Re: [Bf-committers] Trying to build the PACKAGE.vcxproj using VS 2013
Hi, The cmake rpm builder has worked for a while now and I do actively maintain it -- except when one of the major dependencies of blender are out of sync with the distro version *cough*python*couch* and I really don't feel like borking my whole system...again. Not that this has anything to do with the current discussion but thought I'd mention it in case someone was looking for some code to trim while making extensive changes to the cmake buildsystem. Dan On Sat, Nov 15, 2014 at 12:50 AM, Martijn Berger martijn.ber...@gmail.com wrote: Hi Tom, I am having trouble reproducing this, I have made extensive changes to the cmake buildsystem over the last week and would recommend updating the blender source and regenerating the solution and retrying it. For me both the NSIS and the WIX packers work and so does building the PACKAGE target. The package target is not really needed to do blender development and did not work properly prior to my changes anyway. Martijn On Fri, Nov 14, 2014 at 9:10 PM, Tom Dominique tom.domini...@gerbertechnology.com wrote: Hello Everyone, I am new to Blender and I am trying to build using VS2013. I am using Cmake 3.0.2, NSIS 3.0b and git. I am able to create a blender.sln and I am able to build almost all of the 135 projects that were created. The only project I can not build is the one I believe I need, and that is PACKAGE.vcxproj. I have attached my error message that I am receiving. Any help would greatly be appreciated. Thanks, Tom 1-- Build started: Project: PACKAGE, Configuration: Release x64 -- 1 CPack: Create package using NSIS 1 CPack: Install projects 1 CPack: - Install project: Blender 1 CMake Error at c:/BlenderDEV/BuildPath/source/creator/cmake_install.cmake:41 (message): 1ABSOLUTE path INSTALL DESTINATION forbidden (by caller): 1c:/BlenderDEV/BuildPath/bin/Release/2.72/scripts 1 Call Stack (most recent call first): 1c:/BlenderDEV/BuildPath/cmake_install.cmake:35 (include) 1 1 1EXEC : CPack error : Error when generating package: Blender 1C:\Program Files (x86)\MSBuild\Microsoft.Cpp\v4.0\V120\Microsoft.CppCommon.targets(132,5): error MSB3073: The command setlocal 1C:\Program Files (x86)\MSBuild\Microsoft.Cpp\v4.0\V120\Microsoft.CppCommon.targets(132,5): error MSB3073: cd c:\BlenderDEV\BuildPath 1C:\Program Files (x86)\MSBuild\Microsoft.Cpp\v4.0\V120\Microsoft.CppCommon.targets(132,5): error MSB3073: if %errorlevel% neq 0 goto :cmEnd 1C:\Program Files (x86)\MSBuild\Microsoft.Cpp\v4.0\V120\Microsoft.CppCommon.targets(132,5): error MSB3073: c: 1C:\Program Files (x86)\MSBuild\Microsoft.Cpp\v4.0\V120\Microsoft.CppCommon.targets(132,5): error MSB3073: if %errorlevel% neq 0 goto :cmEnd 1C:\Program Files (x86)\MSBuild\Microsoft.Cpp\v4.0\V120\Microsoft.CppCommon.targets(132,5): error MSB3073: C:\Program Files (x86)\CMake\bin\cpack.exe -C Release --config ./CPackConfig.cmake 1C:\Program Files (x86)\MSBuild\Microsoft.Cpp\v4.0\V120\Microsoft.CppCommon.targets(132,5): error MSB3073: if %errorlevel% neq 0 goto :cmEnd 1C:\Program Files (x86)\MSBuild\Microsoft.Cpp\v4.0\V120\Microsoft.CppCommon.targets(132,5): error MSB3073: :cmEnd 1C:\Program Files (x86)\MSBuild\Microsoft.Cpp\v4.0\V120\Microsoft.CppCommon.targets(132,5): error MSB3073: endlocal call :cmErrorLevel %errorlevel% goto :cmDone 1C:\Program Files (x86)\MSBuild\Microsoft.Cpp\v4.0\V120\Microsoft.CppCommon.targets(132,5): error MSB3073: :cmErrorLevel 1C:\Program Files (x86)\MSBuild\Microsoft.Cpp\v4.0\V120\Microsoft.CppCommon.targets(132,5): error MSB3073: exit /b %1 1C:\Program Files (x86)\MSBuild\Microsoft.Cpp\v4.0\V120\Microsoft.CppCommon.targets(132,5): error MSB3073: :cmDone 1C:\Program Files (x86)\MSBuild\Microsoft.Cpp\v4.0\V120\Microsoft.CppCommon.targets(132,5): error MSB3073: if %errorlevel% neq 0 goto :VCEnd 1C:\Program Files (x86)\MSBuild\Microsoft.Cpp\v4.0\V120\Microsoft.CppCommon.targets(132,5): error MSB3073: :VCEnd exited with code 1. == Build: 0 succeeded, 1 failed, 0 up-to-date, 0 skipped == ___ Bf-committers mailing list Bf-committers@blender.org http://lists.blender.org/mailman/listinfo/bf-committers ___ Bf-committers mailing list Bf-committers@blender.org http://lists.blender.org/mailman/listinfo/bf-committers ___ Bf-committers mailing list Bf-committers@blender.org http://lists.blender.org/mailman/listinfo/bf-committers
Re: [Bf-committers] Trouble to build blender-2.72 on Fedora Rawhide
Looks like a missing include from pulling in the COLLADA's COLLADAFWInstanceBindingBase.h. Perhaps adding an #include vector to source/blender/collada/ArmatureImporter.h will fix it, can't really test right now but that's what I'd try first. Dan On Tue, Sep 30, 2014 at 9:19 AM, Jochen Schmitt joc...@herr-schmitt.de wrote: Hello, I have tried to build the last official release of blender (blender-2.72) for Fedora Rawhide on the Fedora build system. Unfornately this build fails with several errors. Because I don't have an idea how to fix the issues I would like to ask for any hint to solve the issues. You can access the log for the builds on [1] [2] Best Regards: Jochen Schmitt [1] http://koji.fedoraproject.org/koji/taskinfo?taskID=7720882 [2] http://koji.fedoraproject.org/koji/buildinfo?buildID=581566 ___ Bf-committers mailing list Bf-committers@blender.org http://lists.blender.org/mailman/listinfo/bf-committers ___ Bf-committers mailing list Bf-committers@blender.org http://lists.blender.org/mailman/listinfo/bf-committers
Re: [Bf-committers] Can hair particles be removed from particle edit mode?
Select the particle you want gone and press the 'x' key? On Thu, Apr 3, 2014 at 12:38 AM, gandalf3 zzyx...@gmail.com wrote: Hair particles can be added in particle edit mode with the add brush, but I can't seem to find a way to remove them. Can they be removed? And if not, why not? I recently posted a question about this on stackexchange http://blender.stackexchange.com/q/8264/599 with more details about what I've tried so far. -- -gandalf3 ___ Bf-committers mailing list Bf-committers@blender.org http://lists.blender.org/mailman/listinfo/bf-committers ___ Bf-committers mailing list Bf-committers@blender.org http://lists.blender.org/mailman/listinfo/bf-committers
Re: [Bf-committers] Revisions for the 2.70a release
On Tue, Mar 25, 2014 at 10:32 AM, Sergey Sharybin sergey@gmail.comwrote: You can create a branch based on the tag. And we never did testbuilds of 'a' releases anyway. Since we're talking about branches and tags... I noticed that if you do make a branch (or just checkout the tag) and then try to build with CMake git complains about the branch having no upstream or somesuch and doesn't produce a commit hash. Not that big a deal but it breaks CPack since the Release tag doesn't properly parse. Dan ___ Bf-committers mailing list Bf-committers@blender.org http://lists.blender.org/mailman/listinfo/bf-committers
Re: [Bf-committers] Import, research on speed improvements
On Sun, Mar 9, 2014 at 12:21 PM, Fredrik hansson fredrikhansson_12...@yahoo.com wrote: hi, i have recently written a obj loader myself just for fun. now i did try to optimize it quite a bit by making it multithreaded and a few other things. snip // Fredrik Kind of curious how multi-threaded parsing would keep the verts in order since the order they're defined in the file is where the face indices come from? I'm pretty tempted to write a C++ .obj parser -- I've written a few python ones in the last few years to test out parsing algorithms since the grammar is simple and there's plenty of example files out there. The last one (testing my py3k+metaclass conversion of spark.py, the earley parser from the python source tree) was something like 100x slower than Campbell's handwritten one but that's to be expected from a slowish algorithm that can parse pretty much any grammar you throw at it. Been wanting to play with Boost::Spirit so maybe I can knock out something fairy quickly using my (mostly complete) grammar from spark.py. Dan ___ Bf-committers mailing list Bf-committers@blender.org http://lists.blender.org/mailman/listinfo/bf-committers
Re: [Bf-committers] Install deps script not working on Fedora 20
Probably the same issue blender was having where fedora 20 doesn't do the -mt suffix on the boost libs anymore. Dan On Mon, Dec 30, 2013 at 7:52 PM, Daniel Salazar - patazstudio.com zan...@gmail.com wrote: This is another log he pasted on IRC which I think is more useful http://paste.fedoraproject.org/65050/55349138/ This looks suspicious? CMake Error at /usr/share/cmake/Modules/FindBoost.cmake: (message): Unable to find the requested Boost libraries. Daniel Salazar patazstudio.com On Mon, Dec 30, 2013 at 8:39 PM, James Wrigley jwrigle...@gmail.com wrote: Hi there, When running the install_deps script on F20 (with all_static set to true, otherwise untouched), it fails at building OIIO. I've pasted the cmake logs here: http://www.pasteall.org/48373/bash. If I run the script with --skip-oiio it fails with the same error when building OSL. Any ideas? Thanks, JamesNZ ___ Bf-committers mailing list Bf-committers@blender.org http://lists.blender.org/mailman/listinfo/bf-committers ___ Bf-committers mailing list Bf-committers@blender.org http://lists.blender.org/mailman/listinfo/bf-committers ___ Bf-committers mailing list Bf-committers@blender.org http://lists.blender.org/mailman/listinfo/bf-committers
Re: [Bf-committers] Fedora 20 linking broken in cmake on libboost_filesystem
Yep, that fixed it Thanks, Dan ___ Bf-committers mailing list Bf-committers@blender.org http://lists.blender.org/mailman/listinfo/bf-committers
[Bf-committers] Fedora 20 linking broken in cmake on libboost_filesystem
Hi all, The fine folks at redhat never fail to disappoint and broke the build again with this update to the boost libs: * Fri Jul 26 2013 Petr Machata pmach...@redhat.com - 1.53.0-12 - There's no physical difference between single-threaded and multi-threaded builds, except some libraries are only built in multi-threaded mode. So build everything in multi-threaded mode, and ditch tagged layout, which we don't need anymore. https://bugzilla.redhat.com/show_bug.cgi?id=971956 The fix is to set Boost_USE_MULTITHREADED to OFF -- suppose we need a flag for that now instead of having it hardcoded to ON. Dan ___ Bf-committers mailing list Bf-committers@blender.org http://lists.blender.org/mailman/listinfo/bf-committers
Re: [Bf-committers] Git migration preview repo available
On Sat, Nov 2, 2013 at 5:08 AM, Campbell Barton ideasma...@gmail.comwrote: @dalai, probably its based on emails, the only info git has is from: https://gitorious.org/blender-git-migration/blender-git-migration/raw/authors.txt Hello, I abandoned my old domain so this should be (assuming it doesn't break anything): dna = Dan Eicher d...@eu.phorio.us thanks Dan ___ Bf-committers mailing list Bf-committers@blender.org http://lists.blender.org/mailman/listinfo/bf-committers
Re: [Bf-committers] Inventing Pseudo House Printer Technology
Always wanted me a pseudo house... Anyways, you'll probably get more interest starting a thread on blenderartists.org since that's the main hangout for the artist types. Dan PS: I also live in downtown Phoenix and haven't seen any of your structures in my travels so am guessing this is a new thing? ___ Bf-committers mailing list Bf-committers@blender.org http://lists.blender.org/mailman/listinfo/bf-committers
