Comments on: So I designed a block chipher. http://trilema.com/2018/so-i-designed-a-block-chipher/ Moving targets for a fast crowd. Wed, 13 May 2026 22:23:28 +0000 http://polimedia.us hourly 1 By: Mircea Popescu http://trilema.com/2018/so-i-designed-a-block-chipher/#comment-126943 Mircea Popescu Tue, 30 Oct 2018 20:31:35 +0000 http://trilema.com/?p=81342#comment-126943 For future record : this <a href=http://btcbase.org/log/2018-10-30#1867696 >actually died the next day</a>, leaving us lonelier and sadder than before. For future record : this actually died the next day, leaving us lonelier and sadder than before.

]]> By: Mircea Popescu http://trilema.com/2018/so-i-designed-a-block-chipher/#comment-126942 Mircea Popescu Tue, 30 Oct 2018 16:26:53 +0000 http://trilema.com/?p=81342#comment-126942 I wasn't multiplying matrices, but bit offsets. "The i-th bit of E is obtained by taking the i-th bit of P, and xoring it with the product of the 1st bit of K and the 1st bit of P, and further xoring it with the product of the 2nd bit of K and the 2nd bit of P..." is what was contemplated. I wasn't multiplying matrices, but bit offsets.

"The i-th bit of E is obtained by taking the i-th bit of P, and xoring it with the product of the 1st bit of K and the 1st bit of P, and further xoring it with the product of the 2nd bit of K and the 2nd bit of P..." is what was contemplated.

]]> By: Diana Coman http://trilema.com/2018/so-i-designed-a-block-chipher/#comment-126941 Diana Coman Tue, 30 Oct 2018 16:16:17 +0000 http://trilema.com/?p=81342#comment-126941 "E, the enciphered message, is the result of the formula E[i] = P[i] xor K[1] * P[1] xor K[2] * P[2] xor ... xor K[j] * P[j] xor ... xor K[n] * P[n] (where P is the plaintext)" -- I don't think this is correctly stated - or I don't get something in it. I think E(i) = P(i) xor (K*Reg)(1) xor (K*Reg)(2)... xor (K*Reg)(n) with Reg the matrix of registries and K*Reg the usual matrix multiplication. With the example from the logs: K=(0,1,0,1) P=(1,1,1,0) and E(2)=P(2) xor P(3) xor P(1) = 1 xor 1 xor 1 = 1 but with the formula above it would be E(2)= P(2) xor 0*P(1) xor 1*P(2) xor 0*P(3) xor 1*P(4) = P(2) xor P(2) xor P(4) = 1 xor 1 xor 0 = 0 . "E, the enciphered message, is the result of the formula E[i] = P[i] xor K[1] * P[1] xor K[2] * P[2] xor ... xor K[j] * P[j] xor ... xor K[n] * P[n] (where P is the plaintext)" -- I don't think this is correctly stated - or I don't get something in it. I think E(i) = P(i) xor (K*Reg)(1) xor (K*Reg)(2)... xor (K*Reg)(n) with Reg the matrix of registries and K*Reg the usual matrix multiplication.

With the example from the logs: K=(0,1,0,1) P=(1,1,1,0) and E(2)=P(2) xor P(3) xor P(1) = 1 xor 1 xor 1 = 1 but with the formula above it would be E(2)= P(2) xor 0*P(1) xor 1*P(2) xor 0*P(3) xor 1*P(4) = P(2) xor P(2) xor P(4) = 1 xor 1 xor 0 = 0 .

]]> By: Mircea Popescu http://trilema.com/2018/so-i-designed-a-block-chipher/#comment-126936 Mircea Popescu Tue, 30 Oct 2018 01:21:41 +0000 http://trilema.com/?p=81342#comment-126936 Well, you do write great code, just... what can I say, not always for the right thing. Well, you do write great code, just... what can I say, not always for the right thing.

]]> By: Stanislav Datskovskiy http://trilema.com/2018/so-i-designed-a-block-chipher/#comment-126934 Stanislav Datskovskiy Mon, 29 Oct 2018 19:53:30 +0000 http://trilema.com/?p=81342#comment-126934 http://btcbase.org/log/2018-10-29#1867216 for fyootoor ref. http://btcbase.org/log/2018-10-29#1867216 for fyootoor ref.

]]>