HOME | SIGN UP | ACCOUNT | SUBSCRIBE | TROUBLESHOOTING | LOGIN
News · Meetings · Jobs · Downloads · Universal Thread articles · Web articles · Conference coverages · Consultants · Blogs · FAQ · User groups · Photos · Videos

DevTeach 2006

DevTeach offers a new breed of conferences. Sessions will include both presentation material and hands-on training. Traditional conferences had session leaders present material using a PowerPoint presentation, usually with some preprinted take away material. DevTeach sessions will add, whenever possible, hands-on training conjointly with the presentation. Attendees, equipped with their own laptop and wireless access card, will be able to try and test the concepts presented at the time of the presentation. This new concept is designed primarily for developers, but will not distract the others who may not be at that level. This conference will promote the latest Microsoft developer's tools.

Date:May 8-12, 2006
Location:Montreal, Quebec

This conference has 55 pictures. We display 12 randomly selected pictures here. You may click here to view the album thumbnails.

DevTeach 2006
DevTeach 2006
DevTeach 2006
DevTeach 2006
DevTeach 2006
DevTeach 2006
DevTeach 2006
DevTeach 2006
DevTeach 2006
DevTeach 2006
DevTeach 2006
DevTeach 2006

Monday, May 8

by Dan LeClair

Ah, Montreal in springtime! Having left a hot and muggy Houston at oh-dark-thirty this morning, it was refreshing to land in cool and breezy weather. After a short taxi ride (with windows down), I found myself at the Marriott Chateau Champlain, sight of this year's DevTeach conference.

This is the third time I've been a guest of Jean-René Roy's hospitality, and from the session schedule it looks like this year will be another great conference. JR puts a lot of effort into bringing great talent to speak here, and a continued rise in attendance (the attendee count has doubled in four years) demonstrates that this show is a winner in the development community. Plus, Montreal is a great city to visit, except maybe in winter...

Monday being devoted to pre-conference activities, I had some time to walk around the town before hooking up with Patrice Paré, my counterpart for reporting at this show. Montreal is full of great restaurants, and since Patrice is from these parts he recommended a small sushi place (Mikado, on St. Denis) that turned out to be superb. Be sure to try the Trust-Me-You'll-Like-It roll. We hope we've laid out a strategy that will cover as many topics as possible for you in the next few days, so stay tuned!

Tuesday, May 9

Keynote: Standardizing access to data - The LINQ project
Y. Alan Griver
by Patrice Paré

The preliminary session common to every attendees who were not late today was given by Alan Griver. Alan is the Group Manager for the Visual Studio Data group. As Group Manager, Alan’s teams produce the tools used inside of Visual Studio .NET, Office and SQL Server that surface data capabilities, as well as Visual FoxPro.

As the first session of the DevTeach 2006, the Keynote made a fabulous impression on people in the ballroom of the Marriott Chateau Champlain today. Alan presented the LINQ project. LINQ stands for Language Integrated Query. It is truly a standardization of query languages that wants to answer the problem of relating data, objects and XML. This unified query language would, on the long term, replace the XQuery and the SQL Queries.

The main innovations are the implicitly typed locals, the extension methods, inline functions and object initialization. The query expressions are like the familiar SQL query statement. Also, classes describe databases. The tables are like collections. Other aspects of LINQ include strongly typed connection and results and integrated query syntax. One advantage I see for the LINQ is the easiness to read code. As for the XML, what I liked best was the possibility to dynamically create XML document with an object model that will fill the XML with data. For more details, please download the slides from the DevTeach website.

Finally, LINQ would have sounded incomplete if Intellisense was not present. The programmer gets full support from Intellisense anywhere in his code, whether he is writing VB, XML or LINQ code lines.

At the end of the presentation, I got the feeling that LINQ would open a new era in querying data. In my mind, I will watch for the next tech reviews.

Introduction to LINQ (Language Integrated Query)
Markus Egger
by Dan LeClair

LINQ seemed to be the la pensée du jour today, as this session followed hot on the heels of YAG's LINQ keynote. Although a feature of the next version of VS.Net, LINQ is already causing quite a buzz in the .Net development community, as it will bring the ability to query data directly into a .Net application.\

Markus usually packs his sessions full of information, and this one was no exception. He covered, among other things:

  • XLINQ
  • Extension methods ("the coolest feature I've seen that you aren't supposed to use")
  • Anonymous types - useful for creating classes for special, single-purpose use without formal definition
  • Lambda expressions - simple way of defining an entire method in a single line
as well as demo-ing some cool code putting these concepts in action.

Some caveats regarding LINQ: right now the CTP versions only supports SELECT statements, but it is still very much a work in progress. When it ships will include INSERT, UPDATE, and DELETE actions as well. LINQ, paired with the CLR integration in SQL 2005, effectively allow developers to create fully managed stored procedures. As an old T-SQL dog, this gives me a bit of concern, as from what I've seen, it is too easy to get carried away with using CLR.

Just my opinion...

Introduction to ADO.NET
Rod Paddock
by Patrice Paré

This presentation was meant for programmers new to the .NET environment or for those who have not yet worked with databases.

The presentation started with the explanation of the architecture of the ADO.NET. Basically, ADO.NET is a set of classes and commands to access data in .NET applications. Rod divided the architecture into two sections: the noun section and the verb section. The first refers to the in memory database or, the dataset. The other section refers to commands to open a channel to the database and query it.

The speaker then did a couple of demos showing how to open a connection to a database, how to retrieve the data. What was the most interesting is in the use of the store procedure. Rod used a lot of stored procedure and he showed how to set parameters. This way, the code was clean from SQL queries.

One other thing he used that I believed is good for windows application is the use of dataset and the default view. His query selected about 100 000 rows from his sample database and it took a couple of seconds to show up on the datagrid. But with the dataset as a global variable, filtering and sorting took only a tick to render the results.

This presentation was definitely targeting an audience not used to ADO.NET. And the speaker had a marvelous pedagogic approach. The use of the Power Point slides and the demo did point to most important thing to remember. While the slides pointed out the theory, the demos showed the mechanism of ADO.NET.

This session was definitely the best I attended today. I learned a lot, and moreover, I retain a lot of information from it.

C# Future and Present
Markus Egger
by Dan LeClair

There's a lot of changes coming in the next version of VS.Net, and in this session, Markus covered several topics regarding improvements in C#. Some of it was new, but having just previously sat in his LINQ session, some of what he covered was a repeat. Regardless, the session was well attended, in fact it was standing-room only.

Markus split his session into two parts: what's new and improved in C# 2.0, and what's coming in 3.0 (code-names Orcas). One of the more interesting 2.0 items was the concept of "partial classes", or classes compiled out of multiple source files. Basically a text-merge feature that occurs before compilation, this allows for some interesting documentation and/or code reuse possibilities for class libraries. Markus suggested that if a single class is so complex or so large it needs to be split, its best re-examine the class design.

Markus had a very whizzy demo form that he used to illustrate all of his samples. Once I get my hands on a copy, it should help me understand some of the new features a lot more.

Other topics he covered included:

  • Generics - one of his favorite new features, allowing the developer to define the type used in each particular instance. compared how to get around problem in C# 1.1 (way too much work), then showed how to do it via generics in 2.0
  • Iterators - a specialized feature, basically a simplification of iEnumerator. An iterator is any object you can use a FOR EACH action on.
  • Nullable objects. Some object types can't be null in 1.x versions of C#. In version 2.0, objects can be specified as nullable. Implementation is done via generics.
  • Static classes - These classes only have static members.
  • Property Accessor Accessibility - helps clean up design

After covering 2.0 enhancements, Markus went on to cover many LINQ-related changes coming up in 3.0. Despite the repetition, I still found this session very useful.

Introduction to C# 2.0
John Bristowe
by Patrice Paré

The following presentation was given by John Bristowe. John is a Developer Evangelist with Microsoft Canada based in Calgary, Alberta.

I attended this session for two reasons. The first was the speaker. I met John Bristowe last year at the Microsoft Visual Studio 2005 launch in Montreal. He made the introduction to .NET 2.0 and I like his style. I was not disappointed today. His factual and W.Y.S.I.W.Y.G. approach make his session an interesting one. The other reason is I have always been fond of C++ and Java language. Since I have been working with VB for some times, it seemed to me like dating again a long lost girlfriend.

This session was presented just before lunch on the first day. Only a few attended to the session, mainly because there where so many conference at the same time and also because of the subject. The introduction was meant for people already using C# in 1.X .NET framework. This fact did not bother me much. The speaker started with some basic examples. It gave me the chance to remember that the C# declarations are done in a specific way. I had a little more hard time with the semi-column (“;”) at the end of line statements though and my notes reflect it. When I read them again, I notice sometimes I forgot the semi-column.

