Dino Esposito's WebLog

It's long way to ... Dublin

despos — Tue, 06 May 2008 13:22:00 GMT

Coming to Dublin in just a few weeks (May 27) for a full day of public, 9-to-5, LINQ, LINQ-to-SQL from ASP.NET and Silverlight. Read all details here.

A pattern-oriented look at LINQ and LINQ-to-SQL from within Web and Silverlight applications

We'll start the day with a 10,000 look at LINQ. Hey, I can hear some of you complaining that it is just the "usual" boring stuff about LINQ being so cool etc. Things that you've probably heard a zillion times. Well, sort of. As Don Box told me once, it's all about perspective. So in the first module you'll hardly find me discussing how to group/join data. I'll focus on the architectural model and drive you naturally to understand the various flavors of LINQ. A full understanding of the IQueryable interface will open up a new perspective on LINQ-to-SQL and LINQ-to-Entities.

In the end, LINQ-to-SQL is two-folded. On one end, it is a programming tool for a DAL, but if you know it well it can morph into a simplified, but not less effective, O/RM tool.

After due lunch, discussing data access strategies from ASP.NET and Silverlight with patterns. Which patterns? MVP and MVVM. And services. And data feeds.

It'll be fun. Don't hesitate to ask more information :)

The "Driving Force" pattern--part 1 of N

despos — Mon, 21 Apr 2008 19:30:00 GMT

Recession is perhaps affecting the economy, but it couldn't be farther from the dazzling world of (Microsoft) software. CTPs are coming out every day and some of them are amazingly morphing into Betas. Which is anyway good. But when a technology turns of age becoming a Beta, you--the developer|architect|software engineer--are no longer authorized to ignore it with the abused excuse that it is only a CTP.

You can easily be flooded with technologies, frameworks, products, and also patterns and paradigms. This is especially true for the Web. Isn't AJAX representative of a "paradigm shift"? Paradigm shifts, though, are an extremely delicate event that occurred in the history of mankind quite a few times already. As emphatic as it may sound, the Web of today is just a special case. Search for "paradigm shift" on Wikipedia (http://en.wikipedia.org/wiki/Paradigm_shift). At some point, you'll read:

Paradigm shifts tend to be most dramatic in sciences that appear to be stable and mature, as in physics at the end of the 19th century. At that time, physics seemed to be a discipline filling in the last few details of a largely worked-out system. In 1900, Lord Kelvin famously stated, "There is nothing new to be discovered in physics now. All that remains is more and more precise measurement." Five years later, Albert Einstein published his paper on special relativity, which challenged the very simple set of rules laid down by Newtonian mechanics, which had been used to describe force and motion for over three hundred years. In this case, the new paradigm reduces the old to a special case (Newtonian mechanics is an excellent approximation for speeds that are slow compared to the speed of light).

The Web is a science that before AJAX appeared stable and mature. AJAX is comparable to Einstein's relativity. It will take years to reach again some technological stability for the Web. Read, some set of frameworks and products that are widely accepted and not put under discussion every week by new CTPs and approaches.

Five years? Well, five years may be nothing compared to the mankind lifespan, but it's a lot of time in software, and for the Web in particular. Five years in software are more or less the equivalent of million years of earth life. And even more if you limit to consider the Web, which is only 20 years old.

AJAX has been a real paradigm shift whose effects--three years after its introduction--are only now starting to become stabler. I have a lot of hopes and expectation from MS and the ASP.NET team in sight of ASP.NET 4.0. I expect it to be built with AJAX in mind from the grounds up, incorporating common Web and AJAX patterns in a brand new set of controls and components.

Oh yes, but I started this post with the concept of "driving force" in mind. How can you find orientation in a world full of CTPs? Like many of you, I can't just keep up with everything that is coming out. So I find it useful to identify the "driving force" of each new thing from Microsoft and see whether it cares me to a decent extent. Today here at DevConnections, Orlando, I finally identified the driving force of a buzzword that hit me recently.

ADO.NET Data Services (aka, Astoria)

Driving force: the need of building richly interactive Web systems.
What's that in abstract: New set of tools for building a middle-tier or, better yet, the service layer on top of a middle-tier in any sort of application, including enterprise class applications.
What's that in concrete: provides you with URLs to invoke from hyperlinks to bring data down to the client. Better for scenarios where a client needs a direct|partially filtered access to data. Not ideal for querying data from IE, but ideal for building a new generation of Web controls that breath AJAX. And just that.

I'm going to apply the "Driving force" pattern to virtually any CTP or buzzword that I happen to hear about. Next one is M-V-VM pattern for WPF (and Silverlight). And next, of course, the ASP.NET MVC FX. Stay tuned.

