3 ms·
Classic failed argument. Service isn't implementing features that are visible to you, so you argue that they aren't doing anything. Sorry, designing highly scal
by timdorr 18y ago
Classic failed argument. Service isn't implementing features that are visible to you, so you argue that they aren't doing anything. Sorry, designing highly scalable systems isn't always about whizbang features. But that's the thing: The whizbang feature is that it all works and stays online, despite their 700% growth.
Another thing to consider is that with a site of this site and activity level, any minor change has to be mulled over quite a bit before implementing. There's question of whether or not people will accept it and of how it will be implemented at scale. You can't always take the Facebook approach and upheave your entire website every few months.
And complaining about searching not being on the home page? 2 seconds in Google gets you: http://blog.twitter.com/2009/02/testing-more-integrated-search.html http://blog.twitter.com/2009/02/testing-more-integrated-sear...
- thorax 18y agoAnd they made an open enough service that whizbang features are coming out of the woodwork from third parties. Like 10 new Twitter apps a week or something, I bet.
- swombat 18y agoTwitter was founded in 2006. According to Twitdom there are 554 twitter applications out there (no doubt it's missing a few). Even assuming that they came out linearly from day one, that would be about 1 application every couple of days. Making more realistic but less provable assumptions, I would assume that Twitdom doesn't know about all the apps - let's say they know about 50% of them. And I would imagine that at least 80% of these apps have come out in the last two years rather than in Twitter's first year. That would mean about 800 apps in 700 days - so just over 1 app a day. Which is about 10 apps a week. Good guess.