What a long Strange Trip it has been.

Talks

Ezra Zygmuntowicz

@ezmobius

#ruby-on-rails

#deployment

#software-development

What a long Strange Trip it has been.

by Ezra Zygmuntowicz

In his talk "What a Long Strange Trip it Has Been," Ezra Zygmuntowicz recounts his journey with Ruby on Rails from its inception in 2004 to the present day, discussing the evolution of the framework and its significance in web development.

Early Days: He began as a webmaster at the Yakima Herald Republic during a challenging time, where he inherited a problematic PHP codebase. Ezra found Rails shortly after its launch, switching from Python to Rails, which became the foundation for various applications, including the newspaper's website.
Deployment Challenges: Ezra explains the complexities of deploying Rails applications at a time when deployment methods were not well-established. He became known as a deployment expert in the Rails community as he tackled various server configurations to handle dynamic hits efficiently.
Formation of Engine Yard: In 2006, alongside Tom Morini and Lance, he co-founded Engine Yard to address the hosting needs of Rails applications at a time when Rails was gaining traction among developers. During a period of intense growth, he managed support and deployment for numerous clients.
Transition to Cloud: By 2009, Ezra transitioned Engine Yard's services onto Amazon Web Services (AWS), signalling a pivotal shift in cloud computing for Ruby on Rails applications, leading to increased scalability and efficiency.
Innovations in 3D Printing: He shares his latest venture, Trinity Labs, focusing on advancing 3D printing technologies, including building his own portable 3D printer. Ezra draws parallels between the early days of Rails and the nascent stage of 3D printing, emphasizing the potential for rapid change in manufacturing.
Future Directions: Ezra discusses his vision for the future of manufacturing, advocating for a decentralized model enabled by 3D printing that allows for micro-customization and on-demand production.

In conclusion, Ezra emphasizes the importance of adapting to changes in technology, reflecting on his own evolution as a developer and entrepreneur. He aims to inspire the community by showcasing how passion and ingenuity can lead to groundbreaking advancements in both software and manufacturing.

00:00:25.519 My name is Ezra Zygmuntowicz, and I'm going to give a bit of a different talk today. It's not super technical; it's more of a story.

00:00:33.120 I don’t know how to put it—it's the story of my life on Rails since it first came out. It's been a long and fun journey since 2004 when Rails first hit the scene.

00:00:46.520 In 2004, I was the webmaster at a little newspaper in Eastern Washington called the Yakima Herald Republic. I was tasked with rewriting their entire website, which was a big, messy PHP kind of soup.

00:01:03.840 Funny story: the previous webmaster had obfuscated his PHP code before he left. He changed all the variables to x, y, and z, and then tried to hold the newspaper ransom for five grand to reveal how to change it back. My first week there, I actually got to do a sting operation to catch that guy.

00:01:19.400 So here I was, the only tech person at this newspaper, with the website falling apart. They said I needed to rewrite the whole site, plus our classified and obituary systems. I only knew a little PHP and wasn’t really a programmer yet, so I grabbed some books on Python and Ruby.

00:01:48.399 It was a choice between Python and Ruby, and I spent a few months researching. Just as I was about to write it all in Python, this guy, David Heinemeier Hansson (DHH), released Rails in August 2004. It was the perfect framework for the newspaper's needs. I threw out my Python book, and that made my decision easy.

00:02:58.480 I think I released the first commercial Ruby app for a newspaper in the fall of 2004. I ended up getting branded as the 'deployment guy' in the Rails community because, back then, nobody knew how to deploy Rails applications, not even the 37signals guys.

00:03:10.159 We were lucky if we had a few scripts that could restart our processes 500 times a day instead of a thousand. The newspaper provided me with a couple of Linux boxes and an Xserve, and I had to manage 400,000 dynamic hits a day from day one without the site crashing. This was Rails in 2004. I set off to figure out what I could do about it.

00:03:42.000 It was chaotic trying to figure out web servers—Lighttpd, Apache, FCGI, SCGI—and I conducted a big research effort on all those technologies. I managed to transform the deployment process without any problems. That first site still runs and I think it’s still on Rails 0.6, and it works just fine.

