Re: Taxonomy draft - web replication problems

From: George Michaelson (ggm@dstc.edu.au)
Date: Sun Aug 22 1999 - 17:05:57 MDT


Here are some others.
  
                 Caching Replication
  
                cgi-bin refers cgi-bin can refer to
                to original replicant.

                cgi-bin output cgi-bin output may need to be
                refers to original munged to refer to replicant.

                unlikely to can exploit BASE override
                set BASE= in with base-relative URLs.
                header

                (ie cache calc (replicant-calc
                 per hit, not at start of nexus of hits
                 per nexus-of-hits) only)

                may not be URI will work for ftp/http forms
                neutral in many cases

                                        will work for alternate replicants
                                        eg if MIRROR.txt is being tried in
                                        order, can cover for 'sister'
                                         replicants
                        
                search db site-managed search DB may
                doesn't require require re-calc to reform resultant
                re-calc URI output

                per-hit ICP or other content assumed current. not checked
                refresh checks at hit-time.

                non-local fetch to non-origin fetch may be from another
                origin URL always replicant

                weak admin controls stronger admin control on replication

cheers
        -George



This archive was generated by hypermail 2b29 : Thu Nov 18 2004 - 11:21:26 MST