Rate-Limit: Thread fix, reset date #160
Conversation
|
Kohsuke Kawaguchi » github-api #284 SUCCESS |
| @@ -18,7 +18,7 @@ | |||
| @Before | |||
| public void setUp() throws Exception { | |||
| Properties props = new Properties(); | |||
| java.io.File f = new java.io.File(System.getProperty("user.home"), ".github.kohsuke2"); | |||
| java.io.File f = new java.io.File(System.getProperty("user.home"), ".github"); | |||
There was a problem hiding this comment.
GitHub.connect() already looks for ~/.github, so I need to keep the original line.
There was a problem hiding this comment.
maybe better cal it .github.test then?
There was a problem hiding this comment.
In general not critical in comparison to thread usage.
|
This issue is blocker for me :( |
|
Added two related to rate limit info commits, can move in separate PR if you want. |
|
Kohsuke Kawaguchi » github-api #287 SUCCESS |
|
Kohsuke Kawaguchi » github-api #288 SUCCESS |
|
Kohsuke Kawaguchi » github-api #297 SUCCESS |
|
Removed one unrelated commit. Also if some trigger plugin uses this library it can stuck with the next execution time: |
|
Another one bug is that if you have bad password + rate limit 0, then rate_limit page http answer is treated as NotFound and sets rate limit to 10000 (it think that its enterprise). |
Strategy pattern is better for API rate limit handling as the current behavior is quite valid for batch applications. I'm not taking OkHttp version change so as not to introduce any unneeded version requirement.
|
I implemented a variant of this in 4093e53 and merged. |
|
Could you also add "bad password" throw before looping? Sent from my iPad
|
|
caused regression #198 |
for #159