|
Diego F. wrote: The idea is having a loop that calls the ThreadPool to execute some method, and then block that main thread until all threads are finished.
Then why would you even use threading at all??
The whole point behind threading is to move work to a seperate thread so your primary thread can continue to respond to requests.
The thread pool doesn't raise any events when a thread is complete. Your threaded code has to supply that, or some other method to signal completion.
|
|
|
|
|
Well, in my case, I don't mind to block the main thread. The help of ThreadPool is that I need to run a method many times (let's say hundreds), and if I can run it separately, I would think that it saves time.
Am I wrong? Other way, you are rigth. It has no sense the use of threading.
Regards,
Diego F.
|
|
|
|
|
Hi,
AFAIK there is no built-in mechanism for this.
I see at least three possibilities:
1.
You could try to count yourself; use Interlocked.Increment each time you send
a piece of work to a thread, and have it call Interlocked.Decrement when it is done.
Then have your main thread do a busy wait (with Sleep!) on the same counter.
disadvantages:
- if a thread exits abnormally (or forgets to decrement) your main
thread will hang forever.
- the wait loop is not nice.
2.
Since .NET 2.0 there is BackgroundWorker class, which supports cancellation and
offers events. You could use its IsBusy property and/or its ProgressChanged event.
3.
And of course you could use regular threads, keep them in a list, and periodically
check their status. (Dont forget to clear the list when you conclude all has been
done; so the threads get garbage collected).
|
|
|
|
|
Thank you for your ideas.
What method do you think is better when there are many threads?
Regards,
Diego F.
|
|
|
|
|
I am not sure:
- I have not used BackgroundWorker yet;
- I have limited experience with ThreadPool; IMHO it lacks some functionality.
- most often I use regular threads, so I can easily manage them (e.g. change their
priority).
Diego F. wrote: when there are many threads?
what is many ?
It is my view that executing the same work in more than a couple of threads does
not improve performance: all these threads start to compete with one another
(they all want CPU cycles, or all want disk access, or...). So having 2 or 3 or 4
seems a maximum.
The one good reason I can come up with for really many similar threads is for
encapsulating independent jobs (as would be the case in a server with many clients).
BTW: when using real Threads, you can also do the following:
- for each new thread you create, add it to a list
remark: you probably want to set IsBackground true
- create a loop over that list, performing a Thread.Join();
when that loop finishes, all threads are done (does not matter in what order!).
advantage: no need to count
remark: clear the list when done
- you might even create an extra thread to run that loop, so it does the waitforall
logic, while your main thread can go on doing whatever it likes to do...
Hope this helps.
|
|
|
|
|
Would this post[^] also be by your hand or is this person just copying and pasting your words?
|
|
|
|
|
Hi, I am trying to write a program in C# that will ask for a password before running any program. Suppose if any one wants to run MS Word he has to enter the password. Now my question is how can I set the password check before executing a program?
Please let me know if any of know how.
Johnny
|
|
|
|
|
That's not possible.
Well, it is, but only if you have kernel development experience and know how to use a technique called "DLL Injection" to hook into the kernel and intercept the calls to CreateProcess. You also can NOT do this in and managed code language, like C# or VB.NET. You'd have to write this hook in C++ only.
This hook would have to have access to a database that contains the passwords for various processes. It would also have to have information on the dozens of other processes to NOT ask for a password for because these processes get launched spontaneously in the normal course of operation of Windows.
You would also need a manager application for this database so you can add/remove/change applications and their passwords.
The fact that you even asked this question shows that you don't have the requisite knowledge of Windows internals to pull this off.
|
|
|
|
|
Dave Kreskowiak wrote:
Well, it is, but only if you have kernel development experience and know how to use a technique called "DLL Injection" to hook into the kernel and intercept the calls to CreateProcess. You also can NOT do this in and managed code language, like C# or VB.NET. You'd have to write this hook in C++ only.
Is this going to remain a viable option going into Win64? I know that MS has sharply reduced the ability to hook into the kernel, but don't have the kernel specific knowledge needed to know if this is an affected area or not.
--
You have to explain to them [VB coders] what you mean by "typed". their first response is likely to be something like, "Of course my code is typed. Do you think i magically project it onto the screen with the power of my mind?" --- John Simmons / outlaw programmer
|
|
|
|
|
dan neely wrote: Is this going to remain a viable option going into Win64? I know that MS has sharply reduced the ability to hook into the kernel, but don't have the kernel specific knowledge needed to know if this is an affected area or not.
Hell, I have no idea!  I'm really no expert on the kernel in's-'n-out's. I've just seen what it takes to hook API functions, I haven't actually done it myself.
|
|
|
|
|
That is quiet hard to achieve. This basically means that you have to override the Windows execute functions. When you click on a program binary ( exe ) Windows initiates the execution of the program and executes the code in that binary. There is no function that can neither interrupt the execution nor interrupt the executed code in the binary.
