|
So what? The compiler can't enforce intents expressed in documentation. All that matters is that the class provides the signatures required by the interface -- the members don't even need to do anything, or they can throw NotImplementedExceptions for all the compiler cares.
A developer can implement an interface any darn way he likes; an IEnumerable can access a database or a Random or any old thing, an IQueryable can access a Web Service or IME or shared memory, the interfaces don't care. It all depends on what the developer needs to do. It is then up to the user of the class to decide whether or not that class provides the desired information in a useful manner. Other than that, the class is a black box and you shouldn't go poking your fingers in it.
|
|
|
|
|
There are two sides to interfaces: one for communicating with a compiler, and one for communicating with programmers who read your code. You keep arguing that there's only the first side, pretending the second side does not exist. However, it's there, and I am sure you've been made painfully aware of it on more than one occasion, when the code compiled but did not work.
There are legitimate concerns that compilers simply cannot enforce, yet these concerns do not become less relevant because of it. For example, compilers cannot enforce exact type casts inside fast enumeration loops, but this means only that the programmers should pay more attention.
There is more to implementing an interface than providing implementations to all its methods so that the compiler stops complaining. For example, if you implement IQueryable in a way that reads everything from the database and then filters in memory, you will "provide the desired information", but not "in a useful manner". That does not make your class useless - others may still use it in their code. It's just that they will not be using it in a way they normally use an IQueryable.
I completely agree with your "class is a black box" comment, yet "poking one's finger into a class" is drastically different from understanding the intent behind its implementation.
|
|
|
|
|
dasblinkenlight wrote: one for communicating with programmers who read your code
Which is the name of the interface and the signatures of the members -- any intent expressed in other documentation, while instructive, is not part of the contract and may be incorrect or may not have been obeyed by the developer of any particular class that implements the interface. Such documentation should be taken with a shot of tequila.
My own pet peeve with this regard is the documentation of IDataReader, which Microsoft's providers implement in surprising ways.
|
|
|
|
|
Name+signatures is a good start, but there are other items there that everyone follows:
- Documentation may ask not to pass null or negative values for certain arguments,
- Documentation may ask to pass only sorted collections to certain methods (e.g. a binary search),
- Documentation may ask not to call a method in a particular state (e.g. creating a statement on a closed connection),
- Documentation may ask to call a method in a context of an ambient transaction,
- Documentation may ask not to call a method before calling another method (e.g. obtaining a write lock without releasing a read lock first)
The list goes on and on. I'm sure there are classes and methods with awful documentation, as well as implementations written without regard to the interface documentation (as you mentioned). But that does not mean that all documentation is somehow bad, only that there is some room to grow 
|
|
|
|
|
That sounds like implementation-specific documentation, not interface documentation. One needs to keep the two concepts separate.
|
|
|
|
|
dasblinkenlight wrote: In case of LINQ to SQL, this means pushing the query logic into RDBMS
Which has no bearing on any other particular implementation the OP may be considering.
|
|
|
|
|
I'm using VS 2010 on Windows 7. I found this code in a Google search, and it compiles and runs, but when I pass in values to SetVolume, there is no change in the sound volume:
public static class Volume
{
[DllImport("winmm.dll")]
private static extern int waveOutGetVolume(IntPtr hwo, out uint dwVolume);
[DllImport("winmm.dll")]
private static extern int waveOutSetVolume(IntPtr hwo, uint dwVolume);
public static int GetVolume()
{
uint CurrVol = 0;
waveOutGetVolume(IntPtr.Zero, out CurrVol);
ushort CalcVol = (ushort)(CurrVol & 0x0000ffff);
int volume = CalcVol / (ushort.MaxValue / 10);
return volume;
}
public static void SetVolume(int Volume)
{
int NewVolume = ((ushort.MaxValue / 10) * Volume);
uint NewVolumeAllChannels = (((uint)NewVolume & 0x0000ffff) | ((uint)NewVolume << 16));
waveOutSetVolume(IntPtr.Zero, NewVolumeAllChannels);
}
}
I found a couple of Mixer type projects here on CP, but they're over the top as far as wat I'm looking for. I simply want to change the volume level of the speakers.
Anyone know how to do this?
>
Everything makes sense in someone's mind
|
|
|
|
|
I've never used the Wave functions but I have extensively used the MIDI functions that are very similar and in the same dll.
