long, not too long

[Ryan Liu]

In Visual Studio 2003, when i debug a C# program, I found:

300L * 10000000l -1294967296 int
210L * 10000000L 2100000000 int
How come? a long type can not handle 300L * 10000000l, the result is too
big for long type?

Thanks!
Ryan

[Jon Skeet [C# MVP]]

Ryan Liu <ad********@onl ine.sh.cn> wrote:

In Visual Studio 2003, when i debug a C# program, I found:

300L * 10000000l -1294967296 int
210L * 10000000L 2100000000 int
How come? a long type can not handle 300L * 10000000l, the result is too
big for long type?

It's a debugger issue, that's all. It works fine in normal code.

--
Jon Skeet - <sk***@pobox.co m>
http://www.pobox.com/~skeet Blog: http://www.msmvps.com/jon.skeet
If replying to the group, please do not mail me too

[Ryan Liu]

I guess not quite so. I found this problem when my code does not work, then
I begin to debug and find this issue.

Regards,
"Jon Skeet [C# MVP]" <sk***@pobox.co m> ????
news:MP******** *************** *@msnews.micros oft.com...

Ryan Liu <ad********@onl ine.sh.cn> wrote:

In Visual Studio 2003, when i debug a C# program, I found:

300L * 10000000l -1294967296 int
210L * 10000000L 2100000000 int
How come? a long type can not handle 300L * 10000000l, the result is too big for long type?

It's a debugger issue, that's all. It works fine in normal code.

--
Jon Skeet - <sk***@pobox.co m>
http://www.pobox.com/~skeet Blog: http://www.msmvps.com/jon.skeet
If replying to the group, please do not mail me too

[Bill Butler]

"Ryan Liu" <ad********@onl ine.sh.cn> wrote in message news:%2******** *******@TK2MSFT NGP12.phx.gbl.. .

I guess not quite so. I found this problem when my code does not work, then
I begin to debug and find this issue.
Are you by any chance stuffing it into an int instead of a long?
Why don't you post the offending code.

Bill


"Jon Skeet [C# MVP]" <sk***@pobox.co m> ????
news:MP******** *************** *@msnews.micros oft.com...

Ryan Liu <ad********@onl ine.sh.cn> wrote:

> In Visual Studio 2003, when i debug a C# program, I found:
>
> 300L * 10000000l -1294967296 int
> 210L * 10000000L 2100000000 int
>
>
> How come? a long type can not handle 300L * 10000000l, the result is too > big for long type?

It's a debugger issue, that's all. It works fine in normal code.

--
Jon Skeet - <sk***@pobox.co m>
http://www.pobox.com/~skeet Blog: http://www.msmvps.com/jon.skeet
If replying to the group, please do not mail me too

[Ryan Liu]

Here is the code. I did make sure are datas are used long type

All it does is scan the ArrayList which is one of the values of a Hashtable.
If it find right AppMessage object in this arraylist, it returns it.
Otherwise it keep scan until time out.

The code makes trouble is the last line -- to check if time out part:

do
{ ......
}
while (!stopReceiving Thread && DateTime.Now.Ti cks - startTime <=
timeoutSecondsL * 10000000L);

I found it returns immediately and tells me it is time out if I put
timeoutSeconds larger then 210.
So in my code, I restrict the time out less than 210 seconds.

Thanks,
Ryan
/// <summary>
/// send a message to AppServer and wait for reply
/// </summary>
/// <param name="appMsg"></param>
/// <returns>the message returned from server</returns>
public AppMessage WaitReply(int cmdType, bool doEvents, int
timeoutSeconds )
{

long timeoutSecondsL = (long)timeoutSe conds;
if(timeoutSecon dsL > 210) timeoutSecondsL = 210L; //when it is 220 ,
*10^7 will get negative value

long startTime = DateTime.Now.Ti cks; //100-nanosecond

do
{
//waitedResponseM sgs is a hashtable, its key is cmdType,value is an
ArrayList

object thisTypeMsgs = this.waitedResp onseMsgs[cmdType];
if(thisTypeMsgs != null && ((ArrayList)thi sTypeMsgs).Coun t > 0)
{

foreach (AppMessage rm in ((ArrayList)thi sTypeMsgs))
{
if( ( rm.CommandType == cmdType) && //check if receiving thread get
the reply for this request
(!rm.IsRequest ) )

