Powerful Stuff: WS-*, Single Sign-On, and InfoCard

It's risky to speculate on the nature of the "powerful stuff" that Steve Ballmer mentioned during a Q&A session at the recent TiEcon 2005 conference:

"We are working on more existing powerful stuff around XML Web services that will address many issues beyond RSS."

However, a likely "powerful stuff" candidate is Microsoft's InfoCard initiative for personal digital identity management and Web-based single-sign-on (SSO) in the forthcoming Windows Longhorn client OS. The first public demonstration of InfoCard occurred in May at the Digital ID World 2005 conference in San Francisco. Microsoft more or less simultaneounsly published Kim Cameron's "The Laws of Identity" white paper and a more extensive "Microsoft's Vision for an Identity Metasystem" article. (Kim Cameron is Microsoft's Identity and Access Architect, and publishes the Identity Weblog). The "Pre-Release Software Code Named “Avalon” and “Indigo” Beta1 RC" download, which appeared on May 23, 2005, runs on Windows XP SP-2 or Windows 2003 Server. This download provides the Indigo runtime infrastrastructure for InfoCard Beta 1. (The only references to InfoCard Beta 1appear in the press release and the main Longhorn Developer page's link to the RC, which also has a link to the Release Notes.) Running the Indigo setup program installs the .NET Framework 2.0 April CTP (Beta 2) version. You also can download and install an updated WinFX SDK as an ISO image from a link on the download page. The runtime and SDK compatible with Visual Studio 2005 Beta 2. Johannes Ernst's Blog provides an independent overview of InfoCard and describes its reliance on the WS-* stack. According to Johannes, InfoCard employs the following WS-* members and related specs:

• SOAP [1.2]
• WS-Addressing
• WS-MetadataExchange
• WS-Policy
• WS-Security
• WS-SecurityPolicy
• WS-Transfer
• WS-Trust
• XML Signature
• XML Encryption
• SAML
• WS-Federation (?, unclear)

[Note that Indigo bindings for WS-* support use SOAP 1.2, which results in Web services that don't meet WS-I Basic Profile 1.1. As Tim Ewald observes, many organizations require that all Web services they publish or consume to claim BP-1.1 conformance.] If processing InfoCard identities requires implementation of the eight WS-* specs from the above list, support for SAML, and the Indigo messaging infrastructure, is InfoCard destined for HailStorm's fate? At this point, only WS-Security is an official OASIS specification; the remaining members are at varying points in the standards process. So far, InfoCard appears to me to be another example of the overly complex "everything at once" syndrome that doomed HailStorm. The preceding Indigo and InfoCard Beta 1 RC release followed a May 13, 2005 joint publication by Microsoft and Sun Micrososystems of the Web Single Sign-On Interoperability Profile and Web Single Sign-On Metadata Exchange Protocol (WSSOMEX) specifications. These specs provide a mechanism for integrating WS-* and Liberty Alliance identity management of Web-based single sign-on technologies. WSSOMEX represents Sun's first—if tentative—committment to the WS-* standards beyond WS-Security. The press release, transcript of remarks by Steve Ballmer and Scott McNealy's comments, and related links are here. WSSOMEX is the first concrete result of the 10-year Sun-Microsoft technical collaboration agreement of April 2004. Paul Madsen posted an early analysis of WSSOMEX and WS-MetadataExchange:

WSSOME[X] defines how WS-MetadataExchange can be used to determine which Single Sign-On protocol suites (SAML 1.1, ID-FF 1.2, SAML 2.0, WS-Federation, etc) your partner is capable of supporting so that the two of you can actually do something interesting (like enabling SSO for your customers, employees, etc). WS-MetadataExchange defines a SOAP-based request/response protocol. Fundamentally, one provider says to the other 'tell me what you can do'. If the returned list includes something that the asking provider can also [do], then we have an intersection of capabilities and we're off to the races. If [there's] no intersection, [there's] no way forward.

