Description:
In earlier releases of 5.0 (5.0.13 and 5.0.18 tested) a string inserted into an integer column would truncate in non-strict mode and give an error in strict mode.
Somewhere along the way this disappeared.

mysql> CREATE TABLE bugtest ( a INT );
Query OK, 0 rows affected (0.28 sec)

mysql> SET sql_mode = 'STRICT_ALL_TABLES';
Query OK, 0 rows affected (0.00 sec)

mysql> INSERT INTO bugtest VALUES ('-');
Query OK, 1 row affected (0.07 sec)

How to repeat:
CREATE TABLE bugtest ( a INT );
SET sql_mode = 'STRICT_ALL_TABLES';
INSERT INTO bugtest VALUES ('-');
DROP TABLE bugtest;

Thank you for a bug report. Verified just as described with latest 5.0.36-BK on Linux:

openxs@suse:~/dbs/5.0> bin/mysql -uroot test
Reading table information for completion of table and column names
You can turn off this feature to get a quicker startup with -A

Welcome to the MySQL monitor.  Commands end with ; or \g.
Your MySQL connection id is 1
Server version: 5.0.36 Source distribution

Type 'help;' or '\h' for help. Type '\c' to clear the buffer.

mysql> CREATE TABLE bugtest ( a INT );
SEQuery OK, 0 rows affected (0.01 sec)

mysql> SET sql_mode = 'STRICT_ALL_TABLES';
Query OK, 0 rows affected (0.00 sec)

mysql> INSERT INTO bugtest VALUES ('-');
DQuery OK, 1 row affected (0.00 sec)

Assigning to maintenance for investigation. Please assign back to runtime if find complex.

A patch for this bug has been committed. After review, it may
be pushed to the relevant source trees for release in the next
version. You can access the patch from:

http://lists.mysql.com/commits/22095

The patch http://lists.mysql.com/commits/22095  is ok to push.

I only found a very small thing, which is not harmful GCC,
but other compilers will likely generate warnings. In this
piece of code:

+bool Field_num::get_int(CHARSET_INFO *cs, const char *from, uint len,
+                        longlong *rnd, ulonglong unsigned_max, 
+                        longlong signed_min, longlong signed_max)
+{
+  char *end;
+  int error;
+  
+  *rnd= cs->cset->strntoull10rnd(cs, from, len, unsigned_flag, &end,

The "rnd" variable is a pointer to "longlong" type,
while strntoull10rnd() return type is "ulonglong".
Adding an implicit cast should make the code safe
against warnings:

  *rnd= (longlong) cs->cset->strntoull10rnd(cs, from, len, unsigned_flag, &end,

Pushed, for 5.0.40 and 5.1.18-beta

Noted in 5.0.40, 5.1.18 changelogs.

String truncation upon insertion into a integer column did not
generate a warning (or an error in strict mode).