zeroZshadow
8abf09e1ce
Allocate the maximum in 1 go for marshaling
2016-06-11 04:20:04 +02:00
zeroZshadow
bb0fb2c0f5
Fix UnmarshalJSON benchmark
...
Optimize MarshalJSON a little more
2016-06-11 03:48:54 +02:00
Bruce Marriner
c676f9e9ea
Correct node error. It's 0-1023!
2016-06-10 18:54:48 -05:00
Bruce Marriner
00c61c48d0
Use nodeShift instead of 12.
2016-06-10 18:51:24 -05:00
Bruce Marriner
94a47d42a7
Don't defer.
...
This dropped (on my system) the Generate benchmark from about
285-330ns/op down to 244ns/op. Yay!
2016-06-10 18:20:30 -05:00
zeroZshadow
7c362b8bf3
oops, left some experiment stuff in
2016-06-10 20:35:59 +02:00
zeroZshadow
d8933e1ea5
Added tests for MarshalJSON and UnmarshalJSON
...
Reduced allocation of MarshalJSON from 65 to 40 B/op
2016-06-10 20:33:36 +02:00
Bruce Marriner
348dc1e55f
Lint.
2016-06-04 15:52:39 -05:00
Bruce Marriner
db0acce1d0
Linting, Optimizations, Cleanup, More Benchmarks.
2016-06-04 15:50:03 -05:00
Bruce Marriner
0fdfe21511
Removed http.Handler interface
...
Including this added 60KB to the compiled library which I do not think
is worth it just to include a helper function consisting of about 4
lines of code.
2016-06-01 16:58:13 -05:00
Bruce Marriner
bdcd2f75e6
Added http.Handler interface
2016-06-01 16:49:10 -05:00
Bruce Marriner
bf4c51fc2b
Rename to snowflake
2016-06-01 16:12:53 -05:00
