There is a strange bug on the homepage of ABC355. If you're in English mode, when you enter the page, your browser begins playing music; after that, there will be some Japanese introduction to this company.

I found that the problem was caused by the video on the homepage (in Japanese mode), and the solution is to ban your browser from automatically playing media on the page. And I hope AtCoder can fix it. This is not a big problem, but sometimes it's annoying.

Misread B :)

Btw, how to solve F? Does it involve link-cut trees?

found very difficult.

My idea for problem-E using dp find the minimum no of moves to reach from L to R, then get the path. I'm getting WA.

Code

SpeedCoder?

Anybody missed reading the constraint on weight in F and then waste time?

Cool.

Is there any other solution apart from link-cut tree? I didn't know about link-cut tree so did a brute force kind of solution.
I processed initial N-1 edges from lowest weight to highest weight. For each weight, I maintained connected component data using DSU.
Then for each additional edge of weight W to add, I checked if nodes U, V are in same component at DSU of weight W. If they are that means, an edge with weight <= W has connected components of these two nodes so the incoming edge can't replace any edge in MST.

If they are in different components, then that means, components to which U, V belong are connected by some higher weight edge and incoming edge can replace heavier edge in MST and update MST edge weight and then do union for U, V for all edge weights greater than or equal to W.

I got an AC with this approach but not sure if this approach is correct as everybody is mentioning link-cut tree. Can someone confirm if this is correct?

Why does this solution to D gives WA (https://atcoder.jp/contests/abc355/submissions/53878525)

For problem E, is there any intuitive understanding about the definition of S(x,y) = -S(y,x) when x > y.

I wonder how could someone come up with this definition and construct a graph.