Mike Lorengo's Weblog

Development Nirvana

mlorengo — Thu, 22 Feb 2007 01:16:00 GMT

Overview

I recently ditched my AMD desktop machine and picked up an ASUS G1 notebook. I like ASUS products, in fact I based my Media Center around an ASUS motherboard. With the recent release of Microsoft Virtual PC 2007, I took the opportunity to build out a W2K3R2 based development server. After insalling Visual Studio 2005, I had to add a bunch of tools to complete my development environment. As I add these items to my VM, I'm going to capture it and update it for future builds. I do anticipate creating a Vista VM as soon as the VS2005 SP is availabe, so I will need to go through this procedure at least one more time.

Visual Studio

Technorati tags: VS2005, AddIn, Development, Programming, .Net 3.0

Visual Studio 2005 SP1: It's important that you read Jon Galloway's Things I wish I had known before installed Visual Studio Service Pack 1, before installing this update. In addition you may run out of contiguous memory space during the install, so you should probably follow these intstructions before attempting it.

.Net Framework 3.0

Microsoft® Windows® Software Development Kit for Windows Vista™ and .NET Framework 3.0 Runtime Components: The Windows SDK includes content for application development with the APIs in Windows Vista, including the .NET Framework 3.0 technologies: .NET Framework 2.0, Windows® Presentation Foundation, Windows® Communication Foundation, Windows® Workflow Foundation, and Windows CardSpace™.
Visual Studio 2005 extensions for .NET Framework 3.0 (Windows Workflow Foundation): Compatible with the released versions of the 2007 Microsoft Office system, Microsoft Windows Vista, and the .NET Framework 3.0 Runtime Components. Windows Workflow Foundation is the programming model, engine, and tools for quickly building workflow-enabled applications on Windows. It consists of a Microsoft .NET Framework 3.0 namespace (System.Workflow), an in-process workflow engine, and designers for Microsoft Visual Studio 2005.
Visual Studio 2005 extensions for .NET Framework 3.0 (WCF & WPF), November 2006 CTP: The Visual Studio 2005 extensions for.NET Framework 3.0 (WCF & WPF), November 2006 CTP provides developers with support for building .NET Framework 3.0 applications using the released version of Visual Studio 2005. This is the last release of this technology that will be in the form of a set of extensions for Visual Studio 2005.

Web Services

Web Service Software Factory: The Web Service Software Factory is a cohesive collection of guidance that enables architects and developers to design and build higher quaility services and achieve more predictable results in less time by applying proven practices and patterns directly within Visual Studio 2005. Requires the Guidance Automation Extensions and Guidance Automation Toolkit Download (GAX) expands the capabilities of Visual Studio 2005 by allowing architects and developers to run guidance packages, such as those included in Software Factories, which automate key development tasks from within the Visual Studio environment.
Web Services Contract First (WSCF): WSCF offers a simple yet powerful WSDL Wizard that abstracts away all the nitty-gritty details of WSDL and therefore does not give room for making errors and wrong assumptions just by trying to use and applying everything that can be done stated by the original WSDL specification.

Unit Testing

NUnit: NUnit is a unit-testing framework for all .Net languages. Initially ported from JUnit, the current production release, version 2.2, is the fourth major release of this xUnit based unit testing tool for Microsoft .NET.
MBUnit: MbUnit is an extensible .Net test framework. As in NUnit, tests are created at runtime using Reflection and custom attributes. MbUnit differentiates itself from NUnit in it's extensibility model. It contains a number of tests that go beyond the simple unit testing, such as a greater assert range, pairwise testing, combinatorial testing, data oriented testing, etc...
NUnitAddIn: The NUnit Addin allows you to run unit tests inside Visual Studio 2005.
JetBrains UnitRun: ReSharper UnitRun is a free add-in for Microsoft Visual Studio 2005 that allows you to automatically run and profile unit tests. This user-friendly tool detects test fixtures of the supported unit testing frameworks and lets you run or profile them right from the code editor or from Visual Studio's Solution Explorer. An MBUnit PlugIn for UnitRun is available as well.
NUnit Code Snippets for Visual Studio 2005: For NUnit users who have switched to Visual Studio 2005 and would like a reasonably decent set of keyboard shortcuts for some of the most common uses of the NUnit framework, you might find these code snippets to be of use.
Simple TDD Visual Studio Templates: To speed up the creation of solutions using NUnit or MbUnit? for unit test, use these two templates to plug into Visual Studio to do it for you. There's one for testing with MbUnit? 2.3 RTM and one for NUnit 2.2.9.
Spell Checker for HTML and ASP.NET pages: Spell checker works in Source view, it is able to extract text from markup elements and use Office 2003 spell checker to check the text. Spell checker is able to handle entities to some extent (they are considered whitespace for now).

Refactoring Tools

Resource Refactoring Tool: The Resource Refactoring Tool provides developers an easy way to extract hard coded strings from the code to resource files.
DevExpress Refactor for ASP.NET Version 2.1.3: Refactor! is freely available to all ASP.NET 2.0 developers and offers a comprehensive suite of tools that enable you and your team to simplify and shape complex code and HTML markup - making your web applications easier to read and less costly to maintain.
Cool Commands 4.0: Lots of helpful shortcuts for accomplishing tasks.
SlickEdit Gadgets: Some useful, some fun utilities to help out your IDE experience
- Editor Gadgets: A Line Ruler, Indentation Guide, Auto Copy selection, and a fun Editor Graphic.
- Command Spy: Learn what commands are attached to what keyboard shortcuts to help you become a keyboarding guru.
- File Explorer: Use the File Explorer rather than the IDE open file dialog to drag and drop files into your project.
- Data Object Analyzer: Inspect the contents of any clipboard operation, or drag-and-drop operation, from any other application.
- The SLOC Report: An easy way to count the lines of code. The line count is divided into three categories: code, comments, and whitespace.
Microsoft Visual Studio 2005 IDE Enhancements: Visual Studio 2005 IDE Enhancements are a set of Visual Studio extensions that are designed to make you more productive. These enhancements are directly integrated into the Visual Studio IDE.
- Source Code Outliner: The Source Outliner tool is a Visual Studio extension that provides a tree view of your source code's types and members and lets you quickly navigate to them inside the editor.
- Visual C++ Code Snippets: The Visual C++ Code Snippets tool lets you insert snippets in your code by using a pop-up menu that contains programming keywords. VB.NET and C# languages have this functionality in Visual Studio 2005.
- Indexed Find: The Indexed Find tool is a Visual Studio extension that uses the Microsoft Indexing Service to provide improved Search capabilities to the integrated development environment (IDE). It sends the results of a search to the Output Window.
- Super Diff Utility: The Super Diff Find tool is a Visual Studio extension that compares text files. It uses color coding and graphics to show the difference between the files in deleted text (red), changed text (blue), inserted text (green).
- Event Toaster Utility: The Event Toaster tool is a Visual Studio extension that notifies users about specific events within the Visual Studio IDE.
XPathMania: An XPath Tool Window for Visual Studio 2005. Produces by the Mvp.Xml project team.

