Windows Phone 8.1 app error calling Windows Phone Silverlight 8.1 new Socket exception - sockets

Environment: Windows 10 eval copy; VS 2013 Express Update3.
I created a Blank WP 8.1 app and a Windows Phone Silverlight Class Library. I added a class to the class library and in the constructor attempt to new a Socket. The Blank WP 8.1 app creates an instance of the class in its OnNavigatedTo event. That's the only change I made to the WP 8.1 app.
The call in the Class1 constructor generates the exception given below. This occurs in the Emulator 8.1 WVGA 4 inch 512MB. Can anyone tell me what I'm doing wrong?
MainPage.xml.cs:
protected override void OnNavigatedTo(NavigationEventArgs e)
{
Class1 class1 = new Class1();
...
namespace SockLib
{
public class Class1
{
public Class1()
{
try
{
// The next line generates the exception.
Socket sock = new Socket(AddressFamily.InterNetwork,
SocketType.Stream, ProtocolType.Tcp);
}
catch (Exception e)
{
System.Diagnostics.Debug.WriteLine(e.ToString());
}
}
}
}
A first chance exception of type 'System.NullReferenceException' occurred in System.Net.ni.DLL
System.NullReferenceException: Object reference not set to an instance of an object.
at System.Net.Sockets.SocketNotificationManager..ctor()
at System.Net.Sockets.Socket.get_SocketManager()
at System.Net.Sockets.Socket..ctor(AddressFamily addressFamily, SocketType socketType, ProtocolType protocolType)
at SockLib.Class1..ctor()

Will,
I suspect you cannot use OnNavigatedTo, because the UI is not built yet (and maybe something else, that have to do with the App capabilities on the phone). Would you try to use a button Click event handler to initialize your socket?
I have a very similar problem. I am trying to create a pure console application for WP8.1, and my Socket constructor fails the same way.

Related

How to re-register refit handlers in MAUI

In my MAUI, I register custom auth handlers using the following code snippet.
public static class MauiProgram
{
public static MauiApp CreateMauiApp()
{
var builder = MauiApp.CreateBuilder();
builder
.UseMauiApp<App>()
.UseMauiCompatibility()
.UseMauiCommunityToolkit()
.RegisterRefitClients()
return builder.Build();
}
}
And my RegisterRefitClients extension method looks like this.
public static MauiAppBuilder RegisterRefitClients(this MauiAppBuilder mauiAppBuilder)
{
mauiAppBuilder.Services.AddRefitClient<IMyApiService>()
.ConfigureHttpClient(c => c.BaseAddress = new Uri(Configuration.BASE_URL))
.AddHttpMessageHandler<AuthHeaderHandler>();
return mauiAppBuilder;
}
Here, the thing is, Configuration.BASE_URL can change at run time. By default, it always points to our production server. But the testers can change it to staging/dev server at run time. I want to invoke this method at run time, in order to set custom auth handler for the newly changed staging server.
But I cannot get the MauiAppBuilder instance anywhere in the app once it is initiated.
How do I call the RegisterRefitClients at run time?
RegisterRefitClients can not be called when the program start running. .NET MAUI enables apps to be initialized from a single location. The MauiProgram class is the entry point to the application, setting up configuration and wiring up services the application will use. Once the program starts, you can not change the MauiProgram.

JBoss 7.1.0: executor.submit() how to debug (no exception but task not started)

