Jonathan Kingston

This was actually mentioned in my article as follow up tasks: https://jotter.jonathankingston.co.uk/blog/2017/04/04/containers-assignment/ This would potentially break some websites and users expectations around how containers work, however I still think this...

Hey @PedroArvela do you think #518 might solve the same use-case? This would be simpler and allow the user to be in control more.

Hey @stefanpenner and @rwjblue https://github.com/jonathanKingston/broccoli-sri-hash/blob/master/index.js#L70 Some questions: - Does inputPaths and outputPath have preset meaning within broccoli caching writer? - Can we get rid of base yet as it's bad?...

Also @corpulentcoffee I'm a little lost why your vendor file doesn't have an integrity on it too.

@corpulentcoffee are you on @2.0.0 ? paranoiaMode should always be off now that Chrome is behaving. (paranoiaMode is the stupid name I gave the disablement you are talking about)

:+1: thanks for checking that seems the correct behaviour for those versions... most vendors have unicode in them (if not all of them). Yeah I think the base check works...

> I'm not sure why the existing !relativePath[0] === '/' check works at all, since !relativePath[0] would return a boolean, which is then strictly compared to a string...? Yup always...

See also: - https://github.com/jonathanKingston/ember-cli-eslint/issues/37 - https://github.com/jonathanKingston/broccoli-lint-eslint/issues/20

@stefanpenner you already have npm and git access, so as and when it makes sense feel free :D. I still will be around for questions about SRI and clarifications on...

I'm happy to accept patches here but I don't even use the extension myself any more.