Day 2 of Flock was just as packed with great content as Day 1.
The first session of the day delivered for me the most interesting and alarming idea. The speaker was Radoslaw Krowiak, the co-owner of Akademia Programowania in Krakow. He’s been involved in teaching kids proramming since 2013. The age of his students is 5 years and higher. He noted that skills like communication, commercial awareness, ability to work in a team, or problem solving, are amongst the top ten skills missing in college graduates.
He went on to say that math, engineering, and the sciences in general have a reputation of being boring, “too hard,” and “not cool.” The fact that these subjects are not taught in creative environments is not helping either. Furthermore, the sciences are often taught in isolation from each other, not to mention their isolation from the arts. Naturally, just as the world is complex, so are all these areas interconnected and should ideally be seen in a more holistic way.
What I am getting at: Radoslaw presented the results of a creative skills test designed by George Land for NASA. The principle of the test was something like “what can you do with an xyz?” The results were given in percents with 100% being the perfect score. The test was given to kids aged 5, 10, 15 years, and to 1,000,000 adults. The data coming out of this test was this:
Age 5 = 98%
Age 10 = 30%
Age 15 = 12%
1,000,000 adults = 2%
Alarming!!! We are un-learning creativity at school. We are given the answers. We are told to do it on our own. We’re told not to make mistakes. But mistakes are the best way to learn!
That is why students at Akademia learn games. Games are the most cool and non-boring subject for them. They are learning how to make games with programs like Scratch, Lightbot, or Blockly. What are a few great things about learning how to code? Creating apps/games makes the participants understand and use math/physics in practice. It is a much more “cool” way to use Boolean, Cartesian coordinates, Gravity and Kinetics, Conditional clauses, problem-solving, implementation… Plus, they are learning not only scientific concepts, but are also allowed to express themselves in a creative way.
Star Trek in Fedora!
The second (10:00) talk of the day was not less exciting than the Keynote. The title was Kirk, McCoy, and Spock build the future of Fedora and the speaker was Matthew Miller. That topic actually deserves its own article, which I’d like to push out to the Fedora Magazine soon. Essentially, it is an analogy for the Program Logic model originally designed by the Kellogg’s company (yes, the breakfast cereal company). The model is being successfully used by the Modularity initiative, which was the topic of the next presentation.
The complete title of the 11:00 talk was Modularity: Why, where we are, and how to get involved, delivered by Langdon White. Langdon has been actively blogging about Modularity for some time. The discussions about it have been ongoing for the last 3-4 years and we are now finally arriving to concrete bits! The project is aimed at “resolving the issue of diverging (and occasionally conflicting) lifecycles of different “components” within Fedora.” If you’d like to learn more, this Langdon’s short post is an excellent start.
The chain of topics logically continued in the presentation titled Factory 2.0. The easiest way to grasp it is this: “Having Modularity without Factory 2.0 is like running barefoot on pieces of lego.” Factory 2.0 is NOT a single web application, a rewrite of our entire pipeline, a silver bullet, a silver platter, just Modularity, or going to be easy. It is redefining how we deliver things in Fedora. The Factory 2.0 team has identified six problems to be solved for the project to get to its resolution. There is an order in which they need to be solved; Modularity is the last one on the schedule because it depends on the other problems’ solutions. You can look at the slides here!
IoT on Fedora
The IoT on Fedora was an update of a talk the authors gave two years ago. Good news are that Fedora does have the basis for IoT: End point devices (ARM devices), gateways, messaging systems, data gathering, storage and analytics, security… Atomic images are an excellent fit for Fedora IoT endpoints, too!
Since it is “internet of other people’s things,” SELinux is *really* necessary, especially because people rarely realize the security risks in IoT. Overall, the awareness about the subject in the community has grown in the past two years. There is diversity in IoT devices and platforms. Yet still, many issues remain open. For example, we need to figure out how to use Atomic for endpoint/gateway deployments, and answer questions like – How do you update your IoT door lock?
Enterprise Fedora Desktop
At 15:30, I attended the Progress on Enterprise Fedora Desktop. It was a follow-up on a one-year-old presentation about a small effort to produce something that’d make Fedora (and other Linux distros) reasonably usable in corporate environments. The talk gave technical details about how to manage different identities at the same time and how to set up an ‘enterprisey’ data center in your home. The advantages of this system are: you can control your own infrastructure, and improve user experience by reducing the number of password/logon interactions.
Spam in Fedora Wiki
How we took care of spam was the name of the 16:30 session I chose. The author, Patrick Uiterwijk from Red Hat, tackled a recent steep increase in spam occurrence in the Fedora Wiki. His first solution was temporary CLA+1 Wiki and Trac. During that setup, only verified Fedora contributors could log in to the Wiki. After that, he restructured the fight against spam from scratch by introducing Basset. Basset gets messages from Wiki, Trac, FAS, and Pagure, and determines a score based on certain modules, such as spam words. Based on the final score, it either: happily accepts your message, deletes your message and account, or if it’s not sure, sends a message to the admin.
How do we ship it?
The last talk of the day in my schedule was Redefining how we deliver Fedora by Dennis Gilmore. Fedora 24 introduced some significant changes in the delivery process: we switched to Pungi 4, started using Livemedia to create live CDs, started nightly composes of Rawhide and branched… and have a continuous stream of install media that is going to be tested on a regular basis.
What are the future plans? A longer term plan is to completely get rid of Alpha (if we do enough testing and gating, it should work). In Fedora 25, we want to have Workstation ostree, and we’ll be making the Cockpit layered image. It’s the first of layered images; a new deliverable type! It allows you to have more container content
What is after Fedora 25? ASBS (layered images) and Docker are coming in. People should be able to build layered images themselves after F25. We’ll also heavily focus on automation, signed repos in Koji, DVD building in Koji, Modularity, and Alternative architectures.
What happened after this heavily infused marathon? You can guess that we went to infuse ourselves with… other type of content. We did the infusion partially on a boat cruise in the Krakow center, which was really amazing. I have to say a big thank you to the organizers who are going out of their way to make a great event!