Re: [Bf-committers] SIGGRAPH observations
snip MULTITHREADED DEPENDENCY GRAPH Pixar right now uses basically one character per thread, then background caches frames for playback to keep cores occupied. It's kind of a workaround, but ensuring fast playback this way is quite nice for animators. They showed their Presto animation software, with fast playback, opensubdiv and hair deformation on the GPU, baked ptex applied to the mesh and lighting and shadows from the key light. All looked quite nice. They are looking into finer granularity too but don't have it yet. Dreamworks is very granular, graphs with 50k-150k nodes. Requires careful design of rigs, but they have very good scaling. Overhead from granularity is reduced by using TBB, and letting threads handle chains of nodes without going through the task system. Pixar uses a system where there is a very clear separation between output data and the depsgraph, for evaluating multiple frames at the same time and to reduce locking, this is something we want in Blender too. They also compile the depsgraph in advance, and Dreamworks caches networks for changing various values. It's unclear if this will help in Blender, maybe with quite complex rigs. I think it's best to leave this as an optimization to solve when it shows up as a performance problem. This paper from the Technicolor folks is two clicks away from the SIGGRAPH papers link so may (or may not) have been noticed already: http://jmarvie.free.fr/Publis/2013_WEB3D/manyCoreEventEvaluation.pdf Two level event scheduler for a multithreaded depsgraph they claim results in near-linear speedups. Dan ___ Bf-committers mailing list Bf-committers@blender.org http://lists.blender.org/mailman/listinfo/bf-committers
Re: [Bf-committers] FCurve Python update patch
Hi, I did a quick check and it does indeed look like a simple matter of missing update calls. It's an easy fix but it's too close to release to be messing around with something like this so... Dan On Wed, Jul 17, 2013 at 10:57 AM, Tom Edwards cont...@steamreview.orgwrote: Here's a script: import bpy o = bpy.context.active_object if not o.animation_data: o.animation_data_create() a = o.animation_data.action = bpy.data.actions.new(Test Action) c = a.fcurves.new(data_path=location,index=2) c.keyframe_points.add(3) c.keyframe_points[0].co = [0, 0] c.keyframe_points[1].co = [4, 1.5] c.keyframe_points[2].co = [8, 0] This is what I see (note the orange handle line): http://img46.imageshack.us/img46/8609/aath.png Editing the curve or even starting an edit then cancelling it causes the handles to pop into the correct position. On 17/07/2013 5:46, Dan Eicher wrote: HI, Do you have some code that shows how this fails because it really should happen automagically, probably just a missing update event somewhere. Dan ___ Bf-committers mailing list Bf-committers@blender.org http://lists.blender.org/mailman/listinfo/bf-committers ___ Bf-committers mailing list Bf-committers@blender.org http://lists.blender.org/mailman/listinfo/bf-committers ___ Bf-committers mailing list Bf-committers@blender.org http://lists.blender.org/mailman/listinfo/bf-committers
Re: [Bf-committers] FCurve Python update patch
HI, Do you have some code that shows how this fails because it really should happen automagically, probably just a missing update event somewhere. Dan ___ Bf-committers mailing list Bf-committers@blender.org http://lists.blender.org/mailman/listinfo/bf-committers
Re: [Bf-committers] Please turn off Auto Run Python Scripts by default
There's a compile time flag that does this if you're into that sort of thing -- WITH_PYTHON_SECURITY I believe it is -- that the linux distros use it for their packaged blender builds Well, except fedora who uses WITH_PYTHON_SAFETY which still auto-runs scripts and gives you much slower python scripts. Filed a bug report a while back but they don't seem to care. Dan ___ Bf-committers mailing list Bf-committers@blender.org http://lists.blender.org/mailman/listinfo/bf-committers
Re: [Bf-committers] Blender developer weekly meeting notes - May 26, 2013; Leap Motion
On Fri, May 31, 2013 at 12:39 PM, Campbell Barton ideasma...@gmail.comwrote: Mine also arrived, I plugged in and the firmware update immediately ran (without asking) and failed to update with error, now the light wont turn on. Searching online and apparently theres some way to repair failed firmware update but I'll need access to a windows system. Possibly my own fault for extracting the deb package and running on an unsupported Linux disro (64bit Arch Linux). However all the applications run, just firmware update problem. Sort of the same here on Fedora, fails with 'Unable to find firmware to install' which I suppose is a little better than bricking it by trying to install bad firmware. The only thing I installed was the udev file since the Leap app wouldn't connect to the device without proper permissions (I'm guessing) and all the programs run fine like you said it just doesn't actually *do* anything. Other than that it makes for a pretty paper weight at the very least. Dan ___ Bf-committers mailing list Bf-committers@blender.org http://lists.blender.org/mailman/listinfo/bf-committers
Re: [Bf-committers] Blender developers meeting minutes, April 21 2013
do you plan to do a full python binding to voro++ ? (Every class, every function and so on ? Yeah...mostly. Some of the stuff doesn't really make sense to call from python (only used internally by the lib) while other stuff isn't really nice from the python side like voronoicell::face_vertices (which needs to be parsed to get the actual vert indexes). But when I'm done is should be a fairly reasonable approximation of the C++ api with some 'bonus' bits thrown in to make it easier to use from the python side. Stuff like automagic python sub-classed wall classes callable from C++ (like from wall_list and container classes) already works. Dan ___ Bf-committers mailing list Bf-committers@blender.org http://lists.blender.org/mailman/listinfo/bf-committers
Re: [Bf-committers] Edit mode - additional selection mode
Something like L key while hovering over a mesh? Dan On Sat, Apr 6, 2013 at 2:49 PM, Hadrien Brissaud hadris...@gmail.comwrote: Hello, I would like to suggest adding a selection mode to the existing vertex, edge and face : element. This would allow for selecting independant, unconnected groups of faces - mesh islands, if you wish. This option exists in 3dsmax and I have long missed it, first in Maya, then in Blender. I always find myself selecting a face and then growing selection (ctrl+plus) a number of times to dodge this situation. If this addition proved difficult to implement or for another reason was out of the question, I suggest adding element or mesh island to the Shift-G (select similar) menu. Thank you, Hadrien ___ Bf-committers mailing list Bf-committers@blender.org http://lists.blender.org/mailman/listinfo/bf-committers ___ Bf-committers mailing list Bf-committers@blender.org http://lists.blender.org/mailman/listinfo/bf-committers
Re: [Bf-committers] Platform maintainers: update of collada libraries
Don't know if it's related but collada building is broken using fedora 18 repo libs due to DocumentImporter.cpp:128:72: error: cannot allocate an object of abstract type 'ExtraHandler' On the plus side it's finding the libs ok. Dan ___ Bf-committers mailing list Bf-committers@blender.org http://lists.blender.org/mailman/listinfo/bf-committers
Re: [Bf-committers] Looking for insight into a problem with setting up a modifier
Looks like you didn't add MOD_kkshapekeymodifier.c to the build system so it doesn't get compiled. Dan ___ Bf-committers mailing list Bf-committers@blender.org http://lists.blender.org/mailman/listinfo/bf-committers
Re: [Bf-committers] Minimal Blender specs - 5 year old systems OS
IDK, I think a lot of users get hand-me-downs and there seems to be a trend of lower power (energy efficient) computer being made these days. Not to mention if blender ever gets fully ported to tablet-like devices. It might be better to target OS versions rather than hardware specs since computers haven't really advanced that much over the last five years other than throwing more cores at the problem. Something produced five years ago can still do 99% of what the average user needs and this target audience is probably a lot bigger than the amount of people that can plunk $1k down on a upscale box. As an aside, both of my working computers (a netbook 5+ year old laptop) don't fall within these specs and aren't really good at *using* blender but are still good for testing. Dan ___ Bf-committers mailing list Bf-committers@blender.org http://lists.blender.org/mailman/listinfo/bf-committers
Re: [Bf-committers] OSL Composite Node
Hi all, Didn't get quite as much done as I'd hoped over the last week but have to go back to work so won't be able to touch this for 2 - 3 weeks -- here's my progress so far if someone else wants to mess with it: www.pasteall.org/38787/diff It doesn't actually work yet (but doesn't crash blender either which is says something), only works with 'External' .oso files, the shader and searchpath:shader are hard coded and to get the node to show the sockets you need to do a node.update() from the py-console. Isn't really a whole lot left to get it working but rent must be paid... Dan ___ Bf-committers mailing list Bf-committers@blender.org http://lists.blender.org/mailman/listinfo/bf-committers
Re: [Bf-committers] OSL Composite Node
Couple OSL questions... From reading the code it looks like in order to set the individual ShadingSystem::Parameters() one needs to clear/redo the shader for every pixel shading operation. Seems like it's setup to shade a whole lot of points using the same parameters and not really made to update them in the middle of a shading pass. What I'd like to do is load the shader and then update the parameter input values on-the-fly, is this even possible? Hmm... totally forgot what my other question was. Dan ___ Bf-committers mailing list Bf-committers@blender.org http://lists.blender.org/mailman/listinfo/bf-committers
[Bf-committers] OSL Composite Node
Hi all, During the Developer Meeting Ton suggested that someone look into creating a comp node based on OSL so I've been (semi-)diligently working on that the last couple of days. So far I have done all the plumbing and am ready to tackle the trick bits of hooking into OSL. From what I can find on the interwebs the way to do it would be like in OSL/testshade ( https://github.com/imageworks/OpenShadingLanguage/tree/master/src/testshade) where you set up a 'fake' plane the size of the image (or tile I suppose) and evaluate the x, y points on said plane after passing in the input/output socket data as parameters. While this is good and all for complex algorithms it does seem a bit overkill for simple pixel filter functions (which the IRC discussion proposed). Something like this would be super trivial using libjit (which is the only other similar library I'm familiar with) so I'm hoping someone knows a method to just call a function using OSL without having to setup a shading context. Ton said he'd read on the OSL website that it is also designed for stuff like this so I'm guessing that it must be possible without too much trouble. Anyhoo, off to plunge into the deep, dark depths of the OSL codebase. Dan ___ Bf-committers mailing list Bf-committers@blender.org http://lists.blender.org/mailman/listinfo/bf-committers
