##Left-off
To-do
- Highlight the 'syntax' better? It wasn't really that useful for Sarah
- The A + B stuff, not sure if it's relatable enough
- The examples are better anyway
##Feature Ideas
Notes & Thoughts
- I should present even and odd stuff better
- Right now it's in nth-of-type, but would work better in nth-child maybe?
- :nth-child(even) is a thing
- Divide the levels into groups?
- Is it daunting? Too many levels?
----- ^ Prioritize this stuff later ------------------------------------------------
Social & Engagemnt
- Email signup via mailchimp
- Do more work to figure out who has already been there, treat them a bit differently?
- "Welcome back et."
Gameplay
- Award stars when a level is completed without commas and using the indicated selector
- This would be in addition to the checkmark we already do
- selector: "plate:nth-of-type(2n+3)"
- Make this level use apples instead of plates? Seems too big..
- Riddle mode
- Difficult selector problems
- Add a new one every week
- Gives me a reason to bring people back
Fun sfuff
- Some kind of bonus level
- Appears randomly and not very often
- Can have different food items
- Or a special riddle that they get two shots on
Aesthetic Cleanup
- Color the hover border according to the element hovered and context
- Apples have a red border...
Branding
- Pull jquery from CDN
- Figure out CNAME setup for github
- Would be cool for it to always show up as cssdiner.com
Questions
- Is it possible to track where & how the diner is being shared?
- Sharetally, or something similar?
Tracking where people have trouble
- Figure out how best to track which levels are the hardest
- Current proposed method is by tracking incorrect guess counts per level
ga('send', {
hitType: 'event',
eventCategory: 'guess',
eventAction: 'incorrect',
eventLabel: 3, // Level Number
eventValue: 10 // Number of incorrect guesses
});
- How will this show up in Analytics?
- Will it be useful?
- Can I look up the average eventValue per eventLabel
- When should we track an incorrect guess..
- When the selector isn't blank
- When the attempted selector isn't the exact same one tried before...
- Don't want to count 10 attempts if someone just hits enter 10 times in a row
- Could just keep track of atttempted selectors and make sure it's unique