Limiting assumptions
--------------------

[Unless otherwise indicated, these issues are NOT considered bugs.]

The system hosting the compiler is assumed to be POSIX, with 8-bit
characters that adheres to the US-ASCII standard as far as the first
128 positions are concerned.

The Unix newline convention is assumed.

The host system's intmax_t and uintmax_t should match the target's
(otherwise preprocessor arithmetic will not meet C99 requirements).


Intentional deviations from the standard
----------------------------------------

[Unless otherwise indicated, these issues are considered bugs and will
be fixed in the future, but they were deemed nonessential features and
thus were left out of the initial revisions.  Items are prefixed by a
reference to the ISO/IEC 9899:1999 (E) standard document.]

5.1.1.2:1(1) Source files are read as sequences of octets.  No
      multibyte sequences are interpreted.

Universal character names are not interpreted.

6.4.6:3 Digraphs are not recognized.

6.10:1 Multiple #else's before #endif are not diagnosed.