I belive you may need to call WaveOutOpen first to get a valid handle, then call SetVolume using that handle and finally call WaveOutClose, again using that handle.
I haven't looked at the documentation at all so I may be way off base here, but this is the way the stuff I have used works.
|
|
|
|
|
Could you share some code? I cannot for the life of me find any C# code that handles getting and setting the speaker volume.
Everything makes sense in someone's mind
|
|
|
|
|
I've had a play and this seems to work. This is the extent of my testing but I hope it helps!
It's quite a lot of code - welcome to the world of PInvoke. There is more that needs doing such as checking the result of each API call and raising a custom exception if != MMSYSERR_NOERROR etc...
using System.Runtime.InteropServices;
[StructLayout(LayoutKind.Sequential)]
internal struct WAVEFORMATEX
{
public short wFormatTag;
public short nChannels;
public int nSamplesPerSec;
public int nAvgBytesPerSec;
public short nBlockAlign;
public short wBitsPerSample;
public short cbSize;
}
using System;
using System.Runtime.InteropServices;
internal delegate void waveOutProc(
IntPtr hwo,
int uMsg,
IntPtr dwInstance,
IntPtr dwParam1,
IntPtr dwParam2);
internal static class NativeMethods
{
public const int MMSYSERR_NOERROR = 0;
public const int CALLBACK_FUNCTION = 0x00030000;
[DllImport("winmm.dll", SetLastError = true)]
public static extern int waveOutOpen(
out IntPtr phwo,
int uDeviceID,
ref WAVEFORMATEX pwfx,
waveOutProc dwCallback,
IntPtr dwCallbackInstance,
int fdwOpen);
[DllImport("winmm.dll", SetLastError = true)]
public static extern int waveOutClose(IntPtr hwo);
[DllImport("winmm.dll", SetLastError = true)]
public static extern int waveOutGetVolume(
IntPtr hwo,
out uint pdwVolume);
[DllImport("winmm.dll", SetLastError = true)]
public static extern int waveOutSetVolume(
IntPtr hwo,
uint dwVolume);
}
using System;
public class WaveOut : IDisposable
{
private waveOutProc callback;
private IntPtr handle;
private int id;
public WaveOut(int id)
{
callback = new waveOutProc(Callback);
this.id = id;
}
~WaveOut()
{
Dispose(false);
}
public int ID
{
get { return id; }
}
private void Callback(IntPtr hwo, int uMsg, IntPtr dwInstance,
IntPtr dwParam1, IntPtr dwParam2)
{
}
public void Close()
{
if (handle != IntPtr.Zero)
NativeMethods.waveOutClose(handle);
handle = IntPtr.Zero;
}
public void Dispose()
{
Dispose(true);
}
protected virtual void Dispose(bool disposing)
{
if (disposing)
{
}
if (handle != IntPtr.Zero)
Close();
}
public uint GetVolume()
{
uint volume = uint.MaxValue;
if (handle != IntPtr.Zero)
NativeMethods.waveOutGetVolume(handle, out volume);
return volume;
}
public void Open()
{
if (handle == IntPtr.Zero)
{
WAVEFORMATEX waveFormatEx = new WAVEFORMATEX();
waveFormatEx.wFormatTag = 1;
waveFormatEx.nChannels = 2;
waveFormatEx.nSamplesPerSec = 44100;
waveFormatEx.wBitsPerSample = 16;
waveFormatEx.nBlockAlign = (short)(waveFormatEx.nChannels * (waveFormatEx.wBitsPerSample / 8));
waveFormatEx.nAvgBytesPerSec = waveFormatEx.nSamplesPerSec * waveFormatEx.nBlockAlign;
NativeMethods.waveOutOpen(out handle, id,
ref waveFormatEx, callback, IntPtr.Zero,
NativeMethods.CALLBACK_FUNCTION);
}
}
public void SetVolume(uint volume)
{
if (handle != IntPtr.Zero)
NativeMethods.waveOutSetVolume(handle, volume);
}
}
using System;
internal class Test
{
private static void ConvertFromVolume(uint volume, out int left, out int right)
{
left = (int)volume & 0xFFFF;
right = (int)(volume >> 16) & 0xFFFF;
}
private static uint ConvertToVolume(int left, int right)
{
return (uint)right << 16 | (ushort)left;
}
public void Run()
{
WaveOut waveOut = new WaveOut(0);
waveOut.Open();
int originalLeft;
int originalRight;
ConvertFromVolume(waveOut.GetVolume(), out originalLeft, out originalRight);
Console.WriteLine(
"Left: {0}, Right: {1}",
originalLeft, originalRight);
int newLeft = 32768;
int newRight = 32768;
waveOut.SetVolume(ConvertToVolume(newLeft, newRight));
ConvertFromVolume(waveOut.GetVolume(), out newLeft, out newLeft);
Console.WriteLine(
"Left: {0}, Right: {1}",
newLeft, newLeft);
waveOut.SetVolume(ConvertToVolume(originalLeft, originalRight));
ConvertFromVolume(waveOut.GetVolume(), out newLeft, out newLeft);
Console.WriteLine(
"Left: {0}, Right: {1}",
newLeft, newLeft);
waveOut.Close();
}
}
using System;
class Program
{
static void Main(string[] args)
{
new Test().Run();
Console.ReadKey();
}
}
|
|
|
|
|
Thanks, I'll have a go at it.
