Got a REALLY weird one for you all:
After version 14, some of our InfoMaker reports that are scheduled to run to PDF still run and display properly, but when text is selected and then copy/pasted into notepad or tessitura, the characters that are pasted are different from what is visually displayed in the PDF. This only occurs on some types of reports; not all. On-Account Tracking and Customer Service Issue tracking are both affected.
As an example, if we see the text 83074, selecting it and copying will actually copy the character codes <0x1b><0x16><0x13><0x1a><0x17> to the pasteboard. (Depending on what PDF reader you are using, this may end up being ????? characters or no text at all.)
Any of you who know your ASCII tables may notice that 83074 would usually be represented as codes 0x38, 0x33, 0x30, 0x37, 0x34. And yes, there is a consistent difference between the character codes being displayed and those being copied, which is 0x1d. This seems rather arbitrary, so no ideas immediately jump out as an explanation.
This behavior started when we moved to v14 and changed servers. At that time, I did rebuild the machine running the scheduled reports, but I am using the same Sybase DataWindow PS driver recommended in the documentation, and I believe it was the same on my old server. Also, the behavior is the same when saving a PDF with the client running on a local machine. Those all have the Sybase DataWindow PS printer configured using the ghostpdf.inf driver, but the behavior is the same.
Anybody ever see something like this before? It didn't occur to me that the GhostScript version might have changed, so I'm going to take a look at that next.
7KDQNV!