The latest programs use only one process for several application instances. An example is Microsoft Word. When you run it for the first time it creates a process and an empty document. When you run the executable for the second time (while first instance still running) then it doesn’t create a new process. It tells somehow to the first process to create just a new empty document. In this manner the program uses lower resources. How to implement it under .NET?
Is there some reading on how to organize the initial ideas, for a fairly large software (a graphics software, estimated to take about six months). I feel at loss, not having a computer science background.
Just a thought: Look for information on the Microsoft Solution Framework: Envisioning phase / Conceptual Design / Logical Design / Physical Design / Developement / Test / Deploy.
MS Exam 70-300 covers this also. (So one of the study guides for that exam could help out - I do NOT recommend the MS Press book for that exam, it is very dry reading and I felt that my major intestine was about to jump into my throat and strangle me in order to prevent me going insane with boredom)
What exception is being thrown exactly? I'm assuming an ArgumentException based on your description, but please correct me if I'm wrong. If a SocketException is being thrown, it's possible that your router does not support multicasts. You should consult your router documentation (or ask an admin) to find out for sure (although it would be safe to assume that it does).
Finally, for sanity's sake, debug this code and after the ip is assigned by IPAddress.Parse, check the IPAddress.AddressFamily and make sure that it is indeed AddressFamily.InterNetwork. An ArgumentException is most likely thrown when the IPAddress does not belong to the AddressFamily specified in the UdpClient constructor.
Hi Folks, I wonder if somebody can either point me in the right direction or explain a certain concept to me... i just can't figure it out (i've browssed through MSDN)...If in an application I have several threads running and while these threads are running something arrives on the serial port of the computer...Using a control I can raise an event to signal this "BUT" will the application be aware that this event has taken place? Or how can I make sure that the application is aware that something has been recieved at the serial port while other threads are running?.... Please can someone help...
A secondary issue also related ... say again multiple threads are running in an application, in one thread I implement a lock on a certain portion of code (so this must complete execution before any other thread can resume control)... if a timer raises an event (or is due to raise an event) while another thread is executing a portion of locked code... will this timer event be lost by the time the locked code in the other thread completes???
can anyone help!!! I'm desperate... can't find this information anywhere...
maria_p wrote: say again multiple threads are running in an application, in one thread I implement a lock on a certain portion of code (so this must complete execution before any other thread can resume control)...
Now, what I remember from my concurrency lectures (8 to 10 years ago) if you lock a piece of code (the "critical section") other code in other threads can still execute as they are not preventing from executing, they just can't execute the locked code.
Think of the critical section as the centre of a highway intersection with an American 4-way Stop sign. (In America the rule is the first to arrive has right of way and can pass through the intersection). So the first car arrives and is allowed onto the intersection. A fraction of a second later another car arrives from a different direction, it must wait until the first car has passed through. This is locking. Now, just because some traffic has to wait to gain access to this intersection doesn't stop the traffic flowing on an interstate a mile away.
Does this help you understand how concurrency (multithreading) works?
Yes thank you Sir, what you have said and the way you put it has given me a totally different way of looking at the issue (Excellent analogy!)... still a little confused about one issue though.... regarding the event at the serial port....
If various threads are executing in the application will the system get a chance to acknowledge the/an event (which I am expecting at some point) that may take place due to something arriving at the serial port? Or when one thread looses control another thread will immediately take control and due to this the possible event will be missed or seriously delayed? Is there a way around this or am I just missing the point about something?
If you could also explain this I'd be most grateful.
Colin Angus Mackay wrote: Now, what I remember from my concurrency lectures (8 to 10 years ago) if you lock a piece of code (the "critical section") other code in other threads can still execute as they are not preventing from executing, they just can't execute the locked code.
This depends on a number of factors. First, it depends on what you're locking against (if using the lock keyword). As I explained to her before, if you want to lock a block of code for all threads, then a static object (say, the Type of the class that contains the definition) should be used. If she uses an instance variable (such as this), then only the method will be locked for that instance no matter how many threads reference the class.
If you use the lock keyword, a Monitor is used which does block successive calls by different threads. The only way for threads to not wait is to use Monitor.TryEnter or use a Mutex and call WaitOne with a value to set the timeout.
Don't forget about all the articles in the .NET Framework SDK besides those in the class library. There is a pretty big selection of articles about threading in .NET at http://msdn.microsoft.com/library/en-us/cpguide/html/cpconthreading.asp[^] which also includes a lot of examples. Many of the overviews should even discuss threading in general for those not familiar with threads.
I'v been converting some classes I wrote in VB.net to stick them in a C# DLL, but have come accross some wierd irregularities in the .net architecture.
Firstly, the C# compiler will not allow me to explicity pass the value of one numerical variable into an enumerated data type, whereas VB allowed it, and MSDN swears blind it should. For instance:
VB: mnFog(x, y) = Convert.ToByte(sChar) - works fine
C#: mcFog[nLocX, nLocY] = Convert.ToByte(sChar); - Cannot implicitly convert type 'byte' to 'ThisEnum'
The enum in question is derived from a byte.
Secondly, C# just doesnt "see" some methods and properties which VB can use. For instance, in Microsoft.Data.Odbc the OdbcDataReader has an Item collection which allows its user to retrieve a field by its SQL name, instead of its column number (like the .Get[Type]() functions).
See for yourself, reference the DLL, find "Microsoft.Data.Odbc.OdbcDataReader.Item" in the object browser, then try and use it.
Has anyone else experienced these sort of problems?
dalm wrote: Cannot implicitly convert type 'byte' to 'ThisEnum'
Cast your value you pass, so if you have a function sig. which has an enum as a parameter such as this:
int SomeFunc(ThisEnum e);
and you know the enum value, simply cast it when you make the method call:
int a = 2;
dalm wrote: Secondly, C# just doesnt "see" some methods and properties which VB can use. For instance, in Microsoft.Data.Odbc the OdbcDataReader has an Item collection which allows its user to retrieve a field by its SQL name, instead of its column number (like the .Get[Type]() functions).
C# doesn't syntactically reference items like VB.NET does, it uses the  syntax instead.
I have a .aspx.cs file and a group of .cs files I need to compile together. When I try this on Visual Studio .NET, there are no compile issues, but when I try to execute the file, it comes across a literal form control ("error_tags") and gives error "type or namespace cannot be found".
The class in the .aspx.cs files inherits from a base class in the .cs file. The base class contains the definition "public Literal error_tags;" Is the issue that the .cs files are not being compiled?
Previously, when I was compiling my .dlls from the command line and moving them to the bin folder I had no issues.
The <%@ Page%> should inherit from the class defined in the .aspx.cs file, as well as contain the name of the assembly in which that class is contained. If the class in that file inherits from another class defined in another .cs file, that latter class should derive from System.Web.UI.Page.
If you right-click on the files, select Properties, and examine the "Build Action" property, these should say "Compile". This is the default build action for .cs files. If the project compiles all the files set for compilation and there are no build problems, then all your Types are being resolved correctly. The only problem would be that the Inherits attribute of the <%@ Page%> declaration probably isn't referencing the right Type, and / or the CodeBehind attribute does not specify the path to the correct .aspx.cs file.
Finally, make sure that the Liberal error_tags class is declared as public, protected (recommended), internal, or protected internal. If it is declared as private no class - including subclasses - can access it without using reflection (which is expensive and not a very good design).