Image painting: in Windows the first pressure value can be too high for
some tablets. Now it ignores that first value if it is >= 0.99. I'm not
sure if this is always the case here, so needs further testing.
commited temporary fix: executable name is quoted for all platforms except Windows now,
nicely wrapped in #ifdefs. Will be doing nice wrapper function BLI_system for system calls in blenlib
after release.
Please test on all platforms!
All of these I think still need work but they at least do not coredump.
Basically they need some troubleshooting.
I'm heading home now so will work on it again tomorrow hopefully.
If anyone is bored feel free to take a look at them and see if you can
find the problems. (dnr.c also needs to be converted to float aware)
Kent
Report had a very strange crash, which in the end appeared to be using
a filename of 165 chars long. This gave two bugs;
- the readBlog() function failed on that
- even worse: the struct Global still stored it in G.sce with 160 chars!
This is definitely an important issue... and (for the devs reading the
64 bits discussion) really not something allmighty coders are needed for!
:)
<sob>the 'time cursor' for bake-render has to go... it is being called
inside of a thread (even whilst thread is mutex locked), but that's not
supported in X11. It might be even instable in OSX/Windows even...
Only way to bring it back is to have the main loop (not in thread) update
the time cursor like each second. Would still mean to add a counter var...
will think it over. First want to have confirmed this is stable.
Composite: when using multiple scene render-nodes, and one of these
scenes got re-rendered (by making scene active temporary), the composite
cache should free the used buffers.
Now, on each render, all scenes in a Blender project are being checked.
Commit of patch #5385, to make forward cycling more
user-controllable. Previously it only worked on one axis,
which was auto-detected from movement. This allows forward
cycling to work in more situations, such as stair
stepping.
Differentiated properties and "game properties" in Object docs.
Also the new NLA/Pose key docs were added in the property class instead of Object, tsk tsk.
Order of freeing of Blender database is important, but the past year(s)
new database entries were added in wrong places, so freeing could go in
wrong order. In this case it was Groups being freed before Objects. But
there are more errors...
It only shows really in Windows, since it's the OS that is sensitive for
writing data in freed memory. For other OSes it doesn't make Blender more
stable, only mem-frees happened, no mem-allocs.
Displace modifier now increases user count for Texture it links with.
Note: user counters for objects should not be increased. This is handled
differently (delete object = clear modifiers)
Compositing, depenency fix, so nodes execute only when really changed.
(Error was in nodes that don't get buffers as input, but only values. These
nodes flooded "execute" events through the node tree.)
Python scripts: when setting a layer for an object, the function assumed
the object had to be in the current scene, thus ignoring layers for other
objects.
Now the object->lay value is set always.
fixes.
gamma.c made it return B_PLUGIN_VERSION instead of a hard coded 4.
I haven't had time to look at the other plugins but I'm guessing
the deal was B_PLUGIN_VERSION was still 3 and gamma.c was float aware
so it was bummped to 4. Now that B_PLUGIN_VERSION is 5 I'm guessing
the other plugins need to be updated so they are at level 5 as well.
(float aware) I'm going to work on it later today. At home with a
sick kid. This is all in reference to bug #5692
Kent
Moved function used for 'Visual' keyframing, to live with the rest of the
API calls I added for Pose-Mode bone snapping. Now, it is in a more
'releasable' state.
In a few cases, it may be helpful to turn off the 'only show objects that are
in visible scene layers in the nla' trick. By default though, this is still on. Find
the switch in the View menu of NLA editor.
[ #5458 ] Some transforms are not accurate while pressing shift key
Shrink/Fatten & Push/Pull: Shift slowed the transformation but didn't add precision. (integer division. fixed by applying the same solution used in Grab)
Warp: Shift was ignored. Fixed by changing warp to use the same input method has To Sphere. That is, the 3D view acts like a big horizontal slider: Left -> No effect, Right -> Full effect. This is somewhat of a "big" change from how warp used to work but this is more predictable and is not limited by the original position of the mouse pointer. (of course, this works with Shift correctly too).
Render: Ambient Occlusion was being triggered now with the Material flag
(Shader panel) "Shadow". Nice idea, but that was not so in 2.42a and
before, breaking how previously renders looked.
Now the "Shadow" button in Shaders panel only triggers shadow again.
(Note: the Ambient slider in material triggers AO)
