It's just
    that the number of OpenGL ES 2.0 tutorials are somewhat few and rare.

    I know that there are websites and blogs that teaches you OpenGL ES 2.0,
    how to use them, and how to work with shaders and things like that. Well, I
    did followed a few of their tutorials, and this is what I got from them.

    The attached ZIP file is my current OpenGL ES 2.0 project that I
    self-taught myself just by reading the online tutorials about OpenGL ES
    2.0. The results aren't exactly like what I anticipated.

    I'm already into lesson 2, displaying a 3D model mesh. It was supposed to
    show up as a regular green cube mesh, with no lightings and textures. The
    actual result I got is a flatten pyramid with its top cut off, like this
    shown below:






    <http://www.sketchup.com/crimages/bm-154069-FlatTopRoofBlock.jpg>

    Since I don't know why a cube looked like in the picture above, and not
    like a regular shaped cube, I figured I try looking up a video to see how
    OpenGL ES 2.0 for Android was coded in real-time.

    It's a letdown when what I (supposedly) typed in doesn't look what I
    expected, and that there's no clue to how it happens like that. Which leads
    me to question the problem with the lack of OpenGL ES 2.0 video tutorials.
    They help tremendously when the program was coded and run in real-time,
    from another developer's perspective.

    So, why the lack of these video tutorials? Thanks in advance.

    --------------------

    Or you know, if someone is willing to help me debug and solve the problem,
    I'll be even more happier to let that person to help me out, and teach me
    something I don't know or didn't grasp when I was learning it myself. That
    would be grateful.

     

    tom_mai78101 <tom.mai78101@gmail.com> Mar 29 12:36PM -0700  

    I have seen it, but it doesn't answer my question. Why aren't OpenGL ES 2.0 video tutorials available on the net?

     

    bergstr <sell.bergstr11@gmail.com> Mar 29 11:17AM -0700  

    ok, I have changed the code such that the frequent calls into native code
    (in method validateKey) don't happen any longer. They were unneeded anyway,
    as the internal cleanup thread can safely assume that the keys are valid.
    With this change, I can no longer reproduce this issue.
     
    However, I am still left with the impression that Android may at any point
    suspend background threads in favor of the UI thread regardless of their
    state and whether they hold locks. If this is true, it imposes very hard
    requirements on the programmer to avoid cases where lock contention between
    the UI thread and a background thread can lead to a deadlock, as in my
    case. At the very least, I would expect this to be documented somewhere,
    preferably in a prominent place..
     
    chris

     

    Pratama Nur Wijaya <set.mnemonix@gmail.com> Mar 29 04:10AM -0700  

    thanks for your advice..
     
    On Friday, 29 March 2013 13:49:56 UTC+7, skink wrote:

     

    skink <pskink@gmail.com> Mar 29 08:35AM -0700  

    Pratama Nur Wijaya wrote:
     
    > map = mapFragment.getMap();
     
    > i'm tested my apps on Emulator api level 17
     
    post your activity_
    bank_location.xml layout file
     
    pskink

     

    Pratama Nur Wijaya <set.mnemonix@gmail.com> Mar 29 08:42AM -0700  

    here http://pastebin.com/DpWSTnr9
     
    On Friday, 29 March 2013 22:35:38 UTC+7, skink wrote:

     

    Jonathan S <xfsunoles@gmail.com> Mar 29 09:15AM -0700  

    Looks like Google Play services is not available in your tablet.
     