Everything makes sense in someone's mind
|
|
|
|
|
Ok, I've coded it.
What should be happening when I run this?
Everything makes sense in someone's mind
|
|
|
|
|
When I run it on my system it
1. retrieves the original volume
2. then changes it and retrieves it agin to show it is changed.
3. It then sets it back to the original and retrieves it again to it has been changed back.
If you remove step 3 it should remain changed to whatever you set it to in 2. Check out the code in Test.Run(), I have commented it all.
|
|
|
|
|
Ok, that's what I figred should be happening.
I'm listening to music right now via headphones. When I run this, the sound level doesn't change.
Everything makes sense in someone's mind
|
|
|
|
|
I haven't actually tested it with audio - just the values from GetVolume so I see that the driver has responded. There could be one of two things here...
1. The device ID needs to be the correct one for the device who's volume you want to adjust - I'm not sure without further research how to determine the correct ID for the main audio out device/master
2. Maybe the API only adjusts the volume for the audio stream you play directly - unlikely though.
Edit: you can use waveOutGetNumDevs to get the number of devices and waveOutGetDevCaps in conjuction with WAVEOUTCAPS to get the correct device. Check out this[^] page. If you can't resolve this, let me know and I'll have another look tomorrow, bedtime for me here in the UK - good luck!
|
|
|
|
|
Are you using Vista or Windows 7? If so, this[^] may explain why it's not working and a possible workaround?
Edit: Tested demo app and it works!
|
|
|
|
|
Yes, I'm Windows 7.
This app worked. Thank you
Everything makes sense in someone's mind
|
|
|
|
|
If the above doesn't help - a search for Audio Mixer C# yielded this[^] - looks good...
|
|
|
|
|
I saw that but it looked far more comples than I wanted to get into just for setting the master volume. But hey, I might have to go there at some point.
Everything makes sense in someone's mind
|
|
|
|
|
Ok so I seem to have my backgroundWorkers worked out. However I can't figure out how to report progress back to the original form. I'll try to pice together the code to have this make sense. FYI I don't have anything in the progressUpdated even yet.
namespace MailTest
{
public partial class Form1 : Form
{
List<BGW> BGWs = new List<BGW>();
public Form1()
{
InitializeComponent();
for (int i = 0; i < 3; i++)
{
BGWs.Add(new BGW());
BGWs[i].ProgressChanged += BGW_ProgressChanged;
BGWs[i].RunWorkerCompleted += BGW_WorkerCompleted;
}
btnGetMessageInfo.Enabled = false;
}
void BGW_ProgressChanged(object sender, ProgressChangedEventArgs e)
{
BGW bgw = (BGW)sender;
}
void BGW_WorkerCompleted(object sender, RunWorkerCompletedEventArgs e)
{
BGW bgw = (BGW)sender;
}
class BGW : BackgroundWorker
{
public BGW()
{
this.WorkerReportsProgress = true;
this.WorkerSupportsCancellation = true;
}
protected override void OnDoWork(DoWorkEventArgs e)
{
base.OnDoWork(e);
WriteMail.WriteAndParseMail();
}
}
private void btnGetMessageInfo_Click(object sender, EventArgs e)
{
processNumberOfWorkers.howMayWorkers();
MessageBox.Show("There are " + GlobalVars.intHowManyWorkers + " workers needed!");
btnGetMessageInfo.Enabled = false;
backgroundWorker2.RunWorkerAsync();
BGWs[0].RunWorkerAsync();
}
class WriteMail
{
public static void WriteAndParseMail()
{
int intEmail;
bool bolSrvMsg;
string txtLogString = null;
TcpClient tcpClient = new TcpClient();
if (percentComplete <= highestPercentageReached)
{
if (percentComplete == 0)
{
percentComplete = 1;
}
highestPercentageReached = percentComplete + 1;
string strCounter = currentMail + " of " + intEmail;
}
I know I'm probably missing something obvious but still new to backgroundWorkers in C# and not seeing many examples for doing this from a class. When I do it from the main form with a different backgroundWorker it works fine, but when I changed it to this "type" it didn't like the function on the same class so I had to create a new one (which I wanted to do anyways).
Thanks in advance!!!!
|
|
|
|
|
MacRaider4 wrote: I'm probably missing something obvious
For some obscure reason there is a BackgroundWorker.WorkerReportsProgress property, and in good Microsoft tradition its default value is the least useful one.
Luc Pattyn [Forum Guidelines] [My Articles] Nil Volentibus Arduum
Please use <PRE> tags for code snippets, they preserve indentation, improve readability, and make me actually look at the code.
|
|
|
|
|
I thought the only two choices for that are true and false? which I have in :
public BGW()
{
this.WorkerReportsProgress = true;
this.WorkerSupportsCancellation = true;
}
is this where I get a lot of snickering behind my back 
|
|
|
|
|
Sorry, I initially didn't read all your code, I just reported the most likely oversight.
I looked through your code now, and I would say:
1. your ReportProgress method should get called, however it is empty. What makes you say it doesn't work?
2. your Mail stuff is static, having multiple BGW's active in WriteAndParseMail() is not safe, it requires data synchronization.
3. I'm not sure you can have multiple TcpClient instances operating concurrently like that.
4. your percentage stuff is incomplete, so I can't be sure it is correct.
5. for overall progress, I often use a timer (Windows.Forms.Timer) which just gathers the number and updates the GUI at fixed intervals, say once a second. That turns out to be easier than having to check for actual percentage changes (which is however more economical).
Luc Pattyn [Forum Guidelines] [My Articles] Nil Volentibus Arduum
Please use <PRE> tags for code snippets, they preserve indentation, improve readability, and make me actually look at the code.
|
|
|
|
|
When I use the hard coded version I have:
private void backgroundWorker1_ProgressChanged(object sender, ProgressChangedEventArgs e)
{
pgbWrite.Value = e.ProgressPercentage;
lblCurrentCount.Text = e.UserState.ToString();
}
Of which does work (when it's not calling from the class), sorry I should have included that.
What makes me say it doesn't work is when I uncomment that section of code out I get a error saying it's not in the current context.
I have had 2 TcpClients running at the same time, so that doesn't seem to be a issue.
Is this what you mean by percentage stuff being incomplete?
percentComplete = (int)((float)currentMail / (float)intEmail * 100);
if (percentComplete <= highestPercentageReached)
{
if (percentComplete == 0)
{
percentComplete = 1;
}
highestPercentageReached = percentComplete + 1;
string strCounter = currentMail + " of " + intEmail;
BGWs[0].ReportProgress(percentComplete, strCounter);
}
|
|
|
|
|
Easy... Your BGWs are defined in Form1, while you're trying to access them from within WriteMail.
I would suggest adding a "sender" parameter to WriteAndParseMail(), so you can pass the worker into it.
protected override void OnDoWork(DoWorkEventArgs e)
{
base.OnDoWork(e);
WriteMail.WriteAndParseMail(this);
}
And in the worker thread...
highestPercentageReached = percentComplete + 1;
string strCounter = currentMail + " of " + intEmail;
(sender as BackgroundWorker).ReportProgress(percentComplete, strCounter);
|
|
|
|
|
General 
News 
Suggestion 
Question 
Bug 
Answer 
Joke 
Praise 
Rant 
Admin 
Submit an article or tip
