Matt: The idea behind Xindice P2P is to provide a JXTA service that more or less provides the same interface as the XML:DB API. But instead of querying or updating a single database, you query and update the entire peergroup.
Peers can replicate popular content, making it faster and more reliable to locate certain documents. If you were getting urn:oid:1.2.3 from a peer running on someone's laptop and they went out of range of their WiFi provider, you could get urn:oid:1.2.3 from another peer that had a copy. You could then host urn:oid:1.2.3, too for other peers. Using JXTA membership services you can create peergroups for specific organizations or interests, and control who can join and thus access or update content within the group. JXTA itself provides secure, TLS-based transport for applications that require secure transfer of documents. JXTA also solves the problems of tunneling through firewalls and over NAT boundaries. --Sean > -----Original Message----- > From: Matt Liotta [mailto:[EMAIL PROTECTED] > Sent: Wednesday 12 June 2002 11.26 AM > To: [EMAIL PROTECTED] > Subject: RE: replication > > > Please do tell more about your ideas for a peer-to-peer XML database. I > am currently using Xindice for web application component persistence. > This means I need to have a single Xindice instance shared by all of my > web application servers. I would rather have a Xindice instance per web > application server that is able to replicate with the other Xindice > instances. > > -Matt > > > -----Original Message----- > > From: Sean Kelly [mailto:[EMAIL PROTECTED] > > Sent: Wednesday, June 12, 2002 11:21 AM > > To: [EMAIL PROTECTED] > > Subject: RE: replication > > > > That sounds like something similar to what I'm working on: combining > > Xindice > > with JXTA. The idea is to make a peer-to-peer XML database. > > > > For my application, globally unique identifiers for documents are > URIs. > > That allows me to use URNs like urn:oid:1.3.1.6.7566... for an > > OID-identified document or urn:uuid:7e1f1218ab800128... for a > > UUID-identified document. > > > > > -----Original Message----- > > > From: Matt Liotta [mailto:[EMAIL PROTECTED] > > > Sent: Wednesday 12 June 2002 10.52 AM > > > To: [EMAIL PROTECTED] > > > Subject: replication > > > > > > > > > I am not sure if anyone has thought about replication for Xindice, > but I > > > can certainly see some applications for it. One of the first issues > for > > > dealing with replication is obtaining a globablly unique identifier > for > > > each document in the database. As far as I can tell, this can be > simply > > > done by concatenating the unique identifier of the document in a > > > collection with the unique identifier of the collection in the > database. > > > Is this correct? > > > > > > -Matt > > > >
