coll feature request: retrieve data at index position versus index number

Hello,

I have a coll with several hundreths of indexes, with frequent updates. Therefore indexes do not have consecutive numbers. Imagine the first indexes are 4, 7 and 10. There is no function to retrieve the data from the third index if I don't know its number.

I could duplicate the coll using dump, then sort and renumber but because of the frequent updates and coll's size it would be very unefficient.

So an additional "nth" function to retrieve data with an index position rather than an index number would be useful.