Page 1 of 2

BF-Windows/Linux (2005-12-20) CVS+AutoKeyAvail+FasterSoft

Posted: Thu Nov 24, 2005 6:04 pm
by ZanQdo
Linux: ( 2005-12-08 ) AutoKeyAvail ... show&id=62

Win: ( 2005-12-20 ) AutoKeyAvail+FasterSoft ... show&id=63

Patched with:
- 10x faster softbody colitions by bjornmose
- AutoKeyAvailable by IGN traker

CVS patched with IGN's auto key available option in the preferences pannel, it makes auto-key to key only existing curves just as the "Avail" option in the insert key menu.

Posted: Fri Nov 25, 2005 10:14 am
by Falgor
Sweet macaroni that's awesome!
(but takes some time to calculate) WOW!

Posted: Fri Nov 25, 2005 11:15 am
by tv
now - if mergetools would preserve UV's - that would be awsome.

Posted: Fri Nov 25, 2005 8:47 pm
by malefico
How do you do an open GL render with this now ? :?

Posted: Fri Nov 25, 2005 10:38 pm
by ZanQdo
I have no idea, I was thinking the same when I try it...

Someone? :?

Posted: Fri Nov 25, 2005 10:43 pm
by blazer003
Good, FREAKING, Mother of mercy. That monkey collision thing is awesome.

Posted: Sun Nov 27, 2005 11:00 pm
by Glaurung

thanks all the coders for your work. I want to report a possible bug. If it is not or if this is not the place to report, I'm sorry. I've searched on about, but I don't find anything. (I'm talking about a bug using testingbuild 24-11-2005)

This happens when I use a colorband in the material settings: When I render static particles using guidelines as a field force, there appears on the low area of the particle system some black squares. If I use unified render, those appears as a very little white stars.

The guidelines has nothing to do with the problem, it occurs also whitout him.

If the emitter has not color band, there are'nt any black squares on the render. The particle emitter is simply a half uvsphere with particle effect default settings and nor=0.110

I'm running a pentium IV at 2'4 ghz, 512 ram on a asus motherboard. And a Canopus Xplode graphic board. Sadly still using Windows 98 SE, until I migrate some day to linux. And Direct X installed, if there are something to do.

I hope this will help you. For questions -or "to noob recomendations" ;-)-, i'm up to you. thank you.


Posted: Mon Nov 28, 2005 10:24 am
by LetterRip
Hi that rendering bug I'm pretty sure has been fixed in the past few days, try to download a more recent test build (rc1 should be in the forums today or tomorrow), and cofirm whether or not it is fixed, if not then please report to the bug tracker.


Posted: Mon Nov 28, 2005 11:00 am
by zupermonkey
Yeah, the white/black dandruft is fixed in the build from this topic.

problem reporting bugs

Posted: Mon Nov 28, 2005 2:10 pm
by Glaurung

the problem with the small black squares still occurs. I've downloaded the build from graphicall following the link in this thread.

i've tried to report bugs but, like other times some months ago, the register and login process is too confusing. I've seen again this message:

Invalid Password Or User Name

Cookies must be enabled past this point.

Due to the move to a new server and switch to ssh, we were not able to migrate the hash tables with passwords.
This means that for a first login here, you have to use the "Lost your password?" option below, which will send you an email with a URL allowing you to re-insert your password.
Sorry for the inconvenience!

For the usage of CVS/SSH authorised keys, see


I don't know CVS working and things like that. I'm only a user that wants to help developpers. Reporting bugs must be a more easy take. With his rules and process, but more easy. If i've created an account here (in the developper forums), should I create another for reporting bugs? Some months ago I create an account but it don't works, like now. I tell you, the black squarism occurs again, if someone can waste his time to investigate, i'm up to him.

Posted: Mon Nov 28, 2005 4:55 pm
by ZanQdo
Building rigth now, lets see if it still hapens, just give me a second :)

Edit: done

Posted: Mon Nov 28, 2005 8:57 pm
by LetterRip

post a link to the blend file and I'll test and file the bug for ya, If you can reduce the amount of 'stuff' in the file, that would make the debugging go faster also.


Posted: Mon Nov 28, 2005 10:30 pm
by lguillaume
ZanQdo wrote:I have no idea, I was thinking the same when I try it...

Someone? :?
Richie's patch is a WIP, he used the bouton to make the render, I send his a patch to his code to enable all render with the same button, I don't know when he will send his new patch with less bug, and border render enable in any view.