Configuration Management

Subversion: Subversion is a free/open-source version control system. That is, Subversion manages files and directories, and the changes made to them, over time. This allows you to recover older versions of your data, or examine the history of how your data changed.
TortoiseSVN: TortoiseSVN is a really easy to use Revision control / version control / source control software for Windows.
Since it's not an integration for a specific IDE you can use it with whatever development tools you like.
TortoiseSVN is free to use. You don't need to get a loan or pay a full years salary to use it.
CruiseControl.Net: CruiseControl.NET is an Automated Continuous Integration server, implemented using the Microsoft .NET Framework.
Trac: Trac is a web-based software project management and bug/issue tracking system. It provides an interface to Subversion and an integrated wiki. Setting up Trac was a little involved, and that will form the basis for another blog post.

My Media Center Specs

mlorengo — Wed, 26 Apr 2006 14:10:00 GMT

I just recently built my own media center from scratch. The goal was to have a Media Center server in my house, on which I would store all of my DVD's, CD's, Photo's and Recorded TV programs. All of this information would be accessible from any room in the house, and on the road via the Internet. A few people have asked me about the specs for my machine, so I thought I would just post them here.

The goal of this computer was to be my Home Media Server, and therefore the specs are a little more than what is needed. I currently have Media Center 2005, Rollup 2 loaded, but my hope is to have Vista running on it soon.

The Media Center server is just the first step in "The Big Plan". I'll be making posts off and on again, as I add to it.

Component	Model
Motherboard	ASUS A8N-SLI Premium Socket 939 NVidia nForce4 SLI ATX AMD
	I chose this motherboard, because it has a heatpipe cooling system on the Northbridge chipset, is SLI capable, accepted the Dual Core AMD processor, and support for overclocking.
CPU	AMD Athlon 64 X2 3800+ Manchester 1GHz HT 2x512KB L2 Cache Socket 939 Dual Core
	I chose this processor because it offered the best price per performance at the time.
Video Card	eVGA 256-P2-N553-AX GeForce 7600GT CO 256 MB 128-bit GDDR3 PCI Express x16
	I chose this video card because it is SLI capable, and wanted some ability to handle Vista when it comes out.
Hard Drive	Western Digital Caviar SE16 WD50000KS 500GB 7200 RPM 16MB Cache SATA 3.0Gb/s
	I chose this hard drive because I wanted to max out storage on my server. I paid a premium for the 500GB capacity, but it's quiet. I just recently added a second one bringing my capacity up to 1TB. I still have room in my case for 2 more drives.
Memory	Corsair XMS 1GB (2x512MB) 184-Pin DDR SDRAM Unbuffered DDR 400 (PC 3200) DUal Channel Kit System
	I like Corsair memory, it has heat spreaders, and the price was right.
CPU Fan	Zalman CNPS9500 LED 92mm 2 Ball Blue LED Light Cooling Fan with Heatsink
	I chose this fan, because Zalman makes great cooling products, in addition I was intalling this in a Zalman case with a Zalman power supply.
Case	Zalman HD160-B Home Theatre Aluminum PC Enclosure w/Remote (Black)
	I chose this processor case because it was all aluminum, looked like a media center component, and got good reviews. I also seriously considered getting a Origen AE Case

I already had a Hauppauge PVR 350, TV Tuner card so I recylced that into the system for watching and recording live tv.

PDC Reflections (Day 1)

mlorengo — Sun, 18 Sep 2005 09:37:00 GMT

Here it is the Saturday after the PDC, and I think I'm finally ready to reflect on what was one of the most intense learning weeks I've had in a really long time. In my opinion this PDC represents some very sweeping changes that will impact .Net developers for years to come.

We've been hearing about technologies such as Longhorn, Indigo and Avalon since PDC03, but this time around the tools and models are nearly complete, add in Windows Workflow Foundation, all of the .Net 2.0 technology, then sprinkle with future technologies like Language Integrated Query (Linq) and you've got quite a dose.

I'll be posting a series of blog entries to help organize my thoughts around the different topics.

This first entry has to do with Day 1 technologies.

The Goods

Amy and I arrived at the Keynote Hall a tad early. We decided to arrive around 7:00am so that we could register and get "The Goods" before the Keynote started. The Goods consisted of 5 DVD's containing Windows Vista Beta 1 and the PDC:CTP, Visual Studio Team System Beta 2, WinFX, SQL Server 2005 CTP's, LINQ Technology, and eventually VS2005 RC and Longhorn Server.

Most of this can be downloaded online

Other technologies that were introduced over the course of the PDC can be found here

Monad Shell (Microsofts new Command Shell)
Atlas (.Net implemenetation of AJAX and more)
Language Integrated Query (Linq, D Developing an Advanced ASP.NET Server Control with Rich Design-TimeLink, XLinq)
SQL Server 2005 Express
Windows Workflow Foundation
Acrylic (Graphics Design Software)

That should be all you need to have your own little PDC in a box. What you can't get is the Windows Vista PDC CTP, or VSTS versions. Now onto the summary.

We were treated with an opening video "Change The World", and then Bill Gates started his Keynote, in which he showed a video about recruiting for Microsoft. Bill and Chris Capossela talked about Windows Vista and Office 12, RSS and IE7. After that Jim Allchin came out and discussed some enhancements to Windows Vista, including SuperFetch with aims to speed up application load times, as well as the ability for Vista to utilize USB Flash Drives for extra memory (too cool!).

For me, the meat of the morning keynote was the demonstration of Indigo, Avalon and Atlas. These are technologies that will have a release around mid/late 2006. The Linq demo while impressive, is still too far away for me to get too excited about.

After that some demo's were shown by North Face marketers (here's a whitepaper on how it was built), and some information on a new Windows Vista sample application Microsoft Max, and the DigitalLocker.

I was planning on attending the MS Build breakout session, but unfortunately the Keynote went long and the session was cancelled. Also, I can't help but mention the Phone Deal that Went Sour.

