I take particular issue with the idea that alignment errors are additive. They are not.
Suppose the mitre/blade alignment is carried out and the error is X. Suppose after that, a fence/mitre alignment is carried out with error Y. Then the fence/blade error is X+Y.
On the other hand,
Suppose the mitre/blade alignment is carried out and the error is X. Suppose after that, a fence/blade alignment is carried out with error Y. Then the fence/blade error is Y, not X+Y.
It's easy to disprove me: offer a counter example. Just draw out an example when they are not additive.
If we boil the arguments to the fundamentals, here is a statement to which all should agree:
The fixed reference points on a table saw are the miter slots and the plane of the table itself.
Nope, there are no fixed reference points. The frame of reference is a choice. If you need one, you can make it whatever.
Again, focusing on fundamentals, here are our objectives as I see them:
* The sides of the rip fence should be parallel to the miter slots and perpendicular to the table. This should remain true at any rip fence position and whether the rip fence is on the left or right side of the blade.
* The miter gauge should guide the wood to be cut in a line that is parallel to the miter slots. The face of the miter gauge should be perpendicular to the table.
* As the blade cuts, the blade tips should form a fixed line in the plane of the table that is parallel to the miter slots. This should remain true at any blade height or angle.
Your three fundamentals are rooted in single fundamental: For less overall tearout, less burning, and lower risk of kickback, the wood ought to move parallel to the blade.
From that fundamental, we derive: the mitre guage and rip fence ought to be aligned so as to constrain wood's movement to that direction.
That being said, there are scenarios where it's desirable for the the wood to move at angle to the blade.
Where most of us seem to disagree with unburled is the alignment process itself.
Actually, the disagreement concerns not the entire mitre/bevel alignment process, just the various techniques for measuring techniques for that alignment.
Related to that, we all recognize that no alignment is perfect,
and we disagree on what to focus on for best results.
Actually, the disagreement concerns the various techniques for measuring techniques for the mitre/bevel alignment.
I agree with most of the others that the best approach is to align the rip fence
Off topic. Read the original post, then look who drifted the thread with fence alignment.
unburled wants to align the rip fence with the miter slots, then align the blade with the rip fence.
Not true. Quote me. Off topic.
The root of this issue is related to unburled's assumption that alignment errors are additive, which they are not.
If a mitre/blade alignment is carried out with error is X, and then a fence/mitre alignment is caried out with error Y, then the fence/blade error is X+Y. To think otherwise is simply wrong.
It doesn't really matter. If unburled does a decent job of aligning the rip fence with the miter slot and the blade with the rip fence, then the blade should be reasonably well aligned with the miter slots, and we can all get back to cutting wood again.
That is a big 'if' ... because I've never done that. Off topic.