Eve Maler, UNIX Software Publications (
Wed, 11 Aug 93 14:31:18 -0400

I wanted to support Bob Stayton's suggestion about storing hierarchy
information in a flat way in links:

>In either case, I still would like to
>provide information to connect the <S3> section to its
>parent, children, and siblings. The very act of
>breaking up the large document means I have to
>provide some mechanism to indicate those connections,
>either with the LINK mechanism in HTML or as some
>IDREF attributes in the <S> tags. There should
>be enough information to reassemble the
>hierarchy starting from any point in the hierarchy.

Given the nature of a web, where you can make every chunk accessible
off the net, if the original document depended on hierarchical context
and containment to provide some information then you definitely don't
want to lose that information.

Hierarchy keeps going out of style (though it never quite gets all the
way there). However, I'll go out on a limb and say I'm still in favor
of it. :-) Putting in specialized IDREF attributes in <S> tags for the
purpose of recording their hierarchical context seems like a good thing
to do. (You could break the hierarchy in weird ways if you really
wanted to, by going around in circles or whatever; this would be
problematic for traditional hardcopy output if the pieces weren't
nested in the original source.)

I do wonder, however, why every last piece of a document needs to be
made accessible separately. Not everything can be written modularly
down to the last paragraph, and sometimes subtrees should simply travel
together. In the terms of Bob's example, why was that fictitious
section at level 3? Would it make any sense if someone were to read it
without reading all the levels above? If not, either it shouldn't be
accessible all by itself or it should be rewritten to be in less need
of its context.


Eve Maler
Digital Equipment Corporation, Nashua, NH, USA