|
Hello,
I currently have my application, which uses remoting to connect to the server. My problem is that if I change the assembly version on the server, the application will have trouble deserializing and the local copy of the assembly must be updated.
However, if I make very small changes to my app, I will only update the build and revision numbers, and not the major or minor versions. I would like my client and server apps to work if the major.minor version is the same, not taking into account the build.revision part.
I have found the bindingRedirect tag in the application config file, but a) it seems from the docs that when the app looks for a range of versions it will be ok to use a new version, and b) it seems to me that it is for loading assemblies locally, and not from remoting (altough I am not so sure about this.)
[EDIT: I tried bindingRedirect and it works as I expected. I will have to use two bindingRedirects, so that lower versions resolve to my version, and higher versions also resolve to my version. Now I'll have to change app.config for every build. No, that won't do it... and I also tried using an * instead of the revision number -- it didn't work either.]
BTW, my assemblies are strong named.
So, any ideas?
Thanks!!
Luis Alonso Ramos
Intelectix
Chihuahua, Mexico Not much here: My CP Blog!
-- modified at 20:59 Tuesday 27th June, 2006
|
|
|
|
|
Hi,
I am getting the name of the class from database as string. I want to create object for this class dynamically.
how do I do it?
|
|
|
|
|
Create a Type object from the name. Use the GetConstructor method to get a ConstructorInfo object for the constructor. Use the Invoke method to run the constructor.
---
b { font-weight: normal; }
|
|
|
|
|
Please can you show a sample code.
|
|
|
|
|
I explained exactly how to do it. Need I actually do it for you also?
---
b { font-weight: normal; }
|
|
|
|
|
|
That wasnt rude. It was simply "Give it a try yourself".
Teach a man to fish and all that.....
Current blacklist
svmilky - Extremely rude | FeRtoll - Rude personal emails | ironstrike1 - Rude & Obnoxious behaviour
|
|
|
|
|
As well as the answer already provided, which will work no problem, you can also look at the CreateInstance method on the Activator class.
The documentation can be found on msdn.
Current blacklist
svmilky - Extremely rude | FeRtoll - Rude personal emails | ironstrike1 - Rude & Obnoxious behaviour
|
|
|
|
|
Hi,
As well as the class name, you need to know what assembly the class is contained in. There are different ways of doing this, but just lifting a bit of code from work:
Assembly assembly = Assembly.LoadFile(@"c:\myassembly.dll");
object dynamicClass = assembly.CreateInstance(classNameFromDb);
Not sure that helps as you end up with an untyped object. The way we do it here is to make the objects we create dynamically implement an interface, so the second line actually looks more like
IDynamicObject dynamicClass = assembly.CreateInstance(classNameFromDb) as IDynamicObject;
This way, we can then call the methods the interface specifies.
After all, creating an instance of a class is little use if you have no idea what the object does. This is why defining an interface is useful.
And don't forget the gotcha, the trap I always fall in with this, its not just the type name you need, but also the namespace. eg. "MyNamespace.MyClass" which you need to pass to CreateInstance.
Regards,
Rob Philpott.
|
|
|
|
|
Hey Everyone,
I'm writing an app to test whether a serial port on a computer will work correctly and I'm using the principle of serial loopback.
I'm using the SerialPort class from System.io.ports and using a serial port object to write to the port and then listen to see if there was any data recieved. However, I always get a timeout for the recieve portion, even when I put the recieve method into a Thread.
Any ideas what I should try? Thanks in advance
-Karthik
|
|
|
|
|
Karthik Gaekwad wrote: I'm using the principle of serial loopback
I don't think you can use a principle I think you have to actually have a loopback. 
"Just about every question you've asked over the last 3-4 days has been "urgent". Perhaps a little planning would be helpful?"
Colin Angus Mackay in the C# forum
led mike
|
|
|
|
|
led Mike,
You are correct. I am using loopback to solve the problem, but my recieve thread won't fire, even if I have the recieve wait for a message.
Hope that clears it up. Any suggestions?
-Karthik
|
|
|
|
|
And the solution is....
After you write to the port using a serial port object, use the readExisting method. DO NOT CLOSE THE PORT and then OPEN the port again to read. There is nothing stored in the buffer at that point.
