11-23-2007, 09:16 PM
|
#22 (permalink)
|
|
The Frequenter
Join Date: Nov 2007
Location: Netherlands
Posts: 460
Thanks: 49
|
Very awesome article. I've read it completely and I must say, bravo.
My system is protected against querystring attacks. Every command MySQL uses, every designated die or exit function is blocked an I left nothing to coincidence there.
Refering back to the mysql_real_escape_string and the sprintf command. I still don't quite understand those. What's up with them. Are they handy to use, easy to filter out and do they provide a safe security and a easy handle?
Next to that, my cookies are stored ... well, of course on a users PC, but stored quite safely.
I used part of a method mr. buddha 'invented'. So to speak anyways. I retrieve the ID from the user, the username, the sha1 (4x) md5 (4x) password, the IP, the browser (incl language) and the time. Next, I am pushing it all into a once again formatted sha1 (4x) md5 (4x) string, which I call the cookies validationkey. For short, unforgeable as I know it. Since you cannot copy the login SECOND, IP, Language, Browser, username and 8 times encrypted password. Next, I push it, when succesfuly logged in, into the users database, where it's stored as user_validationkey. That is the key which will be refreshed every time. The NEXT thing I do, is of course update the userrecords and give the cookie an ID of it's own. It's almost logical that the $_SESSION[ID] and user_id still stored the user's ID, but sessions don't last as long as 30day cookies, or more.
Well, I am quite amazed by the topic but as you can see and if I am correct, my system is at least a little bit foolproof. Tho I bet some of your handy people could still get in.
What do you guys think?
|
|
|