Looks, walks, and quacks like a O/RM

despos — Mon, 21 Apr 2008 18:57:00 GMT

OK, it's not a problem to admit that I have no copyright on it, but I just love to use it.

It looks like a O/RM, it walks like a O/RM, it quacks like a O/RM. Hence, it's O/RM...

What am I talking about? Just, guess what, LINQ-to-SQL.

It's not a full O/RM like NHibernate, Genome, or others. Oh sure, not like EF too :) But it looks like that. And if you're open-minded enough, it also may walk and quack like a O/RM. If this is an acceptable statement for you--it mostly depends on what your definition of an O/RM be--then this might be a pleasant reading: my latest article on DotNetSlackers.

Enjoy and let me hear your thoughts!

More on DataContext in (hopefully) a realistic world

despos — Wed, 19 Mar 2008 08:20:00 GMT

A comment from Bigyan to my DataContext article on DotNetSlackers yesterday made me think that I probably had to put my thoughts down a little more explicitly. The question that Bigyan poses is expressed in detail in this post of his and revolves around the best way of dealing with DataContext in multi-tier apps. My experience suggests that you use a new instance of the DataContext for each unit-of-work you need. The DataContext is a lightweight object to instantiate/dispose just for this. Can't find the link but somewhere on MSDN I've found a note (documentation) along the following, quite explicit, guidelines:

In general, a DataContext instance is designed to last for one unit of work however your application defines that term.

It is not recommended to maintain a DataContext live for more such as stored in a singleton object. Likewise, you don't have to pass it around, because it is not exactly a light object once in use with data attached: data + tracking info + mapping info. If you find it cumbersome to write new DataContext at each step, you can use a bit of abstraction and perhaps a Repository pattern. One more thought about not passing it around tiers and layers. You don't reuse the data context; you'll reuse the data you attach to it. So it makes sense to cache data somewhere and using distinct data context to do manipulation. On the new data context, you just attach any desired data. Yes, again, using the DataContext should be kind of using an ADO.NET connection: you use always new DbConnection objects but reuse command text and parameters. You never keep a connection open for a long time and don't pass it around tiers.

My two (euro)cents.

DataContext and design patterns

despos — Tue, 18 Mar 2008 15:46:00 GMT

An article on the internal structure of the DataContext object of LINQ-to-SQL is out on DotNetSlackers. Enjoy.

XAP has to be a MIME Type

despos — Tue, 11 Mar 2008 08:44:00 GMT

I love the VS embedded Web server, but it too often makes you thinking that things are easier than in the real and bitter world. So a trivial Silverlight 2.0 sample page that works from within the auto-generated ASP.NET Web site stops working as you create an IIS application for it. The symptom? No Silverlight content shows up and any network monitor tool you use reports that no XAP resource was found. But the resource is there. It's a simple matter of configuring IIS, though.

I was going to write a post myself. But then I found out that someone already did it. And it's a great write-up.

Hot AJAX+Silverlight2 in Warm Mallorca

despos — Thu, 06 Mar 2008 08:38:00 GMT

Silverlight 2 Beta 1 is out and I just installed it on my machines. OK--we're now ready for the next stage. A key question to answer for virtually everybody planning RIA development is:

Should I go with a Plain New Web approach and opt for AJAX on top of ASP.NET (or whatever else)?
Should I go with a Silverlight 2 approach in just a few months when it'll be released?

Put down simply, it's a matter of deciding whether you may or may not afford a plug-in of approx 4 MB to be downloaded/installed on client machines. Not an issue for intranet applications and most Web sites. Possibly an issue if you want to maximize (and even more) the audience of your portal. I expect to start seeing on home pages a choice Web|Silverlight in much the same way we see today a choice among languages.

To explore the tools, to refresh the techniques, to learn patterns and models, to grab directions, may I suggest you some hot stuff happening in a warm place this Spring?

What	Next Generation Web Applications
Where	Mallorca, Spain
When	April 9 thru 11
Details and Costs	here NB: all-inclusive-but-flight!!!

If you have a strong interest in sun and beaches in the warm Mediterranian islands, now you can also have some AJAX and Silverlight training in the same package :)

People of Europe (and not just Europe), join us ...

Source code for ASP.NET 3.5 Core Ref

despos — Wed, 05 Mar 2008 09:00:00 GMT

The latest baby, Programming ASP.NET 3.5 Core Reference, is out and you can freely download the full source code (21 MB) from here. It should be noted that for sort legal reasons (sic!), the binaries of the AJAX Control Toolkit couldn't be added to the package. It's therefore up to you to download and install it separately. Then make sure you have the proper DLL in the Bin folder of the book samples application.

