- May
- 02
- 2008
Twitter is not abandoning Rails
By: Robert Dempsey | Tags:TechCrunch posted a rumor yesterday that Twitter was going to ditch Rails for PHP or something else. This is not true. Evan Williams tweeted that this was indeed false:
“FWIW: Twitter currently has no plans to abandon RoR. Lots of our code is not in RoR, already, though. Maybe that’s why people are confused.”
I loved David’s response to the scaling argument in a recent eWeek article:
“This is known as the ‘Last Stance’ defense. When you have nothing of left of substance to argue with, you draw the ‘But does it scale?’ card. This is on page one of the Fear, Uncertainty and Doubt playbook.”
Wikipedia defines scalability as, “ability to either handle growing amounts of work in a graceful manner, or to be readily enlarged.” We can do that with Ruby and Rails, just as you can do that with many other programming languages and frameworks. People seem to forget that scalability is not simply a matter of code but the architecture of the application and the infrastructure the app is sitting on. If you design an application without growth in mind you’ll have issues, hands down. If you host your application on shared hosting and don’t give it the resources it needs you’re going to have issues.
“Can Rails scale?” is the wrong question to be asking. The right question is “how do we scale properly?”
Share this post
Other posts you may enjoy
You can leave a response, or trackback from your own site.
2 Responses to “Twitter is not abandoning Rails”
On May 2nd, 2008 at 12:40 pm Louis F. Springer said:
Well said.
Making scalable and highly available services has nothing to do with the programming language, at least by itself. This area would generally be the *last* place I’d look, not the first thing I’d bash. There *may* be implementation issues with a language, but its impossible to determine based on what I have heard so far how many, if any, of the issues Twitter is having have anything to do with Ruby or Rails.
For example, the 500 errors we saw yesterday were caused by the need to restore a crashed database instance. This doesn’t sound like a Ruby or Rails issue to me.
Architecture and designing services for scaling and availability is about physics. The discipline has a lot more to do with things like how you go about identifying single points of failure, looking for bottlenecks, and tying to understand the implications of maintaining state than anything to do with Ruby or the Rails framework.
On May 8th, 2008 at 3:24 pm Dan Keldsen said:
The “does it scale?” question is similar to the “what’s the ROI?” question.
Both are what ask do when they want to just be able to shoot a project down. Funny how their own pet projects don’t get this sort of nonsensical scrutiny.
It’s not that the questions are unimportant, but they focus on such a narrow aspect and leave truly important discussions and strategy stuck in a corner.