The so-called flag of the active partition is actually also formulated by Microsoft. No matter what Microsoft formulates, it wants to become an industrial standard because everyone uses its system. On the first sector of the hard disk, at the beginning is the MBR program code, and at the end is the partition table. The structure of the partition table is also formulated by Microsoft (or we can at least assume that it is formulated by Microsoft). The so-called flag of the active partition is also only what the MBR program code of Microsoft needs to use. Other boot programs do not necessarily need to recognize this active flag. For example, the GRUB boot code as MBR ignores the active flag and directly starts the GRUB main program. After the GRUB main program gains control, it then looks for GRUB's own MENU.LST file and starts the operating system on any partition (which can be the operating system in the logical partition of the extended partition) according to the instructions inside, completely ignoring the partition active flag placed on the partition table by Microsoft.
What you said about Microsoft's boot process is correct.
The GRLDR boot sector installed by bootlace is usually not installed to a certain partition, but to the MBR. In fact, bootlace can also be used to install the boot code to the boot sector of any partition, but this is an advanced topic (dangerous) and not for ordinary users. Ordinary users will be able to use this function in the future, but that has to wait until these mechanisms are all perfected.
The boot code on the MBR only takes charge of finding and starting the GRLDR file (from each partition) and does nothing else. Once the GRLDR file is started, GRLDR finds its configuration file menu.lst and starts any other operating system according to the instructions in menu.lst. GRLDR can recognize FAT12/16/32/NTFS, as well as various file systems under Linux, so it is not difficult for it to start any system. It doesn't care about the active flag of the partition, whether it is active or not, it can start. It even has a command specially used to change the active flag of a certain primary partition.
grub4dos was mainly maintained by me before. bean is to take over from me, and from now on, bean is the main maintainer, and I am also a participating member. bean's joining should be said to be a big leap in this project. bean is a professional, and I am an amateur, so there are naturally some differences. Of course, most of us are amateurs, and I don't mean that amateurs must be bad and professionals must be good. I just mean that bean is both professional and very good. bean has done a lot, for example, chainloader loading IO.SYS of WinME, which involves Microsoft's secrets (decompression), which is very difficult, and this is what bean did. There are also bean's patches in the int13 emulation code. These are the work that bean did during the period when I was the main maintainer. Now bean has shown strength in the NTFS aspect. NTFS is a tough nut to crack. The world many people are researching NTFS, but because Microsoft does not disclose the technical details of NTFS, it is very difficult for the outside world to fully grasp the access method of NTFS. Although the open-source community can now read NTFS, writing to NTFS is still unreliable. I have written the boot sector of EXT2, so I can understand the difficulty of writing the boot sector. Fairly speaking, comparing the boot sectors of all kinds of file systems, then the NTFS boot sector is the most difficult to write. bean has written it and is now perfecting it. bean has done a lot, and it can't be described one by one here. To sum up, bean will further perfect grub4dos. I believe that through his and our joint efforts, grub4dos will go further.
因为我们亲手创建，这个世界更加美丽。
DigestIII
StickyII