Hello-
  I've set up self-hosted cameras for myself and several others now, here is what I found:

Don't go wireless, even if it's not jammed, it's not reliable. I can turn on an old 2.4G video sender from the '90s and knock out everything. Thieves do this now.

Running the wires is a pain, but worth it. PoE, power over ethernet, is very important. Trying to do without it is a pain, and devices get unplugged. I've also had times where all the cameras act strangely and need to be restarted, so PoE was fantastic. Also you can use a UPS.

Inexpensive cameras are fine if they have RTSP. They WILL try to 'phone home'! I have them on a subnet where outgoing is blocked. When you look at the traffic it's unreal. Block their IP from outgoing. Mine are isolated from the net, and they work fine. RTSP cameras work with open source stuff. Finding the right strings for the streams can be a pain.

They keep coming out with higher resolution cameras. It's a waste, it only increases the load on the video server. I use the low-res stream anyway, it's fine.

ZoneMinder works and looks nice but it's very CPU intensive. You have to mess with the settings a lot, and they're not easy to figure out. Getting recording to work well is tricky. Save still images. Between the cheap cameras and Zoneminder, the delay can be pretty long. Something that does less processing on the streams would be better, maybe Motion, but I haven't used it much. Despite this, Zoneminder is nice, even supports PTZ on strange thrift store IP cams!

My ZoneMinder box has VPN to my small cloud VPS, and there is URL that proxies to it, behind a password. We can look at it on our phones from anywhere, and we have control and final say, and no monthly bill other than the VPS that we had anyway.

I've used the SV3C outdoor, non-wireless PoE units in the lowest resolution that they still sell. This is not an endorsement, they are constantly trying to send packets to strange places, so they're blocked. But they're inexpensive, (2 pack for $100) weatherproof, support open standards and once they're set up they just work, some of them for many years now. I don't bother with things like 'alerts' because like most cameras, they use email or FTP! (can't we just hit a URL?) Internally, they're running linux, but it's locked down and crippled and not worth the effort to hack. Just use 'em for what they're good for, and block the traffic.

P.S. for some things, a webcam on a raspberry pi will do, and you can get it to work with Zoneminder. It's not actually any cheaper, but you might have the stuff laying around. It's great for stuff like bird's nests and pets. However you will find that such things and cameras with cords tend to get unplugged somehow. The mounted PoE ones just always work.

Hope that helps!
=rMortyH

If you're a US credit card user, look at the receipt that you signed. It doesn't say 'I agree to this purchase", it says "I agree to be bound by the credit card agreement."

This is because the statute of limitations on your credit card agreement runs from the date you made the agreement. (Also your last payment, it's complicated..) The new signature resets the statute of limitations on that contract to start at the date of that signature.

You are usually only bound by that agreement for six years. (depends on the state, but it's usually under the rules of the credit card company's state of choice)

The reason that you sign is to set the statute of limitations on your debt contract to start from the date of that new signature. This makes it easier to collect the debt or sell it. (Did you think it was there to somehow protect YOU?)

=rMortyH

While not responding to a call is pretty bad, plenty of cops were playing it when it came out.

Right when it was most popular my partner and I lived on the upper west side of Manhattan, and she was obsessed with it. Someone posted a hack that found all the Pokemons in your area and put them on a map using Google maps. It wasn't really a cheat because you still had to get there before everyone else. A lot of times when we went to get one in our neighborhood, someone else had gotten it, and there wasn't anyone else around, but there were some cops sitting in a cop car.. A few times we ran into cops with their phones out who just said hello but didn't admit they were playing. Then one time we saw a really good one near us, in Riverside park. We were on Riverside Drive looking down on the spot. Could we get down several flights of stairs and get it before someone else got it? Just then we saw two uniforms running full speed towards the spot. There was hardly anyone else around. They stopped on the spot and started messing with their phones. Too late, they got it!

It's really bad.

It's amazingly, mind-numbingly bad.

It's one of the worst things I've ever seen.

It's so bad, that is just might end up being a classic.

In the old days, it made a lot of sense to have hardware raid so that you could offload the work to a dedicated processor. It was still software raid, just on another piece of hardware.

Nowadays, software raid has almost no impact on the load of the system. It's just WAY simpler to use software raid, and most Linux distros will recognize a software raid volume from another system.

I manage dozens of machines with RAID 1 pairs. They read at the maximum speed that the hardware interface will allow on very busy systems. With two drives that both have grub installed and bootdegraded enabled the system boots up just fine with either drive missing. I can take one of the drives out and the system still runs, and put that drive in another system and mount it as a degraded raid array. I don't have to worry about a failed hardware controller and having to find the same one to put my drives in, it all just works.

