Hacker News Reader

Comment by kevingadd | Hacker News Reader

Not to mention that RTs are trivially reversible: You can undo them in the UI, and all tweets can be deleted.

That's true. But my service doesn't detect the retweet right away. So it has to stay retweeted for some (currently about 1 minute) period of time. I can tweak this parameter in future if I see that users misuse the service and undo retweets after receiving a code.

Perhaps at minimum write code to monitor tweets say 24h later so you can measure this.

Surely after a few hours it doesn't really matter if the retweet is removed. Sure, it won't appear on their history but the recent timeline is the biggest use case.

Perhaps send the code after say 15-30 mins, or base the time on their follower count.

Edit: Does this violate any Apple or Twitter ToS? I've a feeling Apple won't like this, but they probably won't care.

That's what I thought, only most recent timeline counts. From my experience anything older than 15-30 minutes is not noticed anymore. So I agree, that's a good amount of time to use as the grace period. Just intuitively I think people wouldn't care to remove their retweets. I might be wrong though. Need to get more data on this. I read Twitter ToS pretty carefully, and I didn't find anything that my service would violate. As for Apple, still need to look into that.. It is something I see all the time though. And I do it for my own apps without any issues so far.

Reply on news.ycombinator.com