hey guys, I’m a hobbyist 3D artist, and I know a thing or two on how to work with blender and some other open source projects that interact with it.
now, I’ve been seeing this software a lot in the open source community named FreeCAD and never got to dip my feet in and explore it, as I’m afraid it’s either a waste of time or too complicated to be a part of hobbyist arsenal. But I’m still attracted to it because of things I’ve seen people make with it -mainly archviz and part-design and some engineering stuff which I think is super cool-.
so here I am asking for your thoughts and opinions about it, thanks in advance.
I personally avoid it as much as possible. It is cool that there finally is a parametric cad system with code available, that runs on linux, but it ends there. I really don’t like the interface, and loathe (When I last tried to use it, about a year ago, if this has changed for the better, and I really hope it has, please let me know, it would make it viable for me to use) the 2d sketch editor.
In terms of complexity, the software is manageable. I have been around cad systems for a long time, despite only working properly with them once across a year, but it is quite easy to grasp the basics of how to create and work on parts in parametric. Where the learning curve comes in is knowing what features to look for, how to find and apply them in the interface of the cad system. For me, this is where freecad fell a long way short, I just don’t get it, it feels like I have to play a game of chess against it every time I try and work out how to use a feature I know I want. Other [parametric] cad systems are much more intuitive and similar about the workflow.
Depending if you are willing to learn the interface patterns of the software (unless changed, not very transferable), the parametric system thought process makes modelling parts for things like a 3d printer easy and very much in control of what your part is. The parametric process won’t change much in other cad systems that use it to my knowledge.
If you want another option to try if you are more interested in individual parts than drawings/assemblies, I loved the feel of SolveSpace when I last used it. You may want to try that too, since it can export models to step (3d mesh, not parametric data), which you can view in other systems to use later.
Please note that I wrote this from the context of using Alibre (Previously Called Geomagic) and Autodesk Inventor (2014 & 2016).
yeah it’s not really viable for real projects. There are commercial CAD programs that run on linux (or in the browser) though if you don’t want to dual boot or run a VM
- Houdini
- Autodesk inventor
- Tinkercad
- ARES Commander
- DrafSight
- QCAD
- CYCAS
most people just use autocad in wine though
i couldn’t agree more on this one, as freecad uses the concept of workbenches as it doesn’t want to be limited to only one workflow but that make it hard for beginners to locate certain tools that complement each other but are put in different workbenches. and the UI is ungodly clunky and and very distracting too (with those icon colors)
as i said, this is just a hobby so i’m not very concerned about special UI quirks just don’t want to waste time on garbage software (as to be honest you some of the works people do in FOSS that looks amazing then you find out that the software isn’t capable and the people who made those just have some serious dedication it’s like they make the monalisa in old microsoft paint)
well to be honest the interface doesn’t look apetising but i’ll keep it in mind, also FreeCAD AFAIK can export 3d meshes in obj format too
well, i’m just hobbyist so as long as it can do the tasks i want in a reasonable time-frame and the outcome is good i’m fine with it (time != money to me)
as for the software you mentioned i love the procedural workflow of houdini but i probably will never master it as i only work a couple of hours here and there on this stuff, and i recognize QCAD and draftsight but i’m not familiar with 2D CAD work
yeah, that’s fair.
I’d say a lot of what works for you in linux depends on time preference. If you don’t have deadlines, for example, Kdenlive or Shotcut might be fine for you, but you’re gonna save a lot of time using Lightworks or Davinci Resolve, especially at 1080p+
same with Photoshop Vs. gimp for cmyk workflows and color management, Blender vs. houdini/maya for pipeline/OFX support, natron vs. Nuke and Fusion, etc. etc.
well it’s not just time as FOSS usually lags behind in features and are not as polished as close-source commercial ones, some times its fine -like blender which is a top grade modeler and the only thing that come close to it in this sense is modo-.
But then you have kdenlive/shotcut which are a horrible experience to me in the past as they keep crashing, and sometimes won’t even install due to package conflicts on top of their redundant workflow or inkscape that needs hours of calibration to work with touch screen.
the graphics apps in linux are getting better and currently a mix of FOSS and proprietary software is the best you can do to have acceptable speed in projects.
from here on the FOSS projects i keep eye on are blender, krita as they seem to be the most actively developed and they keep gaining traction.
and cheers mate for giving your opinion!!
I didn’t mean to imply that all open alternatives are at feature parity by any means, just a simple way to look at it.
kdenlive works great on distributions that respect upstream defaults in my experience, but the problem is with it’s completely single threaded and cpu bound handling of media processing. Totally chokes on 1080P lossless or raw footage on upward. Without a total rework of MLT that’s not going to change, and without OFX support or decent color management you may as well write it off from the get-go.
the fun thing is, blender VSE, which is an abandoned wasteland from the developers point of view, is according to many the best video editor in the free FOSS arena (i remembered that lightworks is open-source)… Which says a lot.
most of these projects could’ve been great but the unfocused management of development and resources and the lack of resolve to make daring changes -mainly to the UI, break backward compatibility and supported platforms/hardware- just because some 10 neckbeards in some basement said it will mess with there workflow are huge hurdles for these projects.
blender really isn’t suitable as an NLE from an interface and UX perspective, but it definitely handles all the realtime stuff much better than MLT. All of its timeline functionality is built for compositing and animation (ala Nuke, Fusion) and it becomes really cumbersome to use for sequences longer than a few hundred frames.
The politics and culture are the primary limiting factors in free software projects though, agreed.
yeah blender VSE was built by a guy and he only implemented things that where needed by him
you should check this guy’s thoughts on the future of blender VSE as i really hope the blender foundation clear up some resources and get a dev ti mirror this design
Honestly it’s not a priority for me, I just use resolve. Blender is a great 3d program, that I use almost daily, but “being less shitty than MLT-based editors” is hardly a benchmark for success.
I do support Sobotka’s ‘filmic blender’ project fully, though.