Randori: “The Skyline Problem” in C (acm.uva.es/p/v1/105.html)
20:00 ~ 21:00 Discussing the Problem
The problem was presented by one of the participants, who proposed an aditional challenge of developing an algorithm that would run in time O(n lg n). The other participants were interested in the challenge, and we discussed a few approaches to the problem. For each approach, we tried to understand if it would meet the additional restriction and why. Some very interesting suggestions came up and we also learned what that restriction really means. After discussing the approaches, we settled on trying to implement one of them, based on divide and conquer strategies.
21:00 ~ 22:00 Coding Session
During the coding session, we started by testing and implementing the structure of a skyline. We didn’t get very far since problem discussion took up so much time. What a dilema! We also managed to change pairs more “time-boxedly”, that is, we respected the 7 minute timer for Randori a little better that usual. This makes the meeting more dynamic and fast paced, and we enjoyed it.
22:00 ~ 22:15 Retrospective
We enjoyed the discussion of approaches, but regreted having less time to code - a dilema we don’t yet know how to solve. We used the C Dojo Tests that has been remade by HugoCorbucci (he owes us a Kata of it, but he has little time to prepare it), and it is improving. We still want a lot more things, though! Someone pointed out that doing TDD in a familiar language like C makes it easier to undestand what TDD really is. Unfortunately we still got problems with the Makefile, but we don’t really want to solve them during the meetings. We have improved a bit the level of conversation in the audience, but it is still a problem. We also don’t like that we haven’t posted last week’s meeting report yet.