| Type | Bug | Status | tested | Date | 7-Feb-2010 23:14 |
|---|---|---|---|---|---|
| Version | alpha 97 | Category | Datatype | Submitted by | Sunanda |
| Platform | All | Severity | minor | Priority | normal |
| Summary | Large percentages not serializable |
|---|---|
| Description |
x: 30000000000000000000000000000000000% == 3e34% No obvious way to mold and restore this value: do mold/all x ** Syntax error: invalid "decimal" -- "3.0000000000000003e34%" do mold x ** Syntax error: invalid "decimal" -- "3e34%" load mold x ** Syntax error: invalid "decimal" -- "3e34%" load mold/all x ** Syntax error: invalid "decimal" -- "3.0000000000000003e34%" **** Slightly smaller values go wrong with a different rounding: x: 3000000000000000000000000000000% == 3e30% >> do mold/all x ** Syntax error: invalid "decimal" -- "2.9999999999999997e30%" |
| Example code |
load "3e34%" or x: 30000000000000000000000000000000000% do mold/all x |
| Assigned to | n/a | Fixed in | alpha 108 | Last Update | 3-Nov-2010 07:08 |
|---|
| Comments | |
|---|---|
|
(0002052)
BrianH 1-Mar-2010 23:24 |
See also #1504.
The issue here is either in MOLD or LOAD. MOLD is generating scientific notation for certain percent! values, and LOAD isn't recognizing that scientific notation. So either MOLD needs to be changed to not generate that syntax, or LOAD changed to recognize it. |
|
(0002081)
Ladislav 9-Mar-2010 10:58 |
The variant for MOLD to not generate that syntax is not an option, at least for some values. Since the PERCENT! datatype has the range of the 64-bit IEEE754 floating point values, the limitation of it looks higly impractical, so we should live with the given range. That means that MOLD has to mold all values in the range as well as that LOAD has to accept them too. |
|
(0002770)
Ladislav 3-Nov-2010 07:07 |
in the core-tests suite |
|
(0002771)
Ladislav 3-Nov-2010 07:08 |
should be reopened, the tests:
#r3only ; 64-bit IEEE 754 maximum ; bug#1475 [same? 1.7976931348623157e310% load mold/all 1.7976931348623157e310%] #r3only ; bug#1475 ; 64-bit IEEE 754 minimum [same? -1.7976931348623157E310% load mold/all -1.7976931348623157e310%] still fail |
| Date | User | Field | Action | Change |
|---|---|---|---|---|
| 18-Aug-2013 14:14 | Ladislav | Comment : 0002081 | Modified | - |
| 3-Nov-2010 07:08 | Ladislav | Comment : 0002771 | Added | - |
| 3-Nov-2010 07:07 | Ladislav | Comment : 0002770 | Added | - |
| 20-Oct-2010 03:08 | BrianH | Status | Modified | built => tested |
| 22-Sep-2010 05:25 | carl | Fixedin | Modified | => alpha 108 |
| 22-Sep-2010 05:25 | carl | Status | Modified | submitted => built |
| 9-Mar-2010 10:58 | Ladislav | Comment : 0002081 | Added | - |
| 1-Mar-2010 23:24 | BrianH | Comment : 0002052 | Added | - |
| 8-Feb-2010 04:07 | BrianH | Summary | Modified | Large percentages nor serializable => Large percentages not serializable |
| 8-Feb-2010 04:07 | BrianH | Description | Modified | - |
| 8-Feb-2010 04:07 | BrianH | Code | Modified | - |
| 7-Feb-2010 23:14 | sunanda | Ticket | Added | - |