Instead I attended the session PRS401 ASP.Net: Developing an Advanced ASP.NET Server Control with Rich Design-Time. Eilon Lipton presented ways to create ASP.Net server controls and at the same time create a Designer for each. It was a good overview. I plan on experimenting with this a little more in the future. (Incidentally you can download the ppt slides from most of the sessions here).

The next session was on TLN303 Monad: Advanced Command Line Scripting. While it was interesting, it wasn't what I was expecting. They gave an overview of Monad, and did a compare/contrast with other scripting languages. I guess I was hoping for a more detailed, getting things done sort of talk.

The final breakout session of the day was PRS404 ASP.NET: Developing Advanced Custom Providers for Membership and Role Manager. This was the best breakout session of the day, Stefan Shackow covered the new 2.0 provider model and covered some limitations and gave possible workarounds for certain situations. He has made the code available here.

After that Amy and I headed to the big room to get our schwag! I think we ended up with about 13 T-Shirts each, plus hats, plus a book and a bunch of other little trinkets. It was kind of embarassing, hauling around our loot in our arms. We finally ducked into the PDC Marketplace to ask for a bag. Amy's bag lasted all of 5 minutes before it busted under the weight. :)

After a quick dinner on appetizers, we hopped a bus to head to the www.pdcunderground.com talks at the Westin. We got there just in time to catch the last 3 speakers. After that it was back to the PDC where we caught a couple of BOF's. Finally we headed back to the hotel around 11pm. It was time to get some rest, for tomorrow we would be doing it all over again.

Blog Clippings #2

mlorengo — Wed, 18 May 2005 21:11:00 GMT

I was having lunch today when it occurred to me that I haven't blogged an entry in at least 3 months. Part of this has been due to my laptop going belly up, my new job, and just plain old fashioned laziness. So until I get the data off my laptop, this post will be in the spirit of laziness.

Powertoys for VS2005 Beta 2

From the Class Designer team blog... "The Design Tools Enhancements add-in provides a common set of features that can be used by both the Class Designer and the Distributed System Designers"

GTDTiddlyWiki

A simple html file based wiki. Simply download to your thumb drive and you have your own personal wiki. For those not in the know GTD stands for Getting Things Done, here's a wiki that talks about GTD.

Smell To Refactorings

When looking at code you begin to notice certain smells. This table organizes the smells, and proposes refactoring operations to clean up the smell.

Designing Great Frameworks

Brad Abrams presents a series of lectures on design guidelines for developing classes and components that extend the .NET Framework.

Microsoft Solutions Framework (MSF) for Agile Software Development, Beta

This download contains the agile process guidance that will be shipping with Visual Studio Team System. MSF for Agile Software Development is a scenario-driven, context-based, agile software development process that utilizes many of the ideas embodied in Team System. This process incorporates proven practices developed at Microsoft around requirements, design, security, performance, and testing.

Nic Wolfe's Password Generator

Now only remember a single pass phrase yet still have highly secure unique passwords for all of your sites.

start.com Starting to Heat Up

mlorengo — Sun, 20 Mar 2005 23:37:00 GMT

I first found out about the www.start.com/1 and www.start.com/2 prototypes from Dare Obasanjo's post on the subject here. It seems that Microsoft is getting into the web-based Rss aggregator market. I was first introduced to RSS aggregator via Dare's awesome RSS Bandit client many moons ago. However as I bounced around from work and home, it became increasingly difficult to keep the feed read states in sync. In addition, if I added a feed at work, I'd have to add it at home as well. I seem to remember hearing something about RSS Bandit providing synchronization to get around this issue, but I abandoned it long before that.

My new favorite aggregator became Bloglines. I'm still using it to this day. I've made my subscriptions public, so you can see what I read. Plus you get to see how disorganized my feeds are. What do I like about Bloglines?

I can read it at Home, Work, at a friends house, the library, or across the world.
I can use the Bloglines toolkit for Firefox and right click on any feed I come across and add it to my subscriptions.
I can share my blog subscriptions
I can search my blog subscriptions or all of the Bloglines subscriptions
I can clip blogs and retrieve them from any other computer
I can read Bloglines on my PocketPC
And many many more cool things

So what does this have to do with www.start.com? Well, as I said, I heard about this a little over a week ago, and the progress that has been made so far is astounding. Not only that but Firefox support is coming along nicely.

What suggestions do I have for the start.com site?

Allow for a public API
Allow me to import my Bloglines subscription (see suggestion 1)
Make it easy for me via a Firefox/IE extension to add feeds to my profile while I'm browsing around the web and find a cool feed
Allow me to create feeds from searches against the start.com catalog of feeds
Allow me to clip a blog entry for later perusal
Make it easy to rearrange feed/categories perhaps via AJAX or is it XAP?

Steve Rider is blogging about the latest goings on here. In addition he's added a feed to help keep us up to date on the latest and greatest changes to start.com here. Looks like Scoble can't fire him now.

All Things Indigo - Indigo Resources I've Found

mlorengo — Sun, 13 Feb 2005 18:19:00 GMT

The big buzz ~~coming~~ is the March CTP Release of Indigo, one of the ~~three~~ two :) pillars of Longhorn. I'm going to be gathering some posts in this entry from around the blogosphere in hopes of relieving my brain of some more details. In addition I won't have to save these links on my machine which gets repaved every three months or so.

News Flash

The Beta1 RC of Avalon & Indigo has been released to the wild, thanks to Drew Marsh for the link, you can get the direct download here Supposedly it will be available via this page.*

Blog Entries

Clemens Vasters creates a simple Hello World Indigo Application.
Bruce Williams takes Clemens' Hello World sample and converts it into a Queued Service.
Shy Cohen demonstrates Indigo's Duplex Channel and use of the config to set endpoint bindings.
And of course Don Box talks about Service Contracts in Indigo.
Clemens Vasters spends the weekend with Indigo (Part 1,Part 2,Part 3)

Indigo Blogs

Resources

Channel 9 Indigo Wiki
March 2005 CTP Release Notes
Indigo Doucmentation for March 2005 CTP
Indigo Portal in the Longhorn SDK
Indigo Section in the Longhorn Developer Center
Distributed Applications Using "Indigo" Services in the Longhorn Development Center
Indigo MSDN Newsgroup

Articles

Introducing Indigo: An Early Look by David Chappell (Feb. 2005)
Inside "Indigo", Chapter 2: The Journey of a Message by Jason Clark (Feb. 2004)
Codename Indigo: A Guide to Developing and Running Connected Systems with Indigo by Don Box (Jan. 2004)
Creating Indigo Applications with the PDC Release of Visual Studio.Net by Yasser Shohoud (Jan. 2004)

Samples

Indigo Transactional Web Services using Microsoft WinFX CTP March 2005* (April 2005) - Implementing WS-AtomicTransaction in Indigo. by Jamesqding

Podcasts

.NET Rocks Mood Indigo* (April 2005) - Carl talks with Richard Turner and Anand Rajagapolan, Program Managers on the Indigo team to really dig into the Indigo experience.

