As I fired up one of my test machines yesterday after replacing a defective motherboard and Windows started to boot, I thought I might actually have gotten away with something grand and glorious. Not!! Although I was able to replace an Asus P5 Pro that had gone wonky on me with a Gigabyte GA-P43-ES3G and get Windows to start seemingly OK, further analysis revealed that I’d been bitten by an interesting pair of BIOS defaults involving:
- USB Keyboard disabled
- IDE disk controller (not AHCI)
Both the source and target systems use a built-in Intel AHCI ICH10 SATA RAID controllers, and I’d hoped to transfer the 64-bit Windows Professional AHCI installation unaltered as well. But because the BIOS refused to recognize the USB keyboard the first time the system booted (so I couldn’t enter the BIOS set-up program by hitting the DEL key), it automatically defaulted to IDE controller status and trashed all the carefully constructed AHCI setup I’d gone through on the previous incarnation.
Is it my fault that I forgot to plug in the PS/2 keyboard for that first boot-up? Technically, yes it is. But dang, nobody uses PS/2 keyboards any more and it ticks me off to no end (especially since I’m now going to have to rebuild this whole installation from the ground up) that the BIOS makers still don’t turn support for USB keyboards on as a BIOS default. C’mon guys: this was OK in the mid-90s when USB keyboards were the exception rather than the norm. But they’ve been the norm for at least 15 years now, so it’s time to change the basic set of assumptions that BIOS builders encode into their platforms to reflect a more current reality now on the ground.
I’d been so happy that Windows 7 didn’t puke on my motherboard swap and make me re-assert my license (though it did ask me to (re-)activate my license, at least I didn’t have to get on the phone to get a 25 digit activation key from somebody, which is usually a 20-minute process or so) that it took me a while to recognize that I’d munged the re-animation of this machine anyway by neglecting to hook up a PS/2 keyboard for its initial reboot. Now that I’ve been forcibly and painfully reminded of this situation once again, I won’t make this mistake for another two years or so. But again I really wish the BIOS builders would catch up with the first decade of the 21st century, rather than maintaining the status quo that prevailed at the close of the last millenium.
I’m waiting for some more memory, a new SSD, and a newer graphics card to replace the fanless 9600GT in this test machine before I re-do the current installation, but it’s already on my to-do list for later this month. In conclusion, all I can say is “Ouch! Ouch! Ouch! My bad…”