Top 10 iNat feature requests of all time

Forum Feedback
1

Since the Feature Requests topic was created back in 2019, some 1,283 feature requests have been submitted. I thought it might be helpful to take a look at what the most popular requests have been and where they stand today. Below are the top ten feature requests by number of votes. It is also, coincidentally, the list of all feature requests that have gotten at least 100 votes.

Make captive/cultivated not automatically “no ID needed”
172 votes
This was responded to by staff (tiwane), and the identify filters were improved. However, this did not directly address the specific request. According to loarie, this idea has been considered by staff, but he characterized it as “ambitious” and “disruptive”, so it seems unlikely to move forward.

Make computer vision include hybrid taxa on an opt-in basis
143 votes
This was recently replied to by staff (alex), who endorsed implementing limited support for hybrids on a trial basis. Because hybrids had caused problems with bird ID suggestions in the past, alex proposed that we enable it one taxon at a time and see what the results are.

Easy way to mark multiple-species observations
128 votes
This was implemented in 2024 as a new Data Quality Assessment to indicate “Evidence related to a single subject”.

Share observations between users
127 votes
This was responded to by staff (tiwane) who said it would require a lot of work and would be unlikely to happen soon. The main problem is that it would require reworking part of iNat’s fundamental data model. The request was tagged as “very-challenging”.

Automatically add a spectrogram view to observations with sounds
115 votes
This was responded to by staff (kueda) who said they would really like to implement it, but it was complicated, mainly due to the wide range of audio that would need to be accommodated (including bats). In the meantime, @japh built a chrome extension that automatically adds spectrograms to audio observations and has offered to let the iNat team integrate the tool directly into iNat.

Search and filter identifications
112 votes
This was responded to by staff (tiwane) who commented that it may eventually be implemented, but not any time soon. That reply was back in 2020, however. In the meantime, @pisum create a 3rd party tool at https://jumear.github.io/stirfry/iNatAPIv1_identifications.html.

Apply community consensus logic to annotations in the same way as for IDs
109 votes
Although this request hasn’t been responded to directly by staff, it has been marked with the “under-review” tag.

Duplicate prevention: Notify observers if their image checksums match others on the site.
105 votes
This was one of the first feature requests ever made. A month after it was posted, staff (tiwane) commented that a spec had been written up for the iNat team to review and see if the feature was feasible. No further updates were provided.

Change account deletion functionality to allow account anonymisation and prevent deletion of IDs
103 votes
This was responded to by staff (tiwane) who endorsed the idea, but stated that it would take a lot of careful planning to implement. He also explained that it would be tricky to anonymize @-mentions since these were just text and not stored in any special way. In a subsequent update, tiwane stated that “Anonymization option(s) is something we want to do, but as to when it would potentially be available I don’t know.”

Create a new taxon rank for iNaturalist: the species group
103 votes
This was replied to by staff (loarie), who stated “If this is an insect only problem, my preference would be to coerce ‘section’ and ‘subsection’ which is restricted to plants into this use case rather than add entirely new ranks.” Commenters then clarified that it was not an insect-only problem and that reusing “section” or “subsection” was not a viable option since those terms have a different meaning in zoological taxonomy. Loarie also brought up concerns with having different taxon nodes with the same name, a problem that already occurs with the taxon rank ‘complex’.

That’s it! Of the 10, 1 was actually implemented and 9 were responded to in some way by staff. Do folks feel like the feature request process is useful? Are there ways it could be improved?

28 Likes
2

I like the feature request category. Looks like most feature requests don’t get implemented because they are incompatible with how iNat is built, rather than staff not caring or not wanting to implement it. As most users, me included, have no idea how the underlying tech and code of iNat works, it’s not surprising to see such a big amount of feature-request “fail” at that hurdle.

My only gripe is the voting system. I’d like a more weighted vote. Perhaps some sort of poll underneath the feature request where users could vote for importance/urgency. A lot of feature requests seem like “nice to have” additions, whereas others are solutions for an issue that maybe should be addressed first.

7 Likes
3

Cool review, thanks for putting this together zygy!

Just noting that the latest discussion on this one was a staff member (alex) restarting experiments with opt-in plants.

Absolutely, I really appreciate the staff taking the time to consider these suggestions and explain why they may or may not be feasible.

4 Likes
4

You’re right! I totally missed that! I’ll fix it in the original post…

2 Likes
5

Then I am - ambitious and disruptive ;~)
Not Wild still Needs ID, and a (new) CID status.

I would like the 2 months back that I wasted on Suspended CNC - but that only has 34 … 36 votes.

4 Likes
6

My own most popular proposal just barely missed the top 10 at 98 votes.

1 Like
7

Note that those with the highest number of votes may have the highest number of votes at least partly because they mostly have not been implemented. They have possibly had more time to accrue votes than hypothetically much more popular requests that were fairly quickly implemented. So, this list may be biased in the direction of popular requests that have not yet been implemented.