{
((ArrayList)thi sTypeMsgs).Remo ve(rm);
return rm; //once received reply, break loop and return
}
}

}

if(doEvents) Application.DoE vents();
Thread.Sleep(50 0); // wait some miliseconds

}

while (!stopReceiving Thread && DateTime.Now.Ti cks - startTime <=
timeoutSecondsL * 10000000L);
throw new Exception("Time out");

}
"Bill Butler" <qw****@asdf.co m> дÈëÓʼþ news:CYKSf.38$y o1.35@trndny09. ..

"Ryan Liu" <ad********@onl ine.sh.cn> wrote in message

news:%2******** *******@TK2MSFT NGP12.phx.gbl.. .

I guess not quite so. I found this problem when my code does not work, then I begin to debug and find this issue.

Are you by any chance stuffing it into an int instead of a long?
Why don't you post the offending code.

Bill

"Jon Skeet [C# MVP]" <sk***@pobox.co m> ????
news:MP******** *************** *@msnews.micros oft.com...

Ryan Liu <ad********@onl ine.sh.cn> wrote:
> In Visual Studio 2003, when i debug a C# program, I found:
>
> 300L * 10000000l -1294967296 int
> 210L * 10000000L 2100000000 int
>
>
> How come? a long type can not handle 300L * 10000000l, the result is

too

> big for long type?

It's a debugger issue, that's all. It works fine in normal code.

--
Jon Skeet - <sk***@pobox.co m>
http://www.pobox.com/~skeet Blog: http://www.msmvps.com/jon.skeet
If replying to the group, please do not mail me too

[Bill Butler]

"Ryan Liu" <ad********@onl ine.sh.cn> wrote in message news:O0******** ******@TK2MSFTN GP11.phx.gbl...

Here is the code. I did make sure are datas are used long type

All it does is scan the ArrayList which is one of the values of a Hashtable.
If it find right AppMessage object in this arraylist, it returns it.
Otherwise it keep scan until time out.

The code makes trouble is the last line -- to check if time out part:

do
{ ......
}
while (!stopReceiving Thread && DateTime.Now.Ti cks - startTime <=
timeoutSecondsL * 10000000L);

I found it returns immediately and tells me it is time out if I put
timeoutSeconds larger then 210.
So in my code, I restrict the time out less than 210 seconds.
First:
Does it return or does it throw an exception?

clipping your method down to a minimum I go the following

public void WaitReplyGood(i nt timeoutSeconds )
{
bool stopReceivingTh read = false;
Console.WriteLi ne("WaitReplyGo od({0})",timeou tSeconds);
long startTime = DateTime.Now.Ti cks;
do
{
Thread.Sleep(50 0);
Console.WriteLi ne(DateTime.Now );
}
while (!stopReceiving Thread && DateTime.Now.Ti cks - startTime <= timeoutSeconds * 10000000L);
return;
}

This does not exhibit the behavior that you described (I even used an int for timeoutSeconds since
that makes no difference).
If you replace the 10000000L by simply 10000000 you do get the behavior that you described.
If stopReceivingTh read is true you get that behavior as well.

See if you can comment out more of your code and still make the error occur.

Keep us posted

Bill

/// <summary>
/// send a message to AppServer and wait for reply
/// </summary>
/// <param name="appMsg"></param>
/// <returns>the message returned from server</returns>
public AppMessage WaitReply(int cmdType, bool doEvents, int
timeoutSeconds )
{

long timeoutSecondsL = (long)timeoutSe conds;
if(timeoutSecon dsL > 210) timeoutSecondsL = 210L; //when it is 220 ,
*10^7 will get negative value

long startTime = DateTime.Now.Ti cks; //100-nanosecond

do
{
//waitedResponseM sgs is a hashtable, its key is cmdType,value is an
ArrayList

object thisTypeMsgs = this.waitedResp onseMsgs[cmdType];
if(thisTypeMsgs != null && ((ArrayList)thi sTypeMsgs).Coun t > 0)
{

foreach (AppMessage rm in ((ArrayList)thi sTypeMsgs))
{
if( ( rm.CommandType == cmdType) && //check if receiving thread get
the reply for this request
(!rm.IsRequest ) )

{
((ArrayList)thi sTypeMsgs).Remo ve(rm);
return rm; //once received reply, break loop and return
}
}

}

if(doEvents) Application.DoE vents();
Thread.Sleep(50 0); // wait some miliseconds

}

while (!stopReceiving Thread && DateTime.Now.Ti cks - startTime <=
timeoutSecondsL * 10000000L);
throw new Exception("Time out");

}
"Bill Butler" <qw****@asdf.co m> дÈëÓʼþ news:CYKSf.38$y o1.35@trndny09. ..

"Ryan Liu" <ad********@onl ine.sh.cn> wrote in message

news:%2******** *******@TK2MSFT NGP12.phx.gbl.. .

>I guess not quite so. I found this problem when my code does not work, then > I begin to debug and find this issue.

Are you by any chance stuffing it into an int instead of a long?
Why don't you post the offending code.

Bill

>
> "Jon Skeet [C# MVP]" <sk***@pobox.co m> ????
> news:MP******** *************** *@msnews.micros oft.com...
>> Ryan Liu <ad********@onl ine.sh.cn> wrote:
>> > In Visual Studio 2003, when i debug a C# program, I found:
>> >
>> > 300L * 10000000l -1294967296 int
>> > 210L * 10000000L 2100000000 int
>> >
>> >
>> > How come? a long type can not handle 300L * 10000000l, the result is
> too
>> > big for long type?
>>
>> It's a debugger issue, that's all. It works fine in normal code.
>>
>> --
>> Jon Skeet - <sk***@pobox.co m>
>> http://www.pobox.com/~skeet Blog: http://www.msmvps.com/jon.skeet
>> If replying to the group, please do not mail me too
>
>

[Jon Skeet [C# MVP]]

Ryan Liu <ad********@onl ine.sh.cn> wrote:

I guess not quite so. I found this problem when my code does not work, then
I begin to debug and find this issue.

Well, 300L * 10000000L works absolutely fine in code.

Could you post a short but complete program which demonstrates the
problem?

See http://www.pobox.com/~skeet/csharp/complete.html for details of
what I mean by that.

--
Jon Skeet - <sk***@pobox.co m>
http://www.pobox.com/~skeet Blog: http://www.msmvps.com/jon.skeet
If replying to the group, please do not mail me too

[Mike]

"Ryan Liu" <ad********@onl ine.sh.cn> wrote in message
news:O0******** ******@TK2MSFTN GP11.phx.gbl...

Here is the code. I did make sure are datas are used long type

All it does is scan the ArrayList which is one of the values of a
Hashtable.
If it find right AppMessage object in this arraylist, it returns it.
Otherwise it keep scan until time out.

The code makes trouble is the last line -- to check if time out part:

do
{ ......
}
while (!stopReceiving Thread && DateTime.Now.Ti cks - startTime <=
timeoutSecondsL * 10000000L);

I found it returns immediately and tells me it is time out if I put
timeoutSeconds larger then 210.
So in my code, I restrict the time out less than 210 seconds.

Thanks,
Ryan
/// <summary>
/// send a message to AppServer and wait for reply
/// </summary>
/// <param name="appMsg"></param>
/// <returns>the message returned from server</returns>
public AppMessage WaitReply(int cmdType, bool doEvents, int
timeoutSeconds )
{

long timeoutSecondsL = (long)timeoutSe conds;
if(timeoutSecon dsL > 210) timeoutSecondsL = 210L; //when it is 220 ,
*10^7 will get negative value

long startTime = DateTime.Now.Ti cks; //100-nanosecond

do
{
//waitedResponseM sgs is a hashtable, its key is cmdType,value is an
ArrayList

object thisTypeMsgs = this.waitedResp onseMsgs[cmdType];
if(thisTypeMsgs != null && ((ArrayList)thi sTypeMsgs).Coun t > 0)
{

foreach (AppMessage rm in ((ArrayList)thi sTypeMsgs))
{
if( ( rm.CommandType == cmdType) && //check if receiving thread get
the reply for this request
(!rm.IsRequest ) )

