/[gzz]/manuscripts/storm/article.rst
ViewVC logotype

Diff of /manuscripts/storm/article.rst

Parent Directory Parent Directory | Revision Log Revision Log | View Patch Patch

revision 1.121 by benja, Sun Feb 9 08:49:00 2003 UTC revision 1.122 by benja, Sun Feb 9 09:40:51 2003 UTC
# Line 40  rendered this assumption obsolete. Struc Line 40  rendered this assumption obsolete. Struc
40  [ref chord, can, tapestry, pastry, kademlia, symphony, viceroy,  [ref chord, can, tapestry, pastry, kademlia, symphony, viceroy,
41  skip graph, swan] allow location-independent identifiers  skip graph, swan] allow location-independent identifiers
42  to be resolved on a global scale.  to be resolved on a global scale.
43  It is now feasible to do a global search for all information  It is now feasible to do a global search to find all information
44  about a given identifier, on any peer in the network.  about a given identifier, on any peer in the network.
45  This, we believe, may be the most important result of peer-to-peer  This, we believe, may be the most important result of peer-to-peer
46  research with regard to hypermedia.  research with regard to hypermedia.
47    
48  We examine how location-independent identifiers can support *data mobility*.  We examine how location-independent identifiers can support *data mobility*.
49  In today's computing world, documents move quite freely between  In today's computing world, documents move quite freely between
50  computers: being sent as e-mail attachments, carried around on disks,  computers: they are sent as e-mail attachments, carried around on disks,
51  published on the web, moved between desktop and laptop systems,  published on the web, moved between desktop and laptop systems,
52  downloaded for off-line reading or copied between computers in a LAN.  downloaded for off-line reading or copied between computers in a LAN.
53  We use 'data mobility' as a collective term for the movement of documents  We use 'data mobility' as a collective term for the movement of documents
54  between computers (or locations on one computer, such as folders),  between computers (or folders!),
55  and movement of content between documents (through copy&paste) [#]_.  and movement of content between documents (through copy&paste) [#]_.
56    
57  .. [#] While the physical mobility of e.g. notebooks may effect  .. [#] While the physical mobility of e.g. notebooks may effect
# Line 68  can be recognized whereever it is moved Line 68  can be recognized whereever it is moved
68  .. [#] It might be more appropriate to speak about *resources*  .. [#] It might be more appropriate to speak about *resources*
69     and *references* instead of *documents* and *links*, but     and *references* instead of *documents* and *links*, but
70     in the spirit of [ref kappe95scalable], we stick with     in the spirit of [ref kappe95scalable], we stick with
71     the latter terms for explanation purposes.     the simpler terms for explanation purposes.
72    
73  *Dangling links* are an issue when documents are moved  *Dangling links* are an issue when documents are moved
74  between servers; when no network connection is available,  between servers; when no network connection is available,
# Line 85  to documents stored on either one's comp Line 85  to documents stored on either one's comp
85    
86  Advanced hypermedia systems such as Microcosm and Hyper-G  Advanced hypermedia systems such as Microcosm and Hyper-G
87  address dangling links through a notification system [ref]:  address dangling links through a notification system [ref]:
88  When a document is moved, servers storing links to it are notified.  When a document is moved, a message is sent to servers storing links to it.
89  Hyper-G uses an efficient protocol for delivering such notifications  Hyper-G uses an efficient protocol for delivering such notifications
90  on the public Internet.  on the public Internet.
91    
92  Location-independent identifiers for documents  Location-independent identifiers for documents
93  make such notification unnecessary; a peer-to-peer lookup system  make such a system unnecessary; a peer-to-peer lookup system
94  can resolve documents whereever they are moved.  can find documents whereever they are moved.
95  Such a system also works for data not publicized on the Internet.  Such a system also works for data not publicized on the Internet.
96  For example, if one email has a document attached to it, and another email  For example, if one email has a document attached to it, and another email
97  links to this document, an index of locally stored documents  links to this document, an index of locally stored documents
# Line 190  Section 8 concludes the paper. Line 190  Section 8 concludes the paper.
190    
191  The dangling link problem has received a lot of attention  The dangling link problem has received a lot of attention
192  in hypermedia research [refs]. As examples, we examine the ways  in hypermedia research [refs]. As examples, we examine the ways
193  in which HTTP, Microcosm [ref], Hyper-G [ref] and URNs [ref]  in which HTTP, Microcosm [ref] and Hyper-G [ref]
194  deal with the problem.  deal with the problem.
195    
196    .. XXX and URNs [ref]
197    
198  In HTTP, servers are able to notify a client that a document  In HTTP, servers are able to notify a client that a document
