Very nice presentation. It would have been better if the camera focused more on the slides. Several times, wanted to look at the board to follow what he was saying.
Great presentation, I thoroughly enjoyed watching it. I've been diving into Neo4j recently, and I loved how this talk not only covered the introductory topics really well but also went into common pitfalls, anti-patterns, and where to look for advanced material (like the graph algorithms). It's clear this guy knows his stuff!
Great Initiative by Microsoft. It was a good interactive session, one guy asked a question which I totally had in my mind. Concise and straight to the point. Really looking forward to Neo4j.
I love the presentation, and I'm interested in using Neo4j. What turns me off a bit (as a researcher) is the Cypher language. It seems to be overcomplicated SQL (but it might be just my "noobish" impression). I want to write simple queries like in 23m:39s : "Who loves Ann?" ;) and db engine figures out > "Ann" probably {Person}, "loves" probably {label} etc... Does such a parser exist?
Not saying nothing possible but it it's difficult don't forget it that way Use match The car gets bound to max vars in Database Merge Get or create Merge = 1 Want to create a unique node property constraints from the start Create constraint on person Assert exists label name Join created at query runtime Live data query, still compute it every single time Need more joins query complexity grows Adding new data, and relationships, where you get into the NoSql sql schema bound is good on data integrity but bad in evolvability and Agility Hard to redefine relationship on fly and redefine a graphive data, everytime you recompure the data doing lot more scans than strictly necessary Data lineage problem Are as acyclic graph Give all info from data sourced by source X Recursive sql using dried procedures Cipher defined relationship hops Problem is grsphy thus intuitive by using neo4js It mimics a read only svheme Values only computed once When data was written, and never again Traversing data is cheap because it's about pointer dereferencing
Great Video. I have a question - Say I have 15 different legacy RDMS database instances where the data related to each other reside and these have millions of records.If I implement Neo4J here, my intention is to traverse the graph efficiently and being able to pull the relationships. I don't want to create a monolith and huge Neo4J database that consists of data from 15 of the legacy databases . I rather want to leverage existing data stores but just want to create a graph that I can traverse for ascertaining the relationships.Please let me know if you can answer this.
This is an excellent presentation, both in content and quality. Only thing is as Ramarao commented, some presentations of this kind have the current slide as a window in window along with the presenter, this is enormously helpful in following and understanding what the presenter is saying. But, nevertheless, an excellent introduction with meat (second part).
Content is excellent...but dont no why the camera is mostly focused in the tutor instead of the slides...it was a bit difficult to go through for this reason.
Basically schema migrations won't have a place here right? only data migration matters which should be done by code which is the case also for entity framework migrations.
I also came across something else, I know the neo4j engine and cypher is really cool but I really don't like to put the queries and commands inside a string variable in VisualStudio with no inellisense support and no syntax highlighting. It will be problematic for a big project and for the beginners. I checked that there is a neo4jclient library available but it doesn't look that much promising really and you have OGM on Java only. Is there any plan to improve on this part?
Why do you keep referring "Google" this and google that while you are presenting on Microsoft world :) LOL !!! Please no offense to anyone, but bing is also an awesome search engine. But man, this is awesome presentation. I am working on a hack project and I have few ideas to build some solution that I can build using Neo4j especially with cypher query language. Thanks for uploading this video.
Google has become kind of standard vernacular for many people to say "search on the web". It's something pretty much everyone does lol. Even if you're at microsoft it's a hard habit to get rid of!
