
Originally Posted by
nevS
Well, copying the assets into your public repository will probably not work - Webzen could just mail to GitHub (or similar service), and your repository is gone. As you said, recreating assets is also a task which is just too big.
So, the project must be limited to be a plain main.exe replacement, which can use all the existing assets without actually delivering them.
You could offer to be compatible to additional formats, though. So your new stuff can use better formats ;)
I think developing such a client is not so difficult, if you have knowledge of all the data formats. The server contains most of the game logic, so the client does "only" have to do:
1. React on packets coming from server, which could change visible aspects.
2. React on user input by sending packets back to the server and possibly do some animations (e.g. walking).
3. User interface stuff (menus)
Generally, I like the idea that it should be compatible with the mu online network protocol - who would use it without compatible servers out there? And rewriting a server can be very time consuming - time you would need to write the client :) And even if you want to rewrite a server (what I do atm ;)), you can still use this protocol, it's a bit limiting, but mostly not that bad.
IMHO, compatibility to existing stuff and openness are the most important requirements for the success of such a project.