Log in

No account? Create an account

Javascript and AJAX

I just spent about half an hour reading up on the basics of AJAX, and have fixed some theoretical gaps in my knowledge as to how the higher levels of webpage dynamism work; I'm right up to the level of where I'd be ready to start writing ECMAScript snippets and actually building those pages. I'm not versed in ECMAScript, but given the variety of languages I know I could still effectively program in it if I had to. I'm going to have to look further into HTML5 before I actually consider doing so.

I have a reluctance to go too far into this though, partly because of how I feel about ECMAScript; I don't believe that the web should be more dynamic than it needs to be, because the more people use general-purpose code rather than plain HTML, the harder it is for tools to mechanically deal with web content (as a perhaps-overly-strong example, webpages done entirely as ECMAScript-code-or-flash or that implement links as onClick events will both keep webspiders out, and may be harder to visit through automatic translators). For the (now mostly personal) web content I administer, actually moving to HTML5 would mean making two versions of a lot of the presentation layer; one for people like me who normally use tools like NoScript to disable Javascript by default for every site we visit (enabling it when necessary), and one for people who normally want more dyanmism. Do I want to rewrite POUND to use HTML5? It seems that HTML5 is crippled without ECMAScript; there's little point in the move without that.

I could of course just tinker with code samples and smaller projects to play with the technology enough to understand it; my database-query-explorer toy would be a prime candidate for conversion to HTML5+ECMAScript, and I could probably use the new canvas element as a much prettier alternative to calling GNUPlot behind the scenes to make images.

I realise that I may be committing myself to being a technological dinosaur on this front (which is safe for me because as a systems wizard, it's far from my intended area of expertise), and that I'm probably missing out on being able to do some really cool things with my existing software base. There is admittedly an appeal in pushing the presentation layers of POUND out into a client-side ECMAScript app and having the policy layers hook into an HTTP-RPC-layer instead. Sigh.