Hi Andreas, As some one whose worked on sites that where ranking around 20,000 at times, I have to agree with Darren. Aside from that my previous Merb recommendation stands. If you really want to talk mad scaling, then you're in the right place =-) Aimonetti was, until recently, the primary developer on a Merb site doing more traffic than any other Merb site in production. I've got a background doing millions of hits per day on Rails apps. These are do-able things. If I was looking to "go big" with a site, I would be seriously thinking Merb and DM right now. Such things often require the bleeding edge of any technology.
From the performance respect I would love to see you're benchmarks published. Not because I don't believe your assertions, but because it would be nice to have clena reliable comparison to mark our improvements against. Most of the people contacting me lately aren't interested in a technology comparison... they want Rails or Merb and they know it, so I don't have time right now to do the comparisions my self. Let us know what you pick, why, and how it goes? =-) Rob PS: Sorry if I sounded a little grumpy yesterday... I had just spend over an hour last week debating the "Rails scales" thing with a Java guy who wanted to base all of his assertions off of "theoretically"s and "what if the devs don't write good code" senarios. On Mon, Sep 1, 2008 at 8:48 PM, Andreas Kirn <[EMAIL PROTECTED]> wrote: > > Rob Kaufman wrote: > > > > > > On Sun, Aug 31, 2008 at 1:21 PM, Andreas Kirn <[EMAIL PROTECTED] > > <mailto:[EMAIL PROTECTED]>> wrote: > > > > > > Rails is great for building a simple "dynamic" website or a quick > > proof-of-concept app, but let's face it, it's too slow and rigid for > > large web applications. At least in my experience. > > > > > > Have to disagree with you here... it's not only something to face, but > > it is something to forcefully stamp out as 'not true'. Sure you can > > write a crappy Rails app, but you can also write a crappy <place > > lanuage or framework of choice here> app. Understanding the Ruby > > framework and learing how Rails is built on top of it allow you to get > > the speed and scalablity out of the framework. This is universally > > true of all computing languages and frameworks. > > In every test I've run so far Rails was more resource-hungry and slower > than similar apps written in PHP and Python, and not by a small margin. > Wonderful and easy to use as it is, ActiveRecord is too slow, > inefficient, and restrictive to serve as an enterprise-level ORM (in my > personal and highly subjective opinion). I'd like to use DataMapper, > but Rails doesn't make that easy. Of course one can cache the bejesus > out of an app or throw more hardware at it or try to rewrite or extend > large portions of the framework, but that's not what I'm getting paid to > do. While it may be fun to hack the hell out of Rails in my free time, > right now I need a tool I can use without major modifications. > > But I don't want this to turn into a religious debate about Rails, so > let me rephrase the question: If one were looking for a Ruby framework > to build a large, high-traffic application, which of the previously > listed frameworks would you recommend and why? > > Btw, do you know of any high-traffic social or e-commerce apps built on > Rails, other than Twitter? > > Andreas > > > > --~--~---------~--~----~------------~-------~--~----~ SD Ruby mailing list [email protected] http://groups.google.com/group/sdruby -~----------~----~----~----~------~----~------~--~---