I can be so dumb sometimes. That's what I get for trying to modularize my program.
|
|
|
|
|
Nice work, thanks for posting the solution. 
|
|
|
|
|
Hello,
Directory.Exists(“\\localhost\c$”) and Directory.Exists(“\\localhost\SomeLocalShare”) and Directory.Exists(“\\127.0.0.1\c$”) and Directory.Exists(“\\127.0.0.1\SomeLocalShare”) all stop returning true once the network goes down. I know this seems obvious but how is it I can still resolve the local instance of SQL server? My question is how do I resolve a UNC name (assuming it is localhost) if the network is down, disabled, or non existent on Windows Xp pro and Better?
I was thinking of a virtual port maybe or a virtual Nic.
Any help will be very much appreciated
Ronald Hahn, CNT - Computer Engineering Technologist
New Technologies Analyst
HahnTech Affiliated With Code Constructors
Edmonton, Alberta, Canada
Email: rhahn82@telus.net
|
|
|
|
|
Okay, maybe this is really a silly question but why do you need to resolve a UNC to the localhost/local system? I'm trying to figure out what purpose this would serve and for the life of me I don't get it.
Here's why I'm a bit lost: The localhost/local system is a known quantity and therefore you can write your code to use local system drive paths for accessing any information.
Mike Poz
|
|
|
|
|
I’m writing a tablet app that has start up parameters in a Db. Some of the parameters are UNC paths to data libraries that may or may not be local. Some tables will have a constant connection to the LAN, others will be disconnected piroticly. I’ll know in advance the ones that are mobile and set the data stores to be local in the Db. I want to use the same logic in both connected and disconnected environments. (I’m able to assume that the app was started with either a connection to the Db or a copy of the data needed locally) Also how Universal is a naming convention if it Only works in a connected environment. In the real world 25 drives are not enough. I think the whole industry should do away with the idea of local letters and move completely to UNC \\Device\share\path.
One idea I had was to find where windows holds its share info and find the local resource that way. In the db I’d have \\computerName\share so I could parse it out and compare to this list. In computer management Win still know what it is sharing even with no NIC
Ronald Hahn, CNT - Computer Engineering Technologist
New Technologies Analyst
HahnTech Affiliated With Code Constructors
Edmonton, Alberta, Canada
Email: rhahn82@telus.net
|
|
|
|
|
HahnTech wrote: One idea I had was to find where windows holds its share info and find the local resource that way.
That's maintained in the registry under HKLM here:
HKEY_LOCAL_MACHINE\SYSTEM\CurrentControlSet\Services\lanmanserver\Shares
You'll find keys with each share name, and their data value contains permissions and local path to the share.
This assumes your use account has the permissions to at least read this information.
HahnTech wrote: I’ll know in advance the ones that are mobile and set the data stores to be local in the Db.
If you will know in advance what is local and what is remote, then it sounds to me like you need to have some error checking with some "if - else" blocks to deal with when your "directory.exists()" fails to return true.
If(Directory.Exists(\\localhost\c$)
{
Connect(\\localhost\c$\datbase);
}
else
{
//try a possible local path using C:\
or
//pop an error stating that the database is unavailable
}
Either way this is basically an error handling issue where you need to fail gracefully when your "directory.exists" returns false.
Mike Poz
|
|
|
|
|
Thanks for the Reg Keys. I'll look into that solution.
This is not an error handling problem. This is a windows problem with not Fully utilizing UNC. There is no way to convert the parts of a UNC name into a local path. A properly normalized db will have separate columns for device, share, and path (UNC: \\DeviceName\ShareName\Path) so its not as simply as If UNCpath exists Else just use local. That would require restructuring the Db. And as you probably know that’s not always an option.
Ronald Hahn, CNT - Computer Engineering Technologist
New Technologies Analyst
HahnTech Affiliated With Code Constructors
Edmonton, Alberta, Canada
Email: rhahn82@telus.net
|
|
|
|
|
Hi,
Is it posible to implement a class inheritet from eg. int string .. types
thx
|
|
|
|
|
int and string are both sealed. You cannot inherit from a sealed class.
You could try implementing a class to wrap an int or a string though that might meet you needs.
|
|
|
|
|
|
string-> NO
public sealed class String
int (Int32) is a struct. Does not support inheritance
Sincerely,
Elina
Life is great!!!
Enjoy every moment of it! 
|
|
|
|
|
Why on earth would you want to?
Try code model generation tools at BoneSoft.com.
|
|
|
|
|
how can i have ellips form
MD_NADA
|
|
|
|
General 
News 
Suggestion 
Question 
Bug 
Answer 
Joke 
Praise 
Rant 
Admin 
Submit an article or tip
