Niels, the old CLA is still valid, great to get this going! Cheers,
/peter neubauer GTalk: neubauer.peter Skype peter.neubauer Phone +46 704 106975 LinkedIn http://www.linkedin.com/in/neubauer Twitter http://twitter.com/peterneubauer http://www.neo4j.org - Your high performance graph database. http://startupbootcamp.org/ - Öresund - Innovation happens HERE. http://www.thoughtmade.com - Scandinavia's coolest Bring-a-Thing party. On Mon, Jun 13, 2011 at 3:03 PM, Niels Hoogeveen <[email protected]> wrote: > > Hi Peter, > Thanks for the positive feed back. > I think it would be best to call this component Neo4J-collections, or > something to that extent, so data structures more constrained than a property > graph can be maintained. The BTree (Timeline index) is one such example, the > RTree from the Neo4J-spatial component is indeed another such structure. > There are many more classic data structures that can be modeled in a graph, > and that have several constrains that can be met in the API. > eg. > Lists: nodes can have at most one incoming and one outgoing relationship of a > certain type.Stacks: Lists that can only be appended on one side and where > nodes can only be removed from that same side Queues: Lists that can only be > appended on one side and where nodes can only be removed from the other side > Trees: ...Binary Trees:...Red-Black Trees: ....etc. > I am willing to separate the Btree related code from the Neo4J-index > component. I haven't looked into the code of Neo4J-spatial, but it would be a > good idea to separate the RTree related code and put it in this new > component. That way, RTrees can be used without having dependencies on Geo > Tools. > In the past I already signed a CLA when I did some work on the Meta Model > component. Please let me know if I have to renew that CLA, or if the old one > is still valid. > Kind regards,Niels Hoogeveen > >> From: [email protected] >> Date: Mon, 13 Jun 2011 14:27:32 +0200 >> To: [email protected] >> Subject: Re: [Neo4j] Neo4J-collections... In-graph Timeline index, again >> >> Niels, >> taht sounds like a great suggestion. I just talked to Johan about >> supporting in-graph indexing structures in the indexing framework. >> Also, the whole Neo4j-Spatial index is essentially an in-graph >> indexing structure. >> >> So, I would suggest that we put a component with in-graph "simple" >> indexing structures into place (maybe just timeline for the time >> being) and take this as a strating point to make the index framework >> respect them. If you send in the CLA according to >> http://wiki.neo4j.org/content/About_Contributor_License_Agreement, I >> can create the component stub adn you can put it in there? >> >> Cheers, >> >> /peter neubauer >> >> GTalk: neubauer.peter >> Skype peter.neubauer >> Phone +46 704 106975 >> LinkedIn http://www.linkedin.com/in/neubauer >> Twitter http://twitter.com/peterneubauer >> >> http://www.neo4j.org - Your high performance graph database. >> http://startupbootcamp.org/ - Öresund - Innovation happens HERE. >> http://www.thoughtmade.com - Scandinavia's coolest Bring-a-Thing party. >> >> >> >> On Mon, Jun 13, 2011 at 2:17 PM, Niels Hoogeveen >> <[email protected]> wrote: >> > >> > As one of the legacy users of the in-graph Timeline index, I have concern >> > over the future of this "component". >> > Right now the in-graph Timeline index is part of the component >> > neo4j-index, which also contains the legacy Lucene index. >> > This component has a dependency on Lucene 3.0.1, while Neo4J 1.4 depends >> > on Lucene 3.10. >> > Of course it makes no sense to keep upgrading legacy code to newer >> > versions of Lucene. >> > As a result of that, applications depending on the in-graph Timeline are >> > stuck with Neo4J 1.3, without there actually being a good reason for that. >> > Last week I already wrote a message about this issue but did so over the >> > weekend, triggering little response. (see: >> > http://lists.neo4j.org/pipermail/user/2011-June/008935.html) >> > In a message early May, I suggested the creation of a component >> > Neo4J-collections, where all sorts of in-graph data structures can be >> > grouped together. (see: >> > http://lists.neo4j.org/pipermail/user/2011-May/008461.html) >> > I hope my suggestion will somehow lead to a solution where the in-graph >> > Timeline index can become part of the standard Neo4J infrastructure. >> > Kind regards, >> > Niels Hoogeveen >> > _______________________________________________ >> > Neo4j mailing list >> > [email protected] >> > https://lists.neo4j.org/mailman/listinfo/user >> > >> _______________________________________________ >> Neo4j mailing list >> [email protected] >> https://lists.neo4j.org/mailman/listinfo/user > > _______________________________________________ > Neo4j mailing list > [email protected] > https://lists.neo4j.org/mailman/listinfo/user > _______________________________________________ Neo4j mailing list [email protected] https://lists.neo4j.org/mailman/listinfo/user
