[Gzz-commits] manuscripts/storm short-paper.rst

[Benja Fallenstein]

CVSROOT:        /cvsroot/gzz
Module name:    manuscripts
Changes by:     Benja Fallenstein <address@hidden>      03/05/29 21:23:08

Modified files:
        storm          : short-paper.rst 

Log message:
        more, now first draft complete, slightly over maxsize

CVSWeb URLs:
http://savannah.gnu.org/cgi-bin/viewcvs/gzz/manuscripts/storm/short-paper.rst.diff?tr1=1.12&tr2=1.13&r1=text&r2=text

Patches:
Index: manuscripts/storm/short-paper.rst
diff -u manuscripts/storm/short-paper.rst:1.12 
manuscripts/storm/short-paper.rst:1.13
--- manuscripts/storm/short-paper.rst:1.12      Thu May 29 18:59:32 2003
+++ manuscripts/storm/short-paper.rst   Thu May 29 21:23:08 2003
@@ -13,18 +13,16 @@
    on the Web, they are a necessity for making Web-like hyperlinks
    useful on the local desktop.
 
-We propose Storm, a storage system identifying documents by
-cryptographic hashes and signatures, independently of their
-location. Our system automatically finds link targets wherever
-they are, on the local system or on the network.
-On the network, our identifiers are resolved 
-through a peer-to-peer distributed hashtable.
+We propose Storm, a storage system for use both on the
+desktop and the public network. Storm assigns a document
+a permanent URI when it is created. Using peer-to-peer
+technology, we can find documents even though our URIs
+do not include location information.
 Thus, links continue to work unchanged when documents are emailed
 or published on the network.
 
-Our system uses URIs to integrate with the Web. We have so far
-extended KDE and Netscape Communicator 4 to understand
-our experimental URN namespace. Most other systems can use Storm 
+We have extended KDE and Netscape Communicator 4 to understand
+our experimental URIs. Most other systems can use Storm 
 through an HTTP gateway.
 
 .. raw:: latex
@@ -44,10 +42,7 @@
 .. documents can be linked like web pages,
    which would make them part of the web:
 
-Bush's Memex [XXX] was to be a hypermedia device
-for personal use, not for hypermedia publication.
-
-Yet, while many Web pages are densely interlinked today,
+While many Web pages are densely interlinked today,
 desktop documents still have rarely any hypermedia
 functionality. Even though documents written with
 Microsoft Word or OpenOffice.org can nowadays be
@@ -59,19 +54,14 @@
 pieces of documents in discussion, or to
 refer from a short memo to a document
 with more in-depth information.
-Both individuals and workgroups could benefit
-from such functionality.
 
 .. links needed that don't break when documents are moved:
 
-We believe that a major burden to hypermedia 
-on the desktop is breaking links. 
-When sending a couple of hyperlinked notes per
-mail, the receiver would have to assemble them
-into the sender's directory structure. A document
-collaboratively authored by sender and receiver
-could not be linked to, except if both parties
-place it in exactly the same directory.
+A major obstacle to hypermedia 
+on the desktop is breaking links. A user receiving
+hyperlinked documents would have to reconstruct
+the original directory structure; otherwise the links
+won't work.
 
 .. using location-independent identifiers for
    non-breaking links:
@@ -79,42 +69,32 @@
 If desktop documents could be refered to independently
 of their location, it would be possible to make
 non-breaking links to them. When a document is published
-on the Web, it would have the same URI [XXX] as on the
+on the Web, it would have the same URI as on the
 author's desktop. When a link to a document is emailed,
 the receiver could follow the link as long as their
 computer can find the linked document-- on its local
 harddisk, attached to another email, or on the
 public Internet.
 
-.. contribution of this paper:
-
-This document describes Storm, a library for
-storing and retrieving data using 
-location-independent Uniform Resource Names [XXX].
-
-
-Related work
-============
-
 .. Freenet's cryptographical identifiers:
 
 Location-independent identifiers aren't a new idea.
 In Freenet [freenet-ieee]_, an anonymous peer-to-peer
-publication system, *cryptographic hashes* [XXX]
+publication system, *cryptographic hashes* [bakhtiari95cryptographic]_
 are used to identify versions of documents. Hashes
 are short, e.g. 20 bytes, yet it is extremely unlikely
 that two different documents will ever have the same hash,
 making them ideal identifiers on the Web.
 