John introduces us to the C# in the 2.0 .NET framework by covering the following topics: generics, anonymous methods, nullable types, iterators, partial class, external aliases, inline warning control, fixed signed Buffer and edit and continue. I found that like others attendees, generics were like templates, although there are some distinction between those concepts. For more details, I enjoin you to look at the Herb Sutter blog on MSDN.

The most interesting thing in these subjects was the “Edit and continue” feature that allows the programmer to modify his code in runtime. This is the nicest thing from my point of view. While you save your code, the framework recompiles your edited code and manages to maintain variable values. This feature improves your productivity in .NET 2.0 because you do not have to stop the debugger, modify your code and fire another debugging session.

Object Modeling with the VS2005 Class Designer
Kevin McNeish
by Dan LeClair

This session was a last-minute substitution for a session by a speaker unable to make the conference, but I'm glad I sat in on it. Easily the coolest feature I've seen in VS 2005, the Class Designer is both a visual designer and object modeling tool for VS.Net classes. Using basically standard UML notation (only differing in a couple of minor points), the Class Designer is a fast and lightweight tool for doing class construction, effectively allowing the designer to do everything through the UI that can normally be done in code. Kevin's biggest beef seemed to be the lack of full UML notation, as it doesn't do sequence diagrams, but I'm willing to bet somebody in Redmond is working on that as you read this...

Kevin showed just about every aspect of the designer you could think of. Creating diagrams, adding and removing classes, showing subclasses (easily done with a right-click menu), adding relationships, adding and modifying all of the properities, events, and methods of a class...you name it. Depending on the level of detail you want to show, you can include field properties, overloaded methods, you can add comment objects - this thing looks like a really great tool, and I'm looking forward to giving it a whirl.

Visual Studio 2005 Datasources and Databinding
Beth Massi
by Patrice Paré

What a great presentation! A lot of senior stuff for advanced .NET users. But for the juniors, they also have something interesting to learn with the wizards.

Has anyone thought doing databinding with no lines of code? With the wizard help, which was instantly denounced by the speaker (we will come over that later) she built a quick database interface on a windows form. After you have set your datasource, you only have to drag and drop columns on the form. You want to see everything from your database? No problem! You drop the table from the server explorer on the form and a datagrid is instantly added. What made me smile was the navigational toolbar that was also added on the form which is a cool feature.

The speaker does not recommend this technique of designing a form though, as the data and GUI are too tightly integrated. The best practice is to build data layers.

The building of a data layer required more work around but at the end, Beth Massi has set up the same database interface on the form with only two lines of code.

The core of the presentation was the BindingSource which allows the programmer to interact with the data collection. What struck me in the demos was the capability to dynamically add labels and textboxes on the form and there text property would be bound to the datasource. So when the user selects a row in the datagrid, the content of labels and textboxes would change accordingly.

Finally, the presentation was a serious one and I got the feeling I was not the only one that had some difficulties understanding the complex concepts the speaker presented. I felt like overwhelmed by the concepts and this feeling was echoed by some attendees. Otherwise, the more seasoned programmers like the dynamic approach and the rhythm of this conference.

SQL Express and Integration into .NET Applications
Cathi Gero
by Dan LeClair

Despite the number of times Cathi and I have been at the same conference, I've never heard her speak. But when I saw the title of this session, I had to sit in. MSDE never really impressed me, with its throttled performance and 2Gb data limitations, but what I've heard about SQL Server Express was very intriguing.

Cathi's session on SE didn't disappoint, either. Starting with a comparison of MSDE and SE (and there is a lot more power in the latter), she went on to discuss some of the particulars about how it works, how you communicate with it, and new features in the latest release of it, SQL Express Advanced.

A connection string for SQL Express is a little different than what you would use when talking to a full version of SQL, as you have to specify the AttachDBFileName property. What this does is tell SE what and where the database file is that will be used. SE attaches the database file on startup, which results in a little bit of a performance hit. The upside, though, is since the database files are not permanently attached, copying the database to a new installation is a breeze.

Unlike MSDE, there is a graphical interface available for managing SE databases (SQL Server Management Studio Express). This, I think, is enough of an improvement alone to make me want to experiment with it.

There are limitations to SE. First, the databases it uses must reside on the same machine as the SE instance itself. You cannot access DB's that reside on other machines, or mapped drives. SE is also limited in the amount of memory it will use (1Gb) as well as having a maximum DB limit (4Gb). SQL Agent has also been disabled in SE (whereas MSDE supported it). Regardless, I think the improvements in SE as a whole more than make up for these limitations.

