We have OECGI2 happily using the Get_OS_File routine to grab a pdf or xls file, append the appropriate content-type headers and return the file. Yes, I have confirmed that Request is set to 1 so it returns binary. I wrote a routine "RUN_MYCGI_REQUEST" which calls "RUN_OECGI_REQUEST" in order to totally get ahead of any content-type changes that RevSoft routines might be making. The very last thing I do before returning the Response to the internet that has been returned from RUN_OECGI_REQUEST is to OSWrite a copy of the data to the server hard disk. It is totally non-corrupted and open just fine. However, after it leaves our hands something corrupts the pdf or xls files.

The character being inserted into the file is hex \fffd\ which (when Googled) says it is the "Unicode Replacement Character". There is also a char( 1) appended to the end which is listed as the "Unicode Substitute Character". This has led us to add the "charset=utf-8" to the Content-Type string, but that has not helped.

So to sum it up, it is leaving our hands completely clean, but something out there in the "ether" is corrupting the file. Guesses anyone?

Jim,

What is the date on the OECGI2.exe?

I think you want wan that is may2008 or newer

- Bob

View this thread on the Works forum...