A possible way around it , is to disable all binaries, or set the permission of the binaries ( eg. winword.exe ) so that only a particular user is allowed to execute it. Then write yourself a little wrapper that is allowed to execute the binaries after you prompted for a password.
Cheers,
|
|
|
|
|
the site appears down at the moment, but the secure process utility from Stardock's Thinkdesk Suite[^] might be able to do what you want.
--
You have to explain to them [VB coders] what you mean by "typed". their first response is likely to be something like, "Of course my code is typed. Do you think i magically project it onto the screen with the power of my mind?" --- John Simmons / outlaw programmer
|
|
|
|
|
Dear all,
I'm trying to insert a double value in a database. To my understanding a double value can exist of eg. 23.45
When I insert a double value ( 23.45 ) into my postgres database, the seperator dissapears and the value in the table becomes 2345. The table in the database is of type Double Precision.
I have to use Double, instead of decimal since a library is demanding a double value as input.
Can anyone point me in the right direction ? How can I store the value so that the dot is there as well. ( 23.45 )
Kind regards,
|
|
|
|
|
Without seeing some actual code, it's hard to diagnose your problem. The only thing I can think of is that maybe there are localization settings coming into play, which would require you to use a comma (,) instead of a decimal point (.)
I would suggest that to avoid this, you use parameterized queries instead of just using string parsing to insert the value.
|
|
|
|
|
Hi Ian,
Thank you even giving it a go
This is basically what I am doing :
private void btnRunwaySet_Click(object sender, EventArgs e)<br />
{<br />
<br />
<br />
RegistryKey key = Registry.CurrentUser.OpenSubKey("Software\\TestApp\\DBSettings", false);<br />
if (key == null)<br />
{<br />
key = Registry.CurrentUser.CreateSubKey("Software\\TestApp\\DBSettings");<br />
}<br />
else<br />
{<br />
object db;<br />
db = key.GetValue("name");<br />
lbl.Text = db.ToString();<br />
}<br />
<br />
<br />
Double a = Convert.ToDouble(textbox.Text);<br />
<br />
<br />
NpgsqlConnection conn = new NpgsqlConnection("Server=" + txtDBip.Text + ";" + "Port=" + txtDBPort.Text + ";" + "User Id=" + txtDBuname.Text + ";" + "Password=" + txtDBPasswd.Text + ";" + "Database=jordan;");<br />
conn.Open();<br />
NpgsqlCommand command = new NpgsqlCommand("INSERT INTO (table(column)) VALUES (" + a + ") WHERE (anothercolumn) = '" + lbl.Text"'", conn);<br />
command.ExecuteNonQuery();<br />
conn.Close();
The column of the table is of type double precision to avoid an invalid cast exception. Now, any value given in textbox will be converted to Double and then inserted in the database. I suspect something is going wrong during the conversion from String to Double.
If I insert the Double value eg. 887.23 into the textbox, the string value 887.23 is converted to Double. The funny thing is, that when I covert it right back to string :
anothertextbox.Text = a.ToString();
or
anothertextbox.Text = Covert.ToString(a);
the values that shows up in anothertextbox is 88723. So I assume this is more of a math thing then a database error. The database is simply accepting and inserting the presented Double value.
I'm quiet puzzled actually.
|
|
|
|
|
Ahhh, ok, I see. Simple problem. Convert.ToDouble is meant to be used to convert from other numeric types, or from a number boxed in an object. It's not a string parser.
What you're looking for is double.TryParse()
Double a;
if (!double.TryParse(textbox.Text, out a))
throw new ArgumentException("Type a number");
If you check the API for TryParse, you'll see it tries to convert a string to a double, placing the result in the second argument. It returns true if it succeeds, so if it comes back as false, the person typed something non-numeric.
Convert.ToDouble was probably just grabbing every digit character and stuffing them together.
|
|
|
|
|
Ian,
Thank you for your reply. it works. I implemented your solution and it all runs smooth now. Thanks a lot !!
However, your comment about Convert.ToDouble not being a string parser pursueded me to go down a different path. I tricked the database by presenting it with a string value for insertion into a double column. I was expecting a Cast Exception, but it went in happily, as long as the string value is in the double form, eg. 88,543. Any other form will be rejected. I'm just wondering if this a bug in Postgres or that the database is unable to distinguish a string value ( written as double ) from a true double value.
After I grabbed the string value back from the database, I converted it into a double :
Double aa;<br />
aa = (Double)cmd1.ExecuteScalar();
And the value was in the correct format ( 88,543 ).
However, I prefer your solution, and I'm using it.
|
|
|
|
|
|
Ankit Aneja wrote: if(filename.indexof("FIX")!=-1)
if(filename.indexof("FIX")!=0)
Now it is Possible.
Regards,
Satips. ![Rose | [Rose]](https://www.codeproject.com/script/Forums/Images/rose.gif)
|
|
|
|
|
Have to tried the possibility of small / caps change.?
Rate This Message.
Thank You,
A.Surenth.
|
|
|
|
|
Hi,
I have a situation where I have to have a check box inside a Data grid.
