Re: [GUSDEV] Preliminary 3.5 Schema

[Michael S. <msa...@pc...>]

Since releasing the preliminary 3.5 schema, the following changes have been 
proposed.  The list below only covers App, SRes, and Core-- other schema 
changes will come early next week.

The scheduled freeze date for the 3.5 schema is 4/29.  If there are no 
objections to the items below, Core and App will be frozen for 3.5 on 4/25 
(more SRes changes may be coming early next week).

Consult the road map for full changes.  The list below are just new changes 
since sending this previous email.  They'll be added to the road map after 
feedback from the list is received.

App:

* The App schema has been deprecated.  The remaining few tables that were 
in app will be moved to non-GUS project specific schema (i.e. a Study 
Annotator schema).

Core:

* The workflow tables have been deprecated as they were never actually 
used.  This includes core.workflowstatus, core.workflowedge, 
core.workflowinit, core.workflowedgesql, core.workflow, 
core.workflowblackboard, and core.workflowstatus

* core.generaldocumentation is tentatively deprecated, pending the lack of 
objections from this list.

* core.userdatabase is tenetively deprecated, pending the lack of 
objections from this list.

SRes (tentative, pending more review)

* SRes.SequenceOntology.SO_ID will be a VARCHAR(12), not a NUMBER(10)

* SRes.MGEDOntologyRelationship will be named SRes.OntologyRelationship

* SRes.MGEDOntologyTerm will be named SRes.OntologyTerm

* SRes.BibliographicReference.authors will be nullable, to support the 
alternative BibRefAuthor linking table. 
SRes.BibliographicReference.External_Database_Release_ID will gain the FK 
the ExternalDatabaseRelease table

* SRes.BibRefAnnotation is tentatively deprecated, pending the lack of 
objections from this list.

Thanks,

Mike


--On April 7, 2005 1:53:44 PM -0400 Michael Saffitz 
<msa...@pc...> wrote:

>
> All,
>
> Development work on the GUS 3.5 Schema is nearing completion, and we'd
> like to take this opportunity to solicit community feedback and provide
> the 3.5 schema for review.
>
> With the 3.5 Schema nearly finalized, we're now turning our attention to
> ensuring compatibility in the object layer and with plugins for 3.5. Once
> done, GUS 3.5 will be ready for release.  It's my hope that that will be
> as soon as possible, but I'm hesitant to give a firm date.
>
> The attached file contains the 3.5 schema, in a simple text summary
> format.  When reviewing the file, please keep the following in mind:
>
> * I've done a very preliminary analysis of the file--- there may be
> glaring mistakes/omissions.
>
> * All tables and views in GUS 3.5 are versioned, and the version
> tables/views are not displayed.
>
> * All tables and views in GUS 3.5 contain the standard housekeeping
> columns, which are also not displayed.
>
> * Subclass tables (marked by "extends superclass" where superclass is the
> name of the parent table) only have the columns added by the subclass
> displayed.  For example, if the superclass has columns a, b, c, and d,
> then the subclass will display the column e but have all five columns in
> the database.
>
> * This is an object-centric display-- the method of subclassing (i.e. use
> of imp tables) is not changing.  As a result, Imp tables are not
> displayed, nor are super/sub classes flagged as views.
>
> * The Type of a column with a foreign key is the name of the foreign key
> table.  For example, a table with a column NA_FEATURE_ID with a foreign
> key to DoTS.NAFEATURE will have a type of DoTSS.NAFEATURE.  This is the
> only way foreign keys are displayed in this view.
>
> * All tables first column serves as the Primary key.  There is no
> denotation for the primary key (although it will also appear as an index
> against the first column).  Some tables are missing primary keys in this
> document-- it will be fixed soon.
>
> * All indexes for a table are displayed.
>
> Please let me know if you have any questions/concerns.
>
> Thanks,
>
> Mike