As you can see Barak Obama is a kitten killer in your demo.
It's pretty fun for users who are using it for its intended purpose, and not very trollable if you use the default URL randomization.
Also bear in mind I made this in a day. It might be neat to add sentiment analysis / negative word flagging in order to weed out all-troll submissions.
I think go to facebook and say "Hey guys! Let me know what you think! [my link]". Now pseudo_friend21 clicks the link and see's they can say whatever they want.
"You're fat and stupid, go die in a fire" - dead_mom_1991
"Take your mom's advice, please!" - boyfriends_sister13
"Hey guys, don't be so hard on her... she IS mentally retarded after all! lol" - long_time_friend19
So now a single user has launched 3 personal attacks that will cause this person a good bit of personal anguish when they stir up those emotions.
1. your mom is dead, let's be reminded of that
2. let's start a fight with your love interest, because his sister is mean
3. I haven't talked to that friend in a while... but NOW I know how they fell about me.
This is just the tip of the iceberg, if pseudo_friend21 sends the link posted to mortal_enemy18 then the proverbial shit will hit the proverbial fan.
edit: good rule to remember when building systems like this. http://www.penny-arcade.com/comic/2004/03/19
This seems like a great fit to use facebook authentication. But you could just post on their facebook wall. Who wants statistics on their praises anyway?
- 25-70 hits for strings containing words like nice/fun/smart/kind
- 1-3 for dumb/ugly/stupid/boring/fag
- no incidences of slut/bitch/hate/crap/... (I searched a LOT)
- one racial slur that I could find
- surprisingly no negative uses of curse words (one "fucking awesome")
This is a way lower incidence of negativity than anticipated. Due to the simplicity of the text inputs, I'm going to hold off on preemptive moderation that would change the feel of the app and see how it works if I just flag blacklisted words.
Some possibilities that come to mind, maybe using a combination of two or more of these:
- Require FB auth. Remove anonymity (to an extent) from the equation.
- Launch with a canned, relatively small number of broad compliments.
- Allow only community approved compliments. If new ones are added, block until they're approved.
- Allow submissions only until the list (without revealing the recipient's name) has been approved by multiple community members. Then lock the list allowing only up votes on existing compliments.
This is an instant dealbreaker for me nowadays. I refuse to have a Facebook account for a number of reasons and know a couple of other people that are the same.
However, requiring FB auth may not make a large impact on the number of users registering for the service and/or users like me may not actually be within the range of users who would be interested in the service.
(Personally, I think it's a really nice idea. A bit of positive feedback can go a long way on some rough days.)
I'm actually more curious about how the site could handle multiple people with the same names, or users who don't really think about it and just put in "John" for the name.
The site has me thinking about aggregating other input info from the internet...
