It is now a month since the Arduino Uno meetup at ITP finished, I have been busy as hell (and I still am), but it feels it is time to report back to the community about my share of what happened there. We -the Arduino team- had to meet for our yearly discussion about how to move on with the project. It is just that it was our 5th anniversary, so it felt like we had to make something special. We prepared a selection of talks and topics we felt had to be discussed with everybody and we brought things up to a series of people. We had to chose between making “the Arduino conference” or making something a little more intimate that would gather representatives from the community, the distribution list, the users, teachers, and people developing products on top of our technology.
We tried to get people from all over to NYU (New York University, our host) and spent two days in a series of marathonian sessions about IDE improvements, API enhancements, library compatibility, board redesign ideas, business models, and web 3.0. I am actually really proud of the Spanish speaking community, represented by Igor Gonzalez, from Spain, who gathered everyone’s opinions throughout the forum and brought up important matters that concern the way we should design our new online services to make them more inclusive to non-English speaking people. The meeting closed up with an open day where anyone was welcome to pass by, bring their project along and share with the rest.
I took the chance to prepare things for the talk about the web, I made sure I had a series of good graphs to show, and those are the ones I really itch to show you guys. The web has experienced an explosion in the amount of visitors since we started, and it feels it will only grow. I think I expressed in an earlier post my personal obsession in having 100% uptime for our web services, so that people can rely on us for education, sharing knowledge with others, and document the outcome of their projects. Looking at this statistics, I started to understand that … if we started to work with high-schools and not only with universities, we should address a completely different set of issues like the size (probably we will multiply by 10 our network use in a matter of months), safety, user privacy (you don’t want your kids’ information to be misused on the internet, right?), and of course: COOLNESS FACTOR!
I added this mission to my list of things to study as part of my PhD at Malmo University and scheduled a meeting at Google, thanks to an ITP alumni (Laura thanks a bunch for arranging it!) where I exposed mine and Arduino’s concerns about our future on the web. During a 1 hour presentation and a 1 hour lunch meeting some people at Google NYC and I came to understand a possible model for bringing Arduino to work on their cloud infrastructure. For those that don’t know about it, Google offers scalable infrastructure on Python and Java for free up to a certain amount of hits/month. As for today we are already over their limit, but they are really flexible when it comes to open source projects.
However, the conversation about the web has just started, and will last for some time. I want to understand the implications of migrating from a VPS/dedicated server structure to a cloud one. I have read a lot in the weeks following to the March meeting in NY and it is my understanding the cloud makes sense up to a point. What would happened if we spent e.g. 1 year developing a whole new web infrastructure and all of the sudden we had to migrate to our own series of dedicated servers because of the size we would have reached. This is when I need to talk more about it, I will go back to the Google people, who have been super helpful so far. But I will also analyze others like Amazon who offer a linux transparent cloud that would allow for faster migrations, at least is how I understood their system.