On Thu, 18 Nov 2004 03:31:49 +0000, Andy Dingley
wrote:

On Wed, 17 Nov 2004 14:41:27 GMT, igor wrote:

I've also seen the $$ that can be spent for a router table fence, e.g. Incra.

Those were some of my wasted $$ 8-(

However, I
have also seen some fences for sale that have no more than jig screws/knobs
in T-tracks.

A fallacy in fences is that they have to slide. A swing is _much_
easier to arrange, and just as good.

I can see that, now that you mention it, except perhaps if one is using a
miter gauge in a table slot, it would seem.

There are different sorts of "accuracy"; precision, accuracy and
repeatability. For physical guides such as fences, we can also add
rigidity. As a user interface question, there's also adjustability.

Precision is the size of the calibrations. It's a good thing if
they're actually in the positions they claim, but that's not what's
measured by precision.

Accuracy is the question of whether the calibrations are really where
they claim. As most fences are uncalibrated, this is simply
irrelevant.

Repeatability is again only a matter for a calibrated fence. It's the
question of whether the same labelled position is in the same place,
no matter whether this is accurate or not.

[snipped for length]

Wow. Thanks for all that. I've read through your post twice and will read
it again. I appreciate both the "academic" points as well as the
practical. Very helpful. -- Igor