## Learning to Read and Write Kanji Row 26

As with other posts in this series, this is the results of a test I took today, July 31st, 2021.

I only got two wrong in this test, one of the better scores in recent memory. This is likely due to the random choice of lower row kanji -- intended to force my brain to remember all 421 kanji I've learned. But in this test, the kanji that are from row 1-25 are: 口 冷 親 届 好 港 実 林 回 二. The rows these are from are 4, 16, 9, 23, 13, 21, 16, 8, 18, 1. Which ones did I get wrong? Row 23 and row 21. Note how I stopped my retesting at row 15 because I got such a bad score on it. Specifically I missed a stroke on 冷 but I got most of it. I looked at the banner in my room that has two penguins and that kanji drawn: 冷月中 to get the ice radical on the left. I don't consider that cheating because in a pinch I'll be able to draw it correctly. I forgot . This kanji was in KARAKARA, so I think about it when I see deliver, but I didn't think of door 戸 and rice paddy 田, which I should have. I forgot the reading of but that's okay for now. I wrote the backwards 5 radical 已 in the lower right of . I am happy that my brain remembers 港 as a backwards 5. If we look at k_rads1.txt (a list of popular radicals for each kanji in my list) we find that 已 is found in , , and so far.

What have I been up to instead of learning kanji? I've been playing a lot of games. Not the ones I should, but that's part of my gaming, getting away from stuff I'm supposed to do. If I spend all my time playing the games I'm supposed to, I'll overflow. Yesterday I mostly completed a task I gave myself May 25th, 2021. Not bad. It's reversibility of the function f(x) = ((x + 1) & 7) ^ ((x & 0xfffff8) - 4) which can be seen below in figure 1. Click the image to get a larger version. If you're a cryptographer, you might recognize this construction. It's hash-related. As you can see, this is a poorly-written hash for a specific reason. Why? It seems that the reversibility of functions relies upon a bit of stuff that seems pretty easy to produce, until you protect something with it. Mixing and obfuscation isn't enough. Symmetric ciphers rely on the fact that knowing a lot of the plaintext won't give you the key. Reversibility of a ciphertext to a key would be a cryptanalysis of a cipher assuming the ciphertext necessary to learn the key is substantially cheaper to obtain than brute force search of the key.

Figure 1: f(x) = ((x + 1) & 7) ^ ((x & 0xfffff8) - 4)

Until next time!

I'll hopefully post all tests as I pass them. Wish me .

Javantea out.