Wednesday, May 10

by Dan LeClair, Patrice Paré

La validation des données en profondeur (Thoroughly Data Validation)
Nicole Calinoiu
by Patrice Paré

This is the second day of the DevTeach Conference. Ho god! What a day we had yesterday. And today’s schedule sounds as busy as yesterday. I started my second day by attending Nicole’s session. Nicole Calinoiu has over ten years of experience in the design and developer of data-driven applications, including over five years working on publicly available web applications. She is a Microsoft developer security MVP. Her session was on data validation.

The information technologies rely on two things basically: data and processing logic. The data process is treating input data and outputting them in an understandable manner for users. But the treatment becomes irrelevant when the data are invalid. Moreover, invalid data may cause software to crash and ends the user experience.

To start with, what is Data? Data are every piece of information unknown by the program at compilation time. It may be operating system data, configuration files, data files or input from the GUI. No data validation may increase time debugging and development cost. Also, it may decrease performance. Nicole pointed out programmers would save time and efforts with validation. Data validation will reduce significantly the number of bugs. It will also improve security and data quality.

There are two ways to do data validation: the hard way and the soft way. The hard validation means you apply strict rules before accepting data. The soft one requires the user confirmation for unusual data. Nicole gave an excellent example to illustrate this with the date of birth. Hard validation would allow anything where the user’s age is between 0 and 130 years old. Soft validation would prompt the user to confirm that he really is his age if he provided something between 90 and 130 years old.

The speaker put some emphasis on four data types that should required more attention. The string type, numbers, enums, and XML data type are often used. Strings should be tested against their length value. Numbers should be parsed using their TryParse method. Enums should be tested with the IsDefined method and the data should be processed through a switch/select case routine. Finally, XML should be tested against a schema. That was the most important part in that part of the session. The schema should be an embedded resource. This way, the schema is known at the compilation time thus it would not be considered as a configuration file.

When validation should be done? Anytime data crosses boundary level. Those boundary levels are application, machine, assembly, class, and member. For more details and to get the Nicole’s bibliography, you are invited to download her files on devteach.com.

SSIS Tips, Tricks and Gotchas
Kirk Haselden
by Dan LeClair

This session was totally worth going to, at least for me. Having worked with SSIS for the last several months, I realized after sitting through Kirk’s presentation that there is quite a bit more to the product than what I have had a chance to use.

The session was a collection of tips, tricks, and “gotchas” regarding SSIS, as well as some interesting demos. Kirk started with a discussion of configurations in SSIS, which provide a way to set up the environment (such as pathing) for a package before it actually runs. There is a configuration wizard that can assist in setting these up.

Configurations and expressions are the way you build packages that are easily deployable, and Kirk stressed that you should understand them before you start building packages!

One tip I can immediately use: SSIS packages with many connections can easily take minutes to open, as the environment tests each one. The SSIS environment has two ways to deal with this, one being a “work offline” setting at the global level. The other is setting the timeout on individual connection objects from the default of zero.

Among the many demos shown was one that emailed the SSIS execution log on package completion. This could be really useful in situations where packages fail.

Here’s a tip for anyone that may be responsible for lots of SSIS package development: use template packages. These help ensure consistency across packages, as well as increasing object reuse.

There was just so much in this session, I could easily spend a lot of bytes telling you all about it. Kirk has a new book coming out in July regarding SSIS, and you can be sure it is on my list of things to order soon.

Programmation Ajax avec ASP.NET (Ajax Programming with ASP.NET)
Dominic Sevigny
by Patrice Paré

L’avenir du web, quel est-il? Sortez votre boule de cristal, votre jeu de tarot ou consulter les étoiles parce que l’avenir du web, bien qu’assuré, peut prendre des tangentes imprévues. Certaines choses cependant doivent évoluer dans le mode du web et avoir un débuguer pour le JavaScript ne serait pas un luxe. Ni celui d’éliminer les post back, à la fois inutiles et nécessaires, vers le serveur. Il semblerait que ces inconvénients soient sur le point de disparaître. Effectivement, avec Ajax, ces maux de têtes pourraient se transformer en un doux orgasme pour les développeurs web. De plus, selon certaines rumeurs, Microsoft aurait dans sa manche prochainement un engin pour débuguer JavaScript qui serait prêt lorsque Ajax aura fait ses preuves.

