Odd problem with a Report

[p.koppers]

Hi,

I've stumbled onto a odd problem with a Report, or the Report Engine ...
There is an application that runs client-server and uses the Report Engine to print a specific report using a filter.

It works fine in the current circumstances, but the same application produces Dr.Watson messages (GPF on prowin32) at another customer. This also occurs at my current customer in a test environment with almost the same conditions as the other customer.

The odd thing is that thr problem appears to be linked to the report number (each report has a number which the filter uses to print a specific report), but I don't understand why ...
When printing a report with a number greater than 9999 (thus 10000 and higher) the next call to the report engine will cause a Dr.Watson.
The number is used in the report, but has display format that can work with 10 digits.

Does anyone recognize this kind of odd behaviour ? Or does anyone know what the real problem is ?

As for some extra info:
client: WinNT 4, Progress 8.3b
server: unix, Progress 8.3b
(The other customer uses 8.3d)
Application runs continuously (that is: until the Dr.Watson) and cals upon the printrb interface of the Report Engine.
(table interface will produce the same error)

 

[p.koppers]

Don't know if it could cause this problem, but I noticed that the test environment has a 8.3c database and the other customer has a 8.3d database.
Compiling the application, however, is done with 8.3b.
Can this have something to do with the problem ?

 

[p.koppers]

problem solved

or so I guess ...

Compiling with 8.3c in the test environment solved the problem there.
Hopefully compiling with 8.3d will do the same for the other customer.

I still wonder what really caused this problem, but that will be unanswered