    On Friday, March 29, 2013 11:11:49 AM UTC-4, Pratama Nur Wijaya wrote:

     

    Pratama Nur Wijaya <set.mnemonix@gmail.com> Mar 29 09:38AM -0700  

    thanks for everyone here.. ^_^
     
    now its work on my tablet..
    i have change my code like these http://pastebin.com/hS0tKLEY
     
    i think i have got some logic error on these..
     
    if (map == null)
    {
    Utils.TRACE("API " + Build.VERSION.SDK_INT);
    FragmentManager fragmentManager = getSupportFragmentManager();
    SupportMapFragment supportMapFragment = (SupportMapFragment)
    fragmentManager.findFragmentById(R.id.map);
    map = supportMapFragment.getMap();
     
    if (map != null)
    {
    setupMap();
    }
    }
     
    On Friday, 29 March 2013 23:30:55 UTC+7, skink wrote:

     

    bob <bob@coolfone.comze.com> Mar 29 07:26AM -0700  

    Perhaps you should see this tip about "*largeHeap*":
     
    http://knowledge.lapasa.net/?p=772
     
    Thanks.
     
     
    On Thursday, March 28, 2013 1:41:27 PM UTC-5, Nathan wrote:

     

    Arun Kumar K <arunkumar.k@npcompete.net> Mar 29 06:18PM +0530  

    Hi guys,
     
    I have one doubt in activity.i want a call back when activity is destroyed
    from the stack.My constrain is
     
    1.Without using any inbuilt function like( onDestroy,finish and all)
     
    My another doubt is For example:
     
    ActivityA,ActivityB if i go to activityB from activityA then i go to
    activityA from activityB how can i find to reached the activityA.My
    constrain is
     
    1.Without using any inbuilt function like( onStart)
     
    --
    *Thanks & Regards*
    *K.Arun Kumar*

     

    bob <bob@coolfone.comze.com> Mar 29 07:17AM -0700  

    Why can't you use a built-in function?
     
     
    Thanks.
     
     
     
    On Friday, March 29, 2013 7:48:00 AM UTC-5, Arun Kumar K wrote:

     

You received this message because you are subscribed to the Google Group android-developers.
You can post via email.
To unsubscribe from this group, send an empty message.
For more options, visit this group.

--
--
You received this message because you are subscribed to the Google
Groups "Android Developers" group.
To post to this group, send email to android-developers@googlegroups.com
To unsubscribe from this group, send email to
android-developers+unsubscribe@googlegroups.com
For more options, visit this group at
http://groups.google.com/group/android-developers?hl=en
---
You received this message because you are subscribed to the Google Groups "Android Developers" group.
To unsubscribe from this group and stop receiving emails from it, send an email to android-developers+unsubscribe@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out.
 
 

[android-developers] Digest for android-developers@googlegroups.com - 20 Messages in 11 Topics

Comments: (0)

Group: http://groups.google.com/group/android-developers/topics

    Jonathan S <xfsunoles@gmail.com> Mar 28 11:11PM -0700  

    GoogleMap is still NULL after setupMapIfNeeded(). I hope you able to
    improve your coding ability.
     
    On Friday, March 29, 2013 12:27:45 AM UTC-4, Pratama Nur Wijaya wrote:

     

    Pratama Nur Wijaya <set.mnemonix@gmail.com> Mar 28 11:19PM -0700  

    if my GoogleMap still null.. why my program running in my phone.. and get
    force close on my tablet..
     
    can u give me some advice
     
    On Friday, 29 March 2013 13:11:30 UTC+7, Jonathan S wrote:

     

    skink <pskink@gmail.com> Mar 28 11:49PM -0700  

    Pratama Nur Wijaya wrote:
    > if my GoogleMap still null.. why my program running in my phone.. and get
    > force close on my tablet..
     
    > can u give me some advice
     
    see http://developer.android.com/reference/com/google/android/gms/maps/SupportMapFragment.html#getMap()
     
    it reads:
     
    Returns
    the GoogleMap. Null if the view of the fragment is not yet ready. This
    can happen
    if the fragment lifecyle have not gone through
    onCreateView(LayoutInflater, ViewGroup, Bundle) yet. This can also
    happen if Google Play services is not
    available.
     
    pskink

     

    Mario Giammarco <mgiammarco@gmail.com> Mar 28 03:29PM -0700  

    Still debugging.
    What I have found:
    1) no deadlocks
    2) stomp library uses hawtdispatch. They then use java NIO and they open
    sockets and channels.
     