What are the perspectives for the web? Take out your crystal ball, your tarot cards or read the constellation of Venus because, while the web healthiness seems pretty good, what curve will the web take is still unpredictable. Some things must change in the web environment and having a JavaScript debugger wouldn’t be a luxury. Neither would it be a luxury to eliminate post backs that are useless yet necessary at the same time. Post back should only happen when something visually noticeable on the form happens. Changes in data are not among them. This is where Ajax intervenes. Ajax stands for Asynchronous Javascript and XML. It is a merge of three technologies: HTML, XML, and Javascript.

Dominic explained how classic HTTP request works and he compared it against the Ajax engine. The Ajax engine interacts between the client’s browser and the web server. On the way back, the signal is sent in XML format from the web server to the Ajax engine. The Ajax engine converts it into HTML/CSS data so the browser can interpret it.

This technology is available in ASP.NET 2.0. ASP.NET 2.0 implements the basics so you don’t have to care about different browser compatibility. It is also easy to use and is ready for Atlas. You probably wondering if some guys have started working with Ajax. In our session, an attendee claimed he started working with it and deployed a solution involving Ajax. But other people around the world think Ajax worth investing time. Frameworks are available; webcontrols using Ajax are also available.

So a bunch of tools and environment is now ready for you. I recommend you to download the Zip and have a look on how Dominic manipulates Ajax.

SQL Server 2005 Security Enhancements
Randy Dyess
by Dan LeClair

Randy made it clear from the get-go that he was only going to touch the highlights of security changes in SQL 2005. His agenda was pretty aggressive, as he wanted to cover security design principals, the new security model in 2005, user-schema separation, managing permissions, catalog security, and more.

The session was very informal, and a little scattered; Randy jumped around from subject to subject, and ended up skipping some portions because he ran out of time. However, there was a lot of good information presented, not only by the speaker, but in response to attendee questions. Randy had a good report with the audience, and everyone there seemed to appreciate what he was presenting.

As an example, he spent some minutes discussing user-schema separation. In previous versions of SQL Server, most objects ended up being assigned to the DBO user, to prevent broken ownership chains when object owners left an organization. SQL 2005 introduced schemas to database construction, and changed the security model to assign ownership to schemas (as well as users). In fact, you don’t have to have a user to create a schema. If the ownership is granted to the schema instead of a user, you won’t have maintenance issues and/or broken ownership chains when users are removed.

As I mentioned, there was a lot of information presented in this session, and even though some of it was rushed over, it gave me enough to answer a couple of my own questions, as well as want to continue research later on.

Introduction to Developing for the Pocket PC
Dan Jurden
by Patrice Paré

Pocket PC, PDA, BlackBerry, PalmPilot… those new toys on the market are marvelous things, aren’t they? And you can develop your own application for your latest acquisition. With the .NET 2.0, it is a piece of cake, I must admit, after I saw Dan’s Session. First, Dan showed us some unknown areas of Visual Studio .NET 2.0. I mean, I did not know I could start programming for Pocket PC or smart phone right away. Every tool, mostly, are already there. The two missing things you absolutely want to download are Windows Mobile 5.0 SDK or SmartPhone SDK, which was made available only after the launch of .NET 2.0 and the ActiveSync 4.1 which is a tool to get your Pocket PC synchronized with your server. Also, you may want to have the closest Pocket PC emulator as you can by looking on your Pocket PC manufacturer web site or on third party web site.

The challenges for Pocket PC developers are well identified. You have to keep in mind two things that have come obsolete for 15 years, memory and battery lifetime. Gee! Memory! The ratio Mb/$ has come so ridiculously low that nobody cares about it now. You need more memory than highly efficient programs. A quick look on internet and you could find a flash memory of 2 GB for $35. Another thing developers should keep in mind is the possibility for the user to switch for the portrait mode to the landscape mode. Avoiding this pitfall from the user’s point of view should win his attention.

The wide range of Pocket PC devices, their unique interface, and their connectivity type provide a rich solution environment for developers.

Dan has one tip for everyone. Closing an application on a Pocket PC does not really close it. It merely sends the program running in the background. You can try this out. Create a form with a combo box containing few elements. Run the program in debug mode. Select an item, say the third one. Click on the closing icon. Have you remark that your VS is still running in debug mode? Ok. Now, go the running programs panel and open it. You will see your program running. Open it now. What do you see? Right! The third item of your combo box is still selected.

