The scariest part is not knowing what is happening while everything is down. Maybe I am in the minority, but I am not well connected with many of the higher ups in the know. So I have to assume the worst and cross my fingers that everybody's accounts are safe.
I am glad it is back, and the glitches are being ironed out.
Hopefully fixing the fork database will prevent this kind of downage again. It is concerning that witnesses testing the new fork can cause the main blockchain to hit a stop error.
Having some communication representatives available from official sources would also be appreciated in the future. Rumors start flaring up like wildfire because most of us have no direct communication lines, and that is bad for value of Steem on the exchanges. I heard there was a twitter post, and that was it. Maybe this can be improved too. I'm sure some of the witnesses would be happy to be trusted with spreading official news through popular discord channels, and on their app sites.
You can follow @SteemNetwork for live updates. We understand how much you care about Steem and the Steem applications you rely on and are continuously working to develop better communication systems so that you can remain up to date, while still allowing the engineers who have to work together to diagnose the problem and develop a fix to do their jobs.
Yesterday during the 12hour / 12second issue, this was actually not true. I found it sad that all of the information about the hardfork that I would have considered relevant and helpful was not accessible through normal Steem front-ends. I am not saying its bad to eat your own dogfood, but if there's suddenly no dog-food to eat then its a hungry pack of wolves... or something.
Perhaps it makes sense to consider IPFS'ing important announcements about things like hardforks as a type of insurance against this kind of thing happening in the future.
Anyway, thanks to all the awesome witnesses who got us back online, their devs and the whole steem community. Great to be here.
I definitely agree with this, but I will point out that myself and many others were indeed spreading the word through a number of the servers where we're active. We do our best to make sure we can explain as clearly as we can wherever we can, while juggling the background stuff, but you're absolutely right that we can't reach every community or even most people as effectively as official channels.
Having compounding errors is one of those things that testing might miss, no matter how much is done — in this case, almost a month of running testnet, plus two simulated hardforks, and a week or so of quite a few witnesses running the new code on mainnet all managed to sneak by before the perfect storm of nodes and transactions hit to kick this all of. It definitely sucks, and goes to show we can basically never sleep... but I'm bolstered working on a chain designed to stop before it starts double spending or compromising wallets. If there's a community that you're in that you feel needs a liaison, please reach out to me on steem.chat or discord and I'm happy to add it to the places I roll through when I'm updating.