Page 2 of 2
Re: PureBasic 5.00 beta 7 is out !
Posted: Thu Oct 25, 2012 2:42 pm
by Guimauve
Fred wrote:No, i don't think we have changed something on linux at all.
It's very strange that the move from Beta 6 to Beta 7 reveal a problem like this. Maybe an update from the GTK libs at the same time provoke this. Anyway, I have study 2 or 3 months ago the possibility to track all event on the GLAreaGadget() require a redraw and doing it only if required.
Thanks for your quick response.
Best regards
Guimauve
Re: PureBasic 5.00 beta 7 is out !
Posted: Thu Oct 25, 2012 4:55 pm
by jesperbrannmark
small humble request for future betas:
the purelibraries have a userlibraries folder...
this is a empty folder but i have lots of stuff in it, so when i install a new beta i always need to fiddle with directories...
why not just skip this folder and when pb starts first time make the directory if it doesnt exist...
Re: PureBasic 5.00 beta 7 is out !
Posted: Sun Oct 28, 2012 1:39 pm
by Axeman
EDIT: Ok, I've posted this request to the 'Feature Requests and Wishlists' forum as suggested by Fred. The link to the new post is below.
http://www.purebasic.fr/english/viewtop ... =3&t=51922
---
Hiya. I've gone over the docs fairly thoroughly and can't seem to find any commands for obtaining collision feedback information. The nearest I could find was EntityCollide which just confirms whether two specified entities have collided, but gives no information about the collision. This is a little sub-optimal.
If two entities collide it's necessary to first find out that those two entities have collided without testing each entity against every other entity. Blitz3D provides a collision list for each collision enabled entity, for example. Once you've discovered which entities have collided with which other entities then it's necessary to obtain information about the collision, such as the position of the collision, the collision normals, etc. For a physics based collision you also need to know how much force was involved in the collision, or at least need to have enough information available to calculate it.
Even if the physics engine is handling how the colliding bodies react physically, there's still a great many other actions that may need to be taken if two entities collide. Things such as placing a 3D sound at the collision point with a volume that reflects the force of the collision, for example. Or calculating impact damage for two colliding game objects. This all requires decent feedback about the collision event.
Seems like this is an area that's been neglected, and it really makes Purebasic useless for a wide variety of games and other projects - which is unfortunate.
Possibly the best approach for providing collision event information would be a collision event que (a linked list containing structured variables) where collision information is deposited in collision event structures. Some information that would need to be provided would be the identifiers for the two colliding entities, the position the collision occurred at, the normal direction for the collision, and the linear and angular forces involved in the collision in the case of a physics based collision (bearing in mind that different sounds, character damage, etc may need to be applied and modulated depending on the nature of the forces involved).
Re: PureBasic 5.00 beta 7 is out !
Posted: Sun Oct 28, 2012 5:46 pm
by Fred
Better post such stuff in feature sub-forum so it can be tracked, as it's no help here.
Re: PureBasic 5.00 beta 7 is out !
Posted: Mon Oct 29, 2012 6:59 pm
by Samuel
Sorry if this is the wrong section to post this in, but I have a quick question on the new beta7. Was there a change to procedures? It seems like all of my procedures no longer work correctly. The last version I was using was beta4 and everything seemed to work fine on that version.
Re: PureBasic 5.00 beta 7 is out !
Posted: Mon Oct 29, 2012 7:41 pm
by netmaestro
Post a sample of a procedure that no longer works in Coding Questions along with your opinion that a bug in b7 might be responsible. If this turns out to be the case, the team will move it to the Bug section and fix it.