itextsharp help required!

Using itextsharp for PDF generation. Software is on about 700 PC's and it doesn't work for TWO of them. This is the stack trace on the two machines where it doesn't work:

Code:

---

Exception Source:      itextsharp

Exception Type:        System.IO.IOException

Exception Message:    Rebuild failed: Input string was not in a correct format.; Original message: Input string was not in a correct format.

Exception Target Site: ReadPdf

 

---- Stack Trace ----

  iTextSharp.text.pdf.PdfReader.ReadPdf()

      xClaim.exe: N 00190

  iTextSharp.text.pdf.PdfReader..ctor(filename As String, ownerPassword As Byte[])

      xClaim.exe: N 00247

---

Things I have tried - none work:

• Uninstalling/Reinstalling Adobe 6 and 7 Professional (not just reader) (yes, one failure is occurring on a patched AA6, another on patched AA7)
• Changing the path of the blank PDF they are opening and populating to a wide variety of places on the OS.
• Added the blank PDF file as a resource to my solution and used the Byte overload ctor (no luck - same error message which surprised me greatly - just the different ctor shown)

The file isn't corrupt, they can navigate to the exact same PDF that I'm trying to load with itextsharp and double click it just fine. At first I thought it might be an application trust issue, but it's deployed through ClickOnce so that shouldn't be it. Perhaps there is an option in Adobe somewhere that is prohibiting opening in this fashion? Why is it working on 698 PC's and not two? Any ideas would be helpful.

I'm guessing it's something to do with this

filename As String

If the filename consists of the path here, and the path is the windows documents and folders path, then look at the usernames involved. Any unusual characters that don't belong in a name such as an apostrophe? Or any other strange characters?

Static path, same for all users. Doesn't work on TWO computers.

Plus, I changed the ctor used and compiled the blank PDF into my app instead of locating it on the file system. Still same error, slightly different stack trace:

Code:

---

Exception Source:      itextsharp

Exception Type:        System.IO.IOException

Exception Message:    Rebuild failed: Input string was not in a correct format.; Original message: Input string was not in a correct format.

Exception Target Site: ReadPdf

 

---- Stack Trace ----

  iTextSharp.text.pdf.PdfReader.ReadPdf()

      xClaim.exe: N 00190

  iTextSharp.text.pdf.PdfReader..ctor(pdfIn As Byte[], ownerPassword As Byte[])

      xClaim.exe: N 00253

  iTextSharp.text.pdf.PdfReader..ctor(pdfIn As Byte[])

      xClaim.exe: N 00007

---

if I had to trouble shoot this issue, I would be looking into what elements that iTextSharp may use that is DIFFERENT on the problem machines, versus the good ones. That would include framework versions, any patches/SP applies to the framework version this app is targetting. After I decided that was all good, I would let the people on this forum know, and ask for more suggestions ;)

Target machines are 2000 miles away and I cannot reproduce problem here. At wits end, gonna have to deploy anyway.

2000 miles away doesn't mean a whole lot in the age of computers.

What OS do they run? You should be able to remote into them and poke around...

all depends on if this is consumer machines, corprorate environment, etc...

Corporate environment, permissions to do that strictly controlled.