06.12.2017 9:45, Alex Peshkoff via Firebird-devel wrote:
Sooner of all due to historical reasons - see xdr_double.
According to wiki https://en.wikipedia.org/wiki/Double-precision_floating-point_format
double is stored as two independent 32 bit pieces on some ARM architectures while on most
On 12/06/17 00:04, Dimitry Sibiryakov wrote:
Hello, All.
Can someone explain why double in clumplets is handled as two LONGs,
instead of single INT64?
Sooner of all due to historical reasons - see xdr_double. And compare it
in master and FB1.0.
--
06.12.2017 00:04, Dimitry Sibiryakov wrote:
Can someone explain why double in clumplets is handled as two LONGs,
instead of single INT64?
And, actually, is there any clumplet that have double value?..
There are no such clumplets, AFAIK. These methods are not used currently.
Dmitry
---
Hello, All.
Can someone explain why double in clumplets is handled as two LONGs, instead of single
INT64?
And, actually, is there any clumplet that have double value?..
--
WBR, SD.
--
Check out the vibrant tec
