Skip to content

Instantly share code, notes, and snippets.

@iKenndac

iKenndac/podcast.md

Last active June 4, 2017 16:52
Show Gist options
  • Star 0 You must be signed in to star a gist
  • Fork 0 You must be signed in to fork a gist
  • Save iKenndac/aa2ef2b2c3c7deba1f85f428b4f43925 to your computer and use it in GitHub Desktop.
Save iKenndac/aa2ef2b2c3c7deba1f85f428b4f43925 to your computer and use it in GitHub Desktop.
Download ZIP
Raw
podcast.md

iKenndac's podcast idea:

Each episode starts with a story, fact or popular issue, followed by some discussion on that story/fact/issue and how it can teach us lessons in modern development. Personal stories are best, or facts that can bring people in that have interesting and relevant experience.

Podcast episodes would likely need guests with expertise in that episode's field to be engaging.

Episode: What if it is actually life and death?

Story: Spotify town hall meetings and the hyperbole of the competition wanting to take away our families and children. Then, the story of the flickering radio cover art bug. Which one is actually more life-and-death — a start-up's competition, or a cover art flickering on a car screen?

Discuss:

  • The hyperbolic life-and-death mentality of start-up culture, a modern-day tribal chant at times. The ridiculousness of the contrast between that and leaving the meeting to continue your task of bringing cache latency down by 15ms.
  • The idea that if you really dig deep, you kind of realise most apps just don't really matter that much in the grand scheme of things. They're encased in glass rectangles, separate from the real world of consequences. If Spotify goes down for a while, nobody dies.
  • This starts to change when you exit the glass rectangle. Hypothetical situation: What if you're driving along through a school district and that cover art flicker happens, causing the driver to look away from the road just as a child runs out from parked cars. Bam, a child dies.
  • "But that's a one in a billion chance!" -> There are over 1.2 billion cars on the road.
  • "Improbable, but plausible"
  • Whose fault is it? Who gets sued?
  • Talk about driver distraction rules, and the stuff I had to implement at Spotify. Ford's repeating but not-too-often "Spotify is not logged in" voice prompt.
  • Is this why car software has a reputation of being bad? Is it actually bad? Are the challenges between bringing this heavily-regulated industry together with the fast-paced development of the modern smartphone, or is CarPlay destined to be closed off to third parties except for a few minor categories of apps?

Episode: The Internet Really Creeps Me Out, Man

Story: Adding an audio route log to the Spotify client and getting millions of data points within a couple of days, and panicking that I'd accidentally screwed up. Turns out, it was just a "drop in the bucket" compared to everything else.

  • The proliferation of in-app analytics and user tracking (mention web analytics and that they're out of scope for this discussion)
  • My own journey with having nothing, then moving to Spotify and seeing the firehose that's there. Then starting Cascable and having in-app analytics.
  • What if your analytics provider gets hacked?
  • Personally I have services in my app, but have no intention of tracking my users. I have ad tracking to find out which of my ads are worth the money, and in-app analytics to find out what's working or not for people. I try my hardest to make the data anonymous. Is that ok? Why? Why not? Does intent matter?
  • Having law enforcement contact Cascable and the perspective shift that brings — suddenly a cache eviction behaviour becomes very important.
  • Does it always become skeevy? E.g. Uber.
  • …but in-app analytics are really useful! Is it possible to have both worlds, or is an on/off user option the best we can hope for?

Episode: Apps And The Kind Of Customer They Bring

Story: Working at KennettNet years ago on Music Rescue, an app for recovering music from an iPod back to a computer after disaster. Before the days of popular streaming services or even re-downloading your purchased music. The customer that screamed at our poor customer support agent until she was in tears.

  • Contrast customers from different types of apps:
    • Customers that are already angry before they start using your software (Music Rescue)
      • In retrospect, a phone line was probably a bad idea
    • Customers of free/cheap apps
    • Customers of 'Premium apps' like Cascable ($25)
  • Customers holding their money ransom "I'd pay if you just implemented this one thing!"
  • Do expensive apps weed out the assholes?
  • Cascable's customers, and how on average I've never had a nicer bunch of people annoyed with my bugs. Apart from that one Fuji Testflight guy.
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment