Open Source...everything?

A few days ago there was a post on Solid Angles Arnold beta mailing list about a Python-based early-stage Blender to Arnold connection. The result was a discussion about the impact of open source software on the 3D industry as a whole, which reminded me on  the moment I first read about the plans for a complete Blender overhaul (Python, Qt, architecture revamp, etc) about two years ago. I thought, that this would change the game in the long run for everybody, if they can deliver.


Meanwhile they have delivered, and it's effects begin to show. More and more people have started using Blender, it's even being used in game production of some mid-sized game companies.
As a life-long user of commercial 3D Software (mostly 3dsmax, Maya, and Softimage), I must say it makes sense. Think about it for a second - it's every larger companies dream to have a scalable architecture with full access to every single line of code! One of the main reasons companies are still struggling with Maya ia because it is very open. You can change almost any aspect of the software and work around issues when something's not working (as it is frequently the case), let alone those who write large parts of their software themselves from the start.
You can do the same with Blender, but even more broadly and low-level because you have access to every line of code. If somethings broken or needs an improvement, you can fix or improve it immediately. Sure, not ye average 3D dude from next door, but companies can. That's how Maya got big too. It started in the companies doing feature production, not with single man shows. After that, everybody went: "Oh, I want to use that Software they did Stuart Little with". For those studios it won't make such a big difference if Blender still has missing features, they will implement them themselves, with and on an architecture they have full access too 24/7/365, without the need to wait for the annual or bi-annual update or bug fix, or an Autodesk technician to implement custom code into their private branch of the software.


I dare to say this:

1. In three years time, Blender will be among the top 3 most widely used 3D apps on the market.
2. In 3 to 5 years, we might not all be using Blender, but we will all be using open source 3D software as our main 3D application.

Yes, sooner or later, Maya, Softimage, Max, Modo, Lightwave, C4D,... whatever they currently put price tags on, will be open sourced.
I don't know if that will also happen to CAD programs so quickly, if ever, but I'm almost 100% certain about DCC.
What's clearly noticable is that Blender is being developed faster than any other software atm. And not just feature wise, the software has seen radical changes of it's architecture on a scale only comparable to what has happened to Softimage over the last four years, only faster.  And those who contribute to Blender do it for mostly no money. How could a company, seeking to sell software for profit, ever hope to compete with this in the long run? I dont think they can, until they change their business model into something along the lines of Red Hat or Suse. That is: making custom solutions for paying customers (those who cannot afford three ore more developers full time), support fees, etc.

The support scheme is something that has been taking shape over the last years anyway, hence all those efforts by Autodesk to get customers to invest in subscription plans.
Sure, they could try to escape the open source competition for some time by providing more and more sophisticated high-level tools, cranking up support quality and stability of their products and update frequency,
but the high tech race is run for the 5% of the upper class customers who need a faster fluid solver, crowd simulation, etc.
For the remaining 95% of all users/customers, it's bread and butter features we keep hearing them asking for, like better polygon modeling tools, etc. But there's not much head room in those departments left,
until a certain level of maturity is reached, and then what? Sooner or later all DCC programs will have the same bread and butter features, and then it only comes down to stability and support/speed of bug fixes, flexibility, extendability, and that's exactly where open source software with a large enough user base like Blender, already shines today.

Personally, tbh, I wouldn't mind. The game will be different for those writing software, and easier for those using it. Open source Softimage any one? I'd not look at Blender for a second longer.