Alex Couture-Beil

Thanks for the report, can you provide a minimal Earthfile that replicates this.

> I agree that is confusing. > From this example its not clear to me if it is caching /the-cache/f? especially if you look at the output: ``` +test |...

relates to #2268 which requires outputting logs like: ``` ::error file=app.js,line=1::Missing semicolon ``` without any prefix, so github can pickup [annotations](https://docs.github.com/en/actions/using-workflows/workflow-commands-for-github-actions)

That's one long error message :sweat_smile: It looks like this only occurs under the `LOCALLY` mode; where when run under `FROM alpine` it's really fast. Thanks for the report; i'll...

I was able to replicate a similar error using `FROM alpine`, when giving it 500 RUNs: ``` FROM alpine:3.13 toomuch: RUN echo hello >> output.txt RUN echo hello >> output.txt...

It appears that docker has a similar limitation. Consider a Dockerfile with 500 RUNs: ``` FROM alpine:3.13 RUN echo hello >> output.txt RUN echo hello >> output.txt ... ``` `DOCKER_BUILDKIT=1...

This would have to be internally evaluated inside earthly rather than inside buildkit, such a feature might relate to Vlad's comment regarding a `IF (...` type of syntax: https://github.com/earthly/earthly/issues/1458#issuecomment-983054322

perhaps another nice idea would be to support: RUN --optional-secret=MKDOCS_INSIDERS_PAT ... which would prevent the `RUN` from failing if the secret doesn't exist.

note that `IF_IS_SECRET_SET` could be syntactic sugar for `IF --optional-secret=mysecret test -n "$mysecret"` (with the exception of not being able to differentiate between not set vs set to an empty...

Thanks for taking a deep dive @dustyhorizon I think it would be better to tackle it under https://github.com/earthly/earthly/blob/34320d5111d7488bee0ecbc961b4e9d51d289e06/earthfile2llb/converter.go#L1708 and https://github.com/earthly/earthly/blob/34320d5111d7488bee0ecbc961b4e9d51d289e06/earthfile2llb/converter.go#L1901 However, there's some additional issues though that we need to...