I have an issue with JBoss EAP 7.1.0 GA. On one server (my DEV) this works like a charm while on the other (TEST environment) the Callable executed using executor.submit() does not seem to be started (I do not see that "This is call" message in log), but no exception or any other clue is given.
The question is - where should I look like / how should I debug this issue?
The calling code:
#Resource(name = "DefaultManagedExecutorService")
ManagedExecutorService executor;
try {
DownloadPlayers dp = new DownloadPlayers();
Future<Queue<PlayerForDownload>> f = executor.submit(dp);
Queue<PlayerForDownload> q = f.get();
L.info(q.size());
} catch (Exception e) {
L.error("EXCEPTION" + e.getMessage());
}
The class it calls:
public class DownloadPlayers implements Callable<Queue<PlayerForDownload>> {
// the constructor gets called, I'm sure as it writes to log
// the call is as simple as this
#Override
public Queue<PlayerForDownload> call() {
L.info("This is call()");
try {
return this.getPlayersForDownload();
} catch (WorkerException e) {
L.error(e);
return null;
}
}
}
As stated above, the code itself seems to be OK as it works in one server but does not work on the other. Both are
7.1.0GA standalone.
Any advice how to debug the ManagedExecutorService?
Thanks.
In this particular case the problem was that on the TEST environment it was only allowed to run two threads which we already running (by some completely different part of application I didn't realize). So the problem is solved now by setting the "Core threads" parameter in ManagerExecutorService to higher value, the tasks are running.
However the tricky part was that there was no obvious visible difference between the JBoss servers (I compared the standalone.xml configs...) just because the ManagerExecutorService in JBoss has some default (blank) values that actually depend on the system config (v-CPU cores in my case). So despite the config being the same, the "Core threads" seem to default to 2 on TEST and some higher (unknown to me) value on my DEV.
So never depend on default settings in ManagerExecutorService if comparing two environments.
I have also rewritten the logic, instead of using blocking Future.get() or checking for Future.isDone() in a loop a do Future.get() with timeout and in the exception handler I decide whether to keep waiting or fail.

java.lang.IllegalStateException: Could not construct instance of helper class class

I used **ormlite greendao** for my Android application, everything works well until my android phone upgrade to Android 6.0 (Marshmallow) these days.
The problem happens at this line in my code:
private DatabaseHelper getHelper() {
if (mHelper == null)
mHelper = OpenHelperManager.getHelper(mContext, DatabaseHelper.class);// Error raise here
return mHelper;
}
And I catch out this causes the issue:
java.lang.IllegalAccessException:
java.lang.Class is not
accessible from
java.lang.Class
I have researched on the internet but couldn't figure out.
It only happens in Android 6.
Update
Cause:
java.lang.IllegalAccessException: java.lang.Class<com.mypackage.DatabaseHelper> is not accessible from java.lang.Class<com.j256.ormlite.android.apptools.OpenHelperManager>
I solved my problem by myself
As the cause problem above I changed class access level of
com.mypackage.DatabaseHelper
into public class.

Why is my CE app refusing to run?

