MHS-DS: Changing the tree (OUCH)

["Harald T. Alvestrand" <[e-mail]>]

With regards to formalizing my suggested "simple" approach:

I checked the drafts text, and I think the place we need
to change *IF* we want to move to the "missing components are
explicitly represented" approach is the "infotree" draft - the
one we thought most stable :-(.

The changes should be (relative to infotree-03.txt):

Chapter 1, section 1, page 1:

Add the paragraph:

 Missing O and PRMD fields are represented by the value " "
 (single space). It is legal to let these RDNs be aliases pointing
 back to the entry above them, so that the Distinguished Name
 PRMD=" ", ADMD=d400, C=DE is an alias pointing to
 ADMD=d400, C=DE

(This is really a hack for backwards compatibility, but lets Michael
off with adding only 1 alias rather than redefining his tree....)

Bullet 7, page 7, is changed from:

 o  ADMD will always be present in the hierarchy.  This is true in the
    case of `` '' and of ``0''.  This facilitates an easy mechanical
    transformation between the two forms of address.

into:

 o  ADMD, PRMD and O will always be present in the hierarchy.
    Missing attributes are represented by a single space.
    This facilitates an easy mechanical transformation between the
    two forms of address, and allows association of routing or mapping
    information with the fact that an attribute is missing.

Other documents should remain basically unchanged, but the examples
need to be examined with care.
What do people think?

                    Harald Tveit Alvestrand