|
Hi,
I made an installer using the VS 2005 setup and deployment project. I made a custom installer class and overrode the Install, Uninstall and Rollback methods. I installed my program, which worked, but when I tried to uninstall it, there was an exception. Now I don't know how to uninstall my program. Also, I don't know how to debug the installer.
If anyone could provide some insight, it would be much appreciated.
Thank you very much
choo_chu
|
|
|
|
|
And the exception would be...
|
|
|
|
|
Hi EliottA,
The exception is "can't write to registry key". I forgot to set the boolean value to true to make the registry key writeable. By default it is false...not writeable.
choo_chu
|
|
|
|
|
To clarify, the problem you're running into is that the custom action throws an unhandled exception, the uninstaller catches it and rolls back the uninstallation process. This effectively disallows you from uninstalling your own package. Is this correct?
If so then what the exception is is irrelevant, the important part is that your custom action needs catch EVERYTHING or else you run into the problem you're experiencing. To get your program uninstalled just fix the code and copy the new build of the custom action with super-duper exception handling over the existing file. When the installer is run again hopefully it won't notice the difference between files and this time the unhandled exception won't be unhandled.
If this isn't your problem then this post probably isn't very helpful 
|
|
|
|
|
Hi Jimmanuel,
You identified my problem correctly. I will try what you reccomend.
Thank you,
choo_chu
|
|
|
|
|
Hi Jimmanuel,
Your solution worked! Thanks for your help
choo_chu
|
|
|
|
|
Glad to help
|
|
|
|
|
I've got several xml data access classes. Each is implemented as two partial class files. One generated by the XSD tool (one of the SDK commandline apps), it applies the attribute to each class it generates, and does not appear to have an option to stop it from doing so. The second is code that I've written that provides add/delete/validation type functions, these I'd like to be able to debug normally.
Deleting the attribute out of the autogenerated code manually is a poor solution since it would just come back each time the file was regenerated due to a schema change. I suppose I could find a tool that does an automated search/replace and call it from the batch file I use to regenerate the source files, but that strikes me as being more cumbersome than changing the code I wrote.
It is a truth universally acknowledged that a zombie in possession of brains must be in want of more brains.
-- Pride and Prejudice and Zombies
|
|
|
|
|
I don't think you can stop it, but you could inherit your class from the auto generated dataSet instead of using a partial. That way the debugger will step through the code in the inherited class just fine.
"An eye for an eye only ends up making the whole world blind"
|
|
|
|
|
I suppose that's an option, but not having to do that with generated code was listed as one of the benefits of partial classes when MS came up with the concept in 05.
It is a truth universally acknowledged that a zombie in possession of brains must be in want of more brains.
-- Pride and Prejudice and Zombies
|
|
|
|
|
I am developing an application in C#.NET , for transferring data between two PC's using RS232 protocol. For communication, i m using "SerialPort" class of "System.IO.Ports" namespace. but now i have a problems in recieveing the data and sending the data. even one major problem is whenever i'll recieve at that time i have to identify the begining of message and ending of message as well as if two messagees are recieveing at thae same time then how to differentiate that messagees. please help me i nthat. i really stuck in that. thank you in advance
Arpita Patel
|
|
|
|
|
Hi,
did you mention to use a start and a stop character, e.g. (char)1 start and (char)2 to stop. 
Then use
recvStr = recvStr.Remove(0, recvStr.IndexOf((char)1)-1);
msg = recvStr.Substring(recvStr.IndexOf((char)1), recvStr.IndexOf((char)2, recvStr.IndexOf((char)1));
Kind regards
|
|
|
|
|
I used this code for sending the data
private void SendData()
{
if (CurrentDataMode == DataMode.Text)
{
// Send the user's text straight out the port
comport.Write(txtSendData.Text);
// Show in the terminal window the user's text
Log(LogMsgType.Outgoing, txtSendData.Text + "\n");
}
else
{
try
{
// Convert the user's string of hex digits (ex: B4 CA E2) to a byte array
byte[] data = HexStringToByteArray(txtSendData.Text);
// Send the binary data out the port
comport.Write(data, 0, data.Length);
// Show the hex digits on in the terminal window
Log(LogMsgType.Outgoing, ByteArrayToHexString(data) + "\n");
}
catch (FormatException)
{
// Inform the user if the hex string was not properly formatted
Log(LogMsgType.Error, "Not properly formatted hex string: " + txtSendData.Text + "\n");
}
}
txtSendData.SelectAll();
}
For the Recieving purpose I used
private void port_DataReceived(object sender, SerialDataReceivedEventArgs e)
{
// This method will be called when there is data waiting in the port's buffer
// Determain which mode (string or binary) the user is in
if (CurrentDataMode == DataMode.Text)
{
// Read all the data waiting in the buffer
string data = comport.ReadExisting();
// Display the text to the user in the terminal
Log(LogMsgType.Incoming, data);
}
else
{
// Obtain the number of bytes waiting in the port's buffer
int bytes = comport.BytesToRead;
// Create a byte array buffer to hold the incoming data
byte[] buffer = new byte[bytes];
// Read the data from the port and store it in our buffer
comport.Read(buffer, 0, bytes);
// Show the user the incoming data in hex format
Log(LogMsgType.Incoming, ByteArrayToHexString(buffer));
}
}
now give me appropriate answer for my question.
thanks a lot fro replying me.
Arpita Patel
|
|
|
|
|
To parse individual messages out of a stream of them there needs to be a start message flag that signals the beginning of a message. Just pick a unique character or pattern of characters and put that at the beginning of each message. Pick a different flag and put it at the end to identify where each message stops.
|
|
|
|
|
Hi,
if you can choose the protocol freely, then this would probably be the easiest set-up with reasonable reliability:
- send everything as text using ASCII, i.e. don't use binary data;
- end each message on a reserved character, \n is a good choice; you can achieve this by setting
SerialPort.NewLine to '\n' and using WriteLine()
- create a receiver thread, which basically is an eternal loop performing a ReadLine(); this one will block until a SerialPort.NewLine was received, then return all the data received so far, which would be exactly one message.
WARNING: the receiver thread cannot directly access any GUI Controls, it needs Control.Invoke for that.
Luc Pattyn [Forum Guidelines] [My Articles]
DISCLAIMER: this message may have been modified by others; it may no longer reflect what I intended, and may contain bad advice; use at your own risk and with extreme care.
|
|
|
|
|
I do a lot of RS232 programming for my job. All of our messages usually start with a special character, such as the <stx> hex 0x02 character, followed by a message identifier that we use to let the program know what to do with the message. Then the message is ASCII encoded text followed by a termination character, such as <etx> hex 0x03. When a message is received, the starting character, message identifier, and ending character are stripped from the message which is then passed to whatever part of the application it is intended for. If you are having trouble with receiving and sending data, make sure that you are using a crossover cable <null modem=""> which has the send and receive wires switched at one end of the cable. If you use a straight through cable then the receiving computer will not be getting the data on the receive pin but the transmit pin, which it won't process. By definition, serial communication means that you cannot get two messages at the same time as you are transmitting the bytes in a 'series'. So as long as you are looking for the stop character to terminate a message then the next message should not get jumbled or lost as long as you don't overrun the buffer.
|
|
|
|
|
thanks for replying me
hey you said you do lot of RS232 programming. can you teach me that because I am new in RS232 programming with multithreading and i have to do it in time so I dont have enough time but I will manage if you are ready to teach me.
I really appriciate for that.
I am waiting for your reply.
thanks.
|
|
|
|
|
hey need help for serial port.
please give me your email addres or something so i can send info about it.
please.
thanks
|
|
|
|
|
I don't mind offering assistance on your project if you have specific questions. I understand deadlines and your issue with getting this done in a certain amount of time but for me to 'teach' you about threading and serial port communication is a considerable task. What do you know already? If you are just starting out then I suggest you read the documentation at MSDN on the Threading class and the SerialPort class or check out one of the fine articles here at CP on the subject. That's how I learned how to do a lot of the stuff you are wanting to do. If you have a specific question feel free to contact me (you can click the email link at the bottom of this post) and I will answer the best I can.
|
|
|
|
|
thanks for help.
and I know basic info about threading and serial port. but the main problem is in my project computer has to recieve signals from device in continous manner. Its like embedded tech. so new for that. thats why confused. cant find your email link so have to put on general blog. if you dont mind give me your email id so i can send you my problem.
but thanks a lot.
|
|
|
|
|
This problem is giving me a lot of headaches and can´t find the solution...