Re: [Bf-committers] OSL Composite Node
On Tue, Jan 8, 2013 at 4:32 PM, Brecht Van Lommel brechtvanlom...@pandora.be wrote: As I understand it, libjit doesn't actually come with any language compiler, it's just a way to generate instructions manually. So I'm not sure how using that would be trivial, LLVM gives you the same functionality as libjit, but the actual implementation of OSL on top of LLVM is still a lot of code? Yeah, one would have write a parser to use libjit, not too terribly hard but... What I really meant was a way to call the compiled functions which it seems is basically what OSL::ShadingContext is, a wrapper around the llvm function calling mechanism with extra stuff tacked on for OSL globals and whatnot. If you're going to allow only basic pixel processors, i.e. just reading and writing to pixels in the same location then it fits quite well already. You can have basic input/output parameters of shaders and it could run for each pixel. If you want to do things like blurring it gets more complicated because there is no concept of image buffers to read/write to, so some sort of mechanism for that would need to be implemented. From the IRC discussion it was decided not to do any complex things like blurs since there's no guarantee that the pixels you want to look up actually exist due to the way the compositor works with multi-threading and tiles. So just simple single pixel stuff like relighting a scene (Ton's example), technicolor, film-grain (I guess would work?) etc. Dan ___ Bf-committers mailing list Bf-committers@blender.org http://lists.blender.org/mailman/listinfo/bf-committers
Re: [Bf-committers] Bug CompositorNodeDilateErode
On Sat, Jan 5, 2013 at 12:06 PM, Sergey Sharybin sergey@gmail.comwrote: Hrm, this is pretty nasty issue -- dilate node has got own type which overrided base node's type. This indeed need to be solved, but afraid it'll break some scripts. What's also interesting for me here -- is it more like a feature or makesrna should notify about such kind of conflicts? That's just regular python isn't it? If you redefine a property in a subclass then you'd have to use super().type to get at the base class property. The most obvious solution is to rename CompositorNodeDilateErode.type to something that describes it like 'foo_type'. Dan ___ Bf-committers mailing list Bf-committers@blender.org http://lists.blender.org/mailman/listinfo/bf-committers
Re: [Bf-committers] [Bf-extensions-cvs] SVN commit: /data/svn/bf-extensions [4138] trunk/py/scripts/addons: Fix [ #33759] Can't use UV image export UV_layout error
This one breaks render_povray/render.py line 1784 (elif with no if) On Fri, Jan 4, 2013 at 9:37 AM, Bastien Montagne montagn...@wanadoo.frwrote: Revision: 4138 http://projects.blender.org/scm/viewvc.php?view=revroot=bf-extensionsrevision=4138 Author: mont29 Date: 2013-01-04 16:37:47 + (Fri, 04 Jan 2013) Log Message: --- Fix [#33759] Can't use UV image export UV_layout error Another glitch from recent alpha work. Also modified povray's code. As usual, maintainders of these scripts, please double check changes are ok! :) Modified Paths: -- trunk/py/scripts/addons/io_mesh_uv_layout/export_uv_png.py trunk/py/scripts/addons/render_povray/render.py Modified: trunk/py/scripts/addons/io_mesh_uv_layout/export_uv_png.py === --- trunk/py/scripts/addons/io_mesh_uv_layout/export_uv_png.py 2013-01-04 07:57:26 UTC (rev 4137) +++ trunk/py/scripts/addons/io_mesh_uv_layout/export_uv_png.py 2013-01-04 16:37:47 UTC (rev 4138) @@ -113,7 +113,7 @@ # scene render settings scene.render.use_raytrace = False -scene.render.alpha_mode = 'STRAIGHT' +scene.render.alpha_mode = 'TRANSPARENT' scene.render.image_settings.color_mode = 'RGBA' scene.render.resolution_x = image_width Modified: trunk/py/scripts/addons/render_povray/render.py === --- trunk/py/scripts/addons/render_povray/render.py 2013-01-04 07:57:26 UTC (rev 4137) +++ trunk/py/scripts/addons/render_povray/render.py 2013-01-04 16:37:47 UTC (rev 4138) @@ -1697,15 +1697,16 @@ if not world.use_sky_blend: # Non fully transparent background could premultiply alpha and avoid anti-aliasing # display issue: -if render.alpha_mode == 'PREMUL': +if render.alpha_mode == 'TRANSPARENT': tabWrite(background {rgbt%.3g, %.3g, %.3g, 0.75}\n % \ (world.horizon_color[:])) #Currently using no alpha with Sky option: elif render.alpha_mode == 'SKY': tabWrite(background {rgbt%.3g, %.3g, %.3g, 0}\n % (world.horizon_color[:])) #StraightAlpha: -else: -tabWrite(background {rgbt%.3g, %.3g, %.3g, 1}\n % (world.horizon_color[:])) +# XXX Does not exists anymore +#else: +#tabWrite(background {rgbt%.3g, %.3g, %.3g, 1}\n % (world.horizon_color[:])) worldTexCount = 0 #For Background image textures @@ -1776,10 +1777,11 @@ # for skysphere..5.5 tabWrite(gradient y\n) tabWrite(color_map {\n) -if render.alpha_mode == 'STRAIGHT': -tabWrite([0.0 rgbt%.3g, %.3g, %.3g, 1]\n % (world.horizon_color[:])) -tabWrite([1.0 rgbt%.3g, %.3g, %.3g, 1]\n % (world.zenith_color[:])) -elif render.alpha_mode == 'PREMUL': +# XXX Does not exists anymore +#if render.alpha_mode == 'STRAIGHT': +#tabWrite([0.0 rgbt%.3g, %.3g, %.3g, 1]\n % (world.horizon_color[:])) +#tabWrite([1.0 rgbt%.3g, %.3g, %.3g, 1]\n % (world.zenith_color[:])) +elif render.alpha_mode == 'TRANSPARENT': tabWrite([0.0 rgbt%.3g, %.3g, %.3g, 0.99]\n % (world.horizon_color[:])) # aa premult not solved with transmit 1 tabWrite([1.0 rgbt%.3g, %.3g, %.3g, 0.99]\n % (world.zenith_color[:])) ___ Bf-extensions-cvs mailing list bf-extensions-...@blender.org http://lists.blender.org/mailman/listinfo/bf-extensions-cvs ___ Bf-committers mailing list Bf-committers@blender.org http://lists.blender.org/mailman/listinfo/bf-committers
Re: [Bf-committers] Linux package maintainers
The arch people do make a valid point, it really isn't that hard to produce a tarball when doing the 'release ahoy' announcement. I asked about this before (on the IRC maybe?), why not just do a svn export and have everyone build from that instead of using specific svn tags/versions? Everyone's happy and it occasionally finds issues like building blender outside a svn checkout was producing wonky version numbers on the splash screen for a while there. For most software I need to update for whatever reason I take the official .spec file, bump the version numbers and build/install the package(s) but the few times I tried that with the blender.spec it didn't work out so well. Since blender updates deps a lot faster than distros this happens more often than you'd think. So I'd like to propose that producing a source tarball become an official part of the release process. As an aside, it would probably help to 'spread the love' if packagers *cough*fedora*cough* didn't complain when you file bug reports against their packages. Dan ___ Bf-committers mailing list Bf-committers@blender.org http://lists.blender.org/mailman/listinfo/bf-committers
Re: [Bf-committers] Linux package maintainers
Looking through http://download.blender.org/release/ the most recent source tarball was blender-2.58-source.tar.gz. Unless they live somewhere else? Dan ___ Bf-committers mailing list Bf-committers@blender.org http://lists.blender.org/mailman/listinfo/bf-committers
Re: [Bf-committers] Linux package maintainers
On Wed, Jan 2, 2013 at 2:06 PM, Thomas Dinges blen...@dingto.org wrote: The sources are here: http://download.blender.org/source Wait, so what's the issue again? ...though perhaps there should be a link from the 'Download Blender' page because if you clink that link and it's not there then it must not exist. The only other arch complaint I see is about the letter suffixes (which do indeed wreck havoc on packaging systems) but that was discussed in depth previously. Dan ___ Bf-committers mailing list Bf-committers@blender.org http://lists.blender.org/mailman/listinfo/bf-committers
Re: [Bf-committers] [Bf-blender-cvs] SVN commit: /data/svn/bf-blender [53515] trunk/blender: add option WITH_SYSTEM_BULLET to link against the bullet installation found on the system .
On Wed, Jan 2, 2013 at 5:23 PM, Campbell Barton ideasma...@gmail.comwrote: Revision: 53515 http://projects.blender.org/scm/viewvc.php?view=revroot=bf-blenderrevision=53515 Author: campbellbarton Date: 2013-01-03 00:23:52 + (Thu, 03 Jan 2013) Log Message: --- -- snip -- if(WITH_PYTHON_MODULE) add_definitions(-DPy_ENABLE_SHARED) -endif() +endif(). #- # Extra compile flags Looks like an extra '.' slipped in there, not sure if cmake cares or not though. Dan ___ Bf-committers mailing list Bf-committers@blender.org http://lists.blender.org/mailman/listinfo/bf-committers
Re: [Bf-committers] Python Access to some properties of nodes
On Sat, Dec 22, 2012 at 12:21 AM, Lukas Tönne lukas.toe...@gmail.comwrote: IMHO storing the links in a separate list like we do now is one of the more well-designed code parts ;) It may be slightly inconvenient if you want to get a list of links for a specific socket, but it has several advantages over storing lists of links in each socket or node: * If each socket stores its own link list that means a lot more overhead for all those lists. Now we have exactly 1 list. * You only ever need a single loop. If each socket has a list of links, to walk over all links in a node tree (e.g. for drawing or the cut operator) you need to walk over all nodes and then all sockets and then all links ... * Storing a link in sockets is ambiguous, you could store the link either in the output or input socket. If you store it on one side only you have to loop over everything again to find links of the other side. If you store it in both you have to keep them in sync, which is error prone and leads to redundant data. * In terms of time complexity having to filter the list of all links vs. a reduced list of socket links does not make much difference, both are essentially O(N+M), where N is number of nodes, M is number of links. In general though the node DNA should not be judged by performance criteria too much, if you need to access the node structure millions of times (e.g. in rendering) you should create an optimized structure with arrays instead of linked lists and can store links in whatever way is needed for fast access. All modern node systems in Blender (cycles, compositor) do this. Bottom line: the single links list is the best all-round solution for the purpose of storing data in .blend files and user manipulation. I was really only talking about how it's exposed to python, not rewriting the way nodes work internally. Script writers shouldn't need to know (or care) how things work inside blender, they just need a quick and easy way to figure out which socket links to which other socket(s) and having to iterate over a list of links and compare with the socket they're trying to find isn't really 'quick and easy'. Basically all I'm saying is move the links from a property of the Node to the NodeSocket where it should really be. Dan ___ Bf-committers mailing list Bf-committers@blender.org http://lists.blender.org/mailman/listinfo/bf-committers
Re: [Bf-committers] Python Access to some properties of nodes
On Fri, Dec 21, 2012 at 4:52 AM, Lukas Tönne lukas.toe...@gmail.com wrote: # List of connected (Node, NodeSocket) pairs for 'socket': input_sockets = [(link.from_node, link.from_socket) for link in socket.id_data.links if link.to_socket == socket] output_sockets = [(link.to_node, link.to_socket) for link in socket.id_data.links if link.from_socket == socket] Somewhat off topic... Looking at this code I'm starting to think that maybe it was a mistake to expose Node.links instead of having it as a property on the sockets themselves, it would make the code a lot clearer and it isn't like you can really do anything with the links independent of the sockets they're assigned to. It may even be better to totally abstract away the NodeLinks too, maybe socket.links just return a list of (Node, NodeSocket) tuples -- have to think on this a bit more though since it would probably require sub-classing NodeSocket into NodeInputSocket, NodeOutputSocket so you'd know if the returned node/socket pair was an input or output for the node. Would also clean up node.links.new(add?) a bunch. Dan ___ Bf-committers mailing list Bf-committers@blender.org http://lists.blender.org/mailman/listinfo/bf-committers
Re: [Bf-committers] Python Access to some properties of nodes
Assuming Campbell's OK with it it wouldn't be too terribly hard to add something like Socket.link to push the iterating over the links into C since IIRC that's how it works internally. I personally won't have time to mess with it until after New Years but if someone else wants to bang their head against makesrna and node tree structs it might turn out to be as simple as wrapping an existing function (though I do seem to recall that the node links were a little tricky to wrap originally). Dan ___ Bf-committers mailing list Bf-committers@blender.org http://lists.blender.org/mailman/listinfo/bf-committers
Re: [Bf-committers] Issues to build blender-2.65 on Fedora 18
Try adding -DBOOST_ROOT=/usr to your cmake command line. /me grumbles about inappropriate flags again... Dan On Wed, Dec 12, 2012 at 2:04 PM, Jochen Schmitt joc...@herr-schmitt.dewrote: On Wed, Dec 12, 2012 at 05:29:16PM +0100, Brecht Van Lommel wrote: On Wed, Dec 12, 2012 at 12:20 PM, Jochen Schmitt joc...@herr-schmitt.de wrote: On Wed, Dec 12, 2012 at 10:52:20AM +0100, Brecht Van Lommel wrote: Could you try editing line 656 in the top CMakeLists.txt: set(__boost_packages filesystem regex system thread date_time) To change the order of the packages, for example: set(__boost_packages filesystem regex thread date_time system) I have tried out the following patch without any success. Ok, this particular change won't work, because systema is not a valid package name. Maybe no order works and it's a different issue. Anyway it could help if you uploaded the CMakeCache.txt, and maybe also the output from make VERBOSE=1 so we can see the exact linker command that is used. It should be linking to /lib64/libboost_filesystem-mt.so.1.50.0, since the filesystem package is in the list, but somehow the error indicates that it isn't. OK, Here the output of the related error messages after a build with make VERBOSE=1: Best Regards: Jochen Schmitt /usr/lib64/ccache/c++-Wall -Wno-invalid-offsetof -Wno-sign-compare -Wlogical-op -Winit-self -Wmissing-include-dirs -Wno-div-by-zero -Wuninitialized -Wundef -Wmissing-declarations -O2 -g -pipe -Wall -Wp,-D_FORTIFY_SOURCE=2 -fexceptions -fstack-protector --param=ssp-buffer-size=4 -m64 -mtune=generic -fPIC -funsigned-char -fno-strict-aliasing -fopenmp -msse2 -msse -pipe -fPIC -funsigned-char -fno-strict-aliasing -O2 -DNDEBUG -pthread CMakeFiles/blender.dir/creator.c.o CMakeFiles/blender.dir/buildinfo.c.o -o ../../bin/blender -rdynamic ../../lib/libbf_windowmanager.a ../../lib/libbf_editor_space_api.a ../../lib/libbf_editor_space_action.a ../../lib/libbf_editor_space_buttons.a ../../lib/libbf_editor_space_console.a ../../lib/libbf_editor_space_file.a ../../lib/libbf_editor_space_graph.a ../../lib/libbf_editor_space_image.a ../../lib/libbf_editor_space_info.a ../../lib/libbf_editor_space_logic.a ../../lib/libbf_editor_space_nla.a ../../lib/libbf_editor_space_node.a .. /../lib/libbf_editor_space_outliner.a ../../lib/libbf_editor_space_script.a ../../lib/libbf_editor_space_sequencer.a ../../lib/libbf_editor_space_time.a ../../lib/libbf_editor_space_userpref.a ../../lib/libbf_editor_space_view3d.a ../../lib/libbf_editor_space_clip.a ../../lib/libbf_editor_text.a ../../lib/libbf_editor_transform.a ../../lib/libbf_editor_util.a ../../lib/libbf_editor_uvedit.a ../../lib/libbf_editor_curve.a ../../lib/libbf_editor_armature.a ../../lib/libbf_editor_gpencil.a ../../lib/libbf_editor_interface.a ../../lib/libbf_editor_mesh.a ../../lib/libbf_editor_metaball.a ../../lib/libbf_editor_object.a ../../lib/libbf_editor_physics.a ../../lib/libbf_editor_render.a ../../lib/libbf_editor_screen.a ../../lib/libbf_editor_sculpt_paint.a ../../lib/libbf_editor_sound.a ../../lib/libbf_editor_animation.a ../../lib/libbf_editor_datafiles.a ../../lib/libbf_editor_mask.a ../../lib/libbf_editor_io.a ../../lib/libbf_render.a ../../lib/libbf_intern_opennl.a ../../lib/libbf _python.a ../../lib/libbf_python_ext.a ../../lib/libbf_python_mathutils.a ../../lib/libbf_python_bmesh.a ../../lib/libbf_ikplugin.a ../../lib/libbf_modifiers.a ../../lib/libbf_bmesh.a ../../lib/libbf_compositor.a ../../lib/libbf_opencl.a ../../lib/libbf_blenkernel.a ../../lib/libbf_nodes.a ../../lib/libbf_gpu.a ../../lib/libbf_blenloader.a ../../lib/libbf_imbuf.a ../../lib/libbf_blenlib.a ../../lib/libbf_intern_ghost.a ../../lib/libbf_intern_string.a ../../lib/libbf_avi.a ../../lib/libbf_imbuf_cineon.a ../../lib/libbf_imbuf_openexr.a ../../lib/libbf_imbuf_dds.a ../../lib/libbf_collada.a ../../lib/libbf_intern_bsp.a ../../lib/libbf_intern_elbeem.a ../../lib/libbf_intern_iksolver.a ../../lib/libbf_intern_memutil.a ../../lib/libbf_intern_guardedalloc.a ../../lib/libge_blen_routines.a ../../lib/libge_converter.a ../../lib/libge_phys_dummy.a ../../lib/libge_phys_bullet.a ../../lib/libbf_intern_smoke.a ../../lib/libextern_minilzo.a ../../lib/libextern_lzma.a ../../lib/libextern_co lamd.a ../../lib/libge_logic_ketsji.a ../../lib/libextern_recastnavigation.a ../../lib/libge_phys_common.a ../../lib/libge_logic.a ../../lib/libge_rasterizer.a ../../lib/libge_oglrasterizer.a ../../lib/libge_logic_expressions.a ../../lib/libge_scenegraph.a ../../lib/libge_logic_network.a ../../lib/libge_logic_ngnetwork.a ../../lib/libextern_bullet.a ../../lib/libge_logic_loopbacknetwork.a ../../lib/libbf_intern_moto.a ../../lib/libge_videotex.a ../../lib/libbf_rna.a ../../lib/libbf_dna.a ../../lib/libbf_blenfont.a ../../lib/libbf_intern_audaspace.a ../../lib/libbf_intern_mikktspace.a ../../lib/libbf_intern_dualcon.a
[Bf-committers] OSL rpm for Fedora
Packaged up OSL and threw it up on the Review Request tracker ( https://bugzilla.redhat.com/show_bug.cgi?id=885357). Not really sure how the whole package review process works so I figured I'd poke the list here where I'm a known entity since there seems to be a few hoops to jump through. Dan ___ Bf-committers mailing list Bf-committers@blender.org http://lists.blender.org/mailman/listinfo/bf-committers
Re: [Bf-committers] Writing a Compositor node plugin?
Did you add your new files to cmake/scons so they get compiled linked? Dan ___ Bf-committers mailing list Bf-committers@blender.org http://lists.blender.org/mailman/listinfo/bf-committers
Re: [Bf-committers] Load an external python file
The easiest way would be to put your scripts in a directory and add a __init_.py then: import sys sys.path.append('/path/to/cycles/cycles_module') from cycles_module import utils Fedora, in their infinite wisdom, doesn't add /usr/local/lib/python2.7/site-packages/ to they python path so I have to do this quite often. Dan ___ Bf-committers mailing list Bf-committers@blender.org http://lists.blender.org/mailman/listinfo/bf-committers
Re: [Bf-committers] Can BGE be relicensed?
On Sun, Dec 2, 2012 at 11:55 AM, Tom M letter...@gmail.com wrote: A second issue is that code ownership isn't solely that of the BF and contributors - technically the original dual licensed code from when blender went open source - the NaN shareholders still have some sort of legal rights. So any other licensing changes other the those transformations allowed under the GPLv2 or the original Blender license could be quite headache inducing. LetterRip I'm guessing The Community didn't purchase the cppyright to the code but merely paid to have it released as GPLv2 then? If that's the case then all these talks on re-licensing are mostly (totally) pointless... Dan ___ Bf-committers mailing list Bf-committers@blender.org http://lists.blender.org/mailman/listinfo/bf-committers
Re: [Bf-committers] Can BGE be relicensed?
On Sun, Dec 2, 2012 at 1:21 PM, Sinan Hassani sinan.hass...@gmail.com wrote: So one idea is to have the BF give people the ability to purchase a license for BGE(Blender) for say $99 US where they can license it under the MIT license. This would generate revenue for the BF, they can have extra money to pay GE developers, and it will attract more developers. This is how the Torque3D model now works, the engine is open source, they make money on services. Once you license it to *one* person under a MIT license it's licensed to *everyone* under a MIT license. I'd just learn how to use (or write an exporter for) a MIT'd game engine myself... Dan ___ Bf-committers mailing list Bf-committers@blender.org http://lists.blender.org/mailman/listinfo/bf-committers
Re: [Bf-committers] [Bf-blender-cvs] SVN commit: /data/svn/bf-blender [51027] trunk/blender/source/blender/imbuf /intern/colormanagement.c: colormanage_colorspace_get_named() can (and does ) return NU
On Wed, Oct 3, 2012 at 11:21 PM, Sergey Sharybin sergey@gmail.com wrote: Function you mentioned should not return null and if if does -- some datablock's color space is configured wrong and that's what should be fixed. This probably prevents crash but currently i don't consider it's a correct fix and would want to have feedback about circumstances when crash happens here. This prevents a crash if the colorspace data file can't be found -- blender just complains about the OCIO environmental variable instead of segfaulting which IMHO is a little more user friendly. To reproduce the crash just move the datafiles/colormanagment folder somewhere blender can't find it. This doesn't fix any color management issues or anything like that, it just checks for a null pointer being returned from a function that *can* return NULL...easy peasy. Dan ___ Bf-committers mailing list Bf-committers@blender.org http://lists.blender.org/mailman/listinfo/bf-committers
Re: [Bf-committers] Python 3.3?
While y'all are messing with the python configuration in the blender.spec maybe you could fix my bug? https://bugzilla.redhat.com/show_bug.cgi?id=821064 And so as to not be totally off-topic, you could get the version and pipe it to cmake like: %define pyver %(%{__python} -c import sys ; print(sys.version[:3])) ... cmake .. -DPYTHON_VERSION:STRING=%{pyver} Dan ___ Bf-committers mailing list Bf-committers@blender.org http://lists.blender.org/mailman/listinfo/bf-committers
Re: [Bf-committers] DragDrop position
Pretty sure you have to do it manually, if you look at the code for the outliner it calculates the 'dropzone' from figuring out what was under the dropped item by looping over the outliner items and comparing the x/y positions. Dan ___ Bf-committers mailing list Bf-committers@blender.org http://lists.blender.org/mailman/listinfo/bf-committers
Re: [Bf-committers] Images in Panels..
