There has been a revival of the data bridge linking iNaturalist records to the UK system iRecord this week which has triggered conversation about potential issues to overcome moving forward.
One crucial problem is that it’s a one-way street - we have the national hoverfly scheme leader (and probably others) checking over 1000s of records but there is no way for their verification to be passed back to iNaturalist. It’s a real shame for the UK iNaturalist records not to receive this feedback I think.
“BRC have discussed the possibility of this with the developers of iNaturalist and a resolution is currently not available.”
Can anyone tell us what limits this part of the data link ?
I would have imagined this was not a problem on the iNaturalist side, as adding comments and IDs appears possible through the API …
The identifiers on iRecord need to have their own iNat accounts and post the IDs to iNat. If they want it to be done from iRecord, they need to use the iNaturalist API to build that pathway to get the identifications back to iNat.
It’s definitely possible to post identifications via the iNaturalist API, so long as the identifiers have an iNat account. As far as the GDPR, iNaturalist is compliant with the GDPR (according to @ahospers). As long as iRecord gets consent from users to communicate their data to iNaturalist, I don’t think there should be a problem. That said, the GDPR is a very complicated law, so it would be good to know what the specific concern is.
@sbushes: This may be a radical suggestion, but has iRecord considered migrating to iNaturalist as a platform? Many years ago, I assisted iSeahorse in migrating from their own homemade observation system to iNaturalist, and they have been happy with the results. Maintaining an observation database and associated interfaces is a huge effort. Plus having observations fragmented across multiple platforms is problematic (as you have already noticed). Are there critical features of iRecord that are not available on iNaturalist?
I had a response on one of the Facebook groups from John van Breda :
“It’s not technically impossible, but iNaturalist would not allow iRecord to connect to the API via a single “system” authentication to post the feedback back. Therefore we would have to write additional code to allow verifiers to add their iNaturalist login details into their iRecord profile. Expecting all the iRecord verifiers to register on iNaturalist and set up the link in iRecord isn’t really feasible, so we’d end up with the situation that some verification feedback would make it’s way back to iNaturalist and some wouldn’t, which would be even more confusing.”
Short version :…Never going to happen! :D - at least not in the near future.
Longer version : …A real breakdown of why would be a bit tangential to this thread I believe, but happy to detail my take on it on another post if you want to know more. I do think how sites like iNaturalist (designed for observers) and iRecord(designed more for verifiers) can or cannot work in tandem is an interesting topic. But essentially, as I understand it, iRecord is a decade or so in the making - trying to bring together 100s of different national recording schemes / records centres under one umbrella to centralise UK data-gathering and verification. When it comes to total migration, I doubt the issues in play would bear much similarity to those iSeahorse encountered tbh - due to the differences in approach, as well as the number and range of different voices powering the system, many of whom seem significantly opposed to use of iNaturalist themselves.
that still might be better than not pushing the identifications over though, right? on the iRecord side, you could just mark which records were sent over and which weren’t. that would help eliminate some of the confusion, and then if a non-linked user established a link, then that could trigger all the unsent records to be sent.
I would have thought so - from my POV anything is better than nothing.
I struggle to see the POV from the verifiers tbh, as it’s not possible to know what this interface physically looks like on their side. It seemed to me as if some were already under the impression their comments get sent back to the recorder. I also don’t see that’s it’s such a big deal to ask verifiers to have an account here if they wish their comments to be sent upstream - it’s mutually beneficial as it will limit mistakes in IDs down the line, help fix the AI, etc.
Last year I sent my records over myself so I was able to see feedback on everything and then pull comments or IDs back to iNaturalist myself where needed. This year, I have no idea which records have and have not been verified. I can search within iRecord and find records with my iNaturalist name on them, but not with any ease as it’s not even possible to search by recorder name at present, only location/taxon. I just realised I can actually look inside these records to see comments though (if there are any). I would have thought it would be possible to connect my iRecord account to my iNaturalist profile from the iRecord side somehow so the data is assigned to me and I can receive notifications on the other platform at least.
Could it be tagged on the iNat side in theory to link to the external datapoint, as the ones sent to GBIF are?
It just seems to be a missed opportunity for both platforms at present.
Hopefully there is a better resolution down the line.
Actually, it wouldn’t have to be that complicated from the user’s POV. iNaturalist is an OAuth2 provider, so in theory, the verifier on the iRecord side could just check a checkbox that says “push this verification to iNaturalist”. The first time they did that, it would take them to the iNat site to log in (if they weren’t already logged in), and then it would offer them the choice to authorize iRecord with access to their iNat data. Then the two accounts would be connected. It’s similar to the workflow you get when you log into a discussion site or comments section using your Google or Facebook account. Same technology. So they wouldn’t have to enter their iNat username and password into iRecord at all.
Integrating OAuth into iRecord is either easy or hard depending on if you have any experience with integrating authentication services. OAuth2 is the industry standard (and a huge improvement over previous systems) and there are lots of software libraries out there you can basically just plug into your website. If you’ve never dealt with 3rd party authentication, though, it can be a little bit of a learning curve. If they’re interested in looking into this further, point them to the Authorization Code Flow documentation.
Nice, thanks for this - I remain a little confused myself around what the exact limitations here are - I will pass this on. I think John is correct that some verifiers will be reluctant though. There is significant opposition to iNaturalist by some members of the community.
I mean that records could be stored on iNat and integrated in their website, or if it’s too hard to code, then it could be done with simplier links. That way all the work (schemes) part is left as it was, but data is in both places without duplicating it, also all ids done on iRecor would actually be on iNat records without doing anything more. But it’d be a real integration of iNat page, not some automatical reposting of records under one account as many websites do, so like if you open a page there’s be a window with this observaion in this page.
Still not sure what you mean exactly (maybe something like iFrame (?)) …
In general though, anything which means the verifiers have to do anything other than what they normally do on iRecord will be a cause of frustration ( some are already burnt out with existing workload, etc)…so there’s just no room to add extra effort into process for some. Although if what @zygy says is correct - that it’s potentially just a single login and then no need to have to do anything again, then I’d imagine it’s something the majority of verifiers would rather have than not.