The vanishing of a fellow iNatter

I know you are already aware of this, but for the benefit of others who may encounter this topic, just want to be clear that sockpuppet accounts are expressly forbidden by iNaturalist, and subjects their creators to having all of their accounts suspended.

11 Likes

How do you know the IDs you are cloning are correct? Are you going to vet the deleted accounts to determine which ones have IDs that are worth preserving and which ones do not?

Not all the people who delete accounts are going to be experts on the taxa they ID. And not all experts are created equal.

Blindly cloning IDs may do as much harm as good.

1 Like

They would be saved as withdrawn, that means you could check observation and see it, it’s what is missed from those delete accounts, there were disagreements, but can’t expect every observer to remember where iders disagreed with each other.

2 Likes

We don’t know if the ID’s are correct when they’re posted now. How would the deleted accounts be worse? Or better? We do know that deleting accounts removes ID’s, in some cases thousands of ID’s and in many cases those ID’s were useful.

7 Likes

Agreed. Also, I see the historic IDs and comments as adding context to an observation, and I use that context as part of my ID decision making even if some of the prior IDs are inaccurate.

So from an ‘information value’ perspective I think an obs is more than the sum of its individual parts.

For example if someone is suggesting an outlandish ID, it will make me at least think why it is not that ID. More thinking is better than less.

I basically see deletion of IDs and comments as removing useful context from the observation. This makes me sad. Well, sadder :upside_down_face:

Therefore I fully support preserving incorrect IDs (and comments) in any future iNat feature to address the overall bulk deletion issue (whatever that may look like.)

7 Likes

I’ve been trying to hold off on replying until I had something more concrete to say about potential options moving forward, but I want to address a few things:


Using software to re-post IDs seems like it would probably violate iNat’s prohibition of “Machine generated observations, identifications and comments”. It also goes against the original content creator’s wishes to have their content removed.


As far as using the API for some sort of backup solution for your observations, that seems fine but note that if a lot of people use it or it becomes a burden it may have to be limited.


GDPR isn’t really a sticking point here, at least as far as account/data deletion is concerned with iNat. I suppose my saying

iNaturalist will always allow users to entirely delete their content from the site, both as a matter of principle and because laws like GDPR would not allows us retain data by users who no longer want us to.

a few years ago was not as clear as it should have been, but I meant that on principle iNat allows users to delete their content from iNaturalist, separate from GDPR. GDPR and other laws can inform us and must be taken into account, but they’re not a reason or cover for this choice. I think there are many legitimate and reasonable views on this topic, but I don’t think they’re particularly relevant to this discussion as the option to completely remove one’s data from iNat isn’t going anywhere. That’s why, much earlier in this thread, I said this is about “[providing] some options for people who don’t want to participate any longer or distance themselves from iNat.”

So if/when we add anonymization options, there will still be an option to delete everything. Whether that would have made a difference with recent account deletions, who’s to say, but those people deleted their accounts knowing what would happen and had the option of just not using iNat, so it was a deliberate choice.

I also do want to apologize for putting anonymization on the backburner for so long, as I said it was something I’d work on. But it’s being discussed and evaluated by iNaturalist staff at the moment.


This is difficult stuff. If someone wants to keep their observations and photos, do we force them to change the licenses to the photos to CC0? Exactly what text will be displayed for an ID from a deleted account, and will it be used as part of the community taxon? How could this happen at scale so that it won’t cause disruptions elsewhere? If a some sort of grace period were given, how can the account be restored with as little impact as possible? Right now it involves taking down a server for a day and things like bug fixes and new functionality can’t really be tested during that time.

Even with simple upates there are often bugs, and bugs for something like this could be really disruptive so it would have to be done right.


This thread is getting really long. I thought about closing it out soon but instead I’m just going to put it on slow mode for now.

17 Likes

Four hours between posts, WOW!

So apparently, “the more vague, emotional side of things” entails squabbling. And a desire to control what other users can and can’t do.

Let us not forget: we each are here of our own free will. We do not owe iNaturalist – or other users – anything other than compliance with TOS as long as we choose to stay. It isn’t as if anyone here has bought the rights to our content.

Honestly, this conversation reminds me of someone contesting a will because they didn’t inherit what they wanted. Maybe they want to change probate law to force decedents to bequeathe things a certain way. “The more vague, emotional side of things” gets pretty ugly when the emotions are “but I wanted…”

This thread has evolved into something that I no longer wish to participate in, so this will be my last reply in it.

2 Likes

If you do end up closing it completely, could you leave it visible, unlike this one which I don’t think even appears as searchable anymore?

With over 300 comments, it’s one of the more popular subjects that people seem very eager about, so there’s clearly something here that may be worth looking at when considering what direction(s) iNat will go in the future.

4 Likes

Can you explain how you are defining “an option to delete everything”? Or, to put it a different way, is staff prepared to consider modifying what “deleting everything” looks like, at least in the case of IDs/comments?

One of the things I have been trying to emphasize is that the current practice causes disruption not just because of the deletion (loss of research grade, etc.), but because of the form that deletion takes: removal without even any record that a comment/ID previously existed. The “invisibility” of the deletion is a problem, not just the deletion itself. It erases the history connected with an observation that was created by someone other than the user deleting their account: it affects other people’s content. As I see it, the repercussions go beyond a user’s right to delete their own data and it is important to minimize the impact on others who had no say in this user’s actions.

