[Gzz-commits] manuscripts/pointers article.rst

[Benja Fallenstein]

CVSROOT:        /cvsroot/gzz
Module name:    manuscripts
Branch:         
Changes by:     Benja Fallenstein <address@hidden>      03/11/03 18:59:14

Modified files:
        pointers       : article.rst 

Log message:
        write out points in appl section

CVSWeb URLs:
http://savannah.gnu.org/cgi-bin/viewcvs/gzz/manuscripts/pointers/article.rst.diff?tr1=1.106&tr2=1.107&r1=text&r2=text

Patches:
Index: manuscripts/pointers/article.rst
diff -u manuscripts/pointers/article.rst:1.106 
manuscripts/pointers/article.rst:1.107
--- manuscripts/pointers/article.rst:1.106      Mon Nov  3 18:44:06 2003
+++ manuscripts/pointers/article.rst    Mon Nov  3 18:59:14 2003
@@ -379,31 +379,35 @@
 in a distributed, decentralized and above all
 *heterogeneous* way.
 
+.. XXX heterogeneous *how*?
 
-- standardizing on pointer records such as ours,
-  can provide full static-page (no CGI, no servlet) 
-  www features set on P2P or just about any type of system
-
-    - independent of underlying system, each client can use 
-      best methods to retrieve page. Analogous to the common HTTP
-      protocol: a language all types of servers and clients
-      can speak.
-
-       - can have normal filesharing (user decides)
-         but also oceanstore or
-         freenet-like semantics which store data decided
-         on by the overlay
-
-    - with ad hoc wireless local networks between mobile devices,
-      getting e.g. the latest slashdot could be cheaper,
-      as the neighbours' caches could be used; trust would not
-      be an issue due to the digital signatures.
-
-- need to use opaque pointer ids as addresses;
-  arguments for and against these are discussed in 
-  [balakrishnan03semanticfree]_
-
-    - should be possible to attach metadata to pointers
+Using pointer records, we can provide the functionality
+of static Web pages. While dynamic Web pages generated
+through server-side scripting and database connections
+are certainly important, this still makes up for 
+a significant portion of today's Web.
+(We may provide tools for handling simple cases
+such as showing a list of the newest articles on a news site
+through client-side scripting.)
+
+The data model used in this P2P Web is independent
+from the underlying mechanism used to find data;
+a client could utilize e.g. different overlay networks
+with different properties (such as speed and anonymity),
+the caches of other machines on the LAN,
+and the complete set of files ever downloaded to the local harddisk.
+This model is analogous to the HTTP protocol: a common language
+spoken by all clients and servers.
+
+In an ad-hoc wireless local network between mobile devices,
+the latest issue of e.g. Slashdot (an popular online news site)
+could be downloaded from other devices' caches, instead
+of paying for wireless Internet access; trust would not be
+an issue as the pointer blocks' signatures could be validated.
+
+The peer-to-peer network we propose would use
+random-looking, semantic-free identifiers; arguments for this
+are discussed in [balakrishnan03semanticfree]_.
 
 **A distributed wayback machine.**
 A surprising benefit of pointer records is that