1. 20 10月, 2018 15 次提交
  2. 19 10月, 2018 4 次提交
  3. 18 10月, 2018 6 次提交
  4. 17 10月, 2018 2 次提交
  5. 16 10月, 2018 10 次提交
    • A
      Interrupt http_server.py by ctrl + c (#1001) · a90cf4c2
      Amos Lim 提交于
      a90cf4c2
    • R
      First pass at http benchmark. · c61a0f2f
      Ryan Dahl 提交于
      c61a0f2f
    • R
      Fix throughput_benchmark · 62962e71
      Ryan Dahl 提交于
      62962e71
    • R
      Format · 590feb1c
      Ryan Dahl 提交于
      590feb1c
    • J
      `deno -v` should report typescript version · 1322f3de
      Jinho Bang 提交于
      Fixes #993
      1322f3de
    • A
      Exit cleanly on unrecognized arguments (#990) · fb90c6f5
      Andy Hayden 提交于
      Refactor set_flags to return a Result
      fb90c6f5
    • J
      Fix a binary size regression · 28682a1d
      Jinho Bang 提交于
      This patch changes Jumbo build to use only in debug mode.
      28682a1d
    • K
      Update to TypeScript 3.1 and ts-simple-ast 17 (#980) · 25c97962
      Kitson Kelly 提交于
      25c97962
    • J
      Improve build performance by using jumbo build · dc8841d8
      Jinho Bang 提交于
      The Jumbo build is the Chromium implementation of a Unity build system
      aimed at dramatically lowering the compilation times. It can be easily
      enabled with `use_jumbo_build=true`.
      
      When you enable this, the biggest problem is that name conflicts can
      occur between multiple c++ files merged, which is no problem for `deno`.
      Because the V8's Jumbo build is managed in upstream, and `deno` will
      just pick up a stable build V8.
      
      So, this patch enables Jumbo build by default. After this patch, the
      total number of build objects are halved. (1014 -> 493)
      
      FYI, the following results are tested in my local environment.
       - IMAC 5K, i5 4-cores, 24GB, 512GB SSD
      
      Test result1: No use ccache
      |   |Jumbo build without ccache|Normal build without ccache|
      |---|--------------------------|---------------------------|
      | 1 |         0:05:26          |          0:11:31          |
      | 2 |         0:05:24          |          0:11:29          |
      | 3 |         0:05:25          |          0:11:28          |
      | 4 |         0:05:24          |          0:11:29          |
      | 5 |         0:05:26          |          0:11:29          |
      |AVG|         0:05:25          |          0:11:29          |
      
      Test result2: Use ccache
      |   |Jumbo build with ccache|Normal build with ccache|
      |---|-----------------------|------------------------|
      | 1 |        0:01:38        |        0:01:13         |
      | 2 |        0:01:45        |        0:01:21         |
      | 3 |        0:01:39        |        0:02:20         |
      | 4 |        0:01:45        |        0:02:16         |
      | 5 |        0:01:42        |        0:02:19         |
      | 6 |        0:01:17        |        0:01:11         |
      | 7 |        0:01:01        |        0:01:48         |
      | 8 |        0:00:54        |        0:01:57         |
      | 9 |        0:01:03        |        0:02:15         |
      |10 |        0:01:36        |        0:02:19         |
      |AVG|        0:01:26        |        0:01:54         |
      
      Refs: https://chromium.googlesource.com/chromium/src/+/master/docs/jumbo.md
      dc8841d8
    • A
      Specify deno_dir location with env var DENO_DIR (#970) · 15590a0c
      Amos Lim 提交于
      (Use C:\deno instead of c:\deno in appveyor config because it's cloned to c:\ by clone_folder variable in .appveyor.yml. On the other hand, build directory is pointed to C:\ by $(APPVEYOR_BUILD_FOLDER) so that test targets are placed on separated partitions.)
      15590a0c
  6. 15 10月, 2018 2 次提交
  7. 14 10月, 2018 1 次提交