Thanks for that information, Dan!

Introduction to Analysis Services 2005
Tamer Farag
by Dan LeClair

This session wasn't exactly what I was expecting. Although I've done some experimentation with Analysis Services, I was hoping for a good overview of the business intelligence features of SQL 2005. Unfortunately, Tamer’s session was a little more basic than that.

The first half of the session was more of an introduction to BI and the history of Analysis Services. Relational DB’s are really two-dimensional, and can’t handle the analysis requirements of multidimensional reporting. Tamer spent some time discussing the Microsoft vision for BI, the tools required for effective BI, and what MS was trying to accomplish with the new features of Analysis Services.

It wasn’t until much later in the session that he began to describe and demonstrate the Unified Dimensional Model (UDM), the core of Analysis Services. He went into some of the mission critical features of UDM (availability, serviceability, manageability), as well as some of the advanced analytics and other features of Analysis Services. Unfortunately, I think he lost some of his audience by then. Of course, the free beer offered in-between sessions today probably didn’t help…

While I think there was some good information presented, both in the session and in the materials, it would have been much more interesting – to me, anyway – if he had skipped what sounded like a Microsoft sales pitch and gone straight into the why’s and wherefore’s of Analysis Services.

Vendor Spotlight: Dynacom
by Patrice Paré

This afternoon I went to see two commercial booth. The first I went to was Dynacom. Dynacom is business offering a complete accounting solution that gives to the users all flexibility needed to adapt the Dynacom software to their needs. When a business buys a license it could customize its software using the Dynacom IDE. The salesmen gave a demo and I could see that visually, Dynacom interface is pretty and reproduces the graphic environment of Windows XP. Also, something else interesting was that the software was developed using .Net. There was 4 things that struck me. First, you can adapt the software to meet your object data model. Next, the bug tracking module got my full attention. Third, it was the security feature throughout the software that uses Active Directory to allow users to get and use different user interfaces. Finally, developers can join the Dynacom community to develop plug-ins on a personal or professional behalf. License cost ranges from free to 50000$.

Thursday, May 11

by Dan LeClair, Patrice Paré

Trucs et Astuces : Améliorer votre productivité (Tips and Tricks : Improve your productivity)
Luc Gauthier
by Patrice Paré

Luc was straightforward right at the beginning of the session. People who don’t care about or don’t like VB (he was talking to the C# programmers) were invited to leave the room anytime during the session; he would not take offence out of it. After the doors closed behind the leavers, Luc tips and tricks were good. I have been working with VS 2005 for some times and there were things I wonder about and other that made my twitch.

When I write my code comments, I would like to add a block of comments automatically with specific comments to put in. This is possible. Type three commas at the beginning of your method and hit [ENTER]. The comments magically appear. They do not fit your enterprise comment policy? You can fix that. You have to edit the VBXMLDOC.xml document. After you are done, you can save it into two directories only. The one containing the original one is a good place, but the best place to save your XML file is in /application data/Microsoft/Visual Studio/8.0. The technology behind it is snippet. Another snippet I like was to qickly add a property in your class. You type the keyword property and then hit [TAB]. Visual Studio magic does the rest. Like the other feature, you can edit it to fit your variable/property naming convention and policy.

There is one more thing I would like to talk about. Luc presented an add-in that was totally cool! It is Refactor from DevExpress. There is a trial version available to download and a full licence version cost 100$. In the trial version, you have access to a dozen of functionality but the full licence edition has more than 900 features. Luc presented two features I believe will improve everyone’s productivity. The first one has the possibility to invert parameters in a method signature. Using the arrows on the keyboard, you can move forward or backward the parameter. When you are done, the Refactor brings you everywhere in your code to ask you if you want to apply the changes when you call that method. Of course, you can apply the modification to all if desired. Another feature I like and would use is the capability to extract a block of code from a method and create a new method with it. The Refactor creates a method with the right number of parameters. I think this is a very cool feature for quick code refactoring.

The visualizer in debug mode is not very explicit when it comes to objects containing other objects. You can create you own visualizer that will show object members.

For more details, you can watch a similar demo presented at the 2005 PDC on this web site: http://microsoft.sitestream.com As for the Refactor tool, you can have more details at: http://www.devexpress.com

The Query Governor – SQL CLR In Action!
Richard Campbell
by Dan LeClair

Going to this session was such a snap decision for me that I didn't bring my laptop. And, since my handwriting has gotten so poor after so many years of using a keyboard, my notes on this session are, well, a mess. But this was easily one of the coolest demos I've seen in a long time.