Videos

BizTalk Server and Indigo* (May 2005) MSDN TV - Scott Woodgate shows a glimpse into the future with a working prototype demo illustrating BizTalk Server and "Indigo" working together.
Indigo Security in a Nutshell* (April 2005) MSDN TV - Doug Purdy provides an overview of the Indigo Security Programming Model.
Introduction to Indigo* (April 2005) MSDN TV - Steve Swartz provides a brief conceptual overview of Indigo, walks through some code, and introduces you to his jackalope.
Via Gene Webs blog entry:Indigo Webcasts including a) Programming Indigo 1, b) Programming Indigo 2, c) Indigo Security, d) Reliability, e) Tooling for Distributed Applications, f) The IT Pro Experience, g) Migration, Upgrade, Integration, Interop, h) Basics, i) Indigo Internals 1, j) Indigo Internals 2.
Don Box - Final Part of Indigo Team Tour via Channel 9 (Sep. 2004)
Don Box - Tour of Indigo Team via Channel 9 (Aug. 2004)
Service Orientation and Today's Technologies via MSDN TV (Aug. 2004)
The Road to Indigo Technology Roadmap

Conference Sessions

Indigo Day at VSLive (Feb. 2005)
Summary via The Server Side
Eric Rudder showed of Indigo in the VSLive! Keynote
Summary of Eric Rudder's Keynote and Don Box's & Steven Schwartz's session on Programming Indigo via Mujtaba Syed
The Videos are now online here

*Updated Entries. If you know of an Indigo resource or blog (preferably with an Indigo category, please send me a note. Also a comment was made that anything written prior to Feb 05 should be disregarded since these papers will have been written about a very, very different Indigo and may end up being more confusing than useful!

Argument Exception Handling Utility Classes

mlorengo — Sun, 06 Feb 2005 11:29:00 GMT

I just spent twenty minutes trying to track down this article in my blog clippings. I'm posting it here so I can find it a little faster next time. Nils Jonsson created a utility class for simplifying argument checking. The post that triggered my search for this article was K. Scott Allen's entry on argument checking in the Enterprise Library blog entry.

Optimizations in Complex Object Creation Part Deux

mlorengo — Thu, 03 Feb 2005 14:32:00 GMT

In comments to my article on Optimizations in Complex Object Creation, Frans Bouma suggests altering my query to use subqueries for the selecting the data rather than inner joins. I must confess, that I am not the all knowing Sql guru, so I appreciate the feedback, and it certainly makes sense.

To quote from his comment...

Method 2 is often the slowest. If you have 10 wines for a producer and per wine 10 varieties, you have 1 + 10 + 10*10 queries.

If you use:
fetch the wines.
Use the wines WHERE clause as the where clause for a subquery to fetch the winevarietal rows and use THAT where clause as the where clause for the subquery for varietal, you'll get 3 queries.
Then merge the results in memory, which is very straight forward. This is different from method 3 in that you don't pull a lot of data you don't want with the join, as joins can produce a lot of redundant data.

Well, wouldn't you know it, I'm using the slowest method currently. But again, it all depends how I want to use the data, let's say I don't want to lazy load the information, Frans suggests redoing my Select cause. So, after doing that, I now have something like this, that I can use for Method 3, which should reduce the amount of data coming back.

SELECT Wine.Id, Wine.ProducerId, Wine.Name, Wine.Vintage, Varietal.Name, WineVarietal.Percent
FROM Varietal, 
  (SELECT * from winevarietal, 

    (SELECT * from Wine Where wine.producerid=[@producerId)
      WHERE winevarietal.wineId = Wine.id) As WineSubQuery
 
    WHERE ((([WineVarietal].[VarietalId])=[Varietal].[Id]))
ORDER BY Wine.Id;

Dru, comments as to why I don't return a DataSet/DataTable to get all of the data and then populate my objects.

My answer is that there is no reason that I couldn't, and in fact I have some methods in my IWineProvider interfaces that return the results as a DataSet. It's just that I don't need (right now) the extra overhead of a DataSet in loading my objects. And again, I can choose to modify the underlying methods as I or other people suggest improvements (as well as using the Sql suggestions).

So, Frans & Dru and anyone else thanks for the feedback, and if I missed the point you were trying to make, please let me know.

Optimizations in Complex Object Creation

mlorengo — Wed, 02 Feb 2005 19:25:00 GMT

In comments to my article on Handling Complex Objects in the GridView Control, Eric Wise mentions that one needs to be careful about Handling Complex Objects in Grids capabilities with complex objects. Why? If I understand correctly, it's a database issue. You want to avoid loading up all of that information up from the database unless your really want to.

He proposes two alternatives

Expose inital load methods that make the huge database call and prepopulate the referenced objects in one shot.
Expose domain level queries that execute single sql statements and return datatables.

I'm going to explore method 1 in this post, because I can already use method 2 with my object model. I simply don't include those properties in my binding list that require complex types. Unless I'm missing something, I don't see any advantages of method 2 over this solution. I can then drill down from the summary information as needed to the detail. This is known as Lazy Loading.

So, with the assumption (mine), that I can do method 2 above with my current object model, and still retain a level of separation, let's look at how we can improve the data loading, *if* we've decided that we *do* want to display "complex" objects on the GridView.

One way to optimize the object loading would be to reduce the number of times we go to the database. Here's a layout (via Access) of my data tables.

If I wanted to load all of the Wines associated with a Producer, I could do it via the following methods

Return a DataReader of Wine records, and then lazy load the WineVarietal (and Varietal) records in another DataReader as needed
Return a DataReader of Wine records and for each one get another DataReader to load the associated Varietal information
Return a DataReader of Wine records with multiple INNER JOIN's and iterate through the records building the Wine and Varietal information
Issue a batch of Sql Statements for each set of information and use the IDataReader's NextResult() method to advance the DataReader to the next result set while building the Wine and Varietal information.

I'm currently using method 2. And it goes something like this:

Get the WineDataProvider
Get an IDataReader from the WineDataProvider for all the Wine produced by the specified Producer.Id

For each record in the IDataReader
    Build a Wine instance from the IDataReader
    Get the VarietalDataProvider
    Get an IDataReader from the VarietalDataProvider for the current Wine.Id

    For each record in the IDataReader
        Build a Varietal instance and add it to the Varietals collection

    Add the Wine to the collection

Return the collection of Wines

Method 1 (Lazy Load) would look very similar to Method 2, with the exception that the second IDataReader would be delayed until the Varietals collection was accessed. In my GridView example, it would be deferred until I bound the Wine instance to the row and accessed the Varietal collection in the <ItemTemplate>. So, again I would expect very similar performance. Of course you don't know until you measure right!

Method 3 would retrieve all the necessary information in one database call. The stored procedure would something like this. Now, there's most likely some optimizations that I can make to the stored procedure, but that's not the purpose of this post. This would give me all of the Wines for a particular Producer. In fact it would give me multiple records for a Wine if the wine had multiple Varietals.

SELECT Wine.ProducerId AS ProducerId, Max(Wine.Id) AS WineId, Max(Wine.Name) AS WineName, WineVarietal.Percent, Varietal.Name
FROM Varietal 
INNER JOIN ((Producer INNER JOIN Wine ON Producer.Id = Wine.ProducerId) 
INNER JOIN WineVarietal ON Wine.Id = WineVarietal.WineId) ON Varietal.Id = WineVarietal.VarietalId
GROUP BY Wine.ProducerId, WineVarietal.Percent, Varietal.Name
HAVING (((Wine.ProducerId)=[@producerId]));

Some sample output for the above statement could look something like this

**Query1**
ProducerId	WineId	WineName	Name	Percent
181	1	2001 Ravenswood Winery Shiraz-Cabernet	Cabernet Sauvignon	75
181	1	2001 Ravenswood Winery Shiraz-Cabernet	Shiraz	25
181	82	2000 Ravenswood Winery Pickberry Vineyards Sonoma Mountain	Merlot	53
181	82	2000 Ravenswood Winery Pickberry Vineyards Sonoma Mountain	Cabernet Sauvignon	45
181	82	2000 Ravenswood Winery Pickberry Vineyards Sonoma Mountain	Cabernet Franc	2

So, now my logic for building up an instance of a Wine would look something like this

For each record in the IDataReader
    If the WineId = current Wine instance Id
        Build a Varietal instance and add it to the Varietals collection
    else
        Build a Wine instance from the IDataReader
        Add it to the Wine collection
        Build a Varietal instance and add it to the Varietals collection

Return the Wine collection

Again, check my logic, but you get the idea. Here we tradeoff making a database call for every object and instead load it up all at once.

Finally Method 4

John Kilgo discusses using the IDataReader.NextResult() method here. In this method I would forego the inner joins of the previous example and instead issue 2 selects, 1 to retrieve the Wine records, and the second to retrieve the Varietal records. I would imagine the logic would go something like this

For each record in the IDataReader
Build a Wine instance and add it to the collection

Advance the IDataReader to the next result set

For each record in the IDataReader
Retrieve the Wine instance from the Wine collection based on the Wine.Id
Build a Varietal class and add it to the Varietals collection

Some optimizations can be made to the above code as well

Conclusion

As you can see there are many different ways to handle the data, it all depends on how you are going to use it. In fact you could implement the data retrieval algorithms based on a strategy pattern, then once you understand what method works best for a specific situation, you simply adjust the data retrieval strategy.

Did I mention how much I love blogging? It's great to get feedback. It's the one thing I've missed since I took about 5 months off. Did I mention I'm starting back to work on Monday? Yup.

Handling Complex Objects in the GridView Control

mlorengo — Wed, 02 Feb 2005 14:41:00 GMT

The new GridView in Asp.Net 2.0 makes displaying business objects just as easy as displaying DataSets. However when it comes to displaying complex business objects (my definition) it isn't that apparent (at least to me) how to go about doing this.

When I say complex business object, I mean a business object that is composed of more than just properties of simple types (int, string, etc), but instead has properties consisting of other objects or even collections of objects. A prime example of this in my case is the Wine class.

In the class diagram above, you can see that the Wine has a Varietals property that is a collection of WineVarietal classes, in turn, the WineVarietal class is composed of an Id, Percent, and a Varietal class. We're talking complex, not rocket science, but complex. The Class diagram illustrates how I chose to model the relationship of a Wine to it's Varietal. In this case a Wine can be composed of one or more varietals of grape in varying percentages. Please see my Varietal is the Spice of Life for more details on the modeling.

In the declaration of the .aspx page GridView control, I have the following fields bound. Notice the Varietals field, it's the same as all of the others.

<asp:gridview id="wineGridView" runat="server" AutoGenerateColumns="False">
  <Columns>
    <asp:BoundField HeaderText="Id" DataField="Id" SortExpression="Id"></asp:BoundField>
    <asp:BoundField HeaderText="Vintage" DataField="Vintage" SortExpression="Vintage"></asp:BoundField>
    <asp:BoundField HeaderText="Name" DataField="Name" SortExpression="Name"></asp:BoundField>
    <asp:BoundField HeaderText="Appellation" DataField="Appellation" SortExpression="Appellation"></asp:BoundField>
    <asp:BoundField HeaderText="Varietals" DataField="Varietals" SortExpression="Varietals"></asp:BoundField>
    </Columns>
</asp:gridview>

However the GridView will display the following. See how the Varietals column simply outputs the class name of the field. What I would really like to do is have it display the list of varietal components, along with there percentage makeup.

Id	Vintage	Name	Appellation	Varietals
82	2000	2000 Ravenswood Winery Pickberry Vineyards Sonoma Mountain	Sonoma Mountain	Lorengo.VirtualCellar.Business.WineVarietalCollection

It turns out that the GridView supports the concept of a <asp:TemplateField> which allows for the customization of the output. What I will need to do is replace the <asp:BoundField> for the Varietals property, and instead insert a <asp:TemplateField>. Inside of the <asp:TemplateField> I will add a DataList control, and give it an id="varietalDataList". For the <ItemTemplate> in the DataList I will use the Container.DataItem property, which should point to a WineVarietal item in my Varietals collection. I can then cast the DataItem to the appropriate type.

For example ((WineVarietal)Container.DataItem).Percent will give me the percentage of the grape varieatal in the current wine. See the code below.

