.NET Architectonics

Crushed by complexity—me too, and for a long time

despos — Thu, 12 Nov 2009 15:37:00 GMT

I had no sessions on ASP.NET MVC in this Fall DevConnections, but I’ll have quite a few in the upcoming Spring 2010 event which will coincide with the Launch event of VS 2010. Yet, many of the questions I got at the end of my sessions (code design, domain model, social network applications, jQuery UI) revolved around ASP.NET MVC and, as one nicely said, my recent conversion to it J

Technically speaking, it is not a matter of a conversion.

I didn’t change my mind about it. More simply, I decided to have, at last, a non-superficial look at it. I got to know it, and I recognized its inherent superior architecture, its core simplicity, its extensibility that largely stems from proper class design. It is so extensible that … it can be made any complex and sophisticated without impacting people just looking for a close-to-the-metal framework.

This is a stunning point compared to Web Forms.

In Web Forms you have to fight to achieve simplicity (the definition of simplicity generally agreed on in 2009), and you can hardly have it cheap, and often you can hardly have it full. In ASP.NET MVC, instead, you currently have to fight to have some good complexity and abstraction in the view. Still provided that wanting a grid can be considered a form of complexity.

So why did I toughly disregarded ASP.NET MVC from about one year since its inception? Let’s look at the dates.

· October 2007: Microsoft releases the Preview 1 of ASP.NET MVC. And I was among the first to write a comment (see DotNetSlackers.com) and to emphasize its real REST nature versus the alleged MVC-ism.

· October 2008: After a wave of Preview X, Microsoft releases Beta 1. And I started on it—very slowly—but started. My articles on aspnetPRO (now DevConnections magazine) witness that, as they start with the January 2009 issue.

· March 2009: Microsoft releases v1.0. And I see the (green) light.

Now what?

Hey, but in this way you didn’t help Microsoft to build to product. I don’t work for Microsoft. I work for myself and another company—and it’s hard enough J

Hey, but in this way you didn’t help the community (and yourself) to have a better product. Well, this is a more delicate point that probably has a lot to do with what I really do every day. I do help clients building their own systems, but I’m hardly sitting writing production code. All I need is, knowing a number of products/frameworks deep enough at their core to deliver classes, write books, create prototypes and proofs of concept, and fix hard design issues. (Little detail—with a bit of money attached.)

Hey, so you want to be paid to share your high thoughts on Preview 345 of any SomePieceOfSoftware framework? As long as previews come out every four weeks or so (not to mention nightly builds you can optionally look at), well, it’s a lot of (my) time. It means subtracting time to projects, clients, family, friends, and worst of all, tennis J Getting a return doesn’t necessarily mean getting some cash, however. It just means getting a return of any type that has value for you. Quite simply, I don’t see any value for me in previewing CTP 4 of the 3^rd week of 7^th month of the year.

There’s a popular proverb we use in Italy (it probably exists in other languages too) that is used to celebrate the old good times. It says “there’s no more the half-season” and it is used typically to complain about the weather too hot in April or too cold in October. The common wisdom expects some sort of transition between changes in season.

In software, we always had sharp changes—from version N to version N+1. But now on the wave of open-source and community-driven projects, in software we are having the concept of the half-version introduced. And worse yet the transition is getting longer and longer. The granularity of the release breaks in smaller pieces every day. The average number of Previews grows, followed by a few CTPs, then Beta 1, then Beta 2, then a few RCs, and finally RTM.

Is it all of it? Not yet. There’s an immediate SP1 for anything that didn’t make it in the RTM and perhaps a R1 or R2 later on.

The software product is like a mutant virus—you don’t know its direction (and it can get worse at any time) until it reaches the Beta 2 stage. After that, it will certainly mutate again but is now harmless for developers J

“Crushed by complexity” is the title of a session that Billy Hollis delivered at DevConnections. Unfortunately, I couldn’t be there but whatever he said I feel I agree J

ASP.NET MVC is MUCH better than MS seems to think

despos — Wed, 11 Nov 2009 06:08:00 GMT

Here at DevConnections I just attended the ScottGu’s keynote on Visual Studio 2010 and Web development. I haven’t had much time to spend on the latest Beta of Visual Studio 2010 yet so I found most of the information quite helpful and interesting. It looks like Microsoft is finally coding some good functionalities from ReSharper and .NET Reflector inside of Visual Studio 2010.

Little gems like “Generate stub method” or “Display the hierarchy of calls” are now available natively. Still ReSharper is a must I think, but it is good to have a really better VS. By the way, I haven’t seen yet a preview of ReSharper for VS 2010 so the gap might still be quite large J

My primary focus these days is for ASP.NET MVC as I’m doing—guess what—a book scheduled for the release of the .NET 4 platform next April. I particularly loved the TDD spin of ASP.NET MVC that is visible from tooling support. I’m not simply emphasizing the possibility of doing unit tests; I’m referring to the possibility of writing tests first and then the code. And here’s that facilities like stub method and types generation come into play.

And finally, I respectfully but strongly disagree on the slant of many Microsoft presentations that touch on ASP.NET MVC. I disregarded ASP.NET MVC for too long. Now that I got to know it from the inside, well, it’s the best (large) piece of code I’ve seen for a long time.

It’s much better than MS seems to think and tell.

In my opinion, it really represents the way to go for most developers in the mid term. It is probably premature to suggest today that Web Forms be abandoned to embrace ASP.NET MVC. But the new framework (in version 2.0 with Visual Studio 2010) drives you toward better code. Don’t get it wrong: ASP.NET MVC doesn’t automagically make your code clean, elegant and flexible. You still have to go a long way ahead but it puts you on the right track and delivers an environment where you can write testable and practice seriously with unit testing.

A point that Scott made about ASP.NET MVC is that it gives you total control over HTML. Not that this is a false statement, but it happens because server controls are marked as evil. They still work (even though their use may compromise the design—hold on) if you want and if you use the increase your distance from HTML. But on the other hand, if you stop using server controls in Web Forms you get closer to the HTML metal also in Web Forms. So in the end the reason for using ASP.NET MVC is Separation of Concerns and testability.

Eventually, Microsoft will be able to add that to Web Forms too, perhaps via the MVP pattern as the Web Client Software Factory framework attempted to do a while back.

The dead-end of Web Forms

despos — Thu, 01 Oct 2009 14:01:00 GMT

I had a talk last week at BASTA about ASP.NET MVC vs. Web Forms and I repeat the same talk today here in London at Software Architect conference. (Well, repeating a session is a big term for me--I'll never be able to repeat the same session the same way two or more times...).

The key question that people ask, the only answer they want to hear, is about which one is preferable to use for the next project. Clearly, the natural answer would be a classic "It depends". My rule of thumb is fairly simple

If ASP.NET works for you, then stay with it.
If you start complaining about limitations you experience (not limitations others say you are experiencing), then look ahead.
Once you decided to take the plunge into ASP.NET MVC go ahead and never hesitate. If you seem not understanding it very well, study it more.

When introduced, Web Forms was a cutting edge solution and it just engineered current best practices. But it was ten years ago. We could argue whether it was the right choice to engineer ASP practices ten years ago. In fact, more or less at the same time Sun did it differently when they architected JSP.

There's not much more you can expert or achieve with Web Forms than you do today. OK, tomorrow, with version 4. This is the dead-end of Web Forms.

If it doesn't serve you any more the way you like, change it. It will be a change for the better. But the better is also different and requires a different approach and skills. Design is design, and with ASP.NET MVC (which is far from perfection, by the way...), you need to gain design and architecture skills.

My next book is just on ASP.NET MVC (February 2010) and will target version 2. Like many other books of mine it won't be an how-to book. And I'm taking architecture and design very seriously as I explain controllers, views and models. Stay tuned. And plan your design training :) Contact me... ha ha ha Smile.

Silverlight Coding Contest

despos — Thu, 02 Jul 2009 15:30:00 GMT

A few days ago, ComponentArt started a Silverlight coding contest. If you submit your Silverlight applications for evaluation by the "Esteemed Panel of Judges" you can win a prize of $10,000 or get free licenses to ComponentArt products if you happen to be one of the two runner-up. Much more details available at the following sites:

- http://www.componentart.com/ - http://www.componentart.com/community/competition2009/ - http://www.componentart.com/BLOGS/miljan/archive/2009/06/22/10-000-for-the-best-silverlight-app.aspx

BTW--I'll be one of the esteemed judges :)

DataForm Control in Silverlight 3

despos — Thu, 02 Jul 2009 15:25:00 GMT

Silverlight 3 comes with a new control—the DataForm control—through which you can design a view model around a data type in what actually results to be a specialization of the MVVM pattern. The DataForm control is also smart enough to analyze the public properties of its data source and generate some UI accordingly. The DataForm will use text boxes for string properties, check boxes for Booleans, and a date picker control for dates. If bound to a collection of objects, it will also display a navigation bar.

Read the full (part I of the) story on DotNetSlackers.com.

ASP.NET 4.0: more control on viewstate management

despos — Sat, 13 Jun 2009 20:04:00 GMT

ASP.NET is a stable and mature platform for building Web applications. Personally, I can hardly imagine a revolutionary set of new and compelling features to be added to it. So what's new in ASP.NET 4.0? Beyond AJAX stuff, there are some interesting enhancements in the Web Forms area. As I see things, all changes in ASP.NET 4.0 can be catalogued under the label of "more control". You get more control over viewstate, script references, ID generation, output caching and even, but in very limited form, over HTML generated by some controls.

Let's briefly focus on the viewstate extended control. In ASP.NET, the viewstate is optional but it is enabled by default. In addition, the viewstate is not simply a way of reducing your bandwidth. It is rather functional to the implementation of the Web Forms model. So just dropping the viewstate in a new version of ASP.NET is simply out of question: either you get a new ASP.NET platform such as ASP.NET MVC or you stick to Web Forms with the viewstate on board.

However, there's a subtle aspect of the viewstate management that has been fixed in ASP.NET 4.0. I said fixed because, well, from my perspective it had to be considered a bug-by-design in previous versions.

All server controls (the Page class derives from Control) have the EnableViewState property through which you can disable the viewstate for that control. What is little known ais that the EnableViewState property is ignored for child controls. In other words, if you take the default value (true) for the page, then whatever value you assign for it to any controls in the page... it is ignored. You can have have TextBox1.EnableViewState = false but still have the text box to read/write state from the viewstate if the viewstate is enabled at the page (or container) level.

This will change in ASP.NET 4.0 thanks to the new ViewStateMode property.

This property indicates whether the viewstate for the control is enabled|disabled|inherit. You can use the ViewStateMode property to enable view state for an individual control even if view state is disabled for the page. This is the great news. Finally, you can now disable the viewstate on the page and decide which controls will have it enabled (opt-in). In earlier versions you could only do the reverse (opt-out): enable the viewstate and then decide which controls will not support it.

Give a chance to prediction

despos — Thu, 21 May 2009 08:04:00 GMT

It is mostly about AJAX applications, but it applies well to any scenario where a smart/rich client is present. I'm talking about the "Predictive Fetch" pattern. Quite simply, it refers to the idea of preloading data that the current user can request in a few moments. It relates to caching--more, it is often implemented through caching--but it is a different kind of thing. It is actually a strategy for certain pieces of the user interface, where you need/want to exceed expectations and provide an output close to (if not under) the threshold of human consciousness (about 10 ms). Immediate response.

Like many other AJAX-related things, it is mostly a matter of tradeoff. You are guessing, that's what you're doing. And the guess can be right or wrong. If wrong, you have just wasted some resources and CPU cycles on both client and server. If right, you astonish users. BUT... because you can hardly afford pre-fetching from every possible use-case, then the open point is: what the user reaction/feelings when one feature is sooo fast and another similar one is slower?

The full story on DotNetSlackers.

Testability vs. Testing

despos — Wed, 13 May 2009 09:26:00 GMT

The first time I heard the expression "unit testing" at a .NET conference was--if memory serves well--back in 2004. Since then, it took probably a couple of more years for the theme of "unit testing" to gain due visibility in the .NET space. I can't mention a date when I heard the word "testability" instead. And still in articles, blogs, and conference talks the prevailing term is "testing". Not that no conferences in the world had the word testability pronounced lately, but it never happened to me to attend a talk or something where I could hear it--maybe I just go to the wrong sessions and miss a lot of great events :)

Seriously, I feel that the word "testing" is used much more frequently than the word "testability". And I do believe that for developers and, especially, architects testability is a far more important aspect. In first place, testability is one of the required attributes of any software system according to the ISO standard for software architecture. Second, testability has a deeper impact than testing on the quality of the code you write. What's testability?

In our book "Architecting Applications for the Enterprise" Andrea and I write it as follows:

A broadly accepted definition for testability in the context of software architecture describes it as the ease of performing testing. And testing is the process of checking software to ensure that it behaves as expected, contains no errors, and satisfies its requirements.

In a nutshell, testing returns you a working application; testability returns you a better designed code. Honestly, can you spot any difference between a unit-tested piece of code that works and a piece code that just works (maybe tested by simply poking around)?

With the excuse of letting you do that cool thing called unit-testing more quickly and effectively, testability silently drives you towards a much better design of the code with plenty of SoC, dependency injection, low coupling, high class cohesion, and minimal responsibilities for classes.

