|
Ran
|
Jobs
5
|
Files
7
|
Run time
1min
|
Badge
README BADGES
|
push
travis-ci
investigated and fixed hanging issue on AppVeyor Also added extensive debugging facilities to metadata.py which can be used in the future. For details on what happened, check the individual commit messages below. Squashed commit of the following: commit 8a2e460b3 Author: Zhiming Wang <zmwangx@gmail.com> Date: Fri May 8 15:47:25 2015 -0700 do not test vp9.webm on AppVeyor There seems to be a bug in the ffprobe installed by AppVeyor (using cinst) that always run into an infinite loop when processing vp9.webm with -show_frames. See https://ci.appveyor.com/project/zmwangx/storyboard/build/1.0.50, and especially the log file https://ci.appveyor.com/api/buildjobs/n7e9dakb5ivsaxee/log for the presence of the infinite loop. commit 3519e048f Author: Zhiming Wang <zmwangx@gmail.com> Date: Fri May 8 14:24:59 2015 -0700 focus on metadata on vp9.webm From https://ci.appveyor.com/project/zmwangx/storyboard/build/1.0.48 and https://ci.appveyor.com/project/zmwangx/storyboard/build/1.0.49 it's pretty clear that metadata on vp9.webm is causing the issue. Focus on it by deleting other samples and dropping print statements everywhere in metadata.py. The debug printings in metadata.py can be kept after debugging. commit 65ce7b750 Author: Zhiming Wang <zmwangx@gmail.com> Date: Fri May 8 13:59:41 2015 -0700 adding back vp9.webm to see what happens Somehow after removing vp9 the build was a success, at least for Python 2.7 and 3.2 (didn't finish the build). Also adding ffmpeg -formats and ffmpeg -codecs to see if there's anything wrong with FFmpeg. commit dd8e82eeb Author: Zhiming Wang <zmwangx@gmail.com> Date: Fri May 8 13:30:28 2015 -0700 removing vp9 sample to see what happens According to https://ci.appveyor.com/project/zmwangx/storyboard/build/1.0.47/job/5xhe8x5m2ribeqac the build seems to hang on the last sample, vp9.webm. Also upped the game by being even more verbose. commit 050178ba8 Author: Zhiming Wang <zmwangx@gmail.com> Date: Fri May 8 13:03:38 2015 -0700 try to find out why appveyor builds hang... using the time-tested method of dropping print statements everywhere. Recently (since yesterday) the AppVeyor builds always hang for no obvious reason at more_test_metadata (tests.more.more_test_metadata.MoreTestMetadata); see https://ci.appveyor.com/project/zmwangx/storyboard/build/1.0.45 for example. I've run the tests on my own Windows 8.1 VM and can't see anything suspicious.
965 of 1065 relevant lines covered (90.61%)
4.51 hits per line
| ID | Job ID | Ran | Files | Coverage | |
|---|---|---|---|---|---|
| 1 | 71.1 | 0 |
90.61 |
Travis Job 71.1 | |
| 2 | 71.2 | 0 |
90.33 |
Travis Job 71.2 | |
| 3 | 71.3 | 0 |
90.14 |
Travis Job 71.3 | |
| 4 | 71.4 | 0 |
90.14 |
Travis Job 71.4 | |
| 5 | 71.5 | 0 |
90.14 |
Travis Job 71.5 |