UEFI evolution in time

Consider briefly the path.

In principle, by 2010 we can recognize the existence of two steps in the boot.

1. The first step is called PI (platform initialization), which includes the start of the underlying platform: the first thing that ran was the starting safety protocol, whether it was positive they proceeded in sequence to start the processor, then the chipset, then the motherboard. Once this was working, they proceeded to load the operating system. Here proposed two ways: loading an operating system transient response, to host applications as web browsers and email clients, and load the operating system of choice, which was previously choosing which device would run the boot (disk, memory card, etc..).

2. After the start, and being raised as the operating system, you may want to interact with this firmware directly. This should be done exclusively by the UEFI interfaces wherever possible. In this section, remember that Windows historically chose not to speak to the BIOS directly, but use the concept of high-speed drivers rather than access the underlying firmware in all modern versions of Windows.

To enter address the underlying issue, say that we can describe, prima facie, four-stroke-cycle main use of a PC:
     on
     Platform Initialization
     Starting the Operating System
     off