is slow but using a carthesian join is fast.

select rownum
from (select 1 from dual group by (cube(1,2,3,4,5,6,7,8,9,10))
, (select 1 from dual group by (cube(1,2,3,4,5,6,7,8,9,10));

and deleivers the same number of rows.
This is faster than

select count(rownum ) FROM dual CONNECT BY ROWNUM

On all three of your suggestions I got: ERROR at line 1:ORA-00920: invalid relational operator
SQL> select rownum from dual connect by rownum
2 /
select rownum from dual connect by rownum
*
ERROR at line 1:
ORA-00920: invalid relational operator

Did you leave something out of the scripts? (on the second one, there is a parenthesis missing)

What also works as a row-generator:

select *
  from (select level l
          from dual
       connect by level < = 6
       )
/

select to_char(add_months(to_date(’01-12-1996′,’DD-MM-YYYY’), r.months), ‘Q’) kwartaal
, to_char(add_months(to_date(’01-12-1996′,’DD-MM-YYYY’), r.months), ‘MONTH’, ‘NLS_DATE_LANGUAGE = dutch’) maand
, to_char(add_months(to_date(’01-12-1996′,’DD-MM-YYYY’), r.months), ‘YYYY’) jaar
–, last_day(add_months(to_date(’01-12-1996′,’DD-MM-YYYY’), r.months)) last_day
FROM ( select rownum months
from ( select 0
from dual
group
by cube (1,2,3,4,5,6,7,8)
)
) r
;