Review TPCH runs

[krlmlr]

Query 10 didn't finish on my machine for sf=1. I suspect this is due to our use of the "full compatibility" operators from the rfuns extension for comparison. We might want to use faster variants in "good" cases (e.g., compare a column with a constant) so that duckdb can optimize.

Related: #270, CC @toppyy.

[krlmlr]

All good now, this was due to undesired materialization, #289.

[krlmlr]

The TPCH data changed in duckdb 0.8.0, in particular, the o_comment column in the orders table. @Tmonster @hannes: does that ring a bell?

0.7.1

$ echo 'INSTALL tpch; LOAD tpch; CALL dbgen(sf = 0.01); SELECT * FROM orders LIMIT 1; ' | ./duckdb-0.7.1 -json
[{"o_orderkey":1,"o_custkey":370,"o_orderstatus":"O","o_totalprice":172799.48999999998449,"o_orderdate":"1996-01-02","o_orderpriority":"5-LOW","o_clerk":"Clerk#000000951","o_shippriority":0,"o_comment":"nstructions sleep furiously among "}]

1.1.3

$ echo 'INSTALL tpch; LOAD tpch; CALL dbgen(sf = 0.01); SELECT * FROM orders LIMIT 1; ' | duckdb -json
[{"o_orderkey":1,"o_custkey":370,"o_orderstatus":"O","o_totalprice":172799.49,"o_orderdate":"1996-01-02","o_orderpriority":"5-LOW","o_clerk":"Clerk#000000951","o_shippriority":0,"o_comment":"ly express platelets. deposits acc"}]