I've been maintaining a Windows CE app for some time now (over a year) and have produced new versions of it from time to time, copying them to the handheld device[s] and running the new versions there.
Today, though, I created a new Windows CE app for the first time. It is a very simple utility.
To create it in VS 2008, I selected a C# "Smart Device Project" template, added a few controls and a bit of code, and built it.
Here are some of the options I selected:
I copied the .exe produced via building the project to the handheld device's Program Files folder:
...but it won't run. Is it in the wrong location? Does it need some ancillary files copied over? Is there some other sort of setup I need to do to get it to run? Or what?
UPDATE
Since there's not much of it, I'm pasting ALL the code below in case somebody thinks my code could be the problem:
using System;
using System.Linq;
using System.Collections.Generic;
using System.ComponentModel;
using System.Data;
using System.Drawing;
using System.Text;
using System.Windows.Forms;
using System.IO.Ports;
namespace PrinterCommanderCE
{
public partial class PrinterCommanderForm : Form
{
public PrinterCommanderForm()
{
InitializeComponent();
}
private void btnSendCommands_Click(object sender, EventArgs e)
{
SendPrinterCommands();
}
private void SendPrinterCommands()
{
bool successfulSend = false;
const string quote = "\"";
string keepPrinterOn = string.Format("! U1 setvar {0}power.dtr_power_off{0} {0}off{0}", quote);
string shutPrinterOff = string.Format("! U1 setvar {0}power.dtr_power_off{0} {0}on{0}", quote);
string advanceToBlackBar = string.Format("! U1 setvar {0}media.sense_mode{0} {0}bar{0}", quote);
string advanceToGap = string.Format("! U1 setvar {0}media.sense_mode{0} {0}gap{0}", quote);
if (radbtnBar.Checked)
{
successfulSend = SendCommandToPrinter(advanceToBlackBar);
}
else if (radbtnGap.Checked)
{
successfulSend = SendCommandToPrinter(advanceToGap);
}
if (successfulSend)
{
MessageBox.Show("label type command successfully sent");
}
else
{
MessageBox.Show("label type command NOT successfully sent");
}
if (ckbxPreventShutoff.Checked)
{
successfulSend = SendCommandToPrinter(keepPrinterOn);
}
else
{
successfulSend = SendCommandToPrinter(shutPrinterOff);
}
if (successfulSend)
{
MessageBox.Show("print shutoff command successfully sent");
}
else
{
MessageBox.Show("print shutoff command NOT successfully sent");
}
}
private bool SendCommandToPrinter(string cmd)
{
bool success = false;
try
{
SerialPort serialPort = new SerialPort();
serialPort.BaudRate = 19200;
serialPort.Handshake = Handshake.XOnXOff;
serialPort.Open();
serialPort.Write(cmd);
serialPort.Close();
success = true;
}
catch
{
success = false;
}
return success;
}
}
}
UPDATE 2
Based on this, I added a global exception handler to the app so that Program.cs is now:
namespace PrinterCommanderCE
{
static class Program
{
[MTAThread]
static void Main()
{
AppDomain currentDomain = AppDomain.CurrentDomain;
currentDomain.UnhandledException += new UnhandledExceptionEventHandler(GlobalExceptionHandler);
Application.Run(new PrinterCommanderForm());
}
static void GlobalExceptionHandler(object sender, UnhandledExceptionEventArgs args)
{
Exception e = (Exception)args.ExceptionObject;
MessageBox.Show(string.Format("GlobalExceptionHandler caught : {0}", e.Message));
}
}
}
Yet running the new build shows nothing - it just "flashes" momentarily with about as much verbosity as Lee Harvey Oswald after Jack Ruby's friendly visit.
UPDATE 3
Could the problem be related to this, and if so, how to solve it?
The circumstance that both my updated version of an existing app AND this brand new and simple app refuse to run indicate there is something fundamentally flawed somewhere in the coding, building, or deployment process.
UPDATE 4
As this is a minimal utility, the reason it (and my legacy, much more involved) app are not working may have something to do with the project properties, how it's being built, a needed file not being copied over, or...???
NOTE: The desktop icon is "generic" (looks like a blank white form); this perhaps indicates a problem, but is it indicative of something awry or is it a minor (aesthetics-only) problem?
UPDATE 5
In Project > Properties..., Platform is set to "Active (Any CPU)" and Platform target the same ("Active (Any CPU)")
I have read that this is wrong, that it should be "x86", but there is no "x86" option available - Any CPU is the only one...?!?
UPDATE 6
In Project > Properties... > Devices, the "Deploy the latest version of the .NET Compact Framework (including Service Packs)" is checked. Is this as it should be?
UPDATE 7
Okay, here's the really strange part of all this:
I have two CF/CE apps that I need to run on these Motorola/Symbol 3090 and 3190 handheld devices.
One is this simple utility discussed above. I find that it actually does run on one of the devices (the 3190, FWIW). So it runs on one device, but not on the other.
HOWEVER, the other (legacy) .exe is the opposite - it runs on the 3090 (where the utility will not even start up), but not on the 3190.
So the utility's needs are met by the 3190, and the legacy util's needs are met by the 3090. However, the NEW version of the legacy app does not run on either device!
I am baffled; I feel as Casey Stengel must have when speaking once of his three catchers: "I got one that can throw but can't catch, one that can catch but can't throw, and one who can hit but can't do either."
UPDATE 8
The 3190 has a newer version of the CF installed; it seems that both the new and the old apps should run on the new device with the newer CE, but they don't - only the one built against/for the new framework does...
