A rite of passage for mill owners is the tramming process, where the head is adjusted to be (almost) perfectly perpendicular to the table. The Homier mill is no exception; my mill's head was slightly off, typical for all mills. The documentation provided did not cover how to tram the mill -- the column and base looked monolithic except that there were 4 large bolts adjacent to the base.
The table was centered to avoid any tilt and the head was set low on the column and locked. A Sharpie was used to mark the fore and aft measurement points on the table(see picture above); the left and right points were judged based on the table slots. To minimize the DTI's movements a piece of glass was placed on the table for the initial work (both the table and the glass must be very clean: even one small piece of swarf could ruin the accuracy of the readings).
A drill chuck was used to hold a rod appropriate for my DTI clamp and the DTI was mounted as shown to allow measuring the variation in distance between the spindle and the table. The accuracy of the chuck and the straightness of the rod do not affect the readings; using a collet to hold the rod will not allow more accurate readings. As with RDM for lathe alignment, taking the difference between readings eliminates errors due to these sources.
Four readings were taken in order to judge the alignment error initially, the "Before" readings in the table (in thousandths). The diameter of the circle touched by the DTI's probe was 8 inches. The readings before and after tramming are shown in the table below:
| Homier MillDrill Tram Readings | |||||||||
|---|---|---|---|---|---|---|---|---|---|
| Front | Rear | Left | Right | ||||||
| Before | 0 | +1.0 | 0 | +7.5 | |||||
| After | 0 | +0.1 | 0 | +0.1 | |||||
The four bolts surrounding the column were loosened and downward pressure was applied on the front of the head. This caused the column to rock forward slightly and cracks opened between the column and base: as anticipated, it was not monolithic. 
The castings seem to have been bolted together, sanded and then filled with a soft material before painting. When the junction between the parts opened, small particles of this filler found their way into the joint without me realizing it. I added paper shims in the appropriate places to reduce the error and re-tightened the bolts. The error lessened but further adjustment was needed so I repeated the process. The error changed but not quite as expected. I did this several times before realizing that bits of filler were disturbing the process. I then scraped the joints clear with a thin metal piece (most of the filler was on the rear) and then lightly gripped pieces of paper in the mating joints and pulled them out to remove the filler dust. The external area of the rear joint was scraped and filed to further remove filler as part of the cleanup process. The picture shows where the filler broke away toward the rear of the right side; the edge of the paper shim is visible peeking out of the joint.
Once the filler was cleaned off, tramming was predictable and quick, being completed in about a half hour. I found an easy method of inserting shims which wasn't obvious to me initially. The head/motor is approximately balanced on the column; by pulling down on the front of the head with the retaining bolts loosened, the column tilts forward slightly. The column base has an area forward of the front bolts around which the column base cantilevers upward, raising both sides and the rear of the column off the base -- paper shims can be slipped into place easily on these 3 sides while applying downward pressure on the front of the head with the other hand. On my machine the mating surface between the mill base and the column base does not touch all the way across on the rear so shims work best on the sides of the column base. If shims are needed at the front of the base, replace the plate glass on the table with a block of wood, then use the quill's fine feed to apply downward pressure via the rod which will raise the head, allowing a shim to be slipped into place on the column front.
Once the alignment was approximately correct I completed tramming with the plate glass removed (more care is needed while turning the spindle to the measurement points because of the T slots) to eliminate that as a source of error. The "After" readings in the table were taken with the glass removed.
Update: the tram drifted slightly over 3 years so I recently re-trammed. Paper shims are a pain to remove because it is difficult to get at the area and they must be scraped off when they've been in place for years. Best to use brass or stainless shims to avoid this.