Is this significantly different in use / operation to --rebuild?
What is the use case for --force-reload over --rebuild?

@attie thanks for your comments. In the mainstream case there is no difference. I'm using this where I supply the bitstream, built outside of glasgow, where the id is not valid and I want it to be reloaded. I guess I could overload --rebuild to reload in the cases where an image was supplied and avoid the extra option, would that be appropriate.

I'm using this where I supply the bitstream, built outside of glasgow, where the id is not valid and I want it to be reloaded.

Ah, I see.

I presume you're using this in conjunction with --prebuilt-at? If so, perhaps something else should be used in place of the bitstream's ID for external bitstreams (e.g: a hash of the binary file)? This way we could remove the extra flag, and it would always reload when required...?

Alternatively a bitstream ID of all ones could be used to indicate "always reload".

Alternatively a bitstream ID of all ones could be used to indicate "always reload".

I'll update the PR to do this, great idea.

Done. Please squash when merging to hide the mess.

@TomKeddie please could you rebase/squash and force-push when you're happy?

(It looks like software/glasgow/cli.py still need to be cleaned up)

Thanks @TomKeddie.

Not sure if there's a preference either way for bytearray('\xff' * 16, 'latin-1') vs b'\xff' * 16 - I'd pick the latter, but take @whitequark's lead.

LGTM.

Thanks!