Even better, these pairs used to read and write at the speed of a single drive, but in more recent versions of linux, it is smart enough to read from both drive in the pair as if they were RAID 0, doubling read speed, without any special hacks. (Thanks Kernel Team!!)

Hardware RAID is still good in the case where you want the OS to see the whole array as if its one single drive, such as for Windows or other less flexible OSs. But, if it's not linux, it's crap. So I run RAID volumes on my real metal, but anything else is virtualized and it sees its storage as one simple volume, and never has to know that the physical storage is actually redundant.

When a drive dies, which happens fairly often, it shows up on a dashboard and everything keeps right on running. I can put a new drive in when I'm ready, add it to the array, and it syncs right up. I don't trust hot swap so I bring the machine down, but otherwise it's just a quick reboot.

So, hardware raid adds complexity and reduces flexibility without giving me any real advantage.

And of course, RAID is not backup. Never say, "What would I do if I lose that volume?" Always assume you may lose that volume. Drives will die. Do regular syncs to another volume. I sync all my stuff to another volume, then sync that one. Not a mirror, just two different syncs. Two backups. One dies, use the other one. Nothing fancy. I never want to try to recover a lost volume. It's gone. No big deal. Build a new raid and restore. Move on with life.

500 million Spanish speakers can't be wrong.

Amazon's RDS offerings are really convenient, from the perspective of making snapshots and setting up replication. But, I have never been able to push their default 'SSD' storage past 60MB/s. (PostgreSQL and Mysql) That's terrible. That's less than USB 2, and even some SD cards can do that! Our on-prem can do 180MB/s on spinning rust and around 550MB/s on (obsolete) SATA SSD. If you want anything better on RDS you have to REALLY pay a premium for IOPs and transfer, or pay a premium for way more ram and a ton of caching, in addition to external caching in the rest of your stack. I have not used Aurora on RDS, so I don't have a comparison, but I have my suspicions. It would be pretty easy to just give you a few more MB/s and make it look a whole lot better. Luckily in our case we could optimize things enough that storage performance didn't matter too much. But RDS storage performance is so pitiful that it's seriously worth considering putting your DB on a bare metal box somewhere with NVME storage and just put up with the network latency and get 50 times the storage performance (and more ram and cpu while you're at it ) at a fraction of the price.

We've gone from 480Mb/sec up to 6Gb, 10Gb, now 20Gb, and we still can't use a native USB cable to network two linux boxes.
Thunderbolt on Mac has this, and there was a very limited solution for linux that was never production ready.
We're still stuck at 1Gb/sec between machines for any networking that isn't cost-prohibitive, impractical, or both.
The cables exist, I look forward to the day when I can plug two machines together and run a network between them over this cable.
Will this ever happen? Or is this a case of hardware vendors blocking open source so that they don't compete with their own jurassic and overpriced 10Gb products?

Going all the way back to USB 2.0, there is a host-to-host mode in the specification, but it's not implemented in any operating system. The best we can get now is USB3.0 to gigabit dongles, which don't take full advantage of the available speed at all.

USB 3.0 is great for storage, the speed is really nice. Then you find that your bottleneck is gigabit ethernet.
Thunderbolt supports network bridging on the mac, but it's not really useable on linux.
What are the roadblocks to IP over full-speed USB 3.0 or 3.1 data link? Is it pushback from the network vendors?

We already have the cheap, fast hardware, but we can't take full advantage of it.

Will there be any effort to use this for networking? It's really about time.

=rmortyh

Two more words: Arbitration Clause. There will be no more class actions.

Wow. It's a shame no one will ever see this.

Yeah, that's crazy. What were they all doing?
They could have just shut down the reactor and had them all pedal!

Look under Settings/Privacy
There is a switch, which reads 'Send Microsoft info about how I write to help us improve typing and writing in the future'

This the collection of keystroke data. They can do anything they want with this. Definitely makes it even more creepy to log in to someplace else on a Windows 10 box.

Another thing which is standard practice is to list all kinds of serious and unlikely reasons they'll use your data, followed by 'or any other legal purpose' which does not mean for some 'legal' matter, which it's meant to sound like, but for ANY purpose which is not SPECIFICALLY ILLEGAL. Which means anything.

You can turn off the keystroke thing, but Microsoft routinely resets preferences, including privacy preferences, when you run an update. So you have to keep checking it and make sure it's off. However, I doubt very much if it matters. You're sending EVERYTHING to Microsoft and they can use it for any purpose.

http://laughingsquid.com/annoy...
Nailed it! 'nuf said...