An official test build for Blender 2.71 will be available this week, and the Google Summer of Code starts today!
Ton Roosendaal writes:
Hi all,
Here are the notes from today's meeting in irc.freenode.net #blendercoders
1) Release status
- Planning didn't change, release targets same.
- Meeting agrees on moving to BCon4 (bugfix only) period, but to make it at least last two weeks. Everyone should focus on bug tracker!
- An official test build will be made available mid this week.
2) Other Projects
- Google Summer of Code officially starts monday. Ton Roosendaal will send the reminder mail to the SoC developer list.
- OpenSubdiv: not much progress to mention this week. For Linux things work, for Windows and OSX we have to wait for libraries to be compiled and added in svn.
Short meeting!
Laters,
-Ton-
15 Comments
"Google Summer of Code officially starts monday. Ton Roosendaal will send the reminder mail to the SoC developer list."
http://yeeeeeeeeeeeeeees.com/ <--- You know what to do.
Good luck to all during Google Summer of Code.
What are the Blender GSOC projects this year?
Just because it's you:
http://www.blendernation.com/2014/04/22/google-summer-of-code-selected-blender-projects/
Here's the details of the proposed projects:
http://code.blender.org/index.php/2014/05/google-summer-of-code-2014/
They're all deserve our full thanks, but I think the two particular headliners are the Interactive Quadrilateral Remeshing that's interactive and provides better quads of topology (similar to ZBrush's ZRemesher), as well as modernization of Blender's NURBS tools (a long-awaited development).
With OpenSubdiv almost ready, I have a question about the older Catmull Clark implementations in Blender.
Has anyone else noticed that the Subsurf and Multires modifiers appear to do their subdividing in different ways? I think that the Multires modifier preserves mesh volume much more than the Subsurf modifier does... meshes really contract a lot more with Subsurf than Multires. This makes me curious about how it works.
I did not notice that. I have to look into this.
What I would love is also the ability to make the mesh go through the vertex points instead of shrinking.
Would be really awesome when you have a 3D scan and you try to rebuild it.
Well it's true that the multires modifier does not perfectly interpolate through the vertices; true interpolation probably carries a performance penalty to calculate. But the edges of separate split meshes stay much closer together with multires than they do with the subsurf modifier. So there are some curiosities and the answers don't show up when I search for them in the wiki pages. Could it simply be that Multires treats all edges as if they were partially creased?
Incidental update: there's so much that we forget about how a tool does what it does after using it for some time. I just now remembered that to subsurf the default cube with the resulting lines going through the vertices, this would actually cause the mesh to grow in volume (due to the equally spaced right angles.) So it really is a bit of a dark art at times, just to figure out what results are the most ideal. Pixar Opensubdiv is going to be such a gift to the community.
what happened to Fluid Designer?
I thought it was to be released with 2.70!
it was released with 2.70 but the new release is coming out with 2.71 and also you have to be in the beta program to receive it (i am ) i asked Mr Andrew peel,
"We will have a new version available soon. We just wanted to make sure all of the parametric object creation tools are working well before we release the next build, but after that we will release regular weekly updates."
This i think is a good reason for the perceived delay.
plus the "delay" is making me want it more ;)
Something about openVDB integration ?
from ton own mouth"OpenVDB is half finished , needs to be picked up but is not highest priority. "
source
http://gooseberry.blender.org/the-beacon-of-hope/#comments
i recall bretch post comment in response guy developing a mesher for particles fluid sim
"I don't think we can accept a patch in the form of a python addon with a
C++ module at the moment. This would need to be integrated into our
build system to compile the module and OpenVDB for all platforms, and
that really makes it a core feature rather than an addon.
One of the Google summer of code projects intends to integrate
OpenVDB for volume rendering in Blender. Once that is implemented it
might be possible to access OpenVDB from an addon, but really I think a
feature like this should be natively in the Blender C/C++ for best
integration and performance."
source https://developer.blender.org/T35565
i visited personal website of GSoc student Rafael Campos and he mention nothing about further progress or work. i guess he hit a road block .
check also fallowing http://blender.45788.x6.nabble.com/Re-Volume-Rendering-for-GSoC-td107608.html
i dont understand why some development is kept in the dark.
i alway hope to read about progress on developer meeting notes.
hope this help
what I would have liked to see is an update to the Remesh Modifier, because although the new remesher that is coming with the GSOC project this year, I have no idea just exactly how that will work but I have a fair estimate that it will work similar to auto retopo tool in 3d Coat just with less options and working in a different way, this sort of remesher would be awesome but I definitely know how the dynamesh works, and I feel that when sculpting a dynamesh style of remesher would be more beneficial as you don't need any user input to set it up, you just need to click remesh, this style of remesher would be awesome for remeshing during the sculpting procedure.
During the sculpting phase, when using Dynamic topology, your mesh can become very polygon heavy, and since the dynamic topology doesn't replace he model with a new triangulated model, it just triangulates the original model leaving you with undesirable topology especially if your model was created really quickly and had undesirable topology in the first place, your sculpts can quickly become very problematic very quickly, for instance if you original model had poles , they would remain there indefinitely and you cant for the life of blender remove them with the smooth brush, which is where hand retopology comes in, since the remesh modifier is useless, you cant use it instead during the sculpting phase, you have to hand retopologise the model which could take 5-0 hours at a time, and if you have to do tis several times during a sculpt, a simple mode hat should have taken 1-2 days quite quickly takes a week or more to get finished on the sculpting, dynamesh would have removed the need to do hand retopology, so I do believe that an upgraded remesh modifier would have been an awesome thing to have.
However the new remesh modifier coming in GSOC 2014 could also be used, its just that remesh modifiers like these take longer to compute which is why Z remesher never took over Dynamesh because it takes longer and is more of a final remesh, something that you would do near the end, but it still could be used instead but would take longer to compute each time.
Just because the remesh modifier is not very good with complex models, doesn't mean that it can not be made better and upgraded to the point that it could produce better quality results.
Well anyways, I can always dream lol
first of I am only goibng to say that what I am going to ay ight ot be possible due to limitations with togays hardware or the implementation of the feature I am going to talk about.
but anyways all ideas start out with a dream and if it were not for dreams we would not have advanced as far as we have .
so here I go, do you guys know the blender game engine.
Do you guys know the freestyle render engine.
fantastic, that makes everything that much easier.
Right my idea is to merge the freestyle render engine with the blender game engine, I mean how cool would that be, yeah the blender game engine is not the best, but freestyle is amazing, and with textured strokes it can make renders look like drawings. Just image that technology added to the blender game engine, blenders game engine would be one of the first game engines to have a cool none photorealistic render that I know of that is. I can just imagine playing a game that really looks like a work of art.
I mean good looking games aren't everything.
But there is a problem, I don't thing that it would be possible due to that render engine not being real time.