Re: Strong/Shared assemblies - C# Discussion Boards

Something unclear to me about threads...

19-Mar-04 6:09

See the HttpWebRequest.AddRange method documentation in the .NET Framework SDK, which includes an example. The HTTP Range header specifies which byte range you want to get. See RFC 2616, Section 14, Subsection 35[^] for more information on the Range header.

This is assuming that the HTTP daemon supports it. This is only supported in HTTP/1.1. Be sure to catch exceptions when it's not supported, though the server will most likely just hand you the entire file without complaining on a single request.

For multi-threading, I recommend asynchronous calls (see HttpWebRequest.BeginGetResponse) as opposed to handling threads yourself. You can implement a ProgressBar by using HttpWebResponse.GetResponseStream and reading and writing the bytes to a file (in that loop is where you'd increment your progress bar).

Microsoft MVP, Visual C#
My Articles

profoundwhispers19-Mar-04 5:43

profoundwhispers

19-Mar-04 5:43

Quoting from MSDN:

Don't control the execution of worker threads from your main program (using events, for example). Instead, design your program so that worker threads are responsible for waiting until work is available, executing it, and notifying other parts of your program when finished. If your worker threads do not block, consider using thread pool threads. Monitor.PulseAll is useful in situations where worker threads block.

What does that mean? Specifically, what does it mean to say that a thread blocks or does not block?

Thank you.

Sammy

"A good friend, is like a good book: the inside is better than the cover..."

Re: Something unclear to me about threads...

Heath Stewart19-Mar-04 6:01

Re: Strong/Shared assemblies

19-Mar-04 6:01

It means that you have the worker threads query for information, or jobs to complete. They wait (block) on a wait handle that is signaled when work is available. They get the work, perform their operations, then block again waiting for the wait handle to be signaled.

Microsoft MVP, Visual C#
My Articles

Strong/Shared assemblies

Kant19-Mar-04 4:57

Kant

19-Mar-04 4:57

To be placed in the GAC, an assembly must have a strong name. I know both strong-named assemblies and shared assemblies can be placed in GAC.

So what's the use of shared assemblies? How can they be in GAC without strong name?

I'm Confused | :confused:

Promise only what you can do. And then deliver more than what you promised.
This signature was created by "Code Project Quoter".

Heath Stewart19-Mar-04 5:59

Re: Strong/Shared assemblies

19-Mar-04 5:59

You're confusing .NET assemblies with Win32 assemblies. An assembly is just a file containing other information and a manifest of that information (like a Java jar file). .NET assemblies contain embedded resources, modules containing IL, and metadata. Win32 assemblies are still native executables that have a manifest file associated with them and are installed in the %WINDIR%\WinSxS directory in Windows XP and above. They are not installed into the global assembly cache for .NET.

Microsoft MVP, Visual C#
My Articles

Kant19-Mar-04 10:36

Kant

19-Mar-04 10:36

I was talking about this:

Shared Assemblies
Assemblies shared by multiple applications should be installed in a centralized repository called the global assembly cache. .NET clients can access the same copy of the interop assembly, which is signed and installed in the global assembly cache.

Link...[^]

Promise only what you can do. And then deliver more than what you promised.
This signature was created by "Code Project Quoter".

Re: Strong/Shared assemblies

Heath Stewart21-Mar-04 12:52

21-Mar-04 12:52

Then apparently there's no different regarding .NET assemblies. I apologize, but I've never heard them referred to as "shared" assemblies. The terminology makes sense, though. By installing assemblies into the GAC they are shared, but they must still be signed. Assemblies should be signed anyway for any production code. It's not like it's hard. Just run "sn.exe -k KeyFile.snk" to generate a key. Refer to that for all your assemblies for either a product or your organization. Consider it a satefy mechanism to safe-gaurd against spoofying attacks (replacing your assemblies with malicious assemblies - unless they replaced them all).

