Use your sense of SMELL to diagnose computer errors: the new “smell checker” spell checker is a revolution in error notification!

Background:

In programming, there is the notion of “code smell”—a subtle indication that something is terribly wrong in a piece of source code, but without any (obvious) actual mistake.

For example, if you saw the following:

print("E");
print("RR");
print("OR");
print("!");

instead of

print("ERROR!");

that would be a good indication that something extremely bizarre was going on in a codebase.

The issue:

Unfortunately, in order to notice “code smell,” a person must actively review the source code in question.

Proposal:

But what if code smell could ACTUALLY generate a strange or horrible smell (Figure 1)? Then a person wouldn’t have to actively look for problems—the horrible smell of rotting meat would indicate that there was a problem in the codebase.

This smell-based notification method wouldn’t need to be restricted to programming errors, either: spell checking notifications, software updates, and other information could all be conveyed by smell.

 

1-code-smell.png
Fig. 1: This bizarrely-formatted source code might cause the laptop to emit a boiled-cabbage smell.

Details:

  • A computer could have an incense-burner-like attachment that would allow it to emit various smells.
  • For example, a spellchecking warning could emit the smell of recently-touched copper coins (Figure 2), while “you have 100 unread emails” could emit the smell of curdled milk.
  • This would allow a user to know what items require attention on their computer without even having to turn on the screen!
  • This smell-dispensing attachment could be refilled just like printer ink, making it extremely eco-friendly.
2-smell-check-spell-check.png
Fig. 2: Different warnings and errors could have different smells of various degrees of noticeability and/or unpleasantness. Here, the user might know that they have both a spelling error AND a grammar error by the mix of the spelling-smell (dog that has spent one hour in the rain) and grammar-smell (recently-touched pennies).

PROS: Allows computer errors to be conveyed without requiring the user to actively look at a screen.

CONS: People get used to strange smells fairly quickly, so these smell-based warnings would need to be addressed quickly, before the user adjusted to the smell and stopped noticing it.