Re: Using random() in update produces same random value for all

Yep, interesting.  Checked with PostgreSQL 10.1.
=> select *,random() from generate_series(1,10);
 generate_series |      random
-----------------+-------------------
               1 | 0.308531506918371
               2 | 0.126279713585973
               3 | 0.984668150078505
               4 | 0.884970095474273
               5 | 0.692738385871053
               6 | 0.290897831786424
               7 | 0.914066118188202
               8 | 0.031909613404423
               9 | 0.574441066011786
              10 | 0.631192437838763
(10 rows)
=> select *,(select * from random()) from generate_series(1,10);
 generate_series |       random
-----------------+--------------------
               1 | 0.0718352268449962
               2 | 0.0718352268449962
               3 | 0.0718352268449962
               4 | 0.0718352268449962
               5 | 0.0718352268449962
               6 | 0.0718352268449962
               7 | 0.0718352268449962
               8 | 0.0718352268449962
               9 | 0.0718352268449962
              10 | 0.0718352268449962
(10 rows)
=> select *,(select random()) from generate_series(1,10);
 generate_series |      random
-----------------+-------------------
               1 | 0.848611807450652
               2 | 0.848611807450652
               3 | 0.848611807450652
               4 | 0.848611807450652
               5 | 0.848611807450652
               6 | 0.848611807450652
               7 | 0.848611807450652
               8 | 0.848611807450652
               9 | 0.848611807450652
              10 | 0.848611807450652
(10 rows)

Looked like random() is "volatile", but in subselect it works like "stable".