8f104396ec
All successful, non-garbage-collected builds in the evaluation are passed in a attribute set. So if you declare a Hydra input named ‘foo’ of type ‘eval’, you get a set with members ‘foo.<jobname>’. For instance, if you passed a Nixpkgs eval as an input named ‘nixpkgs’, then you could get the Firefox build for x86_64-linux as ‘nixpkgs.firefox.x86_64-linux’. Inputs of type ‘eval’ can be specified in three ways: * As the number of the evaluation. * As a jobset identifier (‘<project>:<jobset>’), which will yield the latest finished evaluation of that jobset. Note that there is no guarantee that any job in that evaluation has succeeded, so it might not be very useful. * As a job identifier (‘<project>:<jobset>:<job>’), which will yield the latest finished evaluation of that jobset in which <job> succeeded. In conjunction with aggregate jobs, this allows you to make sure that the evaluation contains the desired builds. |
||
|---|---|---|
| .. | ||
| Hydra | ||
| Hydra.pm | ||
| Makefile.am | ||