4 ms·
Hi Iian, great work! I built recruitment CRM with cached 1,000,000+ nodes and can imagine how much time and efforts you invested in your project. I assume tha
by ilya1 11y ago
Hi Iian, great work!
I built recruitment CRM with cached 1,000,000+ nodes and can imagine how much time and efforts you invested in your project.
I assume that .NET MVC and KnockoutJS was your first choice as you get used to it ... and it's fine.
I personally found a lot of frustration dealing with Neo4J full-text search in multiple languages English, Russian and Japanese. There was no clear guideline on Lucene integration at the time.
How is your experience with Neo4J? Do you have to sync Neo4J data with traditional databases like M$ SQL?
Do you want to make your project as community efforts or see it as potential business?
The last question is much more important. I assume that a lot of developers won't be excited about .NET or KnockoutJS staff, however they might consider to share some crawling data, D3.js graph visualizations code, etc with you.
- argimenes 11y agoHi Ilya, Thank you for your kind words and your excellent feedback re: the chosen platform and the OSS community. To be honest I hadn't thought about the language and platform angle until you mentioned it, but it is definitely something I may need to consider. C# is a mature language comparable to most other quasi-imperative/quasi-functional ones so conversion I don't think would be hard so much as time consuming. The brain of the system is ultimately in the data structures, the Neo4j Cypher queries, and the product design/architecture. Can I ask what technologies you think would be most likely to attract community involvement, but wouldn't sacrifice static typing? Perhaps some Java framework for the back end, Angular or React for the clientside framework? I would not be able to immediately convert the project to those technologies, though, given the code count, but something to keep in mind ... Many thanks, Iian
- argimenes 11y agoHi again, To answer your other questions, I have no real plans to monetise this. I did try to raise funding for an 'art/culture travel app' that underpinned Codex but the focus was too niche and I didn't have a graphic designer. I think an open community driven project would be more useful, overall. Regarding Neo4j itself, most of my queries are run across the node paths rather than text so searches as such. So I haven't had to deal much with internationalisation issues (e.g., and searching for text with or without acutes, graves, etc.) The data in the system is pure Neo4j and not sourced from SQL of any kind. ;-)