This is likely the most up to date, bug-free, and featureful TinyFugue 5 out there.
I am no longer actively developing this, but I will still be responsive to bug reports.
Check out my tinyfugue-scripts repository for a lot of great stuff.
Unicode support is now on by default, so read below on that, as you almost certainly want this feature, you might just need to install a package on your OS.
General feature list and installation follows. The UTF8 section is what you probably what everyone should be aware of, so starting with it for those that tend to skip reading these.
autoconfig ./configure (To see a variety of compile time options: ./configure --help) make (Optionally: sudo make install)
UTF8 Widechar support is enabled by default, but you need to install libicu to get this to work. It is highly recommended that you use this as it makes the game experience much much nicer. If this is working, you will notice immediately as the default status line goes from a dashed line to a solid line. While I am working on many things with this client and happy to take requests or bug reports, proper UTF8 is my main personal interest.
On Debian this requires the following packages: icu-devtools libicu-dev libicu63
You might also need to enable UTF8 support for individual worlds. For example, when connecting to a RhostMUSH server, this requires you to set:
@toggle me=utf8
There are tends of thousands of characters this enables, and no one font supports them all, though the most common core ones tend to be supported by all. I've found a powerline patched font gives me the best experience, and there are a number of options available here:
https://github.com/powerline/fonts
This site rates the support of various fonts and allows you to test specific characters out ahead of time:
https://www.fileformat.info/info/unicode/font/index.htm
Personally, I use DeJa Vu Sans Mono with the powerline patch
Setup:
-
autoconf
-
./configure --enable-python
Python 2.7 or Python 3+ are required for support, as older versions are missing certain features that this version of TF is targetting. If you need to run Python 2.1-2.6, it's recommended to use the the stock 50b8 with the Python patch, available here:
In the future, Python 2.7 support might be removed if the required featureset begins to diverge too much.
To run Python, you will need a few pieces of additional software targetting the specific version f Python you wish to run. On Debian for example, these would be variants of:
libpython-dev python-cxx-dev python python python-config
Usage:
-
run TinyFugue (tf) and then /help tf python
-
scripts are installed in the tf-lib/python directory for a few packaged ones, but you can obviously ad our own.
-
Check out tf-lib/python/urlwatch.py for a simple scripting example. This uses only a tiny fraction of what you can do, but I've been so busy making this patch I haven't had time to write all the scripts I want!
This list covers the changes in this version, though some minor details might have been missed. It will be cleaned up once we're out of pre-release. Game related changes are documented in tf-help
UTF8 support (--enable-widechar=yes) on by default UTF8 now default_charset when UTF8 compiled (previously US-ASCII) PYTHON interpreter support for interfacing with client and macros PYTHON 2.7 and 3 are both supported (--enable-python) off by default PYTHON interpreter debugger available (/help tfpy_debug) off by default /HELP has been spell checked, along with everything else HTML documentation and html2tf found and returned to repo ANSI log output (/help ansi_log) off by default LOGs output prefix, such as timestamp (/help log_prefix) off by default ERROR output prefix now adjustable (/help error_prefix) '%' by default UNNAMED worlds not wrapped in () (/help oldunnamed) no-() by default FG world switching can use TF4 style (/help oldstylefg) off by default @ALERT picks up on messages that could have been lost / as a prefix to /command sends to MU, as opposed to ///command SSL looks for GNU TLS ATCP support (--enable-atcp=yes) off by default GMCP support (--enable-gmcp=yes) off by default OPTION 102 support (--enable-option102=yes) off by default MCCP supprt (--enable-mccp) changed to off by default MCCP protocol lockup issue resolved RFC 2066 compliance for Telnet Charset Negotiation PCRE unbundled and now can use system libraries GCC 10 and CLANG support 64-bit crash issues resolved MULTI-address hostname crash issue resolved OS X compile issues resolved BUFFERS increased to reflect more modern computing resources GZIP now standard for file compression (COMPRESS_SUFFIX) TF-LIB being updated and modernized (see tf-lib/README)
Plus numerous bug fixes, clean-up, and needed code base clean-up and modernization.
Significant effort has been put into finding the various community patches for TF50b8 and integrating them all into one place. There are still several patches in the queue waiting to be examined and then potentially applied.
In addition, new work has been added to this code base and will probably continue to be for some time. The intention is to make sure a well loved piece of software is still able to be run on modern platforms and future platforms, increased stability, and then lastly to add some additional features in-line with the existing codebase.
Similar efforts are being done with TinyFugue Macros, and some may later be integrated into tf-lib:
https://github.com/warmstarter/tinyfugue-scripts
Deprecations:
NO_FLOAT deprecated. This has been broken for over a decade and has little purpose on modern computer systems.
OS/2 deprecated. This was not being maintained and likely not used by anyone. If needed the old sources are in the git history.
CYGWIN deprecated in favor of WSL for Windows, to better keeps platofmrs in-sync and remove issues.
BeOS deprecated as likely no longer used. Should still compile, but might require some modifications for installation of binary, no way to test.
SOCKS deprecated as it was broken and mostly removed anyway. Very little use in modern world and can be handled by a separate application.
NO_HISTORY and NO_PROCESS options deprecated as modern systems can handle them just fine. If you don't use these features, you can simply ignore them or use the existing tfrc options to set things like histsize to 0.
systype is deprecated and removed, as all modern platforms support unix-like features (this mostly relates to OS/2 deprecation)
auto_fg was already deprecated, now fully removed.
oldstyle filenames are deprecated in favor of newstyle, can always be changed manually, newstyle was the default.
ancient bad keyboards no ones uses anymore no longer supported. If needed the .tf files can be founds in the git archive
.tinytalk deprecated and seemed to be broken anyway.
RECEIVELIMIT check no longer works, as there's really no meaningful way to do it. This number was raised so high for servers using 102+ACTP+etc.. That it's no longer meaningful. On some servers 256+ bytes is broken subnegotiation, on others 32K+ is broken subnegotiation. Likely someday it will be 64K+.
======
Unicode notes: "Release" is the branch you want. On Debian or Ubuntu, make sure you apt-get libicu-dev before running ./configure. There is a known and easily triggered bug: The backspace, delete, and arrow keys move by bytes, not by characters. For now, editing a line with multibyte characters will send XML hexadecimal escape codes if unencodeable bytes are in the input box. I will gladly accept pull requests.
Setting your terminal to send ^H instead of ^? for backspace mitigates some, but not all of the backspace issues.
For servers that don't negotiate charsets, /set default_charset=NAME before defining a world with /addworld or /test addworld(). I recommend putting /set default_charset=Latin-1 at the top of your worlds.tf.
======
This fork of TinyFugue is currently being maintained by wstarter. It also includes assorted patches developed by various community members in the years since TF saw saw it's final release. More complete credits will be added as this moves out of pre-release
TinyFugue is Copyright (C) 1993-2007 Ken Keys ([email protected])
For a long time TinyFugue was written, maintained, and copyrighted by Ken Keys ("Hawkeye"), who can be reached by email at [email protected]. His final release was 50b8 in January 2007, but he is still active in giving support at https://sourceforge.net/p/tinyfugue
TinyFugue versions 1.5.3 and earlier were written by Greg Hudson ("Explorer_Bob"), without whom I would have never gotten into this mess.
TinyFugue was initially derived from tinytalk, a unix client written by Anton Rang ("Tarrant") and modified by Leo Plotkin ("Grod").
Additional thanks to Jeff Prothero ("Cynbe"), Bruce Sterling Woodcock ("Bruce"), David Moore ("OliverJones"), Ben Jackson ("Ben"), the mudders of glia.biostr.washington.edu, and everyone else who tested, found bugs, or came up with suggestions, and to Jim Lick ("Jingoro") for providing tcp.com services.
============================================ TinyFugue for Mac OS X
You must have developer tools installed on your system, and of course this is setup and used through Terminal or an equivalent program. I hear most people think iTerm2 is the way to go, should you want more features in your terminal emulator.
TinyFugue for Win32
The likely best option for running TinyFugue under Windows is to get 'Windows Subsystem for Linux'. This should be straight-forward and give you an almost identical experience to a TinyFugue on UNIX like systems, but it has not yet been tested.
https://docs.microsoft.com/en-us/windows/wsl/install-win10
======================================================================
TinyFugue should work on any Unix-like system that has an ANSI compiler and BSD-style select(). This includes almost everything except some old SVR3-like systems which either do not have select(), or have a select() that works only on STREAMS devices but not terminals.
Installing TinyFugue is quite easy on most Unix-like systems. In the top directory, execute these commands: autoconf ./configure make make install
The configure script will attempt to figure out everything it needs to know about your system and install all files in their proper location. If you have any problems, or want to change a default installation option, be sure to read the rest of this file.
In addition to the usual GNU configure options, TinyFugue's configure script has these options:
Optional Features: --disable-option-checking ignore unrecognized --enable/--with options --disable-FEATURE do not include FEATURE (same as --enable-FEATURE=no) --enable-FEATURE[=ARG] include FEATURE [ARG=yes] --enable-version insert version number into names of installed files --enable-symlink=NAME make a symlink NAME to the executable --enable-core enable core dumps for debugging --disable-ssl disable SSL connections --disable-getaddrinfo disable getaddrinfo implies --disable-inet6 --disable-inet6 disable IPv6 network support --disable-256colors disable 256 color support --enable-python enable python interpreter for scripting --disable-widechar disable wide character UTF-8 support --disable-termcap use hardcoded vt100 codes instead of termcap --enable-termcap=LIB enable termcap with library LIB (e.g., "ncurses") (needed only if configure guesses incorrectly) --disable-mailcheck disable mail checking --enable-mailcheck=DIR enable checking for mail in directory DIR (needed only if configure guesses incorrectly) --enable-atcp enable ATCP protocol support --enable-gmcp enable GMCP protocol support --enable-option102 enable option102 protocl support
For a complete list of options, run "./configure --help".
