from web site
PEMCO Realty Handling Broker Claxton Real Estate, Inc. Broker/Owner Century 21 Realty LLC Service Consultant Keller Williams Real Estate Peachtree Roadway.
Aha you just found the bug! I changed from "creole" to "reverse"! I simply compared the output of both schema files, and the "reverse" target develops a VARCHAR not a TIMESTAMP. This is a postgresql database. I am using "timestamptz" columns. I included the following to database/reverse/pgsql/ 57 'timestamp' => Propel, Types:: TIMESTAMP, 58 'timestamptz' => Propel, Types:: TIMESTAMP, *** INCLUDED *** 59 'tinyblob' => Propel, Types:: BINARY, And that made it work.
Maybe you also need to add one for "timetz"? I am not 100% confident of my knowledge of this problem, so I'll let you carry out the specific details, but I think we have actually discovered the bug. Thanks! Alan On Feb 27, 2008, at 6:10 PM, Hans Lellelid composed: Hi Alan, I'm confused; I see it's a timestamp in your database, however it looks like a varchar in your schema.
did something modification there? Alan Pinstein wrote: Alan Pinstein wrote: When utilizing beta3, the following code worked: $obj- > set, Dts( time() ); Whereas in beta4, it broke, and had to be converted to: $this- > set, Dts( date(' r') ); Can someone please discuss what took place? I don't always mind it, however I have looked through the code and the wiki and there is no explanation of temporal * mutator * behavior, just temporal * accessor * habits.
There was an unit test testing this-- and now I just added a couple more assertions so that we're evaluating not just time columns, however likewise timestamp and date columns. This is passing fine. I can quickly duplicate the bug. Not able to carry out INSERT statement. [wrapped: Found Here : Invalid datetime format: 7 MISTAKE: void input syntax for type timestamp with time zone: "1204153379"] and, from propel.
Can you inspect whether you have actually set: propel. use, Date, Time, Class = true That needs to be the default, and should not really affect the results here, however is the only thing I can consider straight off. I do not have it set to false, and I also attempted it with the specific setting (it didn't change the code, so the default was correctly working).