This is my fault. Boom – there it is. No excuses, no lame attempts to laugh it off and blame Phil (even though it really is his fault). A lot of people have been messaging me via Twitter and leaving comments on my blog with the same question:

Dude, where’s the source! You promised 1-2 weeks! It’s been a month!

My response has been the same every time: the source will be ready by MIX. As you can imagine, people haven’t taken to this very well and have decided that:

1. This is yet another, typical Microsoft project with flaws and endless delays
2. This is just a marketing vehicle with no substance
3. I must have better things to do and I clearly don’t care

I can understand the frustration – honestly I can because I told you something that’s not happening and that’s not cool. Well I’m here to tell you that nothing’s changed really – I still can’t load the source (but I’ll tell you why in just a second) – but it’s not for any of those reasons (obviously).

Good Things Happen
The Storefront was probably one of the most organic applications I’ve ever worked on – something that was allowed to morph and change based on community feedback and changes in technology allowed. It changed faces twice:

• When I decided to refactor with an eye on harnessing the complexity using DDD
• When I decided (recently) to refactor yet again, to reduce “concept count” and to massage out a bit of extra poof put in there by my initial attempts with DDD

What you don’t know is that a 3rd refactor has just happened – and for a good reason: Microsoft is taking this effort seriously and it is now my 100% full time commitment (as opposed to before – the Storefront was almost completely an extracurricular thing for me). Now I know what you’re thinking:

OMFG Great!1!1! Another Pet Shop or Fitch and Mather!1!1!

I have nothing against those sample apps – nor do I have an issue with Starter Kits. But this is neither of those things. It has changed, it has matured, and it’s becoming the application I wanted it to become when I started this effort last spring: it’s becoming relevant.

We’re not all evil here. It’s not always about marketing, and it’s not always about pimping our stack – sometimes Good Things Happen and I’m asking for your patience for just another 4 weeks, while we tidy this thing up and I get ready to show it off (and some other things) at MIX.

So Why No Source!1!1!
Because it’s not working yet and we’re still nailing out concepts – architectural shifts (slight, but awesome) that will make this thing work in a relevant way. It’s always been in the spirit of the Storefront to do these kinds of things in the open and to remain transparent – but we have a “concept shift” happening that I’ve been pushing for, and we’re “spiking it” right now to see if it makes sense (which I think it really, really does!) – it really helps if you can do these kinds of things in small groups.

So I’m taking this 45 days to work this thing and I’m asking for your patience – it’s coming together and I think it will rock at MIX. This has been a great experience in learning in every way, and if there’s something specific you’d like to know more about (in terms of patterns or code I’ve shown with the Storefront) I’m more than happy to show you and write a blog post on it. Someone just today asked for more information about the Presentation Model Pattern – so I’m working up a post on that currently.

I’m not playing you and I have no other agenda right now except to devote as much of my time as possible to getting this right. It could flop, or it could fly. I just need my head down right now with the rest of the team to nail this – and I’m asking you to be patient with me. It’s only another 4 weeks.

I felt pretty good about where things were a few weeks ago, but as I started tightening things up and showing the app to others (doing walkthroughs and code talks), I began to feel… well a bit “meh” about it. Something was bugging me about the way it’s come together. Some people refer to this as “smell” – when things work but just don’t seem right. I’m about to change that.

UPDATE: Kevin Dente suggested I might have put my foot in it with my statement below that DDD is simply a collection of patterns. That’s what it means to me, at this point, however that’s not what DDD is. It’s a bit more, obviously, and involves working closely with your client as well as testing the tar out of your code. I’d already done that (arguably) and I should have been more clear.

Enter DDD
Please don’t hang up yet. Hear me out. I knew this day was coming, but I wanted to stay clear of “dogma” and let my instincts and tests dictate the application design. People had mentioned DDD, but to be honest I didn’t know enough about it to impose it’s strictures on the application.

Over the last few weeks I’ve been demoing the app to various people and walking through the code with them and to be honest, the story I’ve been telling is a tad fragmented. I don’t think it’s bad, but there’s a common understanding that’s missing. I’ve tried to pay attention to patterns where applicable, and I’ve also tried to avoid “crufting up” the application with dogmatic adherence to a line of thought.

I think I just got caught in the middle. 

Truth be told I sensed this about 3 months ago and at the time I thought “I need to stop and redo something here” – but I wasn’t sure just what it was that I’d redo. So I picked up copies of Nilsson’s and Evans’ DDD books (at the behest of many) and began to read them on the weekends and evenings.