The following environment variables are meaningful to configure. Their default values are listed in brackets after the description. They can be set on the command line like this: env CPPFLAGS=-I/my/include ./configure
CC C compiler ["gcc" or "cc"] CFLAGS Options for C compiler ["-g -O2" or "-O2" for gcc, "-g" for others] CPPFLAGS Options for C preprocessor (e.g. -I) [""] LIBS "-l" and "-L" options for the linker (in addition to the ones determined automatically by configure).
Example: If your system has SSL libraries and headers installed in /usr/local/SSL/{lib,include}, use the "--with-inclibpfx=/usr/local/SSL" option with configure.
If you do not define the file prefix with ./configure --prefix=PREFIX, configure will choose a prefix:
"/usr/local" if you can write to /usr/local/bin and /usr/local/share; otherwise, $HOME.
The executable is installed in
If --enable-version is given, the tf executable and library directory will have a version number embedded in their names (e.g., PREFIX/bin/tf-50a1 and PREFIX/lib/tf-50a1-lib). This is useful if you want to keep old versions around. To create a symlink from "tf" to the actual executable, use --enable-symlink (this is not implied by --enable-version, as in earlier releases).
There is no man page for tf. For basic usage options, run "tf -?". For more information, use "/help" within tf.
If you move or rename any files after installing (this includes copying TF to another machine where the file paths are different), you must let TF know. To do this without reinstalling, you can set the environment variable TFLIBDIR to the new path before running tf, or run tf with the -L option. This can be done with a shell script like this:
#! /bin/sh
exec /new/location/bin/tf -L"/new/location/lib/tf-lib" "$@"
The variables TFHELP and TFLIBRARY can also be set in the environment, but this should not normally be necessary, since their default values are based on TFLIBDIR.
If you want build tf for installation in one directory ("{FINAL}"), but you want to be able to test it by installing it in another directory ("{TEMP}") first, do: ./configure --prefix={FINAL} make env prefix={TEMP} make -e install Then, you can test tf by running {TEMP}/bin/tf -L{TEMP}/lib When you are satisfied it works, do make install to install tf in {FINAL}.
Options are not available on all systems. To use an option, give it as an argument to the installation program (e.g., "make clean").
all Compile, but do not install (the "install" option should be run later to install the files). install Compile (if you haven't already) and install. The default installation locations depend on the operating system. clean Remove object files and other junk from source directory. uninstall Remove tf executable, help files, and library from their installed locations.
Some features of TF can be disabled for secure public installation, by using one of these /restrict commands in %{TFLIBDIR}/local.tf:
/restrict SHELL Prevents all access to shell or external commands.
Disables TF builtins "/sh" and "/quote !", and
uncompression during /load and /help.
/restrict FILE Prevents reading and writing of files. Disables
TF builtins "/load", "/save", "/saveworld", "/log",
and "/quote '", and sockmload feature.
Implies /restrict shell.
/restrict WORLD Prevents the user from defining new worlds and
connecting to undefined worlds. TF builtins
/addworld and the "/world <host> <port>" semantics
are disabled. Implies /restrict file.
If you have write permission in /usr/local/bin and /usr/local/lib, TinyFugue will be installed there by default. Or you can change the locations by using the options described above. The Makefile will then put all files in their proper location with the proper permissions.
Some features of TF can be disabled for secure public installation, by using the /restrict commands in %{TFLIBDIR}/local.tf.
If the default terminal handling option does not work, tf can be configured to use vt100 codes or nothing at all.
On a UNIX-like system, ./configure will try to find a termcap-compatible library in this order: the one specified in --enable-termcap=LIB, if any; termcap; ncurses; curses. If none of those work, tf will use hardcoded vt100 codes.
You will be much, much happier not using a hardcoded situation. The proper libraries should be available on any modern system. If for some reason you find you are stuck in hardcoded mode after having installed appropriate libs, create an issue with system details so this can be looked into. hardcoded is ancient and undesirable enough that it might likely stop being supported in the future.
If you are short on disk space, you can compress the helpfile and library files (except stdlib.tf). Make sure the COMPRESS_SUFFIX and COMPRESS_READ macros are set correctly; set them in the %{TFLIBDIR}/local.tf file if needed. Note that compression can not be used if /restrict is used.
TF can be made to connect through a generic proxy server by setting the %proxy_host variable at runtime. See "/help proxy".
Also see the README file in the subdirectory corresponding to your operating system to see if tf has transparent firewall support on your system.
The ./configure script can usually figure out everything it needs to know to configure TinyFugue for your unix-like system. But if it fails because your system is non-standard, you may need to change some environment variables or disable some features.
If you run into issues with a specific platform, please put details into the issue tracker. I will work to make everything work more automatically, but some issues are outside the scope of TF. Common problems and their fixes will be noted.