IETF 95 Hackathon
Authored by: Agustín Formoso
Great f2f meeting, worth attending. Get to know new people coding similar stuff that you might be coding. Get to feel the IETF spirit.
Still, I think it needs some intermediate level between high-barrier-and-focused, and low-barrier-but-unfocused groups.
Essential things for the hackathon to feel comfortable were there: the physical space for all the participants, food and drinks (actually there were some pretty good meals and snacks), and enough power outlets to charge all of our nerdy devices.
One thing that could had been of use are additional elements for sprint planning such as post-its, big papers to set up sprint backlogs, and some paper and pens to sketch and communicate with other strange-language-speaking group mate.
At the kickoff, the group champions explained what each group is about and what the main objective is. As a first timer, looking for a suitable group to work with during 2 days was a critical point. My previous experience with the hackathon was literally none, as I had no relation with any work in progress so far. I had to find a group which topic was of interest to me and which I had at least some vague concept about the technologies involved. Fortunately, I ended up in a group which required no previous experience with, no specific tools to do what was planned, just my laptop and my will to solve a problem. The problem was to find any metrics that helped discover anomalies or patterns in a dataset of traceroutes, it was a cool project and I enjoyed being the two days coding stuff for it. Besides, you get to feel the nice IETF spirit of building things for a better Internet.
The group deliverables
Probably as the group I ended up in was very generic and the entry barrier was pretty low, setting specific deliverables for the 2 day sprint was difficult. Other groups had more specific tasks to accomplish, such as building features for a YANG modelling application or implementing TLS 1.3 in a Firefox browser. I guess that by following some of the groups closer and collaborating remotely prior to the hackathon might end up in getting the most out of the hackathon, as in that case the hackathon would just be a team f2f meeting, without the hassles of setting up development environments, getting to learn new concepts, or new technologies entirely (nor was my case, but I easily see that situation as a very common scenario as IETF hackathons get more popular).
I still think that the hackathon should consider having low, intermediate, and high entry barrier groups. There's room to set an intermediate level in each hackathon where the group would require some things such as tools (virtual machines, IDEs) and some follow up (probably in the order of weeks time). In that case I think a better and more focused deliverable might be reached. In this consideration, the low entry barrier groups should approach newcomers with no experience or contact with the IETF hackathon, and the high entry barrier group should approach the usual hackathon participants.
The IETF hasn't finished yet and I think is a very good instance to get to know new people and establish links to motivate future work. The coding done in the group I was part of is of special interest to me and hope that some months from now we have a nice tool to analyze network graphs which benefits our champion (Pål) and the rest of us.
[Editors Note: For more information on the IETF hackathon series, see IETF Hackathon]