Koolaid or Kool Kid?
I’m sure I’ve just polarized my readership – but I think, perhaps, there’s some common ground here that will 1) help me and 2) help you . If you don’t know DDD (and perhaps don’t want to know it because of… well shall we say getting beaten over the head with it), I think watching as I refactor the application with DDD in mind may really help you. There is a lot of goodness there.

If you’re a DDD wonk and have been waiting for me to “come to [Deity]“, well I’m almost there, but not quite. I really like the DDD approach for the way it deals with complexity – but complexity doesn’t exist when the app is small. This is something I’ve struggled to understand with respect to TDD and DDD.

In fact, so does Nilsson as he starts off his demo app, using TDD and DDD. He literally hams his way through his initial unit tests, implementing patterns and principles that will help him, ultimately, deal with an application he knows will get complex. I almost jumped out of my chair saying “how do you KNOW you need the FACTORY PATTERN YET!” and then BAM, he backs his way out of it and out of all DDD implementation until he really needs it (YAGNI baby!). I’m still unclear on this – I know it needs to be introduced at a refactoring point, but at the same time I also trust my instincts to know that all applications will end up being complex if they are successful.

Anyway, I’m now at a point in the life of the app where I can clearly see that it’s complex and needs to be bound by some common patterns and principles. That’s all that DDD really is – a collection of well-known patterns, many of which I’m already using.

A Little Help From My Friends
I’ve asked a few folks I know in the community if they’d be willing to come on and help me out here with implementing some core DDD principles, and also their thoughts on DDD as a whole. I think this will be a very good episode.

I’m never sold on “only one approach” to anything, and I’ll press these guys to help me implement what makes sense. I also know that not everyone is sold on DDD – even some of the more seasoned folks in the community. I was chatting with one of them today and asked what they thought and the answer still has me giggling:

Domain Modeling: Good
Jargon Explosion: Irritating

So very, very true.

I think this will be a really good episode, and to be clear, my goals are:

• Add some elegance using some well known patterns (Factory, Unit of Work, Repository)
• Reduce complexity and duplication (Services that just pass on queries, etc)
• Simplify

I’m aiming for Thursday/Friday for this episode to be ready.

There is a lot of work left to be done but I thought it would be a good idea to package up a Preview so that people can play with the site and the code.

Episode 24 Has Been Cut
I’ve been working on Episode 24 for the last few days, which deals with how to implement coupons and incentives. I finished the system (and I’m happy with it) but the screencast put me to sleep. It was pretty dull, to say the least.

So I’ve decided to focus everything on cleaning this up and getting it ready for general use. If you’re really interested in how the coupons were created, I’ll dive into it at a later time. For now I really want this site to be finished!

You can get Preview 1 Here.

To install, you’ll need to unzip into a directory of your choice and then restore the DB backup file that’s in the “Commerce.Web\App\DBFiles” directory (we’ve removed the DB from the App_Data directory).

Make sure you’re using Visual Studio 2008 with MVC Beta 1, and you’re good to go.

A few months back I got to talking to J Sawyer who is a Developer Evangelist in the Texas area about the work I’ve been doing with the MVC Storefront. Before I worked at Microsoft, J was my “main man” with the Commerce Starter Kit – the very first Open Source project I ever worked on.

During the course of the conversation J asked me if I had any plans to do any “WebForm-y” stuff; and I said “no” since this was supposed to be an ASP.NET MVC project. But then I started thinking…

ASP.NET 3.0 doesn’t have any Starter Kits out there (that I know of) and it’s time we had one . If I’ve done my job right – J should be able to take the back-end goodness (Services, Data Access, Unit Tests, etc) and put them behind an ASP.NET WebForms site. And he did!

And thus CSK 3.0 is born.

But wait! There’s more!
Part of J’s job at Microsoft will be to support this new project – something that I think is really, really groovy. More time, more love, more awesomeness .

In addition to the work he’s done to get the site rolled to WebForms – J’s also created a DynamicData admin site! I’ve been wanting to do just that for a long time – and I think this is a perfect synthesis of how MVC, DynamicData, and ASP.NET WebForms can all play happily in the same sandbox.

I sat down with J and talked about this and some other things – so have a look-see!

Watch it here. (15 minutes, 16 Mb)

The CSK 3.0 Project is here

J’s blog is here