Ian Campbell

> Fixing this directly would involve making the hash algorithm somehow configurable. #2586 requires something similar. > An alternative might be to rebase onto the existing linuxkit/alpine base So this...

> The console scrolls by too fast to capture anything usable. I suppose I could take a video but that seems overkill. Is serial console (including iO, AMT, IPMI SOL...

I think I must misunderstand how firmware loading works, perhaps some one can enlighten me. My understanding was that firmware loading was an asynchronous process (triggered by drivers while they...

> here's a few things you're missing: Thanks. Please can you arrange for the core bits of your second bullet to make it into the commit message... I'm afraid I...

> In fact with the most recent master it seems the WARNING has gone, even though llb-gobuild is still used -- the usage is in a example command (so a...

Having noticed the `// +build ignore` I think this is now only relevant as it relates to #63.

Ah yes, I forgot that the cgo stuff would be opaque to the Go AST.

Looking at [`git-check-ref-format(1)`](https://git-scm.com/docs/git-check-ref-format) I see (just quoting the relevant ones) > 3. They cannot have two consecutive dots .. anywhere. > > 4. They cannot have ASCII control characters (i.e....

Hard> Yes :-/ A separate tool seems like a reasonable solution. It might be useful to that goal to have vndr not strip the vendor.conf files out of things which...

I don't know how one would translate this into a Windows equivalent but LinuxKit calls the credhelper itself to extract the values: https://github.com/linuxkit/linuxkit/blob/master/scripts/push-manifest.sh#L52...L54 If you could replicate that on Windows...