    The problem is:
    - the thread waiting for a reply is doing the right thing: infact nothing
    is sent so nothing can be received.
     
    So the real problem is: the thread that must send data with nio send
    nothing.
    But if I go in debugger stepping mode it sends data!
    I am getting crazy, since I am 99% sure that there is no deadlock.
    Does the debugger disable jit? Change bytecode?
     
    Il giorno domenica 24 marzo 2013 23:49:40 UTC+1, Kostya Vasilyev ha scritto:

     

    Jungle Jim <jjjunglejim@gmail.com> Mar 28 02:41PM -0700  

    I am developing an app that will have both free content and the option to
    receive additional content by purchasing an annual subscription. I am using
    Version 3 API, which has only recently supported subscriptions.
     
    I am getting error messages that say my devices do not support
    subscriptions. I doubt that is the case and that I'm probably doing
    something wrong in my code.
     
    Does anyone have an app published on Google Play that has a subscription
    available that uses Version 3 of the API? Or know of one available? I would
    like to test my devices to determine whether or not they truly have the
    support for subscriptions. If your subscription price is reasonable, I will
    buy one to test my devices.
     
    Thanks.

     

    Tobiah <toby@tobiah.org> Mar 28 02:31PM -0700  

    I'm doing a startActivityForResult() to launch a barcode scanner.
    The problem is that if the user presses home from the scanner, then goes back to
    my app and launches the scanner again, it appears to be launching a new
    instance, and the camera is already busy, so it fails.
     
    Is there a way to launch another app in this way, while insuring
    that I will only launch one instance of the other app?
     
    Thanks,
     
    Tobiah

     

    Heidi Zhou Nolan <heidi.zh.n@gmail.com> Mar 28 01:41PM -0700  

    Hello!
     
    I like to try Open Bluetooth Low Energy API (
    http://android-btle.github.com/framework/) in order to work with Bluetooth
    4.0.
     
    However, I can't download the library
    from https://code.google.com/p/broadcom-ble/. I got permission denied
    error.
     
    Any idea what to do in order to get the library?
     
    Thank you in advance.
     
    Heidi

     

    Nathan <nathan.d.mellor@gmail.com> Mar 28 11:49AM -0700  

    Currently my manifest is wide open.
     
    Let's suppose I want to exclude a device that I don't think is working for
    many people anyway.
     
    What will happen to users that have it installed?
     
    Will Google Play immediately zap the app from everyone, including those who
    have paid for it?
    Will it stay there forever and and just not get updates?
    Will I get angry emails from people who say it actually was working and
    where did it go? Or that gee, it never really worked but it must be really
    defective if it uninstalls itself?
     
    Nathan

     

    RichardC <richard.critten@googlemail.com> Mar 28 12:17PM -0700  

    No Zaps - just no updates.
     
    On Thursday, March 28, 2013 6:49:32 PM UTC, Nathan wrote:

     

    Nathan <nathan.d.mellor@gmail.com> Mar 28 12:36PM -0700  

    On Thursday, March 28, 2013 12:17:04 PM UTC-7, RichardC wrote:
     
    > No Zaps - just no updates.
     
    Thanks - I'll try it with one and see how it goes - whether it worth adding
    more exclusions.
     
    Nathan

     

    Nathan <nathan.d.mellor@gmail.com> Mar 28 11:41AM -0700  

    Does anyone know the specs on the Transformer TFT700T? Particularly the VM
    heap size so I can set up an emulator.
     
    It seems that device is forgetful - ie runs out of memory more often than
    other devices - according to some error logs I have.
     
    I have a theory that a large pixel screen combined with a low heap space
    number could cause problems for my app.
     
    I've tried setting up one with 24M, with 1920x1200, and other settings that
    look reasonable. It comes up black and never finishes booting. I could be
    doing other things wrong, of course. I've only setup a very few emulators
    since working on a MacBook.
     
