|
I figure it's going to have something to do with the initial program that calls it. Something in that program is screwing it up in some major ways.
I ended up trying to run it as an external process. And, it starts to work. It starts the form up, starts loading the database in the background. The events from the loading are firing correctly...and then for no apparent reason, it just stops loading the database. The UI is still functional and still works, except for the parts relying on the parts of the database that were not loaded.
I made a few changes to the code as far as what was output to the StatusLabel so that I could track the progress and it almost seems like a memory issue...as in, the program that called the process set some limits on the memory available to the program. I have 2GB of RAM and the program gets to about 34 MB and then quits the loading. Running the exact same program from a test app runs it fine. The only difference is the program that called it.
I ended up forgoing the whole thing and just loading the database and serializing the object that I created. Doing that and deserializing the object works fine. It just means that I have to implement a user option for updating the object.
The program running it just seems to screw up anything I write for it to do. It's quite frustrating. Anyway, the newest update is supposed to support multi-threading better so maybe with the newer version it will work the way that I want. Until then, a redesign of the process was the only solution.
Thanks for the help.
|
|
|
|
|
It still sounds very mysterious. 34MB isn't really that big, I would suspect something else is going wrong. Could it be you are getting a time-out, maybe on the database connection? Maybe not being the foreground process makes it get less of the CPU cycles (assuming some other process also wants them).
Anyway, thanks for the update, and good luck with it.
|
|
|
|
|
oh...and I appreciate the insights...
|
|
|
|
|
I am having an issue with event 1000 showing up when I start a service I created. It drops in a matter of seconds after starting, and completely ignores exception handling within the app as it does so.
It does not exhibit this behavior on Server 2003, only on 2008. Both are 64-bit implementations.
Any advice would be welcomed. Also, is there a good reference for looking up the exception code listed below? Thank you.
Faulting module name: KERNELBASE.dll, version: 6.1.7600.16385, time stamp: 0x4a5bdfe0
Exception code: 0xebf00baa
Fault offset: 0x000000000000aa7d
Faulting process id: 0xc4c
Faulting application start time: 0x01cadb380757df32
|
|
|
|
|
|
|
Actually, yes. As I recall, it was because I did not have .NET 3.5 enabled on the server, only 2.0. (Unless I am confusing it with a similar issue.)
|
|
|
|
|
Hi,
I'm getting 'Only one usage of each socket address (protocol/network address/port) is normally permitted' exception during invocation of remote object method. I know what causes this problem but I don't know why .
My app registers 2 TcpChannel, ports 9001 and 9002. Ports are used for different remote objects. In such configuration this exception occurs when I'm trying to invoke method of the second object first time. This issue disappears if I use only one TcpChannel with any port number and both remote objects use it.
Thx.
Clarification.
Both remote objects are client and service in the same time i.e. registered as a WellKnownClientType and WellKnownServiceType. Two computers run the same application and talk to each other via remoting. Registration of channels are done only when objects registered as a Service.
modified on Tuesday, April 13, 2010 4:41 PM
|
|
|
|
|
I don't know if this is related but I once had an issue where I was trying to open two TcpChannels and I couldn't because they didn't have unique names. I ended up having to set the name property on the channel like so:
TcpChannel channel;
...
Hashtable properties = new Hashtable();
...
properties.Add("name", "UniqueStringHere");
channel = new TcpChannel(properties, clientProvider, serverProvider);
ChannelServices.RegisterChannel(channel, false);
Good luck!
