Skip to content

Character Set (CP437, UTF-8 or ASCII)

The game can send its screens in three character sets: CP437, UTF-8 or plain ASCII. This page explains the difference, which one to use, and how to choose.

Quick answer

  • By default the game sends CP437 (what traditional BBS terminals expect). Use -utf8 for UTF-8, -cp437 to force CP437, or -ascii if neither looks right.
  • -local auto-detects from your locale — UTF-8 when your locale is UTF-8 (common today) — so local play usually needs no option.

The three character sets

  • CP437 is the classic BBS and DOS character set. Traditional terminals such as SyncTERM and NetRunner expect it. The box-drawing lines and block shapes in the game's screens come from CP437.
  • UTF-8 is the modern standard. Most terminals use it today.
  • ASCII is the plain 7-bit set every terminal agrees on. The game substitutes look-alikes for what ASCII lacks: box rules become -, | and +, shaded blocks become #, and an accented letter loses its accent. The artwork is plainer, but nothing can be mis-decoded.

CP437 and UTF-8 are not the same at the byte level. If the game sends one and the terminal expects the other, the screen looks wrong: garbled boxes and stray symbols instead of clean lines. ASCII is the way out when you cannot tell which one a caller has, or when their terminal is set to something else entirely (Latin-1, say, which is a common default in modern telnet and ssh clients).

The options

  • -utf8 — force UTF-8 output.
  • -cp437 — force CP437 output.
  • -ascii — force plain ASCII output.
  • None of them — the game sends CP437, unless you use -local, which auto-detects from your locale (see below).

You can use only one at a time. A separate option, -no-ansi, controls whether the game sends colour and cursor control at all; the two questions are independent, and a terminal can need both. See the Command Reference for the full list of command-line options.

Auto-detection

The game auto-detects the character set only with -local: it reads your terminal's locale (LC_ALL, LC_CTYPE, then LANG) and uses UTF-8 when the locale is UTF-8, so local play usually needs no option. To see CP437 on a UTF-8 system, run with -cp437 (or set a non-UTF-8 locale, for example LANG=C).

Without -local — for example when the game runs as a door, serving a remote caller — there is no local terminal to read, so the game uses the default (CP437) unless you pass -utf8 or -cp437.

Note: the Linux text console (a raw virtual terminal, not a terminal window) is UTF-8. Use -utf8 there if the game does not pick it — a bare console login sometimes has no locale set.

If the game could better fit how you run it — for example, by detecting the character set when it runs as a door — please open a feature request.

Languages

The game ships translations in other languages. Which ones you can pick depends on the character set.

  • UTF-8 — every language is available. UTF-8 can show any character.
  • CP437 — a language is available only if all of its characters exist in CP437. German works, for example, because CP437 has the letters ä, ö, ü, and ß.

Languages that use characters CP437 does not have are not offered in CP437 mode. This includes Russian and other Cyrillic languages; Czech, Polish, and similar languages (their accented letters); and Chinese, Japanese, and Korean. To use one of those, run with -utf8.

  • ASCII — the same rule, more strictly. A language whose letters merely lose their accents is still offered, since it stays readable; one that would come out as question marks is not.

The game works this out for itself. It checks each translation, and if even one character in it cannot be shown in CP437, that language is left off the CP437 list — so the screen never shows broken characters. Nothing is hard-coded: if a translation later adds a character outside CP437, that language drops from the CP437 list on its own.

In CP437 or ASCII mode the language menu (under Preferences) lists only the languages that character set can display. If a language was already set in a UTF-8 session and it does not fit, the session falls back to English rather than showing broken characters.

Testing a character set

  • To see CP437 the way a classic caller would: connect with SyncTERM, or wrap your connection with luit -encoding CP437.
  • To see UTF-8: use any modern terminal set to UTF-8.
  • To see ASCII: run with -ascii on any terminal. It looks the same everywhere, which is the point of it.