Sun's Hubert Le Van Gong posted a response to Paul's post and added his own initial InfoCard analysis and a follow-up in response to Kim Cameron's comments. InfoWorld's Jon Udell also weighed in with a post about Web single-sign-on with client-side certificates, a much simpler technology that never caught on, versus InfoCard. In fairness to InfoCard, the Liberty Alliance lists a large number of "Liberty-Enabled Products," but, according to Web service analyst Ron Schmelzer, "[T]here are still very few products, if any, that implement Liberty Alliance on the desktop client, and so Microsoft has a distinct advantage."

RSS and Web Services: Keep It Simple, Steve.

Amit Malhotra's May 19, 2005 post on therssweblog includes a transcript of his unscheduled Q&A session with Microsoft's Steve Ballmer at the TiEcon 2005 conference in Santa Clara:

Q1. How important is RSS? A fad, important, huge or will [it] replace the Web/HTML dominance of the internet? A1. We believe RSS is important and will be around for a while, but it is not going to change the world. It is a little too simple; that is also the reason everyone’s using it. We are working on more existing powerful stuff around XML Web services ... that will address many issues beyond RSS. RSS will be around, but whatever we are working next will be cooler and more prevelant. [Emphasis added] Having said that, there are groups in MS that believe RSS has the potential to change everything and many future technolog[ies] will be built around RSS, the internal debate goes on.

Q2. How do you, or do you, see Google/Blogger and similar tools being a threat to MS Office dominance? A2. Not at all, people will always need Office for the complexity of tasks they perform and, as such, Google’s offerings in the strain of Gmail/Blogger will not replace it. We think it will be part of what we offer in future versions of Office. Besides the next release of MS Office will have the tools to publish blogs as part of its collaborative tools; watch for them.

Amit's preceding transcript, to which I've made minor edits and emphasized text, generated a brief flurry of reaction by other well-known bloggers, including Microsoft's Robert Scoble, who left a comment to a related posting on Steve Rubel's Micropersuasion PR blog. A reader named Bud left the following comment that's a propos my preceding post on the issue of overcomplicating Web services and SOA:

When I talk to people who are doing web services and show them RSS, they say something like, "Hey, we could achieve a heck of a lot with just that. We should just implement it. Web services is so complicated and requires so much effort just to work." And, these are fairly sophisticated developers and architects. Simple, easy to implement technologies where developers can quickly do it and people can immediately see the value are going to win the day. That's things like AJAX as well as RSS.

Steve Ballmer's first answer indicates that Microsoft has a Web service-based replacement in mind for RSS that has a higher "coolness quotient," but I question whether any Web service API or toolkit can come close to competing with RSS's current prevelence in Web developer mindshare. [The assumption is that preceding references to RSS also encompass ATOM.] My bookshelf (as well as Don Box's) has a copy of the .NET My Services Specification (517 pp.) from the 2001 Microsoft Professional Developer Conference (PDC). .NET My Services, better known by its "HailStorm" codename, was intended to evelope XML-based identity, address-book, Web site favorites, calendar, travel and much other personal data with a digital wallet in a Web services wrapper. The essence of HailStorm, in the words of Mark Lucovsky, then Microsoft's Distinguished Engineer and Chief Software Architect of .NET My Services, was:

HailStorm embraced the idea of decoupling the data from the application. The idea was to allow a variety of applications to process and manipulate your calendar data, your address book, your email, your favorite web sites, your travel preferences and itineraries, etc. This is not a new, novel idea, but was certainly something that was important and core to the system. Simple applications that we were trying to enable included the ability to overlay your personal calendar with the calendar of your favorite band, or your favorite sports team, or your spouse, etc. We wanted to enable a unified "address book" where your contacts could be used across applications written by any vendor.

The HailStorm announcement aroused a call-to-arms among Microsoft's competitors to prevent Passport from becoming the world's default identity management and authentication protocol. The Liberty Alliance, championed by Sun Microsystems, proposed a "federated identity management system" and successfully thwarted serious consideration of Passport and HailStorm technology by IT management. Another issue with HailStorm was that client applications weren't really "simple." For example, the introductory chapter for the .NET Calendar service specificaton was 68 pages long. Lucovsky subsequently moved on to Google and recently posted a comparison of the core HailStorm concepts with those of RSS 2.0 and Atom. Mark cites the common core concepts as:

• Network Centric, Extensible Data Model, for Everyday Data
• Data Decoupled from Applications
• Anytime, Anyplace, and from Any Device Access
• Identity Centric Data Access

A major difference I see is that HailStorm proposed to standardize a wide range of data structures and methods for reading and updating personal information. HailStorm relied on the proprietary .NET Passport service (now "Passport Network") for authentication, which Microsoft's Kim Cameron now admits is out of context for authentication by non-Microsoft sites. As one commentor observed, HailStorm proposed to do "everything at once." RSS 2.0 and ATOM concentrate on public content sydication and have proven very successful at their assigned tasks, while only the successor of .NET My Alerts—SQL Server Notification Services—has achieved any semblance of industry adoption.

Microsoft Web Services DevChannel Opens at FTPOnline

Fawcette Technical Publications recently added a new Microsoft Web Services DevChannel to the FTPOnline portals collection. The DevChannel consists of links to Microsoft white papers, artlcles from FTP print and online publications, video clips from FTP conferences and MSDN TV, plus related resources. Service-Oriented Architecture (SOA) gets the primary emphasis, as expected, but many articles include sample C# and VB .NET Web service code. My "Build Real-Time Web Images" article from the August 2004 issue of Visual Studio Magazine was the lead link for the initial DevChannel page. The article describes how to write VB. NET 2003 Windows form clients for Microsoft's SQL Server-driven TerraService and MapPoint Web services. TerraService is an example of a set of freely-accessible Web services that enable .NET developers to display tiled aerial/satellite photographic images of most of the earth, and USGS topographic map images of North America. MapPoint Web services render bitmaps from vector-based maps of the U.S. and many other countries. Using MapPoint Web services requires obtaining a commercial or developer license and executing a simple username/password authentication request prior to invoking the desired Web method.

Update 6/14/2005: InfoWorld's Jon Udell commented on developers use of GoogleMaps versus TerraServer:

"Years ago an early reviewer of Visual Basic 1.0 (Steve Gibson, I think) said that VB increased the software developer's leverage by an order of magnitude. That was true, but you can't keep going back to the same well. In a column on Google Maps I wrote:

Developers haven't flocked to TerraServer. What's Google's secret? Web DNA and no Windows tax. Responding in email, Jim Gray reminded me that TerraServer does offer SOAP interfaces [1, 2]. And yet those interfaces demonstrably have not inspired a flurry of innovation. Why not? Microsoft is obliged to portray the Web-based user experience as a dead end that can never be improved, and Windows as the only way forward. So it's going to be Visual Basic and client/server all over again: Windows applications will control the user experience; servers will dish out the data; developers will connect the dots."

For a preview of the TerraService and MapPoint Web service client projects, check out the illustrated online help page on the OakLeaf Web site.

Like OakLeaf's public CFR Web services, the TerraService and MapPoint Web services take full advantage of Visual Studio's Add Web Reference Wizard to autogenerate .NET 1.0+ C# or VB .NET Web service proxy classes. Simplified programming of data-driven Web services and the capability to autogenerate Web service client proxy classes were Microsoft's primary marketing topics for .NET 1.0. VS 2005's new built-in Web server simplifies development and publication of basic SQAP 1.1 and 1.2 Web services, but offers few other Web service enhancements. SQL Server 2005 boasts the capability to host native (in-process) SOAP Web services that substitute Windows Server 2003's and Windows XP's built-in HTTP.sys driver for IIS and VS-generated .asmx files. Promotion of the WS-Security and other WS-* specifications, Web Services Enhancements (WSE) 2.0+, and the forthcoming Indigo message bus appears to have increased the FUD factor surrounding SOA in general and Web services in particular. This uncertainty has reduced IT management and developer interest in implementing basic .NET and Java Web services as the first step in demonstrating the practicality of enterprise-level SOA to corporate and line-of-business management. Perhaps the Web services developer community would be better served by more publicly accessible, data-intensive, real-world Web services and fully implemented client examples, such as the TerraService and MapPoint projects, rather than white papers that describe esoteric SOAP headers and emerging messaging "standards." RSS 2.0 and ATOM adhere to the Keep It Simple, Stupid (KISS) principal, which is the secret to their current success in content syndication. --rj

P&P Enterprise Library, VB, and Data Access Application Block Bugs

Microsoft's patterns & practices group (PAG) released an updated version of seven Application Blocks for the .NET Framework 1.1 in mid-January. The new Enterprise Library (EntLib) Application Blocks are based on the Avenade Connected Architecture for .NET (ACA.NET). Avenade Inc is a joint-venture software consulting company formed by Accenture and Microsoft in 2000. I had created a sample VS 2005 front-end to the earlier (Microsoft-only) version (2.0) of the Data Access Application Block (DAAB) for the book's "Best Practices" chapter, and I expected the new EntLib blocks to be similar to their predecessors. Instead, I discovered that the source code for all new blocks was C# only. (The earlier blocks had C# and VB implementations, and I had no problems compiling them with VS 2005 May 2004 CTP.) The missing EntLib VB implementations have evoked the expected howls of protest from the developer community. As an example, vice-president of the Microsoft Developer Division S. "Soma" Somasegar's "Enterprise Library" blog post of 3/13/2005 (from India) drew several comments regarding the missing VB block source code. Michael Kropp, who runs the PAG team, replied to Bill McCarthy's complaint with the following promise:

While we made a decision not to provide all the application blocks in multiple languages we intentionally did a number of things to specifically help VB developers successfully use Enterprise Library (API Reference documentation, code samples, quickstarts and documentation). Going forward, we plan to make additional investments in Enterprise Library to include end-to-end reference implementations in both VB and C#.

Hopefully, VB block implementations will be included in the promised upgrade for the upgrade to .NET 2.0 and VS 2005. Attempts to build the EntLib source code with VS 2005 February CTP failed as a result of Upgrade Wizard errors. In addition, the architectural changes caused the new blocks to be totally incompatible with their predecessors. So, I compiled the EntLib DAAB source code with VS 2003, upgraded the DataAccessQuickStart.sln client project to .NET 2.0, added the .NET 1.1 Common.dll, Configuration.dll, and Data.dll assemblies as references, created the sample database in SQL Server 2005 February CTP, and attempted to run the client. I had to make a minor modification to the app.config file (removing the <keyalgorithmstorageprovider nil="true"> element) to eliminate a runtime error when creating a new abstract Database object. At first glance all seven test buttons delivered the expected results. A second look at the Update a Database Using a DataSet button's result ("2 rows were affected") piqued my curiosity. The SalesData class's UpdateProducts function includes a Dim rowsAffected As Integer = db.UpdateDataSet(productsDataSet, "Products", insertCommandWrapper, updateCommandWrapper, deleteCommandWrapper, UpdateBehavior.Standard) instruction that executes three commands and should return "3 rows were affected." Inspecting the underlying Products table last row(s) indicated that the deleteCommandWrapper command didn't behave as the Microsoft developers had expected. They had expended large amounts of energy on unit tests for the block code but didn't even run cursory tests on all functions of the QuickStart client code. The basic problem was an attempt to add and delete the same record in a single operation. The required @ProductID identity value for the DeleteProduct stored procedure isn't available until the the AddProduct stored procedure executes and updates the Products DataTable. The DataRowVersion.Current parameter value of the cwDelete.AddInParameter("@ProductID", DbType.Int32, "ProductID", DataRowVersion.Current) instruction always is DbNull.Value. The moral of this story is that developers who thoroughly test only the "interesting" parts of their projects end up with egg on their faces.

Another Alternative to AdventureWorks

Before the days of COM-based VB6 Web Services, I wrote a VB6 client/server demonstration project that emulated early Web-based retail supply chain management (SCM) applications. The "smart client" front end generated orders for consumer electronics products, and included full support for drop shipments from one of three distributors. Unlike AdventureWorks and Northwind, the project included invoicing, credit-card processing, and drop-shipment purchasing from and invoicing by the distributors. Freight charges were obtained from the UPS Web site. The unified client contained about 20,000 lines of code. The SQL Server 2000 back end consisted of six databases: OCE_Cust for sales order processing and invoicing, OCE_Prod for warehousing and shipping, OMB_Network (OmegaBank) for credit card processing, and three distributors: AlphaDist, BetaDist, and GammaDist. The design was based on separate physical locations for each database. For an early book on .NET Web services, I upgraded the Windows client to VB 2002 (with the Wizard) and converted the credit-card and distributor classes to .NET Web services. A flow diagram, description of the project, typical XML document instances and schemas, and links to WSDL documents and database diagrams are here. The final step was conversion of the client—again with the Upgrade Wizard—to a single ASP.NET Web page, as shown here (click the image to open it in a new window):

The page won't garner any design awards, but it displays all information required to process an incoming order, which takes from 0.5 to slightly over 1 second when generating XML purchase and invoice documents for outsourced items. A live demonstration of the project is available here. My April 2002 "Speed SQL Server Data Access" article discusses relative performance of the three successive versions of the OCE project (see Table 1 on page 2).

Note: The Web Services Interoperability Organization (WS-I.org) subsequently created a sample Retailer SCM interoperability test application that's similar to—but much simpler than—the OCE project. WS-I requested permission to republish the flow diagram and some content from the OakLeaf site, which I granted, but I haven't seen any evidence of their use of the material.

AdventureWorks consolidates human-resources, sales, purchasing, and production data in a single database. In the real world, the tables for each schema undoubtedly would reside in separate databases that are under the control of the related groups. Microsoft is heavily promoting service-oriented architecture (SOA) with autonomous Web services, but AdventureWorks' single-database design isn't suited to SOA. A version with four (or more) individual databases, simplified table structures, and more complete sample data would be better suited to the needs of VS 2005/SQL Server 2005 developers and book writers.

VB6 to VB .NET Upgrade Wizard Gets a Bum Rap

As of today, 273 Microsoft MVPs and 4,152 other interested parties have signed a petition to "further develop VB6 and VBA." The petition's request to include "VB.COM" (a.k.a. "Classic VB" and "vbclassic") in future versions of the Visual Studio IDE has fallen on deaf ears in Redmond. One of the petition's FAQs states:

"Microsoft provided a migration wizard intended to ease the transition from VB6 to VB.NET, true. There is near-universal agreement that using this tool to port code assets is an incomplete solution at best, in that leaves myriad "TODO:"'s scattered throughout the translated code. At worst, the migration wizard is an extremely poor choice in that, unlike a complete rewrite, it doesn't take full advantage of all that the new platform offers."

Migrating production VB6 code to VB 2002 or later is a controversial topic, but the VS 2002 Upgrade Wizard worked fine for me. In 2001 I wrote a large-scale ASP Web application (VBScript) with three VB6 data access layer classes (ActiveX DLLs) that connected to a 1-GB demonstration SQL Server database containing the entire contents of the U.S. Code of Federal Regulations (CFR) obtained from the GPO's beta version of the eCFR. (It's still in the beta stage.) Another VB6 project translated the CFR sections' SGML to XHTML and stored the data in a varchar or text column, depending on the size of the section. Subsequently, I used the SOAP Toolkit 2.0 (beta 2) to wrap the classes as RPC/Encoded Web services. Here's the history of the initial version and the first upgrade process:

• CFR-COM (mid-2001) used Active Server Pages and conventional Visual Basic 6.0 ActiveX DLLs for TOC navigation, text display, and full-text search operations. Microsoft Data Access Components (MDAC) 2.7 handled database access.
• CFR-SOAP (late-2001) was CFR-COM upgraded to XML Web services with the Microsoft SOAP Toolkit 2.0 Beta 2 and MDAC 2.7 for database access.

When Visual Studio .NET, as it was called at the time, went gold, I rewrote the client in ASP.NET 1.0. Then I gave the Upgrade Wizard a shot at upgrading the three classes, which totaled about 10,000 lines of code, to VB .NET for conversion to ASP.NET Document/Literal XML Web services. Here's the history:

• CFR-ASPX (January 2002) was CFR-SOAP with all pages upgraded from ASP to ASP.NET with Visual Basic .NET managed code and MDAC 2.7.
• CFR-Final (February 2002) is the final version that replaced COM-SOAP components with ASP.NET XML Web Services for text display, navigation, and search operations. ADO.NET SqlConnection, SqlCommand, and SqlDataReader objects handle database access.

Here are links to the Wizard's Upgrade Reports for the CFRTocWS (table of contents navigation), CFRSectWS (formatted sections), and CFRSearchWS (full-text search). I spent about two or three hours searching for and fixing known VB6 upgrade issues and than ran the Wizard. Each upgrade report had one warning. Not bad for version 1.0 of a very complex code conversion utility, at least in my opinion. I switched from VB6 to VB .NET and never looked back. Writing VBA code to update clients' Access applications now seems to me like trying to write this blog in Brasilian Portuguese.

Note: In August 2002, the OakLeaf CFR Web Service project won the charter Microsoft .NET Best Award for horizontal solutions. The "Use XML Attributes to Navigate Data" .NETInsight article explains the CFRTocWS service's navigation features and "Run Full-Text Database Searches on a Shoestring" describes the CFRSearchWS service.

Northwind vs. AdventureWorks as a Sample Database - Round 2

Here's more on the issue of AW vs. NW as a sample database for books about Visual Studio and SQL Server 2005. This post uses AW's Sales.SalesOrderHeader and Sales.SalesOrderDetail tables as examples.

It's easy to create a master-details form from these two tables but providing the user readable values from foreign-key fields is a pain. As an example, The Sales.SalesOrderHeader table has nine foreign-key fields: CustomerID, ContactID, SalesPersonID, TerritoryID, BillToAddressID, ShipToAddressID, ShipMethodID, CreditCardID, and CurrencyRateID.

Here's a screen capture of a sample project that loads the last 1,000 Sales.SalesOrderHeader rows and related Sales.SalesOrderDetail rows in descending order. Click the thumbnail to open it in a resizable window.

Note: VS 2005 introduces "smart captions" for autogenerated detail view controls. Spaces substitute for underscores and are inserted between a sequence of lower-case and upper-case characters. Unfortunately, DataGridView column headers lack this convenient feature.

Each foreign-key field requires a lookup operation to display readable data from the related table; some values require joins to related sub-tables. Lookups require adding 11 or more TableAdapters to the DataSet. Northwind requires only three or four added lookup tables: Employees, Shippers, Products and, optionally Customers.

Some text boxes, such as Sales Order Number and Total Due are bound to computed columns, and the uSalesOrderHeader trigger updates Revision Number, Sub Total, and Modified Date text box values. Thus, these six text boxes are read-only. The DataGridView's Line Total column also is computed, and the uSalesOrderDetail trigger updates the DateModifed column. However, the autogenerated UpdateCommand.CommandText instructions attempt to update the computed columns, which throws an exception. Elmininating the exception requires editing the dataset code to remove the column(s) from the UPDATE statement and the corresponding parameter(s). Partial classes let you supplement autogenerated code but not alter it directly.

Most AW tables have guid columns, which complicates INSERT operations. You must add code to the DataGridView's UserAddedRow event handler to add a Guid.NewGuid() value to the rowguid Cell. The DateModified column also requires Now or the like as a datetime value, which the iSalesOrderDetail trigger overwrites. Similar modifications are required for the corresponding SalesOrderHeader columns.

AW is a useful sample database for demonstrating gotchas with UPDATE and INSERT operations on tables that have computed columns and values inserted or updated by triggers. However, creating a real-world project with dropdown lists to enable editing the master form and detail DataGridView with readable values is more than a challenge-and-a-half, especially for users new to VS and VB. I'll probably include sample projects similar to these in an advanced chapter and a forthcoming Visual Studio Magazine article. Like the VS 2005 help files, I plan to stick with good old Northwind for the majority of the book's sample code and project examples.

Northwind vs. AdventureWorks as a Sample Database

Microsoft "strongly urges" book authors to abandon Northwind as the default sample database in favor of SQL Server (SQLS) 2005's updated AdventureWorks (AW) OLTP database. AW has a very complex, highly normalized structure of 68 tables. The Visio 2003 and HTML versions of an AW database diagram require printing to at least a 17-inch x 22-inch sheet to be readable. AW uses SQLS 2005's new user-schema separation feature, which lets you substitute an arbitrary prefix for the traditional database owner's name (dbo by default). Thus the fully-qualified name changes from SQLS 2000 and earlier's ServerName.DatabaseName.OwnerName.TableName to ServerName.DatabaseName.SchemaName.TableName. The SchemaName lets multiple users own a single schema based on their membership in roles or Windows groups. AW has five schemas: HumanResources, Person, Production, Purchasing, and Sales.

