|
Have a central repository where the application logs in/out when starting/terminating.
E.g. by having a central database or using a web service.
|
|
|
|
|
Thar sounds dangerous: what if a logout gets missed (e.g. client app crashed).
After some of these, the server will report the maximum allowed number of users
has been reached though no one is actually using the app ?? SO there is need for
a way to get out of there, and once that solution is known, the whole thing becomes
pointless.
I prefer a broadcast + timeout solution where a starting app tries to get a response
from it's sibblings on all other machines; only the machines running the app would
repond and it there are too many of those, the most recent client would refuse to continue.
AFAIR FileMaker has used such a scheme, and it did not need anything special set up
on any Firewall or so (I guess they used TCP/IP with one of the ports that are open
by default...).
I havent seen any source code on this though...
Luc Pattyn
|
|
|
|
|
How to implement single sign on in c#?
please help
Thanks & Regards,
Venkat
|
|
|
|
|
You'll need to give a bit more detail as to your intended usage, environment, etc. before that question can be adequately addressed.
However, a simple search of Google yields 816,000 hits. Maybe you could read a little first.
only two letters away from being an asset
|
|
|
|
|
I need to implement single sign on for an intranet application using c#
Thanks & Regards,
Venkat
|
|
|
|
|
I'm curious as to the which method one would choose, and why. My preference for this would be Method B, it seems cleaner, easier to read and runs faster (though moderatly depending on the code inside the block). I also try to avoid conditions like while(true) in multi-threaded situations.
Method A
<br />
int x = 0;<br />
do<br />
{<br />
if( x < some_number )<br />
break;<br />
<br />
...do some work...<br />
<br />
x++;<br />
}while(true);
Method B
<br />
int x = 0;<br />
do<br />
{<br />
...do some work...<br />
}while(x++ < some_number);
only two letters away from being an asset
|
|
|
|
|
Mark Nischalke wrote: I also try to avoid conditions like while(true) in multi-threaded situations.
Interesting. A multi-threaded situation is usually the only time when I'll use a while(true). Typcially it's in a thread procedure that can in theory run forever and is only broken out of when the containing object is disposed.
while(x++ < some_number);
Personally, I don't like this. I try to avoid statements that have side effects. I think I'm not seeing your point, though. The example can easily be rewritten as:
int x = 0;
while(x < some_number)
{
x++;
}
Is there a reason you want to use the do/while loop construct? It can come in handy when you need to run the loop at least once, but for this example, I'm not seeing it.
|
|
|
|
|
Leslie Sanford wrote: it's in a thread procedure that can in theory run forever and is only broken out of when the containing object is disposed.
Certainly. This example though was used in a situation where the method containing it could be accessed by multiple threads but should not be locked in an infinate loop or race condition.
Leslie Sanford wrote: Is there a reason you want to use the do/while loop construct? It can come in handy when you need to run the loop at least once
Nothing significant in the meaning of the code snippet, just the example I was working with needed at least one execution.
only two letters away from being an asset
|
|
|
|
|
I'd always use a for loop, if I am iterating from a known value, to a known value.
Christian Graus - Microsoft MVP - C++
Metal Musings - Rex and my new metal blog
|
|
|
|
|
That's true, forgot that. I meant to imply that some_number was a variable, possibly set in the loop.
Just looking for opinions and ideas to improve my skills.
only two letters away from being an asset
|
|
|
|
|
So, perhaps say some_test instead:
Method A
<br />
do<br />
{<br />
if( some_test )<br />
break;<br />
<br />
...do some work...<br />
<br />
}while(true);<br />
Method B
<br />
do<br />
{<br />
...do some work...<br />
}while( ! some_test ) ;<br />
As long as there's only one test I'd choose B.
A may be better if there are several tests, but I'd make it a while, rather than a do/while.
|
|
|
|
|
Aside from the fact that your two examples are using conditions that are the opposite of each other  , I agree with Christian that a regular for loop would be the obvious choise:
for (int x = 0; x < some_number; x++) {
...dostuff
}
---
Year happy = new Year(2007);
|
|
|
|
|
Hi,
I wanna know how we can read an Excel file and then write it into a table in Access Database.
After that, we export our data into an Excel file.
Please help me, thanks so much
|
|
|
|
|
You want to read an Excel file to import into Access to then export it back into Excel?  What would be the point?
only two letters away from being an asset
|
|
|
|
|
that's right. But i want my program can do that when i click a button. If u know how, share for me, please
|
|
|
|
|
I have to traverse object graphs from a text files that look similar to this simplified example:
ClassA.Method(ClassB.Method(),null,1.7F).Property
I am passed ClassA and ClassB as objects, so have all I need, but...
Without traversing recursively, I don't think I can easily use reflection and guarantee the right members are used. Am I correct in this assumption?
If so, should I consider a runtime compile and invocation of the code snippets instead?
Hope someone can help!
Thanks, MikeS
|
|
|
|
|
|
Thanks for that - I did look at C-Script, which is where the compile thoughts originated.
I guess I wanted feedback on the limitations/simplicity of reflection a text 'dot notaio'n path, compared to the flexibility of compiling the text as code (losing casting etc).
much as I'd love to play with compilation, I've decided to go with reflection and live with it's limitations. I'll keep my restricted syntax as 'valid' c# script/code so I can move to compilation.
