[Ericson2314]

I think it's definitely possible. Look at Blender and GRASS GIS, some of the most successful open source projects among non-programmers of all time. Architecture tool is merely the midpoint between those too domains, right? :D

[wolfgke]

Blender was originally a closed-source product by Not a Number Technologies (NaN). After NaN went bankrupt, the creditors agreed to let Blender become released as open source for 100,000 €.

[nybble41]

True, but not really relevant. (Unless you're implying that you think Autodesk is likely to go bankrupt and might be open to a similar arrangement?)

The Blender we have today is much improved compared to the last version developed by NaN. Most of its major features today did not exist when it first become open source. It was certainly helpful to have a working core at the start—I'm not saying that the 100k € was a poor investment, by any means—but I would be surprised to find that there is much left of the original code by this point.

[danShumway]

Agreed. Of course I don't think this is what GP was implying, at all, but for the record it would devalue a ton of work to say that Blender is where it is because it started as a closed source product.

Blender is where it is because the core organizers over the past decade have had a clear vision, great community management, and have worked their butts off to make the software as good as it is today -- in terms of UX, capabilities, performance, marketing... everything across the board.

When I started using Blender, it was not as powerful or usable as other software in the field. It was impressive, and having a good core did help, but the program today has just advanced so far, I'm not sure it's really comparable.

See also Krita, for another Open Source project I think is headed in the same direction as Blender. None of this stuff happens by chance, developing Open Source software that's popular with a general non-programmer audience is really stinking hard, and the teams that can pull it off deserve all of the praise they get.

[IshKebab]

Architecture tools are way more niche. I think a better place to look is mechanical CAD: how many decent open source mechanical CAD programs are there? Approximately one: SolveSpace. And that is quite basic - even simple stuff like bevels is not supported. Absolutely nobody whose livelihood depended on it would use it.

It's just too niche and too complicated for there to be enough developers willing to writing it for free.

[Ericson2314]

My uneducated guess was that engineering tools have a lot more hard requirements than architecture CAD. That's why architects higher engineers to say whether their buildings will fall down!

[rbanffy]

How much money would be needed to turn SolveSpace into something usable?

[neutronicus]

Call Siemens and ask what they charge for Parasolid. Hint: you have to ask, so...

Computational Geometry is hard. FOSS CG Kernels are not going to spring from the ether. CGAL exists, but its target audience is CG researchers. OpenCASCADE exists, but it's really a loss-leader for proprietary extensions and consulting contracts.

[rbanffy]

Maybe they can spring from a consortium of interested architects.

[IshKebab]

At a complete wild guess I would say something like $50m.

(Obviously it depends what you mean by "usable" - I use it now, so in a sense it is already usable. I guess you mean to get to the point that businesses would rely on it.)

[neutronicus]

Blender is programming-adjacent (a lot of Blender users are employed by software businesses to create assets for distribution with software products). In particular, I think that's where a lot of the funding for it comes from.