Visual Studio Team System 2008 Database Edition GDR - RTM

Finally the moment is there, the final version of the Visual Studio Team System 2008 Database Edition GDR has been released to the web.

Although the name "GDR", which stands for "General Redistribution Release", implies otherwise, this really is a complete new release, based on a new architecture. We have taken the architectural changes from the Rosario project (Visual Studio Team System 2010) and rolled them forward in time. Rolling forward these changes help us align the GDR release with the upcoming Visual Studio Team System 2010 release and provide numerous new features and extensibility points in this release. If you would compare the GDR release with the "Visual Studio 2005 Team Edition for Database Professionals" or "Visual Studio Team System 2008 Database Edition" you will agree this is a completely new product!

The key architectural changes

  • Model based architecture - Everything underneath the covers is based on a true model representation of the SQL Server schema. This facilitates a true offline declarative database development system where the source code defines the shapes of the schema objects.
  • Models are implemented by Database Schema Providers, DSP's for short. - The introduction of a provider model enables multiple things at once. First of all the decoupling of release vehicles. For example when SQL Server will release a new version or adds new functionality in a service pack, like they did in SQL Server 2005 SP2 when adding "vardecimal" support, we can simply update an existing provider or provide a new one in case of a new SQL Server release. Overtime we will go to a model where the SQL Server providers will be an integral part of the SQL Server release. The provider model is also a key extensibility point, allowing 3rd parties to extend database project ecosystem and add support for other database platforms in Visual Studio Team System 2010.
  • Tool extensibility - In this release external tool writers have full access to the T-SQL parsers (for SQL Server 2000, 2005 and 2008), the SQL Script DOM and the schema model (when inside Visual Studio). This adds on top of the ability to write extend the tools inside the platform. You can extend T-SQL refactoring by writing your own refactoring types (operations) and refactoring targets; add T-SQL Static Code Analysis rules, data generators, data distributions and test conditions. Database Projects (.dbproj) now provide a truly extensible declarative database development platform.
  • Separation of BUILD and DEPLOY. - The separation of build & deploy makes it possible to deploy the output of your database project to many different targets and different points in time. Build now produces a single artifact file hat describes the schema inside your database, called a .DBSCHEMA file. This file is used by the deployment engine to deploy your schema.
  • Standalone Deployment Engine. - The inclusion of a standalone and redistributable deployment engine makes it possible to deploy the output of your database projects (.DBSCHEMA files) to a target database without the need of having Visual Studio Team System Database Edition installed. This enables key scenarios like the inclusion of database schema deployment as part of your application installation.

All these architectural changes enable many new usage scenarios and interesting new features. So lets walk through the list of new features and major changes.