<asp:gridview id="wineGridView" runat="server" AutoGenerateColumns="False" OnRowDataBound="WineGridView_RowDataBound">
  <Columns>
    <asp:BoundField HeaderText="Id" DataField="Id" SortExpression="Id"></asp:BoundField>
    <asp:BoundField HeaderText="Vintage" DataField="Vintage" SortExpression="Vintage"></asp:BoundField>
    <asp:BoundField HeaderText="Name" DataField="Name" SortExpression="Name"></asp:BoundField>

    <asp:TemplateField HeaderText="Varietals" SortExpression="Varietals" >
      <ItemTemplate>

        <asp:DataList id="varietalDataList" runat="server">
          <ItemTemplate>
            <%# ((Varietal)((WineVarietal)Container.DataItem).Varietal).Name %>
            (<%# ((WineVarietal)Container.DataItem).Percent %>%)
          </ItemTemplate>
        </asp:DataList>

      </ItemTemplate>
    </asp:TemplateField>

    <asp:BoundField DataField="Appellation" SortExpression="Appellation"></asp:BoundField>
    <asp:BoundField HeaderText="ItemState" DataField="ItemState" SortExpression="ItemState"></asp:BoundField>
  </Columns>
</asp:gridview>

Ahh, but how does it know to bind the items in the DataList to the WineVarietal object of the current wine? You may have noticed the added attribute in the <asp:GridView> control. To be more specific, OnRowDataBound="WineGridView_RowDataBound".

This let's us add some special code to handle the wiring up of the Wine's Varietals collection in the pages codebehind.

protected void WineGridView_RowDataBound(object sender, GridViewRowEventArgs e)
{
  Wine w = null;

  GridViewRow r = e.Row;
  if (r.DataItem != null) // Make sure we're not in the Header or Footer row
  {
    w = r.DataItem as Wine;
  }

  DataList varietalDataList = e.Row.FindControl("varietalDataList") as DataList;
  if (varietalDataList != null && w != null)
  {
    varietalDataList.DataSource = w.Varietals;
    varietalDataList.DataBind();
  }

For each row in the GridView (including the Header and Footer rows), we will find "varietalDataList" control and update it's DataSource to the Varietals collection of the current Wine object. Then we call it's DataBind() method and woo hoo! Here's what we get!

Vintage

Name

Varietals

ItemState

2000

2000 Ravenswood Winery Pickberry Vineyards Sonoma Mountain

Merlot (53%)

Cabernet Sauvignon (45%)

Cabernet Franc (2%)

Sonoma Mountain

Clean

I like it. Now, I'm not sure if this is the best way to do it. It just works for me. If anyone has any other suggestions as to accomplish this, I would sure be interested in hearing about it. My next step is to code up a general Producer and Wine search page, so I'll take what I've learned today and apply it to those pages.

Further Refinement with the ObjectDataSource Control

mlorengo — Wed, 02 Feb 2005 13:48:00 GMT

In order to take advantage of the new GridView control in Asp.Net 2.0, I've been researching how too use the new ObjectDataSource control, and add support for it in my Virtual Cellar application. I now have two classes in the Virtual Cellar Model that will serve as the candidate classes for use as an ObjectDataSource. A lot of Asp.Net 2.0 examples use the SqlDataSource control to illustrate the GridView features, this entry will list some resources that specifically apply to using the ObjectDataSource

Paging with the ObjectDataControl (Frederik Normen)
Discusses how to implement paging support.
Sorting Objectively (Jonathan Goodyear)
Discusses how to add column sorting support.
MSDN2 Overview of the ObjectDataSource

As I mentioned in my last post, I was trying to determine how to prevent the GridView from being populated on the first page load. I found the following note in the ObjectDataSource documentation.

The ObjectDataSource control retrieves data whenever the Select method is called. The Select method is automatically called by controls bound to the ObjectDataSource when their DataBind method is called. If you set the DataSourceID property of a data-bound control, the control automatically binds data when a Web page loads. However, if you set the DataSource property, you must explicitly call the DataBind method of the data-bound control. You can call the Select method programmatically at any time to retrieve data.

So, it turns out that specifying the DataSourceId property autobinds the control. This is more of that "no-code" model in Asp.Net 2.0. In order to suppress the autobind, I simply removed the DataSourceId attribute from my .aspx page. This prevents the GridView from being populated on the page load. It also removes the link between the GridView and the ObjectDataSource, so I now have to setup a TextChanged handler do to the binding.

protected void ProducerSearchTextbox_TextChanged(object sender, EventArgs e)
{
	producerGridView.DataSource = producerDataSource1;
	producerGridView.DataBind();
}

The above snippet sets the DataSource property to my ObjectDataSource control's Id property. This is equivalent to setting the DataSourceId = "producerDataSource1" property in the .aspx page, except that it doesn't autobind on page load. Then I call the DataBind() method.

I refactored a lot of the Repository code right out of my model. My reason for doing this was to satisfy a requirement that the ObjectDataSource have a parameterless default constructor. It turns out that you can overcome this requirement by handling the ObjectCreating event on the ObjectDataSource and performing my own construction. As it turns out, the new code model seems to fit better, and I eliminated a lot of classes that I just don't need yet if ever. My new class diagram looks like this.

Bigger picture here

The WineDataSource and ProducerDataSource replace the WineRepository and ProducerRepository classes. The Builder classes are responsible for creating actual instances of the Wine and Producer classes, thereby segregating the responsibilities of each class. In order to make the *DataSource classes more friendly to the DataSource configuration wizard, I've decorated them with the following attributes.

[DataObject]
public class WineDataSource
{

  public WineDataSource()
  {
  }

  [DataObjectMethod(DataObjectMethodType.Select, true)]
  public IList<Wine> SelectByProducerId(int id)
  {
    return GetWineList(this.wineProvider.GetByProducerAsDataReader(id));
  }

  [DataObjectMethod(DataObjectMethodType.Select)]
  public IList<Wine> SelectById( int id )
  {
    return GetWineList(this.wineProvider.GetAsDataReader(id));
  }

  //...
}

The DataObject attribute on the class level tells the wizard that this is a DataSource class. The DataObjectMethod attribute exposes the SelectByProducerId(int id) as a DataObjectMethod, and more specifically a "Select" method. The second boolean argument specifies that this is the default "Select" method. Here's how it looks in the designer.

The only items that will display in the drop down list are those that have been marked with the [DataObject] attribute (as long as the Show only data components checkbox is selected). This eases the clutter when your project consists of a lot of classes.

The next step is to choose the Select method. Again, only those methods with a DataObjectMethod and a DataObjectMethodType of Select will be shown.

In conclusion, I've illustrated how to prevent the autobinding from occurring on the page, refined my class diagram to reflect the addition of the new xxxxDataSource classes, and demonstrated the use of the DataObject attributes (which are in System.ComponentModel).