James
|
|
|
|
|
Does anyone know how to not show weekend dates in NPLOT charts. In Zedgraph one would use something like myPane.XAxis.Type = AxisType.DateAsOrdinal;
costPS.Clear();
CandlePlot cp = new CandlePlot();
cp.DataSource = dt;
cp.AbscissaData = "Date";
cp.OpenData = "Open";
cp.LowData = "Low";
cp.HighData = "High";
cp.CloseData = "Close";
cp.BearishColor = Color.Red;
cp.BullishColor = Color.Green;
cp.Style = CandlePlot.Styles.Filled;
costPS.SmoothingMode = System.Drawing.Drawing2D.SmoothingMode.AntiAlias;
costPS.Add(new Grid());
costPS.Add(cp);
costPS.Title = symbol;
costPS.YAxis1.Label = "Price [$]";
costPS.YAxis1.LabelOffset = 40;
costPS.YAxis1.LabelOffsetAbsolute = true;
costPS.XAxis1.HideTickText = true;
costPS.Padding = 5;
costPS.AddInteraction(new NPlot.Windows.PlotSurface2D.Interactions.HorizontalDrag());
costPS.AddInteraction(new NPlot.Windows.PlotSurface2D.Interactions.VerticalDrag());
costPS.AddInteraction(new NPlot.Windows.PlotSurface2D.Interactions.AxisDrag(false));
costPS.InteractionOccured += new NPlot.Windows.PlotSurface2D.InteractionHandler(costPS_InteractionOccured);
costPS.AddAxesConstraint(new AxesConstraint.AxisPosition(PlotSurface2D.YAxisPosition.Left, 60));
costPS.Refresh();
Hopefully someone has experience with NPlot and can help on the issue. Thanks for any help!
|
|
|
|
|
Part of my application uploads files to an FTP server and reports progress. Currently, the progress is simply calculated by number of files remaining vs. files total in order to generate a percentage and update a progress bar. I want to detail out a little bit more and have a file upload progress bar as well as an overall progress bar.
The entire operation is on a background worker, upload is a function within the process, not on it's own thread. Do I need to employ another background worker in order to update the file progress bar (Which would update at byte chunk intervals) while the main background worker updates the overall progress bar (Which updates on a per file basis)?
Would it be better to calculate the total bytes of all files and then update both progress bars at the same time with the percentage of progress on the overall bar being bytes sent vs. total bytes while the file bar being file bytes sent vs. total file bytes?
I think I answered my own question there. But I would like to know what others think.
The other question is this: I am storing the names of the files to be uploaded in a file queue DB table and accessing it with a data reader. In order to calculate the total size of all files do you think it would be better to store the file size of those files in the table at the time they are inserted and run a simple SUM() operation when I get the data in the first place? Or perhaps write a routine that loops the datareader once and calculates the total file(s) size with an IO operation?
Thank you very much, --EA
|
|
|
|
|
Hi EA,
1.
BackgroundWorker.ReportProgress() takes either an int, or an int and an object of your choice.
So you could pass two ints to the ProgressChanged event, where the second one would get boxed.
Or you could create a class or struct holding whatever you want to pass, and just pass that.
Or you could use one or more class members.
The only thing you must keep in mind is two threads are active here: the BGW making progress, and the GUI thread displaying progress; so you must avoid inconsistent progress data; not by using locks, but in some other way, e.g. by using value types.
An alternative would be this:
pass the progress within a file as the int, and pass the filename as the object;
let the display routine figure out the overall progress, which can be organized to require only a very simple addition and division.
2.
if the file sizes don't change (much), I would consider storing them in the table; getting them from the file system may be too expensive, depends on how many and how far off (network drive, FTP site, ...).
3.
Warning: don't update progress for every tiny amount of data handled, since that might end up slowing it all down. Updating more than say 20 times per second really makes no sense, the human eye won't follow.
Occasionally I update progress by polling rather than pushing it out: just have a timer (a System.Windows.Forms.Timer is the best fit) cause the display to be updated periodically.
|
|
|
|
|
Thank you Luc, you are a well of knowledge. Do you think there is value to creating a nested class vs. an on the fly array?
Example:
protected class myHandler
{
public int fileProgress;
public int overallProgress;
public object Clone();
}
bgWorker.reportProgress(0,myHandler.Clone());
vs.
Int32[] myProgress = new Int32[];
bgWorker.reportProgress(0, myProgress);
I had some ongoing errors with the protected handler class, so right now I am using the on the fly Int array, but if it is just plain wrong...
|
|
|
|
|
If you only have two progress bars, one overall, one for current file, all you need to pass is two percentages. So a simple struct, or an int[2] array would do. However, I would be inclined to make it even simpler:
foreach (string filename in filenames) {
ReportProgress(101, filename);
... some kind of loop {
...
