Info on coordinate interpolation mitigation test?

Today I noticed that an opt-in for interpolation testing has appeared at the bottom of observation pages.

I opted in and I did notice a couple of UI changes that I assume are intended to make it harder for someone with bad intentions to discover the true location of an observation whose location is obscured.

Is there any info on what’s being tested here, or a thread discussing the pros and cons? There’s been a fair amount of discussion of obscuration in the forum and the key principle always seems to be achieving the best balance between strong protection for the small number of organisms where precise location info is a genuine threat while retaining usability and openness for all the rest.

2 Likes

I don’t think there is any particular announcement about this test yet, and I would encourage the discussion here to not go into any details that would make it easier to circumvent location obscuring.

4 Likes

Yeah, I’m trying to think of a good way to talk about the pros and cons of various changes without providing a guide to defeating obscuration. More likely, it won’t get discussed until after the changes are moved into production and then we’ll see if any of them have impacts that far outweigh their protective value.

2 Likes

this is a breakdown of what’s been changed: https://github.com/inaturalist/inaturalist/commit/938c742703d301d8639bf0d77de376eee18ff123.

(i don’t think there’s anything particularly secret about what’s described there, but feel free to flag / hide this post if you think otherwise.)

i’m just guessing, but just based on what’s been changed, there’s probably not a lot of harm to discussing the impact of any particular change on your particular workflow (ex. i will no longer be able to do this), as long as you’re not trying to compare it to how much protective value you think a particular change provides.

3 Likes

Thanks for finding that! From what I can tell, all those changes make sense. They improve the implementation of location obscuration.

We still need to find better ways to accurately assess which taxa require obscuration (and in which parts of their range). Protection status is a good start, but only a minority of protected species are put at greater risk through precise location info. In many cases, better location data can result in more interest and better protection. Accurately determining which approach to take often requires labor-intensive enquiries.

2 Likes

Sorry, I meant to make some kind of announcement but then realized there’s no way to communicate just with curators in the Forum (the test announcement is only visible to curators on the site). If you find bugs, please feel free to report them. The problems these changes are trying to mitigate are kind of an open secret, but I’d rather not have this thread serve as clear directions for how to circumvent them, documented for perpetuity in public where search engines can index them. Perhaps I should start a DM thread in the Forum and include anyone here who’s either a curator or who I know who wants to participate?

5 Likes