Microsoft Architecture Journal 4 Out Now

mlorengo — Wed, 26 Jan 2005 00:59:00 GMT

From the Blog Clippings Category The Microsoft Architects Journal 4 is now available here as a pdf. The table of contents lists these articles.

Choosing the Right Presentation Layer Architecture
Information Bridge Framework
Benchmarking a Transaction Engine
Enterprise Architecture Alignment Heuristics
Razorbills
Next Generation Tools for OOD

Adapting The Virtual Cellar Domain Objects to .Net using the ObjectDataSource

mlorengo — Tue, 25 Jan 2005 17:05:00 GMT

I've been ever so slowly adding new .Net 2.0 features to the Virtual Cellar code. The most apparent changes are driven by the use of ASP.Net 2.0 features, most notably the GridView control and the new suite of DataSource controls.

I wasn't too happy to see the introduction of the SqlDataSource control, I thought it just encouraged bad layering practices, allowing the population of the new Data Controls through straight ahead SQL statements. Sure, I could setup a stored procedure and call that with the SqlDataSource control, but I was still programming to the database. It just felt wrong.

So, I was rather dejected, I really wanted to use the new GridView control in my application. That's when I found out about the ObjectDataSource control. There's a great MSDNTV Episode that talks about this. This was the perfect solution. I would be able to use my Domain Objects directly with the GridView control. There was only one problem, the object used as the data source needed to have a default contructor.

As you may (or may not) remember, I based the Virtual Cellar model on the Repository Pattern. Unfortunately, I based my domain objects on a base class, RepositoryItemBase, that does not allow for a default constructor, instead it includes a reference to the RepositoryFactory, should any of the classes such as the Wine class need to expose properties of another aggregate root (Producer). Part of this was to get around the problem of having a reference to the ProducerRepository in the Wine class, instead a request would be made to the RepositoryFactory to generate a Factory instance (in this case a ProducerFactory) for re-hydrating the Producer class from the ProducerRepository, from information stored in the Wine class.

The short answer, is that I am unable to use the ProducerRepository as an ObjectDataSource, because it does not support a parameterless default constructor. So my immediate fix is to create an adapter class ProducerDataSource that wraps the functionality of the ProducerRepository, yet has a parameterless default constructor to satisfy the requirements of the ObjectDataSource.

Here's a shot of the class diagram modeled in the new VS2005 diagram tool.

As you can see, I've defined only one method "SelectByName" for the purpose of my testing. The method takes a string that is used for matching on the Producer.Name field. It simply calls the QueryByName method in the repository, which maps to the GetByNameAsDataReader in the IProducerProvider interface. This example is using the OleDbProducerProvider for the purposes of acting on an OleDb data source.

I've used another 2.0 feature (Generics) to return a IList<Producer> from the ProducerDataSource.SelectByName method. The GridView can now take this IList<Producer> and map it to grid columns.

Here's a quick code excerpt that illustrates the ProducerDataSource object and the ProducerGridView interaction.

<form id="form1" runat="server">
  <asp:TextBox ID="TextBox1" runat="server"></asp:TextBox>
  <asp:Button ID="Button1" runat="server" Text="Button" />
  <asp:GridView ID="ProducerGridView" runat="server" DataSourceID="ProducerDataSource" AutoGenerateColumns="False">
    <Columns>
      <asp:BoundField HeaderText="Id" DataField="Id" SortExpression="Id"></asp:BoundField>
      <asp:BoundField HeaderText="Name" DataField="Name" SortExpression="Name"></asp:BoundField>
      <asp:BoundField HeaderText="Address" DataField="Address" SortExpression="Address"></asp:BoundField>
      <asp:BoundField HeaderText="City" DataField="City" SortExpression="City"></asp:BoundField>
      <asp:BoundField HeaderText="State" DataField="State" SortExpression="State"></asp:BoundField>
      <asp:BoundField HeaderText="Country" DataField="Country" SortExpression="Country"></asp:BoundField>
      <asp:BoundField HeaderText="PostalCode" DataField="PostalCode" SortExpression="PostalCode"></asp:BoundField>
    </Columns>
  </asp:GridView>
  <asp:ObjectDataSource ID="ProducerDataSource" runat="server" TypeName="Lorengo.VirtualCellar.Business.ProducerDataSource"
    SelectMethod="SelectByName">
    <SelectParameters>
      <asp:ControlParameter Name="name" ControlID="TextBox1" PropertyName="Text" DefaultValue="Robert Mondavi"
        Type="String" />
    </SelectParameters>
  </asp:ObjectDataSource>
</form>

You'll notice that the ObjectDataSource has a SelectMethod attribute which maps to the SelectByName method on my ProducerDataSource. Also the <SelectParameters> node maps the TextBox1 controls Text property to the "name" parameter of the method. This allows the user to enter a value in the text box to search for.

The resulting pages looks like this

Unfortunately, the page does the query when it is loaded, so the DefaultValue attribute of "Robert Mondavi" is used and displayed. What I would like to do is prevent the initial lookup from occurring, presenting the user with just a search box and button. Does anybody know of hand how to accomplish this?

In the meantime, I'm going to revisit my Repository / DataSource classes and see if I can't come up with a better solution. Now that I know there is an easy way to utilize my business objects with the data controls, I'm excited.

Creating a MasterPage for the Virtual Cellar

mlorengo — Wed, 19 Jan 2005 23:30:00 GMT

I've begun creating the website that will house the Virtual Cellar data model and form the main user interface for the application. My next couple of entries will pertain to some initial tasks that need to be done so that I start hanging pages on the website. In this entry I'll discuss the setup of the site master template for the Virtual Cellar.

The General Layout

I plan on having a three column layout along with a header section, as illustrated in the picture below.

There are four major sections

Banner
The Header section will contain the application title bar as well as the main menu navigation for the application
ContextBar
The ContextBar will contain information that will vary depending on the context of the application. For example, if the user is not logged in, it will contain a username, password prompt. If the user is logged in, it could contain statistics about her account, an overview of her cellar statistics or latest headlines via rss feeds.
Main
The Main section will be where the user spends most of her time interacting with the application. The content of the section will be the data pulled from the Virtual Cellar catalog in the form of reports or where the user will enter new information into the Virtual Cellar.
SideBar
The sidebar section is where items of interest will be displayed depending on the type of information being displayed in the main section. These could be Google AdSense links, or links to items for sale on Amazon and eBay.

Creating The MasterPage

In Visual Studio, I create a new MasterPage template by Right-Clicking on the web project and select to add a New Item. I'm then presented with the following dialog box.