Note: An Accounting schema and Accounting.Invoices table are conspicuous by their absence. Apparently, the firm accepts and fulfills orders but doesn't issue invoices or charge credit cards for goods shipped. Northwind also lacks an Invoices table.)

The Person.AddressType table has six types of addresses: Billing, Home, Main Office, Primary, Shipping, and Archive. The table contains foreign-key values for StateProvince and CountryRegion tables. Thus a T-SQL query to return all columns for an individual customer address is (from the Sales.vIndividual view):

SELECT i.[CustomerID], c.[Title], c.[FirstName], c.[MiddleName], c.[LastName], c.[Suffix], c.[Phone], c.[EmailAddress], c.EmailPromotion], a.[AddressLine1], a.[AddressLine2], a.[City], StateProvinceName = sp.[Name], a.[PostalCode], CountryRegionName = cr.[Name], i.[Demographics], c.[AdditionalContactInfo]

FROM [Sales].[Individual] i INNER JOIN [Person].[Contact] c ON c.[ContactID] = i.[ContactID] INNER JOIN [Sales].[CustomerAddress] ca ON ca.[CustomerID] = i.[CustomerID] INNER JOIN [Person].[Address] a ON a.[AddressID] = ca.[AddressID] INNER JOIN [Person].[StateProvince] sp ON sp.[StateProvinceID] = a.[StateProvinceID] INNER JOIN [Person].[CountryRegion] cr ON cr.[CountryRegionCode] = sp.[CountryRegionCode]

WHERE i.[CustomerID] IN (SELECT [Sales].[Customer].[CustomerID] FROM [Sales].[Customer] WHERE [Sales].[Customer].[CustomerType] = 'I')

Creating (or attempting to create) an updatable typed DataSet for individual customers requires six DataTables and generates a DataSet designer file of more than 6,000 lines. I haven't yet attempted to write FillBy... methods for a parameterized details form that would enable updates and inserts with bound text boxes and dropdown lists.

Following is a simple DataGridView of the Sales.vIndividual view. Click the thumbnail to open it in a resizable window.

Without a parameterized FillBy... method, the Fill method returns 18,508 records and populates the lookup lists. Thus, it takes about 10 seconds to open the release version of AppName.exe with the client app and SQLS 2005 instance running under Windows 2003 Server on a 2.6-GHz Pentium 4 box with 1 GB RAM. TaskMan shows that the AppName.exe process consumes 101 MB. Parameterizing the seven SELECT queries with individual FillBy... methods can solve the opening time and resource-consumption problems but doesn't address the complexity issue.

Northwind's dbo.Customers table provides readable BillTo addresses, and the dbo.Orders table supplies ShipTo information without requiring a series of INNER JOINs. It's my contention that AW's complexity will distract readers when learning typical tasks, such as creating a Customers-Orders-LineItems master-details-subdetails form. Thus, I'm using old-timey Northwind for most design and coding examples in my book. Another reason for using Northwind is it's popularity in VS 2005's offline help files. Searching the February CTP with 'Northwind and Walkthrough' returns 62 hits; 'AdventureWorks and Walkthrough' returns 3. Similarly, 'Northwind and HOWTO' returns 47 hits; 'AdventureWorks and HOWTO' returns 0. If the SQLS team "strongly urge[d]" the User Ed(ucation) folks who write the VS 2005 help files to use AW, the recommendation appears to have fallen on deaf ears.

Note: One of the book's sample projects lets readers populate a Northwind.SalesOrders table—an updated clone of the Northwind.Orders table—with an unlimited number of rows containg randomized data for sequential dates, a specified average number of line items per order, and random ProductID values.

The Opening Salvo - Visual Studio 2005 and SQL Server 2005

Welcome to the OakLeaf Systems blog. I'm in the process of writing a new book about database programming with Microsoft Visual Basic 2005, the successor to Visual Basic 6.0 and Visual Basic .NET. I plan to share my experiences—good and not so good—with the latest Community Technical Previews (CTPs) and betas of Visual Studio 2005 and SQL Server 2005/SQL Express. Stay tuned ...