High speed keyboard data entry leads to mis-selection

Website with Firefox 93 on Windows 7

Several thousands of African plant ids are currently at level “Plantae”. I tried to quickly go thru the list and give ids. Most of them are Magnoliopsida, and I cannot give a better id than that.
See https://forum.inaturalist.org/t/mission-impossible-identify-plantae-in-africa/43528

How to reproduce:
I open the URL for Plantae outside the cape region:
and identify quickly by using the keyboard.
I.e. hit i, wait shortly, type zw (German Zweikeimblättrige means Magnoliopsida), wait shortly, cursor down, enter, enter, cursor right.
As you can reproduce, “Zweikeimblättrige” shows up as the first item in the list. Oddly, rather often, the 3rd or 4th item of the list (cicadas, a rabbit) get selected, e.g. in

This may be related to bug report

That explains why I found one of your rabbits up a tree.
But perhaps instead of powering thru dicots?

1 Like

This has been a consistent problem for me on touchscreen. It seems like it’s not about the tap speed, but that the web page has decided to interpret the selection tap as pointing to a different item in the pulldown, other than the one visually displayed under the tap location.

I find I can reproduce it over and over within the same page at times, which is bewildering when it’s at like, species level and some subspecies gets chosen for me instead repeatedly, when it’s the only other item in the pulldown.


I’m not able to replicate. @bernhard_hiller would you be able to make a screen recording of this? You can share it here or with help@inaturalist.org. I was able to detect a replicate a bug I was seeing in Identify the other by making a screen recording. Sometimes it captures clues that aren’t really explainable in words.

Oh, wow, that happens to me all the time! Touchscreen “always” selects a different option to the one I touched, so I can’t use it. Have to stick to the mouse. From what I’ve seen, it selects whichever part of the drop-down list of options is displayed over the top of the “Save” button, rather than the option I actually touched.


Oh wow, I will watch out for that on mine, thanks!

Sorry, I do not have screen recording software.
And beyond that, the bug generally gets not visible to me during the identification session, only much later on when someone else goes thru the observations at “Matter of Life” level, gves a correct identifictaion, and then I receive a notification.

I type with really high speed - that was the point why reproducing the Therac bug failed for the developer who never typed as fast as the technicain operating the system.
Sometimes the list is not yet rendered when I hit cursor down, and selection fails (that’s expected behavior then); also during the next steps, some key strokes seem to get lost sometimes, and the page does not move ahead. And only on such occasions do I see when a selection error has occured before which can still be corrected before submitting the id.

Where does the list get sorted - on server side or client side?
What is the sorting key? It is not alphabetcially sorted.
Does it get transfered as a whole, or as a couple of sets?
How is rendering done? By binding the complete sorted list, or item after item?

How does list replacement work? After typing z, a request is sent for taxa starting with z, and when w gets typed, a new request is sent for taxa starting with zw. What about when the list for z has not yet arrived when the list for zw gets requested?

Also, today, the list shows 10 items, while the days before it was shorter.
I’ll try a workaround: keep Magnoliopsida in the clipboard, and type shift v or shift insert instead of zw, because the list will then contain only the correct item.

I don’t know where the problem lies. But I often see comments from people who are mystified by the ID attached to their name. ‘Rabbit up a tree’ had me wondering if it was a deliberately wrong ID.
Can it be a slowly sliding pixel problem like the @mention list lurking just off the bottom of my screen, after a few obs. I can see the blue border, perhaps the tops of the letters.

(But still Dicots in Africa - we have 33K already - we need help to move them to family)

Our internet, especially for a data hog like iNat is spectacularly slow, as two of the undersea cables are damaged and won’t be repaired till mid to late September. I type fast, click enter, and wait … till I know that iNat and I agree? Then the next one.

SInce I do not have much experience with African plants (more with european or south-east asian plants; and then also with flowers visible…), I won’t make many improving ids. But among Plantae, it is some 20-25% which I can identify to something different (either better or a different area like ferns, monocots, birds, grass hoppers, etc), mainly Fabaceae followed by Asteraceae, and many Amorphophallus (if those ids are correct - no notifications on further ids yet).


Pre-Maverick Fabaceae in Africa ?