I have an idea for a predicate that would require automated submission via the API, as such a predicate would need to be updated often, for example once a day. I’m wondering if such predicates would be in scope for Golden? As an example, those could be statistics about a company that can be tracked on a time interval basis. I can think of some ideas where it could be useful knowledge for end users, that could be tracked fully automatically, by stitching together a few APIs. But I’m not sure if Golden protocol is the right place to track such statistics over time?
Good question. Right now Golden is focused on building up a set of canonical, verifiable data. While the protocol will support mutation, working with high-frequency data is likely further down the roadmap. I could imagine something like company quarterly filings being possible without much trouble, where pricing tick data is likely much later (if ever).
Are there specific examples that you have in mind?
I was thinking about things like the size of the Discord community or number of stars on GitHub. Those might be interesting signals for cryptocurrency related projects, which are as strong as their community.
Such signals can be easily collected through a set of API calls, but they mostly make sense when collected at regular time intervals, to create a growth graph.
Yeah, I agree that is interesting data and it makes sense to capture it in some form. When time-series data is involved or when data is expected to change the current stance is to use qualifiers on those predicates. It is a good idea to specify the minimum length of time to allow between qualifier dates at the time of the predicate’s proposal. This could be voted on later to be made shorter as protocol capability allows.
Another potentially valuable qualifier would be ‘expected lifespan’ which could specify how often a predicate should be checked for updates. This could apply to longer-term information like employment records as well.