Thought the community might be interested in some historical information about how the MVC2 project was kicked off. This first part is related to building the game from the Dreamcast source code. I may add more info depending on community reaction.
Email chains posted below (read from bottom up).. I've edited only to remove very specific sensitive information:
-----------------------------------------
From: I have received the Dreamcast source code - you can grab from [details removed]. Caveats from Capcom This is the Japanese version of the game. The network code is based on the DC modem and KDDI matching service. The game also shipped with a different set of locked characters and a purchase system that required points from playing online and in the arcade in order to unlock items. I already looked this up, full details here - includes info on how to hex edit the save game file to unlock all 56 characters, 7 stages, and level. - From: Gary-san, But, another e-mail regarding the source code request for DC(JP) version came from R&D. The person who controls all archive is asking specific reason or plan how we utilize DC Japanese source. His questions are; - He only have Japanese version source code and wondering is it useful to our developer. - DC(JP) version has linked feature that use DC Visual memory to unlock certain characters and is not in PS2 and Xbox version. - If developer use the code from DC may cause the issue that certain character could not unlocked. - Also network component is designed for Modem and sync logics may be outdated. If you have any idea to explain would be helpful. Reason such as DC has less bug means more complete game than Xbox version that we plan to use for the base of porting etc. I understand that DC version is useful or not can not be determined unless our programmer review the code. Thank you for your help. ----------------------------------------- From: > We think that Japanese version data is not needed to port > you have US version source code. Please clarify the reason you need the data. The PS2 and Xbox versions of the game contain bugs and defects that make these versions not as desirable as the Dreamcast and Naomi versions. We would actually prefer the Dreamcast home version of the game - if this is available, arcade source code would not be necessary. > We supplied PS2 and Xbox version source backup and porting target is for PS3 > and Xbox360 download delivery. Of above situation we think you should utilize > PS2 and Xbox source materials, not necessary have Japanese version. We plan to base the port on the Xbox version of the source. However, the arcade version would be very helpful to use as reference, and it would aid bug-fixing. There may be reasons to analyze the original arcade code to determine how online networking was added. The arcade code will be used as reference and encryption will not be needed for the port. If the Dreamcast source was available, this would be better. - ----------------------------------------- From: Hi, I'm sending e-mail back and forth with CJ character rights department to get additional backup for Below is several question came back from CJ and they are saying without additional information, they are not willing to supply Arcade version source backup because of that backup is for Japanese version and they don't have US version backup other than they already supplied. Questions from character rights department: - We think that Japanese version data is not needed to port - We supplied PS2 and Xbox version source backup and porting target is for PS3 and Xbox360 download delivery. Of above situation we think you should utilize PS2 and Xbox source materials, not necessary have Japanese version. Comments from Software engineering group: - We are wondering why you need AC version source code, the project is for consumer platform since the beginning. - We also afraid that you are not understand the complexity of the difference between - Why asking AC version instead of DC version? (Kubozono: This caused by some miscommunication, I already answered.) Addition to above, I mentioned this MVC2 project is important to us and explained we would like to have any available materials regardless the languages. Christian-san, At this moment, CJ does not want supply Japanese version backup material. I can get more backup materials, If CJ satisfied with our answer and plan. Thank you ----------------------------------------- From: Great, thanks for the quick response. I can explain that the project is Marvel vs Capcom 2 (www.system16.com/hardware.php?id=800&pag...). The contract we have [details removed] stipulates that this will be a direct port to XBLA and PSN, with all graphics and gameplay remaining the same as the Xbox1 home version. The main exceptions are that we will be supporting online multiplayer, adding stats tracking, and updating the front-end to MS/Sony standards. We can provide the source archive for the PS2/Xbox1 version of the game. This is derived from the original arcade code, but has stubs for some online multiplayer (that was pulled out of the game at the last minute due to support issues). I have enclosed a specific ‘source review’ NDA, which you’ll need to fill out and send back [details removed] For online multiplayer, we would want you to use a middleware library that we have licensed - see GGPO technology (ggpo.net/blog/) for details. It would be great if the additional stats tracking required could be done through the normal Leaderboards/Ranking scoreboards, but if not, we also have a license for GameSpy available. I don’t think the game side port from Xbox1 will give you much trouble. We will provide a list of bugs that should be fixed in the original, but this should not be too significant. The bulk of the busy work will be finding and replacing all menus with TRC/TCR compliant versions, adding Achievements, etc. I expect the online multiplayer will be the most significant task. GGPO works very well for hiding latency, but it requires that the game state be known - game responds to joystick input with only 1 frame fixed delay, game state progresses using basic prediction, joystick data is transferred over the net, if there is a mistake in the prediction the game state must be rewound to the last known good frame. Based on the scope of work, I am looking for a developer that has the following types of experience - Arcade hardware and/or emulation, particularly Dreamcast/Naomi/Atomiswave - Reverse engineering skills - XBLA and PSN game development with good submission track record - Network multiplayer I’ll be up front about this - the timeline requires that the game must be on digital store shelves by [details removed] 2009, so I’m cautiously looking for other options. It would be helpful if you could provide any detail on experience or skills related to the above areas. Thank you, -