-Hashes allow reference to a single version of a
-document. Documents are refered to by public keys [XXX].
-The private key corresponding to a document's public key
-can update the document by digitally signing a new version
-of it.
-
-However, finding a document on Freenet can be
-quite slow, taking many seconds. This is not
-acceptable for general Web publishing.
+.. Hashes allow reference to a single version of a
+   document. Documents are refered to by public keys [XXX].
+   The private key corresponding to a document's public key
+   can update the document by digitally signing a new version
+   of it.
+
+   However, finding a document on Freenet can be
+   quite slow, taking many seconds. This is not
+   acceptable for general Web publishing.
 
 .. non-breaking links seem not globally resolvable:
 
@@ -122,10 +102,6 @@
 assume that identifiers either have to
 say where a document can be found on the network, or they
 cannot be efficiently resolved on a global scale.
-URLs, location-dependent identifiers, break when documents are
-moved. Link services often query only a select set of link
-servers, not the whole network [hill94extending-andalso-carr95dls]_.
-
 Berners-Lee [name-myth]_ argues that for this reason,
 using unique, random-looking numbers to identify documents
 is not possible on a global scale.
@@ -133,9 +109,8 @@
 .. but DHTs can do it:
 
 However, recent developments in peer-to-peer systems have
-rendered this assumption obsolete. Distributed hashtables
-(DHTs) and other structured overlay networks
-[AspnesS2003-andalso-bonsma02swan-andalso-malkhi02viceroy-andalso-maymounkov02kademlia-andalso-ratnasamy01can-andalso-rowston01pastry-andalso-stoica01chord-andalso-zhao01tapestry]_
 
+rendered this assumption obsolete. Distributed hashtables (DHTs)
+[malkhi02viceroy-andalso-maymounkov02kademlia-andalso-ratnasamy01can-andalso-rowston01pastry-andalso-stoica01chord-andalso-zhao01tapestry]_
 
 allow location-independent identifiers
 to be resolved on a global scale.
 
@@ -144,13 +119,16 @@
 DHTs can not only be used to efficiently retrieve documents
 given their location-independent identifier, but also
 to retrieve metadata about the documents. For example,
+unlike many link services that only query a limited
+set of servers for links [hill94extending]_,
 all external links to a document can be found using
 a DHT.
 
-.. brief summary how DHTs work? XXX
+.. contribution of this paper:
 
-This, we believe, may be the most important result of peer-to-peer 
-research with regard to hypermedia.
+This document describes Storm, a library for
+storing and retrieving data using 
+location-independent URIs.
 
 
 Storm
@@ -164,11 +142,17 @@
 Our system, Storm (for STORage Modules),
 uses Freenet-like cryptographic identifiers.
 
+On the user level, Storm could be used as a
+replacement for both file names and HTTP URLs
+as document identifiers (although our implementation 
+uses both files and HTTP internally).
+
 .. part of the web -- URN scheme (so far experimental, 
    targetting registration):
 
-We are using an experimental URN (Uniform Resource Name,
-[XXX]) namespace and are preparing the registration
+We are using an experimental URN namespace
+(Uniform Resource Name, a persistent kind of URI)
+and are preparing the registration
 of an informal namespace with IANA. This allows our
 system's identifiers to be used in Web hyperlinks.
 
@@ -198,7 +182,8 @@
 Storm through the HTTP gateway. OpenOffice.org
 can load documents through the HTTP gateway,
 but cannot save to it, because the gateway
-does not support WebDAV.
+does not support WebDAV (HTTP-based distributed
+authoring and versioning).
 
 .. web interface for searching etc.
 
@@ -208,12 +193,37 @@
 to conduct searches over the titles of all documents
 published on the network.
 
+Our system implements append-and-delete-only
+storage; past versions of documents are stored
+unless deleted explicitly. It is possible to
+store only the differences between versions.
+
 
 Conclusions
 ===========
 
-..
+We have presented Storm, a storage system unifying
+the namespaces for private and published documents.
+This may help in making hyperlinks between desktop
+documents more convenient.
+
+A distributed hashtable is used to find published documents.
+While the core of Storm has been in use for almost
+three years, the peer-to-peer subsystem has only recently
+been implemented.
+
+Currently, a number of systems can use Storm
+through the HTTP gateway. A next step will be
+WebDAV support, allowing many more systems to
+save data in Storm.
+
+Storm can also be used to implement the more ambitious
+Xanalogical hypertext model [#]_ [lukka02guids]_.
+Storm is available as Free Software from
+``http://sv.gnu.org/projects/storm``.
 
+.. [#] Our implementation of Xanalogical storage 
+   is available at ``http://sv.gnu.org/projects/alph``.
 
 Acknowledgements
 ================