Forum: Rails-core (closed, excessive spam) Failing activerecord tests on mysql 4.1

There are some activerecord tests that are failing on my machine. I
suspect this is because I'm still using mysql 4.1 (since that's what I
deploy to).
In particular, in the active schema tests we check that a column of
the correct type has been added with "SHOW FIELDS FROM delete_me where
FIELD='created_at' AND TYPE='datetime'"
Mysql4 doesn't allow you to have a where clause there and the test
fails. the test would pass on 4 & 5 if it didn't used the where and
just inspected the results of the query a little closer.

So should I submit a ticket here or is the answer 'get with the 21st
century' ?

Fred

> So should I submit a ticket here or is the answer 'get with the 21st
> century' ?

If there's something minor like that that prevents the tests from
passing, then yeah, send us some patches :)

--
Cheers

Koz

On 21 May 2008, at 10:27, Michael Koziarski wrote:

>
>> So should I submit a ticket here or is the answer 'get with the 21st
>> century' ?
>
> If there's something minor like that that prevents the tests from
> passing, then yeah, send us some patches :)

Here we go then : 
http://rails.lighthouseapp.com/projects/8994-ruby-...

in the same category there's a test in migration_test that fails
because on ruby 1.8.6 DateTime formats timezone offsets as hh:mm (eg
+05:00) but ruby formats it as +0500. What's the preferred approach
here? assert it's /\+05:?00/ ?

Fred

> in the same category there's a test in migration_test that fails
> because on ruby 1.8.6 DateTime formats timezone offsets as hh:mm (eg
> +05:00) but ruby formats it as +0500. What's the preferred approach
> here? assert it's /\+05:?00/ ?

Sounds reasonable to me.



--
Cheers

Koz

On 22 May 2008, at 00:04, Michael Koziarski wrote:

>
>> in the same category there's a test in migration_test that fails
>> because on ruby 1.8.6 DateTime formats timezone offsets as hh:mm (eg
>> +05:00) but ruby formats it as +0500. What's the preferred approach
>> here? assert it's /\+05:?00/ ?
>
> Sounds reasonable to me.
>
fixed that (and another one) at 
http://rails.lighthouseapp.com/attachments/24122/0...

Fred