Correct, I also read that the .ml addresses were free, but in hindsight this downtime is a bigger cost to bear than the $10/year or whatever it would cost to properly pay for a domain.
In hindsight, yes. But there was no indiciation ahead of time that this situation would happen or was likely to happen. In fact, there was no more reason to believe a free ccTLD was any more likely than a paid ccTLD to cause a problem. The problem arises because a ccTLD's host country can choose to remove any domain it wants, paid or not. One could argue that using a ccTLD at all was a mistake, but you'd have to look at precedent for ccTLD's country's doing this and see if it happens often or not.
Correct, I also read that the .ml addresses were free, but in hindsight this downtime is a bigger cost to bear than the $10/year or whatever it would cost to properly pay for a domain.
In hindsight, yes. But there was no indiciation ahead of time that this situation would happen or was likely to happen. In fact, there was no more reason to believe a free ccTLD was any more likely than a paid ccTLD to cause a problem. The problem arises because a ccTLD's host country can choose to remove any domain it wants, paid or not. One could argue that using a ccTLD at all was a mistake, but you'd have to look at precedent for ccTLD's country's doing this and see if it happens often or not.
Ah, good point.
I 100% agree with that.