On Wed, Aug 1, 2012 at 1:45 AM, Campbell Barton ideasma...@gmail.com wrote: Could you report a bug in gl_load() crashing blender?, I tried to redo this but couldn't get blender to crash. IIRC there's a bug in the mipmaps due to a somewhat recent change that I was meaning to look into but haven't had the time. Don't think I actually confirmed its existence just that someone on either BA.org or the IRC was talking about it. Dan ___ Bf-committers mailing list Bf-committers@blender.org http://lists.blender.org/mailman/listinfo/bf-committers
Re: [Bf-committers] Temporary global storage
On Thu, Jun 21, 2012 at 3:00 PM, Jorge Rodriguez jo...@lunarworkshop.com wrote: This would be perfect but I can't find a place to use it that's not part of an operator. There's no single operator responsible for handling these functions. I'm not able to find any way to attach a SKIP_SAVE property to, for example, wmWindowManager or wmWindow. I'd rather avoid resorting to a static variable in a function. Maybe there's some way to do it with rna that I'm not seeing? If these variables aren't of interest to people in general it's overkill to wrap them in rna and (automagically) expose them in the py-api. Why would anyone care when the last time a tooltip was closed or what the last highlighted button was when writing a script? Dan ___ Bf-committers mailing list Bf-committers@blender.org http://lists.blender.org/mailman/listinfo/bf-committers
Re: [Bf-committers] rigging [was: Re: The Future of Blender Projects WAS meeting notes]
The Big Buck Bunny rigs use the mesh deform modifier which I'd imagine isn't supported by any exporters -- could be wrong though. Basically rig--deform mesh--character mesh so you'd have to export the def mesh to other program def mesh and have the armature drive that. I'm guessing that no game engine uses a harmonic-mesh whatever to drive a fps animated character? Dan ___ Bf-committers mailing list Bf-committers@blender.org http://lists.blender.org/mailman/listinfo/bf-committers
Re: [Bf-committers] Blender File reader SDK (Dan Eicher)
On Mon, Jun 11, 2012 at 1:04 AM, k...@polarlights.net wrote: Thanks a lot Dan, But what are you talking about ? Which C++ code ? Where can I found it or got some information about it ? Thanks Krys ___ Bf-committers mailing list Bf-committers@blender.org http://lists.blender.org/mailman/listinfo/bf-committers Makesrna generates a *huge* C++ wrapper file (RNA_blender_cpp.h) in the build directory (source/blender/makesrna/intern on cmake, not sure what scons does). It doesn't get installed anywhere (anymore) so you either have to do complex include paths into the build dir or move it (them, you need a couple other headers too) by hand. If you make a custom python module and use the C++ api then everything just works, haven't tried outside python so can't say how that works -- blender doesn't have any other way to dynamically load libs so python is pretty much it these days. Just look in intern/cycles to see how the professionals do it... Dan ___ Bf-committers mailing list Bf-committers@blender.org http://lists.blender.org/mailman/listinfo/bf-committers
Re: [Bf-committers] Blender File reader SDK
You could also look into the C++ api generated by makesrna that cycles uses. Dan ___ Bf-committers mailing list Bf-committers@blender.org http://lists.blender.org/mailman/listinfo/bf-committers
Re: [Bf-committers] [Bf-blender-cvs] SVN commit: /data/svn/bf-blender [46643] trunk/blender/source/blender/ makesrna/intern/rna_sequencer_api.c: SequenceElements.pop() -- added 'index ' argument inste
IIRC I just took the code from the copy operator and used that. It'll be a few days before I can get around to this since I'm on the road helping my sister move so not much time. Dan ___ Bf-committers mailing list Bf-committers@blender.org http://lists.blender.org/mailman/listinfo/bf-committers
Re: [Bf-committers] Add dragdrop box for python operator
A third way would be to add a poll() method to operators that, when defined, registers a new dropzone type. Plus a bl_poll type to know into which dropzone list to add it to. Then would just need to have a generic wrapper function (per dropzone region type) to call the py-function like; poll(cls, drug_ob, target_ob) and another generic dropzone_copy() function that copied the drug target objects into op[object] op[target] custom props if poll() returns true...something like that, haven't quite figured out all the details yet. Wouldn't be too terribly hard to implement -- one of the main reasons I've been poking around in the DD code the last couple days is to figure out how to get at this stuff from python. Dan ___ Bf-committers mailing list Bf-committers@blender.org http://lists.blender.org/mailman/listinfo/bf-committers
Re: [Bf-committers] Meanless using of space in skin and simple deform modifiers
On Sat, May 26, 2012 at 1:16 PM, Knapp magick.c...@gmail.com wrote: Negative space is often the key to good design! Indeed... http://en.wikipedia.org/wiki/File:HKBuildingFengshui.jpg ___ Bf-committers mailing list Bf-committers@blender.org http://lists.blender.org/mailman/listinfo/bf-committers
Re: [Bf-committers] Freestyle branch status report May 2012
On Wed, May 23, 2012 at 11:43 PM, Campbell Barton ideasma...@gmail.com wrote: @Tamito, I figured auto generation wasn't really an option here so didn't mention it - of course if it was practical it would be a nice time saver. I wouldn't be too much trouble to come up with some pybindgen scripts to auto-generate the py-api, from what I've seen in the current python bindings there doesn't seem to be any gotchas. The pros: * automagic sub-classing support of C++ classes * can convert C++ getter/setter methods to py-attributes * easily rename C++ methods to make them more pythonic * much simpler to do stuff like __call__ -- operator() mapping, well...automatic really. The cons: * time -- mostly it's just a bunch of grunt work to cleanup the output from the header file scanner tbh * Yet Another Build Dependency * py3k support isn't in the main repo, only in my branch * need to remove python stuff from C++ classes (in the long run this is probably a 'Pro') * need to change base class methods to pure virtual for the python sub-classing to work correctly (ie, not let you create an instance of a pure virtual base class) * tp_iter tp_iternext support is lacking -- could probably be made to work with a little hacking though All in all time is the biggest issue, I can't guarantee to have this done in a timely manner since I'll be taking off the middle of next week for my sister's graduation and to help her move then when I get back I have to bust-ass to make my rent. I do think with a little careful planning it's possible to make a 100% drop in replacement for the current hand-written bindings with some 'free' stuff thrown in like pythreads support and more robust sub-classing plus all the benefits of auto-generation. Dan ___ Bf-committers mailing list Bf-committers@blender.org http://lists.blender.org/mailman/listinfo/bf-committers
Re: [Bf-committers] Freestyle branch status report May 2012
On Thu, May 24, 2012 at 7:01 PM, Tamito KAJIYAMA rd6t-k...@asahi-net.or.jp wrote: Dan, Auto-generation of the Freestyle Python API would be great if that is feasible and practical. There are about 150 Python wrappers of C++ classes, and maintaining them manually has been a kinda painful excercise. Just for your information, I send you a link to a blog post that summarizes problems and implemented solutions for improving the Freestyle Python API: http://freestyleintegration.wordpress.com/2009/08/04/recent-commits-on-the-freestyle-python-api/ I forward this to you, since there were several subtle issues that needed to be addressed carefully. Honestly speaking, for this reason I am a bit reluctant to do a major revision of the present Python API implementation. As implicitly mentioned in the blog post, the C++ class system is quite static (i.e., user extensions of the present class hierarchy are very unlikely). There are a limited number of C++ classes and their methods that are supposed to be subclassed/overriden by Python subclasses (see Section 3.4 of the blog article). Exploiting these characteristics of the C++ class system might help you simplify the Python API auto-generation. Regards, -- KAJIYAMA, Tamito rd6t-k...@asahi-net.or.jp Looking over that blog post, freestyle code and pybindgen (which can indeed do iterators, who knew?) the only thing that would really need to change for it all to work would be adding methods like Interface1D.getVertices() which could be wrapped in python as Interface1D.vertices but would need to return the collection instead of the actual iterator. Then python would do the right thing if one were to write: for vert in i1D.vertices: do_stuff(vert) For the borrowed/owned reference stuff, pybindgen throws an exception if you don't specify that on function args and return values that are pointers or references so either way you have to tell it if it's owned or borrowed, no choice. Finally, for python sub-classing, that Just Works. Python can call C++ (non-pure virtual) base class methods, C++ can call overridden py-methods, everyone's happy. I'm not suggesting you stop work on the current py-api and wait on my slow ass or anything, just seeing if there's any interest in this before starting any work on it. Dan ___ Bf-committers mailing list Bf-committers@blender.org http://lists.blender.org/mailman/listinfo/bf-committers
Re: [Bf-committers] Why is not BLENDER_VERSION_CHAR bumped to a?
On Wed, May 23, 2012 at 4:39 AM, Brecht Van Lommel brechtvanlom...@pandora.be wrote: I've bumped the version character to a now. It's not really clear to me why this would be needed, but it shouldn't hurt either. Brecht. The package build systems turn it into an int so it becomes 2.63.1 while trunk was still at 2.63.0 which makes the packaging system always prefer 2.63a over trunk builds and try to 'upgrade'. Just so you know... Dan ___ Bf-committers mailing list Bf-committers@blender.org http://lists.blender.org/mailman/listinfo/bf-committers
Re: [Bf-committers] Split Collada into a python frontend and a C++ backend
You wound need to write some C helper functions and either add them to makesrna or hand write your own python wrapper to be able to access them from python. A while back I was working on some auto-generated python bindings for collada-dom and all I can say about that is it's not for the faint of heart -- them folks aren't too worried about fixing bugs and the whole spec is a crazy, complex beast. Pretty sure that was the makefile that borked my /home dir so maybe I'm a little gun shy. I did have some good luck with generate-ds which can take the whole spec/xml definition/whateveryoucallit and output a pure python implementation. Not that it has anything to do with the question though. I do tend to agree with the others, there's already the option to turn it off in the build system and if it's already in there why bother hiding it? Dan ___ Bf-committers mailing list Bf-committers@blender.org http://lists.blender.org/mailman/listinfo/bf-committers
Re: [Bf-committers] Proposal on Default Settings
Can't we just not mess with the default settings unless there's a really, really good reason to change something? Seems like every time I want to model something I have to go poking around to figure out how to turn on/off something so it was the same as the last time I used blender and some of the stuff isn't quite as obvious as it seems to find out how to get it back to the way it was. Unless, of course, someone proposes to change the panel collapse to *only* happen when you click on the little triangle because it annoys me to no end to click on blender from another active app and have the all the little panels dance around for no reason -- I'd be all for that change or reversion, whatever you want to call it. Dan ___ Bf-committers mailing list Bf-committers@blender.org http://lists.blender.org/mailman/listinfo/bf-committers
Re: [Bf-committers] Numpy integration
On Mon, Apr 30, 2012 at 8:16 AM, Eugenio Pignataro i...@oscurart.com.ar wrote: I have my python 3.2.3 compiled from source in Ubuntu 12.04. Cmake can't find numpy. My folder : /usr/local/lib/python3.2/site-packages is empty. What I need? Your OS most likely installs the python packages in /usr/lib while cmake is apparently looking for them in /usr/local/lib -- where I'd imagine you have cmake looking for your custom installed python. Probably just a matter of making (or fixing) a FindNUMPY.cmake module (which IMHO all libs should have *cough*ffmpeg*cough*) Dan ___ Bf-committers mailing list Bf-committers@blender.org http://lists.blender.org/mailman/listinfo/bf-committers
Re: [Bf-committers] Movie distortion algorithm
These guys describe a fast preview mode (if that's the problem): http://www.vassg.hu/pdf/vass_gg_2003_lo.pdf ...not that I understand the math behind it. Dan ___ Bf-committers mailing list Bf-committers@blender.org http://lists.blender.org/mailman/listinfo/bf-committers
Re: [Bf-committers] 3DsMax editable poly like toolset
If I understand how this works then having an undo stack wouldn't cut it. From videos it looks like it applies the whole stack (I'm sure optimized in some way) to the base mesh which is why you can go back N steps and change a parameter which then gets applied up the whole chain. Main use case: * add a mesh * make a loop cut * make a loop cut across the new faces * move some of the newly created geometry * delete the first loop cut * ??? * mesh explodes Dan ___ Bf-committers mailing list Bf-committers@blender.org http://lists.blender.org/mailman/listinfo/bf-committers
[Bf-committers] py-constraints proposal
Added a patch to the boneyard with the back end changes needed to implement pyconstraints -- http://projects.blender.org/tracker/index.php?func=detailaid=30903group_id=9atid=127 My plan is to take the currently broken ScriptConstraint and make it work like panels, operators and my super sexy py-listeners (poor, poor dead py-listeners *sniff*) where you sub-class it, register it and it automagically Just Works(tm). Easy, peasy...just don't want to spend a whole lot of time on this only to have it die on the tracker so figured I'd 'propose it' before going any further. Dan ___ Bf-committers mailing list Bf-committers@blender.org http://lists.blender.org/mailman/listinfo/bf-committers
Re: [Bf-committers] node size
Huh, guessed I missed those (width, height) in my big ol' nodetree py-api patch I worked on quite a while ago. To answer your question, yes it would be possible (and easy) to add those -- I just don't have the authority to do stuff like that so... Dan ___ Bf-committers mailing list Bf-committers@blender.org http://lists.blender.org/mailman/listinfo/bf-committers
Re: [Bf-committers] node size
On Mon, Apr 9, 2012 at 2:36 PM, François T. francoistarl...@gmail.com wrote: oh do you have a patch that we can try or submit to someone who has authority ? Nah, it would literally take five minutes (including compile time on my slow-ass netbook) to add but I haven't messed with it yet. While I'm at it : len(scene.node_tree.nodes['RGB'].outputs) = 1 len(scene.node_tree.nodes['RenderLayer'].outputs) = 28 RGB which is a rgb input node return 1 output which is correct. But the RenderLayer return 28 even if I only have one pass selected. Is that normal or is it a bug ? Don't think those are generated on-the-fly so is probably correct, the draw function decides which ones are visible IIRC. Dan ___ Bf-committers mailing list Bf-committers@blender.org http://lists.blender.org/mailman/listinfo/bf-committers
Re: [Bf-committers] User Interface Tasklist Proposal
I think most of the stuff on your list is either a simple fix or could be implemented in a short time with a little python-fu... and a couple are either 'features' (color picker) or already exist (context sensitive menus in addons). Couple comments (in random order): Floating buttons support I could knock that out in an hour (at most) as a python operator using bgl, did something similar with my pie menu script... took a lot longer than an hour though since I had to learn both opengl and some linear algebra to get it to work (and fix some bugs in blf along the way). FPS-style navigation Pretty sure that already exists as 'fly-mode'. Camera view navigation Simple to write as a py-operator... with the above and 'set camera to view' chained together would take like 5 minutes to implement. Non-obvious progress bar Also not too hard, what would be a project would be to unlink the progress bar from the jobs system so anyone could use it... thought I had it once but turns out all I did was py-wrap the OSX specific progress bar on the dock or whatever they call it. Separators do not drag closed I kind of like that they don't close actually. Face-preserving edge delete Think that's called 'dissolve an edge' in bmesh. Right click - rename/delete in the material/texture editors Left clicking on the little text box is really that hard? Right click - cut/copy/paste for text fields As in ctrl+C and ctrl+V? While I'm just an Innocent bystander in all this I really don't see a laundry list of (mostly) simple fixes as in the spirit of the GSOC, I don't think they just want a code monkey for the summer either way. Dan ___ Bf-committers mailing list Bf-committers@blender.org http://lists.blender.org/mailman/listinfo/bf-committers
Re: [Bf-committers] patch tracker
Three things I've learned about getting patches into trunk: 1) you have to pester people to get them to look at your stuff 2) the bug tracker is the most direct route to trunk since people actually care about the bug count 3) the patch tracker's main use is somewhere to put code you're pretty sure will never get into trunk after you give up on #1 and can't frame as a bug fix I'm not complaining or anything (having gotten something like 50 patches into trunk) and this doesn't always hold true but seems like a good general rule of thumb to go by. Not sure there's any easy fix since module owners have real lives that force them to disappear for months at a time and the resident BF coders usually have more important things to deal with (like keeping the bug count down) than code review even after opening up svn a bit since people still have to track down a module owner/core dev to look at their changes. Anyhoo, just my 2c...and apparent confession to gaming the bug tracker on occasion. Dan ___ Bf-committers mailing list Bf-committers@blender.org http://lists.blender.org/mailman/listinfo/bf-committers
Re: [Bf-committers] Compositing View Port Proposal
Part of the code cleanup could involve porting the widget drawing code to openVG http://www.khronos.org/openvg/ (accelerated 2d drawing with software fallback or an external pure opengl wrapper lib like shviavg) which the mesa gallium drivers now support in the newest release. This would also make it easier to port to other platforms like android (or my poor little netbook) which need all the help they can get with their limited gfx cards. An added benefit would be that google put a bunch of work into this for intel cards for tablets and whatnot (which my netbook just happens to contain, yay) so it would be a more 'google friendly' project than your usual gsoc by shaking out driver bugs and such. It also kind of sounds like this would give you the structure you're wanting without having to come up with a new sub-language for blender drawing code. I did some initial tests with my python wrapper around shivagv in the 3d view port and it (mostly...grr, bugs) worked as expected. Once FC17 comes out I plan on fully testing the mesa hardware implementation but have to wait since I'm way too chicken to upgrade my mesa libs to the newest on an older fedora release. Dan ___ Bf-committers mailing list Bf-committers@blender.org http://lists.blender.org/mailman/listinfo/bf-committers
Re: [Bf-committers] VSE Scene strip shortcut idea
You'd think this would be a simple three line operator but it turns out you can't set a screen from within an op (and have it updated) so all you do is end up updating the current context.window.screen.scene then when the op finishes it switches to the desired 'Default' screen (which also can't have its scene set from within an op). From rna_Window_screen_update: /* exception: can't set screens inside of area/region handers, and must use context so notifier gets to the right window */ Trying to set bpy.data.screens['Default'].scene first also does the same thing (but doesn't ever update) so you can't set this and then set the window.screen to it after. This one seems like a real bug though since rna_Screen_scene_update() is setting the scene on the context instead of the screen which is what's being asked of it. Dan ___ Bf-committers mailing list Bf-committers@blender.org http://lists.blender.org/mailman/listinfo/bf-committers
Re: [Bf-committers] VSE Scene strip shortcut idea
Ok, now it's a simple three line op... http://projects.blender.org/tracker/index.php?func=detailaid=30481group_id=9atid=498 I think tab is taken though, didn't work for me but I didn't spend any time investigation. Dan ___ Bf-committers mailing list Bf-committers@blender.org http://lists.blender.org/mailman/listinfo/bf-committers
Re: [Bf-committers] gsoc 2012
Totally off-topic (and I apologize for that) but... I looked at the files you can download from the physbam website and there indeed is a limited subset of the modules available to the public, they merely promise to open it all up sometime in the future but for now you need to jump through some serious hoops to even get cvs read-only access and I doubt the resulting (non-released) code would be free/libre. Unless you have a link because I'd love to get my hands on that facial muscle sim code? Dan ___ Bf-committers mailing list Bf-committers@blender.org http://lists.blender.org/mailman/listinfo/bf-committers
Re: [Bf-committers] 3D Cursor and Border Select
I use the 3d cursor all the time -- it's a very powerful feature once you know how to 'correctly' use it -- but rarely, if ever, use LMB to set it. Mostly I just use my pie menu script that mirrors the shift+S menu when I'm in that mode since there isn't really a good/fast way to set it and the built-in menu kind of sucks ass when doing a lot of 3d cursor work (for stuff like snapping bones' head/tails and such). My opinion is that it shouldn't be hidden for the 'benefit' of noobs because there's a lot of power that will be lost to them without having to go through all the trouble of finding some check box in some random panel to turn it on. Kind of like the 'lock view to cursor' feature, how many people can say they even know it exists since it's hidden away in the N panel View settings -- honestly I totally forgot about it myself before reading this discussion and I wrote the patch (not that I'm saying it should be more visible/accessible, just using it as an example of a 'hidden' feature)... OTOH LMB select for cursor placement should be reconsidered since using the LMB to set the cursor would suggest it is a *very* important operation (LMB being one of the main UI input methods) while it is mostly not a practical method for accurate/usable placement (2d mouse pointer + 3d view matrix + 3d cursor previous position == not usually what you want unless you specifically know that's what you want (ie, in one of the named view modes)). Setting it to double-click LMB would open up that UI input method to something which really is a *very* important operation (like regular select without all the RMB Select Dogma -- not that I care about that though). Dan ___ Bf-committers mailing list Bf-committers@blender.org http://lists.blender.org/mailman/listinfo/bf-committers
Re: [Bf-committers] 3D Cursor and Border Select
PS: How the hell do you set Double-Click for it?! kmi = km.keymap_items.new('view3d.foo', 'LEFTMOUSE', 'DOUBLE_CLICK') That *should* set something to left mouse double-click it but, you know, doesn't -- the set cursor op still traps it and sets the cursor position twice without letting the double-click through. IIRC there's always been problems with double clicking and in the (two?) places it's used needed some extra love to get working. Dan ___ Bf-committers mailing list Bf-committers@blender.org http://lists.blender.org/mailman/listinfo/bf-committers
Re: [Bf-committers] CMake packaging errors
CPack (aka 'make package') only works for rpms -- if it got that far building and linking the actual blender binary probably worked. Someone could get it to work with NSIS though... Dan ___ Bf-committers mailing list Bf-committers@blender.org http://lists.blender.org/mailman/listinfo/bf-committers
Re: [Bf-committers] Help with the source code of Blender
One of the main problems I found from using the doxygen docs (a long, long time ago so may be fixed now) is that quite often it would parse the 'bad level calls' file before the real one so you would trace a path through the code and end up at a function like 'int foo(void) {}' which is less than helpful. Much easier to use 'grep -R foo .' methinks. Plus, with all the fancy call diagrams and inline code the thing weighed in at ~300mb... Dan ___ Bf-committers mailing list Bf-committers@blender.org http://lists.blender.org/mailman/listinfo/bf-committers
Re: [Bf-committers] Developer meeting notes, sunday january 1st 2012
Hi all, Got my patch all fixed up for trunk a week or so ago but haven't heard anything back on the 'status' other than the initial mention in the meeting notes. http://projects.blender.org/tracker/?func=detailaid=22084group_id=9atid=127 Thanks, Dan ___ Bf-committers mailing list Bf-committers@blender.org http://lists.blender.org/mailman/listinfo/bf-committers
Re: [Bf-committers] [Bf-blender-cvs] SVN commit: /data/svn/bf-blender [42860] trunk/blender/source/blender/ blenkernel/BKE_customdata.h: OSX/gcc-4.6, fix for typedef uin64_t
On Sat, Dec 24, 2011 at 1:42 AM, Campbell Barton ideasma...@gmail.com wrote: I'd rather not have ifdef's like this in every header wherever int64_t/uint64_t is used. perhaps BKE_customdata.h should include BLO_sys_types.h? long term IMHO we should do away with BLO_sys_types.h MEM_sys_types.h (which are the same file), and have some generic header that can be included anywhere that deals with stdint.h types like intptr_t/ssize_t/int64_t etc. Clang building is still broken in trunk because of the redefine of u_int64_t thing. This block is the problem: /* XXX */ #ifndef uint64_t typedef uint8_t u_int8_t; typedef uint16_t u_int16_t; typedef uint32_t u_int32_t; typedef uint64_t u_int64_t; #endif I pulled it out of BLO_sys_types.h MEM_sys_types.h and clang is happy now. Not sure if this is a proper fix (I don't really understand why it's needed in the first place since u_int64_t is gone) but it lets me build up to the next error: In file included from /home/dan/blender-svn/blender/extern/libmv/third_party/glog/src/utilities.cc:334: /home/dan/blender-svn/blender/extern/libmv/third_party/glog/src/stacktrace_x86_64-inl.h:36:10: fatal error: 'unwind.h' file not found #include unwind.h // ABI defined unwinder which is progress I suppose... Dan ___ Bf-committers mailing list Bf-committers@blender.org http://lists.blender.org/mailman/listinfo/bf-committers
Re: [Bf-committers] Blender developers IRC meeting, 18 dec 2011
- Meeting briefly discussed Node plugins. Would be a good way to prevent node-creep! Lukas will look into it. I still have the libplugin (http://libplugin.sourceforge.net/) patch floating around on my hard drive which could easily be used for this with the new node registration system. After the first of the year I should have some time to kick it around a bit and get a demo up and running...and quite possibly replace the xml-based parser with a python one since the last time I brought this up there were some objections to the whole xml thing. But, as it stands, it is a fully functional plugin system base which can be extended to handle nodes and such. Dan ___ Bf-committers mailing list Bf-committers@blender.org http://lists.blender.org/mailman/listinfo/bf-committers
Re: [Bf-committers] Release targets for 2.62
I'd like to get my fcurves easing algorithm patch included in the next release: http://projects.blender.org/tracker/?func=detailaid=22084group_id=9atid=127 Have the patch updated for trunk but it still needs ~20 minutes more work to fix a few compiler warnings about unused variables and add the new file to the build system since it doesn't glob the directories anymore -- should be finished tomorrow since I have to run off to work right now. I might will also make it possible to add new curve functions via python if I'm feeling extra-saucy tomorrow...assuming this is acceptable with The Powers That Be of course. Dan ___ Bf-committers mailing list Bf-committers@blender.org http://lists.blender.org/mailman/listinfo/bf-committers
Re: [Bf-committers] Change Representation and Initialisation of Matrices to Conform with Standard Notation
On Wed, Dec 7, 2011 at 11:05 PM, Andrew Hale trumanblend...@gmail.com wrote: Hi All, I have written a patch which rectifies a cause of confusion among Blender uses when starting out with matrices and vectors. The patch ensures consistency between the matrix representation and matrix/vector products so that they behave in the way that people familiar with math expect. This is achieved by adjusting pythons access to the internal storage, while keeping the internal data the same. Note that this does not interfere with how matrices are stored internally or how matrix products, additions etc are performed internally. The patch can be found here: http://projects.blender.org/tracker/index.php?func=detailaid=29534group_id=9atid=127along with a more in depth explanation. Thanks, Andrew I thought blender does it this way is because opengl also does row (or column, dunno really) major representation? ...and this patch would break pretty much every script out there. Dan ___ Bf-committers mailing list Bf-committers@blender.org http://lists.blender.org/mailman/listinfo/bf-committers
Re: [Bf-committers] Why was BLENDER_VERSION_CHAR cleared?
