# About ScriptFu's TinyScheme
This version of TinyScheme has diverged from upstream.
Upstream is at SourceForge.
Occasionally fixes upstream are manually backported here.
## Changes
These are not upstream:
1. changes enabling wrapping by ScriptFu
2. i18n internationalization
3. unicode via UTF-8
4. byte type
5. string-port changes
## Build differences
This builds differently from upstream.
It might not build standalone anymore.
1. depends on GLib types, superficially
2. depends on external i18n support
3. some files have been extracted from the original scheme.c
4. depends on GLib for unicode support
## Unicode
This discussed implementation of unicode in this TinyScheme.
Scheme strings and string-ports are UTF-8 encoded.
Each char is one to four bytes.
Reserved words in the language are still ASCII.
Symbols may be unicode.
A string's length is in characters, not bytes.
Strings, and the contents of string-ports, are internally still NUL-terminated.
The implementation continues to pass gchar pointers.
Processing scheme strings may require unicode conversions.
The string length value stored in a scheme string's data cell is the length in *characters*.
Most functions taking a string length take the number of characters.
The number of bytes in a string can be had with libc strlen().
## Other forks
Another significant fork of upstream TinyScheme is in the gnupgp project, where it scripts tests.
