Bad Design, Good Solutions
Hey (no|every)body,
I wanted to share a particular difficulty I encountered while developing How Long. My original design had a timeline of the ten most adjacently released friends on each user’s profile box. As simple as that may sound, its a source of countless problems.
Problem A: Suppose Alice is not a How Long user, but has 100 How Long user friends. At some point, Alice adds the application and enters her release date. The application then has to check if an update is needed to 100 profiles. Suppose Alice, Bob and Charlie all have 100 friends and all add the application at the same time and you have 1 dead server, 3 frustrated users, and at the very least 100 users with inaccurate profiles. FAIL
Problem B: Suppose Alice was drafted on the 3rd of December 2006 (respect!) and so were 40 of her friends, 20 of whom use How Long. Alice and her 20 friends therefore have a very boring 10 closest friends list and are all quite unsatisfied. FAIL
Problem C: Suppose Alice and Bob are How Long users with the same release date, they are not currently friends. At some point Alice and Bob befriend each other. Facebook does not send out new friend events to user application, and rightfully so, honestly. Therefore, nothing is there to trigger an update to their profiles, and they’re both left with an inaccurate top 10 list. FAIL
Problem D: Alice is a How Long user with 100 How Long user friends. At some point she decides that because How Long has so many failures, she wants to remove the application from her profile. 100 friends need to be updated, the same problems arise as in scenario A. FAIL
There is only one solution to all this. Dump the idea and replace it with something different. I’m thinking of letting users choose 10 friends they want to display. This opens a new path to monetization. 10 friends for free, more for a symbolic price. I highly doubt this will work, seeing as how this application is targeted towards poor Israeli soldiers who make 700NIS ($208 as of today) a month at most, and belong to a population who generally does not like making transactions online. I’m not expecting a profit, I’ll almost surely dump the idea, but if the non-monetized application is a hit, I might just try it as a sort of monetization experiment and/or Spare Change API familiarization.
Wish me luck!