| bio | website | bernd.eckenfels.net |
|---|---|---|
| location | Karlsruhe, Germany | |
| age | 41 | |
| visits | member for | 6 months |
| seen | Jan 29 at 18:36 | |
| stats | profile views | 3 |
Chief Architect at a german ISV. Java Guru, IT-Security freak, married and pirate.
|
Jan 5 |
comment |
Preventing user supplied javascript from posting to external server Actually CSP has the "connect-src" attribute, but yes the best separation can be achieved by downloading the plugin from its own origin/domain. |
|
Dec 21 |
comment |
How can I explain SQL injection without technical jargon? I typically try to skip the details if it is not a technical user but only describe the risk and effect. "If $software has a bug which allows SQL injection an attacker can smuggle commands into your database to destroy or modify data or passwords." Why would you want to go to the details of sql parsing, prepared statements and quoting. |
|
Dec 21 |
comment |
Resource consumption attacks against algorithms Actually the Apache Range Header bug is not a hash collission problem. Apache had hash problems as well. |
|
Dec 21 |
comment |
How secure is opening an incoming port for a java application? No, it is not the "Java vulnerability" you need to care here about but the actual java application which is started with java. For example you can write a Java appication which deletes a file every time you receive a UDP message. Or allows upload of files and executes it. So basically the risk depends on the java program. Java Runtime Bugs would come on top of that (but are rather unlikely for simple udp ports). |
|
Nov 22 |
comment |
How to securely hash passwords? It is not "most common" or "best". Most common are crypt() family hashes or even unsalted variants. KDFs are not designed for that and therefore be used with care. Dont know what the "best" is (it depends on your goals I guess), but bcrypt and scrypt or even sunmd5 is more purpose build. |
|
Nov 22 |
comment |
Is PBKDF2-based System.Cryptology.RFC2898DeriveBytes() “better” for Unicode Password hashing than traditional methods? For most salted encryption schemes which are used out there in the wild the salt is not hidden. Thats not a helpfull restriction. You cannot legitimately re-produce the hashing if you dont know the salt. |
|
Nov 22 |
comment |
Is PBKDF2-based System.Cryptology.RFC2898DeriveBytes() “better” for Unicode Password hashing than traditional methods? I disagree on the "nist-sp800-132" approves PBKDF2 for "hashing and storing passwords". It actually talks about using PBKDF2 as a KDF for deriving data protection keys which are used to encrypt stored data. |