errors reloading with pg_dump

Howdy:

Running PostgreSQL 7.2.1 on RedHat Linux 7.2 kernel 2.4.7-10.

This past weekend, I was trying to test PostgreSQL 7.2.2 on a backup
box.  I did a pg_dump of my production database (7.2.1) and tried
to load it on my development DB (7.2.2).  When I did, I got
a slew of errors pertaining to invalid data inside the tables.

For example, if I tried to restore a table that has an invalid date in
the date column, an error is generated.  My questions are:

1) If the errors are generated, does that mean that the record is
skipped each time there is invalid data in the table or the table is
skipped?

2) How can I check the integrity of the new database if pg_dump
is not loading all of the information from the production database?

3) I tried doing something like:

[snip]

pg_dump -i -Fc -S postgres -d testdb | pg_restore -v -h 192.168.0.2 -S \
postgres -d testdb

[/snip]

And got a list or errors - mostly because a few tables did not have read
permissions set to them.  I have to go threw them by hand to change permissions
(It's odd: how is it that you can create a table but the table will NOT have
the owner or it's permissions on it?  like so:

[snip permissions]

testdb=> \dp forums_auth
  Access privileges for database "testdb"
       Table            | Access privileges
-----------------------------------+-------------------
 forums_auth            |
 forums_auth_id_seq     |

[/snip permissions]



I'm sure I can do a count or something of each table, but, this also
means my backup / recovery method is screwy because I can't
recover anything that PostgreSQL throws errors at.

How is everyone else in the world reloading tables that may have
invalid data?

Thanks!

-X