From my experience (and much reading) you will certainly need to set the column and the head carefully, and this can take both time and ingenuity. The process is called 'tramming', and means that the column and the quill are to be set truly at right angles to the bed, both sideways and fore & aft. Unless this is done, the cutter will not follow a hole but will skew to one side as the head is moved up and down the column. Once this verticality is established, you can set about ensuring that the head itself is accurately set to the column. This ensures that movement of the quill within the head is also truly at right angles to the bed.
There may be no 'official' way to do this adjusting in practice. You can use an engineer's square to check, or rely on setting the bed level and then checking the quill with a level. Or, use a DTI carried in the chuck and rotated in contact with the bed. But it is too easy, unless you really think about the process, to end up with a vertical quill, but a cock-eyed column, so that quill movement is true but lowering the whole head on the column moves the tool sideways.
How you adjust the column itself is normally a matter of sliding strips of shim steel (or brass) into the spaces where the bolts attach, for the fore & aft measurement, and then ensuring as the bolts are tightened that the column is held vertical (for the sideways measurement). Yes, it can be a bit precarious. The real bad news is that should you ever suffer a serious tool jam, or a similar accident, you need to recheck and perhaps readjust these settings.
A simple way to check the state of a mill is to use a fly-cutter on a flat piece of, say, aluminium plate held firmly down on the bed. Adjust it to take the merest hint of a cut, and traverse the work across beneath the rotating cutter. It should show the same even cutting pattern over the whole cut surface.
Then do the same thing and move the bed towards you instead of sideways. Any hint of hit-&-miss cutting in either of these tests shows that the cutter is not working exactly parallel to the bed.
Then use a new rod of silver steel in the chuck, and check with an engineers square in both directions. No daylight should show over the full length of the rod, if the column and head are set correctly.
I hope this helps.
Regards, Tim