Steve Singer was reviewing changes I'm making to DDL handling, and
noticed something a bit anomalous about the usages of types:

--------------------------------------------------------------------------------------------------------
In sl_log_script definition we are using bigint for log_txid and int8 for
log_actionseq.  I see sl_log_1 sl_log_2 are defined the same way.  Is there a
difference between bigint and int8?  Should we be consistent? Should we be
changing sl_log_1 and sl_log_2 (since the 2.2 upgrade code is dropping sl_log_1
anyway)?
--------------------------------------------------------------------------------------------------------

I note that...

- txid functions tend to use bigint

oxrsdb at localhost->  \df *txid*
                                     List of functions
   Schema   |           Name           | Result data type |  Argument
data types  |  Type
------------+--------------------------+------------------+-----------------------+--------
 pg_catalog | txid_current             | bigint           |
           | normal
 pg_catalog | txid_current_snapshot    | txid_snapshot    |
           | normal
 pg_catalog | txid_snapshot_in         | txid_snapshot    | cstring
           | normal
 pg_catalog | txid_snapshot_out        | cstring          |
txid_snapshot         | normal
 pg_catalog | txid_snapshot_recv       | txid_snapshot    | internal
           | normal
 pg_catalog | txid_snapshot_send       | bytea            |
txid_snapshot         | normal
 pg_catalog | txid_snapshot_xip        | SETOF bigint     |
txid_snapshot         | normal
 pg_catalog | txid_snapshot_xmax       | bigint           |
txid_snapshot         | normal
 pg_catalog | txid_snapshot_xmin       | bigint           |
txid_snapshot         | normal
 pg_catalog | txid_visible_in_snapshot | boolean          | bigint,
txid_snapshot | normal
(10 rows)

- Sequence column types are generally bigint, too:

oxrsdb at localhost->  create sequence foo;
CREATE SEQUENCE
oxrsdb at localhost->  \d foo
             Sequence "public.foo"
    Column     |  Type   |        Value
---------------+---------+---------------------
 sequence_name | name    | foo
 last_value    | bigint  | 1
 start_value   | bigint  | 1
 increment_by  | bigint  | 1
 max_value     | bigint  | 9223372036854775807
 min_value     | bigint  | 1
 cache_value   | bigint  | 1
 log_cnt       | bigint  | 1
 is_cycled     | boolean | f
 is_called     | boolean | f

In 2.2, we're doing some substantial changes to tables (e.g. -
dropping and recreating sl_log_1 and 2), so if consistency of types
would be valuable, it's opportune to change types now.