| 1 |
On Wed, 25 Nov 2009 16:49:17 -0800 |
| 2 |
Zac Medico <zmedico@g.o> wrote: |
| 3 |
> > Not non-preservation. Partial and inconsistent corruption. |
| 4 |
> |
| 5 |
> Wouldn't "loss of precision" be a more accurate description? Of the |
| 6 |
> known packages which require timestamp preservation, do any of them |
| 7 |
> use sub-second precision in their timestamp comparisons? |
| 8 |
|
| 9 |
No. When you're dealing with decimals, floating point rounding isn't a |
| 10 |
simple truncation: |
| 11 |
|
| 12 |
>>> 1234567890.999999999 |
| 13 |
1234567891.0 |
| 14 |
|
| 15 |
nor is it a rounding: |
| 16 |
|
| 17 |
>>> 1234567890.111111111 |
| 18 |
1234567890.1111112 |
| 19 |
|
| 20 |
The former is particularly bad, since POSIX deals with legacy functions |
| 21 |
by ignoring the nanosecond part, not by rounding it. |
| 22 |
|
| 23 |
-- |
| 24 |
Ciaran McCreesh |
Archives