    I saw some info that implies the heap space is 24M out of 1G total Ram. It
    also suggested people can change this setting, but on further searching
    that appears to be only with Root and selected apps.
     
    Nathan

     

    Marta Ribeiro <martacunharibeiro90@gmail.com> Mar 28 09:11AM -0700  

    Yes it is.
     
    On Wednesday, March 13, 2013 8:21:22 PM UTC, bob wrote:

     

    Marta Ribeiro <martacunharibeiro90@gmail.com> Mar 28 09:12AM -0700  

    I did. But I still have the problem. So I'm guessing the problem lays on
    Eclipse.
     
    On Saturday, March 16, 2013 7:54:37 AM UTC, Jason wrote:

     

    Marta Ribeiro <martacunharibeiro90@gmail.com> Mar 28 09:14AM -0700  

    Yes that works. Actually as a workaround I've been copying the app apk to
    the phone and installing it by clicking on the apk file. But still it would
    be faster to run it directly through eclipse.
     
    On Friday, March 15, 2013 2:31:57 PM UTC, bob wrote:

     

    bob <bob@coolfone.comze.com> Mar 28 10:54AM -0700  

    What happens if you do this?
     
    *adb shell*
     
    *am start -n com.example.test/com.example.test.MainActivity -a
    android.intent.action.MAIN -c android.intent.category.LAUNCHER*
     
     
    Thanks.
     
     
    On Thursday, March 28, 2013 11:14:05 AM UTC-5, Marta Ribeiro wrote:

     

    Mahmoud Hadad <mahmoudkatkoot@gmail.com> Mar 28 08:23AM -0700  

    useless replies tbh
     
    On Tuesday, September 13, 2011 4:34:41 PM UTC+3, Conny wrote:

     

    ankit the OPIUM <ankitgoyal1009@gmail.com> Mar 28 06:16AM -0700  

    If this not a web app what you can do is when your app launch retrieve data
    from server and save in device and fetch from there. After that check your
    server regularly if something has been changed over server and download
    only those change files not whole data. in this way only first time it will
    take more time.
     
    *For example* :
     
     
    - you server has 100 records when app launch first time(just after
    installation) you download 100 records
    - after 5 min you check your server and found 5 entries has been changed
    so you will download only 5 records.
    - next time when app launches again you will check only changed records
    on server and download them only.
     
     
     
    On Monday, 18 March 2013 16:56:54 UTC+5:30, sree wrote:

     

    bob <bob@coolfone.comze.com> Mar 28 07:04AM -0700  

    This is called caching.
     
    On Thursday, March 28, 2013 8:16:29 AM UTC-5, ankit the OPIUM wrote:

     

    abhijit <achakra21@gmail.com> Mar 28 01:38AM -0700  

    I have a ViewPager in which i am displaying 5 Fragments like f1,f2,f3,f4,f5when i am pressing back button from in f2 it coming to f1(f2->f1) but when
    i am trying to come from f3 to f1 the back key is not responding.f3->f1 is
    not happening.Why f2 to f1 is possible and why after f2 its not possible
    *i am using this code:*
     
    public boolean onKeyDown(int keyCode, KeyEvent event) {
    if (keyCode == KeyEvent.KEYCODE_BACK) {
    pager.setCurrentItem(0, true);
    return true;
    }
    return true;}

     

You received this message because you are subscribed to the Google Group android-developers.
You can post via email.
To unsubscribe from this group, send an empty message.
For more options, visit this group.

--
--
You received this message because you are subscribed to the Google
Groups "Android Developers" group.
To post to this group, send email to android-developers@googlegroups.com
To unsubscribe from this group, send email to
android-developers+unsubscribe@googlegroups.com
For more options, visit this group at
http://groups.google.com/group/android-developers?hl=en
---
You received this message because you are subscribed to the Google Groups "Android Developers" group.
To unsubscribe from this group and stop receiving emails from it, send an email to android-developers+unsubscribe@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out.