[Release] 6685 Server Source

[no_1]

Is there any intrest in this project?
Downside is focus on the server only not the client side. We would only be able to manage the classic server. Newer server have packet encryptions.

Havent tried running it but looking at it with provided screenshot looks good and already did a nice job.

[DuaSelipar]

Originally Posted by no_1

Is there any intrest in this project?
Downside is focus on the server only not the client side. We would only be able to manage the classic server. Newer server have packet encryptions.

Havent tried running it but looking at it with provided screenshot looks good and already did a nice job.

yeah got several people working on this project personally. but not me xD


you can decrpyt that encryptions using ai.. ive manage use client 2260 ( few patch after ranger release)

[no_1]

Ai to the rescue then. So they not going to use the github? its kinda silent there. So endgoal here is to even use latest clients with own server files then?

[Demonology]

Originally Posted by no_1

Is there any intrest in this project?
Downside is focus on the server only not the client side. We would only be able to manage the classic server. Newer server have packet encryptions.

Havent tried running it but looking at it with provided screenshot looks good and already did a nice job.

I am working on my own version with someone else at the moment too but for now it is not public work but that may change when we are closer to being done. We have deviated quite significantly from the style of the source in my original post. We opted to not use the .csv style action system and are adapting our version to use a more traditional eo db. In terms of implementation, we have implemented all class abilities, hatching, monster ai, composing logic(to the best of our ability), vampire and shadow knight sacrifice functions, full wardrobe functionality, and a significant chunk of action logic. We have also re-enabled mannequin functionality because in this chinese client it was completely disabled due to a mass amount of botting around the time period that this client came out. Overall, the source that I originally released is missing quite a lot and it mostly functions as a proof of concept and hopefully inspiration to someone. Mainly, this release is meant to show an alternate way of creating your own eo source. I think it could be a good starting point for getting into reversing eo but I don't think I would ever open a server without seriously reworking most of it.

Also, as for a very long term goal, I believe I could potentially backport a lot of things from the current eo client. I was experimenting with it a bit with the wardrobe function. I have added almost all of the casual clothing from the current eo client. A few of the casuals have a rendering issue occuring that would need to be worked out. This and the mannequin reimplementation made me pretty confident I could backport entire classes into this client but it would be a ton of work and not something I would consider working on until I have a completed source.

[no_1]

Originally Posted by Demonology

I am working on my own version with someone else at the moment too but for now it is not public work but that may change when we are closer to being done. We have deviated quite significantly from the style of the source in my original post. We opted to not use the .csv style action system and are adapting our version to use a more traditional eo db. In terms of implementation, we have implemented all class abilities, hatching, monster ai, composing logic(to the best of our ability), vampire and shadow knight sacrifice functions, full wardrobe functionality, and a significant chunk of action logic. We have also re-enabled mannequin functionality because in this chinese client it was completely disabled due to a mass amount of botting around the time period that this client came out. Overall, the source that I originally released is missing quite a lot and it mostly functions as a proof of concept and hopefully inspiration to someone. Mainly, this release is meant to show an alternate way of creating your own eo source. I think it could be a good starting point for getting into reversing eo but I don't think I would ever open a server without seriously reworking most of it.

Also, as for a very long term goal, I believe I could potentially backport a lot of things from the current eo client. I was experimenting with it a bit with the wardrobe function. I have added almost all of the casual clothing from the current eo client. A few of the casuals have a rendering issue occuring that would need to be worked out. This and the mannequin reimplementation made me pretty confident I could backport entire classes into this client but it would be a ton of work and not something I would consider working on until I have a completed source.

I noticed the coding style you referring to. I messed around with the Xena project what was released a long time ago. This was a very different type of code.
The classic eo db? what you mean with this? i noticed you also have a file for your database you want to change this ?
The client you releasing here. is this one of the last version without the encryption? Have you tried with other clients to make it work?

And you already did a nice job with what is already done. but i also belive like you mention it stil going to be a masive job to finish it

[Demonology]

Originally Posted by no_1

I noticed the coding style you referring to. I messed around with the Xena project what was released a long time ago. This was a very different type of code.
The classic eo db? what you mean with this? i noticed you also have a file for your database you want to change this ?
The client you releasing here. is this one of the last version without the encryption? Have you tried with other clients to make it work?

And you already did a nice job with what is already done. but i also belive like you mention it stil going to be a masive job to finish it

So, in this version it only kept player data in a mysql database like cq_user and all of the player associated data. The rest of the database was controlled through various .csv files that are kept locally with the server. This source was originally meant to be a locally hosted server and was never meant to be taken online. So I guess that's why the original dev opted for .csv style. The database I am using is very similar to 1655 database but its not the same one it is a bit older. So we have reworked the entire database system to work very similarly to how 1655 and mannequin db works. The only thing we are handling differently is we are using lazy loading to load actions rather than loading the entire action db on server boot. It saves us time for testing but also in a live server environment it cuts down on memory usage since the entire action db isn't being loaded into memory each time the server starts.

As for the packet encryption, the packet cipher has been completely reversed with significant ai assistance. It uses a custom, table driven byte cipher. It is similar in general design to crazy tao's two table XOR encryption but much more elaborate. Each connection has an encryption table and a separate decryption table, a hardcoded 1024 byte starting table which is divided into four 256 byte sections and independent incoming and outgoing byte counters. You can find the implementation for this in GamePacketKeyEx.cs. The client I am using here is a chinese client but this source works with english version 1838 which uses the same networking but that english version does not have fullscreen functionality which is why I opted to just translate the chinese client to english.

It is a massive job but I have been working on it consistently since June. I have no plans of stopping until I consider it complete. I also have plans for lots of custom features after the server is completed.