Richie's old patch and my add for the old opengl render, how it works: if you push the button on the view3d header with:
- nothing, you make in OpenGL render (preview)
- Ctrl, you make an animation in OpenGL render
- Shift, you make a render with the render engine in the render panel
- Ctrl+Shift, you make an animation render with the render engine.

Code: Select all

Index: source/blender/src/header_view3d.c
RCS file: /cvsroot/bf-blender/blender/source/blender/src/header_view3d.c,v
retrieving revision 1.144
diff -u -r1.144 header_view3d.c
--- source/blender/src/header_view3d.c	27 Nov 2005 20:49:25 -0000	1.144
+++ source/blender/src/header_view3d.c	28 Nov 2005 21:14:18 -0000
@@ -3909,9 +3909,8 @@
 		allqueue(REDRAWHEADERS, 0);
-		if (curarea->spacetype==SPACE_VIEW3D) {
-			BIF_do_ogl_render(curarea->spacedata.first, G.qual!=0 );
-		}
+		if ((curarea->spacetype==SPACE_VIEW3D)&&((G.qual & LR_SHIFTKEY)==0)) BIF_do_ogl_render(curarea->spacedata.first, G.qual!=0 ); 
+        else View3DRender(G.qual & LR_CTRLKEY);
 		if (select_area(SPACE_VIEW3D)) {
Index: source/blender/src/renderwin.c
RCS file: /cvsroot/bf-blender/blender/source/blender/src/renderwin.c,v
retrieving revision 1.60
diff -u -r1.60 renderwin.c
--- source/blender/src/renderwin.c	24 Nov 2005 21:53:51 -0000	1.60
+++ source/blender/src/renderwin.c	28 Nov 2005 21:11:45 -0000
@@ -60,7 +60,7 @@
 #include "MEM_guardedalloc.h"
 #include "BMF_Api.h"
+#include "DNA_camera_types.h"
 #include "DNA_view3d_types.h"
 #include "DNA_screen_types.h"
 #include "DNA_vec_types.h"
@@ -1138,6 +1138,63 @@
 			ofsx += stepx;
+void View3DRender(int anim){
+ Object	 oldobj;
+ Camera	 oldcam;
+ Camera *cam;
+ float   save_dist;
+ short   save_x, save_y;
+ short   save_persp;
+ save_x=  G.scene->r.xsch;
+ save_y=  G.scene->r.ysch;
+ save_persp= G.vd -> persp;
+ G.scene->r.xsch=  curarea->winx;
+ G.scene->r.ysch=  curarea->winy;
+ save_dist= G.vd->dist;
+ if(!G.vd -> camera) return; /* we need one cam for the trick */
+ /* render current view */
+ memcpy(&oldobj, G.vd -> camera, sizeof(Object));
+ memcpy(&oldcam, G.vd -> camera -> data, sizeof(Camera));
+ cam= G.vd->camera->data;
+ if (G.vd->persp==0) {
+ 	cam->type = CAM_ORTHO;
+ 	cam->ortho_scale= G.vd->dist*2;
+ 	G.vd->dist = 1000.0;
+ 	cam->clipsta= 1000.0 - 0.5*G.vd->far;
+ 	cam->clipend= 1000.0 + 0.5*G.vd->far;
+  	}
+ else  {
+ 	cam->type = CAM_PERSP;
+ 	cam->lens= G.vd->lens/2;
+ 	cam->clipsta= G.vd->near;
+ 	cam->clipend= G.vd->far;
+ 	}
+ G.vd -> persp = 2;
+ setcameratoview3d();
+ scrarea_do_windraw(curarea);
+ BIF_do_render(anim);
+ /* back to cam settings */
+ memcpy(G.vd -> camera, &oldobj, sizeof(Object));
+ memcpy(G.vd -> camera -> data, &oldcam, sizeof(Camera));
+ G.vd -> persp = save_persp;
+ G.scene->r.xsch= save_x;
+ G.scene->r.ysch= save_y;
+ G.vd->dist = save_dist;
+ scrarea_do_windraw(curarea);
 /* -------------- API: externally called --------------- */

Posted: Mon Nov 28, 2005 10:59 pm
by ZanQdo
Thanks! I will use that :shock:

BTW, graphicall is down, it should be up in a few minutes

Edit: yup up again 8)

Posted: Tue Nov 29, 2005 12:20 pm
by raygun11
I think there is a bug in view rendering, when camera is tracked to an object.
look at this

result of rendering is different from the view in 3D View, and also after rendering, view in 3D View changes ( to match the result in render window).