And the datagrid will be filled by set of rows from a database. selecting a check box should enble me to find the exact row in the databse.
I tried a number of articles. But nothing could do what exactly I want.
Can any one help me out.
Environment: VS2003.
Thanks in advance
Rate This Message.
A.Surenth.
|
|
|
|
|
Anthony Surendran wrote: selecting a check box should enble me to find the exact row in the databse.
This is easiest when your table has a primary key defined.
Make sure the data that populates the grid also has the primary key returned. You can bind the primary key to a hidden column in the grid and use that as a future reference to find the corresponding row in the table.
SG
|
|
|
|
|
hellow friends....
i have developed ASP.NET Services and i want to use this services in my VC++ client but i m not getting result although i m able call method correctly using SOAP .. but result is not proper.?
plz tell me where i m making mistake...
//ASP .NET services..
using System;
using System.Web;
using System.Web.Services;
using System.Web.Services.Protocols;
[WebService(Namespace = "http:/www.WbService07Name.com/webServices/")]
[WebServiceBinding(ConformsTo = WsiProfiles.BasicProfile1_1)]
public class ClsService07 : System.Web.Services.WebService
{
public ClsService07()
{
//Uncomment the following line if using designed components
//InitializeComponent();
}
[WebMethod]
public double DoAdditon(double a, double b)
{
return (a + b);
}
}
///vc++ client
void DisplayAdd()
{
try
{
ISoapSerializerPtr Serializer;
ISoapReaderPtr Reader;
ISoapConnectorPtr Connector;
cout<<"test display:main\n";
// Connect to the service
Connector.CreateInstance(__uuidof(HttpConnector30));
Connector->Property["EndPointURL"] ="http://localhost:2014/WebService07/Service.asmx";
Connector->Connect();
Connector-> Property["SoapAction"] = "http:/www.WbService07Name.com/webServices/DoAdditon";
// Begin message
Connector->BeginMessage();
// Create the SoapSerializer
Serializer.CreateInstance(__uuidof(SoapSerializer30));
// Connect the serializer to the input stream of the connector
Serializer->Init(_variant_t((IUnknown*)Connector->InputStream));
// Build the SOAP Message
Serializer->StartEnvelope("","","");
Serializer->StartBody("");
Serializer->StartElement("DoAdditon",L"http://localhost:2014/WebService07/Service.asmx/","","");
//Serializer->startElement("Add","uri:Calc","","m");
Serializer->StartElement("a","","","");
Serializer->WriteString("5.9");
Serializer->EndElement();
Serializer->StartElement("b","","","");
Serializer->WriteString("10.8");
Serializer->EndElement();
Serializer->EndElement();
Serializer->EndBody();
Serializer->EndEnvelope();
// Send the message to the web service
Connector->EndMessage();
// Let us read the response
HRESULT hr = Reader.CreateInstance(__uuidof(SoapReader30));
// Connect the reader to the output stream of the connector
Reader->Load(_variant_t((IUnknown*)Connector->OutputStream), "");
printf("Answer: %s\n",(const char*)Reader->RpcResult->text);
BSTR br;
Reader->RpcResult->get_text(&br);
cout<
|
|
|
|
|
I have an app.config file that looks like this:
<configuration>
<appsettings file="ClassLibrary1.dll.config">
<appsettings file="ClassLibrary2.dll.config">
This use to work fine in VS 2003, when I migrated to VS 2005 I started getting this error :
" Configuration system failed to initialize"
"Sections must only appear once per config file"
Anyone knows how can I fix this???
Thanks in advance.
|
|
|
|
|
ttohme wrote: " Configuration system failed to initialize"
when I changed the settings available in My.Settings - what I found in my case was that one of the changes I had made had added a StringCollection and it needed a try-catch block to create a new instance the first time around if it didn't exist.
That, plus I had corrupted the user.config file.... when I deleted that it wrote a new one and started working fine.
ttohme wrote: "Sections must only appear once per config file"
To get your code to compile without the warnings, you'll need to replace your call to
System.Configuration.ConfigurationSettings.GetConfig("systems")
with
System.Configuration.ConfigurationManager.GetSection("systems")
This new type lives in the System.Configuration.dll assembly so you'll also need to add a reference to this assembly to your project.
To migrate your v1.1 configuration section handler to v2.0, you'll need to re-write your sectionHandler using the new Configuration system base classes which live in the System.Configuration namespace in the System.Configuration.dll assembly. You'll need to use the following classes: ConfigurationSection, ConfigurationElement and ConfigurationElementCollection.
Here's a pretty good link that blogs how to create your own sectionHandler:
Get it[^]
Regards,
Satips. ![Rose | [Rose]](https://codeproject.global.ssl.fastly.net/script/Forums/Images/rose.gif)
|
|
|
|
General 
News 
Suggestion 
Question 
Bug 
Answer 
Joke 
Praise 
Rant 
Admin 
Submit an article or tip
