Perhaps this may give those aussies a run for the money now.
Ryzom resources code#
Seeing their code in not only a technical education on their architecture but you can see the results of a commerical development process had on the code base versus say an emulator like ArcEmu or any open source driven backend. mind boggling the data reduction, data isolation techniques needed.
Ryzom resources full#
With a full commerical release it allows people to view the strengths and weakness of a particular implementation and see what optimzations can be made.ĬCP right now with Eve Online has one of the most exotic database architectures I've seen to date, I can only imagine the code behind it. A math geek friend of mine rewrote it from a mathmatical point of view and reduced the map generation time from about 4 hours to 2 1/4th hours. Now based on how the two engines worked my map either took up 6mb of ram or 12 mb of ram. I grabbed ArcEmu (a wow emulator) as well as EQ and a few other emulators and stitched a basic randomly generated map in there to test out the algorithm.
Ryzom resources series#
Using that heatmap village markers are deployed then a series of passes are made that merge nearby villages into town, towns into cities, and cities into capitals leaving behind unmerged locals (somewhat like evaporation). One of the algorithms I have been working on\researching is a random city seeding algorithm (I am interested in procedural MMO world development) that takes either a pregenerated world map or proceedurally generated world map and scores the "desirability" of terrain. (In full disclousure I have been working on a MMO from a design standpoint for about 3 years) The real dirty work is in the optimizations of data storage and hard core mathmatics in optimizing game logic for execution efficency. The basics of an MMO, front end or back end are rather simplistic. The real detail is in the backend which are largely proprietary. DAOC\Warhammer uses Gamebryo (same front end framework that Civ4 uses.) The front end of an MMO is relativly canned.
Large-scale balancing and other major changes should be limited to a few people, less than a hundred. If you limit most contributors to only making new quests and dungeons, it might work. Perhaps, however, an MMO could be made to work. If you were to open-source a game without a strong player base with strict ideas of what belongs in the game and what does not, you will end up with a jumbled mess of ideas. These were cut not because they were themselves bad, but because they conflicted with the other elements of the game. When making Wolfenstein 3D, they originally implemented things like dragging corpses into corners and searching through pockets. It takes experience and judgment, knowing not only what to add but what NOT to add. Game design is not, as most people imagine, a simple task.
I will not, however, be making it an open-source project, because then, instead of one unified artistic direction, there will be dozens, pulling the game in different ways. I'm working on a game myself right now, and I fully plan to release the engine code as open-source. Open-source fails at artistic tasks simply because the end result is designed by a committee, not a single vision. The Linux kernel is one of the most stable ever, Apache is the best web server I know of, and Firefox is my preferred browser. Open-source works well for technical tasks.
Game design is an art, and a complex one at that. Open-source clones are often superior, purely on technical grounds, but fully original open-source games tend to be less fun than commercial ones. Game design is one of the things open source does not do well.
0 kommentar(er)
