- From: <Dmitry_Trifonov@exigengroup.com>
- Date: Mon, 11 Mar 2002 18:51:14 +0300
- To: xmlschema-dev@w3.org
Hello,
I have readed W3C_XMLSchema_TestCollection/diffNIST_double
NISTXML-double-minExclusive-1- NIST File: Test files Expecte
1 NISTSchema-double-minExclusive-1.xs NISTSchema-double-minExclusive-1.xs d
d This schema specifies a datatype, d S: 1
NISTSchema-double-minExclusive-1, NISTXML-double-minExclusive-1-1.xml I: 1
which is derived by restriction of
built-in type double by facet
minExclusive with a value of:
1E-1075
Why value of minExclusive is 1E-1075?
By my mind it is misspelling, because IEEE Standard 754 floating point
defines:
Ranges of Floating-Point Numbers
|-----------------+--------------------+--------------------+---------------------|
| | | | |
| | Denormalized | Normalized | Approximate |
| | | | Decimal |
| | | | |
|-----------------+--------------------+--------------------+---------------------|
| | | | |
| Single | ± 2(superscript: | ± 2(superscript: | ± ~10(superscript: |
| Precision | -149) to (1-2 | -126) to (2-2 | -44.85) to ~10 |
| | (superscript: -23 | (superscript: -23 | (superscript: |
| | ))x2(superscript: | ))x2(superscript: | 38.53) |
| | -126) | 127) | |
| | | | |
|-----------------+--------------------+--------------------+---------------------|
| | | | |
| Double | ± 2(superscript: | ± 2(superscript: | ± ~10(superscript: |
| Precision | -1074) to (1-2 | -1022) to (2-2 | -323.3) to ~10 |
| | (superscript: -52 | (superscript: -52 | (superscript: |
| | ))x2(superscript: | ))x2(superscript: | 308.3) |
| | -1022) | 1023) | |
| | | | |
|-----------------+--------------------+--------------------+---------------------|
Best Regards,
Dmitry Trifonov
Exigen Group (Saint-Petersburg)
Received on Monday, 11 March 2002 10:51:29 UTC