D is for Data Science

[Dmitry Olshansky via Digitalmars-d-announce]

25-Nov-2014 02:43, bearophile пишет:
> Dmitry Olshansky:
>
>> Which is 1:1 parity. Another myth busted? ;)

 > dmitry at Ubu64 ~ $ time ./my2 log
 >
 > real    0m0.065s
 > user    0m0.042s
 > sys    0m0.023s
 > dmitry at Ubu64 ~ $ time ./my2 log
 >
 > real    0m0.063s
 > user    0m0.040s
 > sys    0m0.023s
 >

Read the above more carefully.
OMG. I really need to watch my fingers, and double-check:)

dmitry at Ubu64 ~ $ time ./my log

real	0m0.156s
user	0m0.130s
sys	0m0.026s

dmitry at Ubu64 ~ $ time ./my2 log

real    0m0.063s
user    0m0.040s
sys    0m0.023s

Which is quite bad. Optimizations do help but not much.

>
> There is still an open bug report:
> https://issues.dlang.org/show_bug.cgi?id=11810
>
> Do you want also to benchmark that byLineFast that for me is usually
> significantly faster than the byLine?
>

And it seems like byLineFast is indeed fast.

dmitry at Ubu64 ~ $ time ./my3 log

real	0m0.056s
user	0m0.031s
sys	0m0.025s
dmitry at Ubu64 ~ $ time ./my2 log

real	0m0.065s
user	0m0.041s
sys	0m0.024s


Now once I was destroyed the question is who is going to make a PR of this?

-- 
Dmitry Olshansky