Neural review hangs up SuperMemo

Problem

Szymon reported:

When I try to use the 'Go neural' function, 'neural queue' or anything of the sorts it crashes my app. What is interesting is, it all works but in just one (my first) collection, when I try to use it on a newly made one (with few elements) SuperMemo freezes and closes after a few seconds.

Is there anything I can do to make it work? I tried repairing the new collections but there was no difference.

I'm asking because the neural learning style is one of my most desirable functions of the app and I would be really sad if I couldn't use it

Bug fix

This problem should not re-occur in SuperMemo 18 dated Feb 23, 2021 or later

Workaround

Add more memorized elements. The problem occurs only in very small collections (below 20 memorized elements)

Technical

When the neural queue fills up to a certain size, it is built no longer. Small portions might be added at each repetition. However, when the collection is very small, the process needs to be aborted if the queue includes all memorized elements. The problem was that the abort process was wrongly coded ("greater than" instead of "greater of equal than"). This is why collection with fewer than 20 memorized elements could easily run into an infinite queue expansion (which freezes SuperMemo).