The reasons are:
1. To acesss data from the object model in a stuctured way so XSLT can transform it.
2. To hide some complexity from users, so they define business data paths rather than code.
3. To keep the option of using a basic reflection UI to build the references.
4. To ensure reusable features are part of the object model, not in script files.
There is one part I'm thinking of making very un-like c# code for data output reasons: Iterate through collections and invoke the next member in the dot notation path on each item. Given this requirement, I'm investigating if I could use HQL in the file instead...
I'd appreciate any other comments!
Ta, MikeS
|
|
|
|
|
Hey, I'm trying to open notepad from my own application when i click a button.
I already posted this, but the answer i got had a few problems, and there were no other answers. Thanks for the help[^] Luc Pattyn;), but I didn't get it. Well, actually, I think it was my fault it didn't work. (as in i screwed up when modifying it to be used on button_click).
I'd figured I would need a reference, and it says I do ("the type or namespace 'Process' could not be found.( are you missing a using directive or an assembly reference?)") this error occured twice. Any help on correctly doing this or what reference I am missing would be greatly appreciated.
Thanks!
|
|
|
|
|
Hello again,
the Process class is defined in namespace System.Diagnostics
you can verify this in the MSDN help;
it shows "System.Diagnostics.Process" in the section on Inheritance Hierarchy
(at the top of the manual page when using Visual Studio 2003, at the bottom when VS 2005).
This means:
1) you must provide a using statement at the beginning of every source file that wants
to use the Process class, like so:
using System;
using System.Diagnostics;
2) you MAY have to add a reference; you dont if the Intellisense showed the right
stuff while you were typing the using statement (popup list at every period with
the required item in it).
If it does not show, then you must add a reference (right click your project's name
in the Solutions panel).
For System.Diagnostics it is automatic though.
Once you took care of these issues, the Process constructor, methods and other class
members should have full Intellisense support inside Visual Studio, and you are
ready to (iterate on) edit, build and run. Same holds true for every other class
you want to use (of course the most basic ones are always accessible, such as String).
Hope this helps you on the right track.
Luc Pattyn
|
|
|
|
|
OK. Now it doesn't have a problem with the Process being unidentified. But when i run the code the program has a problem with the proc.Start(); part of the code. This happens after i click the source button. It says "Win32Exception was unhandled" and then below that "Access is denied". Thanks for the help!
(And what is a Win32Exception?)
|
|
|
|
|
Hi,
Win32Exception is "The exception that is thrown for a Win32 error code...", see MSDN help
(e.g. inside Visual Studio).
It means something deep down went wrong, a HRESULT error code was generated and rather
than a regular Exception with speciifc meaning, a general Win32Exception was generated
(with a field containing the HRESULT value, in this case, since access violation,
is will end on hex 0005).
In general, it means you try to access something you are not entitled to.
One possible scenario is you want to open a file that is already open somewhere,
or you want to open a file you are not allowed to open at all (someone elses file,
a system file, ...).
I fail to see what the connection is with your Process.Start() stuff, either
it launches Notepad or it doesnt. Anything from there is Notepad's matter
(e.g. if file does not exist or cannot be read, it should tell you).
What makes you say the problem is with Process.Start() ? Did you catch an exception
and got a reference to that specific line ? or did you single-step the code just till
there ? or what.
If it is and you need more help, one would have to see the source code...
Luc Pattyn
|
|
|
|
|
ok. here is the source:
Process openSource = new Process();<br />
openSource.StartInfo.FileName = @"C:\WINDOWS\system32\notepad.exe";<br />
openSource.StartInfo.Arguments = webBrowsing.DocumentText;<br />
openSource.StartInfo.UseShellExecute = true;<br />
openSource.Start();
when i click the button a window pops up saying what i said about the Win32Exception and it hightlights the "openSource.Start();" part of the code.
thanks for the info and help.
|
|
|
|
|
OK, I suggest you make your program show the value and some content of
webBrowsing.DocumentText right before you create the new Process(),
something like:
string fileSpec=webBrowsing.DocumentText;
log("fileSpec="+fileSpec);
try {
using (StreamReader sr=File.OpenText(fileSpec)) {
for (int i=0; i<3; i++) {
string s=sr.ReadLine();
log(s);
}
}
} catch (Exception exc) {
log(exc.Message);
throw;
}
Process openSource=...
in the above replace log() by whatever function can show you a string;
this could be Console.WriteLine()
this will tell us:
- whether the file spec is a valid windows file spec
- whether the file exists
- whether it is really readable
Good luck !
Luc Pattyn
|
|
|
|
|
I'm coming up against cases where the order of events - or even whether events ever fire - depends on whether my app is running "in debugger" (ie, F5 in Visual Studio), or running in it's release version.
So, some of my code needs to be aware of which environment I'm in. Does .NET / VS offer a nice structured way of doing this, or do I need to start messing with process names, etc?
Thanks!
|
|
|
|
General 
News 
Suggestion 
Question 
Bug 
Answer 
Joke 
Praise 
Rant 
Admin 
Submit an article or tip
