FreeCAD 0.19 Released For Advancing Open-Source CAD Software

[stiiixy]

FreeCAD deal with liquid/gas flow rates etc, for integrated verm/aquaponics in my case, or am I looking in the wrong news stand with CAD?

[tchiwam]

FEM should be possible to benchmark. Assembly time to load etc etc... How stop the watch is the problem

[uid313]

I've started to learn to code with the Go programming language and am starting to feel comfortable with it and have for a long time wanted to try to add CAD as a hobby and saw there was OpenSCAD. But then it dawned on me that what the OpenSCAD language does can probably already be replicated with other programming languages and sure enough I saw there were projects such as OOML, SolidPython, and sdfx.

Now I'm somewhat conflicted. Do I go ahead and learn the CAD specific language OpenSCAD or just use one of the projects that allows you to model in a more general language and cut OpenSCAD out of the picture?

I'm going to pick up other languages after Go but not sure OpenSCAD really needs to be one of them. Looking at an example on the SolidPython git page:

from solid import *
from solid.utils import *
d = cube(5) + right(5)(sphere(5)) - cylinder(r=2, h=6)

does indeed look cleaner than

difference(){[INDENT]union(){[/INDENT][INDENT=2]cube(5);
translate( [5, 0,0]){[/INDENT][INDENT=3]sphere(5);[/INDENT][INDENT=2]}[/INDENT][INDENT]}
cylinder(r=2, h=6);[/INDENT]
}

For now I'm going to push that decision down the road because I'm actively working my way through exercism and have other things I want to pursue before CAD.

Maybe you want a OpenSCAD library in Go?
Well, it depends on. With OpenSCAD you also have a software with a GUI and render preview, with Python or such you have to open the code, then save it, then run it, then look at the output or something to see if it is what you want. But if you think the Python syntax is more easy and cleaner, then go for it. But then it also depends on if the libraries are alive and well-maintained and such.

[kenjitamura]

Hopefully it can soon be ported from Qt 5 to Qt 6.

Another interesting CAD software is OpenSCAD in which you don't draw with designer tools but you declare the drawing with code.

I've started to learn to code with the Go programming language and am starting to feel comfortable with it and have for a long time wanted to try to add CAD as a hobby and saw there was OpenSCAD. But then it dawned on me that what the OpenSCAD language does can probably already be replicated with other programming languages and sure enough I saw there were projects such as OOML, SolidPython, and sdfx.

Now I'm somewhat conflicted. Do I go ahead and learn the CAD specific language OpenSCAD or just use one of the projects that allows you to model in a more general language and cut OpenSCAD out of the picture?

I'm going to pick up other languages after Go but not sure OpenSCAD really needs to be one of them. Looking at an example on the SolidPython git page:

from solid import *
from solid.utils import *
d = cube(5) + right(5)(sphere(5)) - cylinder(r=2, h=6)

does indeed look cleaner than

difference(){[INDENT]union(){[/INDENT][INDENT=2]cube(5);
translate( [5, 0,0]){[/INDENT][INDENT=3]sphere(5);[/INDENT][INDENT=2]}[/INDENT][INDENT]}
cylinder(r=2, h=6);[/INDENT]
  }

For now I'm going to push that decision down the road because I'm actively working my way through exercism and have other things I want to pursue before CAD.

[kpedersen]

Engineering in general has a dire need for tools such as FreeCAD to further develop to the point they can easily and viably replace commercial offerings. While many solid free solutions exist, they generally lag behind the better proprietary solutions- at least in my limited experience.

In indie games development I see this as an issue too. In the past we used 3D World Studio to knock up some maps (Blender is not particularly well suited for constructive solid geometry (CSG ), it works but has a poor workflow). As with most commercial stuff, this tool is pretty much unmaintained and there was no way we were going to get it on modern platforms.

In the end I spent far too long hacking on the old Quake III level editor (GtkRadiant) to make it suitable for generic modelling (i.e not tied to BSP, q3map, etc) http://thamessoftware.co.uk/openradiant.html

But this is unlikely to replace commercial offerings. There just isn't enough of a market, people either hack on the open-source stuff to make it suitable for their niche or they just make do with slightly awkward open-source tools. In some way, this is part of the "fun" of computing I suppose.

10% of a project is making it good and working to "open-source standards". The other 90% is making it suitable for replacing the leading commercial tool. Most of this is probably marketing too.

Edit: I actually considered using FreeCAD as a level builder but the codebase was too much to maintain for a single person. There is a very delicate balance here. This possibly also contributes to the issue of open-source projects not being commercial replacements.

[M@GOid]

Engineering in general has a dire need for tools such as FreeCAD to further develop to the point they can easily and viably replace commercial offerings. While many solid free solutions exist, they generally lag behind the better proprietary solutions- at least in my limited experience.

The LWN article, rightfully so, expresses some reservations about the learning curve. I feel that this is the most critical aspect to improve, and I'm certain it will be resolved in due time.

While I'm presently in a situation where the non-profit organization I sysadmin for can easily benefit from generous sponsorships by major proprietary vendors, that won't last, and locks people into learning the ins and outs of software they can't necessarily bring with them into future workplaces or their home. I'm eagerly awaiting further improvements.

I would be nice if they could replicate the financing model of Blender or Krita. Those two really took-off after they started doing sponsorship work.

[fkelava]

Engineering in general has a dire need for tools such as FreeCAD to further develop to the point they can easily and viably replace commercial offerings. While many solid free solutions exist, they generally lag behind the better proprietary solutions- at least in my limited experience.

The LWN article, rightfully so, expresses some reservations about the learning curve. I feel that this is the most critical aspect to improve, and I'm certain it will be resolved in due time.

While I'm presently in a situation where the non-profit organization I sysadmin for can easily benefit from generous sponsorships by major proprietary vendors, that won't last, and locks people into learning the ins and outs of software they can't necessarily bring with them into future workplaces or their home. I'm eagerly awaiting further improvements.

[woife]

I'm really looking forward to the Engraving support in the Path workbench. Currently I have to rely on other tools for that type of stuff, having that integrated in FreeCAD would be nice

Another interesting CAD software is OpenSCAD [...]

There were recently two interesting articles on lwn.net about FreeCAD and OpenSCAD:

• FreeCAD: https://lwn.net/Articles/828037/
• OpenSCAD: https://lwn.net/Articles/829489/

[uid313]

Hopefully it can soon be ported from Qt 5 to Qt 6.

Another interesting CAD software is OpenSCAD in which you don't draw with designer tools but you declare the drawing with code.

[Jedibeeftrix]

Great news.

But becomes truly interesting if paired with libredwg to add support for Autocad's DWG format.
They can't be distributed together due to GPL restictions on libredwg, but nothing to stop an organisation packaging up both in AppV I think...?