4 ms·
Somehow I find it hard to imagine that even if the new management and technical team had gelled that they would have been able to keep up with Facebook. Their
by dasil003 17y ago
Somehow I find it hard to imagine that even if the new management and technical team had gelled that they would have been able to keep up with Facebook. Their codebase was a flaming hulk of crap to start with, and you can just imagine how much improvement it would get from a biz guy "hiring a team to clean it up". Compared to Facebook which is one of the most amazing engineering organizations of all time.
The only opening would have been if MySpace was somehow smarter about what users wanted, but outside of customizable profiles and music they didn't have any new tricks.
- ngsayjoe 17y agoI think Facebook architecture is bloated as well ... /home.php? come on ...
- dasil003 17y agoIs the critique that they are using PHP or that they aren't using some kind of extensionless routing? Either way it's silly to use superficial factors like that to judge large websites. Given the performance profile, data interconnectedness, complex public API, and interface richness of Facebook, the rate at which they push changes (small changes daily, weekly major pushes) and advance the product is astonishing.
- simonw 17y agoCouldn't agree more. MySpace has always stood out for me as the popular site with the most obviously terrible implementation. Their key "personalisation" feature was originally an accidental security hole! Meanwhile, Facebook were hiring smart engineers and building generally innovative features. I remember the first time I typed a URL in to a Facebook message - a little panel popped up showing me thumbnails that had been scraped from that page, asking me to select one to illustrate the link. It's also interesting to contrast the process side of things. The FT article talks about MySpace having to take lists of feature proposals to News Corp for approval. Meanwhile, Facebook are leading the industry in their adoption of A/B testing and letting their own users' behaviour direct their engineering efforts.