00:04:27.240 Anyway, moving on, deploying applications became my thing. If you look back at the Ruby on Rails mailing list from those years, you’ll find thousands of posts from me as I navigated through my learning process, trying to help others.

00:04:45.720 The Pragmatic Programmers asked me to write a book on deployment because there weren't any. I accepted, which turned out to be a bad decision. Writing books is hard, and it takes a long time. I started writing the book and managed to finish it three years later with the help of a couple of co-authors.

00:05:06.919 This delay was mainly because, in 2005, Rails was exploding onto the scene while I was busy trying to write the book and fix various deployment issues. Things were getting better, but everything still sucked back then.

00:05:42.080 In 2006, I received a call from Tom Moen, who is still the CTO of Engine Yard, along with Lance, who helped co-found Engine Yard with me. They said that the deployment situation was terrible, but that Ruby on Rails was really cool, and suggested starting a company to address that. We figured that starting a Rackspace for Rails would be a great idea.

00:06:22.200 So, we brainstormed this company as Rails began gaining significant traction. It finally passed the fearful stage of ‘First they laugh at you’ and began to mature, with developers writing business-critical applications. There were now larger applications costing tens of thousands of dollars to develop, and people needed a reliable place to host them.

00:07:30.440 I was the sole support person for Engine Yard and also managed deployments for everyone’s apps, including our first 150 customers. The first RailsConf was in 2006, where I gave a talk on deployment and launched Engine Yard.

00:08:32.000 From that fall in 2006 until the end of 2007, I rarely left my house. Working remotely, I lived in my office, virtually full-time, since I was the support guy for all our early customers. I barely got a couple of hours of sleep as I bounced back and forth helping people. That year was when I started getting gray hair, so it was quite an adjustment from nothing to having gray sideburns!

00:09:48.640 Building Engine Yard predated the emergence of services like Amazon EC2. We had to create something that looked like EC2 from a distance, but without the cloud infrastructure that emerged later which really doesn’t mean much at all. In 2007, Amazon launched its services, which was a game-changer. Had our timing been different, we may not have done our hardware business.

00:10:48.240 In 2007, we took a venture capital round, which was another significant change for me, considering my background—I didn’t even graduate high school, just earned my GED. I learned a lot through trial and error.

00:11:32.640 We all moved to San Francisco, the heart of Silicon Valley, and we even moved next door to Twitter on South Park during the early days of the cloud era. In 2008, I started working on a little web framework called MB, which was more of an experiment.

00:12:16.000 The Ruby community tends to generate rivalry out of anything remotely similar, even if there isn’t any real competition. At the end of 2008, we merged projects, in a sense; I handed over MB to DHH and moved on.

00:12:49.839 By the end of 2008, AWS matured significantly, transforming its ephemeral services into robust solutions. I took some time away in a cave to rethink our strategy for Engine Yard's infrastructure and ended up migrating everything to Amazon, marking a pivotal moment for us.

00:13:37.000 In 2010, I left Engine Yard, feeling I had grown as much as I could there. They were moving into a substantial growth phase, and I shifted my focus to another project at VMware. I helped architect and build Cloud Foundry, an open-source platform-as-a-service framework written in Ruby.

00:14:09.480 As I dive deeper into my own journey with Ruby in the cloud, I've found myself wanting to explore other hobbies. I experience this 'itch' every ten or eleven years, where I crave new challenges. From glassblowing to programming in Ruby, it seems each transition leads me to something new.

00:15:01.440 This year, I’m launching a new company called Trinity Labs, focused on building 3D printers, laser cutters, and software that integrates with them. I’m excited to engage with the DIY and maker communities, many of whom still tinker with these technologies in their garages.

00:15:55.000 I've built several 3D printers and realized that the market is ripe for someone to make this technology easy to use, similar to what Ruby on Rails did for web development. I see programming and atoms as the new frontier, and I’m still going to be a programmer but now I’m just fiddling with physical objects.

