Hi guys. This is great! I have previously lurked on here and indeed read the document about the PNP drivers paths/ordering the other day prior to this thread starting. I then questioned in my mind whether it was valid or not, so am very happy this thread has started, to the extend I finally signed up!!! Back on topic: 1. Yes, you can use subfolders, but they need to be specified in you unattend file individually. At least we have some choice over folder design since the old W2k pre SP2 96 character limit for the PNP drivers path! However, I don't understand your logic of having the top level folders numbered. Number ing Subfolders makes sence ie Video\000_Trident Video\0001_Trident In this example if you NEED to include two varying Trident Video drivers for some sort of compatability issue, you know the newer driver you place in the 000 folder will be seen 1st. However, how does numbering the top level folders help, apart from perhaps a very negligable performance increse? The model below is what I mean 000=smbus/chipset 001=lan 002=usb/firewire/serial i-o 003=audio 004=video Surely Windows only even looks for a driver, if it finds a device in the first place?? The example I am questioning seems to infer Windows looks at what drivers it has and then looks to see if it has some hardware to match it?! Surley it wouldn't matter if you had at the top level: 000_Sound 009_Chipset If Windows finds chipset hardware 1st it will still go looking for a chipset driver, in whatever paths you have chosen. Sure it will look in 000_Sound first, as it routes through folders in alphabetical sequence, but it doesn't mean it will install the Sound drivers first just as they are in the first folder does it??? Likewise a chipset driver would not definately install 1st just as it was in a 000_chipset folder? To me, the folders should be designed logically, but have no actaul impact on the older devices are detected or installed. What do you think? Graham
