Photo: Twitter / @debergalis
Building a startup is hard.But there’s a new trend in startups coming out that make it much easier to make an application.
The latest example is Y Combinator graduate Meteor — which just raised $11.2 million in funding from big-time investors like Andreessen Horowitz. It’s a startup that specialises in speeding up the process of making an app.
Meteor takes away a lot of the technical challenges of building a rich web experience with live updates — like comments on Facebook, or a Twitter feed or collaboration service like Asana — by packaging it into a single piece of software you can drop into any app.
Now a less-technical developer can access those really complicated features that users essentially demand, thanks to Facebook, without having to spend months developing them.
We caught up with co-founder Matt Debergalis to find out why Meteor just received a huge round of funding. Here’s what we learned:
- Startups typically spend a lot of time building protocols and software for their applications, and not the actual applications. For example, a small team might spend months building the underlying architecture for a live commenting system on photos, like what you might find on Facebook.
- Meteor essentially takes away that development time. All these protocols are available through Meteor, so app developers don’t have to spend a ton of time building them from scratch. They can focus on building the actual service.
- After building an open-source community, Meteor plans on building a programming infrastructure for enterprise companies. That’s a pretty typical route with most open source projects.
Here’s a lightly-edited transcript of the interview:
BUSINESS INSIDER: Tell me a little bit about Meteor.
MATT DEBERGALIS: We’re building is a new platform for writing rich client applications, things like the photo-viewer in Facebook or Quora or other sites that actually run inside the browser as a program. That’s a big break form the old model, where you would build a website that ran in a distant web server far away from the user that would deliver a rendered page to the browser. The reason this is happening is the same reason we moved from the mainframes in the 70s and 80s to desktop software on Windows — you can build a richer interface, a more fluid experience for the user, if the software is physically right next to he person using the software.
BI: What kind of pain point are you attacking?
MD: Because the technique is new and because the infrastructure doesn’t exist yet, building those rich web services takes teams of experts months to build today. Only larger companies like Twitter and Quora and Asana, people who are well-funded, can do that. Your typical developer is left between a rock and a hard place because users expect the interface to be fluid and interactive like Facebook. It’s a technique that’s difficult for a developer and small team to master. Meteor is the first off the shelf answer to that problem. It ties together all the techniques and technologies to build the rich applications so the weekend coder and typical developer can focus on the application, not the ni tty gritty
“It ties together all the techniques and technologies to build the rich applications so the weekend coder and typical developer can focus on the application, not the nitty gritty.”
BUSINESS INSIDER: What do you mean by “techniques”?
MD: Here’s the challenge: if someone makes a change on a site like website, how do we get that information on to everyone else’s screen. You might have a million people using that app if it’s successful, everyone is using that information. To write software like that you have to build a bunch of programs that coordinate the traffic, you invest in protocol that allows your client to coordinate with the server you wrote, which is your air traffic controller, and the server decides which data to send to which people. You’re investing a whole layer to connect those things and forced to come up with a solution on how to tell each client when their data is out of date, and keep every single interface up to date.
A good analogy is the relational database. Before Oracle provided a database you could buy, any software you wrote you had to include that functionality by hand. Companies used to hire database hands that would build fragile database systems tied to applications. But once we had this off the shelf SQL technology, every developer could start with that off the shelf instead of focusing on the nitty gritty of figuring out how and when to put information on a disk.
BI: How does this help a startup?
MD: If you want to develop an idea as quickly as you can and you’re a small team, you want to iterate and try it out and see how people that feel. If the beginning of that process is a month of underlying technical work and you don’t have any product to show for it, you’re in a tough spot. Meteor lets a startup iterate their product in days instead of weeks or months. It lets people immediately works with graphic design, with a rich application. Meteor is a technology that lets a graphic designer begin building a complete application. It levels the playing field and it lets a single designer or an artist to be able to construct that next blockbuster app.
BI: Tell me a little bit about your team.
BI: You just raised a huge chunk of cash. What are you going to use it for?
MD: The reason we wanted to work with Andreessen Horowitz and Matrix Partners is that they are experts at taking successful open source technologies into the enterprise. Andreessen Horowitz has made some high-profile commitments to developer-centric technologies, we think they’re a great fit. Marc Andreessen basically invented the web. It’s a firm that I think is patient and understands the value of a vibrant open source community.The first thing we’re gonna do is use this to continue to grow the Meteor project, an open source project that anyone can use for free and modify it how they want. We ant to help the community grow. That means we’re going to do whatever we can do to help people build the meteor project — find jobs, build apps, all these things successful to an open source ecosystem.
In the long term, the funding will help build product that commercial enterprises need so they can integrate Meteor into their existing IT infrastructure. The challenge with the enterprise is you have requirements around how you want to allocate computing resources. Those are products we can develop around the Meteor core that we can sell to the enterprise. The team we’ve put together has done this repeatedly. We have to be patient and focused on the open source project, once you have that, we have to figure out how to take that technology into the enterprise, which has this very different set of interests.
BI: How does this fit into the big push into mobile?
MD: You see applications built for all these platforms at the same time. You may build a native mobile application for iOS or Android, and you’re usually gonna have an HTML5 equivalent. The real problem is how to have a consistent architecture that lets you share data across those two things at the same time and how you do it in a way that saves engineering time. Meteor is a bunch of individual packages that fit together. It’s designed to plug your native mobile application into the exact same server as your HTML5/desktop application and have a consistent standard protocol, so when you make a change on your phone or your native app, it appears on the desktop.