How was this done in previous releases? Do kind of agree that it should be 'set(PATCH_VERSION ${BLENDER_SUBVERSION})' because BLENDER_VERSION_CHAR doesn't really give as much useful information (and caused problems with the debian packager in the past). Plus that change would make CPack just a little more bulletproof when dealing with other fedora repos... Dan ___ Bf-committers mailing list Bf-committers@blender.org http://lists.blender.org/mailman/listinfo/bf-committers
Re: [Bf-committers] noise module loaded in driver_namespace by default
On Sat, Sep 3, 2011 at 1:04 AM, Campbell Barton ideasma...@gmail.com wrote: On Sat, Sep 3, 2011 at 5:27 PM, Daniel Salazar - 3Developer.com zan...@gmail.com wrote: I want to add the fantastic noise module to driver_namespace by default. After all, what good is an expression without noise! This module adds the next awesome functions: cell cell_vector distance_metrics fractal hetero_terrain hybrid_multi_fractal multi_fractal noise random random_unit_vector ridged_multi_fractal seed_set turbulence turbulence_vector types vector vl_vector voronoi this + the math module + frame (both already there) = expressions useful for everyday tasks! Daniel Salazar 3Developer.com I'm reluctant to import these directly because some of the names don't make a lot of sense on their own. - random: a python module - types: also a python module - seed_set: doesn't make sense to import this IMHO - vector: similar to mathutils.Vector - cell: not very meaningful on its own. Suggest we import noise but not its members, this way you have to type noise.random() rather than just random() but at least its clear whats being accessed. ___ Bf-committers mailing list Bf-committers@blender.org http://lists.blender.org/mailman/listinfo/bf-committers IIRC my original patch imported noise like is suggested, the code is in there (from 2.4x) but commented out. Dan ___ Bf-committers mailing list Bf-committers@blender.org http://lists.blender.org/mailman/listinfo/bf-committers
Re: [Bf-committers] Compositor nodes-rerouting
On Wed, Aug 31, 2011 at 5:51 PM, Daniel Salazar - 3Developer.com zan...@gmail.com wrote: I'm moving more towards segmentation to manage complexity in nodetrees, islands of nodes separated in space and connected by one or two lines, how ever navigating this widespread node systems is hard. a kind of outliner view for frames would not only clearly separate this island/context but also provide quick navigation. Also frames could be represented in a thumbnail navigator, click on a frame in the navigator and the view takes you there Daniel Salazar 3Developer.com Isn't that basically how group nodes work? Dan ___ Bf-committers mailing list Bf-committers@blender.org http://lists.blender.org/mailman/listinfo/bf-committers
Re: [Bf-committers] Blender incorpore encryption???
Blender uses the industry standard ROT26 encryption scheme IIRC. On Sun, Aug 21, 2011 at 2:27 AM, Tom Edwards cont...@steamreview.org wrote: He might mean does Blender incorporate file encryption. On 21/08/2011 10:14, Damir Prebeg wrote: ??? On 21 August 2011 11:10, Fabio Gonzalezfabiojo...@gmail.com wrote: Blender incorpore encryption??? ___ Bf-committers mailing list Bf-committers@blender.org http://lists.blender.org/mailman/listinfo/bf-committers ___ Bf-committers mailing list Bf-committers@blender.org http://lists.blender.org/mailman/listinfo/bf-committers ___ Bf-committers mailing list Bf-committers@blender.org http://lists.blender.org/mailman/listinfo/bf-committers ___ Bf-committers mailing list Bf-committers@blender.org http://lists.blender.org/mailman/listinfo/bf-committers
Re: [Bf-committers] Libraries missing target_link_libraries(...
Are you talking about the cmake make package rpm builder or making a .spec file from scratch? The cmake one should work in cycles though may need an entry for libcycles.so so as to not complain about an unlisted install target or some such thing. Dan On Sat, Aug 6, 2011 at 5:21 AM, Richard Shaw hobbes1...@gmail.com wrote: Not having gotten any replies to my last few questions I'm starting to think I'm becoming a bother, but I'll try again... I've been working on a blender-cycles build for Fedora and I thought I was getting pretty close until I ran rpmlint on the installed blender package. I noticed I had seemingly 1000+ errors of: undefined-non-weak-symbols which basically means the libraries built by blender are not being linked with all the libraries they should be. I've started adding target_link_libraries(... after blender_add_lib to fix this, for example: From my installed blender instance: $ ldd -d /usr/lib64/libbf_avi.so linux-vdso.so.1 = (0x709d1000) libgomp.so.1 = /usr/lib64/libgomp.so.1 (0x7fc621e67000) libpthread.so.0 = /lib64/libpthread.so.0 (0x7fc621c4c000) libc.so.6 = /lib64/libc.so.6 (0x7fc6218b2000) librt.so.1 = /lib64/librt.so.1 (0x7fc6216aa000) libgcc_s.so.1 = /lib64/libgcc_s.so.1 (0x7fc621495000) /lib64/ld-linux-x86-64.so.2 (0x7fc6222aa000) undefined symbol: jpeg_resync_to_restart (/usr/lib64/libbf_avi.so) All I did was add: target_link_librarires(bf_avi ${JPEG_LIBRARIES}) after the appropriate blender_add_lib(bf_avi...) And now I get this from my local build lib directory: $ ldd -d libbf_avi.so linux-vdso.so.1 = (0x7fff48fff000) libjpeg.so.62 = /usr/lib64/libjpeg.so.62 (0x7f57dc894000) libgomp.so.1 = /usr/lib64/libgomp.so.1 (0x7f57dc687000) libpthread.so.0 = /lib64/libpthread.so.0 (0x7f57dc46b000) libc.so.6 = /lib64/libc.so.6 (0x7f57dc0d2000) librt.so.1 = /lib64/librt.so.1 (0x7f57dbeca000) libgcc_s.so.1 = /lib64/libgcc_s.so.1 (0x7f57dbcb4000) /lib64/ld-linux-x86-64.so.2 (0x7f57dcd16000) So I have two questions: 1. Is this the right way to fix this? (I would assume yes from a cmake perspective, but wasn't sure if there was a more blender way to do it) 2. Am I wasting my time fixing this? In other words. If I can supply a diff when I'm done, will you use it? Respectfully, Richard ___ Bf-committers mailing list Bf-committers@blender.org http://lists.blender.org/mailman/listinfo/bf-committers ___ Bf-committers mailing list Bf-committers@blender.org http://lists.blender.org/mailman/listinfo/bf-committers
Re: [Bf-committers] Cycles and Boost 1.44 (Fedora 14 and above)
I was building cycles on fedora 14 for a while, OIIO needed a reverse patch (on boost maybe?) to build IIRC. Still haven't gotten my f15 install set up with the deps for cycles yet so can't say what's involved with that...been busy with work lately and just haven't had the time to get it working. Dan ___ Bf-committers mailing list Bf-committers@blender.org http://lists.blender.org/mailman/listinfo/bf-committers
Re: [Bf-committers] Some ideas to improve the Add-Ons section
On Mon, May 23, 2011 at 1:58 AM, Jass gaia.cl...@machinimatrix.org wrote: --snip-- - add an API to query (a third party website) for new versions of an addon module, and possibly add a button for retrieving the addon directly from a (third party) download server instead of first downloading from the web, then updating. This would fit nicely into the list of buttons: [Wiki], [Bug Report], [Check for Update] I think I (and others) would be kind of weary of blender being able to download and install software from random websites. For customizable persistent settings, addon authors could use _bpy.user_resource('CONFIG', 'plugin_name') to get a (hopefully) writeable folder to store settings in. As a bit of an aside, I think _bpy.user_resource() should throw an exception instead of returning an invalid path when the folder doesn't exist so one wouldn't have to test for a valid path after calling the function. Dan ___ Bf-committers mailing list Bf-committers@blender.org http://lists.blender.org/mailman/listinfo/bf-committers
Re: [Bf-committers] [Bf-blender-cvs] SVN commit: /data/svn/bf-blender [36779] trunk/blender/source/blender/ blenkernel/intern/fcurve.c: modify fcurve evaluation for bool/enum/ int values.
