Upgrade to Pro — share decks privately, control downloads, hide ads and more …

CSS Scraping and Analysis

CSS Scraping and Analysis

This talk was given at Analyze Boulder on January 7th 2015. It announced some work I had done scraping and analyzing CSS files from 20,000 sites and my findings.

I scraped CSS files from 20,000 websites and compiled them into some charts and graphs that will affect my web development work at Quick Left. I learned about responsive breakpoints, color usage, progressive enhancement, and more.

Alex McPherson

January 04, 2015
Tweet

Other Decks in Programming

Transcript

  1. I Studied How Websites Are Made • 20,000 random sites’

    CSS Files • Put into PostgreSQL • Explored with RStudio
  2. Findings! If you’re a web developer, these are interesting. VERY

    If you’re a not… they’re interesting. sort of
  3. Finding #2: CSS Quality We found people tried to use

    300! There are about 80 valid things to control with CSS
  4. Finding #2: CSS Quality We found people tried to use

    300! There are about 80 valid things to control with CSS
  5. Finding #2: CSS Quality We found people tried to use

    300! There are about 80 valid things to control with CSS
  6. Finding #2: CSS Quality is Lacking People are writing junky

    CSS This leads to junky websites We use tools to write less junky CSS
  7. Finding #3: Breakpoints It’s good for the viewer! Looks right

    on any device. It’s called ‘responsive’
  8. Finding #3: Breakpoints I need to pick the widths where

    it ‘jumps’ from one layout to another. The more breakpoints I have to take into account, the harder styling a page is.
  9. Finding #3: There is only 1 breakpoint Most sites respect

    ~768px (This is where most tablets start)
  10. Finding #3: There is only 1 breakpoint I can tell

    my designer to start with just 2 layouts