Return to Question

From this excellent answer I learned (correct me if I am wrong) that when writing a block cipher with say key size 128 bit, one has to pad the password given (variable size) so that it becomes exactly the 128 bit that is needed. I had thought that one would just add zeros or repeat the password until the length becomes 128 bit. However, after implementing these options, I can see that the key doesn't look quite random.

I understand that one might use PBKDF2. From what I understand, one has to choose a hash function and that the SHA might be a good choice.

I would like to write my own block cipher from scratch, and having to write all the code for SHA seems a bit much for me right now.

(1) Is there a nicer simpler hash function that could semi-securely be used in PBKDF2? Or alternatively a simpler, but still nice way of padding a password that would be "easy" to do "by hand"? At least, I would like something simple so that I can finish an implementation of TEA.

[(2) Bonus question: I noticed that the hash functions SHA, MD5,... seem to produce a string that is made up of characters with ASCII codes in printable range. Is that a problem when hashing passwords that are used to encrypt plaintext?]

EDIT: I realized that I was wrong about what I had "noticed" in questions 2.

From this excellent answer I learned (correct me if I am wrong) that when writing a block cipher with say key size 128 bit, one has to pad the password given (variable size) so that it becomes exactly the 128 bit that is needed. I had thought that one would just add zeros or repeat the password until the length becomes 128 bit. However, after implementing these options, I can see that the key doesn't look quite random.

I understand that one might use PBKDF2. From what I understand, one has to choose a hash function and that the SHA might be a good choice.

I would like to write my own block cipher from scratch, and having to write all the code for SHA seems a bit much for me right now.

(1) Is there a nicer simpler hash function that could semi-securely be used in PBKDF2? Or alternatively a simpler, but still nice way of padding a password that would be "easy" to do "by hand"? At least, I would like something simple so that I can finish an implementation of TEA.

[(2) Bonus question: I noticed that the hash functions SHA, MD5,... seem to produce a string that is made up of characters with ASCII codes in printable range. Is that a problem when hashing passwords that are used to encrypt plaintext?]

EDIT: I realized that I was wrong about what I had "noticed" in questions 2.

From this excellent answer I learned (correct me if I am wrong) that when writing a block cipher with say key size 128 bit, one has to pad the password given (variable size) so that it becomes exactly the 128 bit that is needed. I had thought that one would just add zeros or repeat the password until the length becomes 128 bit. However, after implementing these options, I can see that the key doesn't look quite random.

I understand that one might use PBKDF2. From what I understand, one has to choose a hash function and that the SHA might be a good choice.

I would like to write my own block cipher from scratch, and having to write all the code for SHA seems a bit much for me right now.

(1) Is there a nicer simpler hash function that could semi-securely be used in PBKDF2? Or alternatively a simpler, but still nice way of padding a password that would be "easy" to do "by hand"? At least, I would like something simple so that I can finish an implementation of TEA.

[(2) Bonus question: I noticed that the hash functions SHA, MD5,... seem to produce a string that is made up of characters with ASCII codes in printable range. Is that a problem when hashing passwords that are used to encrypt plaintext?]

EDIT: I realized that I was wrong about what I had "noticed" in questions 2.

From this excellent answer I learned (correct me if I am wrong) that when writing a block cipher with say key size 128 bit, one has to pad the password given (variable size) so that it becomes exactly the 128 bit that is needed. I had thought that one would just add zeros or repeat the password until the length becomes 128 bit. However, after implementing these options, I can see that the key doesn't look quite random.

I understand that one might use PBKDF2. From what I understand, one has to choose a hash function and that the SHA might be a good choice.

I would like to write my own block cipher from scratch, and having to write all the code for SHA seems a bit much for me right now.

(1) Is there a nicer simpler hash function that could semi-securely be used in PBKDF2? Or alternatively a simpler, but still nice way of padding a password that would be "easy" to do "by hand"? At least, I would like something simple so that I can finish an implementation of TEA.

[(2) Bonus question: I noticed that the hash functions SHA, MD5,... seem to produce a string that is made up of characters with ASCII codes in printable range. Is that a problem when hashing passwords that are used to encrypt plaintext?]

EDIT: I realized that I was wrong about what I had "noticed" in questions 2.