Tuesday, May 13, 2003
Goodness gracious, great flaming CSS
I often stare skywards and say silent thanks for not being too technical. There's a whole lot of noise these days about the rights and wrongs of CSS (Dave Winer, as you'd expect, explains CSS far better than I can so, check out his pages if you don't know what it is, and that was posted way back in June 1999). Not being sure whether I had indeed drunk the CSS kool aid inadvertently, it merited some research to see what the fuss is about. Searching Winer's scripting.com shows that he has a lot to say about it, positive and negative, which may have been the start of the fuss. But, thankfully, it seems to be a techno-weenie (like a geek, only smaller) thing, perhaps best encapsulated in this. from the link above ... The huge sense of self importance you detect in most CSS/XHTML evangelists stems from their utter elation at finally pummelling their design expectations and wrestling with buggy CSS implementations until they finally compromise in some Israel/Palestine sense of the word. They become one of the few, and enjoy rubbing our noses in it. "Well I did it, so you must be able to! Unless you're stupid, that is.. you're not stupid, are you?" and this from the same place The point is that CSS won't let you do what you want I'm not sure any of that matters to me. We've used CSS in our DotP platform to what I think is great effect - you can see it working smoothly on ukonline. The pages load fast (faster than anything else I've seen bar google), they render on multiple devices and in multiple browsers more simply. Every technology project I've touched since 1999 has had to assemble a vast factory of browsers and operating systems to see if everything worked in every version of every broswer. That created an "n-squared" problem - a complex matrix of does/doesn't work ticks and crosses - many of which we either couldn't get to and some of which we couldn't fix even if we could. CSS seems to be the way around that now - as long as we use the same style sheet, then we know that every page using it works fine; if we change the style sheet then we retest just a piece of the original load. And style sheets are much simpler to edit it seems than the old nested "if this browser then do this" code that we had before. And grasping that we could make a single change and watch it sweep through the whole site (or even sites with DotP) was a seminal moment for me. I know of departments who have spent tens of thousands on their website just changing their logo, name or whatever. That's all gone. The purest and best demo I have seen of this is at the CSS Zen Garden - give it a go and you'll get the message really fast. So, whatever the pros and cons of how hard it is to write in (and I doubt for a minute that it is any harder than any of the other options), what CSS seems to get me and mine, is rid of a world of pain. That means we can free up time to worry more about what we say on the web, how we say it and where we say it than how it will look on any one screen or how it will be read in a given screen reader. That's a big plus to me. I'm religious on this too. The technically aware can debate this back and forth endlessly (viz dotnet, sun one, java, jini, bluetooth and any other set of standards, implementations and whatnot). What I know is that it works for me. I think we've pushed the envelope with DotP - we may even have broken it at an edge or two. For the first time, we've deployed something in Central Government technology that can be rapidly accepted and adopted by a departments, and that can give them an edge in delivering services to citizens. It makes things better for citizens by improving the experience, the way that we interact, the way services are offered and the ease with which they can be found. Not everyone sees it that way, of course, and (for a while at least) some folks will want to religiously debate the feature set versus perfection. But the bottom line is it delivers now and if you start now, you get benefit now. Or you can debate for a lot longer and get no benefit now and probably none in the future. Seems an easy choice for me.
Posted by Alan at Tuesday, May 13, 2003