The premise behind Richard's session was this: how to create a tool that would either allow or deny query execution, based on the conditions of the server, the expected execution plan, and the size of the expected result set. Although there is a query governor native to SQL 2005, it doesn't quite fit the bill. SET QUERY_GOVERNOR_COST_LIMIT x sets the upper limit of a query's execution time in seconds, but is simply a barrier to execution, nothing more. In addition, it only really works if you actually try to execute the query.

Richard's solution was much more elegant, in that it takes advantage of a new feature of SQL 2005 to help make a more informed decision. SQL will now return query plans in XML format, allowing you to estimate cost of execution. This, coupled with code in the CLR, allows both estimated cost AND counters available through the Windows PERFMON utility to be measured. Results can be returned back to the query or stored procedure that is about to be run, which can then either execute or return an error message if execution would be too costly.

So much cool information was presented in this session that I couldn't write fast enough to keep up. I'm going to definitely take apart the sample provided when I get home!

Adopting Visual Studio Team System
Etienne Tremblay
by Patrice Paré

I attended this presentation mainly because my employer is thinking about moving in Team System. I heard a lot of it but I do not really know the beast. Talking about Team System as a beast is not a euphemism. The product presentation leaded me to think this is a very heavy framework that requires a machine with a lot of resources and, if your business has many developers, architects and project managers, it could require two machines to run it properly. Also, while the speaker ending words were still echoing in my head, I thought that the learning curve was abrupt.

The speaker spent a lot of time explaining what Team System is. Team System (TS) targets software enterprises, whether the enterprise is a consultant or a multinational. If your organization has a well structured hierarchy, then TS will suit you better because it embraces all roles (developer, architect, tester and DBA).

You must absolutely evaluate your development process before tending your credit card to the salesclerk. Etienne repeated several times throughout his presentation that the TS would not answer everyone needs. I got the feeling this investment will be time-consuming if the proof of concept is not done. Etienne said it could eventually lead developers to reject TS because it is not a viable solution.

Champagne to everyone, TS is coming! You make your mind and TS Foundation (TSF) is the best choice for you. Congratulation! Let’s open the box and see what’s in it. Aside TSF and SQL server 2005, you get predictability with report and metric features. You also get something to build the future on. But what do you need? You will need a solid server running Windows 2003 SP1, SharePoint 2.0 SP2, Visual Studio 2005 and Office 2003 SP2.

If my employer asks me my opinion on buying Team System, I will have to say that I have not been convinced Team System is the right choice for us at this point. It just seems too big for our 15 employee business.

Writing .NET Code in SQL Server 2005
Peter DeBetta
by Dan LeClair

One of the most intriguing changes for SQL 2005 is the ability to use .Net code instead of T-SQL. Peter presented a lot of information in this session on how to do it, as well as important information on when to do it. CLR integration is supported for all coded objects: stored procedures, user-defined functions, and triggers. In addition, there are two types of new objects where CLR is the only answer: user-defined aggregates and user-defined types (not to be confused with user-defined data types).

Peter spent a portion of the session going over some of the basics regarding CLR assemblies in SQL. This included a discussion regarding security levels (safe, external-access, and unsafe), where assemblies reside in SQL, how they are injected into a database, and do forth. Two important points stand out. One is that while a lot of the CLR namespaces and functionality is supported, some portions (notably any visible functionality) are disabled. The second is that it is possible to create unsafe assemblies via CLR.

The meat of the session, though, was directed at examining sample assemblies. Peter walked through examples of all five types of CLR objects that can be created, including his favorite, the user-defined type. Of all the examples presented, this one illustrated (to me, at least) the power of using CLR, as he had created a UDT for a date – not datetime – datatype, capable of supporting a date range from 01 Jan 0000 through 31 Dec 9999.

Following this sample was a rather frank discussion of the pros and cons of using CLR instead of T-SQL. Basically, it boils down to this: T-SQL should still be the language of choice for most data-handling operations, as calls to CLR assemblies do carry a performance penalty. However, some things are just better done with the CLR.

SQL Querying Tips and Techniques
Richard Campbell
by Patrice Paré

Tips and Techniques for better and more efficient SQL queries are things I definitely need. Querying is something you do everyday. So the better you do it, bigger gets the pay check as Richard humorously pointed out. The crowded room felt it needed some advices on that particular matter and Richard was to please them with a series of SQL queries.