Enjoy :)

Next Generation Web Applications

despos — Sun, 02 Mar 2008 17:34:00 GMT

The Web was first instantiated in the August of 1991 in one of the CERN labs. All of it was in two machines connected to the same network and exchanging HTML documents over the HTTP protocol. The HT in both names is not surprising—it stands for HyperText. And, at the time, the Web was all about hypertext and logical linking of related documents. HTTP was a mere (and simple) transportation protocol designed to be stateless and with no security pattern in it. HTML was designed to be a basic syntax to describe documents with just a bit of formatting, images, and hyperlinks.

Two years later, in the spring of 1993, I was still engaged on a COBOL/CICS/DB2 project, but actively managing to keep my head above water (read, get another job). At the same time, CERN waived any copyrights on the Web thus making it public domain. And the ball started rolling down the slope.

In less than two years, we had browsers and then JavaScript joined the group, then new tags were added to HTML, and CSS made its debut. In 1997, Microsoft came out with something called ActiveX. Horror story. You, the giant, willing to stop us, the people, from freely enjoying the Internet. It was clear already TEN years ago, that the only way to do serious RICH Web programming was extending the browser capabilities. But ActiveX had no security model (was security really a concern TEN years ago?) and no interoperability. And it was rejected.

Then Flash came in with a sandbox, full cross-platform support, and a poor programming model. For years, Flash has been the only tool available to incorporate rich graphics and media on the Web in a cross-platform manner. It was based on a plug-in, but who ever cared about it? J

In 2005, the AJAX revolution begins. Why in 2005? Because it was probably only then that the critical mass of browser adoption triggered the reaction. More than 90% of browsers happened to support the same set of features. And this set of features was powerful enough to build pages capable of bypassing the browser and make their own calls to the server.

So the whole AJAX thing began. And it’s a process whose termination is far from arriving. Not AJAX itself, but the paradigm shift it represents is revolutionizing the Web. As the mankind history demonstrates, a paradigm shift in any context is always a dramatic change. Have you ever heard of industrial revolution? Or perhaps relativity in physics? What happened next? Everything had to be reconsidered and redone.

Just like with AJAX today. Surely emphatic, but true.

What should we rethink, exactly? The browser, the JavaScript language, the HTML markup. More in general, the interaction model and page model.

Uhm. It’s getting tougher and tougher every word I type. Is it realistic to build true AJAX applications using the SAME browser, JavaScript, HTML language invented for the OLD paradigm? It’s like wanting to explain the Einstein’s relativity with the Newton’s law. Mechanics had to be reconfigured in light of relativity. Likewise, Web application models are to be reconfigured in light of AJAX.

Partial rendering is not the point. Script services are just one minor detail. UI extenders are only whistles and bells.

Browsers and, at least, the page model have to be reconsidered in light of the new AJAX paradigm. How can you describe it? We are transitioning from “I send you a form; you send me a page” to “I send you data; you send me data”.

To implement this new model, we have raw materials (script services, JSON), but we desperately lack technology. Building true AJAX applications is horribly hard. Which is not an excuse not to incorporate at least a bit of partial rendering just in EVERY Web page.

Desperately seeking to replenish creativity. Thoughts?

Latest and upcoming book(s)

despos — Sat, 01 Mar 2008 14:31:00 GMT

As per Thomas request, I'm posting here the TOC of the latest baby, that is Programming ASP.NET 3.5 Core Reference.

Chapter	Title
1	The ASP.NET Programming Model
2	Web Development in Visual Studio 2008
3	Anatomy of an ASP.NET Page
4	ASP.NET Core Server Controls
5	Working with the Page
6	Rich Page Composition
7	ADO.NET Data Providers
8	ADO.NET Data Containers
9	The Data Binding Model
10	The Linq-to-SQL Programming Model
11	Creating Bindable Grids of Data
12	Managing a List of Records
13	Managing Views of a Record
14	The HTTP Request Context
15	ASP.NET State Management
16	ASP.NET Caching
17	ASP.NET Security
18	HTTP Handlers and Modules
19	AJAX Partial Rendering
20	AJAX-Enabled Web Services
21	Silverlight and Rich Internet Applications (v1.x)

With boldface, the chapters that are either new or significantly reworked from the ASP.NET 2.0 edition. All the content in Programming ASP.NET 2.0 Applications Advanced Topics is not touched by the changes to ASP.NET 3.5. So it remains valid even today, in spite of the fact it has a 2.0 label.

