Comment 2 for bug 500696

Still seeing this bug on an up-to-date Lucid system.

Is there a way to test this outside of usb-creator? I'm not familiar with how the boot sector in a DOS filesystem is modified, let alone what the "proper" way of doing it (that would update the backup as well) would be. But if usb-creator is writing the boot sector by doing the equivalent of a dd(1) directly to the device, then I don't think the filesystem driver can be faulted.