From 596c8fce393e7d0be1dbd154aad2c5cb3f4042b6 Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?C=C3=A9dric=20Krier?= <ced@b2ck.com>
Date: Thu, 11 Mar 2010 12:02:29 +0100
Subject: [PATCH] Use double instead of float to store Float fields in MySQL
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

The precision of MySQL float is not enough to handle all Python floats.
From http://docs.python.org/tutorial/floatingpoint.html:
"...almost all platforms map Python floats to IEEE-754 “double precision”"
---
 trytond/backend/mysql/fields.py | 2 +-
 trytond/backend/mysql/table.py  | 2 +-
 2 files changed, 2 insertions(+), 2 deletions(-)

diff --git a/trytond/backend/mysql/fields.py b/trytond/backend/mysql/fields.py
index 3f2840ba9c..f3fbd42c68 100644
--- a/trytond/backend/mysql/fields.py
+++ b/trytond/backend/mysql/fields.py
@@ -55,7 +55,7 @@ class Float(fields.Float):
 
     @staticmethod
     def sql_type(field):
-        return ('float', 'float')
+        return ('double', 'double')
 
 
 class Numeric(fields.Numeric):
diff --git a/trytond/backend/mysql/table.py b/trytond/backend/mysql/table.py
index e9de67c856..4df897b8ef 100644
--- a/trytond/backend/mysql/table.py
+++ b/trytond/backend/mysql/table.py
@@ -175,7 +175,7 @@ class TableHandler(TableHandlerInterface):
                         ('varchar', 'text'),
                         ('text', 'varchar'),
                         ('date', 'timestamp'),
-                        ('bigint', 'float'),
+                        ('bigint', 'double'),
                         ('int', 'bigint'),
                         ('tinyint', 'bool'),
                         ('decimal', 'numeric'),