I think it is awesome that there actually is a process and people can discuss, even if many requests are never implemented or don’t even make it past the first review to get posted.

For an assessment of how useful it is, it would be good to know how many feature requests actually have been implemented by iNat, how many a user has created a work-around for because it was suggested, how many are being considered for the near and far future, and how many have been rejected. I suspect those numbers would show that it has been a very useful process. A summary of that might be an interesting post someone could do to show both how far iNat has come and how many exciting improvements there may be in the future.

4 Likes
8

Couple of things would be helpful in the fungi space.

Ability to have numbers / letters automatically assigned to photos (if box ticked). This is because fungi have both macro photos and then microscope work. And when talking about the photos it is important to say what one is looking at eg chilocystida or pilocystida etc. And when commenters make comments one needs to know what photo they are talking about. Currently I get around this by manually entering letters if they are required.

More robustness to the AI suggestions. People now think AI is good at ID’ing fungi, since it has gotten good with plants and other things. With clear photographs and unique macro features the AI is excellent.

However many people don’t take good photos, especially beginners. And it is beginners who are most ‘gullible’ to AI suggestions and have strong belief that species level ID is possible just on photos. Which isn’t the case.

So having it suggest genus instead of species would be helpful. This would also enable googling / searching / drilling down into the genus to see what species exist and what it could be.

The other suggestion is to only have the AI suggest something once the geographical location is known. Since if geography isn’t know the suggestions often include species nowhere near my country and never likely will be. This would cull a lot of poor AI suggestions.

Another helpful thing would be to have the default setting for the AI to suggest only based on the specific country you are part of (or bio-region). Again a toggle option would enable people to choose between narrower and wider suggestions.

Part of the issue in NZ is that AI suggests things from Australia. This is much better than suggesting Northern Hemisphere species, but there is a lot of differences in fungi between the two countries (which for those who don’t know is minimum 2-3 hour flight one way over sea all the way, so we are not close in that regard.

9

In my first year on iNat we had this feature request. Activated. In two weeks (I remember it as overnight!) And I appreciate it every time I use iNat!

2 Likes
10

I just wish that the maps in different places (upload, explore, observation) weren’t “inter-sticky”. I only use satellite for uploading, for everything else I prefer OSM since my internet is sometimes too slow for satellite (and I also just really prefer the look of OSM).

11

It’s interesting to see the overlap (or lack thereof) with the top 10 feature requests based on number of replies. Some topics get lots of discussion but fewer votes. Here’s a listing of the top 10 based on activity/discussion (number of replies).

Change wording used by the system when downgrading an observation to an higher level taxa
254 replies / 18.1k views / 38 votes (not in top 10 above)
Obviously a perennially frustrating and controversial issue based on how frequently it pops up in discussions on the forum.

Easy way to mark multiple-species observations
218 replies / 11.7k views / 128 votes (#3 in the top 10 above)
Implemented/closed

Implement Photo Blur on observations annotated as “Dead”
149 replies / 7.4k views / 39 votes (not in top 10 above)
Declined/closed

Automatic iNat suggestion for “unknown” observations that reach a certain age
148 replies / 11.3k views / 57 votes (not in top 10 above)
Under review

Make captive/cultivated not automatically “no ID needed”
130 replies / 9.2k views / 172 votes (#1 in the top 10 above)

Don’t let an observation attain Research Grade if its location is very imprecise
126 replies / 6.4k views / 6 votes (not in top 10 above)
Lots of discussion compared to very few votes in favor.

Change account deletion functionality to allow account anonymisation and prevent deletion of IDs
121 replies / 2.3k views / 104 votes (#9 in top 10 above)

Leading subspecies IDs should change the obs taxon like leading IDs of other ranks
118 replies / 10.2k views / 76 votes (not in top 10 above)
Implemented (git-hub issue made) in 2023

Use OpenStreetMap maps
109 replies / 13.4k views / 77 votes (not in top 10 above)
Implemented (web) in 2024

Bulk Selection of Observations to Add to a Project
83 replies / 15.1k views / 34 votes
Declined for bulk additions other than one’s own observations, but implemented adding observations to a traditional project from the Identify page

And just to be thorough, I also checked for the top 10 based on number of views, assuming that maybe some topics may rise up that are of interest to the “silent majority” that neither votes nor participates in discussions but still searches for answers on the forum and reads those threads. Based on that, the following not captured by looking at number of replies would be in the top 10 list based on views:

Separate ‘Like’ and ‘Bookmark’ buttons
12.1k views (19 votes) - closed in 2019, not moving forward

Limit Cache Size or Allow Cache Clearing for Apps
11.8k views (32 votes) - closed in 2023, new request should be made for new apps if this is still an issue

Search and filter identifications
11.0k views (#6 in the top 10 by votes)

Recognize sounds automatically
11.0k views (4 votes) - closed in 2020 with a note that it might be implemented if feasible but improving image recognition will be given priority

8 Likes
12

This topic was automatically closed 60 days after the last reply. New replies are no longer allowed.