Roadmap Discussion (0.16.x and beyond)

[adsharma]

Thanks for contributing to the previous round of discussions in #122

Here's what I'm thinking:

Queries

• subgraphs + open type support can be made transparent. CREATE + MATCH should output the same regardless of the type of the subgraph.

• Improve duckdb as a foreign table #272

Storage

• Feature: stabilize on-disk storage #35
• Optimization: reduce space amplification #19
• Import duckdb source as a library into ladybug so we can perform joins without going through SQL

Icebug Integration

• Define and document icebug memory format
• Streamlined query to create Icebug memory from a subgraph
• Implement notebooks to demo how cypher on Icebug memory works

CI

• Reduce the number of times C++ sources are built (once for each lang, platform, python version). Reuse as much as possible

[adsharma]

Some optimization ideas from #301

[aheev]

Icebug Integration

working on it

[aheev]

#328
#329