00:16:13.640 I brought some pictures of our printers that we're going to sell kits for today. It's fascinating to see what these printers can create, and I personally wanted to showcase one of my projects—a portable 3D printer I designed in a James Bond-style briefcase.

00:16:53.000 I attempted to carry the briefcase onto my flight yesterday, wanting to demonstrate it on stage, but the TSA didn't approve. It’s a compact device that runs on batteries and can be taken anywhere. My intention was to print live during the talk, which would have been a fun demonstration.

00:17:51.800 I aim to pique the interest of fellow Ruby community members and stir excitement about this emerging technology. I believe 3D printing has the potential to revolutionize our current manufacturing systems, allowing for localized, customized production.

00:18:09.200 This could shift our reliance on centralized manufacturing—typically produced in factories overseas—into a just-in-time model, where products are printed as needed and tailored specifically to consumers.

00:18:54.600 For instance, imagine someone wanting to create a customized toy for their child, tailored just for them—with a personal touch. This adaptation can take place right at a local shop, altering the dynamics of distribution.

00:19:24.200 I envision a future where hobby shops can turn into small-scale fabrication plants, equipped with 3D printers that locally produce items instead of needing vast inventories. The scope of potential applications is incredibly exciting and may disrupt traditional manufacturing.

00:20:13.920 Currently, these printers can produce items using biodegradable plastic made from corn, which is safe and recyclable. Printing won't just generate waste; it will contribute to a circular economy where materials can be reused.

00:21:14.720 We're launching the printers soon, and the initial entry-level cost will be around $1,200 for a DIY kit. Our designs offer a range of products, including more accessible and affordable options.

00:22:13.000 Much like Rails in its early days, the 3D printing technology we have is groundbreaking, but the current software is cumbersome. Our mission is to simplify the user experience, making it more intuitive and seamless for people to create.

00:23:38.480 We intend to produce an all-in-one solution where users can design, print, and customize without juggling multiple applications. As we innovate, we’re considering possibilities for add-ins to offer services where models can be uploaded for printing, akin to existing models like Shapeways.

00:24:53.680 But our approach is about integrating a smarter model that empowers users. Managing libraries of designs and supporting user-generated content will be key to broadening the 3D printing ecosystem.

00:26:02.920 As we continue to develop software and design, we’ll focus on user-friendly options to promote widespread adoption and ensure users feel supported. We believe the future of 3D printing will allow people to repair, customize, and create their unique products easily.

00:27:22.080 Our current technology allows printing with two types of plastics, including a biodegradable option that smells like popcorn during production. This unique approach will help eliminate concerns about plastic waste.

00:28:41.840 Our printers are designed to offer versatility, allowing for projects ranging from hobby applications to professional use. The development pipeline also includes exploring materials for specific needs, including wax for casting and other innovative possibilities.

00:29:46.600 User-defined supports are built into our software, allowing automation for optimal printing results without manual intervention. The technology in play ensures high-resolution output while simplifying complex operations.

00:30:33.000 We advocate a transparent process where users can follow along as machines print their components. We prioritize open-source applications so that everyone can benefit from community enhancements.

00:31:27.040 Finally, the quality of printed parts is crucial, with current materials delivering durability and strength. Our vision includes easy access to upgrade pathways for replacing components, ensuring longevity and adaptability in a fast-changing technology landscape.

00:32:19.999 At this juncture, I want to thank everyone for listening and exploring this exciting journey in 3D printing and its parallels with the growth of Rails.

00:32:36.640 If there are any questions, I’d be glad to answer!

00:32:52.560 Yes, we're looking to expand our materials for various applications beyond just plastics.

00:33:13.640 We're developing new materials over the summer that will be available soon.

00:33:29.200 I apologize for not bringing more visual aids, but I have plenty of photos and can share videos!

00:33:43.360 The entry-level price for our 3D printer kits will start at around $1,200, designed to be assembled in just a few hours.

00:34:06.280 Once again, the technology dimensions echo what we've seen with Rails. There is huge potential for this to transition from a messy process into something truly remarkable.

00:34:53.080 Thank you, everyone, for your time!

RailsConf 2012