TQServer/TQClient prefix

[Super Aids]

Is the TQServer/TQClient prefix included in the size or not?

[pro4never]

No. Packet size does not include the seal.

Prefix would also imply it came before the packet but I'm sure you know it goes at the end ahah.


The reason packet size doesn't include the seal is that it's only used to signify the transfer of data is over. If you're combining packets I'm under the impression TQServer is only needed at the end of the large packet block but I could be wrong on that side of things.

[Super Aids]

Yeah, I knew it was at the end of the packet of course. I just wanted to know whether it should be included in the size or not.

[Best Coder 2014]

Originally Posted by pro4never

No. Packet size does not include the seal.

Prefix would also imply it came before the packet but I'm sure you know it goes at the end ahah.


The reason packet size doesn't include the seal is that it's only used to signify the transfer of data is over. If you're combining packets I'm under the impression TQServer is only needed at the end of the large packet block but I could be wrong on that side of things.

Aren't the suffixes just used for packet validation? That's what I always thought. It wouldn't make sense to use the suffix for packet splitting since the prefix/header already takes care of that.

[.Ocularis]

Originally Posted by Best Coder 2014

Aren't the suffixes just used for packet validation?

They may or may not be... When sending a single tcp frame containing multiple server messages the client will still split them properly..

[Best Coder 2014]

Originally Posted by .Ocularis

They may or may not be... When sending a single tcp frame containing multiple server messages the client will still split them properly..

Multiple server messages without the suffixes, you mean?

[.Ocularis]

No, actually. Multiple complete server messages; each including the TQServer suffix, all encrypted and sent as single packet. The client splits them properly.. It could be the size, but it could be the suffix.

What I know for certain is if a message's array size is not 8 bytes larger than the size written to the header, the client disconnects from the *game server.

[Best Coder 2014]

Originally Posted by .Ocularis

No, actually. Multiple complete server messages; each including the TQServer suffix, all encrypted and sent as single packet. The client splits them properly.. It could be the size, but it could be the suffix.

What I know for certain is if a message's array size is not 8 bytes larger than the size written to the header, the client disconnects from the *game server.

Oh, yeah, I'm aware of that. The client is definitely using the size in the header to split packets though, which is why I'm saying I think the TQServer/TQClient suffixes are only used to validate packets, but that conflicts with that P4N is saying about the game server sending multiple "combined" packets with only one TQServer suffix.

[pro4never]

Originally Posted by Best Coder 2014

Oh, yeah, I'm aware of that. The client is definitely using the size in the header to split packets though, which is why I'm saying I think the TQServer/TQClient suffixes are only used to validate packets, but that conflicts with that P4N is saying about the game server sending multiple "combined" packets with only one TQServer suffix.

I'm not working with a version where TQServer/TQClient is used. The necessity of it being needed per packet when merging them together was what I said I didn't know.

The short version of my earlier answer.

No: TQServer is not included in packet length
Client requires TQServer at the end of each group of packets at very minimum or else it will dc.
TQServer MAY not be needed on each packet when you're combining them together (not sure as I'm not working with a version where server seal is used).