Enterprise Manager

Well, data is case sensitive (unless you've done a whole lot of fiddling with NLS parameters to force case-insensitive queries), so a lower case letter sorts after an upper case letter.

  1* select (case when 'j' > 'J' then 'TRUE' else 'FALSE' end) from dual
SCOTT @ jcave102 Local> /

(CASE
-----
TRUE

That means that 'JUSTIN' comes after 'JAMES' and before 'James'. The same goes with any name that comes between 'J' and 'JAMES' in alphabetical order.

Justin

As an addendum to Justin's response ... depending on your character
set the following should give you a good idea of how things will sort:

SELECT 'MORGAN', dump('MORGAN') FROM dual;

SQL> SELECT 'MORGAN', dump('MORGAN') FROM dual;

'MORGA DUMP('MORGAN')
------ -------------------------------
MORGAN Typ=96 Len=6: 77,79,82,71,65,78

SELECT 'MORGAn', dump('MORGAn') FROM dual;

'MORGA DUMP('MORGAn')
------ --------------------------------
MORGAn Typ=96 Len=6: 77,79,82,71,65,110

Note that 78 will precede 110

Though for those with a more technical bent it is likely that what Oracle is doing
is more like this:

SQL> SELECT 'MORGAN', sys_op_map_nonnull('MORGAN') FROM dual;

'MORGA SYS_OP_MAP_NON
------ --------------
MORGAN 4D4F5247414E00

precedes

SQL> SELECT 'MORGAn', sys_op_map_nonnull('MORGAn') FROM dual;

'MORGA SYS_OP_MAP_NON
------ --------------
MORGAn 4D4F5247416E00

But I wouldn't bet my reputation on it.

As an addendum to the addendum, single letters are always smaller than words beginning with that letter. So, your third query would find all employees with a name that began with J and ignore anybody with a name beginning with W or later.

SQL> SELECT CASE WHEN 'WARD' > 'W' THEN 'TRUE' ELSE 'FALSE' END
  2  FROM dual
 
CASEW
-----
TRUE

John

-- REGEXP_LIKE --
function REGEXP_LIKE (srcstr   VARCHAR2 CHARACTER SET ANY_CS,
                      pattern  VARCHAR2 CHARACTER SET srcstr%CHARSET,
                      modifier VARCHAR2 DEFAULT NULL)
  return BOOLEAN;
  pragma FIPSFLAG('REGEXP_LIKE', 1452);

function REGEXP_LIKE (srcstr   CLOB CHARACTER SET ANY_CS,
                      pattern  VARCHAR2 CHARACTER SET srcstr%CHARSET,
                      modifier VARCHAR2 DEFAULT NULL)
  return BOOLEAN;
  pragma FIPSFLAG('REGEXP_LIKE', 1452);

begin 
  if (xor(true,false)) then null;
  end if;
end;
/

PL/SQL procedure successfully completed.

SCN_TO_TIMESTAMP            
STATS_T_TEST_INDEP          
STATS_T_TEST_INDEPU         
STATS_T_TEST_ONE            
STATS_T_TEST_PAIRED         
TIMESTAMP_TO_SCN            
XMLCOLATTVAL          

select 1 from dual 
where (date '2000-01-01', date '2001-01-01') 
overlaps (date '2000-07-01',date '2001-07-01'); 

1

select sys.sys_nt_collect_imp (cast ('01' as raw(1))) from dual
/
select sys.sysevent from dual
/
select sys.instance_num from dual
/

satyaki>
satyaki>create table meeting
  2    (
  3       id      number(5),
  4       st_dt   date,
  5       en_dt   date
  6    );

Table created.

satyaki>
satyaki>
satyaki>insert into meeting values(1,sysdate,sysdate+5);

1 row created.

satyaki>
satyaki>
satyaki>insert into meeting values(2,sysdate+4,sysdate+7);

1 row created.

satyaki>
satyaki>
satyaki>insert into meeting values(3,sysdate+8,sysdate+9);

1 row created.

satyaki>
satyaki>
satyaki>commit;

Commit complete.

satyaki>
satyaki>select * from meeting;

        ID ST_DT     EN_DT
---------- --------- ---------
         1 24-AUG-07 29-AUG-07
         2 28-AUG-07 31-AUG-07
         3 01-SEP-07 02-SEP-07

satyaki>

satyaki>select 1 "Ovrlap"
  2  from meeting
  3  where (st_dt,en_dt) overlaps (st_dt,en_dt);

    Ovrlap
----------
         1
         1
         1

satyaki>

select T1.ID, T2.ID, 
  CASE WHEN (T1.st_dt,T1.en_dt) overlaps (T2.st_dt,T2.en_dt) THEN 'YES' ELSE 'NO' END OVERLAP
from meeting T1, MEETING T2;

        ID         ID OVE
---------- ---------- ---
         1          1 YES
         1          2 YES
         1          3 NO
         2          1 YES
         2          2 YES
         2          3 NO
         3          1 NO
         3          2 NO
         3          3 YES