In either case you will need to have a program on the server whose job is to process the submission, so something with the PDF or FDF, and return a response. In the cse of the complete PDF, this program (Perl script, PHP, ASP, Cold Fusion script, etc.) would take the entire contents of what is submitted and save it to a file on the server. The point is is doesn't happen automatically. When you set up the Submit Form action, you specify the URL of the server-side program that does the processing.

In the case of FDF, the server could get it and merge it with a PDF, but this is normally done with a server-side component such as Appligent's FDF Merge or some component of Adobe's LiveCycle offerings.

Finally, the server has to return a response that can be understood by the client. In the case of a submission from Acrobat/Reader, this can be FDF, which can provide some user feedback, redirect to a static or dynamically generated HTML page, or an empty response.

This all means that you or someone else needs to do some programming. It not necessarily all that complicated, but it helps a lot to have some expertise in this area.

The submission arrives on the server as the input stream, so in PHP you can simply read it and write it back to a file.

See the example in http://tv.adobe.com/watch/acrobat-community-tv/working-with-form-data/

PHP used to include support for FDF parsing, but it's been depreciated. You can still get the libraries though, and put the classes back (along with a copy of the FDF Toolkit binaries) - see http://php.net/manual/en/book.fdf.php - but if you send XFDF data you can parse it easily as it's just XML.

Also, bear in mind that recent security changes in Acrobat and Adobe Reader mean that if you're viewing the PDF on the desktop, FDF responses from the server will be blocked if they try to change the document or execute a script (such as filling in a form value or opening an alert box). If the PDF is running in a browser and the submission is to the same domain as the PDF was served from, the response will be allowed to run.

Ok, when you typed in that PHP it was stored on your web server (I assume it was, since that's where it will need to be to make it work for anyone but you). So that's what I meant by "install PHP scripts on the server".

Before talking any more about how to get the full PDF, I think it's worth talking more about the FDF. FDF files are actually more useful than you perhaps realise. When you have collected a lot of them, they are a whole lot easier to work with and analyse than the filled in PDF. It can save you a bunch of money in some cases too.

What is your objection to FDF files? What is that you (believe that) you can't do with them that you can do with a filled PDF?