Related: 100299K - Digraphs, 290508668

The first case in test 3 was:

676
aa
ab
ac
...

My output was:

`

That was an illegal character, yet the checker gave me an accepted???

It's been 10 (11?) years yet I'm the first one who discover this...

MikeMirzayanov fix this please...

Recently, Luogu has started to provide official English statements from LGR-195.

As a tester, I with someone else am preparing a new round on Luogu. However, currently we have conflicts over English statements: I suggest them provide statements using PDFs that are generated in Polygon (Note that such PDFs slightly differ from those in LGR-201), yet others insist that statements should be provided in the tradition form like LGR-195.

Here I'll ask for your advice. Please, vote:

• The current one (like this)
• The new one (PDFs, like the one below)
• I suggest providing both of them
• I have another plan (We'll be grateful if you share that)

The picture below shows (a tiny part of) the PDF of Problem C in our round.

It has been well-known that submissions can be sorted by execution time. Buy why not sort the submissions by execution memory? Although I finally did it, I found it difficult to look for someone to hack to ML, but it's easy to look for someone to hack to TL.