|
Hi,
Please help me with this weird problem:
- I have a C# application that works fine from Visual Studio (both Debug & Release).
- Works fine on some client machines.
- But hangs at a point (where a VB DLL is called) on some some machines.
- I was trying to research and made some changes like 'Enable ClickOnce' and removed all references to external dlls and re-built the project again.
. Now it still works from Visual Studio
. but hangs when installed and run from Program Files on my machine as well.
Basically, I was able to reproduce what is happening on some client machines. But I'm lost as to what is making it to wait. And how to fix it. The VB dlls are from legacy applications and I'm referencing them by browsing and copying to local.
I'm using XP, 32bit, Visual Studio 2005.
Thanks!
Ap
|
|
|
|
|
If they are legacy VB DLLs, then they are going to be COM DLLs. Issues that can arise from this is that you need to register the DLL on the client machine (and then you can have other issues relating to permissions, DCOM settings and the likes).
|
|
|
|
|
|
Thank you from your responses.
Ok, here is what I did, and found out, but problem still exists..
I dug into the source code of the dlls which basically have some computations involving heavy mathematical calculations + condition checking (to the precision) and iterating
- something that tries to converge and interpolates...
I made them into .NET dlls re-built them with my project (adding some debug statements & logging some output). Now I see that:
1. When I run it from 'Visual Studio', it calculates well and returns with results.
2. While from 'Program Files', it is running in a infinite loop (which I assume needs an additional check and needs to break - I can fix that!).
But my question is:
- with the same input data, how can the calculated results be different and follow two different paths altogether?
- Does VS use a different math library?
- Now, to me it is a bigger problem. Because users are going to run from 'Program Files' and I'm happy with the VS results??
I'm lost!! Please help!!
|
|
|
|
|
Hi,
I have a program for work that uses a formula to calculate the refurb on a unit(parts replaced on cableboxes that were damaged) divided by total units(cableboxes that went through refurb, but did not have any parts replaced). I looked up casting online, and the format for it is intvaluetoconvert = Convert.toint32; I'm doing that , but still get the error Cannot implicitly convert type 'double to int. An explicit conversion exists(are you missing a cast?) What am I doing wrong? Can someone please help? Thank you.
Justin
Here's some of my code to look at?
private int GetRefurbRate()
{
string sql = "";
double Refurb_Rate;
int totalRefurb = 0;
int totalUnits = 0;
string error_msg = "";
sql = "SELECT COUNT(rp.repair_ord) " +
"FROM " + schema + ".repair_part rp " +
"WHERE rp.repair_ord = '" + repair_ord + "' ";
while (true)
{
if (!myDb.RunSql(sql, true))
{
error_msg = "DBError for getting Refurb Rate";
break;
}
if (myDb.dbRdr.HasRows)
{
if (myDb.dbRdr.Read())
{
try //Try and Catch are here b/c I originally had everything ints, and and they just caught the 0 exception.
{
Refurb_Rate = Convert.ToInt32( totalRefurb / totalUnits * 100); //This is where I try to perform the cast.
}
catch (Exception e)
{
Console.WriteLine(e);
}
}
//int Refurb_Rate = Convert.ToInt32(Refurb_Rate);
}
break;
}
myDb.dbRdr.Close();
if (error_msg != String.Empty)
{
MessageBox.Show(error_msg, "Get Refurb Rate",
MessageBoxButtons.OK, MessageBoxIcon.Asterisk);
}
return Refurb_Rate;
}
|
|
|
|
|
You're returning a double. Try converting that to an Int32.
Justiin1265 wrote: try //Try and Catch are here b/c I originally had everything ints, and and they just caught the 0 exception.
How about "If (totalUnits > 0)", since you're not allowed to divide by nothing?
I are Troll 
|
|
|
|
|
That's what I am cofused on. I thought I was inserting the formula into the conversion statement. What is the correct way, b/c the way I am trying to do it gives me an error?
Justin
|
|
|
|
|
You can check whether totalUnits is zero, before attempting a divide;
Refurb_Rate = 0;
if (totalUnits > 0)
Refurb_Rate = Convert.ToInt32( totalRefurb / totalUnits * 100);
You're returning a double, called Refurb_Rate, where the method is expected to return an int, on the last line;
return Convert.ToInt32(Refurb_Rate);
Hope this helps 
I are Troll
|
|
|
|
|
|
I know I'm sorry, someone told me about it last time. I poseted this in a rush and forgot.
Justin
|
|
|
|
|
Pro-tip: you can still edit your question to fix it.
This is not the age of reason, this is the age of flummery, and the day of the devious approach. Reason’s gone into the backrooms where it works to devise means by which people can be induced to emote in the desired direction.
|
|
|
|
|
On which line does the compiler show the error? My guess is it goes wrong at
return Refurb_Rate; as the method return type is Int32.
If you were to change the declaration of Refurb_Rate from Double to Int32 then the problem would go away.
I'm not sure what you are trying to do in the line
Refurb_Rate = Convert.ToInt32( totalRefurb / totalUnits * 100);
It doesn't make much sense as the result of the expression on the RHS is already Int32 making the explicit conversion redundant.
Alan.
|
|
|
|
|
it will also result in a zero value most of the time.
Luc Pattyn [Forum Guidelines] [My Articles] Nil Volentibus Arduum
Please use <PRE> tags for code snippets, they preserve indentation, improve readability, and make me actually look at the code.
|
|
|
|
|
It shows the error at Refurb_Rate = Convert.ToInt32( totalRefurb / totalUnits * 100); I'm trying to convert the doulbe back to an integer, so I figured use the conversion syntax to return to an integer.
|
|
|
|
|
Refurb_Rate is a double, so when you return it later on, it is still holding a double (even though that double may happen to be integral). Since your function is of return type "int", you need to ensure you return data type "int".