I´m reading this article Custom Configuration Sections for Lazy Coders[^]
I´m reading the part titled "The Hello, World ConfigurationSection" , using exactly same code (copy paste directly) and always obtain an error when execute the program (it not returns error/warnings on compiling time).
The error says that can´t load config.SomeSettings from asembly...
My app.config:
<?xml version="1.0" encoding="utf-8" ?><br />
<configuration><br />
<configSections><br />
<section name="SomeSettings" type="config.SomeSettings, config" /><br />
</configSections><br />
<SomeSettings FillColor="LightBlue" TextSize="9.5" FillOpacity="50" /><br />
</configuration>
Proyect --> properties , shows me that the name of my assembly is "config"
I´m totally lost, found more codes similar to the one explained into this article and always, ALWAYS, always obtain the same error.... perhaps someone can give me a hand, thanks
|
|
|
|
|
Just a quick check, but your namespace that the class sits in is definately config right?
Eg, the code should look something like this:
using System;
using System.Configuration;
using System.Collections.Generic;
using System.Linq;
using System.Text;
namespace config
{
public class SomeSettings : ConfigurationSection
{
private SomeSettings() { }
[ConfigurationProperty("FillColor", DefaultValue = "Cyan")]
public System.Drawing.Color FillColor
{
get { return (System.Drawing.Color)this["FillColor"]; }
set { this["FillColor"] = value; }
}
[ConfigurationProperty("TextSize", DefaultValue = "8.5")]
public float TextSize
{
get { return (float)this["TextSize"]; }
set { this["TextSize"] = value; }
}
[ConfigurationProperty("FillOpacity", DefaultValue = "40")]
public byte FillOpacity
{
get { return (byte)this["FillOpacity"]; }
set { this["FillOpacity"] = value; }
}
}
}
|
|
|
|
|
I solved it definetly... i not defined any namespace, so the problem was in App.config...
<?xml version="1.0" encoding="utf-8" ?><br />
<configuration><br />
<configSections><br />
<section name="SomeSettings" type="SomeSettings, config" /><br />
</configSections><br />
<SomeSettings FillColor="LightBlue" TextSize="9.5" FillOpacity="50" /><br />
</configuration>
And all works good
|
|
|
|
|
|
Hi.
I'm having a problem and cannot find the solution. trust me, i've read lots on msdn and google'd, in the wrong places obviously but i hope someone here can help me.
I'm porting an application to compact framework 3 and to reduce client side load i created a webservice for the BL (also, else i would have to copy all code and compile to CF since i cant use "regular" dll's in CF)
this webservice exposes some custom classes and a couple of methods. a simple example would be
(in webservice)
public enum ToStringType {
FirstName,
LastName
}
public class person
{
public string firstname {get;set;}
public string lastname {get;set;}
public ToStringType stringtype {get;set;}
public person(){
stringtype = ToStringType.FirstName ;
}
public override ToString(){
public override string ToString()
{
string tostr = string.Empty;
switch (stringtype )
{
case ToStringType.FirstName :
tostr = FirstName;
break;
case ToStringType.LastName:
tostr = LastName;
break;
}
return tostr;
}
and the ws has a method returning person[]
on the client. when i run the method and get a list of persons
first of all:
myperson.ToString() returns the object class name. not first/last-name
so i add a method:
tostr (){
...
}
with the logic in tostring(),
this works. returns firstname. but
if i (on the client) myperson.stringtype =ToStringType.LastName;
and debug the application, i see that myperson.stringtype IS ToStringType.LastName
but the method returns firstname.
i would really get the tostring to work since i'm adding the person[] to a combobox and i prefer not to use datasource/bindings. (tried those to. same problem, Displaymember has no effect)
how come this split personality disorder in my classes, and why does ToString return the class name instead of my overloaded ToString method?
I hope someone can help me
regards Tomas
There is no spoon
modified on Tuesday, June 30, 2009 10:54 AM
|
|
|
|
General 
News 
Suggestion 
Question 
Bug 
Answer 
Joke 
Praise 
Rant 
Admin 
Submit an article or tip
