RN> After disk 4 you reinsert the disks installation and one
 RN> before continuing with the rest.
 RN> I wondered why the disk images weren't stored in a temp directory to 
 RN> avoid this?

Because OS/2 only reaches the stage where it is executing user-mode programs,
such as SYSINST*.EXE, several floppy discs into the whole process.

The installation program is just an ordinary native OS/2 program.  The loading 
of the operating system kernel *itself*, and of all of the the BASEDEV base
device drivers that are loaded (just in case), is split over two discs.  It is 
only by the time that one has reached the second disc that the operating
system is properly up and running and executing application programs.  It is
*then* that the installation program is in control.  But by that time it is
too late to have made copies of the installation floppy disc when it was
inserted.  Indeed, when the installation floppy disc is in the drive, not
enough of the operating system is actually loaded for application programs,
supposing that they could run at that point, to properly see any hard discs
*at all*, so one couldn't make temporary disc images even if one wanted to.  

This is aside from the fact that the user doesn't select a destination disc
drive until the installation program has run and displayed several prompts,
which happens well after the installation floppy has been inserted and then
removed; and that there may not even *be* a destination disc drive at all
since the hard disc may be blank, requiring FDISK to be run first and the
installation procedure to then be restarted from the beginning.

 ¯ JdeBP ®

--- FleetStreet 1.22 NR
633/260
2501/209