ReportProgress(percentage, filename);
}
ReportProgress(102, filename);
}
and let the ProgressChanged event do the math:
long totalDone;
long overallTotal;
Directory<string filename, long filesize> namesToSizes;
void ProgressChanged(object sender, ProgressChangedEventArgs e) {
int p=e.ProgressPercentage;
string filename=e.UserState;
long filesize=namesToSizes[filename];
if (p==101) {
labelIndividual.Text="now dealing with file "+filename;
} else if (p==102) {
totalDone+=filesize;
} else {
ProgressBarIndividual.Value=p;
ProgressBarTotal.Value=(totalDone+filesize*p/100)*100/overallTotal;
}
}
You have to set up overallTotal and namesToSizes before entering the loop.
Note: While the documentation calls the first parameter "percentProgress", it is not limited to the range [0, 100] so I used two values outside that range to request special actions.
|
|
|
|
|
That is some great information all the way around. Mission accomplished, thanks Luc!
--EA
|
|
|
|
|
Hi all,
I have the following code that I need to return dblRx our of DisplayText in order to use it in another function. Problem I'm having is that if I try to return a double from DisplayText (private double DisplayText), I get the EventHandler error from this.Invoke that I have the wrong data type. Any help would be appreciated.
private void DisplayText(object sender, EventArgs e)
{
RxString = RxString.Remove(0, 3);
double dblRx = double.Parse (RxString);
RxString = dblRx.ToString();
textBox1.Text = RxString;
}
private void serialPort1_DataReceived(object sender, System.IO.Ports.SerialDataReceivedEventArgs e)
{
RxString = "";
try
{
RxString = serialPort1.ReadTo("\r");
serialPort1.WriteLine("*X01\r");
}
catch (Exception Exception)
{
return;
}
this.Invoke(new EventHandler(DisplayText));
}
|
|
|
|
|
Not sure I understand your question very well, however here are some remarks:
1.
mprice214 wrote: RxString.Remove(0, 3); //This removes the first 4 characters
Not true. At best it removes 3 characters; and if there are fewer, expect an exception.
2.
mprice214 wrote: double.Parse (RxString);
that will fail if the data isn't really a string representation of a double. Don't do this on external data, it will bite you. Either add try-catch or better yet use TryParse().
3.
why does DisplayText() have parameters which make it look like a real event handler, you are not using those parameters, instead you are using some class member. It would seem logical to give it one parameter: the text it should display.
4.
mprice214 wrote: serialPort1.WriteLine("*X01\r");
very strange. You want a carriage return plus whatever the system uses for a newline (maybe "\r\n"). Either you like what the system gives you, or you don't; I would never mix the two.
5.
mprice214 wrote: catch (Exception Exception)
{
return;
}
This is not acceptable; whatever goes wrong, you ignore it. So when it goes wrong, you will have an impossible job to diagnose and fix the problem. And believe me, serial communication always goes wrong sooner or later. You should NOT swallow exceptions, either catch a very specific one and add a comment as to why you really don't care about it, or log it somewhere so it leaves a trail.
6.
mprice214 wrote: this.Invoke(new EventHandler(DisplayText));
That does not work, as the parameter list does not match. First make sure which parameters the method should have, then make sure to provide them, using an overload of Invoke().
|
|
|
|
|
Luc Pattyn wrote: maybe "\r\n"
How about using Environment.NewLine?
|
|
|
|
|
If he wants "\r\n" then that is what he should set as SerialPort.NewLine
One should not rely on Environment.NewLine for strings that go beyond the current system (files that get exported, serial communication, etc).
