Vince, does RICH have a debugging mode where it creates a trace file that delineates the steps taken from input to output? I know that it won’t help with this particular situation, but it may help to diagnose situations of this sort in the future.

If there are currently no debugging options available for RICH, I may be able to help set something up that could help. You have my email.

Thanks Dave,

I use a variety of IDE(s) for different stuff (although I still do a lot of coding in notepad LOL)  and in this case Im using Visual Studio which can step through debug, output stack trace, etc.  The modules themselves have an internal logging system that logs exceptions and application data, both to file and to the OS application log. I have logs that log information about the inrormation thats being logged. In this case, the response will not duplicate under any condition I can set up on the development machines. The concatenation\truncation of the string is simply not something that should be able to happen in the code, so at this point Im thinking that there was some sort of “one in a bazillion” problem that happened in the unmanaged (native) libraries. At the precise moment that Arts data was being processed, something pushed onto the stack invading that allocation and caused it to truncate.  Im going to move to the production server and see If I can get it to duplicate there. Its been acting “wonky”. For the last few weeks its been restarting at irregular intervals which I thought might be a CPU thermal issue, but that wasnt it.  The simple fact is that it needs to be replaced.

Vince

Vince,

I tried the chat experiment again with RICH
and I gor the same results but this time, I think
I might know why.

Please contact via email@stoney535atgmaildotcom

I also took some screen captures to share with you.

Thanks,
- Art -

LOL And this gets more interesting. Saw your log entry again

http://ai.r-i-software.com/TypeVI/log.aspx?file=conversation.232013.xml&id=qc2xm4rsbmzw04eu3eygug45

and heres me entering the same string on the production server

http://ai.r-i-software.com/TypeVI/log.aspx?file=conversation.232013.xml&id=xm1yx445ulypkp55akfdasie

No problem posting here Art. I just tried to login to the production server and it had once again performed an auto shutdown. Im still leaning towards the idea that there are problems with the server itself, but Im definitely interested in any ideas that you may have. Im going to go through the system logs.

Vince

@Art, and in general
[chuckles]
The problem wasnt in the namefunctions module, thats why I couldnt reproduce the error. And when I tried entering the text on the production server, I only tried adding the carriage return at the end of the entire block. When I saw the screen shots I knew what was happening. I didnt understand that you meant that you were using carriage returns after each line. I had just changed the way the pre-processor handles linefeeds by causing them to be replaced with NULL rather than SPACE. The was to facilitate changes I had made when addressing how the pre-processor deals with multiple spaces. What was happening was the last word in the sentence was in effect being concatenated with the first word in the next sentence.

So

My name is art
yes
an asteroid was supposed to pass very close by the earth

became

my name is artyesan asteroid was supposed to pass very close by the earth

So Ive fixed the problem and will upload it tomorrow.
Ive said it before, Ill say it again. 5 minutes of Beta testing will uncover more than 6 months of debugging!

Thanks!

Vince