I've learned a ton of CSS in the last two days, by reviewing 7 Chapters of CSS the Definitive Guide 3rd Edition. I certainly feel more comfortable with the idea of CSS Layout. And I begin to see the elegance of the architecture. The 'sticky' part for me, is that CSS has no concept of an object that combines both text and graphic metaphors, outside the basic HTML controls.

For example, I cannot easily define this widget, in a reusable manner.

My brain describes this widget as a set of elements that are encapsulated in an container, but can be manipulated independently

A web based user agent can describe the button in two ways:

A Single Image

Pro: Simple to assemble and position on the page.

Con: Image and Text cannot be easily replaced, providing for reusability

Con: Global changes cannot be impacted by a simple modification to a style sheet. For example if the button background no longer fit with the theme of the site, all buttons would need to be replaced.

A grouping of independent elements that are 'rendered' by the user agent to appear as a button.

Con: There is no explicit definition of the container.

Con: Because there is no container, the 'container' is not reusable.

Con: Images, as border agents, restrict container's ability to be resized.

Pro: If correctly defined, global changes could be applied via modification to the style sheet.

Eureka

In fact, this second description is the way I need to think about all HTML/CSS layout. Three dimensional layering. The <div> and CSS definition provide the wireframe, skeleton, etc. of how the information will be displayed on the screen. A set of images are used to decorate the wireframe layout. Finally, images and text are used to convey the linguistic and visual metaphors for action and information.

When viewed from the Y-Axis Apex, the layers form a single structure. Yet, when viewed askew, there is no containment.

For someone who has been programming in OO for 15 years, the lack of visual containment is very difficult to wrap my mind around.  But by 'layering' the implementation, I get back containment through the use of <div>(s).  I just need to let go of visual containment.

Now that I think of it, this is probably why we programming types are clutching desperately to tables.  <td> gives me containment. 

I cannot wait for Silverlight / Flash ... somebody to rebuild the metaphor.  But until then, at least I get it.  When viewed from this skewed perspective the architecture is quiet beautiful.  As an engineer it makes me cringe, but as someone who loves art, I find it lovely.

It's been 10 1/2 months since I left Microsoft.

The first part of this time was a period of well spent self discovery.  Connecting with the Graphic Design and Web Development community in Jacksonville, helped me ground myself to the local area.  The wonderful trip out West, reminded me how to dream.  And the countless days enjoying the beauty of Riverside, brought me a sense of peace and place.

The next part of my journey was the creation of Jax EBikes.  I've learned so much about business in these last 6 months.  The store is beautiful and my neighbors are a community of entrepreneurs, that inspire me everyday.

But it's time to return to my roots, and get back to developing the kind of cutting edge applications I've dreamed about.  I had hoped that I'd be able to leverage the local .NET Developers Community.  But the Users' Group seems more focused on Enterprise Applications.  While the Enterprise can be a rich and wonderful space, it's complexity is more about maintenance and structure than the kind of creativity and beauty that inspires me.

So over the next few months, I'm hoping to build the 'Creative Coop' part of the space.  I hope to foster the kind of 'StartUp' energy we all felt in the 90(s), by funding a space for people to come and build on their ideas.  I'm hoping to connect back with some of my Microsoft friends to explore what we can create with Microsoft Expressions, WPF and Silverlight. 

We'll see, no matter what the outcome, it should be a wonderful experience.