Conference azur::mcc

1022.0. "ASN.1 REAL support progress report" by TOOK::KOHLS (Ruth Kohls) Thu May 16 1991 11:06

The mcc_asn_put_ref and mcc_asn_get_ref routines now successfully
encode and decode the ASN.1 REAL form on VAX and MIPS hardware.

This code will be reviewed and in the cms class ut1.2.0 for builds that
take place after 5:00 pm Thursday, May 16, 1991.

Work is underway to make the mcc_asn_xxx REAL truly platform-independent.  
Until that happens, the mcc_asn_xxx routines will use G-float on VMS, D-float 
on Ultrix-VAX and the standard double (IEEE) on Ultrix-MIPS.  

	NO qualifiers are needed in ultrix Makefiles to use reals.

	To use G-Float on VMS, you must compile your module with
	the /g_float/DEFINE="CC$mixed_float" switches. (and use VAX C
	v 3.1 or better).

Note that C compilers and C math libraries default or promote floating point 
numbers to doubles.

If you have a real with an exponent outside of the range �127, and you
decode it on Ultrix-VAX, you will get a CVR MCC_S_ASNEXPOUTOFRANGE,
the returned value will be 0.0 (for exponent -127 or smaller) or HUGE_VAL 
(for +128 or bigger--taken from math.h).  This CVR is a warning--if 
the values returned  are acceptable for your calculation, ignore it.

Note:  the above CVR and values are also used to represent � infinity.

Ruth