199  has been moved, and redirect it accordingly [ref spec?]. However,  has been moved, and redirect it accordingly [ref spec?]. However,
200  this is not required, and there are no facilities for  this is not required, and there are no facilities for
# Line 208  a document is moved or deleted, a messag Line 210  a document is moved or deleted, a messag
210  to the filters. Linkbases implemented as filters can  to the filters. Linkbases implemented as filters can
211  update their links accordingly. A client selects a set  update their links accordingly. A client selects a set
212  of remote filters to use. Only links stored by one  of remote filters to use. Only links stored by one
213  of these filters can be found.  of these filters can be found by the client.
214  [HymEbook?]  
215    .. [HymEbook?]
216    
217  .. Microcosm systems can independently choose  .. Microcosm systems can independently choose
218     whether to import filters from other systems, and whether     whether to import filters from other systems, and whether
# Line 218  of these filters can be found. Line 221  of these filters can be found.
221     for example in a workgroup.     for example in a workgroup.
222    
223  In Hyper-G, documents are bound to servers, and a link  In Hyper-G, documents are bound to servers, and a link
224  is stored on the servers of the two documents it connects  between documents on different servers is stored by both servers
225  [kappe95scalable]. This ensures that all links to a document  [kappe95scalable]. This ensures that all links from and to a document
226  can always be found, but requires the cooperation  can always be found, but requires the cooperation
227  of both servers. Hyper-G employs a scalable protocol  of both parties. Hyper-G employs a scalable protocol
228  for notifying servers when a document has been moved or removed.  for notifying servers when a document has been moved or removed.
229  A server hosting links to this document can then ask  A server hosting links to this document can then ask
230  the link's author to change the link, or at least the link  the link's author to change the link, or at least the link
231  can be removed automatically. The *p-flood* algorithm  can be removed automatically. The *p-flood* algorithm
232  employed by Hyper-G for this purpose guarantees that a message  employed by Hyper-G guarantees that a message
233  is delivered to all interested servers, but requires that each  is delivered to all interested servers, but requires that each
234  interested server keeps a list of all the others.  interested server keeps a list of all the others.
235    
# Line 239  knowing only a document's location-indep Line 242  knowing only a document's location-indep
242  it would be possible to find both the document and links to it,  it would be possible to find both the document and links to it,
243  no matter which peer in the network they are stored on.  no matter which peer in the network they are stored on.
244    
245  XXX Say something about the usual resolvable URN approaches  .. XXX Say something about the usual resolvable URN approaches
246    
247    
248  2.2. Alternative versions  2.2. Alternative versions
# Line 262  strong security mechanisms and uses peer Line 265  strong security mechanisms and uses peer
265  as basis of communication channel among limited amount of participants.  as basis of communication channel among limited amount of participants.
266  Neither of these systems supports the immutability of data.  Neither of these systems supports the immutability of data.
267    
268    .. thesis-benja: remove paragraph above
269    
270  [ref HTML version format proposal] Alternate versions important for  [ref HTML version format proposal] Alternate versions important for
271  authoring process [search refs]. (Note: Keeping track of versions  authoring process [search refs]. (Note: Keeping track of versions
272  structure is also \*hyper*media. Refs?) (WebDAV!)  structure is also \*hyper*media. Refs?) (WebDAV!)
# Line 270  structure is also \*hyper*media. Refs?) Line 275  structure is also \*hyper*media. Refs?)
275  2.3. Peer-to-peer systems  2.3. Peer-to-peer systems
276  -------------------------  -------------------------
277    
278    .. thesis-benja: check what needs to be rewritten below
279    
280  During the last few years, there has been a lot of research  During the last few years, there has been a lot of research
281  related to peer-to-peer resource discovery, both academical and in the industry.  related to peer-to-peer resource discovery, both academical and in the industry.
282  There are two main approaches: broadcasting [gnutella1, kazaa, limewire,  There are two main approaches: broadcasting [gnutella1, kazaa, limewire,
# Line 360  responsible for a hashtable item its 'ho Line 367  responsible for a hashtable item its 'ho
367  .. Should we discuss applications of p2p systems (CFS, OceanStore, Squirrel, ...)  .. Should we discuss applications of p2p systems (CFS, OceanStore, Squirrel, ...)
368     here? If so, which ones?     here? If so, which ones?
369    
370    .. thesis-benja: remove paragraph below
371    
372  CFS [ref], which is built upon Chord DHT peer-to-peer routing layer[ref], stores  CFS [ref], which is built upon Chord DHT peer-to-peer routing layer[ref], stores