UPDATE 9
Here is what the 3090 looks like:
UPDATE 10
So I have two exes, one that runs on the devices (both of them now), and the other that will run on neither of the devices. The two exesw seem almost identical. I compared them with three tools: Red Gates' .NET Reflector; JetBrains' dotPeek, and Dependency Walker.
Here is what I found:
Dependency Walker
Both seem to have the same errors about missing dependencies (I didn't have them in the same folder with their dependent assemblies is probably the problem there)
.NET Reflector
The nonworking file has this entry that the working file does not:
[assembly: Debuggable(0x107)]
Is this the problem and, if so, how can I change it?
JetBrains dotPeek
The References in the working copy of the exe are all version 1.0.50000.0
The non-working exe has an identical list of References, and the same version number.
There is this difference, though:
For the working .exe, dotPeek says, "1.4.0.15, msil, Pocket PC v3.5"
For the non-working .exe, dotPeek says, "1.4.0.15, msil, .Net Framework v4.5"
Is this the problem and, if so, how can I change the non-working .exe to match the working one?
This last is disconcerting, primarily because I see no place in the non-working (newer) version of the project where a "4.5" string exists. Where could dotPeek be getting that information?
UPDATE 11
I do know now that the problem is somewhere between these two MessageBox.Show()s, because the first one I see, but not the second:
public static int Main(string [] args)
{
try
{
// A home-brewed exception handler (named ExceptionHandler()) is already defined, but I'm adding a global one
// for UNHANDLED exceptions (ExceptionHandler() is explicitly called throughout the code in catch blocks).
MessageBox.Show("made it into Main method"); // TODO: Remove after testing <= this one is seen
AppDomain currentDomain = AppDomain.CurrentDomain;
currentDomain.UnhandledException += new UnhandledExceptionEventHandler(GlobalExceptionHandler);
string name = Assembly.GetExecutingAssembly().GetName().Name;
IntPtr mutexHandle = CreateMutex(IntPtr.Zero, true, name);
long error = GetLastError();
if (error == ERROR_ALREADY_EXISTS)
{
ReleaseMutex(mutexHandle);
IntPtr hWnd = FindWindow("#NETCF_AGL_BASE_",null);
if ((int) hWnd > 0)
{
SetForegroundWindow(hWnd);
}
return 0;
}
ReleaseMutex(mutexHandle);
DeviceInfo devIn = DeviceInfo.GetInstance();
Wifi.DisableWifi();
// Instantiate a new instance of Form1.
frmCentral f1 = new frmCentral();
f1.Height = devIn.GetScreenHeight();
f1.Text = DPRU.GetFormTitle("DPRU HHS", "", "");
MessageBox.Show("made it before Application.Run() in Main method"); // TODO: Remove after testing <= this one is NOT seen
Application.Run(f1);
devIn.Close();
Application.Exit();
return 0;
}
catch(Exception ex)
{
DPRU.ExceptionHandler(ex, "Main");
return 0;
}
} // Main() method
UPDATE 12
More specifically, I've got infinite looping going on somehow; By mashing the "Ent" pill on the handheld device (that's what the button looks like - a "lozenge") - it sounds like gerbils tap-dancing (as debugging MessageBox.Show()s in two methods pop up and are dismissed over and over ad infinitum ad (literally) nauseum).
If an application does not start it is mostly missing something. As you compiled for WindowsCE and CF3.5, the Compact Framework 3.5 runimes have to be installed on the WindowsCE device.
Normally Compact Framework is part of Windows CE images, at least version 1.0, but who knows for your test device? If at least one CF is installed, an app requiring a newer CF version will show that on start by a message stating about the missed version. So either no CF is on your device, or something is goind real wrong.
You can run \Windows\cgacutil.exe to check the CF version installed on the device. The tool will show the version of installed CF.
You can debug using a TCP/IP connection or ActiveSync connection. See remote debuggung elsewhere in stackoverflow, I wrote a long aanswer about remote debug via TCP/IP. Or does your device neither have USB and WLAN or ENET?
Update: Here is the answer for remote debug via tcp/ip: VS2008 remotely connect to Win Mobile 6.1 Device This will also enable the remote deployment "In Project > Properties... > Devices, the "Deploy the latest version of the .NET Compact Framework (including Service Packs)" is checked. Is this as it should be?"
Are the earlier apps you wrote also written .NET? Compact framework does not care about the processor architecture, only the CF runtimes have to match the processor. So you do not need an x86 target as if you write a native C/C++ SmartDevice project.
To your comments:
a) CF1.0 is installed on the device.
b) the exe built on the colleagues computer seems to be built for CF1 and therefor runs OK.
c) your exe is built for CF 3.5 and does not run as there is no CF3.5 runtime on the device.
d) most CF exe files are very small as long as they do not include large resources or ...
Conclusion so far: Install the CF3.5 runtime onto the device: http://msdn.microsoft.com/en-us/library/bb788171%28v=vs.90%29.aspx.
To run the legacy app on both devices, the referenced Motorola or other 3rd party runtimes must also be installed. I stringly recommand to setup your environment so you can use ActiveSync/WMDC for development, deployment and debugging of the device. If you are unable look for some more experienced colleague.
Can you try to run it inside the debugger and check where it fails?
Can you place a breakpoint right at the beginning of Program.main and check if it's reached?
Debug output may also give you some interesting hints.

