I have an application that uses Crystal Reports for all reports. Each Crystal .rpt file has a subreport that contains the customer’s logo and address information that is used as a header for the report. The app has a routine that allows the users to edit the header subreport and then runs a loop through all existing .rpt files to replace the old header subreport with a new one. That calls Crystal RDC methods to accomplish the task for each report file. The whole thing works beautifully in most cases, i.e. here on my development machine, and, in another test case, in a runtime environment under Terminal Services on a hosting server. However, on a different hosting server, in a near identical configuration, using exactly the same app, exact same list of .rpt files, and so forth, the routine bombs after just one or two report files with the error message “Unexpected automation error: OLE exception error: Exception code c0000005. OLE object may be corrupt Error #: 1440 in frmSplash.reviseheader at line 0.” C5 error, right? Or maybe not. 1440 indicates an automation error coming from Crystal. Or what? I have splattered SYS(1104) around liberally in an attempt to free up memory; to no avail. Deleted the FOXUSER file, ditto. Does anyone have an idea what might be going on in this one server that causes this issue?