2,875 reputation
821
bio website touset.org
location San Francisco, CA
age 31
visits member for 2 years, 9 months
seen 2 hours ago

Cyclist. Rubyist.


6h
comment Can I use AES CTR mode to encrypt files with same key and nonce?
This is an incredibly dangerous thing to do, and the only discernible benefit is saving 32 bytes per file. I have to ask: Why?
May
18
comment Are there any security implications in creating an SSH key for the root user?
I suspect your basic premise is wrong. You do not need to sudo to the root user to accomplish this. Change the ownership and permissions on the subdirectory of /srv you're deploying these sites to, do the same for the directory where Apache configurations are stored in, and configure /etc/sudoers to allow this user to restart apache.
May
11
comment How are digital certificates compared?
If someone has the level of access on your computer necessary to overwrite your TLS certificate chain, you have already lost. They can almost certainly record your keystrokes, view and edit the entire contents of your hard drive, etc.
May
11
comment Which passwords to exclude from a password manager?
This sort of scenario is precisely why you should always have backups. That said, Google allows you to print out recovery codes that can be used in this type of event. Additionally, you can link your cell phone number (which you can use once you've acquired a new phone) to use to recover you account as well.
May
9
awarded  Nice Answer
May
8
answered Which passwords to exclude from a password manager?
May
5
comment Can I initially hash passwords with SHA instead of hashing them with bcrypt to decouple requests from slow crypto functions?
Slow-hashes like bcrypt are designed to be tuned to meet your specific performance needs. If the default number of rounds is taking too long (minutes for 10 passwords is absurdly long), then simply reduce the work factor until you find a better balance.
May
1
comment Is it a secure encoder and hasher?
You appear to be reinventing your own password hashing algorithm. Worse, it appears at first glance that your function is reversible, which is a cardinal sin for password hashing. Please just use an existing, heavily-cryptanalyzed function purpose built for this use case such as bcrypt, scrypt, or PBKDF2.
Apr
29
comment putting my password in a php file (phpmailer)
To what end? Whatever accesses this configuration file must also have access to the key that decrypts it, gaining you little. If anyone gets access to this file as the app user or root, they'll have access to the password no matter what you do.
Apr
29
answered putting my password in a php file (phpmailer)
Apr
28
comment Encoding Same Message with Different IV (AES/CBC)
Essentially the entire point of an IV is to make this exact operation safe.
Apr
24
answered How can I stop an account I don't control from sending spam email in my name to my contacts?
Apr
13
comment More secure curve than Curve25519
We don't always use 256-bit ciphers. AES-128 is highly used, and is still (and will for the foreseeable future continue to be) considered a perfectly reasonable option for any new cryptosystem. In the absence of any specific 256-bit requirement, many would argue that AES-256 over AES-128 is simply wasted CPU cycles.
Apr
10
comment Source code as password
Just use a password manager like KeePass, LastPass, or 1Password. This is essentially a solved problem.
Apr
8
revised Is there any reason for using private key 2 times when creating security hash?
added 11 characters in body
Apr
7
answered Is there any reason for using private key 2 times when creating security hash?
Apr
7
comment Is there any reason for using private key 2 times when creating security hash?
This is a crude attempt at constructing an HMAC. I would hesitate to use such a payment system; if this is the quality of the cryptography they display publicly, the quality they use internally is certainly worse.
Apr
7
comment Is there any reason for using private key 2 times when creating security hash?
This is not strictly true. This construct (crudely) guards against length extension attacks, where an attacker knowing H(m) but not m can generate H(m+p+a) where p is the padding used by the hash function and a is attacker-controlled. Since the attacker doesn't know the secret key being used, he can't ensure his message ends with it. The real solution to this is to use the correct cryptographic construct: a MAC. More specifically, they probably want an HMAC.
Apr
7
comment How secure is MD5 for TLS message authentiaction?
RSA is not being removed from TLS 1.3. Static key exchange with RSA (and DH) is.
Apr
7
answered How secure is MD5 for TLS message authentiaction?