Home > Atmosphere, Websocket > Can I make a living out of my open source projects in 5 months?

Can I make a living out of my open source projects in 5 months?

I’m in the open source world for the last 12 years…it all started in 2002 as a member of the Tomcat team, then in 2005 I have created the Grizzly framework, lead the team that open sourced GlassFish, released the Grizzly Comet Framework, created the Atmosphere Framework in 2007 and its related projects. I also created the AsyncHttpClient library in 2009 and recently designed and implemented the Swagger Socket Protocol in 2012. I think I can say I have a little experience with open source development and management  🙂

All of that work was done under several companies umbrella: Sun Microsystems, Ning, Sonatype and more recently Reverb. In all cases, I was allowed to work on those frameworks because they were used internally. For example, I’ve spent the last 18 months working at Reverb, significantly improving the Atmosphere Framework and even deploying it on Wall Street Journal, one of the biggest/live deployment I’ve ever made so far supporting WebSockets and Internet Explorer 6/7/8/9 at the same time. Phew!!!

During that 18 months at Reverb, I was finally able to release the 1.0 version of Atmosphere, which was a major accomplishment for me after working on the framework for the last 4 years. Then, the popularity of the framework sky rocked and everything jumped: download are almost 200% up every months, Github “watcher” doubled, mailing list is now close to 500 users, Twitter account reached 1100 followers and emails traffic are stable months after months. More important, in December I’ve started getting requests for professionnal services with Atmosphere, contract support, etc.

So, with the current Atmosphere success, I’ve had a hard decision to take: work at Reverb and let the Atmosphere’s community grow and “leave by itself”,  or resign from Reverb and try to build a company to support Atmosphere. Last January was difficult for me as I was constantly under fire by having to work on Atmosphere and at the same time help scaling the Reverb’s platform. Not a nice situation!

So, I’ve decided to tale a leave on absent from Reverb and try to build my own company around the Atmosphere Framework, called Async-IO.org. The website is live and the traffic is already passing 500 unique hit per day, so this is positive…but, It’s a major challenge for the following reasons:

  • I cannot stay unemployed for several months. Hence I allocate myself 5 months to succeed with my crazy decision. I know it’s a very  small amount of time, but that’s all I can allocate.
  • I’m based in Canada (in Québec), not in California or in a city like London or Berlin where I can have funds from Venture Capitals Companies. Montréal is unfortunately stuck in the early 200o where IBM and Microsoft rules the city, so no luck on that side as well!
  • I want to try to do everything all by myself, from designing the website (thanks Github) and learn from existing open source projects that succeeded. I’m event sketching the official logo this week!
  • I don’t want to spend large amount of money on infrastructure. Instead, I would like to reward people that have been working with me in Atmosphere as soon as I can. But I need a solid foundation before I can do that.
  • I don’t have a large corporation financing my work. Since Atmosphere compete with framework like Socket.IO/Node.js, Vert.x and Play 2 (they are all powered by a big company), it’s a real challenge!
  • Atmosphere work with all browsers, mobile browsers and Java WebServers, and from version released sometimes 5 years ago. The testing matrix here is really complex and right now I  strongly need to rely on the community to help, and in return I’ve adopted a short release cycle to make every body happy. That way too much time consuming so I need to improve that situation.
  • Continue being active with the community like I have been for the last 4 years!

Now, after reading this you can call me crazy and I would fully understand! Interested to help? Pretty simple:

So, for the next 5 months I will use this blog to share my experience with this colossal adventure! Stay tuned!

Categories: Atmosphere, Websocket
  1. March 4, 2013 at 8:05 pm

    wow! that’s a difficult situation. Your move was really bold. I used your framework for a few projects and it was very interesting. What about websocket? Isn’t websocket making more general your ideas? and being the next being thing after Atmosphere?

    • March 4, 2013 at 8:34 pm

      Thanks! Atmosphere brings a lot of functionality to WebSockets. It transparently support them and clustering, unify javascript API etc. Much more than what you get with the basic.

  2. March 4, 2013 at 9:23 pm

    Congratulations! Wishing you the best, Jeanfrancois.

  3. Evgeny Kochnev
    March 14, 2013 at 12:25 am



    I am glad to heard that you decided to begin your startup!!!

    Actually I decided to return back to using the Atmosphere from Kaazing JMS Gateway Edition since they have changed their price policy 😦
    They don’t offer monthly subscription anymore 😦 Just the JMS Edition for $24K per year on an annual basis, per Gateway.
    I understand that all of us try to make money but it is too big for me now for my small startup :).

    So, probably I will try to do something like this over Atmosphere. Please, think about – to add additional functionality like this too.

    So, Anyway Congratulations! And Good Luck!!!


    • Evgeny Kochnev
      March 14, 2013 at 12:32 am

      And If it possible – could you add some exactly price for support/training.
      It is very disappointed than there is no any price just “Please contact us!!!” At least me 🙂

  4. March 16, 2013 at 10:37 am

    First of all, congrats on this move. It’s a bold one but a good one, as websocket is such a good technology and big companies are definitely on the lookout for some reference implementation.

    Now if I may share some constructive feedback, I’m sort of new to websockets and I have been trying to figure out for a few days how to use the Atmosphere framework to develop a simple websocket-only (no fallback) server application… and I just can’t. In my humble opinion, if you want Atmosphere to succeed, spread, and eventually be acquired, you have to work on documentation. Right now, it’s a mess. All your tutorials completely bypass the project bootstrap phase and jump right into interface implementations. And all the code samples have so many Maven dependencies that they’re just impossible to run standalone without downloading the internet and spending hours trying to understand how to get it running. I mean my team has been trying out other websocket implementations like Play2, Tomcat and Grizzly/Glassfish and yes, they are not as flexible and omnipotent as Atmosphere seems to be, but they all have straightforward step-by-step tutorials that allowed my guys to get a websocket service running in a couple of hours. I haven’t managed to do that yet with atmosphere. Just my 2 cents, from an open source enthusiast to another.

  5. Steve
    March 18, 2013 at 9:01 pm

    I got to this post by looking for Java/WebSocket libraries and found Atmosphere. I wish you the very best of luck Jeanfrancois!

    You have a lot of experience behind you and deserve to succeed at this. The world needs more small, focused companies. Congratulations on your brave move.

  6. Mete
    September 9, 2014 at 9:29 pm

    Curios how things turned out for you?

  7. January 14, 2015 at 8:08 am

    I conceive you have noted some very interesting details , appreciate it for the post.

  1. No trackbacks yet.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: