dependabot[bot] on npm_and_yarn
adityamukho on development
Bump path-parse from 1.0.6 to 1… Merge pull request #6 from Reca… (compare)
adityamukho on tests
Added ArangoDB v3.8 to build ma… (compare)
@adityamukho This may be a silly question but where is is the "path" field described? I'm trying to figure out what to put there.
@ggendel Please have a look at https://github.com/RecallGraph/RecallGraph/wiki/Terminology#path
can i write RecallGraph like queries myself , so i can make AQL queries work?
@kapilpipaliya You can theoretically write AQL queries directly on RG collections once you're familiar with their document schema and relations, but I would advise against it since this structure is likely to undergo significant alterations during these early days of feature and performance enhancements. On the other hand, the API would remain relatively stable, and does support quite a few powerful filtering capabilities. The upcoming version will make this even more so, by integrating filter expressions into almost every read endpoint.
This is not yet reflected in the documentation, but if you're interested in trying out the absolute bleeding edge, try working with the dev-raw branch. Let me know if you need help with your queries.
@adityamukho Thanks. Performance is always a hot-spot in my work. As long as you keep it backwards compatible, I'm thrilled. For our regression testing, I can truncate your tables at the start of each test so I'm not too concerned. I'm concerned about resources in an active database that runs for years so we may want to have a way to (at least) archive ancient data. I'm not concerned about someone trashing the tables since users have to go through another application layer to get to the database so we can lock down this feature absolutely necessary.
@ggendel Understood. I will take up archiving as a roadmap item, once the internal event log structure has been stabilized (should be by the next 2 releases).
create. It might reject updates otherwise (haven't verified behaviour for pre-existing nodes). What you're describing would eventually be handled by 'explicit commits' feature I have mentioned in the roadmap. How much of a priority is this for you? I'm currently working on enabling APM traces in the application, following which I was planning to focus on the 'valid time' dimension. But, I can re-prioritize (once the tracing instrumentation is finished).
path = '/') and update its event log with entries to represent the existing object graph (
CREATEevents in this case)
@/all The new documentation website is live!
Check it out at https://docs.recallgraph.tech/ and do leave your feedback!