[ome-devel] DB_UPDATE

[Ilya Goldberg]

On Jul 7, 2005, at 11:07 AM, Chris Allan wrote:

> On Thu, Jul 07, 2005 at 10:56:09AM -0400, Ilya Goldberg wrote:
>>
>> On Jul 7, 2005, at 10:26 AM, Josh Moore wrote:
>>
>>> Ilya Goldberg wrote:
>>>
>>>> And at the same time, did something change recently. I'm now missing
>>>> the
>>>> "path" field on "image_pixels"?? No "version" for the update to Bug.
>>>>
>>>>> This was a 2.10 update (we're at 2.16 in this branch, and 2.15 in
>>>>> 2.4.0).  I don't know where you're at, but you gots to keep up,  
>>>>> man.
>>>
>>> ?? Is the path field on image_pixels supposed to be there or not?  
>>> (and
>>> when naturally). What happened? Both DBs (that with and that without)
>>> have configuration.db_version = '2.16'.
>>>
>>> Here's my process:
>>>
>>> Slightly older database from Chris (nightshade) was updated to the
>>> OME_2_4_1_DB_UPDATES branch and _had_ the field.
>>
>> Well, that would be a mistake since it was supposed to be dropped in
>> 2.10.
>>
>>>
>>> This week I installed fresh from HEAD. "cvs up -r
>>> OME_2_4_1_DB_UPDATES"'d and ran "perl install.pl". All the other  
>>> fields
>>> (well, at least most) in the previous DB were there. Path was gone.
>>> Explicame por favor si posible.
>>
>> Somewhere along the line the 2.9 upgrade script wasn't run on the DB
>> you got from Chris.  You have to get Chris to explicame-you.
>> At any rate, the DB you got from HEAD with the applied branch is
>> correct.  No more image_pixels.path.
>> -I
>>
>
> The following is a 2.4.0 stable box that has only ever updated using
> stable packages:
>
> ome=# SELECT * FROM configuration WHERE name = 'db_version';
>  var_id | value | configuration_id |    name
> --------+-------+------------------+------------
>       5 | 2.15  |                1 | db_version
> (1 row)
>
> ome=# \d image_pixels
>                            Table "public.image_pixels"
>        Column        |  Type   |                    Modifiers
> ---------------------+--------- 
> +-------------------------------------------------
>  attribute_id        | integer | not null default  
> nextval('attribute_seq'::text)
>  repository          | integer |
>  pixel_type          | text    |
>  size_y              | integer |
>  module_execution_id | integer |
>  size_z              | integer |
>  file_sha1           | text    |
>  path                | text    |
>  size_t              | integer |
>  image_id            | integer | not null
>  image_server_id     | bigint  |
>  size_x              | integer |
>  size_c              | integer |
>  bits_per_pixel      | integer |
> Indexes:
>     "image_pixels_pkey" primary key, btree (attribute_id)
>     "ome__index_96" btree (module_execution_id)
>     "ome__index_97" btree (image_id)
> Foreign-key constraints:
>     "$1" FOREIGN KEY (module_execution_id) REFERENCES  
> module_executions(module_execution_id) DEFERRABLE INITIALLY DEFERRED
>     "$2" FOREIGN KEY (image_id) REFERENCES images(image_id) DEFERRABLE  
> INITIALLY DEFERRED
>
>
> You'll notice "path" exists. Ideas?


Well, crap.  I had a feeling I would have to explicame-this.  Why must  
I suffer so?
And stop ganging up on me you two!  I've done nothing but write emails  
for like hours and hours.  You're killing me!

-I


>
>>
>>>
>>> -J.
>
> Ciao.
>
> -Chris
>