BTW: MSDN is confusing about the default value of SerialPort.NewLine (it says it is "\n" but refers to Environment.NewLine)
|
|
|
|
|
Thanks.
|
|
|
|
|
you're welcome.
|
|
|
|
|
I use it for most things, but when communicating externally I tend to create a constant set to whatever value needs to be defined and use that. I've recently been writing an IMAP thing for work - it's in the RFC that it uses \r\n at the end of every command so that's what I use, just in case it ever ends up being ported to a system that doesn't use it!
DaveIf this helped, please vote & accept answer!
Binging is like googling, it just feels dirtier. (Pete O'Hanlon)
BTW, in software, hope and pray is not a viable strategy. (Luc Pattyn)
|
|
|
|
|
1.
mprice214 wrote:
RxString.Remove(0, 3); //This removes the first 4 characters
Not true. At best it removes 3 characters; and if there are fewer, expect an exception.
Yes, understand. I originally was removing 4.
2.
mprice214 wrote:
double.Parse (RxString);
that will fail if the data isn't really a string representation of a double. Don't do this on external data, it will bite you. Either add try-catch or better yet use TryParse().
Understand this too
3.
why does DisplayText() have parameters which make it look like a real event handler, you are not using those parameters, instead you are using some class member. It would seem logical to give it one parameter: the text it should display.
see #6
4.
mprice214 wrote:
serialPort1.WriteLine("*X01\r");
very strange. You want a carriage return plus whatever the system uses for a newline (maybe "\r\n"). Either you like what the system gives you, or you don't; I would never mix the two.
The device only requires a cr and as result, it is fine with just \r or \r and \n. Why would a newline be necessary here?
5.
mprice214 wrote:
catch (Exception Exception)
{
return;
}
This is not acceptable; whatever goes wrong, you ignore it. So when it goes wrong, you will have an impossible job to diagnose and fix the problem. And believe me, serial communication always goes wrong sooner or later. You should NOT swallow exceptions, either catch a very specific one and add a comment as to why you really don't care about it, or log it somewhere so it leaves a trail.
Understand. I am trying to get things working first and then am going back to clean up. As I'm relatively new to c#, I'm sure I'll miss some things.
6.
mprice214 wrote:
this.Invoke(new EventHandler(DisplayText));
That does not work, as the parameter list does not match. First make sure which parameters the method should have, then make sure to provide them, using an overload of Invoke().
When I run this, it works. Correct me if I am wrong, but doesn't this pass RxSring back to DisplayText? Are you saying there is a better way to do this?
Regarding the initial question, I need to pass RxString out of DisplayText as a double to use for graphing purposes.
|
|
|
|
|
mprice214 wrote: The device only requires a cr
so use serialPort1.Write("*X01\r"); instead of WriteLine() and you're in charge, now it doesn't matter what SerialPort.NewLine contains any more.
mprice214 wrote: get things working first
sure, so am I. By seeing exceptions right away. As a minimum, add Console.WriteLine(exception.ToString());. Either you don't have exceptions, then it makes no difference; or you have, and now you can see them.
mprice214 wrote: it works
Yes, I forgot: Invoke without parameters provides some defaults, and since you don't really use the parameters inside DisplayText, there is no problem. However, I wouldn't do it like that.
mprice214 wrote: Regarding the initial question
OK, if you are not really interested in the string, why not use a class member "double Rx;" instead of "string RxString;". Have DataReceived read the string, convert it to double, and store the value in Rx; then have DisplayText and any other interested method just read the Rx variable.
|
|
|
|
|
Luc Pattyn wrote: Yes, I forgot: Invoke without parameters provides some defaults, and since you don't really use the parameters inside DisplayText, there is no problem. However, I wouldn't do it like that.
Do you have any other suggestions?
Luc Pattyn wrote: OK, if you are not really interested in the string, why not use a class member "double Rx;" instead of "string RxString;". Have DataReceived read the string, convert it to double, and store the value in Rx; then have DisplayText and any other interested method just read the Rx variable.
Thank you for this and also the the comment on the exceptions!
|
|
|
|
|
General 
News 
Suggestion 
Question 
Bug 
Answer 
Joke 
Praise 
Rant 
Admin 
Submit an article or tip
