lukechilds / base64-async / 72

Builds Branch Commit Type Ran Committer Via Coverage
72 limit-concurrency Limit how many chunks can be worked in each setImmediate loop Otherwise it can still block via lots of concurrent jobs. e.g 100 concurrent jobs with a chunk size of 250KB each means that 100 250KB chunks will be worked in each setImmediate loop. ... push lukechilds travis-ci pending completion   set done
70 master Don't test on latest and 7 push lukechilds travis-ci pending completion   set done
69 master Run benchmarks on Travis push lukechilds travis-ci pending completion   set done
68 master Add "realworld" benchmark push lukechilds travis-ci pending completion   set done
65 decode-in-js Allow var for performance reasons push lukechilds travis-ci pending completion   set done
63 master Remove eslint-disable for no-return-assign push lukechilds travis-ci pending completion   set done
62 master Title benchmark table with chunk size push lukechilds travis-ci pending completion   set done
61 master Increase default chunk size to 250KB push lukechilds travis-ci pending completion   set done
60 master Save decoded buffer chunks in array and concat them at the end Gives a good performance boost. push lukechilds travis-ci pending completion   set done
59 master encodeChunk => decodeChunk push lukechilds travis-ci pending completion   set done
58 master Use var over let for performance sensitive variables Variables incremented in the loop are much slower to update if declared with let. This should be resolved in V8 5.9. push lukechilds travis-ci pending completion   set done
56 master encodeChunk => decodeChunk push lukechilds travis-ci pending completion   set done
55 master Benchmark up to 100 MB push lukechilds travis-ci pending completion   set done
54 master Use non utf8 buffer to help catch decoding errors e.g When using a utf8 string buffer we could store the buffer as a utf8 string and tests would pass. If you then tried to encode/decode an image you'd get incorrect results. push lukechilds travis-ci pending completion   set done
53 master Use time-span for benchmarking, don't roll our own We were incorrectly interpreting process.hrtime anyway push lukechilds travis-ci pending completion   set done