I’m a mechanical engineer myself though I am more orientated towards the construction side of the industry and work almost exclusively 2D and I haven’t really used a solid modeler since my university days.
I have always been a bit on the skeptical side of using Blender as a CAD tool but I have been wanting to get back into 3D CAD and learning 3D printing so not having to learn another program would be a big plus for me.
Good luck guys I will definitely be keeping an eye on this
piotr its hard to get stuff into trunk, this way they can work at their own pace, not be told to wait because some dev is working on something else and doesn’t want interference, etc…the guy that was working on blendercad even stuck with 2.49. if the bf finds it usefull even as a fork it will be gpl so they can use what ever they want, they wont be deprived of anything. anything that would have made it into trunk probably still will. but this way they aren’t taking away from the bf releases which the devs already have their hands full without the additional reviews and the new crew can do their thing, neither side has to sit and wait on the other. it has the chance to be more productive for both than branching. and dont forget the bf has to babysit the google summer of stuff too. the bf can wait till they have the time to check it out.
The main reason for developing in a fork, is that is the best way to show up what can be done for technical (not artists) users, on a development that affects lots of areas.
Most of the development could be interesting for all users, but some maybe will be more specific, and this may make harder to be included.
Woudldnt it be possible to these kind of things using python addon’s ?.
Cause then you wouldnt drift away of from main branche, and you more likely will get more people around it.
As i remind in the past people made measurement stuff using python, and snapping stuff.
Maybe some of features could, others not in the way that exists. The project aims to get productive tools built in. Eg, new improvements allow to modify a dimension value if selected using only the numpad.
Also allowing the user to move around the scene during transform will not be able to be performed with addons.
Imho addons are great because it allows to expand the aplication using a high level api. So you are limited to it, and also it does not guarantee it’s compatibilty due to api and python changes.
While i have nothing against projects like this (can be interesting), i wonder why not helping more mature projects like Freecad (which as o version 0.15 have far more advanced features and are more ready to use for mechanical work than this) to be complete, instead of trying to convert a mesh modeler with no intelligence at all (e.g. Parametric) into a CAD tool.
I still believe that one should use the right tool for the right use. Being a Industrial Designer myself (and having teaching Blender some time ago, and still using and teaching CAD software) i’m somewhat skeptical about this. After all, i tend to use Blender more in the final stages of a project more like a visualization tool (CAD -> Blender -> 3D rendered visualization).
But maybe i’m just getting old. Nowadays i’ve seen guys using Excel to write documents, insted of solving formulas, so probably is just me keeping out of touch with the times…
This is an awesome development, I need this build now!
The best part is the dimension which influences, on the mesh, awesome, I see already fully integrated solutions obtained with Blender, all very nice. wait anxious.
We do not want to convert blender in the way it works, so all added features will work over mesh, we will not make a different application. The whole idea of the project is have a powerful tool, that can be used from begging to end. Of course you can develop with FreeCad and then export your data to Blender (for eg) but the fact of exporting and adding things to the mesh on blender, is not a great thing, if the base model can be changed (maybe in dimension) and reused.
We do it in Blender (we like it, of course!) because working on the mesh give you a lot of freedom that you will never get with applications that build an object based on an stack of operations.
i think from what i’ve seen that 2.8 will have something like that too (some new wiggely widgets).
Maybe create a build to show of some of the things you can code, then work with the devs to extend it to the main branch ?.
I mean there is now a blender build specifically for game assets (all python based).
There is a version for doing collisions with smoke.
Then there is another branch Bofartis or so.
A branch with object nodes
A gooseberry branch (with some extra hair tricks).
Then there are the experimental branches (cycles), from who i got a few versions.
It becomes a bit fragmented, it is ok to fragment, but i think combining would make blender stronger.
What i’ve seen is that the widgets allows to modify mesh, with a visual helper. The main difference here is that dimensions are an instance. At the moment they are all related to vertex, but in a near future they will can be related to geometry. and this is the start point to get constraints about them and also try to implement a constraint a solver.
Some people and us, they start developing on a fork of BF because it’s hard to get some improvement on master, especially if affects lots of areas. So having a fork, is the way to show up your work.
Also git allow to working with forks, in a similar way that branching does.
