>Hi James,
>
>Tried this on Friday as well and problem did not go away. I finally got someone quite helpful (if not knowledgable) at HP and we worked through all that I have tried. He went back and forth to a few people when I was on line. In the end though, since Word and Excel work with the driver, they are not willing to do anything until MSFT looks at it since "it must be a VFP problem". I was able to convince him that that line of reasoning is not necessarily true as it might be that VFP just calls the driver in a different way thus exposing an error in the driver. He said "that's true" but you will need to go to MSFT first to establish that.
>
>So I called MSFT Friday late and got someone who is at least willing to try the app on one of their test machines. We'll see what they come up with. Not being a C++ programmer, I cannot give them details on how they need to trap the error at the OS level in order to determine where the error actually is. I plan to suggest that once they get back to me that they were able to reproduce the problem.
>
>I cannot remember if I said here before but for anyone following the thread, here is a short list of what does or does not work:
>
>HP 4050 PCL 5e driver - no workee, "Pure Virtual Function" error every time
>HP 4050 PCL 6 driver - same problemo
>HP 5 "Standard" driver - better, prints about 75% of the time
>HP 4 driver - better yet, just about always prints (but not always...what did you expect, huh :-)
>
>So we wait to see if a guru from MSFT can help. Most likely scenario is for them to say "you've got to call HP about this" to which I will count to 10 and say (very cheerfully), "Done that!".
>
>Thanks.
>
Albert,

It occurs to me that one of the things that might make this happen is the way VFP interfaces with winspool.drv. VFP doesn't call the HP driver at all, it calls winspool which in turn calls the driver. This is more fodder for the argument that it's the driver and not VFP, BTW.

I've run across a KB article that might be of help, Q175083 ("HOWTO: Print Line-by-Line to a Printer"). While it's in VB, it'd be a simple matter to test since all the source code is provided and it can be copied and pasted as required. My thought here is that perhaps this might emulate how VFP does it's printing. It occurs to me that one major difference between VFP and an application like Word may be the fact that in Word, the document already exists, while in VFP it's generated at runtime. This is a pure SWAG on my part, but if you got the same error in the VB simulation, it would pretty much nail the driver as the culprit.