@klimbot
I 100% agree on having a documented timeline for a migration. This timeline needs to be measured in weeks or months, not hours or days.
While it is easy to say “update NOW” the fact of the matter is that our devices are customer-owned and we cannot force them to go online on short notice. We have already scrambled to update everything that we can, but it takes time for customers to act - even if we try to instill the fear of bricks into them. 
Thank you for your suggestions.
We will proceed with the timeline that will be published later.
@cesanta - can you advise the expected behavior after this change? Our older devices could not connect at Apr-8, 16:40 UTC but were suddenly able to do so after about 22:10 UTC.
We saw a similar behavior yesterday.
Devices with old certs have not been able to connect for the last 24~36 hours. It seems the old mDash root has not been active per the previous announcement above.
@cesanta can you advise? We need to get our last customers updated before doomsday.