RE: Proposal for matching negated sets (was Re: New Public Review Issue: Proposed Update UTS #18)

From: Philippe Verdy (verdy_p@wanadoo.fr)
Date: Sun Oct 07 2007 - 07:49:20 CDT

  • Next message: Michael S. Kaplan: "Re: Proposal for matching negated sets (was Re: New Public Review Issue: Proposed Update UTS #18)"

    > De : Dominikus Scherkl [mailto:lyratelle@gmx.de]
    > Envoyé : dimanche 7 octobre 2007 14:04
    > À : verdy_p@wanadoo.fr
    > Cc : 'Mike'; 'Andy Heninger'; 'Mark Davis'; 'Unicode'
    > Objet : Re: Proposal for matching negated sets (was Re: New Public Review
    > Issue: Proposed Update UTS #18)
    >
    > Philippe Verdy schrieb:
    > > Mark, the question is not "if sets don't have an implied ordering". By
    > > definition a set is a unordered thing. There does not exist any ordered
    > set.
    > Of course there exists ordered sets!
    > If you put an ordering on a set, it still remains a set.

    No! It becomes a vector or list and behaves very differently : you need
    special code to handle insertions (that require reordering, or inefficient
    representation as a list through costly pointer indirections, or costly
    copy-on-move operations).
    And you do not need maintaining the order offsets at runtime for allthe
    internal steps of computing the output. It'sbest to sort the output only at
    end.



    This archive was generated by hypermail 2.1.5 : Sun Oct 07 2007 - 07:52:32 CDT