take a look at the screenshot below. those were the requests that my browser was making before i killed the connection to the tab. looks like almost 60 requests within about 10 seconds, leading to a HTTP 429 (too many requests) error from the iNat servers:
i didn’t check your code to see what it was doing, but if it’s doing the same thing for other folks, that seems like a fast way to get folks potentially blocked by iNat.
also note that there’s a suggested daily rate limit, too.
…
even once you’ve addressed whatever is the rate limit issue here, it kind of seems to me like you should probably make this additional functionality optional, only making all these extra requests if the user really wants to see this information.
Oh, thanks, that’s a good idea - I made it like that. After looking a bit I don’t think the “top observer” is included in any of the endpoints taking multiple taxon ids, so there’s probably no easy way to get it other than one request per taxon.
that’s not the issue i’m describing. the issue is that the page didn’t seem to be limiting itself to roughly 1 request / sec. if you look at my screenshot from above, the page was basically kicking off another request as soon as the response for a particular request was received.
you said that you did implement some sort of limiting though. maybe it waits a few seconds when it encounters an error from the server before retrying? if that’s your approach, that still would be a less desirable implementation of rate limiting than implementing a delay before each request, since you could end up hitting the server with excess requests that it ends up refusing. more importantly, the guidance from iNat is that they’ll potentially block addresses that consistently exceed limits. so if your approach is to wait until you’ve exceeded the limits to slow down, that’s probably not a great thing.
Thanks! Yeah, on my slow internet each request already takes at least a second (or maybe inat servers have been rate limiting me for past abuses when I made a flower checklist generator in 2020 and made a lot of requests…) - but I made it wait an extra second after each request now (and at most 50 requests without clicking the button again). Should now be hard to hit any kind of limit
Mine is a gall midge, Neolasioptera portulacae. Two observations globally, and guess whose is the default picture? I don’t think it’s all that rare; but how many people pay attention to the galls on weeds?
Now my next one, Lipochaeta remyi – three observations globally – is rare (and critically endangered). It is a Hawaiian native plant, and to ID it, I keyed it out in Wagner’s Manual of the Flowering Plants of Hawaii. Unfortunately, I didn’t write down the keying steps I took, so nobody has seconded my ID.
In total, I have nine species with fewer than 10 total observations: 5 plants, 3 insects, and 1 bivalve. Interestingly, the bivalve has seven total observations, of which six were made by just one person – and one by me.
I only have one below 10 total, but part of that I think is due to me focusing on plants and revisiting them for additional observations. For some rare plants I have multiple observations but they are all the same individual during different seasons, so my obsession with documenting them multiple times over for phenology is ruining the stats…
My least observed non-captive observations appear to be lichens, which I usually try to make a species guess for but probably difficult to confirm without microscopy or some chemical testing. My least globally observed at Research Grade is a plant pathogen - seems there’s a common theme with parasites/pathogens not getting observed very often, or possibly getting observed but not getting ID’d to species very often.
Thank you for asking the question, @Ajott and @elias105 for the tools, and everyone for sharing your finds. I seem to be the sole observer for three plant species:
Chaerophyllum sessiliflorum, on the summit of Cradle Mountain, the best-known mountain in the state of Tasmania. It’s an 8 hour walk to get there, including scrambling up a scree slope, but in summer at least 100 people a day would walk within a few metres of these ornate little rosettes.
Ozothamnus ericifolius near a lake in the highlands. This was on a track that is popular with fishers, but I think bushwalkers ignore it because it goes past a dam, farms and an area recently damaged by bushfire. (Quite likely the plants in question only had room to grow because of the fire damage.)
Thelymitra imbricata (a type of orchid), in a modest little reserve next to the busiest highway in the state. I often stop there to walk around in the middle of a long drive and see what is flowering.
None of these three involved exotic feats of walking or navigation to get to. However, none of them stood out as unusual in the field, at least not to me, and presumably not to most other passers by. In each case I nearly didn’t upload them to iNaturalist at all! Two of the three identifications come from the state’s most qualified botanists, so I trust their accuracy, and thank you to all the identifiers out there who make our observations as accurate as possible. And keep documenting the odd corners where you all live, you never know what you’ll find…