373  data as blocks. However, CFS *splits* data (files) into several miniblocks and  data as blocks. However, CFS *splits* data (files) into several miniblocks and
374  spreads blocks over the available CFS servers. Freenet [ref] and PAST [ref],  spreads blocks over the available CFS servers. Freenet [ref] and PAST [ref],
# Line 370  property which is used in Storm blocks. Line 379  property which is used in Storm blocks.
379  Recently there has been some interest in peer-to-peer hypermedia.  Recently there has been some interest in peer-to-peer hypermedia.
380  Thompson and de Roure [ref ht01] examine the discovery  Thompson and de Roure [ref ht01] examine the discovery
381  of documents and links available at and relating to  of documents and links available at and relating to
382  a user's physical location. For example, this could include  a user's physical location. An example would be
383  a linkbase constructed from links made available by different  a linkbase constructed from links made available by different
384  participants of a meeting [thompson00weaving].  participants of a meeting [thompson00weaving].
385  Bouvin [ref 02] focuses on the scalability and ease of entry  Bouvin [ref 02] focuses on the scalability and ease of publishing
386  of peer-to-peer systems, examining ways in which p2p can serve  in peer-to-peer systems, examining ways in which p2p can serve
387  as a basis for Open Hypermedia, while our own work has been  as a basis for Open Hypermedia. Our own work has been
388  in implementing Xanalogical storage [ref 02].  in implementing Xanalogical storage [ref 02].
389    
390  At the Hypertext'02 panel on peer-to-peer hypertext [ref],  At the Hypertext'02 panel on peer-to-peer hypertext [ref],
391  there was a lively discussion on whether the probabilistic access  there was a lively discussion on whether the probabilistic access
392  to documents offered by peers joining and leaving the network  to documents offered by peers joining and leaving the network
393  would be tolerable for hypermedia publishing. For many documents,  would be tolerable for hypermedia publishing. For many documents,
394  the answer probably is no; however, for personal links,  the answer is probably no; however, for personal links,
395  comments and notes about documents, this behavior may be acceptable,  comments, and notes about documents, probabilistic access may be acceptable,
396  especially since this kind of publication would not require  especially when seen as a trade-off against
397  setting up a webspace account first and could therefore  having to set up a webspace account before publication.
 encourage publication.  
398    
399  In the end, some peers will necessarily be more equal than others:  In the end, some peers will necessarily be more equal than others:
400  Published data will be hosted on servers  Published data will be hosted on servers
# Line 415  we are able to check we have the correct Line 423  we are able to check we have the correct
423  the cryptographic hash in the identifier. Therefore, we can  the cryptographic hash in the identifier. Therefore, we can
424  safely download blocks from an untrusted peer.  safely download blocks from an untrusted peer.
425    
426  While digital signatures also allow for self-certifying identifiers,  While digital signatures also allow for self-certifying identifiers
427    (the identifier could contain a public key, and we could check
428    the signature after downloading a block),
429  they raise the need for a public-key infrastructure (PKI)  they raise the need for a public-key infrastructure (PKI)
430  and for a timestamping mechanism in order to be reliable  and for a timestamping mechanism in order to be reliable
431  for more than a short time.  for more than a short time.
432    
433  When we make a reference to a block, we can be sure  When we make a reference to a block, we can be sure
434  that even the original author of the target will not be able  that even the original author of the target will not be able
435  to change it. For example, if a newspaper refers to a letter  to change it (contrast this with a signature-based scheme).
436    For example, if a newspaper refers to a letter
437  to the editor this way, the letter's sender won't be able to change  to the editor this way, the letter's sender won't be able to change
438  the reference into an advertisement for a pornographic web page.  the reference into an advertisement for a pornographic web page.
439    
# Line 595  hitting a 'publish' button moving it to Line 606  hitting a 'publish' button moving it to
606  possibly also uploading it to a server permanently connected  possibly also uploading it to a server permanently connected
607  to the Internet, if one is available.  to the Internet, if one is available.
608    
609  Unfortunately, we have not found a satisfactory representation  We have unfortunately not found a satisfactory representation
610  of zones yet. In particular, how do we decide which zone  of zones yet. In particular, how do we decide which zone
611  a new block should be in? Probably in the private zone  a new block should be in? Probably in the private zone
612  in many cases, but if we have been editing a document  in many cases, but if we have been editing a document

Legend:
Removed from v.1.121  
changed lines
  Added in v.1.122

savannah-hackers-public@gnu.org
ViewVC Help
Powered by ViewVC 1.1.26