It seems to me that replacing deleted IDs/comments with a message “an ID/comment by a deleted account has been removed” (or some other formulation suggested in this thread) would still fulfill the criteria of allowing users to delete their content while preserving a record and providing an explanation for users who are currently left quite puzzled about what happened (Wasn’t there an ID or am I misremembering? Why are there suddently a bunch of old “needs ID” observations? Why does this comment thread seem incoherent?).

I would argue that a message indicating “an ID of x species was suggested by a deleted account” would also fulfill the criteria of “deleting everything”, but if staff disagree with this interpretation I would consider the first option an improvement over the status quo.

8 Likes

It sounds like they are going to preserve the current nuclear option, but add more options (I would assume including anonymisation-related ones) to allow departing members the ability to take a less dramatic approach to departure. They are probably thinking that while some people who leave will still use the nuclear option, some will not, instead opting for a less destructive approach if offered, and therefore the damage to the data set will be reduced. However they will continue to offer the nuclear, maximum damage option out of principle.

Unless there are some meaningful metrics on the reasons why people have left in the past, or their preferred departure action (do nothing, anonymise my data, delete all my data) this doesn’t have any guaranteed outcomes. It might work out better than the current situation where the only 2 options are walk away from the account or delete the account. Or it might not. Time will tell.

4 Likes

Yes, I understood that allowing users to delete everything is going to continue to be an option. It was pretty clear to me from the beginning of the discussion that this would likely be the case. It’s not what I would necessarily prefer, but I accept it. I am concerned with something slightly different here.

What I do not consider tenable is the policy of deletion without a trace – i.e., pretending it didn’t happen and acting as though the IDs or comments were never there in the first place.

The fact of deletion itself (not the specific deleted content) should be a part of the record. This is quite normal practice both in internet communities and research/archival contexts.

Or, to continue your metaphor: if the nuclear option is possible, then I want the holes and rubble that it leaves behind to be visible. Not covered up and erased from the map.

5 Likes

The nuclear option is the nuclear option. Anything that leaves an audit trail is not nuclear, therefore it is a different, new option. They may temper the existing nuclear deletion but I doubt it. Therefore it will probably mean the introduction of additional options along a spectrum in between the current nuclear ‘total wipe’ and the current ‘leave all my data and user ids as they were’. One of which would probably include your desired ‘audit trail’.

I think it was a good idea to slow this thread down. Personally, I think on the whole it’s been a good discussion of a complex and important issue. I’ll put my view on record because @sedgequeen has done an excellent job of holding the fort for my own view and I should give some support!

Any person leaving is a problem and will cause problems. We like people, and we want them to stay :)

This is partly about giving people more choices when leaving, but partly also about reducing the negative impact on the community. Nevertheless there will inevitably be problems, so any changes to iNat should be about choosing the best problems.

  1. The current nuclear option is the worst of all worlds. If it needs to be retained as an option on principle (which I can understand) it should at least result in a note “ID/comment by deleted user” on observations so that the remaining part of the interactions remains at least partially intelligible. (Not - Why was this person debating with themselves?)

Assuming people are given the choice to retain at least their IDs/annotations/DQA in some form…

  1. The option of leaving "Deleted user ID’d this as X-us y-us, and removing it from the community taxon calculation is an improvement, but not by much. It still results in the total number of their IDs needing to be replaced, potentially vastly increased NeedsID piles overnight.

  2. Leaving the IDs in the community taxon calculation means that extra work will be needed on some observations whose ID needs to be turned around. But this will only affect IDs that were wrong or become subject to the minority of taxon changes that don’t shift the IDs automatically. This is a problem - but in my view it is by far the smallest and best problem; and it is a problem we already have to tolerate in the much more common situation of users who are simply inactive.

I do fear that to genuinely anonymise an account the content of comments may have to go, as we can’t be sure they haven’t said something in a comment that might identify them…

Therefore I support more options for departers, evidence of deletion where the nuclear option is chosen, any content that remains with consent (especially IDs) to continue to ‘count’. (annotations are a problem because they can’t be outvoted - perhaps annotations by ‘deleted user’ could be removed in the event of a thumbs down).

Over and out.

18 Likes

I am very sad to hear this. I heard it a while back too from a presentation I was on (about using iNat to examine the biodiversity of beetles in Texas) and someone said something about this – I am deeply sad that he chose to leave, especially because he had so many valuable posts and identifications. Beyond his contributions, too, he was a really nice guy to talk to and I talked to him a few times. I hope whatever he’s doing now his life is going well.

3 Likes

P.S. Could a mod/the OG poster add the name of the iNatter in again? I heard it was removed, but this is not a bad instance in which to use a name to talk about somebody. His name was [username removed by moderator] by the way.

Respectfully, when someone chooses to leave, they have made that choice for a reason, so to undermine their personal agency by coming on after they have left and saying their name and rehashing their situation is poor form, even if done with kindness and friendship.

Please remove the name @silaseckhardt . When in doubt, do not post the personal situations of others.

7 Likes

Yes, when in doubt it’s best not to name individual users in regards to personal choices and issues that could be construed as negative. The OP made the decision not to do so and keep the thread focused more generally on thoughts inspired by a deleted account and the issue more broadly. I would suggest deleting the user’s name from your post.

7 Likes

Is that the papal or editorial “we”?

Please look at their profile; their pronouns are listed at the top (we/us-they/them pronouns).

17 Likes

“That being said i do periodically download a spreadsheet with all my species IDs and location, because i am paranoid i would lose that due to some unforeseen issue.”

Does iNat provide a link to do this or do you have to go through all your IDs and put them in a spreadsheet one by one

I am also kind of paranoid