< prev 22 Dec 2014 to 22 Dec 2014 next >

Posts from 22 December 2014

  • Week 6 Day 5 - The hero of the Greeks

    Today's partner: Max

    We spent the day implementing a lot of the things that are now standard on the modern web. Notably, modifying webpages while they're open without executing a page reload… and sending and receiving data in the background asynchronously by calling on the Homeric hero Ajax to carry our data.

    Wait, no. AJAX (Asynchronous JavaScript And JSON, originally XML) is indeed a means of sending and receiving data in the background while on a website, and it is somewhat heroic in terms of making the web work, but it's mostly a marketing term. What you're actually doing when you make an AJAX site is decoupling page rendering, moving the processes responsible for making your website look and operate onto the client side by leveraging small data transfers and pre-baked rendering operations that live on the client's computer.

    Someone else remarked that they would have felt capable of making a website (with server-side rendering) last week, but now they can't imagine not using AJAX, and I agree. The magic of the web today is being able to do things without having to transition, having access to all your tools through DOM elements appearing on page and stuff just getting whisked away to the cloud without your intervention. Everything feels snappier that way, and for someone like me (who gets angry with a 20ms lag while word processing) it's practically essential. I can already think of the sorts of things I want to implement and test with AJAXy techniques, but as we keep seeing, the class rapidly moves past "the coolest thing we know". The first week, I wanted to make an app in Ruby, then I realized what huge advantages SQL offers, then I thought Rails was the bee's knees, then I thought Javascript was the coolest thing we'd see… now I'm just waiting for the next thing.

    One thing, as an aside, that I can see is that there's virtue to practicing the things you're learning, but there's only so much to be gained. If the pet project you want to do is mostly nitpicking and trying to accomplish something the world has really moved past in the meantime, and all details of implementation… I can't necessarily see the virtue. There's always something to learn that will be maximally useful (of all the things you could be doing, and given your learning style) and it's not necessarily the obvious thing, or tying off the previous lesson.

    Learning relies on practice and scaffolding, but I'm getting the feeling that a lot of technologies require total comprehension of the basics (what's a closure? what's the difference between a proc and a lambda?), some understanding of the intermediate ideas, and total ownership of the thing that makes you money. To get to that point, you have to jump to the next thing while you still have the previous thing fresh in your mind, and have everything gel as a collection of "correct practices" and good habits, … and eventually you can make progress, actually breaking new ground after reaching the top of the (first) cliff. (The rest of the cliffs, of course, are "real" problems to be solved.)

    Learning all the nuance about everything is just pointless, though. There are just too many libraries and frameworks to learn back to front, and generally (successful) people build on what came before, and so the new thing is more or less like the old thing in structure and/or syntax.

    I can foresee, or imagine, programming always tickling the dopamine receptors in my brain because there's always some new framework… in the 50s, academic languages were the forefront; in the 60s it was all about business-specific languages; the 70s saw C gain popularity, with its general purpose systems programming capabilities… but there's not so much that's new in the history of computing. And there's never a point where you can sit still, content with what you know. Maybe that's a good thing.

< prev 22 Dec 2014 to 22 Dec 2014 next >