Travis-CI, mac, clang and permanent failure.

Here's the place for discussion related to coding in FreeCAD, C++ or Python. Design, interfaces and structures.
Forum rules
Be nice to others! Respect the FreeCAD code of conduct!
tom
Posts: 165
Joined: Sun Mar 29, 2015 9:20 pm

Re: Travis-CI, mac, clang and permanent failure.

Post by tom »

PrzemoF wrote: Mon Dec 03, 2018 10:04 pm Edit: I used time stamp - if that fails I'll got for dots or a time stamp once in a while
https://travis-ci.org/FreeCAD/FreeCAD/jobs/463054982
The problem with the timestamp is that it also creates over 1 million lines of log. Thus, there will be no improvement compared to the original log. The trick with the dots is that they are printed on ONE line (using the "tr" command).
Further I suggest to add the "-a" flag to the "tee" command since it avoids that the build_output.txt file of one command gets overwritten by the next command.

However, the latest failure was not related to too long logs or to too silent output. CMake could not find OCC. I'm also using Homebrew as package manager on my Mac, but I could compile FreeCAD today. So, I don't really know, what happens on the Travis machine.
User avatar
PrzemoF
Veteran
Posts: 3520
Joined: Fri Jul 25, 2014 4:52 pm
Contact:

Re: Travis-CI, mac, clang and permanent failure.

Post by PrzemoF »

tom wrote: Tue Dec 04, 2018 1:20 am
PrzemoF wrote: Mon Dec 03, 2018 10:04 pm Edit: I used time stamp - if that fails I'll got for dots or a time stamp once in a while
https://travis-ci.org/FreeCAD/FreeCAD/jobs/463054982
The problem with the timestamp is that it also creates over 1 million lines of log. Thus, there will be no improvement compared to the original log. The trick with the dots is that they are printed on ONE line (using the "tr" command).
Further I suggest to add the "-a" flag to the "tee" command since it avoids that the build_output.txt file of one command gets overwritten by the next command.

However, the latest failure was not related to too long logs or to too silent output. CMake could not find OCC. I'm also using Homebrew as package manager on my Mac, but I could compile FreeCAD today. So, I don't really know, what happens on the Travis machine.
OK, I thought that the total length of the log is the problem, not the number of lines. I'll switch to your solution.

P.S. This build [1] from the main repo was OK, just 20h ago?
[1] https://travis-ci.org/FreeCAD/FreeCAD/builds/462752817
User avatar
PrzemoF
Veteran
Posts: 3520
Joined: Fri Jul 25, 2014 4:52 pm
Contact:

Re: Travis-CI, mac, clang and permanent failure.

Post by PrzemoF »

sgrogan wrote: Mon Dec 03, 2018 11:05 pm
PrzemoF wrote: Mon Dec 03, 2018 2:34 pm My macos knowledge is limited to the fact that it's related to unix somehow.
Me too ;)
I think the root cause is here: https://forum.freecadweb.org/viewtopic.php?f=4&t=31694
Related: https://docs.travis-ci.com/user/reference/osx/
I think Peter might be lurking: https://www.somsubhra.com/github-releas ... ry=freecad
OK, thank you! I'm giving up for now.
User avatar
sgrogan
Veteran
Posts: 6499
Joined: Wed Oct 22, 2014 5:02 pm

Re: Travis-CI, mac, clang and permanent failure.

Post by sgrogan »

PrzemoF wrote: Tue Dec 04, 2018 9:18 am OK, thank you! I'm giving up for now.
Back to your OP. Do you know how to selectively stop the OSX builds? I don't think anything is being learned from the failed build logs, so we are just wasting resources.
"fight the good fight"
User avatar
sgrogan
Veteran
Posts: 6499
Joined: Wed Oct 22, 2014 5:02 pm

Re: Travis-CI, mac, clang and permanent failure.

Post by sgrogan »

PrzemoF wrote: Fri Dec 07, 2018 11:02 pm Clean PR: https://travis-ci.org/FreeCAD/FreeCAD/b ... tification
Thank you.
"fight the good fight"
Post Reply