1.1

PMS currently has the following:

> The package manager must not impose fixed limits upon the number of
> version components. No integer part of a version specification may
> contain more than eight digits. Package managers should indicate or
> reject any version that is invalid according to these rules.

Historically, Portage had weird bugs for excessively long version
parts, especially when leading zeroes were involved, although as far as
I'm aware it's clean with arbitrary lengths now. Paludis is clean (and
issues a QA notice for violations), but portage-utils fails the
whole version handling thing in an epic fashion.

Given this, do we still need that restriction in place? There're quite
a few violations in the tree.

-- 
Ciaran McCreesh