Ultimately,design for testability means applying a software contract (code-contract in .NET 4.0) to classes and methods and keep classes cohesive, loosely coupled, and with dependencies clearly listed. Once classes are testable, unit-testing them is really a little detail.

Things to say

despos — Sun, 03 May 2009 07:33:00 GMT

Two interviews out at about the same time. One is on mythical DotNetRocks; one is on SodThis--brain burps for the tech savvy. Love the payoff :)

Enjoy!

AJAX Architectures Condensed

despos — Fri, 17 Apr 2009 14:40:00 GMT

One thing is using AJAX to dynamically refresh a small piece of a single page; all another thing is designing a whole presentation layer to be partially refreshed in every possible operation against the server. An individual feature can be happily and nicely coded using a smart piece of JavaScript; a whole Web presentation layer will cost you a lot more if done entirely in JavaScript. And from scratch.

Sure, new productivity tools are created every day (from the popular jQuery library to the upcoming ASP.NET AJAX 4.0 framework), but the most effective way of adding AJAX to applications continues to be the subject of research and begins to look like the Holy Grail of Web software.

I just published an article on DotNetSlackers, that summarizes the options you have when it comes to AJAX architectures. It is an excerpt from Chapter 3 of my latest Web architecture book mentioned in this recent post of mine.

A lovely couple (of architecture books)

despos — Thu, 16 Apr 2009 18:15:00 GMT

It was brought to my attention that there's no way to have a (free) look at the TOC of Microsoft .NET: Architecting Applications for the Enterprise other than holding the book in hands and flip through pages. Sorry about that, this post is to make up for the omission. There's no reason and no intention to keep it secret :) Here it is:

Architect and Architecture Today
UML Essentials
Design Principles and Patterns
The Business Layer
The Service Layer
The Data Access Layer
The Presentation Layer

The book develops its content more in terms of general patterns than concrete technologies. It is, however, concretely bound to the MS platform and it is not hard to recognize products and technologies behind the scenes. At the same time, it discusses other mostly open-source alternatives that fit nicely in the .NET stack.

Mid April 2009 is also the release date of the Web counterpart of the architecture book. I'm talking about Microsoft ASP.NET and AJAX: Architecting Web Applications, always from MS Press. Below, you find the TOC. I particularly recommend chapter 3 where I generalize most common approaches to AJAX today coining two (new?) terms: AJAX Server Pages and AJAX Service Layer.

Under the Umbrella of AJAX
The Easy Way to AJAX
AJAX Architectures
A Better and Richer JavaScript
JavaScript Libraries
AJAX patterns
Client-side Data Binding
Rich Internet Applications

Chapter 6 and chapter 7 contain concrete stuff about AJAX patterns with a lot of references to existing frameworks. Chapter 5 doesn't miss some jQuery coverage. Finally, Chapter 8 is about programming the Silverlight 2 model. Nothing on animation and graphics but everything a .NET developer/architect needs to know.

Web Forms vs. ASP.NET MVC

despos — Sat, 11 Apr 2009 11:03:00 GMT

Probably I'm a bit late to the party, but as I gain confidence with ASP.NET MVC I feel I have my cents to share :-)

ASP.NET MVC is not the anti-Web Forms and Web Forms is not the anti-pattern of ASP.NET development. ASP.NET was created 10+ years ago and in the late 1990s the MS platform was VB-oriented. So it was a natural choice to design ASP.NET as a stateful framework over a stateless medium--the Web. Many of the hot features in ASP.NET (postback, viewstate, forms authentication/authorization, server controls and abstraction over HTML) were eagerly welcomed by the community because they were time-saving facilities just implementing common features that everybody would have aboard.

Over years, MS probably failed drawing Web people attention on architecture and software design. There was an interesting attempt made with the Web Client Software Factory to implement MVP and workflow-based page navigation in Web Forms pages, but that turned out to be too far complex. At least, in my opinion. So in some way instead of driving Web Forms towards a better design it seemed preferable to introduce a new framework on the wave of the success gained by RoR and MonoRail. To many people, ASP.NET MVC looks like the new way to go and the only way to achieve SoC, testability, better design.

For sure, ASP.NET MVC is a "new" ASP.NET designed ten years later looking at the current state of the industry and evolution. Comments on Web Forms vs. ASP.NET MVC are often biased and stuffed with pieces of personal experience that is elevated to the rank of absolute, objectives facts.

