> A note on point 15:
>
> «15) env.ecfgUnlocked = ecfg
>
> «There should be some validation that the two configs match, per thumper's
> branch. If we forgot it elsewhere, which we might have done, don't consider
> this a blocker.»
>
> The preceding call to maasEnvironProvider.newConfig() calls Validate(), which
> in turn will validate the state transition. So that should do the trick,
> right?
I don't think so... ISTM that we would validate against nil, rather than against the old config.
> A note on point 15: ider.newConfig( ) calls Validate(), which
>
> «15) env.ecfgUnlocked = ecfg
>
> «There should be some validation that the two configs match, per thumper's
> branch. If we forgot it elsewhere, which we might have done, don't consider
> this a blocker.»
>
> The preceding call to maasEnvironProv
> in turn will validate the state transition. So that should do the trick,
> right?
I don't think so... ISTM that we would validate against nil, rather than against the old config.