John, I don't think it really matters if the URL's are relative. One strategy I would consider is to use VFP to access IE as an automation object, tell the browser to navigate to the page, and use the automation object reference to access the browser's DOM, which will give you a traversable hierarchy as well as very useful collections of the contents of the web page. You might even be able to dynamically set the base URL in the DOM and have it then return fully qualified URL's from the hyperlinks within the document.

>I have a VFP application that needs to read the current web page in Internet Explorer. The user navigates to an ASP web page where there are almost 20 files that need to be downloaded. I have been succesful in having the user enter the file name (the 1st 7 bytes are common to all files) and then downloading the files directly into my VFP application by using some setup vars to set the needed internet path. The problem is that now the ASP service provider is using 3 servers and the files could be on any one of the three. To make matters worse, the WRONG version of the files could be on one of them so I can't just look for the file (I was till I cought heck from the users). I need to be able to read the html of the page the user has on his/her screen so I can get the correct server.
>
>I have some API calls that tell me the name and handle of all open windows but I don't know where to go from there.