« first day (841 days earlier)      last day (934 days later) » 

00:02
Also, we can add a flag to do that (run over range)
 
13 hours later…
12:56
@Deadcode it’s possible the algorithm is really bad, I’ll try to copy someone’s efficient algorithm today. Like emanresu said though, it might just be python anywhere (we’re using the free tier)
Hopefully a fixable issue
13:10
@Steffan #1376 (the rounding issue) isn't actually sympy's fault
it's python's fault
isn't that just round-towards-even
looks like it
>>> sympy.Number.round(sympy.nsimplify(86.5))
86
sympy rounding does it too
same as default python
it's called bankers rounding apparently
i've never actually heard it called that before; interesting
i guess the idea in that case would be that if you have a ton of halves and round them all individually they'll probably add up more or less the same
versus the sum being higher if halves uniformly round up or lower if they uniformly round down
what's interesting is that it's the default rounding mode for IEEE 754 operations for results in binary floating-point formats
 
3 hours later…
16:03
!!/info
 
5 hours later…
20:42
!!/run `abc
@emanresuA abc
!!/status
@emanresuA I am doing well.

« first day (841 days earlier)      last day (934 days later) »