Where communities thrive


  • Join over 1.5M+ people
  • Join over 100K+ communities
  • Free without limits
  • Create your own community
People
Repo info
Activity
  • Feb 23 2020 15:28
    Xuanwo commented #1
  • Feb 23 2020 15:18

    Xuanwo on add-docs

    (compare)

  • Feb 23 2020 15:17

    Xuanwo on master

    docs/design: Add proposal for t… (compare)

  • Feb 23 2020 15:17
    Xuanwo closed #7
  • Feb 23 2020 15:17
    Xuanwo commented #7
  • Feb 22 2020 04:45
    codecov[bot] commented #7
  • Feb 22 2020 04:44
    codecov[bot] commented #7
  • Feb 22 2020 04:44
    Prnyself synchronize #7
  • Feb 22 2020 04:44

    Prnyself on add-docs

    docs: Modify the task progress … (compare)

  • Feb 21 2020 11:47
    Xuanwo closed #10
  • Feb 21 2020 11:47
    Xuanwo opened #10
  • Feb 21 2020 04:43

    Xuanwo on master

    Add a Gitter chat badge to READ… (compare)

  • Feb 21 2020 04:43
    Xuanwo closed #9
  • Feb 21 2020 04:43
    Xuanwo synchronize #9
  • Feb 21 2020 04:42
    gitter-badger opened #9
Xuanwo
@Xuanwo
Hello, world!
1 reply
Lance
@Prnyself
Ping
Lance
@Prnyself
How to define the progress?
Firstly, it is easy to treat file-transfer-type task's total size
divided by read bytes as the progress. However, what if other type
tasks? Such as remove a dir or sync a dir?
So, how about treating the whole, including sub-task, giant task as
one, and the total task divided by task-done count as the progress?
In this way, we could unify the stream transfer task, file transfer
task and other time-consuming task. All we need to do is, get the
count of total sub-task, add after each task done, and finish the
task as usual.
Lance
@Prnyself

From my point of view, noah's task should:

allow us pass a callback or something and noah will update it regularly
allow us call Progress() float or something so we can get current state
Maybe we should take them both into consideration?

Lance
@Prnyself
但是以 noah 现有的体系,如何在 task 开始就获取 sub-task 的个数呢?是不是会对现有的架构产生较大的变化?
32 replies
而且 noah 中现有的 new() 方法我看都是空的,是不是可以对现在的 task flow 来一波重构?
Lance
@Prnyself
之前分割 new 与 run 应该是为了处理 static data 和 runtime data 而设置的吧
Xuanwo
@Xuanwo

之前分割 new 与 run 应该是为了处理 static data 和 runtime data 而设置的吧

我都快忘了- -

Xuanwo
@Xuanwo
test
atest
matrixbot
@matrixbot
@xuanwo:matrix.org test
@xuanwo:matrix.org test
@xuanwo:matrix.org atest
matrixbot
@matrixbot
@xuanwo:matrix.org code example
Xuanwo
@Xuanwo
come on
matrixbot
@matrixbot
xuanwo test
Xuanwo
@Xuanwo
去各种平台体验一下之后居然觉得 gitter 还过得去
Lance
@Prnyself
是吧
Xuanwo
@Xuanwo
能直接 github 帐号登陆也很方便
Lance
@Prnyself
感觉这个看起来比 slack 还清爽些
Xuanwo
@Xuanwo
嗯,也不用着那么多功能
Lance
@Prnyself
切入正题?
Xuanwo
@Xuanwo
Xuanwo
@Xuanwo
但是可以粘贴过去,太蠢了
主题无关的在这边说吧
Lance
@Prnyself
哈哈哈
Xuanwo
@Xuanwo
不知道是什么考虑,挺奇怪
Lance
@Prnyself
那我觉得按你发的 md 这个可以搞,把调用 sdk 的部分去做对接
一般也只有这些过程会耗时多一些
另外也更方便我们去 debug
或许就像你说的,用户根本不会在意我们是否显示的是完全的
Xuanwo
@Xuanwo

或许就像你说的,用户根本不会在意我们是否显示的是完全的

实际上也做不到

同步 1 T 文件,先 list 一遍开销也太大了