Feature list

  • Project System
    • Full support for SQL Server 2000, SQL Server 2005 and SQL Server 2008 support.
      • All database objects are now modeled, therefore they are compared and part of deployment. The side-effect of this is that there are no more objects that need to be stored in pre- and post-deployment scripts.
    • Server vs. Database projects
      • This change allows you to model the different behavior of shared and server level objects inside SQL Server vs. user database level objects. 
    • Partial projects
      • Partial project allow you to share implementation between projects, with a single definition of the code
    • Composite projects
      • Composite projects extend database references, allowing you to add objects from other project to the name database schema namespace. So you can compose your database from other projects and/or DBSCHEMA file references
    • Database references using literals, in addition to SQLCMD token support
      • The addition of literal support facilitates that you do not have to change your code of 3-part names, when do do not require the flexibility of name independent deployment
    • Reference support for XML Schema Collections to include .XSD files
      • Enables single sourcing of XML Schema Collections from an .XSD file which itself can be used by other developers or tools.
    • Reference support of SQL-CLR projects and / or SQL-CLR assemblies
      • Enables the inclusion of a SQL-CLR project or a binary, to represent a SQL-CLR assembly. Like XSD references, the goal is single sourcing of artifacts inside the development environment
    • Import Schema and Import Script is now available through DTE
      • This allows the creation of macros or programmatic add-in to drive these common tasks.
    • No more "DesignDB"
      • The change to a fully model driven implementation makes the need for a local SQL Server database instance obsolete. Having a sandbox for your development environment is still am advised practice though, so you can test the changes independently and in an isolated environment.
  • Build
    • Build validates the consistency of the schema model and produces all the files that you need to deploy the database
  • Deploy
    • Deploy takes the output from the build stage and deploys this information based on the settings supplied
    • Deployment settings are now isolated in to the .SQLDEPLOYMENT file, which allows you to standardize and re-use the deployment settings
    • Most of the "Schema Compare" comparison options are now available as the deployment options, to increase the symmetry between the two
    • VSDBCMD.EXE, is a command line deployment tool, which enables the standalone deployment of Database Project
  • Schema Compare:
    • Compares any combination of live databases, Database Projects and .DBSCHEMA files
    • Session level option support, which can be persisted and re-used
    • Object type level filtering allow you to exclude object types from the comparison
    • You can now quickly navigate back and forward between difference via the toolbar and optionally configurable keyboard shortcuts
    • Substitution of SQLCMD variables, allows for the comparison of projects that rely on the use of SQLCMD variables
    • Schema Compare can now be executed through DTE
  • Data Compare
  • Refactoring:
    • New refactoring types: Expand Wildcards and Fully-qualify Names (in addition to Rename, Move Schema and 3/4 part name refactoring)
    • The patented "preservation of intend" functionality makes is possible that the Deployment Engine will deploy renames and move schema operations as actual renames and move schema operations, instead of a DROP / ADD operation.
    • Extensibility enables the creation of your own refactoring types (operations) and refactoring targets.
  • Database Unit Testing
    • Added support for SQL Server 2008
  • Data Generation
    • BulkCopy based inserts in addition to the existing INSERT data sync
    • Sequential Databound Generator is now included in the product
  • T-SQL Static Code Analysis
    • Execute analysis using MS Build and/or as part of the VS IDE build process
    • Including the ability to write you own T-SQL Static Code Analysis rules
  • Extensibility
    • Public full fidelity .NET based T-SQL parser for SQL Server 2000, 2005 and 2008.
    • Public .NET based SQL Script DOM, for programmatically generation of SQL code (equivalent to the Code DOM in .NET)
    • Access to the Schema Model from within Visual Studio, this enables for example ability to write code generators based on the schema model.
    • Feature extensibility:
      • T-SQL Static Code Analysis rules
      • Refactoring types, these are the refactoring operations
      • Refactoring targets, these are the sources that you want to update as part of a refactoring operation
      • Data Generators
      • Data Distributions
      • Test Conditions

As you can see the GDR release is chockfull of new and improved functionality!  And the good things is you can get it now.

Installation

Uninstall old GDR versions

If you have a previous version of the GDR installed, you will have to uninstall these first. You can do this via Add/Remove Programs or from the command line using:

  • msiexec /X {DDF197C6-4507-3A19-A4B5-0E17CC931370}

Prerequisites

Before you start downloading and installing please check if the following pre-requisites are present on your machine!

Installation

You can download the Visual Studio Team System 2008 Database Edition GDR from the following location:

Upgrading projects

If you have existing Database Projects look at the following matrix:

Version Project upgrade supported
Visual Studio 2005 Team Edition for Database Professionals Yes
Visual Studio 2005 Team Edition for Database Professionals GDR Yes
Visual Studio Team System 2008 Database Edition Yes
Visual Studio Team System 2008 Database Edition SP1 Yes
Visual Studio Team System 2008 Database Edition GDR CTP 17 and before No *
Visual Studio Team System 2008 Database Edition GDR RC0 Yes

* If you need to upgrade GDR CTP earlier projects you can use the Import Script add-in.

Before you upgrade your project, make sure to create a backup of your project!.

During the upgrade process will update your project file (.DBPROJ) and will try to move schema objects that were earlier defined in the the pre- and post deployment scripts in to the project.

Questions

Like always we are looking forward to your feedback which you can post to our MSDN forum and please file bugs and improvement requests using our Microsoft Connect site.

On behalf of the whole "DataDude" team we want to thank you for your support and wish you a happy Thanksgiving!
-GertD