5 Foot Step wrote on Apr 30^th, 2015 at 5:13am:


I found where those settings are stored, but thought it couldn't possibly work that way and ignored it.
Must be worth a try.

Shirley even Turbine aren't that stupid...?

5 Foot Step wrote on Apr 30^th, 2015 at 5:13am:


That...makes me curious.

Fucking Post-MOTU

Need a packet sniffer, time meow

Puppies wrote on May 2^nd, 2015 at 12:41pm:


wireshark is your friend... there's also that nice thingie from sysinternals that trace system calls ( and other calls ).

I think you are way off the mark: The server side is where the certs are located and tokens get generated. You would need to do MitM to get an admin cert to copy from and compare to a token.

Not like you need 256 AES as 128 will sufffice if you just generate one for each session (which would still take billions of years to decipher). I would think they have a commercial product that does the encryption so that it isn't open to side channels.

IMNSHO, the reason people see things like this is that Turbine has had -most probably - memory issues. There were several times where people where getting grouped together without being in the same party. On three occasions, I have found myself at someone else's login screen with their characters. There is code to double check this when you try things such as purchase stuff. Forcing a memory overflow via their client is highly improbable.

As for the flag for F2P, what it does is suppress the request for the license from the client in the first place. The flags are there to make requests from the servers - not actual settings.


...unless they have severely mucked up the code of course.

...I return you to rest of the speculation...

Dark_Helmet wrote on May 4^th, 2015 at 4:00am:


As far as encryption, you're right AES 256 or 128 would take forever to decrypt, but they might not even encrypt it...or at least not in that traditional sense.

If they're using KGV-175D TACLANEs (Tactical LAN Encryption Device...google it, it's OPSEC shit) then we're fucked. I've seen FFK vectors get written...fuck that.

This just leaves more questions...less answers.

Back to the Lab.

I've had to put Google translate on hotkey for this thread...   

Puppies wrote on May 4^th, 2015 at 11:48am:


You are jumping the shark so go back to google some more terms.

No, they wouldn't be using KGs - just simple VPN tunnels would suffice between data centers.

I did have rooms full of HAIPEs, so I do know what I am talking about.

Dark_Helmet wrote on May 5^th, 2015 at 4:17am:

I would NEVER use highly encrypted VPN tunnels between locations where accessing (reading and writing to) databases was going on. The performance would just totally suck.

Better off with private point to point OC-3 or faster connections between such sites.

Puppies wrote on May 5^th, 2015 at 12:07pm:



Just... wow.  Stop.

Arkat wrote on May 5^th, 2015 at 11:19am:


*Le Sigh*

L2VPN... or L3VPN is you really want to bother with it ( but for a point to point, L2VPN is way better )... No point into encrypting. MPLS will to the flow segregation for you.

Your OC-3/STM-1 is so has been when you have Terabit connectivity ( and Turbine... well PNAP... has Terabit connectivity ).
Come on in a few months I'll have almost the equivalent of an OC-12 at home. ( 500Mb down, 200Mb up... that is If I don't want to switch operator for my broadband )

L2VPN, from the user point ( aka the company with two sites separated by lots of distance ), is a seamless LAN...

L3VPN, is two separate LANS tied with two routers...

But in between you can have 10 or more routers carrying the packets along with packets from other companies...

yeah it was a parsing error, I noticed it after posting and I was too lazy to fix it.