Tuesday 30 August 2011

I <3 WWW

As I'm not an iOS or Android developer I attended the August "Mobile Focused" Melbourne YOW! Night with a certain degree of reluctance.


I had visions of skinny-jeaned, black-plastic-spectacled iOS-developing hipsters telling me how, like, awesome their, like, app was because like, it totally like mashed up the tweeting paradigm by like ironically cross-pollenating the artisan tumblr mixtape vibe (thanks Hipster Ipsum!).


And while there was a certain amount of that, I was somewhat refreshed to hear from a couple of speakers (native mobile app developers, mind you) that a great deal of the time, when you strip away all of the hype, what the customer needs can be addressed with a mobile-focused website.


I've seen first-hand the hype-storm that (in particular) the iPad has created. One of the speakers at YOW! painted a pretty accurate picture:

  • CEO's wife buys CEO a new iPad because they're shiny
  • CEO can't really use iPad but notes these apps on the home screen have pretty icons
  • CEO sees other companies have apps
  • CEO decides his company needs an app
  • CEO doesn't exactly know what app is for, but knows that it's needed NOW


What follows (MyCompanyApp v1.0) is usually an app-ification of what the company's current website offers. It is a pointless, money-wasting exercise that gives the CEO his app icon that he can show off in the golf club, but actually has zero advantage over his website (actually negative advantage because it doesn't self-update like a web page).


If the iPad had a way to add a shortcut to a webpage as a home-screen icon, the whole thing could have been done in 30 seconds.


As mobile devices get increasingly capable, well-written, lightweight web pages with advanced features like HTML5 Geolocation can get pretty damn close to the "native app experience" - with none of the approval-process or old-installed-version hassles.


So that's the basket where I'm putting my eggs for the next little while; I just hope the egg doesn't end up on my face :-)

Monday 22 August 2011

The CSS F(l)ail

I've written before about CSS and why it seems to be so hard for a Typical Java Developer (TJD) to craft "nice" HTML+CSS presentation layers.


Here's the behavioural pattern I've observed when a TJD has to do some front-end web-work

:
  • Panic - rather than embrace the opportunity to learn a new paradigm/technology, as they would probably enjoy doing if it was a Java API

  • Copy-and-paste from the existing codebase - despite knowing that this leads to unmaintainable code no matter what the language

  • Copy-and-paste from the web - without understanding what is going on

  • Randomly tweak values in the browser using Firebug or Chrome Developer Tools until it "looks right" (I've started to call this CSS flailing)

  • Give Up And Use Tables (for non-tabular data). The ultimate fail

In posts-to-come I'm going to try and identify some CSS "patterns" that a software developer can actually understand (as opposed to copy-and-pasting) which hopefully will put an end to the above behaviours, any of which would be regarded as unforgivable if it was perpetrated in the squeaky-clean corridors of Java-land ;-) but are somehow considered par-for-the-course in the unfamiliar yet forgiving world of UI...