Fork me on GitHub
#babashka-sci-dev
<
2022-06-20
>
borkdude18:06:04

@rahul080327 > the runners take a bit of time to provision what's that?

lispyclouds18:06:27

So we are using the same infra but shorting the steps, so it could take a while for example the mac one to be provisioned. Then it shorts. So whole run could last upto few minutes

lispyclouds18:06:49

This is to keep the structure same as the required checks in GitHub

lispyclouds18:06:22

We can think of a way to have smaller runner configs for shorted pipelines in a follow up pr

borkdude18:06:35

yeah, e.g. we could provision the cheapest runner possible

lispyclouds18:06:59

Yep. It's all pulled into its own fn, should be easy to implement

borkdude18:06:19

let's wait and see if there are no regressions

borkdude18:06:50

@rahul080327 docker step is failing now

lispyclouds18:06:48

Right, will take a look as soon as I'm back home

lispyclouds18:06:05

This wasn't running in the Pr 😕

borkdude18:06:25

not trying to release now, luckily ;)

lispyclouds19:06:47

well not much use without changing the docker script too when not on master, but should fix it. the name of the artifact was wrong, checked it now on the PR that its fixed

lispyclouds19:06:18

its interesting that the release and setup musl scripts have different default values for BABASHKA_ARCH env var

lispyclouds19:06:47

things i learn about the codebase everyday

lispyclouds19:06:22

this can be merged if youre okay with it, otherwise i can modify the docker script to run on non master to test too

borkdude20:06:08

@rahul080327 can you check on dockerhub if everything looks ok over there too ?

lispyclouds20:06:54

well nothing was pushed there, the last run failed before pushing anything, this one was skipped

lispyclouds20:06:43

i have to change the docker script a bit to make it do things on a branch

lispyclouds20:06:25

issue was that the artifacts were named _x86_64_ and the docker script wants _amd64_

borkdude20:06:28

ok, we can just merge to master and see if it works there

1
borkdude20:06:41

I wonder why this was an issue since the dynamic config should be equal to the previous config when there are things to build, right?

lispyclouds20:06:54

the last commit has a [skip ci] 😕

lispyclouds20:06:07

> I wonder why this was an issue i missed the fact that babashka_arch env var is only needed for aarch64, others use the default values and the default values are different for the release and musl scripts

👍 1
lispyclouds20:06:20

manually triggering pipeline from circle

lispyclouds20:06:51

hub looks okay

🎉 1
lispyclouds20:06:33

killing yaml is probably as satisfying as killing bash with bb :satisfied:

🪓 1
🍠 1