About
Why maux exists
maux comes from a place of deep respect, and honestly of love, for two systems that shaped how a generation thought about computers: the Apple II and Unix.
The Apple II was the machine on the desk: the one you could open up, wire into, and know down to the last soft switch. Unix was the system in the machine room: pipes, a shell, and small tools that each did one thing well, running on a PDP-11 that most of us only read about. Plenty of people writing Apple II software in those years would have given a great deal for that PDP-11 beside them as a development host — a real editor, a real shell, make, and an assembler that could see the whole project.
maux is that wish, granted on the Apple itself. It is a sanctuary where you can develop for your Apple II as if the PDP-11 were in the next room: edit in vi, assemble with nas, link with ld, drive it all with make, then run the result on plain ProDOS with bye and come straight back to where you were.
It is meant to be a happy place. A place where a program is something you can hold in your head, where every byte has a reason to be where it is, and where you can remember what real programming was like — and do it again.
Thanks
Nothing in maux is new. All of it was given to us first, by people who never knew we would be here.
Steve Wozniak designed the Apple II nearly single-handed, and designed it to be understood: open slots, soft switches with names, a disk controller made from a handful of chips, a monitor in ROM, SWEET16 tucked in beside Integer BASIC. A IIe is still his machine at heart, and every byte maux moves goes through his decisions. Steve Jobs wanted that machine to be a finished thing — a quiet box in a plastic case that anyone could buy, carry home and switch on — and made sure it was. Between them they put a real computer on the kitchen table.
At Bell Labs, Ken Thompson and Dennis Ritchie built Unix on a little-used PDP-7 and then a PDP-11, and made a system small enough to hold in your head and strong enough to outlive every machine it ran on. Ken wrote ed, and grep fell out of its g/re/p; Dennis gave Unix C and the manual's plain, exact voice, which the pages on this site try to borrow. Around them: Doug McIlroy, who argued for pipes until they existed, and with them the habit of small programs joined together; Steve Bourne, whose shell maux's sh follows; Stu Feldman, who wrote make; Joe Ossanna, whose nroff is why a manual page still looks like a manual page; and Brian Kernighan, whose Software Tools, with P. J. Plauger, showed that the ideas could travel to machines that were never meant to run Unix — which is all maux really is. And from Berkeley, Bill Joy, who gave us ex and vi.
Then the line runs on, through the people who wrote for the Apple II itself. Glen Bredon, a topologist by profession, wrote Merlin, the assembler many of us actually used, and ProSel, and KIX — a Unix-flavoured shell for ProDOS, with its own ls, grep and ed. He had this dream before we did. maux still carries his Merlin as as, and its own assembler, nas, speaks his dialect. After he died in 2000, his widow released all of his Apple II work, source and all, into the public domain, and that gift is part of why maux can exist. The source as is built from is the documented version 2.59 release that Willi Kusche assembled and Call-A.P.P.L.E. published: a disk put together so that anyone can rebuild Merlin from the published sources and check the result against the original binary. Releasing a thing is one job and curating it so it still works thirty years later is another, and maux depends on the second one. David Schmenk gave the Apple II PLASMA, a real language with modules on a virtual machine small enough to live beside your program; nas is itself a PLASMA program, so when maux rebuilds itself, every byte passes through his work.
Then the archives. Andy Nicholas wrote ShrinkIt and designed NuFX, the format nearly every Apple II file you can download today is still wrapped in; maux’s nulib2 was checked against ShrinkIt 3.4 itself, and unpacks what it packs and packs what it unpacks. Andy McFadden’s NuLib2 and nufxlib are why those archives stayed readable off the machine in all the years since, and his LZW code and notes are what maux’s own encoder was written against — the name nulib2 carries on the disk is his.
None of them had anything to do with maux, and none of them is answerable for it. All of it stands on what they made. Where something here feels right, it is because they made it so; the mistakes are ours.
Licences
maux's own code is MIT licensed. Vendored or externally obtained ROMs, ProDOS, PLASMA and emulator components keep their own licensing and distribution conditions; no ROM bytes are in the tree or in a release. Merlin-8 is public domain: Glen Bredon's widow released all of his Apple II software and source on 2000-08-24.
The manual pages are a mix. Some are original to maux and MIT. Many are adapted from
the oldest publicly available, generously licensed page: Research UNIX V1, V6 and V7
pages under Caldera International's 2002 "Ancient UNIX" grant, and 4.3BSD-Reno and 4.4BSD
pages under their file-local permissive notices. Each derived page keeps its provenance
(source, retrieval date, checksum of the retrieved response, licence) in its source and
at the foot of its page here. The notices themselves are on the disk as
man licenses and on this site as licenses(1),
and the register is docs/man/SOURCES.md; the
grant letter is CALDERA-LICENSE.
This product includes software developed or owned by Caldera International, Inc. This product includes software developed by the University of California, Berkeley and its contributors.
Tools and pins
Beyond the names above, maux is built with merlin32 and ACME on the host, a2til for disk images, and mandoc for these manual pages. It is tested against MAME and against verilogapple. Their licences and exact revisions are pinned in dependencies.lock.json.