How can we distinguish MILESTONEs from NOTEs? They seem identical.
I asked my colleagues how to distinguish a milestone from a note and they all threw up their hands. So it may not be easy. And it may not even be possible to distinguish them in a way that allows for their unambiguous inclusion in production process design. So what follows is an attempt at explanation, rather than an instruction. If you find that you can make use of it, do so. If not, we'll live without milestones. Personally I find them handy for things that look like they outline a structure, but which can't be tagged with the standard <HEAD> <TRAILER> etc. tags that mark the beginning and end of "real" (i.e. recognized) structural divisions.
<MILESTONE> tags can be used and abused in many ways. We prefer to restrict their use to cases in which there is a running alphanumerical sequence outside the main text block, containing no significant text other than that which explains the sequence, serving to break the text up into chunks, but not coincident with any structural divisions.
Because they appear outside of the text flow, they may be viewed as essentially a kind of <NOTE>, but a note of a very restricted repetitive, recurrent, and probably quite rarely used type. Since they are a kind of note, one can always fall back on the default and simple encode them as <NOTE>.
Here are some examples of things that I have treated as <MILESTONE>s in the past:
1. In Sidney's Arcadia, there are simple numbers (1,2,3,4...) that appear in the margins. They do not coincide with paragraph divisions, or any other division. (In fact, they correspond to numbers in the <ARGUMENT> at the head of the chapter, index numbers as it were.) They are not happily tagged as <NOTES> (though they could be), since they contain no content, just a numerical sequence that begins again at "1" with each chapter. They are not any kind of <HEAD> or <CLOSER>, since they don't appear at the beginning or end of any clear divisions.
2. Several books place numbers corresponding to points of sustained logical argument in the margins, to enable one to follow the argument better: the text makes the argument, the margin ticks off the points. Usually these are just numbers, or numbers combined with some kind of repeating text ("point 1, point 2, ..."). Though there is some correspondence here with the underlying intellectual structure, there is none with the typographic structure (points 1,2, and 3 may all appear in the same paragraph), so we can't use them as <HEAD>s. Again, we could tag them as notes, but this loses their distinctive quality of marking out chunks of the text in a series.
3. The 1850 edition of the Wycliffite Bible, which we captured in a previous project, treated the Biblical books and chapters as divisions, with <HEAD>s, but the numbers belonging to the Biblical verses were simply printed in the margin, with no other typographic indication that the verses were regarded as <DIV>s, or even as <P>s. So the verse numbers could not easily be treated as <HEAD>s; instead, we asked that they be treated as <MILESTONES>, since they were a numerical series outside the text block that served to break the text up into sections not corresponding to structural divisions.
4. Many books contain running marginal cross-references to the page numbers or chapter numbers of another edition. Since these are not the page numbers or chapter numbers of the book being captured, they can't be tagged as <PB> or <HEAD>. They could, again, be captured as <NOTE>, but this would ignore their simple, sequential quality--their function as providing a kind of alternative structure. We normally ask that these, if they are simple, recurrent, and sequential, be treated as <MILESTONE>s.