From: Marcin Kalicinski (kalita_at_[hidden])
Date: 2006-04-20 14:20:28
> You are right. In my initial enthusiasm, I overvalued per-tree default
> characters, and the problem you mention is indeed grave. I withdraw my
> Perhaps as a constant in the traits, then? I don't think the same
> problems apply.
But then trees with different constants would have different types and be
incompatible. That is good in the sense that you would be unable to insert
one into another, but it would seem weird not to be able to compare them,
for example. I think separator is a property of path dissection algorithm,
not the tree (not even the path), so any attempt to artificially affix it to
the tree will lead to trouble, sooner or later.
Boost list run by bdawes at acm.org, gregod at cs.rpi.edu, cpdaniel at pacbell.net, john at johnmaddock.co.uk