3 ms·
Twitter must have been misleading when they communicated the reasons for this change since they did not exceed 2^53-1, nor do they expect to exceed this in the
by gilgoomesh 9y ago
Twitter must have been misleading when they communicated the reasons for this change since they did not exceed 2^53-1, nor do they expect to exceed this in the near future.
From a (former) Twitter dev:
> Given the current allocation rate, they'll probably never overflow Javascript's precision nor get anywhere near the 64-bit integer space.
https://twittercommunity.com/t/discussion-for-moving-to-64-bit-twitter-user-ids/9890/2 https://twittercommunity.com/t/discussion-for-moving-to-64-b...
- syncsynchalt 9y agoYour link discusses 2^64, which applies to languages that have native integer types. The 2^53 problem was for Javascript, which has no native integer type, and is thus limited by the mantissa size of Number (which is defined as an IEEE double-precision float). Twitter ids are unsigned 64-bit, since they're generated using Snowflake. That link must pre-date the move to snowflake ids, and is speaking to the count of tweets instead.
- MikeHolman 9y agoI'd be incredibly surprised if they overflowed 9 quadrillion tweets. That's like a million tweets per person on earth.
- quakeguy 9y agoIts the retweets, dude. /s On a more serious side, that number won't be reached anytime soon..
- syncsynchalt 9y agoHaha, sorry, that thought was in reference to the link saying they hadn't overflowed 2^31 yet. Two billion seems believable for tweets as of 2013.
- madcaptenor 9y agoIn the dystopian future where everyone's IoT devices communicate by tweeting at each other this might become a real issue.