Microsoft MVP, Visual C#
My Articles

Loading assemblies

Craig G. Wilson19-Mar-04 4:55

Craig G. Wilson

19-Mar-04 4:55

I am having an issue loading an assembly. Two days ago, this exact same code was working. Today, I cannot, for the life of me, get the module to load. Here is the relevant code:

public ArrayList FindModules(string path,Type type)
{
	ArrayList modulesFound = new ArrayList();
	foreach(string file in Directory.GetFiles(path))
	{
		FileInfo fileInfo = new FileInfo(file);
		if (fileInfo.Extension == ".dll")
		{
			Assembly assembly = Assembly.LoadFrom(file);
			foreach(Type t in assembly.GetTypes())
	                {
				if(modulesFound.Contains(file))
				 break;
				foreach(Type iFace in t.GetInterface())
				{
					if(iFace.Equals(IModule))
					{
						modulesFound.Add(file);
						break;
					}
				}
			}
		}
	}
	return modulesFound;
}

The problem is with the if(iFace.Equals(type)) line. It doesn't recognize that iFace is the same as IModule. If the line is changed to iFace.Name == "IModule" , the it loads the module, but returns a null instance.

Any Ideas????

Thanks,
Craig Wilson

Heath Stewart19-Mar-04 5:56

19-Mar-04 5:56

Use iFace.Equals(typeof(IModule)) instead.

Microsoft MVP, Visual C#
My Articles

Craig G. Wilson19-Mar-04 8:03

Craig G. Wilson

19-Mar-04 8:03

Thank you for your response. I must apologize because I typed my code in incorrectly. I am doing just what you said and it still does not work. See the bottom of my original post. That is what confuses me. The names of each type are the same, but the types are not...

Thanks,
Craig Wilson

Craig G. Wilson19-Mar-04 8:34

Craig G. Wilson

19-Mar-04 8:34

I'm not sure what I did, but it works now. Thanks for your help. Somehow, I think it was getting assemblies mixed up or something. Anyhow, Thanks.

Craig Wilson

Heath Stewart21-Mar-04 12:36

combobox databinding from webservice in mobile application

21-Mar-04 12:36

That's what I figured and was about ready to state. Even though a type may have the same namespace and name, a Type is also comprised of assembly information, consisting of the name of the assembly, the culture info, version, and public key token (if strongly named, which assemblies always should be in any production code).

Microsoft MVP, Visual C#
My Articles

Littlefool19-Mar-04 2:43

Littlefool

19-Mar-04 2:43

Hi, I'm developing a smart device application in c# at the moment.
I have made a webservice who access my SQL database and returns a DataSet with data (a name and an ID).

The meaning is that the first combobox will fill with the associations, and by choosing an association, the second combobox will fill with the institutions of that association.

In my webapplication, I have databinded this DataSet white success to my dropdownlist, but in this mobile application, I always have an error. (but finally I also receive the right information, my combobox will fill with the right information)
I've catcht the error in a messagebox so my programme won't stop during running, and I found out that during Page_Load() he gives the error catched in ComoboBox1_IndexeChanged().

By starting the application, I get first a few "error2-FormatException" and than an "error1-ArgumentException". Afther that my combobox is filled with my associations. By selecting another association, I get an "error2-ArgumentException" before my combobox is filled with the institutions of that association. Confused | :confused:

I know it's a long explanation but I've tried everything and all the examples I've found on the internet gives exactly what I have programmed.
Can anyone please help me with it?? D'Oh! | :doh:

Thank you very much Rose | [Rose]

private void Form1_Load(object sender, System.EventArgs e)
{
try
{
Service service = new Service();
DataSet dataset = new DataSet();
dataset = service.getAssociations();
comboBox1.DataSource = dataset.Tables["associations"];
comboBox1.DisplayMember = "cName";
comboBox1.ValueMember = "ID";
comboBox1.DataBindings.Clear();
comboBox1.DataBindings.Add("SelectedValue", dataset, "associations.ID");
}
catch (Exception error)
{
MessageBox.Show(error.ToString(), "error1");
}
}

private void comboBox1_SelectedIndexChanged(object sender, System.EventArgs e)
{
try
{
Service service = new Service();
DataSet dataset = new DataSet();
dataset = service.getInstitutions(Int32.Parse (ComboBox1.SelectedValue.ToString()));
comboBox2.DataSource = dataset.Tables["institutions"];
comboBox2.DisplayMember = "cName";
comboBox2.ValueMember = "ID";
comboBox2.DataBindings.Clear();
comboBox2.DataBindings.Add("SelectedValue", dataset, "institutions.ID");
}
catch (Exception error)
{
MessageBox.Show(error.ToString(), "error2");
}
}

Creating virtual File Item in C#

Tristan Rhodes19-Mar-04 2:17

Tristan Rhodes

19-Mar-04 2:17

How would I go about creating a virtual file item in C#?

