I’m not sure if I will be able to put the thoughts floating around in my head down into words in a concise manner or not, but here goes:
After having put forth the effort to start becoming familiar with Blender’s interface & capabilities I feel an odd loyalty to it, and a desire to see it succeed.
From recent posts it seems that some forum members are interested in seeing a period of growth - not only for the software, but of the user base and level of expertise as well.
I don’t think anyone on these boards would be upset to see Blender do to the 3D world what Linux has done to the computing world.
So my question is how can we all work together to help achieve this?
The current 3D playing field has some pretty stiff competition, with well established players. But on the other hand, interviews I have read with 3D studios say stuff like “<insert 3D app name> is crap, but it is the best crap right now.”
So on a software level I wonder what it would take to turn some professionals heads? Linux has not completely displaced Windows, but it has made a big mark in the server market by being really good at that. I wonder likewise if Blender development especially focused on making one aspect of the software exceptional - perhaps by enlisting the input from a studio - if it could not accomplish the same thing as Linux. Professionals make their living with their software. To get professionals using Blender it has to be more than free, it has to give them an edge over the competition.
On a user level, I think some of the recent suggestions, such as adding a critique forum & modeling forum, might be useful. There is also ongoing work on providing better access to available tutorials. IMO, solid documentation and official tutorials (that are updated as necessary with software changes) are going to be important for Blender’s future.
Oh well. Not sure if this post really has a point or not. Guess I’ll just see what happens with it.
So, what would it take to make Blender a real name in the pro 3D field?
It’s been close to three months since I used Blender the last time. And I was thinking - what kind of changes would need to be implemented in order to make me switch back from C4D and start Blending again? That might be a starting point… so, here’s what I’ve come up with:
Real workmodes for faces, edges and vertexes with good visual feedback (think Wings here)
Selection edge loop/ring
Hierarchical grouping functions
Real polytools like bevel, knife and bridge etcetera
Extrude/move/rotate/scale along face normals
Undo/redo (yes… =) )
A standard transform gizmo
Simplified plugin handling. Pythonstuff should simply go in a special folder, automatically identified and placed in a menu when Blender starts
Reliable export/import for OBJ, 3DS, XSI etcetera. The well-known 3D formats easily accessible from the filemenu, no fuzz…
Realtime texture manipulation in viewport
Context sensitive menus for everything
A new core render engine (scanline/raytrace hybrid), not as an external thing, but as the default no-messing-around-renderer.
N-gons
User definable hotkeys and button palettes
User definable colors for all GUI-elements
There’s a lot of things to add here, but it’s a start.
I think this one could be solved by sorting out the OO representation window, and providing a “select all children” or such like. I find proper use of empties and layers actually means I don’t miss this at all, it’s just a different approach.
There’s some interesting discussion over on blender.org about stuff involving a transform gizmo etc., but the key thing is how to unify that with Blender’s general approach. The reason I say that is that I have a choice between using Blender and C4D (thanks to the mag it came on!), and I actually choose to use Blender (and Wings in combo) as I prefer it, and don’t want to see that essence diluted.
It’s certainly good to look at other programs, and things they might do better, but ripping features out wholesale and Blender will never be anything more than an imitation.
Pure imitation isn’t going to get Blender anywhere. Imitation always put you at least one step behind.
On the other hand, in a slightly related way, I have heard that the way to be successful in business is to look at what everybody else is doing - and then come up with a unique slant, a way to do it better, faster, cheaper, something.
Blender models, rigs, textures, lights, renders, edits, composites?, animates, and has a game engine. That is a pretty wide spread. What is the chance of putting special emphasis on excelling at one of those, without letting the others slip in quality?
Is the NLA pretty standard among other 3D apps, or is that somewhat special?
from a non-programming point of view:
artists should get everything out of blender to make this great piece of software more well-known… break your own bounds and those of blender.
the quality of art currently produced is amazing, but can’t we just go further?
Good point. Most of us aren’t programmers, and it isn’t realistic to lay all the work in the laps of the coders.
This is an open source project. A community effort. So, really, whether or not we are programmers or not - there is plenty of work to be done to help promote / improve blender.
Any suggestions on how us non-programmers can coordinate / organize our efforts to accomplish something larger than our individual abilities?
I’m not a programmer… more of a hacker (I’m the guy editing the tooltips. %| ), but I can tell you this:
A clearly defined statement of what tool or function you want, where it should go, and what exactly it should do is GOLD to a programmer. If you’re not sure where to post it, try this mailing list:
If you see a new tool introduced, test the heck out of it!! Bang on it with a hammer. Test it in every possible way! Then, if you find something that “broke” (a crash, or more subtly an incorrect result) it can be reported & fixed. If everything worked great… tell someone! Post a pic showing what you achieved with New Function X! Write a tutorial on how you did it!
Just two things off the top of my head. Other ideas welcome!!
