Re: chains vs. directed sets

• Subject: Re: chains vs. directed sets
• From: Paul Taylor <pt@doc.imperial.ac.uk>
• Date: Tue, 26 Nov 91 22:46:03 EST
• Cc: types
• Sender: meyer@theory.lcs.mit.edu

Date: Tue, 26 Nov 91 13:27:08 GMT
To: "John C. Mitchell" <jcm@cs.stanford.edu>

The reason for using directed sets rather than chains is much the same
as the reason for using typed programming languages rather than Kleene
indices: the class of them is closed under the useful operations
(products, for instance) whereas indices involve coding which is quite
irrelevant to the problem.

For example, in the limit-colimit coincidence, compare the notation of
Smyth & Plotkin [78] or [82] with my paper
        Homomorphisms, Bilimits & Saturated Domains - some very basic domain th.
[1987], which was unpublished, but is available by anonymous FTP from
        theory.doc.ic.ac.uk  (146.169.2.37) as theory/papers/Taylor/bilimits.dvi

• Prev: Re: chains vs. directed sets
• Next: re: more about kingdoms
• Index(es):
  • Main
  • Thread