Quote:
Originally Posted by Troubrin
its 3:30am i didnt read the whole thread but have a few questions about it.
1. Has anyone figured out HOW and WHO were hacking the accounts.
2. WTF is sony doing about it.
3. Please read question 1 again.
|
1) From the usage of the attacked accounts, the WHO would be someone with an interest in plat.
2) Only SOE knows, but thus far only putting out generic 'don't get hacked and protect yourself' 'information'.
It is a difficult situation for any company - it is, from what is known, not their software nor their site that is vulnerable (well they are vulnerable, but that is not what is seen in the current context) so they can't do much to protect the users (right now). Also SOE is not a security specialist company and to handle this you really need specialists - they are expensive. It would be nice if they would come forward and disclose how it is happening and what they are doing internally - that is what most respectable companies do these days.
It is clear however that MMO companies have to address this issue, because the offenders will be able to keep preying on users not patching their machines.
3) ok
Quote:
Originally Posted by Niber
I'd be weary of all forums/wiki's/blogs .. whatever else, that allow (public) execution of html codes.
For instance, if you can type this <b><u> HI </u></b> and it actually turns bold and underlined. Then you should never visit that forum/wiki/blog again. 
|
Definitely, however that is just in the encoding you have set - some might be vulnerable to simple hex encoding (the above HTML would be: %3Cb%3E%3Cu%3E%20HI%20%3C/u%3E%3C/b%3E ).
If the forum 'programmer' only sanitize for clear text input in the standard encoding scheme, something like this would be allowed through.
Well-developed wiki/forum/blog software would take a defensive approach and encode the output from their data storage, so no matter what is injected into it, it will be harmless when it comes out.
I think the issue in these attacks (but difficult to say with certainty without any 'source') is that the attacke injected data into the application through SQL injection and the resulting data is not encoded when displayed to the users. So they 'appear' safe, from a usage perspective, but are really not.
Quote:
Originally Posted by Niber
But back to the keylogger issue. Assuming they're logging keystrokes, how are they getting the account name? My client saves my station name.
|
EQ2 has to store it somewhere to put it up there. I don't have EQ2 anymore, but if the input box with the username is just a standard windows control it can be read with a simple GetText message - which is usually what keyloggers do (read all controls on a window/form and store the data with the logged information). If not, then it probably reads it from the same source as the EQ2 binary does.