Re: 25 years later: Bill Schelter and the resilience of Maxima

Dear David,

Thank you for reminding us all of the date of William Schelter's 
passing. We indeed have a lot to thank him for, as well as those who 
worked on Macsyma in decades prior.

I am sure I also speak for others when I thank you also personally for 
the incredible amount of bug fixing you have done lately (I know the AI 
helped but I also know it took someone to guide and use the AI properly, 
efficiently but not blindly, as an assistant not as an all-knowing 
oracle -- a subject we discussed here before.)

Maxima is worthy of our effort I think. It may not appear as polished as 
some of the commercial packages but it has some remarkable, unique 
practical strengths. For what it's worth, I use Maxima intensively on a 
daily basis, and -- even if I occasionally invoke, say, Maple for a 
nastier-than-usual integral or ODE/PDE -- it remains by far my preferred 
CAS.


Viktor




On 2026-07-30 01:09, David Scherfgen via Maxima-discuss wrote:
> Hi everyone,
>
> Today, July 30th, marks exactly 25 years since William Schelter passed 
> away. Even though I never knew him (I was 16 years old back then and 
> the only CAS I knew was "Derive" from our Calculus class), his name 
> has appeared on my screen countless times when starting Maxima or GCL.
>
> I was recently browsing the mailing list archives from August 2001 
> (https://maxima.sourceforge.io/ext/list_archives/2001/index.html), 
> right around the time the sad news broke. People were obviously very 
> unsure about the future of Maxima. Reading that with 25 years of 
> hindsight feels surreal.
>
> I found it especially funny to read a message from Camm Maguire 
> stating he didn't know much about Lisp. Fast forward to today, and 
> he's been the one steering GCL through two and a half decades, not 
> only maintaining it, but making it (almost) ANSI-compliant and a real 
> competitor to SBCL when it comes to the performance of the generated code.
>
> Also, think about the hardware evolution that Macsyma/Maxima's code 
> has survived. Macsyma started out on room-sized PDP-10 mainframes. 
> Today, we can compile the code to WebAssembly and run Maxima natively 
> in a web browser on a mobile device that has 10,000x the memory, 
> 100,000x the computational power and weighs not several tons, but 200 
> grams.
>
> There's also the irony regarding Artificial Intelligence. Lisp was 
> originally thought to be the language for AI. The big recent 
> breakthroughs with LLMs didn't happen in Lisp at all, but rather in 
> Python and C++. But things have come full circle: We now have these 
> modern AI tools helping us debug and optimize the ancient Lisp 
> codebase, often finding bugs dating back to Macsyma.
>
> I sent my first message to this list in October 2012, asking something 
> about my method of repeatedly applying various simplification 
> functions. Since then, as I've spent more time digging into the core 
> to fix bugs and improve performance, I've really come to appreciate 
> the culture here. This list is friendly, honest, professional and helpful.
>
> A big Thanks to the veterans and everyone who kept the engine running 
> after 2001, be it by helping users and developers, reporting bugs, 
> fixing them, refactoring code or making Maxima more usable, accessible 
> and powerful in general and in specific areas. I think Bill Schelter 
> would be very proud of the state of Maxima and its community today. 
> Glad to be a part of this crew and give something back to a project 
> that has been hugely important for my professional career.
>
> I can't imagine what the world will be like in another 25 years, given 
> how fast things are moving right now.
>
> Best regards,
> David Scherfgen
>
>
> _______________________________________________
> Maxima-discuss mailing list
> [email protected]
> https://lists.sourceforge.net/lists/listinfo/maxima-discuss