To clarify a bit more:
I have built a web page that references statically named .js files, for example js1.js. These names will always stay the same, so the web page code is never changed. Both of these files are local on a pocket pc 2003 "PIE" browser. The js1.js file is supposed to point to a file on my ftp server named js1.0001.js. This file contains a link to a CAB file (like a setup.exe for desktop windows) that will install documents onto the pocket pc and change the contents of js1.js to point to the next sequential file on the ftp server, js1.0002.js. If I have not created js1.0002.js, it will just have a blank space on the web page. If js1.0002.js does exist, then an animated gif will appear that is also a link to the aforementioned cab file.
I try to think of it as a multiple redirect, something like passing the buck, so to speak.
web page "look at js1.js to fill this spot in my page"
js1.js "don't look at me, look at js1.0001.js. Insert it's code into your page"
js1.0001.js "Ok, here's a hyperlink with a nifty flashing gif pointing to a cab file"
I hope that makes sense.