On Thu, May 19, 2011 at 5:34 PM, Matt Ebb m...@mke3.net wrote: On Fri, May 20, 2011 at 10:17 AM, Campbell Barton ideasma...@gmail.comwrote: Aligorith and I discussed this before committing, and are aware of the implications. To me if you are editing floats, but have them converted to ints later it makes most sense (on a user level), to round them. If the GUI makes it unintuitive then it should be made intuitive (grid lines when zoomed in for example), if you want I can add this. Rounding them to whole numbers, sure, I totally agree, and if floor() works better then I'm all for it. But when animating bool values it makes perfect sense that as soon as your curve crosses from 0 to 1 (not 0 to 0.5), the value is switched on. Same goes for integer values, animating my curve from 4.0 to 4.51 should not switch the value to 5. cheers Matt ___ Bf-committers mailing list Bf-committers@blender.org http://lists.blender.org/mailman/listinfo/bf-committers I totally agree, 1 == false and = 1 == true is how I always thought it worked. Dan ___ Bf-committers mailing list Bf-committers@blender.org http://lists.blender.org/mailman/listinfo/bf-committers
Re: [Bf-committers] CPack rpm patch
Small tweak to the previous patch to 'fix' the revision number -- http://www.pasteall.org/21511/diff Don't really know how to check if CPack is running to warn if CMAKE_INSTALL_PREFIX or WITH_INSTALL_PORTABLE is set to a bad value without messing up a regular cmake build. Dan ___ Bf-committers mailing list Bf-committers@blender.org http://lists.blender.org/mailman/listinfo/bf-committers
Re: [Bf-committers] translate to 2.5
2011/5/9 Yuniel Castro González ycast...@estudiantes.uci.cu: 1- nuobs = Create(10) 2- scn = Scene.getCurrent() 3- num = Blender.Noise.random() 4- randobjects = Blender.Object.GetSelected() -- Yuniel ___ Bf-committers mailing list Bf-committers@blender.org http://lists.blender.org/mailman/listinfo/bf-committers http://www.blender.org/documentation/blender_python_api_2_57_1/ ___ Bf-committers mailing list Bf-committers@blender.org http://lists.blender.org/mailman/listinfo/bf-committers
Re: [Bf-committers] Camera Guides
You can currently load up an image into the 3d viewport in either the foreground or background using the py-api. I wrote an operator a while back to test image.gl_load (or whatever it ended up being called) that did just that and it was pretty easy -- aside from all the opengl voodoo. You can even display it in 3d space as I (accidentally) found out while playing with artoolkit. Dan ___ Bf-committers mailing list Bf-committers@blender.org http://lists.blender.org/mailman/listinfo/bf-committers
[Bf-committers] CPack rpm patch
Got the rpm package builder working again in CMake. http://www.pasteall.org/21439/diff Dan ___ Bf-committers mailing list Bf-committers@blender.org http://lists.blender.org/mailman/listinfo/bf-committers
Re: [Bf-committers] Camera Guides
With the py-rna attributes view_camera_zoom and view_camera_offset from the cycles branch one could easily implement this as a modal operator without having to patch the source. Dan ___ Bf-committers mailing list Bf-committers@blender.org http://lists.blender.org/mailman/listinfo/bf-committers
[Bf-committers] Cycles rna c++ render extension code patch
Yanked out the C++ API and RenderEngine code from the cycles branch if anyone wants to apply it to trunk: http://www.pasteall.org/21286/diff Needs some scons love still. Dan ___ Bf-committers mailing list Bf-committers@blender.org http://lists.blender.org/mailman/listinfo/bf-committers
Re: [Bf-committers] cycle todo list
Small problem with external_libs.cmake, FIND_PATH(BLENDER_INCLUDE_DIRS RNA_blender.h PATHS ${CMAKE_BINARY_DIR}/include) can't be set because makesrna generates RNA_blender.h after CMake is run. Dan ___ Bf-committers mailing list Bf-committers@blender.org http://lists.blender.org/mailman/listinfo/bf-committers
Re: [Bf-committers] cycle todo list
Also... This boost-1.44 patch breaks the compile: https://svn.boost.org/trac/boost/changeset/62245/trunk/boost/detail/sp_typeinfo.hpp ___ Bf-committers mailing list Bf-committers@blender.org http://lists.blender.org/mailman/listinfo/bf-committers
Re: [Bf-committers] Should CMAKE_INSTALL_PREFIX work?
As a side note, this also breaks the CPack RPM builder. Haven't quite gotten around to working on a fix for it yet though -- just recently moved to an rpm based distro so... Dan On Mon, Apr 18, 2011 at 3:19 AM, Paul Melis paul.me...@sara.nl wrote: Hi, I've been using an SVN checkout of the 2.5 branch and corresponding build directory for some time now, happily doing SVN update and make, make install to keep track of Blender progress. However, I just noticed make install no longer installs files in the correct location anymore. Installation used to go to CMAKE_INSTALL_PREFIX, but now seems to end up in build-dir/bin/2.5x. Has the installation procedure changed at some point? Reading http://wiki.blender.org/index.php/Dev:2.5/Doc/Building_Blender/Linux/Ubuntu/CMake leads me to believe that filling in CMAKE_INSTALL_PREFIX should work, together with make install. This is with cmake 2.8.1, Ubuntu 10.04. The last time installation worked correctly seems to be around February, 28th Thanks, Paul ___ Bf-committers mailing list Bf-committers@blender.org http://lists.blender.org/mailman/listinfo/bf-committers ___ Bf-committers mailing list Bf-committers@blender.org http://lists.blender.org/mailman/listinfo/bf-committers
Re: [Bf-committers] Experimental pydna
Plus it doesn't work on windows which would frustrate people developing on linux and trying to release their script to the public. On Sun, Apr 3, 2011 at 6:32 AM, Ton Roosendaal t...@blender.org wrote: Hi all, Last september, Revision 31766, Campbell added this: --- ./intern/tools/pydna.py Experimental module (for developers only), exposes DNA data via python uses no blender/python modules, pure python + autogenerated ctypes api. --- Since it allows full Blender data access, scripters are using it to cover up for missing RNA parts. A 2nd Life developer in IRC asked if we could enable it on the 2.57 release, by ensuring it works for Windows as well (needs a small #ifdef DNA hack). My suggestion would be to really not do this, it doesn't belong in releases. By fixing this backdoor to work in all released binaries we only will regret it later on. We don't have our RNA project for a good reason... Because 2nd Life really needs image access, I advised them to provide temporarily a special build for their users to survive the period until we have our own decent working RNA level image access. Please advise if that's acceptable? -Ton- Ton Roosendaal Blender Foundation t...@blender.org www.blender.org Blender Institute Entrepotdok 57A 1018AD Amsterdam The Netherlands ___ Bf-committers mailing list Bf-committers@blender.org http://lists.blender.org/mailman/listinfo/bf-committers ___ Bf-committers mailing list Bf-committers@blender.org http://lists.blender.org/mailman/listinfo/bf-committers
Re: [Bf-committers] Experimental pydna
On Sun, Apr 3, 2011 at 8:33 AM, Erwin Coumans erwin.coum...@gmail.com wrote: Dan, Ton mentioned the fix that makes it work for Windows. Doh, replied before the coffee fully kicked in... What is the problem of adding such feature in 2.57, if it is marked clearly as experimental/temporary fix? When is a permanent fix ready? I still don't understand why having no solution is better than a 'too slow' solution? Dan ___ Bf-committers mailing list Bf-committers@blender.org http://lists.blender.org/mailman/listinfo/bf-committers
Re: [Bf-committers] pixel get image processing
On Tue, Mar 15, 2011 at 2:47 AM, Domino Marama dom...@dominodesigns.info wrote: --snip-- Has anyone successfully tried this on Windows or know a way to fix it? Cheers, Domino ___ Bf-committers mailing list Bf-committers@blender.org http://lists.blender.org/mailman/listinfo/bf-committers The ctypes.CDLL() part doesn't work on windows due to the way msvc exports symbols (or more specifically doesn't unless told to). Dan ___ Bf-committers mailing list Bf-committers@blender.org http://lists.blender.org/mailman/listinfo/bf-committers
Re: [Bf-committers] Paging Dave Plater
On Thu, Mar 3, 2011 at 9:47 AM, Richard Shaw hobbes1...@gmail.com wrote: I'm working on getting a build of 2.56 beta or newer built for Fedora and I was told that I should seek out Dave Plater as he's the Suse maintainer which also uses RPMs. I've downloaded the Suse src RPM and I have modified the spec file but I'm running into issues that are beyond my ability to fix and could use some help. Thanks, Richard ___ Bf-committers mailing list Bf-committers@blender.org http://lists.blender.org/mailman/listinfo/bf-committers Don't know if it helps any but with cmake you can do a 'make package' and it spits out an rpm -- which also probably needs someone familiar with rpm based distros to look over/fix. Also needs to be tested on a red hat based distro since (I think) it's only been tried on suse. Dan ___ Bf-committers mailing list Bf-committers@blender.org http://lists.blender.org/mailman/listinfo/bf-committers
Re: [Bf-committers] Paging Dave Plater
On Thu, Mar 3, 2011 at 12:11 PM, Richard Shaw hobbes1...@gmail.com wrote: I tried 'make package' but got the following: $ make package make: *** No rule to make target `package'. Stop. I'm using the source from the Suse src package that appears to be SVN 34784. Any ideas? Richard You have to set up everything like a regular cmake build first then do the 'make package' thing. Also requires a fairly recent svn checkout, not too sure when it was committed. Dan ___ Bf-committers mailing list Bf-committers@blender.org http://lists.blender.org/mailman/listinfo/bf-committers
Re: [Bf-committers] Paging Dave Plater
There's also an osx package building ability which is completely untested if some mac person wants to play too Dan ___ Bf-committers mailing list Bf-committers@blender.org http://lists.blender.org/mailman/listinfo/bf-committers
Re: [Bf-committers] blender-snapshot
Here's a patch (http://www.pasteall.org/19364/diff) that does the rename plus a couple other little things to make a valid and signed package from a 'make package_debian' -- assuming one has their gpg keys set up properly. I've been using it successfully with those changes and it should also work if someone wanted to upload to a launchpad ppa using 'dput ppa:foo/ppa blender_2.56+svn35006-bf_i386.changes'. Dan On Mon, Feb 21, 2011 at 12:24 AM, Campbell Barton ideasma...@gmail.com wrote: Dan, since you added the debian package spec best you to make decisions on this. If its a simple rename I can commit otherwise send in a patch. On a related topic, just moved release/VERSION vars into BKE_blender.h This way we can construct version strings from scanning the header in either format... 2.56a-beta - external release name we use on blender.org. 2.56.2 - internal version string using subversion, internal blender use. the string thats currently 'beta' could be used to detect if blender is a release or not for generating packages. the value could be set to 'alpha/beta/rc/release', and check for 'release', to made a package without the svn revision included. I'd rather not become too involved in packaging discussions, devs who maintain package spec's can choose which name, version string to use which fits with their own distro/package guidelines. - Campbell On Sun, Feb 20, 2011 at 9:21 PM, Dan Eicher d...@trollwerks.org wrote: I'm pretty sure simply changing the name of the .deb package to blender-snapshot doesn't actually fix the 'bug' that was reported. From what I understand the 'a' -- i.e. 2.49a -- was the issue and not the package name which the debian folks solved by converting the subversions to numbers, 2.49a = 2.49.1, 2.49b = 2.49.2, etc... And, to top it all off, the blender-snapshot package conflicts with the regular (stable) blender package so you can't even have two different versions installed which is the only valid reason I can come up with to have a different name. To summarize, the issue is the way *blender* names releases and has nothing to do with the debian package (which is what I said multiple times in the bug report). So, not a bug and as author of the original patch I'd appreciate if the name were changed back. Dan ___ Bf-committers mailing list Bf-committers@blender.org http://lists.blender.org/mailman/listinfo/bf-committers -- - Campbell ___ Bf-committers mailing list Bf-committers@blender.org http://lists.blender.org/mailman/listinfo/bf-committers ___ Bf-committers mailing list Bf-committers@blender.org http://lists.blender.org/mailman/listinfo/bf-committers