A good resource on the legality of reverse engineering can be found at Chilling Effects: Reverse Engineering.
There were several complaints in the lawsuit, IIRC.
One was that BnetD infringed the trademark of VU/Blizzard by using "BnetD" as their name. That claim seemed fairly strong, as I think Blizzard has "B.NET" trademarked, and obviously BnetD and B.NET operate in the same market: there is a fair case that the names are confusingly similar.
Another was that the BnetD software circumvented content protection mechanisms (that is, the CD-KEY check). The argument goes basically along these lines: Blizzard encouraged people to obtain the software legally by offering a service to those who can verify the legality of their copies. When an independant provides a similar service and does not verify the legality of their copies, they are allowing those without legal copies to circumvent the barrier of needing a valid CD-KEY.
Whether or not this would stand up in court is yet to be determined; I can see it going either way. One could easily argue that the implication of this argument is that under this principle, any company would essentially be free from any kind of competition relating to matchmaking services. Since the CD-KEYs are verified only server-side at BNet (and nobody has found such code in the client-side binary), it is impossible to reverse engineer, and thus necessarily one cannot implement the access protection mechanism without the blessing of Blizzard. However, the court may find that the right to make interoperable products through reverse engineering precedes application of the anti-circumvention clauses. There is at least one late-80s ruling regarding access protection for consoles that would support this -- consoles have forever been exersizing technological lock-out -- but I can't remember the details, except that Nintendo partially lost. I think the court found a 3rd party developer had the right to reverse engineer the lock-out scheme, and I believe defeating the lock-out scheme even mean including a "copyrighted" bank of data. However, the court also found that the 3rd party developer obtained the bank of data through duplicitous means, so they lost anyway. If this is a proper precedent, the courts could find that if Blizzard refuses to allow 3rd party matchmaking services to verify CD-KEYs, then the 3rd party matchmaking services no longer need to fulfill this requirement, and thus are not circumventing the access protection mechanism.
A third argument VU/Bliz put forward was that the developers of BnetD copied code and distributed it. I've had a good, long look at the BnetD code, and the only place I can see that may fit that bill is the password verification scheme. When you enter your password and send it off to the realm for validation, the Diablo software encrypts it client side. This is to prevent someone discovering passwords in a network by listening to packets which aren't theirs.
The funny thing is that the password encryption scheme isn't really necessary for matchmaking services. As far as I can tell, it is only to enable telnet or other "plain-text" entry. This branch of code is not used at all for the enabling of the matchmaking services. Furthermore, I think it is kind of rediculous to suggest the code was copied. For one, the code in the Diablo2 executable is in machine-code format, and the similar code in BnetD is in C. There isn't a simple bijection between the 2: obviously the algorithm represented in the machine code has been re-implemented in C, not copied from the executable. Secondly, it is easy enough to reverse engineer (it took me 8 hours to reverse it from scratch, and I'm not fast or experienced at reverse engineering) so I seriously doubt that the BnetD developers obtained it through an insider, or hacking the Blizzard servers to get the source, the only other ways the code could be found to be "copied". Finally, algorithms do not qualify for copyright, only for patents, and I am unaware of any patents on their method of encrypting passwords.
I'm sure there were a dozen more claims they made. However, the bulk of them were against the people that developed BnetD, not the code itself. It may well be that BnetD is legal, while the actions of the developers were not.
I wish this lawsuit would hurry up, though. With every day that passes, VU/Blizz win: who cares if they lose in court if they can kill it through litigation?
There were several complaints in the lawsuit, IIRC.
One was that BnetD infringed the trademark of VU/Blizzard by using "BnetD" as their name. That claim seemed fairly strong, as I think Blizzard has "B.NET" trademarked, and obviously BnetD and B.NET operate in the same market: there is a fair case that the names are confusingly similar.
Another was that the BnetD software circumvented content protection mechanisms (that is, the CD-KEY check). The argument goes basically along these lines: Blizzard encouraged people to obtain the software legally by offering a service to those who can verify the legality of their copies. When an independant provides a similar service and does not verify the legality of their copies, they are allowing those without legal copies to circumvent the barrier of needing a valid CD-KEY.
Whether or not this would stand up in court is yet to be determined; I can see it going either way. One could easily argue that the implication of this argument is that under this principle, any company would essentially be free from any kind of competition relating to matchmaking services. Since the CD-KEYs are verified only server-side at BNet (and nobody has found such code in the client-side binary), it is impossible to reverse engineer, and thus necessarily one cannot implement the access protection mechanism without the blessing of Blizzard. However, the court may find that the right to make interoperable products through reverse engineering precedes application of the anti-circumvention clauses. There is at least one late-80s ruling regarding access protection for consoles that would support this -- consoles have forever been exersizing technological lock-out -- but I can't remember the details, except that Nintendo partially lost. I think the court found a 3rd party developer had the right to reverse engineer the lock-out scheme, and I believe defeating the lock-out scheme even mean including a "copyrighted" bank of data. However, the court also found that the 3rd party developer obtained the bank of data through duplicitous means, so they lost anyway. If this is a proper precedent, the courts could find that if Blizzard refuses to allow 3rd party matchmaking services to verify CD-KEYs, then the 3rd party matchmaking services no longer need to fulfill this requirement, and thus are not circumventing the access protection mechanism.
A third argument VU/Bliz put forward was that the developers of BnetD copied code and distributed it. I've had a good, long look at the BnetD code, and the only place I can see that may fit that bill is the password verification scheme. When you enter your password and send it off to the realm for validation, the Diablo software encrypts it client side. This is to prevent someone discovering passwords in a network by listening to packets which aren't theirs.
The funny thing is that the password encryption scheme isn't really necessary for matchmaking services. As far as I can tell, it is only to enable telnet or other "plain-text" entry. This branch of code is not used at all for the enabling of the matchmaking services. Furthermore, I think it is kind of rediculous to suggest the code was copied. For one, the code in the Diablo2 executable is in machine-code format, and the similar code in BnetD is in C. There isn't a simple bijection between the 2: obviously the algorithm represented in the machine code has been re-implemented in C, not copied from the executable. Secondly, it is easy enough to reverse engineer (it took me 8 hours to reverse it from scratch, and I'm not fast or experienced at reverse engineering) so I seriously doubt that the BnetD developers obtained it through an insider, or hacking the Blizzard servers to get the source, the only other ways the code could be found to be "copied". Finally, algorithms do not qualify for copyright, only for patents, and I am unaware of any patents on their method of encrypting passwords.
I'm sure there were a dozen more claims they made. However, the bulk of them were against the people that developed BnetD, not the code itself. It may well be that BnetD is legal, while the actions of the developers were not.
I wish this lawsuit would hurry up, though. With every day that passes, VU/Blizz win: who cares if they lose in court if they can kill it through litigation?