Fujii-san, Thanks for reviewing and making patch! On 2020-02-13 22:42, Fujii Masao wrote: > The patch changed all calls of pg_gin_pending_stats() so that > they return only tuples column. But what we should change to > stabilize the test is only first call of the function, isn't it? > Since other three calls of the function don't return fragile > result value, I'd like to keep them as they are and change only > the first of the function. Thought? Patch attached. I agree to change only one test. > Also my patch includes the comment about why pg_gin_pending_stats() > returns only tuples column not pages in the regression test. Without > this kind of comment, I'm afraid we may forget why we do that. I think that's a good idea! Regards, Torikoshi Atsushi
Fork