into mysql.com:/usr/local/home/marty/MySQL/mysql-5.0

into deer.(none):/home/hf/work/mysql-5.0.clean

into mysql.com:/usr/local/home/marty/MySQL/mysql-4.1

into mysql.com:/home/mydev/mysql-5.0-5000

into deer.(none):/home/hf/work/mysql-4.1.7344

One more free_charset call added

into rurik.mysql.com:/home/igor/dev/mysql-5.0-0

into rurik.mysql.com:/home/igor/dev/mysql-4.1-0

* Add 0.01 to cost of 'range'+'using index' scan to avoid optimizer choice races with 
  'index' scan.

into mysql.com:/home/mydev/mysql-5.0-5000

into mysql.com:/home/pem/work/mysql-5.0

  by simply disallowing alter procedure/function in an SP (as for drop).

  Bug in the optimizer caused an infinite loop for weird code.

into mysql.com:/home/mydev/mysql-4.1-bug7806

into mysql.com:/home/mydev/mysql-4.1-bug7806

into mysql.com:/usr/local/home/marty/MySQL/mysql-4.1

into mysql.com:/home/psergey/mysql-4.1-bug9103

  Fixed bug #9681.
  The bug happened with queries using derived tables specified by
  a SELECT with ROLLUP, such as:
  SELECT * FROM (SELECT a, SUM(a) FROM t1 GROUP BY a WITH ROLLUP) t2,
  if column a of table t1 is declared as NOT NULL.
  This was to the fact that the first column of the temporary table
  created to contain the derived table erroneously inherited the NOT NULL
  attribute from column a.
olap.result, olap.test:
  Added a test case for bug #9681.

Modified the check for the timestamp field so that the flags for
the automatic for inserts and updates are cleared independently.

into mysql.com:/M50/mysql-5.0

 - Segfault because of passing a 64-bit pointer to 32 bit integer.
 - Add new include  config_readline.h which will calculate some new defines 
   based on what is found in config.h
 - This file was originally included in readline 4.3 as config.h.in, only the part that
    makes new defines have been moved to this file. 

into hundin.mysql.fi:/home/jan/orig/mysql-5.0

into hundin.mysql.fi:/home/jan/mysql-4.1

into mysql.com:/usr/local/home/marty/MySQL/mysql-4.1

into xiphis.org:/.amd_mnt/bk.anubis/host/work-acurtis/bug9496