By: Tom Sydney Kerckhove <syd@cs-syd.eu>
Rip out the x86, the spare instances and the separate packages MikroTik build no x86 hardware: their chassis are arm, arm64, mips, tile and ppc, and RouterOS x86 is a licence for a PC, which is not a chassis anything here describes. So the x86 image, the x86 branch of the vm test and the architecture field are gone, and with them the trap that the x86 branch used -enable-kvm and so could not have run in the build the arm64 one runs in. `GenLocated l a` was instantiated once and its Functor instance never used. `Show` on Validation and the derivations on the spans were never used either. The rejection regex was a list of the words RouterOS had been observed to complain with, which is a list of the ones it has not been seen to use waiting to be read as success. It matches broadly and without regard to case now, which is safe because success is asserted separately by RouterOS saying it executed the file. That change is also how the test found that `(?i)` is not valid in the middle of a Tcl pattern. No licence is claimed anywhere: there is no LICENSE in this repository, so saying MIT was making one up. Addressing the review: everything is a passthru attribute of declarative-routeros rather than also a package in its own right, and routeros-check is a cabal package built through callCabal2nix, so `--version` comes from Paths_routeros_check rather than a number I invented.
Total
16s
Created Started Finished
| Time to Start | Worker time | Duration | Time to finish | Idle | |
|---|---|---|---|---|---|
| Config | 0s | 1s | 1s | 1s | 0s |
| Eval | 1s | 11s | 11s | 13s | 0s |
| Build | 11s | 4s | 4s | 16s | 0s |
| Suite | 0s | 18s | 16s | 16s | 0s |