Changes between Version 2 and Version 3 of Ticket #3344, comment 5


Ignore:
Timestamp:
Oct 24, 2015, 8:16:24 AM (9 years ago)
Author:
postgispaul

Legend:

Unmodified
Added
Removed
Modified

  • Ticket #3344, comment 5

    v2 v3  
    22
    33But I still stand by my assumption that being a primary key does in fact *not* change the end-result in case of a left-join.
     4
     5Edit2: just tested this with a left join lateral icw limit 1, but I guess that's a bit too tricky for postgres's query planner to understand (still uses names table)
    46
    57I assume there isn't a mechanism present where st_extent publishes this kind of information (duplicates won't change the result) so that the query planner of postgres knows it can skip scanning that table?