[FTEQW] rev 5041Traceline MOVE_HITMODEL crash with non-bones
Posted: Fri Jan 20, 2017 9:48 am
Hi guys, I append here a bug trace that makes FTE crashes when you make a traceline in FTE v 5041 using MOVE_HITMODEL and pointing to a IQM model with no bones (a static mesh for example)
This is the file responsible for the crash, line 6978 and 7211 (I tested every other file and this is the one who, reverted to 5040, makes FTE not crash anymore). In my very very humble opinion, here it should be put a if block that says: "if IQM is static mesh, then jump all the events block" and then add a qc.field like "modelType", which gamedevs need to set to MODEL_IQM_TYPE_SKELETAL or MODEL_IQM_TYPE_STATIC, to speed up the process. This could be also useful because this will led to something I always loved to see: a system to manage 2 types of static mesh: 1 for appeareance (the model players see) and 1 for collision (very very low poly that's used by traceboxes for ramps, stairs, etc) and engine managing totally by itself.
I tried to contact Spike but with no success, and now we're at revision 5044, so I thought it was better to repair it now before revisions become newer and newer with a lot of efforts on IQM events(which they're great, by the way).
At the present time any of the new releases (until 5044) doesn't solve the crash
This is the file responsible for the crash, line 6978 and 7211 (I tested every other file and this is the one who, reverted to 5040, makes FTE not crash anymore). In my very very humble opinion, here it should be put a if block that says: "if IQM is static mesh, then jump all the events block" and then add a qc.field like "modelType", which gamedevs need to set to MODEL_IQM_TYPE_SKELETAL or MODEL_IQM_TYPE_STATIC, to speed up the process. This could be also useful because this will led to something I always loved to see: a system to manage 2 types of static mesh: 1 for appeareance (the model players see) and 1 for collision (very very low poly that's used by traceboxes for ramps, stairs, etc) and engine managing totally by itself.
I tried to contact Spike but with no success, and now we're at revision 5044, so I thought it was better to repair it now before revisions become newer and newer with a lot of efforts on IQM events(which they're great, by the way).
At the present time any of the new releases (until 5044) doesn't solve the crash