After adding the VirtualCellar.master file to the project, I'll go ahead and create the following <div> tags that define each of the sections.

<body>
<form id="MainForm" runat="server">
  <div id="Banner">Banner
  </div>
  <div id="ContextBar">ContextBar
  </div>
  <div id="Main">Main
  </div>
  <div id="SideBar">SideBar
  </div>
</form>
</body>

The next step is to add the <style> section to arrange the <div> sections in the order above

<head>
<style type="text/css">
	body {
		margin:10px 10px 0px 10px;
		padding:0px;
	}

	#Banner {
		background:#fff;
		height:70px;
		border-top:1px solid #000;
		border-right:1px solid #000;
		border-left:1px solid #000;
		voice-family: "\"}\"";
		voice-family: inherit;
		height:69px;
	}
	html>body #Banner {
		height:69px;
	}

	#Banner p {
		font-size:14px;
		padding:10px 10px 0px 10px;
		margin:0px;
	}

	#ContextBar {
		position: absolute;
		left:10px;
		top:80px;
		width:200px;
		background:#fff;
		border:1px solid #000;
	}

	#Main {
		background:#fff;
		margin-left: 199px;
		margin-right:199px;
		border:1px solid #000;
		voice-family: "\"}\"";
		voice-family: inherit;
		margin-left: 201px;
		margin-right:201px;
	}
	html>body #Main {
		margin-left: 201px;
		margin-right:201px;
	}

	#SideBar {
		position: absolute;
		right:10px;
		top:80px;
		width:200px;
		background:#fff;
		border:1px solid #000;
	}
</style>
</head>

You may notice the use of the infamous "voice family" hack, this is to get around the various browser bugs or interpretations of the css box model. This code also allows us to view our layout semantically. For the longest time I did layout using tables, little did I know that was a major crime according to the css folk. This article on retooling slashdot with css got me hooked on using css for layout. Another great site is the CSSZenGarden, but I digress. The above css code will give us the nice layout we set out to construct.

Creating a New Page Based on The Master

Now that I have the VirtualCellar.master page created, I can begin using it in new pages that I create. So I'll right click on my web project and select Add New Item. I'm present with the same dialog box as before, but this time I choose, Web Form. I also check the box that says Select Master Page. After selecting Add, a listbox containing all of the defined MasterPages are shown, and I select VirtualCellar.master. A new file Default.aspx is created that contains the following code.

<%@ Page Language="C#" 
    MasterPageFile="~/VirtualCellar.master" 
    CodeFile="Default.aspx.cs" 
    Inherits="Default_aspx" 
    Title="Untitled Page" 
%>

And the output looks like this!

Next time I'll look at adding a menu to the application using the new <asp:Menu /> and <asp:SiteMapDataSource /> controls

Setting Up The December CTP VSTF

mlorengo — Fri, 07 Jan 2005 16:36:00 GMT

1.1 Setting Up The December Team System Preview

I've recently downloaded the Microsoft Visual Studio Team System December CTP from MSDN, and have begun the process of setting up the full blown system. I’m using a Dell PowerEdge 2550 Dual P3-1Ghz system with 2GB of memory running Virtual Server 2005. The plan is to setup 2 virtual machines running Windows Server 2003 each having 640 Mb of memory.

Here’s how it will break down

Ø Hercules: The Dell rack server configured with Window Server 2003, running Virtual Server 2005, DNS, and Active Directory for the “LanThrash” domain.

Ø Hades: The first Virtual Server that will function as the “Data Tier” for the Visual Studio Team System.

Ø Hermes: The second Virtual Server that will function as the “Application Tier” for the Visual Studio Team System.

I'll be following the instructions found here http://go.microsoft.com/fwlink/?LinkId=40042&clcid=0x40

1.1.1 Configuring Hercules for VSTS

Not much needs to be done since Hercules is already setup as the Domain Controller for the LanThrash domain.

1.1.1.1 Create New Domain Users in Active Directory on Hercules

Start/Administrative Tools/Active Directory Users and Computers/Domain/Users/New/User

Add TFSSetup

Add TFSService

1.1.2 Configuring Hades for the Data Tier

1.1.2.1 Add New Domain Users to Local Administrator Group on Hades

Start/Administrative Tools/Computer Management/Local Users and Groups/Groups/Administrators/Add to Group...

Add LanThrash/TFSSetup

Add LanThrash/TFSService

1.1.2.2 Install IIS 6.0

Start/Adminstrative Tools/Manage Your Server/Add or Remove a Role

Add Application Server Role

Enable ASP.Net

1.1.2.3 Install Microsoft SQL Server Beta 2

Download Microsoft SQL Server Beta 2 from msdn subscriptions

Use the Virtual Server 2005 Administration website to mount the .iso file to the Hades CD/DVD Virtual Drive

Run the Install Program

Follow the instructions "How to: Install Microsoft SQL Server 2005 for the Team Foundation Data Tier" in the setup guide

Reboot the server

Finish installation of SQL following the "How to:"

1.1.2.4 Install VSTS Data Tier on Hades

Logon to Hades with the TFSSetup user

Run the Data Tier Setup program following the prompts. (Note: Apparently you can’t mount an .iso file larger than 2.2Gb under Virtual Server, so I used the Virtual CD-ROM Control Panel to mount the .iso file on Hercules, and then copy the /VSTS/DT folder to Hades directly. Apparently you cannot share a mounted .iso either, the other work around was to add the mounted drive on Hercules as a resource to the Hermes machine, but I couldn’t get that to work on Virtual Server)

1.1.3 Configure Hermes for the Application Tier

1.1.3.1 Add New Domain Users to Administrators group on Hermes

Same as Hades

1.1.3.2 Install IIS 6.0

Same as Hades

1.1.3.3 Install Windows SharePoint Services

Download and run Windows SharePoint Services with SP1

Choose the Typical Install

1.1.3.4 Logon to Hermes under the LanThrash/TFSSetup user

1.1.3.5 Run the Setup program according to the directions

It should be noted that I had a couple of problems while doing the install. First this one, while trying to find the Data Tier databases, but it went away, for whatever reason. The next one was an Error 3200, which someone else reported here, I simply hit Retry and it worked, although another solution or cause was having a website running on the same port 8080, that Team System was attempting to use.

1.2 Conclusion

All in all it went rather smoothly, of course this was over a period of a couple of days as I had to configure my new domain, and all my other miscellaneous client computer, etc. I’ll now perform the straight ahead installation of the Client Tier on my Dell Optiplex, but that should be easy J