HTML5 value semantics: a pragmatic perspective.

In recent weeks there has been a healthy and lively debate about value semantics in the HTML5 spec. Divya Manian on her "Our Pointless Pursuit Of Semantic Value" opinion article decries and points out the problems that can arise if we give HTML5 semantics too much importance. A bit later, Jeremy Keith's riposte to Divya's article brought forth the reasons why semantics are important in HTML5.

I will begin by saying that, in my opinion, HTML5 value semantics are purely academic. I don't mean that in a derogatory manner, but rather, the people who ultimately matter (i.e. the end user) could not care less about how semantically correct the markup behind a web page is. Will a semantically correct HTML5 page make you a better designer or developer? Probably not. Will a semantically correct HTML5 pages bring more views to your website? No, by itself it won't. Will your end-users have an improved experience by having semantically correct HTML5 page(s)? In my opinion, no. Will your customer/client approve your work because it is semantically correct or because it looks good and works well? I think we know the answer.

Also to bear in mind is that these semantics are shallow shells. That is to say, behind the scenes, these semantic tags are converted to spans, divs, h1s or other "primitive" html elements by the rendering engine and once painted on the page, they are invisible to the user. So, in many ways, I have to agree with Divya: time spent on figuring out the correct semantics is pointless. Not to say that you shouldn't use them, but they should not be an excuse or an obstacle to your creativity. Semantic markup shouldn't be the basis for design decisions and certainly shouldn't be a reason to hold a meeting about.

In short, life's too short; don't sweat the markup semantics.

 

html5/js-based Analog Clock

Last night I felt both bored and industrious. One of this things I've been meaning to do as of late is to mess around with html5 canvas (2d) object. So, I headed to http://jsdo.it which is a Japanese site with a take on "social" sharing of code. There I saw several "digital" clocks, but saw none that were analog (at least not in the front page or through searching). So, I thought it would be a good exercise to make one. Two hours later it was out for public view/critique. It was really fun and a bit challenging ( figuring out the methods available in the canvas 2d context was a bit of a pain) -- mostly remembering my high school trig :) Once I had the seconds "hand" figured out it was easy peasy thereafter.

Next, I plan to whip out couple of 2d fractal generators ... nothing Mandelbrot or Julia though (those get plenty of limelight and mentions).  If I see the canvas 2d platform lends itself to it and if I feel gumptious enough I'm going to start a math graphing and visualization html5/js library. If you are interested in collaborating in such library, drop me a line via Twitter.