[aeonik]

His work is definitely worthwhile. Wolfram Language by itself is actually very cool. I really wish he would open source it. I'm not going to use any math package that isn't open source for anything.

I think his physics work is pretty interesting as well.

I think the reason he aggrandizes, and this is only speculation, is because he is pushing the envelope in many fields simultaneously and wants to "credentialize" himself, and give people context on his background.

He might also be insecure self-absorbed. Which, while annoying, doesn't disqualify someone's ideas by itself. It just means you might need to watch for ego-based confabulations in their work.

It might also mean he's a pain to work with, if it is indeed an ego thing. But I don't know the guy... so it's all speculation.

[aleph_minus_one]

> His work is definitely worthwhile. Wolfram Language by itself is actually very cool. I really wish he would open source it. I'm not going to use any math package that isn't open source for anything.

Here is the official statements of Wolfram Research on this topic:

> Why Wolfram Tech Isn’t Open Source—A Dozen Reasons

> https://blog.wolfram.com/2019/04/02/why-wolfram-tech-isnt-op...

See also

> Episode 1: Thoughts on Wolfram & Open Source

> https://soundcloud.com/wolframresearch/wolfram-on-open-sourc...

If you nevertheless insist on having an open source implementation of the Wolfram language, have a look at the answers at

> https://mathematica.stackexchange.com/questions/4454/is-ther...

[aeonik]

I don't have an answer for him. I don't mind paying money for software, I get his points, but at the same time, it's a non-starter for me.

When it comes to math. Especially a system as cool as Wolfram Language, I want to hack on the insides.

I don't publish things, or have time right now. But in 10 years, I estimate I'm going to be wanting to compile math software on radically different chips and ISAs.

Without an open source language, everything I build on top of the system will be incompatible with hardware specific instructions, unless they compiled it for that chip, often missing very specific optimizations (which could be improved even today). Or I would need to virtualize.

All this math code will the most important and hottest running code paths in my system, and won't even be able to compile to experimental ISAs or their extensions to take advantage.