better inat widgets

[rmcminds]

would like to highlight eg the rarest taxa i have observed, or anything that i specifically mark as interesting. maybe some summary stats. things that are new to me.

other people's work examples:

https://forum.inaturalist.org/t/using-atom-feed-on-website-need-help/25610

https://github.com/jumear/stirfry

https://observablehq.com/@robin-song?tab=notebooks&type=public

[rmcminds]

probably best practice to use data fetching github action, not live javascript with api. would result in only daily retrieval and wouldn't rely on user allowing scripts in browser.

[rmcminds]

Other ideas:

most phylogenetically distinct taxa

Most recent new taxa

[rmcminds]

https://www.inaturalist.org/observations.json?user=rmcminds&updated_since=2023-07

sync full dataset export single time, then just run this each time to fetch updates? would want to record the time used each time so that the next update gets everything after that. also want to ensure observations that are deleted entirely are recorded.

[rmcminds]

the above request also doesn't seem to include brand new observations, only old ones updated. so a separate call with something like:

https://www.inaturalist.org/observations.json?user=rmcminds&d1=2023-07-01

would be needed to get those (need full date including day)

[rmcminds]

This is a fun tool that I could find some inspiration from: https://forum.inaturalist.org/t/a-tool-to-help-you-fill-local-data-gaps-easily-missed/37575/65?page=3

• Maybe do something similar but with time instead of or in addition to space (eg things that used to be common that no longer are...). Might be legitimately interesting for finding orgs that have disappeared?

Also inverse. Find mis ids or local things that may have been recently introduced?

[rmcminds]

Better formatting with this:

https://developer.mozilla.org/en-US/docs/Learn/CSS/CSS_layout/Grids

Use for both iNat and bookwyrm

[rmcminds]

either the inat api has been updated, or i missed some things in the past. apparently 'v1' is the old one, but is faster...? and both apis seem to have plenty of parameters accepted for the 'obserations/species_counts' endpoint, which does a lot of what my own code does. probably much faster to use it. there's even an option that seems to filter observations to just the first time they're observed - exactly what the output is for my site.