This just reminded me of a book I read years ago about a near disaster at a Titan missile silo back in 1980. Human error led to a fuel leak and subsequent explosion that blew the blast door off the silo and ejected the warhead.
A google search for the Damascus Titan explosion will find details about it.
Take away everybody else’s social security & Medicaid, but leave ours alone!!!
We invested in the Stihl kombitool system 10+ years ago. I think we have 6 attachments for it now. The old 2 stroke motor finally gave out, and it was a no-brainer to replace it with an electric motor, especially since they were running a special for a free battery when you buy a motor plus a battery & charger.
My wife especially loves it since she always had trouble starting the gas engine…
And likely armored as well to protect against underground rock slides, dragged anchors, etc.
It’s been literally a couple decades now but I once had to troubleshoot multiple RAID failures in a number of identical servers that were all running 6 disk RAID-5. Long story short the power supplies in each server was slowly losing its ability to power all the drives at the same time, so random drives started throwing errors. By the time we figured out the root cause, most of the drives had generated enough errors that the RAID controller couldn’t rebuild the volumes.
So, no, as others have said RAID is not a backup and should never be treated as such. A single point of failure like the power supply can easily cause the loss of the entire volume without warning.
Any army that treats their troops as “cannon fodder” deserves not only all the casualties they rack up, but the long term social, political, and economic hardship that is pretty much a guaranteed result of such a policy.
The constant rounding up & minimal training of “cannon fodder” is expensive both in the short and long term. Better to protect well trained resources and have them continue to gain experience by using more advanced weaponry that minimizes risk to them.
Good to know. Thanks for the insight!
But how does it handle issues like retrying until one and/or the other person is alerted, without erroneously alerting the other at a later time when they get home? And pausing until the next morning and picking up where it left off?
I’m sure you know what xkcd has to say about standards…
Back in the 90’s before this whole internet thing started taking off I was heavily involved with Microsoft’s effort to create a telephony API (TAPI) that was meant to standardize all manner of telephone equipment. The problem is that it has to be overly broad in order to support everything from a dial-up modem to fax machines to the telephone systems used in large corporate offices, and everything in between.
I remember testing a TAPI program I wrote on different types of hardware. I wrote and tested it on a handful of smaller systems that handled a dozen or so phone lines. The first time I tested it on a large enterprise phone system it failed miserably. That enterprise system had a feature that I never anticipated so my code didn’t handle it properly. In a nutshell, if you placed a call on hold then that system assumed you were placing a new call and you immediately got a dial tone. My code assumed when a call was placed on hold that that was all that happened.
I can see similar issues with a broad standard like Matter/Thread. There will likely be devices out there that behave in unanticipated ways, and testing them will be difficult unless you have the physical device. But hopefully, given the backing of all those big companies, they’ll have a good handle on this. It should be able to let end users gracefully handle edge cases, etc.
I took a very cursory look at HomeKit a while ago and found its ability to create complex automations rather limited. For example, our washer & dryer are in our basement, and we can easily forget we have loads of laundry being washed/dryed when we get busy with the rest of our days.
We now have an automation that will text me and/or my wife when a laundry cycle finishes. But it only alerts us if we’re home, and only whoever is home so can go take care of it. If nobody is home when the cycle completes then it waits until one or both of us is home, and then it alerts us. It also won’t alert us overnight but will wait until morning. So if we start a load of laundry at 10pm it doesn’t wake us up at midnight but instead waits until 7am to alert us.
I’ve implemented this in both Home Assistant and Indigo without too much difficulty. Not sure how easy it would be to do in HomeKit though…
That’s one of the more complex automations I’ve created, but I have a few others that are up there as well.
Reddit content fed through Google “AI”. What’s not to trust about that?
AKA greed. Why license your content to Netflix when you can have your own streaming service and lock your viewers into your piddly little hoard of content?
Just how many streaming providers are there today? That number likely changes almost daily at this point…