There are some other ramifications i'd like to bring to like for others
that haven't yet tried the code
- Eliminates all extraneous usage of id's
- allows for refactorable presentation without changing references in
methods
- provides two way communication both up and down the chain by way of
the ._.parent attribute
- uses a seperate tag so that attributes used to name waypoints are
clearly seperable from "name" attributes
Here's a fun note; because eval() is not supported in DHTML (right?) i
removed the "at" attribute that would let you specify a waypoint to be
defined on something other then a nodes waypoint parent to support
aspect-oriented namespacing
i'm still wondering how mixins are progressing; i'm waiting on that to
consider how and where i use it in the library i'm offering to the
community.
Anyone else try it out?
P T Withington wrote:
I think you have an interesting idea here that addresses a deep problem.
`classroot` was an attempt to mitigate the issue somewhat, but does not solve
the general problem that you would like, in an arbitrarily complex nest of
nodes, to be able to address any lexically apparent name without having to
navigate explicitly up and down the node hierarchy. I have a placeholder bug
for this that I can't seem to locate just now...