Richard started his presentation with subqueries. His examples showed numerous situation where subquerying is involved. He made a good example with ranking. He use two methods to illustrate ranking and he pointed out a limitation of a method. The result returned values ranking from 1 to 6 with a tied at the second place. Instead of reading 1 – 2 – 2 – 4…, the result was 1 –3 – 3 – 4... His second method corrected this glitch. Richard mentioned in his session limitations about using cursors and he suggested to use CTE as the most effective way to do recursivity. His example used an option OPTION (MAXRERCUSION = 2) (for instance) where the process would dig a maximum of two time in recursivity. If the exiting condition is not met yet, but the process went through the recursive method two times, it will exit anyway. His presentation also included crosstab and pivot, whether they are dynamic (a situation where you don’t know the number of columns) or static (you know the number of columns). To conclude this fast time going session, Richard entertained his audience with error handling. You can insert a Try catch SQL block:

BEGIN TRY 
	Put some SQL here..
END TRY
BEGIN CATCH
	Put some SQL here..
END CATCH

Although Richard still suffered for jet lagged, he was very entertaining and very funny. The room, one of the largest at the DevTeach, was crowded and paid attention from the beginning to the end. It was a real pleasure to sit there and listen to him. Jean-René must definitely re-invite him for the next DevTeach.

Working With Master Pages and Themes in ASP.NET 2.0
Jim Duffy
by Dan LeClair

Jim Duffy said at the end of this session that he always hopes to accomplis three things: to educate, motivate, and entertain in any of his presentations. That he certainly did. I've seen Jim present at a number of conferences, and he never fails to both fill the room and make good on his three-way promise.

Althought the premise of this session - how to use master pages and themes in ASP.Net 2.0 - might seem a little tough to fill 75 minutes with, Jim was pretty successful at doing so. I'm a relative novice with it comes to ASP.Net, having just enough knowledge to be really dangerous, so what he talked about was really useful. Beginning with an overview of the new productivity features related to ASP.Net 2.0, he went into master pages themselves. These basically are pages that have development "holes", or holder places for content areas or pages. Master pages are a way of easily ensuring a consistant look-and-feel to a website, as well as breaking down development into multiple areas.

One of the important concepts he went into depth on was the firing order of ASP pages when they load. This is important to know, because you can reference master pages within master pages, override certain portions of master pages and/or content pages, and so forth, but unless you do it correctly, execution of an object later on in the firing order will cause unexpected behavior.

After much discusson and several demos of how to use master pages, Jim went on to discuss the useage and creation of themes via .SKIN files in a project. Themes can be applied at the application, page, or control level, and he walked through examples of how to create the files, add them to the THEME portion of a project, and how to apply them.

Although I don't do a lot of web development, I came away knowing a lot more about a potentially useful technology than I went in with. The other attendees also seemed to get a lot out of it as well, so score another one for Duffy.

Daniel LeClair
Daniel LeClair is a Senior Developer with Sogeti USA in Houston, TX., and owner of GSF Development Consortium, also in Houston. He has many years of database development experience, dating back to dBase II, and has been constructing systems with Visual FoxPro from its introduction in 1995. In 1997, he began working with MS SQL Server 6.5, using VFP as a front end; currently he specializes in SQL Server database development and data migration. Daniel was the project leader for Eagle USA’s TALON system, winner of the 1999 VFP Excellence Award in the “Client-Server” category. He has written several articles on VFP development for FoxPro Advisor, and has spoken internationally on VFP and SQL Server development. He has a BBA in MIS from the University of Houston, and currently holds MCSD Certification in VFP.

Patrice Paré, Netappsid
Patrice Paré worked as Team Leader for Galaxia, in collaboration with Enovia – Dassault Systèmes, for the development of a ERP / EPM with IBM technologies. Now, Patrice is a member of NetAppsID and he is working as web based transactional application developer. He has been using the .NET 2.0 technologies since September 2005. His activities cover the maintaining and development of object models, data bases, web based applications, and the training of new employees. He also shares his knowledge with community organizations. Patrice has graduated from the Institut Supérieur d’Informatique. Also, he got a degree in Political Science from University of Montreal and a degree in Chinese from the Shenyang Institute of Technology, China.

Copyright © 1993 - 2014 Level Extreme Inc., All Rights Reserved · Telephone: 506-783-9007 Email: info@universalthread.com · Privacy & Security · Copyright · Terms & Conditions