Hello guys.
I'm building a really simple app which uses the DatagramSocket to send small text messages between two clients. in fact, I'm doing it all in one app, where I'm instantiating 2 DatagramSockets which listen to different ports to simulate the two clients.
Now, I'm only sending and receiving strings. Here's the code I'm using to send a string:
enviaMsg: function (msg) {
if (!this.prontoEnviar) return;
var writer = new Windows.Storage.Streams.DataWriter(this.socket.outputStream);
writer.writeString(msg);
writer.storeAsync()
.done(function () {
writer.detachBuffer();
});
},
And here's the code I'm using to receive it:
mensagemRecebida: function (e) {
var buffer = e.detail[0].getDataReader();
var numeroChars = buffer.unconsumedBufferLength;
var str = buffer.readString(numeroChars);
}
And now, the question. The docs say that unconsumedBufferLength (http://msdn.microsoft.com/en-us/library/windows/apps/windows.storage.streams.datareader.unconsumedbufferlength.aspx) returns
the size of the buffer that has not been read, in bytes. But if I'm not mistaken, the JS standard says that each char in a string is stored in 16 bits (ie, 2bytes). That being the, shouldn't the number of chars I recover from the unconsumedBufferLength property
equals the number of chars of my string multiplied by 2? My tests show that this is not happening and that when I write a string, the unconsumedBufferLength property will always return the number of chars in that string...Interestingly, writing integers
I've probably missed something, right?
Btw, and since I'm asking about the DataReader class, I've got another question. It introduces readint16, readint32 and readint64. I understand that those make sense in C#, but in JavaScript, aren't all numbers represented by 64 bit floats??
Luis Abreu
