Performance test unexpected success in scrolltable

[mithrandi]

[SUCCESS!?!]
Reason: 'Index scan to find appropriate storeID starting point once the value index has been used to seek to /near/ the correct starting place causes this to be O(N) on the number of rows with duplicate values.'

xmantissa.test.test_scrolltable.InequalityPerformanceTests.test_rowsAfterItemWithDuplicatesAtStart
xmantissa.test.test_scrolltable.InequalityPerformanceTests.test_rowsBeforeItemWithDuplicatesAtStart

This is probably not worth spending any brain cycles on since I think the plan was to deprecate scrolltable anyway.

[mithrandi]

Oh, it was tdb that was going to be deprecated, not scrolltable. Well, maybe the .todo attribute should just be removed from these tests? I guess the SQLite query optimizer got smarter or something, although I think that just means the test isn't testing what it thinks it is anymore :/

[mithrandi]

Turns out these tests don't failunexpectedly succeed on the older version of SQLite that Travis has.