How do I know when user minimizes / maximizes Eclipse?

I need to respond to the events of minimizing / maximizing Eclipse window. How do I do that?
I can suggest a way: you can write a plugin for it.
For example see this improvized "tutorial", I made it, tried it works on Ganymede. A bit ugly at the final Shell variable, but working. If you know nicer solution just shoot :)
((actually there is a way: to extend your own ControlListener class, but that needs more coding :))
Create a new Plug-in Project, name it as you want, create it from a template named: Hello World Command
Open the SampleHandler class, and then replace the execute() function with this code.
public Object execute(ExecutionEvent event) throws ExecutionException {
IWorkbenchWindow window = HandlerUtil
.getActiveWorkbenchWindowChecked(event);
final Shell s = window.getShell();
window.getShell().addControlListener(new ControlListener() {
#Override
public void controlMoved(ControlEvent e) {
// TODO Auto-generated method stub
}
#Override
public void controlResized(ControlEvent e) {
MessageDialog.openInformation(s,
"WindowEventHandler Plug-in", "RESIZED: "
+ e.toString() + "\nHello, Eclipse world");
}
});
MessageDialog.openInformation(window.getShell(),
"WindowEventHandler Plug-in",
"Hello, Eclipse world, resize will be taken care of.");
return null;
}
now. Start the project (Run As-> Eclipse application), and you'll se an Eclipse button on the toolbar. Click on it! It triggers the running of the above code where the essence is that the window.getShell() returns with the main window component so you can add listeners to it.
If you want it to run automatically, not just for a button, you have to find out a plugin where the entry point is connected to the starting of the application.
Hope this helps.
b
Found a way to do it easily: you have to create a ShellListener or ShellAdapter, which have methods that are called when the shell is iconified, deiconified, activated, deactivated and closed.
After creating it, add it as a listener with the following line:
int i;
PlatformUI.getWorkbench().getActiveWorkbenchWindow().getShell().addShellListener( yourListenerHere);
If you ever remove it from the shell's listeners list, be sure that Workbench, ActiveWorkbnchWindow and Shell are not null.