Changes between Initial Version and Version 1 of Ticket #6700, comment 5
- Timestamp:
- Oct 9, 2010, 10:16:21 PM (14 years ago)
Legend:
- Unmodified
- Added
- Removed
- Modified
-
Ticket #6700, comment 5
initial v1 1 Another alternative: if the columns being sorted on don't provide uniqueness between two matches, would it make sense to compare the other available data for that pair of rows until something not-identical is found? At least in theory that should work, since even a BQuery won't allow the same exact entry into the list more than once, though there is the question of the performance impact. I'd estimate the latter to be negligible though since this code path would only have to be taken in the case where the default set of comparisons can't find any ordering between a pair of items, which should be rare. Thoughts?1 Another alternative: if the columns being sorted on don't provide uniqueness between two rows, would it make sense to compare the other available data for that pair of rows until something not-identical is found? At least in theory that should work, since even a BQuery won't allow the same exact entry into the list more than once, though there is the question of the performance impact. I'd estimate the latter to be negligible though since this code path would only have to be taken in the case where the default set of comparisons can't find any ordering between a pair of items, which should be rare. Thoughts?
