Issue on latest CO version.

Quote:

Originally Posted by pro4never Ok so that's why in my source we never did full login on creation. We tried a few random ways of dealing with the login sequence and did one test where we directly cloned the ENTIRE login sequence from TQ (hard coded except for uids and important data) and we still ended up with no interface. If you do find a solution to this I'd love to hear it cause, while we never spent that much time looking into it, it was definitely an annoying issue when we were working on our server.

Quote:

Originally Posted by m7mdxlife i didn't know your source was running on the latest patch

Quote:

Originally Posted by Lateralus 1033 - I think it's the date/time packet sent at login to synchronize the server's time with the client. Edit: Or maybe not. 5 different subtypes of this packet. Sent client->server and server->client.

word - len, word - type
word(0) - not used
word(1 or 0)  ;client sends 1, server replies with 0
dword(1) ; anyone got any idea what is this for?
dword(1002) ; anyone got any idea what is this for?

word - len, word - type
dword(0) ; always

word - len, word - type
dword(timeGetTime() or 0) ; sometime 0 and sometime timeGetTime() [serverside] -- when it's about creating character it seems to be 0, otherwise timeGetTime()
dword(256)

C:S-1052
C:S-1044(1)
S:C-1044(0)
S:C-2079
S:C-2078
S:C-1033
S:C-1004(ANSWER_OK)
S:C-1006
S:C-10017(25, 39)
C:S-1134... idk how many times
C:S-10010(set loc)
S:C-10010(set loc/tele with proxy(?))
.
.
.

Quote:

Originally Posted by Kiyono It was the latest at the time of release =X

Quote:

Originally Posted by pro4never Yes it was obviously working on older patches for trinity. Something appears to have changed in the newer patches though.

Quote:

Originally Posted by -impulse- Lol, I have to use more the "find" function -.-. Sorry for even setting a question on that packet. Makes sense now, though for some reason TQ sends it twice. For other people... yes, in current's Trinity version (the one online) creating the character gets you in game without any problems and nothing is missing. However, TQ somehow found a way to check if you connect by proxy or not, or atleast that's how it seems to me, cause whenever I create a character when I am sniffing the packets, the server sends 0x54 in the data packet that should contain the set location id... eh funny stuff, I'll just override it, but still... It's driving me crazy, I updated my login sequence and I still can't find what's wrong. There is this 1044 packet that looks like: Code: word - len, word - type word(0) - not used word(1 or 0) ;client sends 1, server replies with 0 dword(1) ; anyone got any idea what is this for? dword(1002) ; anyone got any idea what is this for? 2079 is Code: word - len, word - type dword(0) ; always 2078 ... : Code: word - len, word - type dword(timeGetTime() or 0) ; sometime 0 and sometime timeGetTime() [serverside] -- when it's about creating character it seems to be 0, otherwise timeGetTime() dword(256) Which looked like client sends this packet after 1052, waits server's response and then the server continues with 2079, 2078, 1033, ANSWER_OK, hero info, 10017 (status effect, vip level), 1129 (vip options), client sends bunch of 1134 (quest packet), then the client sends 10010 for set location and the server answers with teleport(?) if I am on my proxy and set location if I am not on it... did they find a way to check? Login Sequence: Code: C:S-1052 C:S-1044(1) S:C-1044(0) S:C-2079 S:C-2078 S:C-1033 S:C-1004(ANSWER_OK) S:C-1006 S:C-10017(25, 39) C:S-1134... idk how many times C:S-10010(set loc) S:C-10010(set loc/tele with proxy(?)) . . . In case someone has more info please post here.

Quote:

Originally Posted by m7mdxlife Impulse said that TQ changed the login sequence so you wont have to relogin after your char is created, and he never got throw with it fully working, he said that he's having some trouble with it, he didn't specify what patch he's using and TQ did that ...

Quote:

Server -> Client,Type: 2078 0C 00 1E 08 45 45 2C 4F 00 01 00 00 Server -> Client,PacketType: 2078 0C 00 1E 08 73 45 2C 4F 00 01 00 00 1328301381 1328301427

Quote:

Originally Posted by pro4never If you do find a solution to this I'd love to hear it cause, while we never spent that much time looking into it, it was definitely an annoying issue when we were working on our server.

if (user.Create(msg))
                                                    if (Database.CreatePlayer(user))
                                                    {
                                                        var response = new TalkPacket("SYSTEM", "ALLUSERS", "ANSWER_OK", string.Empty, 0xffffff, ChatType.Register);
                                                        user.Send(response);
                                                    }
                                                    else
                                                        Console.WriteLine("Failed making char");

Quote:

Originally Posted by Kiyono I removed a few lines of code from your char creation handling and it logs in straight after char creation with interface: Code: if (user.Create(msg)) if (Database.CreatePlayer(user)) { System.Threading.Thread.Sleep(150);//delay to make sure database has finished updating var response = new TalkPacket("SYSTEM", "ALLUSERS", "ANSWER_OK", string.Empty, 0xffffff, ChatType.Register); user.Send(response); } else Console.WriteLine("Failed making char");

Quote:

Originally Posted by -impulse- It's not a bot check as far as the client is concerned... the client doesn't return a packet back... and the client wouldn't need to receive the timeGetTime() for the char creation. Beside this packet is always sent after 1004 answer_ok (2100 or 2101)... If you are using flatfile database or mysql then that Sleep is not needed... I have no idea how mssql works but,l mysql always returns something for every query you make so, until the query is finished, you wont get to the Sleep(150). You put your thread to sleep for nothing.