From: Radu B. R. <ru...@cs...> - 2006-07-11 19:21:11
|
Okay then. Time to make a new interface! :) I think it would be safe to begin with: #define PLAYER_3DPOINTCLOUD_MAX_POINTS ... (what would be an appropriate value here?) #define PLAYER_3DPOINTCLOUD_DATA_STATE 1 typedef struct player_3dpointcloud_data { uint32_t points_count; player_point_3d_t points[PLAYER_3DPOINTCLOUD_MAX_POINTS]; uint32_t points_colors_count; player_color_t colors[PLAYER_3DPOINTCLOUD_MAX_POINTS]; } player_3dpointcloud_data_t; #define PLAYER_3DPOINTCLOUD_GET_GEOM_REQ 1 typedef struct player_3dpointcloud_geom { player_pose3d_t pose; // or, as Raymond suggested in a previous e-mail (with PLAYER_3DPOINTCLOUD_MAX_POSES // defined earlier somewhere): // // uint32_t poses_count; // player_pose3d_t poses[PLAYER_3DPOINTCLOUD_MAX_POSES]; } Please feel free to commend, fix or add additional structures/definitions. Brian Gerkey wrote: > On Jul 9, 2006, at 11:30 PM, Radu Bogdan Rusu wrote: > >> Right now, I am looking at something that spits out arrays of X,Y,Z >> and Graphics3D seems >> like a good candidate. I was already thinking when I sent the e- >> mail yesterday, that I >> wouldn't mind a player_3dmap interface, which could be even more >> specific for these kinds >> of tasks. I have no problem maintaining it. > > I agree with Toby and Geoff that a new interface is warranted here. > It's tempting to reuse an existing interface, but when the new use is > quite different from the original intent of the interface, confusion > is sure to follow. PS. Brian, you can remove the graphics3d patch from SF then. Thanks. Best, R. -- | Radu Bogdan Rusu | http://rbrusu.com/ | http://www9.cs.tum.edu/people/rusu/ | Intelligent Autonomous Systems | Technische Universitaet Muenchen |