4 ms·

Friendster: http://highscalability.com/blog/2007/11/13/friendster-lost-lead-because-of-a-failure-to-scale.html http://highscalability.com/blog/2007/11/13/frien

by chaz 16y ago

Friendster: http://highscalability.com/blog/2007/11/13/friendster-lost-lead-because-of-a-failure-to-scale.html http://highscalability.com/blog/2007/11/13/friendster-lost-l... VB: Can you tell me a bit about what you learned in your time at Friendster? JS: For me, it basically came down to failed execution on the technology side — we had millions of Friendster members begging us to get the site working faster so they could log in and spend hours social networking with their friends. I remember coming in to the office for months reading thousands of customer service emails telling us that if we didn’t get our site working better soon, they’d be ‘forced to join’ a new social networking site that had just launched called MySpace…the rest is history. To be fair to Friendster’s technology team at the time, they were on the forefront of many new scaling and database issues that web sites simply hadn’t had to deal with prior to Friendster. As is often the case, the early pioneer made critical mistakes that enabled later entrants to the market, MySpace, Facebook & Bebo to learn and excel. As a postscript to the story, it’s interesting to note that Kent Lindstrom (CEO of Friendster) and the rest of the team have done an outstanding job righting that ship.

lallysingh 16y ago
To be fair to the rest of the industry, a single glance at the front page shows size of the TRANSITIVE CLOSURE of your social network. Why would you do that? This is bread & butter algorithm analysis. Don't put O(n^2) analyses on your most-loaded page.
- shalinmangar 16y ago
  You are assuming that it is computed on page load.
- spitfire 16y ago
  I actually respect that. While facebook brags about their "social graph" (IE: A list of friends they don't do anything with), friendster tried to take on the hard CS problems. Sure they failed, but they were never fake about it. And they were one of the first.
- aamar 16y ago
  I have heard from insiders that the founder viewed this feature--the count of people 3 degrees away--as the central embodiment of the magic of Friendster, an absolute must-have. I.e. people added friends in great part in order to feel more connected to 100ks of people. Others in the company begged to at least back it down to a count of people 2 degrees away; but the founder insisted that the magic was the 3-degree version. So instead the company pursued technical strategies to speed up the 3-degree count; don't know what those were precisely, but it seems that they were not pursued as zealously as they could be (due to VoIP, etc. distractions). My understanding, by the way, is that the network size was computed on page-load until surprisingly very late, due to the perceived need for real-time. Even after it was cached, it was still computationally expensive, as your numbers were computed (roughly) every time your 3-degree network added a link. In retrospect, A/B testing could have been used to test the executive vision. So although my first reaction upon hearing this story (years ago) was: "that's big-O insanity," now I think that it's just as much a story about willingness to subject vision to empirical data and performing clinical cost/benefit analysis (when appropriate).
- kragen 16y ago
  Because Sixdegrees did that. In fact, it was the entire marketing point of Sixdegrees. I don't think Friendster was trying to imitate Sixdegrees generally — after all, it was a failed bubble business — but unavoidably you tend to model what you're building after things that you've seen before.
silverlake 16y ago
From http://www.nytimes.com/2006/10/15/business/yourmoney/15friend.html?_r=1 http://www.nytimes.com/2006/10/15/business/yourmoney/15frien... But the board also lost sight of the task at hand, according to Kent Lindstrom, an early investor in Friendster and one of its first employees. As Friendster became more popular, its overwhelmed Web site became slower. Things would become so bad that a Friendster Web page took as long as 40 seconds to download. Yet, from where Mr. Lindstrom sat, technical difficulties proved too pedestrian for a board of this pedigree. The performance problems would come up, but the board devoted most of its time to talking about potential competitors and new features, such as the possibility of adding Internet phone services, or so-called voice over Internet protocol, or VoIP, to the site. The stars would never sit back and say, ‘We really have to make this thing work,’ ” recalled Mr. Lindstrom, who is now president of Friendster. “They were talking about the next thing. Voice over Internet. Making Friendster work in different languages. Potential big advertising deals. Yet we didn’t solve the first basic problem: our site didn’t work.”
- prostoalex 16y ago
  It's unclear why that should be a board-level discussion in the first place. I doubt board meetings at Amazon or Google involve people benchmarking performance.
  - kragen 16y ago
    Because it was killing the company. The reason board meetings at Amazon and Google don't involve people benchmarking performance is that the sites are competently run. But anything that is putting the company's future at risk is a legitimate question for the board to talk about.

by chaz 16y ago

lallysingh 16y ago

To be fair to the rest of the industry, a single glance at the front page shows size of the TRANSITIVE CLOSURE of your social network. Why would you do that? This is bread & butter algorithm analysis. Don't put O(n^2) analyses on your most-loaded page.

shalinmangar 16y ago
You are assuming that it is computed on page load.
spitfire 16y ago
I actually respect that. While facebook brags about their "social graph" (IE: A list of friends they don't do anything with), friendster tried to take on the hard CS problems. Sure they failed, but they were never fake about it. And they were one of the first.
aamar 16y ago
I have heard from insiders that the founder viewed this feature--the count of people 3 degrees away--as the central embodiment of the magic of Friendster, an absolute must-have. I.e. people added friends in great part in order to feel more connected to 100ks of people. Others in the company begged to at least back it down to a count of people 2 degrees away; but the founder insisted that the magic was the 3-degree version. So instead the company pursued technical strategies to speed up the 3-degree count; don't know what those were precisely, but it seems that they were not pursued as zealously as they could be (due to VoIP, etc. distractions). My understanding, by the way, is that the network size was computed on page-load until surprisingly very late, due to the perceived need for real-time. Even after it was cached, it was still computationally expensive, as your numbers were computed (roughly) every time your 3-degree network added a link. In retrospect, A/B testing could have been used to test the executive vision. So although my first reaction upon hearing this story (years ago) was: "that's big-O insanity," now I think that it's just as much a story about willingness to subject vision to empirical data and performing clinical cost/benefit analysis (when appropriate).
kragen 16y ago
Because Sixdegrees did that. In fact, it was the entire marketing point of Sixdegrees. I don't think Friendster was trying to imitate Sixdegrees generally — after all, it was a failed bubble business — but unavoidably you tend to model what you're building after things that you've seen before.

silverlake 16y ago

From http://www.nytimes.com/2006/10/15/business/yourmoney/15friend.html?_r=1 http://www.nytimes.com/2006/10/15/business/yourmoney/15frien... But the board also lost sight of the task at hand, according to Kent Lindstrom, an early investor in Friendster and one of its first employees. As Friendster became more popular, its overwhelmed Web site became slower. Things would become so bad that a Friendster Web page took as long as 40 seconds to download. Yet, from where Mr. Lindstrom sat, technical difficulties proved too pedestrian for a board of this pedigree. The performance problems would come up, but the board devoted most of its time to talking about potential competitors and new features, such as the possibility of adding Internet phone services, or so-called voice over Internet protocol, or VoIP, to the site. The stars would never sit back and say, ‘We really have to make this thing work,’ ” recalled Mr. Lindstrom, who is now president of Friendster. “They were talking about the next thing. Voice over Internet. Making Friendster work in different languages. Potential big advertising deals. Yet we didn’t solve the first basic problem: our site didn’t work.”

prostoalex 16y ago
It's unclear why that should be a board-level discussion in the first place. I doubt board meetings at Amazon or Google involve people benchmarking performance.
- kragen 16y ago
  Because it was killing the company. The reason board meetings at Amazon and Google don't involve people benchmarking performance is that the sites are competently run. But anything that is putting the company's future at risk is a legitimate question for the board to talk about.