That's one of a series of posts Mr. Arrington made. They also fit in a general trend of manufacturing a dramatic store with no source and no supporting evidence.
Blaine's slide was perhaps poorly chosen, but it's also entirely correct: scaling the rails portion of twitter is trivial, just as it would be with php, .net, or countless other technologies. The stateless tier isn't interesting. But for whatever reason, the meme was never "mysql can't scale at twitter", but rather "rails can't scale at twitter", even though the app server stack had little to do with it.
Also, those of us who've been through the pain of a rapidly growing startup, even if it wasn't growing as fast as twitter, understand that the real challenges are people, politics and vendors. Rarely is there a deep technological problem that requires a novel solution. Typically the fight is over getting the things you need to implement a well known solution.
In particular I had a few calls with early twitter folks at the time and know first hand that the people involved were not clueless. They were quite humble and helpful in providing frank discussions of vendors and technologies we were considering.
Just because it's a blog doesn't mean it's a new medium or that bloggers shouldn't expect to be challenged on the accuracy of their claims. I don't see any reason that the expectations I'd have of reasoned discourse in person, by phone, video or print don't also apply to the techcrunch blog.
Yes it is hard, but Twitter was down a lot. Was Blaine fired? I met netik at a conference just before he was hired by Twitter - great guy, knows his shit.
I am not saying that they shouldn't be challenged on the accuracy of their claims, I am saying that they are learning as they go. That post was a bit personal, the next one won't be.
Judging by the hit rate outlined in the post above Arrington seems to have learnt a lot about how to handle sources, stories etc. But ignoring him completely based on a few examples from years ago seems wrong.
By your reckoning I seriously would have nothing left to read atm (especially the NYTimes) - I still read it all though because I take everything at face value.
I am not going to exclude myself from an entire slice of online media just because 2,3,4 or 5 or 10 blog posts out of thousands that provide me so much information and value.
Blaine's slide was perhaps poorly chosen, but it's also entirely correct: scaling the rails portion of twitter is trivial, just as it would be with php, .net, or countless other technologies. The stateless tier isn't interesting. But for whatever reason, the meme was never "mysql can't scale at twitter", but rather "rails can't scale at twitter", even though the app server stack had little to do with it.
Also, those of us who've been through the pain of a rapidly growing startup, even if it wasn't growing as fast as twitter, understand that the real challenges are people, politics and vendors. Rarely is there a deep technological problem that requires a novel solution. Typically the fight is over getting the things you need to implement a well known solution.
In particular I had a few calls with early twitter folks at the time and know first hand that the people involved were not clueless. They were quite humble and helpful in providing frank discussions of vendors and technologies we were considering.
Just because it's a blog doesn't mean it's a new medium or that bloggers shouldn't expect to be challenged on the accuracy of their claims. I don't see any reason that the expectations I'd have of reasoned discourse in person, by phone, video or print don't also apply to the techcrunch blog.