This is continuing my interest in developing a drag drop feature. Generaly, files are dropped by handling their HDROP, and therefore, i would need to create a file to do this.

However, for dragging and dropping, the process of creating large files prior to the drag proceess would be horribly costly, so i would like to instead create a virtual file type structure that contains binary serialized data, and can be handled like a system file by OLE methods.

Can this be done?

Cata

Heath Stewart19-Mar-04 4:34

19-Mar-04 4:34

You're answer lies in the documentation for the TYMED data type, which is used in the FORMATETC and STGMEDIUM structures which I mentioned to you months ago when you started this venture. You're only three real options besides creating a physical file to hold the file contents is to lock the file contenxt in global memory (bad for large files), or implement IStream or IStorage to pull the file data when requested.

Say you use an IStream (TYMED_ISTREAM) implementation to "store" your file contents. This is independent of the clipboard format (for which you could still use the CF_HDROP clipboard format). When a drop target says, "hey, I can take a file to be moved or copied", it gets the storage type (TYMED) from the format info (FORMATETC) and gets a pointer to that from the storage medium information (STGMEDIUM). So long as it understands the storage type (which things usually do), it starts making requests on your IStream implementation, which is where you start feeding it information about your file.

That's about as "virtual" as you're going to get, unless you want to go to the practically impossible task of creating a block device in Windows (easy in *nix, NOT easy in Windows).

Microsoft MVP, Visual C#
My Articles

Tristan Rhodes19-Mar-04 6:04

Tristan Rhodes

19-Mar-04 6:04

Ahhhh. I thought i should have done it that way, but it wasn't too sure if it was expecting a stream or an actual file object handle.

So the IStream interface can be dropped to shell, and if i handle it with HDROP, it will construct the file for me?

In which case, i need to set the file name(s) via _DROPFILES. (Where is this object placed? Are there two STGMEDIUM objects, one for the stream, one for the descriptor? I'm a bit fuzzy on that issue, i didn't find the documentation on it that helpful)

I need to implement whatever version of IStream i want to use in a C# based object and define the interface behaviour from the managed end. IStream would in this case tie into a stream object created by serializing an object.

However, to handle this custom IStream, i will also need to build a managed DataObject that implements the IDataObject interface, unless I can get a random IDataObject from the shell and use SetData to set my IStream to the item.

That about right?

Dropping INTO managed code is a piece of cake, shame the reverse is such a pain in the a*se.

Cata

Heath Stewart19-Mar-04 6:19

19-Mar-04 6:19

All the IStream implementation (and it does have to be the actual COM IStream interface, or use the UCOMIStream interface in .NET) has to do is provide an implementation that gets the file bytes. The drop target doesn't care where they come from, as long as the object you pass it supports that interface. Your implementation could simply return the file contents in the IStream::Read method.

And yes, you still need to use the various COM interface for DnD because the similarily-named .NET interfaces don't expose what you need. IIRC, you can still use the managed DoDragDrop but you need to use the COM IDataObject interface.

Use .NET Reflector[^] or ildasm.exe (if you can read IL) to look at how Control.DoDragDrop works. It eventually uses the COM interfaces, otherwise it couldn't communicate with anything that is not a .NET application.

Microsoft MVP, Visual C#
My Articles

Tristan Rhodes19-Mar-04 8:15

Tristan Rhodes

19-Mar-04 8:15

What about the IOLEDataObject?

I have implemented my own managed version, and this one appears to be an unmanaged type that is already registered in the managed DoDragDrop method through casting. If i register my own version of IOLEDragDrop, will it be cast properly by the dodragdrop method? Or do i need to implement a whole new DDD?

Also, where the hell do i declare the name of the file to be constructed? (I think this is the final element of the puzzle, everything else I can do).

As always, thanks for your help.

Cata

Heath Stewart21-Mar-04 12:43