The table below lists my top-ten undisputable facts about ASP.NET MVC and Web Forms. These are facts--everything else, I believe, are opinions. And like all opinions they are fully respectable. At the end of the day, Web Forms and ASP.NET MVC is like car or motorcycle and ... it's up to you, your skills, your education, your attitude, and your project requirements.

Web Forms is hard to test
ASP.NET MVC requires you to specify every little bit of HTML
ASP.NET MVC is not the only way to get SoC in ASP.NET
Web Forms allows you to learn as you go
Viewstate is not the evil and can be controlled/disabled
Web Forms was designed to abstract the Web machinery
ASP.NET MVC was designed with testability and DI in mind
ASP.NET MVC takes you towards a better design of the code
ASP.NET MVC is young and lacks today a component model
ASP.NET MVC is not the anti-Web Forms

On this topic, an article on MSDN Magazine is coming out in the July 2009 issue. A Web architecture book is being published these days and I'll start soon on the MS Press big book guide on ASP.NET MVC to be available towards end of the year.

iBrii is gaining ground

despos — Sat, 04 Apr 2009 20:19:00 GMT

IBrii is a Web application that I've been working on in the past weeks. It's not my baby, and I'm not one of parents, but I can definitely be called as an hooked-on uncle of the baby :-)

So I'm even more proud of the work we did when I read what's on in the Wired How-to Wiki. IBrii is recommended as the best tool to replace and extend the just retired Google Notebook. IBrii is even more, tough.

It is a virtual notepad through which you to write your personal notes, organize and share them with your friends. Your personal notes can contain just everything you can find in the Web. And you can clip-and-share as you browse the net. IBrii is still a beta but is going to support documents (PDF, Office documents) and RSS feeds meaning that you can publish your notes of choice to your blog.

First read this, then rush to get your account today! You'll love it.

PS: Technically speaking, IBrii is a relatively simple ASP.NET application with a lot of jQuery to shape up the UI and calling into an AJAX Service Layer for server-side logic.

Native, Immigrant, or Practitioner?

despos — Tue, 31 Mar 2009 06:48:00 GMT

Marc Prensky coined the term digital native to refer to persons who have grown up with digital technology including computers, Internet sites, and applications. In particular, Marc said that today's "students are no longer little versions of us, as they may have been in the past. In fact, they are so different from us that we can no longer use either our 20th century knowledge or our training as a guide to what is best for them educationally."

Admittedly, terms digital natives and digital immigrants shocked me when I first heard of them.Last night I caught my son (11) in his room doing "some urgent work" and precisely:

Chatting over latest Messenger with a friend about the day and the tennis lesson they just had.
Imparting instructions on how to configure the Windows desktop over Skype.
Using iBrii to do his homework in collaboration with another couple of friends.

And my 11-years old son just self-installed Windows 7 and is actively using it and learning most of available configuration aspects. Which is definitely a good statement about Windows 7. The end of Windows hell? In some way, I glimpse similarities between Windows 7 and the first OS I really enjoyed so many years ago--the mythical Windows 95.

When I mentioned all of this with my wife, she said "How do you call yourself? Are you a native too or are you an immigrant?" Well, I believe that, all in all, I'm just a practitioner :-)

There’s a lot of Software in Our Future

despos — Fri, 27 Mar 2009 14:24:00 GMT

It’s quite a bit of time that Juval Lowy—a Microsoft software legend—talks about energy and its relation to software and calls this the next bubble prefiguring the next boom for developers. To paraphrase John Lennon, he may be a dreamer, but he’s not the only one.

Weeks ago, having heard me a bit concerned about our future, my kids (11 and 8) started their own conversation about a world without computers and programs. How could do school homework without Word to help us put down writings? How could we search for details without Google? How could we get in touch without email? And how could we chat without Messenger?

And I also wondered what else I could do in a world without computers? If nothing happens, if the drift is unremitting, is the Western society destined to collapse?

It’s certainly possible. Never set limits to the human stupidity.

But, beyond temporary depression J, the software is the key to the future. Not a remote future, but the future that begins tomorrow morning.

A lot is being said about alternative energy, but I feel that real point is optimizing the demand/response pattern through which we get energy for our everyday life. I totally buy Juval’s statements in the DevConnections Spring 2009 keynote “The Energynet: the next boom in software”. We need smart power grid as opposed to today’s dumb grids. It is impossible to imagine a physical replacement of the grids all over the world. But we can make the grid smart. In just one way—the software.

It seems like we are witnessing an era of epochal changes—for our generation, the second big change after Internet. We’re not there yet. But we’ll get there. And the key of everything is – and will remain – the software.

Thanks God, kids, dad is in the right business J