Public Comment Number PC-UK0271 ISO/IEC CD2 9899 (SC22N2794) Public Comment =========================================== Date: 1998-09-22 Author: N.M Maclaren Author Affiliation: Self Postal Address: University of Cambridge, Computer Laboratory, New Museums Site, Pembroke Street, Cambridge CB3 3QG, United Kingdom E-mail Address: Telephone Number: +44 1223 334761 Fax Number: +44 1223 334679 Category: Procedural Committee Draft subsection: various Title: Comments on CD1 that have not been responded to Detailed description: The following is a summary of the author's comments on CD1 that have had no response, an inappropriate response, or an inaccurate response. GENERAL COMMENTS ---------------- No response seen, and ambiguous, inconsistent, unimplementable etc. ------------------------------------------------------------------- PC-UK0082 - Source character set values PC-UK0083 - Multibyte characters and C89/C9X changes PC-UK0086 - Invalid but skipped pre-processor directives PC-UK0088 - Direct input/output functions PC-UK0089 - Requiring line buffering on stderr PC-UK0092 - Hard I/O errors and EOF PC-UK0096 - Signal handlers and signal classes PC-UK0098 - Compliance footnote PC-UK0106 - Function specifiers PC-UK0108 - Ill-definition in array declarators PC-UK0114 - Ambiguity in multibyte character functions PC-UK0117 - Major incompatibility with C89 PC-UK0122 - Problems with multiple zero representations etc. PC-UK0171 - Ambiguities and traps in VLAs No response seen ---------------- PC-UK0079 - Source line splicing PC-UK0085 - The escape character PC-UK0090 - Determining whether a stream is interactive PC-UK0091 - Truncating a file PC-UK0093 - I/O error diagnosis function PC-UK0094 - The srand function PC-UK0095 - BCPL muldiv for multi-precision arithmetic PC-UK0097 - Reliable signal handling PC-UK0102 - Linkages of identifiers PC-UK0112 - Symbol providing alignment of objects No response (and extremely serious) ----------------------------------- PC-UK0118 - Enhancements to restrict for optimisation FLOATING-POINT ISSUES --------------------- Simple (and erroneous) denial of ambiguity ------------------------------------------ PC-UK0120 - Rounding in ambiguous cases without IEEE PC-UK0141 - Which functions can set errno PC-UK0142 - Extraordinary roundoff error PC-UK0144 - Serious confusion in the signbit macro PC-UK0155 - The status of __STD_IEC_559_COMPLEX__ Totally inappropriate response (the first is serious) ----------------------------------------------------- PC-UK0126 - Which functions can raise floating-point exceptions PC-UK0150 - Mathematically incorrect behaviour of fmax/fmin PC-UK0151 - Why mention degrees versus radians? PC-UK0154 - Is the cproj function useful? Unimplementable or IEEE-dependent (and inappropriate response) -------------------------------------------------------------- PC-UK0129 - Failure when handling exception flags PC-UK0133 - fesetround is often unimplementable PC-UK0134 - fegetenv, fesetenv and feupdateenv are often unimplementable PC-UK0135 - feholdexcept is sometimes unimplementable PC-UK0138 - Provide IEC 559 required support for signalling NaNs PC-UK0147 - Ambiguity in the copysign function Less severe dependencies (response appropriate, if canned) ---------------------------------------------------------- PC-UK0137 - Allow infinities and NaNs for each precision PC-UK0143 - and floating-point signals Incompatibilities with LIA-1 ---------------------------- PC-UK0156 - Permitting diagnostics for uncleared exception flags PC-UK0158 - Floating-point exception flags and errno More complicated problems (often with an inappropriate response) ---------------------------------------------------------------- PC-UK0152 - Error handling for complex functions PC-UK0153 - carg() is undefined for a complex zero PC-UK0159 - atan2(+-0,+-0) and IEEE PC-UK0161 - pow(+-inf,+-0) or pow(NaN,+-0) and IEEE PC-UK0162 - Complex infinities and NaNs PC-UK0163 - Imaginary, complex infinities, NaNs and inexact PC-UK0164 - Complex multiplicative operators PC-UK0165 - Incorrect claims of LIA-1 conformance Accepted but not done --------------------- PC-UK0130 - fegetround and non-standard rounding modes