I don’t have a whole lot of experience with other tools and such, but one thhing that I could really use in Blender would be a set of sort of ‘mesh sculpting’ tools… that act like clay tools almost. You have your mesh an you take a brush to it in edit mode and deform it in a smooth, elastic way. Like one of the tools could be a thing that has a brush like gimp’s with a diameter and a hardness. THe amount it deforms the mesh as you draw it across the mesh depends on how ‘opaque’ the brush is as it hits the vertexes. It could lead to some very intuitive mesh editing.
and the studios that don’t say that develop their own software on the best to develop on. (as in for example ILM has their own “distribution” of maya)
in that case blender is not the best. It doesn’t have many features and is not documented. (and doesn’t have a paid team of programmers that support the studios like maya does)
also, blender is missing a lot of features (other posts mention many, my point has been proven) and technoligies found in other programs. Not the least of which is cloth simulation.
Agreed. I guess my ultimate question is this: What do the developers, the guys like ton, who are making the decisions want the future of Blender to be? That’s what I want to know.
If the goal is to be a good OSS 3D package, great - because I think they have already achieved their goal. In which case I will use Blender and learn as much as I can, but as others have stated, if I get SERIOUS, I may need to move onto some other package eventually.
If the goal is to make a OSS 3D package that can compete with other apps in a commercial production environment, that is great too. And I would want to know how I could help.
Your post seemed to indicate that you don’t think the second option is possible. What about CinePaint, formerly Film Gimp? It is OSS and IS currently being used in commercial major film productions despite the issues that you raised.
Personally speaking, the feature I would dearly love is real UV unwrapping. I have this bizarre dream that somehow it’s possible to totally automate UV unwrapping, and that this is how every non-Blender person does it (I’m dreaming remember) and I want that magic “do my UV for me” button.
The Wings AutoUV is very close. Blender would need a way of flagging edges as cuts. Apparently the new Houdini has absolutely staggering UV tools . . .
As regards the cloth, there used to be a Python binding to some really quite good cloth sim, the downside being you couldn’t save the settings. Of course, now Blender’s open source maybe that could be melded into the main package? A full dynamics side would be awesome, and honestly I can see it coming as a nice side effect from the work on the game engine.
Other things
more flexible render pipeline (I’ve ranted about this before) where you can mix sequence plugins/external renderers etc. all in one. This would involve an all new sequence editor.
OO window where you can do more than select and move things (“group under empty” would be handy, for example, as would the possibility of a ‘tree view’)
flexible procedural shaders (I can see this evolving out of the current setup which is a fantastic step forward! so thanks cessen and ton!)
separate render passes (this would be solved by my above pipeline idea, but it would be neat if you could get passes of diffuse and specular reflections separately for compositing, and maybe even an automatically generated ambient occlusion pass which uses the ‘amb’ as a ‘ref’ value and automatically creates the dome etc . . .
opening up a pre-render mesh operation sequence, with things like cc-subdivision, booleans, old smesh etc. all chucked in, could even be used to make the procedural shells for fur if done right etc.
python access for practically all existing mesh operations so you can perform them procedurally to whole objects to create new meshes, with booleans, subdivision, smooth, etc. (do that and I would be a very very happy man)
When I think about it there’s loads I’d like, most to do with procedural things (mmm, curious). As you’ve said I think Blender is already very good indeed, and to be honest, for a lot of 3D work, ideal. I’m probably one of the strongest advocates in these parts for Blender as it stands now (not totally keen on these new GUIs etc.), but I’m willing to be convinced otherwise.
If I get the chance I will look into doing that render pipeline stuff myself if necessary. It’s one severe overhaul of that whole system that I’m thinking about though, and it really needs some consideration and planning.
Just reading James’s post,…James, have you bothered to read the release notes for 2.28a at all? Not to mention, there is stuff on that list that is already in Tuhopuu, and will likely be featured in 2.29. Seen the new metashapes yet? Quit y’r b*tchin. Blender rules.
I think they’ve got some valid points. People are constantly saying, “Wouldn’t it be great if blender…” and it makes perfect sense that there should be an organized effort in the community to think up the features, work out how they’d function, perhaps even do some of the math for the coders, and then submit an organized and well documented PROPOSAL to the development teams. That way a bunch of the work is done for them already. Thats not “b*tching” thats productive.
Well the funboard was already mentioned. The documentation team led by our hero S68 is still very much in progress and I’m sure would benefit greatly from more writers/native english speaking editors/etc. Especially when the 2.3 UI changes happen, they’ll need more help taking new screenshots, updating references etc. It’s also disappointing to see things like this get overlooked and forgotten when it’s something perfect for Blender artists to help out with. It seems it’s much easier for people to write enthusiastic words about helping Blender progress than actually doing something about it.
Last month I only read the Blender General forum as I was strictly trying to figure out how to work the basics of this program, so I didn’t see the referenced post in News & Chat.
I would be happy to work on a demo file. It just has to be simple as I barely know squat about Blender. I could probably handle the Parenting one - maybe.
Are the old versions available so I can get a better idea of what the finished product is supposed to look like? Thanks!
Also, I would be happy to help with the documentation. English is my first language, and while my grammar is far from perfect I can do O.K. - I’ll post a note to the documentation mailing list.
?