{
((ArrayList)thi sTypeMsgs).Remo ve(rm);
return rm; //once received reply, break loop and return
}
}

}

if(doEvents) Application.DoE vents();
Thread.Sleep(50 0); // wait some miliseconds

}

while (!stopReceiving Thread && DateTime.Now.Ti cks - startTime <=
timeoutSecondsL * 10000000L);
If you do this, it'll be much more clear, and you'll avoid whatever int/long
issue is facing you:

DateTime dateTimeEnd = DateTime.Now + new TimeSpan( 0, 0,
timeoutSecondsL );
do
{ }
while (... &&(DateTime. Now <= DateTime.End)
m


throw new Exception("Time out");

}
"Bill Butler" <qw****@asdf.co m> дÈëÓʼþ news:CYKSf.38$y o1.35@trndny09. ..

"Ryan Liu" <ad********@onl ine.sh.cn> wrote in message

news:%2******** *******@TK2MSFT NGP12.phx.gbl.. .

>I guess not quite so. I found this problem when my code does not work, then > I begin to debug and find this issue.

Are you by any chance stuffing it into an int instead of a long?
Why don't you post the offending code.

Bill

>
> "Jon Skeet [C# MVP]" <sk***@pobox.co m> ????
> news:MP******** *************** *@msnews.micros oft.com...
>> Ryan Liu <ad********@onl ine.sh.cn> wrote:
>> > In Visual Studio 2003, when i debug a C# program, I found:
>> >
>> > 300L * 10000000l -1294967296 int
>> > 210L * 10000000L 2100000000 int
>> >
>> >
>> > How come? a long type can not handle 300L * 10000000l, the result
>> > is
> too
>> > big for long type?
>>
>> It's a debugger issue, that's all. It works fine in normal code.
>>
>> --
>> Jon Skeet - <sk***@pobox.co m>
>> http://www.pobox.com/~skeet Blog: http://www.msmvps.com/jon.skeet
>> If replying to the group, please do not mail me too
>
>

[Jon Skeet [C# MVP]]

Mike <vi********@yah oo.com> wrote:

If you do this, it'll be much more clear, and you'll avoid whatever int/long
issue is facing you:

DateTime dateTimeEnd = DateTime.Now + new TimeSpan( 0, 0,
timeoutSecondsL );
do
{ }
while (... &&(DateTime. Now <= DateTime.End)

Or to make it even clearer (and avoid problems with knowing how many 0s
to put in):

DateTime end = DateTime.Now.Ad dSeconds (timeoutSeconds );

--
Jon Skeet - <sk***@pobox.co m>
http://www.pobox.com/~skeet Blog: http://www.msmvps.com/jon.skeet
If replying to the group, please do not mail me too

[Rick Lones]

Ryan Liu wrote:

In Visual Studio 2003, when i debug a C# program, I found:

300L * 10000000l -1294967296 int
210L * 10000000L 2100000000 int
How come? a long type can not handle 300L * 10000000l, the result is too
big for long type?

It doesn't look like it *is* going into a long. It looks like the result (or
maybe an intermediate result) is being stuffed into an int, in which case (210 *
10000000 = 2100000000) would fit, whereas (300 * 10000000 = 3000000000) would be
an overflow (greater than 2147483647, which is the maximum positive integer).
But as has been noted - if you want a more definitive answer you should post a
short complete program . . . your snippet does not show the data type of
timeoutSecondsL , for example.

In any case recalculating (timeoutSeconds L * 10000000L) on each iteration is
kind of bad form - calculate it once before the loop, stuff the result into a
long, test against the long in your loop and likely your whole problem will go
away anyhow . . .

HTH,
-rick-