As per Ricky request, instead, about the upcoming architecture books, details should be finalized this week. However, we're talking about two books of some 300 pages each to be available in the Sept/Oct timeframe. More details (and an outline) to come in just a few days.

Now I-Design Code

despos — Fri, 29 Feb 2008 09:47:00 GMT

It's been a while since last post, and it's only few weeks since I officially joined IDesign.

My debut with the new logo on the badge is this week in Frankfurt, BASTA conference. This year the conference is largest than ever regardless of the Launch events around it that are taking place in all countries in Europe and worldwide. Today, I ran the keynote—I believe this is my first time ever I'm invited to give a keynote presentation.

What else is going on these days? Uhm, an important thing that I really don't know if I have to catalog as a "change" or simply an "extension/enhancement". Either way, I realize that I'm slowly but steadily moving my interest towards architecture rather than nitty-gritty code tricks. The fact that I joined IDesign—the .NET architect company—should be seen in this perspective. Also, with the latest baby just hot from printers (Programming ASP.NET 3.5 Core Reference), I'll soon start on a new book project for the first time since Wrox books with a co-author.

I actually have two books in preparation to hit bookshelves this fall--both on .NET architectural topics. One is general .NET architecture for layered systems (patterns, principles, implementation, trade-offs) and will be authored with Andrea Saltarello. Andrea is new to the international audience, but is even more popular than me in Italy and especially on architecture. The second book, simultaneously developed, will be entirely mine and will be the Web complement to the other covering Web architectures and trends for both pure Web and Silverlight.

It's not just wording and smoky concepts. To be a real-world (and credible) architect, you need to know how code works, how code is written, and how technologies are implemented. You basically only save yourself to enter (ever) in code-monkey mode.

Later this year, I'll also have a brand new dazzling Architecture Master Class ready to "ship" at your own company or in public classes worldwide. Really worldwide :) If you're interested, either contact me or make sure you check frequently the IDesign Web site.

How ASP.NET devs are building AJAX apps

despos — Fri, 21 Dec 2007 09:49:00 GMT

Ran into this post and felt like I had to spread it around.

It discusses the results of a survey regarding how developers are integrating AJAX in their applications. I can't say much about the scientific or non-scientific nature of the survey, but it makes anyway for an interesting reading.

A first comment I have is that I'm kind of surprised to see that the best ranked product in the category of "Powerful suites of controls" is only the 5th favorite option (Telerik). It seems that .NET Web developers prefer tools that help to ajax-ify ASP.NET pages rather than the rich and powerful new frameworks that you have to marry and keep for a project-life-time :)

Thanks to Simone Chiaretta for the nice work.

Aspects of AJAX

despos — Fri, 21 Dec 2007 09:00:00 GMT

You probably know already about Gaia AJAX Widgets.

Featured on dnrTV, Gaia widgets provide a comprehensive set of ASP.NET server controls with rich AJAX capabilities--from Button to AutoCompleter and from modal dialog boses to panels. If you're a "classic" ASP.NET developer just should feel just at home. The programming model is kind of familiar and little intrusive--I'd say--in your coding practices and habits. Gaia also pushes to the limit the idea of aspect-oriented AJAX. What's that? It's code, in the end, based on the principle that you can dynamically configure "aspects" of a given control. Yes, it's close enough to what MS calls extenders. But in Gaia it is a native part of the framework.

A few days ago, the Gaia people (btw, a young company based in Norway) launched a competition with an interesting award: €10000 (sort of $14K). The deal is to create an open source AJAX collaboration web site (or similar) to make the world a little better place. Definitely, an ambitious project :)

And, of course, you must use Gaia AJAX Widgets for the purpose.

Instead of just focusing on food (and gain weight) in the forthcoming holiday season, why not just trying something else? :)

History article

despos — Mon, 17 Dec 2007 15:27:00 GMT

ASP.NET Extensions 3.5 is here and I'm slowly talking a look at it. You can read some considerations about the History feature on DotNetSlackers.

JavaScript exception instead of a message box

despos — Tue, 11 Dec 2007 20:56:00 GMT

The AJAX integrated support in ASP.NET 3.5 is nearly identical to ASP.NET AJAX Extensions with the significant exception of WCF services. I noted, though, also a slightly different behavior in the ScriptManager class as far as error handling for partial rendering operations is concerned.

When an exception is thrown during a partial rendering operation the HTTP request returns a regular HTTP 200 status code but instead of the updated markup, it includes a full description of the error. In ASP.NET AJAX Extensions for ASP.NET 2.0, the default error handler pops up a client-side message box with the exception message or any text you assign to the AsyncPostBackErrorMessage property. In ASP.NET 3.5, instead, you get a JavaScript exception.