Luc addresses some other issues with your code below.
This is not the age of reason, this is the age of flummery, and the day of the devious approach. Reason’s gone into the backrooms where it works to devise means by which people can be induced to emote in the desired direction.
|
|
|
|
|
I have a program for work that uses a formula to calculate the refurb on a unit (parts replaced on cableboxes that were damaged) divided by total units(cableboxes that went through refurb, but did not have any parts replaced). I looked up casting online, and the format for it is intvaluetoconvert = Convert.toint32; I'm doing that, but still get the error "Cannot implicitly convert type double to int. An explicit conversion exists (are you missing a cast?)" What am I doing wrong? Here's some of my code to look at:
private int GetRefurbRate()
{
string sql = "";
double Refurb_Rate;
int totalRefurb = 0;
int totalUnits = 0;
string error_msg = "";
sql = "SELECT COUNT(rp.repair_ord) " +
"FROM " + schema + ".repair_part rp " +
"WHERE rp.repair_ord = '" + repair_ord + "' ";
while (true)
{
if (!myDb.RunSql(sql, true))
{
error_msg = "DBError for getting Refurb Rate";
break;
}
if (myDb.dbRdr.HasRows)
{
if (myDb.dbRdr.Read())
{
try
{
Refurb_Rate = Convert.ToInt32( totalRefurb / totalUnits * 100);
}
catch (Exception e)
{
Console.WriteLine(e);
}
}
}
break;
}
myDb.dbRdr.Close();
if (error_msg != String.Empty)
{
MessageBox.Show(error_msg, "Get Refurb Rate",
MessageBoxButtons.OK, MessageBoxIcon.Asterisk);
}
return Refurb_Rate;
}
This is not the age of reason, this is the age of flummery, and the day of the devious approach. Reason’s gone into the backrooms where it works to devise means by which people can be induced to emote in the desired direction.
|
|
|
|
|
Congrats on the PRE tags, however the percentage formula isn't quite OK... see my other msg.
Luc Pattyn [Forum Guidelines] [My Articles] Nil Volentibus Arduum
Please use <PRE> tags for code snippets, they preserve indentation, improve readability, and make me actually look at the code.
|
|
|
|
|
Justiin1265 wrote: Refurb_Rate = Convert.ToInt32( totalRefurb / totalUnits * 100); //This is where I try to perform the cast.
this does not make sense, as all variables are declared integers, therefore the quotient will be integer too, and the Convert method does nothing for you.
Furthermore, the result will not be what you expect:
for totalRefurb=2, totalUnits=10, the formula looks like 2 / 10 * 100, unfortunately in an integer world 2 / 10 is zero.
What you need to do is one of the following:
- multiply first, i.e. 100 * totalRefurb / totalUnits
- use real numbers (i.e. data types such as float or double), this will avoid overflow (when your integers near the max value), and perform rounding.
Luc Pattyn [Forum Guidelines] [My Articles] Nil Volentibus Arduum
Please use <PRE> tags for code snippets, they preserve indentation, improve readability, and make me actually look at the code. modified on Monday, April 11, 2011 3:25 PM
|
|
|
|
|
I deleted my crap responses. Yours is more... what is the word? Oh yes, "correct".
This is not the age of reason, this is the age of flummery, and the day of the devious approach. Reason’s gone into the backrooms where it works to devise means by which people can be induced to emote in the desired direction.
|
|
|
|
|
"More correct"? A purist might object, some characteristics really are binary. 
I'm not saying it is correct now (I still don't get what the OP is after), I do think one more problem has been addressed though.
Luc Pattyn [Forum Guidelines] [My Articles] Nil Volentibus Arduum
Please use <PRE> tags for code snippets, they preserve indentation, improve readability, and make me actually look at the code.
|
|
|
|
|
Hey now, neither float nor double can represent all reals. They can only represent (some) rationals (and -0, NaN, +/-Infinity).
|
|
|
|
|
David1987 wrote: neither float nor double can represent all reals
I agree (and I'll add nothing can represent just any real number's value); I also didn't say such thing.
Luc Pattyn [Forum Guidelines] [My Articles] Nil Volentibus Arduum
Please use <PRE> tags for code snippets, they preserve indentation, improve readability, and make me actually look at the code.
|
|
|
|
|
No.. but you said "real numbers" and "float" in the same sentence, it might give people the wrong idea. I've known plenty of people who were told by a teacher that doubles are "real numbers" (which is almost true if it's put like that - except for NaN and infinities and -0) and they were all confused and surprised when their calculations gave the "wrong" results because they expected all reals to be representable by doubles.
|
|
|
|
|
Yes, it would have been better to put float and real in code-style. I'll fix that.
I use "real numbers" for non-integer numbers (or better yet, numbers not restricted to integer values), and not for "exact numbers". Most programmers do it that way; it may originate in the old days of Fortran, where they had type REAL (there wasn't a FLOAT time then).
And calculus (as well as programming) courses should teach people about number representations, errors, accumulation and propagation of errors, etc. Most programmers understand the square root of two isn't exact inside a computer, yet many have a hard time understanding why one third times three isn't always one.
Luc Pattyn [Forum Guidelines] [My Articles] Nil Volentibus Arduum
Please use <PRE> tags for code snippets, they preserve indentation, improve readability, and make me actually look at the code.
|
|
|
|
|
Give this a try this should fix your problem;
return (int)Refurb_Rate;
|
|
|
|
|
General 
News 
Suggestion 
Question 
Bug 
Answer 
Joke 
Praise 
Rant 
Admin 
Submit an article or tip
