DebConf8 This tshirt is 12 years old and from DebConf8. DebConf8 was my 6th DebConf and took place in Mar de la Plata, Argentina. Also this is my 6th post in this series of posts about DebConfs and for the last two days for the first time I failed my plan to do one post per day. And while two days ago I still planned to catch up on this by doing more than one post in a day, I have now decided to give in to realities, which mostly translates to sudden fantastic weather in Hamburg and other summer related changes in life. So yeah, I still plan to do short posts about all the DebConfs I was lucky to attend, but there might be days without a blog post. Anyhow, Mar de la Plata. When we held DebConf in Argentina it was winter there, meaning locals and other folks would wear jackets, scarfs, probably gloves, while many Debian folks not so much. Andreas Tille freaked out and/or amazed local people by going swimming in the sea every morning. And when I told Stephen Gran that even I would find it a bit cold with just a tshirt he replied "na, the weather is fine, just like british summer", while it was 14 celcius and mildly raining. DebConf8 was the first time I've met Valessio Brito, who I had worked together since at least DebConf6. That meeting was really super nice, Valessio is such a lovely person. Back in 2008 however, there was just one problem: his spoken English was worse than his written one, and that was already hard to parse sometimes. Fast forward eleven years to Curitiba last year and boom, Valessio speaks really nice English now. And, you might wonder why I'm telling this, especially if you were exposed to my Spanish back then and also now. So my point in telling this story about Valessio is to illustrate two things: a.) one can contribute to Debian without speaking/writing much English, Valessio did lots of great artwork since DebConf6 and b.) one can learn English by doing Debian stuff. It worked for me too! During set up of the conference there was one very memorable moment, some time after the openssl maintainer, Kurt Roeckx arrived at the venue: Shortly before DebConf8 Luciano Bello, from Argentina no less, had found CVE-2008-0166 which basically compromised the security of sshd of all Debian and Ubuntu installations done in the last 4 years (IIRC two Debian releases were affected) and which was commented heavily and noticed everywhere. So poor Kurt arrived and wondered whether we would all hate him, how many toilets he would have to clean and what not... And then, someone rather quickly noticed this, approached some people and suddenly a bunch of people at DebConf were group-hugging Kurt and then we were all smiling and continuing doing set up of the conference. That moment is one of my most joyful memories of all DebConfs and partly explains why I remember little about the conference itself, everything else pales in comparison and most things pale over the years anyway. As I remember it, the conference ran very smoothly in the end, despite quite some organisational problems right before the start. But as usual, once the geeks arrive and are happily geeking, things start to run smooth, also because Debian people are kind and smart and give hands and brain were needed. And like other DebConfs, Mar de la Plata also had moments which I want to share but I will only hint about, so it's up to you to imagine the special leaves which were brought to that cheese and wine party! Update: added another xkcd link, spelled out Kurt's name after talking to him and added a link to a video of the group hug.
My work on Google Summer of Code is to create a new strategy on AppRecommender, where this strategy should be able to get a referenced package, or a list of referenced packages, then analyze the packages that the user has already installed and make a recommendation using the referenced packages as a base, for example: if the user runs "$ sudo apt install vim", the AppRecommender uses "vim" as the referenced package, and should recommend packages with relation between "vim" and the other packages that the user has installed. This work is done and added to the official AppRecommender repository.
During the GSoC program, more contributions were done with the AppRecommender project helping the system to improve the recommendations, installation and configurations to help Debian package.
The following link contains my commits on AppRecommender:
During the period destined to students get to know the community of the project, I talked with the Debian community about my project to get feedback and ideas. When talking to the Debian community on the IRC channels, we came up with the idea of using the popularity-contest data to improve the recommendations. I talked with my mentors, who approved the idea, then we increased the project scope to use the popularity-contest data to improve the AppRecommender recommendations.
The popularity-contest has several privacy political terms, then we did a research and published, on the Debian Planet, a post that explains why we need the popularity-contest data to improve the recommendations and how we use this data. This post also contains an explanation about the risks and the measures taken to minimize them.
Then two activities were added to be made. One of them is to create a script to be added on popularity-contest. This script is destined to get the popularity-contest data, which is the users' packages, and generate clusters that group these packages analyzing similar users. The other activity is to add collaborative data into the AppRecommender, where this will download the clusters data and use it to improve the recommendations.
The popularity-contest cluster script was done and reviewed by my mentor, but was not integrated into popularity-contest yet. The usage of clusters data into AppRecommender has been already implemented, but still not added on official repository because it is waiting the cluster cript's acceptance into the popularity-contest. This work is not complete, but I will continue working with AppRecommender and Debian community, and with my mentors' help, I will finish this work.
The following link contains my commits on repository with the popularity-contest cluster script's feature, as well as other scripts that I used to improve my work, but the only script that will be sent to popularity-contest is the create_popcon_clusters.py:
The following link contains my commits on repository with the AppRecommender collaborative data feature:
Google Drive folder with the patch: