userland: 268 applets, an editor, names that resolve, and a web server

Four complaints, and only one of them was the thing it looked like.

**"toybox is too small."** Half true, and the fix cost no licensing at all:
236 of toybox's applets are what defconfig builds, and the rest live in
toys/pending/ -- awk, bc, diff, more, expr, tr, hexdump, stty among them. Same
0BSD tarball, switched off by a config default. 26 of them are on now, 236 ->
268. Not everything: strace wants ptrace, syslogd wants BSD's facilitynames[],
git wants OpenSSL, and those three break the link.

The toybox build also works again on this machine. Its scripts start
#!/bin/bash, which does not exist on NixOS, so the binary in the tree was one
nobody could reproduce; the shebangs are rewritten to /usr/bin/env bash on the
way in. Pending applets are enabled by editing .config directly, because
`make oldconfig` is the obvious way and undoes it -- every pending symbol is
`default n`.

**BusyBox was asked for and is not here.** It is GPL-2.0-only, and `make
ramroot` links the root image *into the kernel binary*, so it would put GPLv2
inside an Apache-2.0 executable -- which is the exact case ports/Makefile's
NOTPORTS filter and docs/licensing.md §4 exist for. nano is the same answer one
version later. Both recorded in licensing.md with what was done instead, since
"no" without an alternative is how a rule gets quietly broken later.

**An editor that is not vi.** kilo, BSD-2-Clause, one 1300-line file, no
dependencies beyond termios and TIOCGWINSZ, both of which zsh already needed.
Installed as /bin/edit as well as /bin/kilo, because somebody with a file to
fix will type `edit`.

**"wget doesn't work."** Not a userland problem, and BusyBox's wget would have
failed identically: nothing on this system could turn a name into an address,
so every program taking a hostname failed in its own way and each looked like
a separate bug. There is still no DNS client in SchOS and there does not need
to be -- musl has a complete resolver and it works over ordinary UDP sockets.
The one missing fact was which nameserver, which the kernel has had all along
from DHCP and never published. /proc/resolv.conf publishes it in resolv.conf's
exact syntax; cmd/resolvconf decides whether it becomes /etc/resolv.conf, and
will not overwrite a hand-written one without -f.

  # wget -O - http://example.com/ | head -1
  <!doctype html><html lang="en"><head><title>Example Domain</title>

**Something to host with.** cmd/httpd: GET and HEAD, static files, index.html
for a directory and a listing otherwise, a child per connection. No CGI, no
keep-alive, no TLS, no config file. The path check is the only
security-relevant code in it and it refuses rather than normalises -- no
component may be `..`, and percent-decoding happens first, because %2e%2e is
`..` and a check before decoding does not see it. Both forms answer 403.

And the disk is 100 MB, from 32. It is sparse, so the image is 21 MB on the
host. RAMDISK_BLOCKS went 6000 -> 8000 with it: the ramdisk was at 86% and the
next thing added would have broken `make ramroot`, which nobody would have
noticed until they used it.

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01VQ7Qx8E1q1M1bbCq8QLHS2
This commit is contained in:
Nexoniarz 2026-09-14 19:24:55 +02:00
commit b08432b2c3
17 changed files with 994 additions and 21 deletions

View file

@ -212,6 +212,43 @@ with SchOS. Keeping the distinction sharp is what stops it eroding:
rebuild never needs the network; not part of any image, and not all of it is
redistributable.
### The userland this rules out, and what was done instead
The rule costs something real and it is worth naming the bill rather than only
the principle.
**BusyBox** is the obvious userland for a system this size — about four hundred
applets, including a shell, an editor, `less` and an `httpd`. It is
**GPL-2.0-only**, so §2 applies, and here the consequence is sharper than
"cannot be linked": `make ramroot` links the root filesystem image *into the
kernel binary*, so a BusyBox in `sysroot/` would put GPLv2 code inside an
Apache-2.0 executable. Excluding it from the ramdisk avoids that and leaves the
ramroot build with no userland at all, which is not a trade worth making.
**nano** is the same answer for the same reason, one version later: GPL-3.0,
which is incompatible with Apache-2.0 in the other direction.
What was done instead, when the complaint "the userland is too thin" came up:
| wanted | answer |
|---|---|
| `awk`, `bc`, `diff`, `more`, `expr`, `tr`, … | already in toybox, in `toys/pending/`, switched off by `defconfig`. Turning 26 of them on took the shipped set from 236 applets to 268, at no licensing cost — it is the same 0BSD tarball. |
| an editor that is not `vi` | [kilo](https://github.com/antirez/kilo), BSD-2-Clause, one file, no dependencies. Installed as `/bin/edit` as well as `/bin/kilo`. |
| `wget` that works | not a userland problem at all. See below. |
| something to host with | `cmd/httpd`, written here, Apache-2.0. |
The `wget` one is worth recording because it was diagnosed wrongly at first,
and switching userland would not have fixed it: **SchOS had no DNS reachable
from a process**, so every program taking a hostname failed and each looked
like its own bug. BusyBox's `wget` would have failed identically. musl already
contains a complete resolver; the only missing fact was which nameserver, which
the kernel had all along from DHCP and never published. See
[architecture/networking.md](architecture/networking.md).
The general shape: **a permissive answer existed for every one of these**, and
three of the four were already in the tree. That will not always be true, and
when it is not the answer is still no.
The mechanisms that keep this from eroding, since good intentions have already
failed once here: