Tax revenues are always a percentage of GDP. Because of that, GDP is absolutely relevant.
In a nation that is not tied to a monitory standard (like gold), the government controls the inflation rate. Thus the rate of inflation and the rate of GDP growth are not the same.
Reducing packet loss is not always a good thing. Packet loss is mechanism that an IP network uses to indicate a lack of capacity somewhere in the system. BufferBloat is one attempt to eliminate packet loss with very bad consequences, never throw packets away by queueing them for a very long time. Pacing can be the opposite side of that coin, send packets so slowly loss never occurs, but that also means the transfer happens very slowly.
When many TCP connections are multiplexed onto a single link the maximum aggregate throughput is achieved with 3-5% packet loss. Less and there are idle points on the line, more and there is self synchronizing congestion collapse.
What's really amusing here is the notion that pacing is the fix for BufferBloat. That's sort of the two wrongs make a right theory; break per link queueing with BufferBloat and then break senders by making them all painfully slow.
Please try again. This time plot revenue and spending as a percentage of GDP. I'll save you some time, go here to see it.
You are correct that spending is up, even as a percentage of GDP. The budget should be reviewed, as some of the causes are cyclical (the recession) and will "self solve" as the economy improves, while others are structural issues, like devoting an ever larger chunk of the budget to military and war expenditures over the past decade.
But it's just as important to realize that as a percentage of GDP revenue is down. Those tax cuts mean the government is taking in a smaller percentage of economic output. So when inflation drives up the cost of guns/tanks/healthcare/office space/contractors for the government there isn't a corresponding increase in revenue to off set it, because we've chosen to end taxes on a number of things that get inflated (like the wealthiest 1%'s salaries).
Your bottom line is wrong. Revenue is up in dollar amount, but down as a percentage of the economy. Spending is up by both measures. Revenue has not kept pace with economic growth. To solve the debt and deficits we must both lower spending and raise tax revenue, ideally by closing loopholes and credits, rather than raising the marginal rates.
The OS (at least, Linux and FreeBSD) enumerate the motherboard ports in order. So if your motherboard has 8 ports you can discover the order and number them 1-8.
If you plug in a device to port 1, and then later plug the next device into port 2, there is no chance the first device will be renumbered.
However, if you plug into port 2, and later add a device on port 1, it will get renumbered.
So when I get a new motherboard, I take a pile of old 16M (yes, Meg) flash drives I have, and name them A-G or whatever, then plug them into every port and reboot the box. By checking which one is on which device, I can label the USB ports 1-N. All that has to be done then is always plug the new device into the next lowest port and there is no renumbering ever.
Yes, if one comes unplugged AND you reboot, the ones past it will renumber. However, plug it back in and reboot and it all numbers back just fine, no config work necessary.
I've deployed a few hundred of these, and maybe once every 2 years or so have a minor issue where a colo tech moves cables wrong. It's never taken more than 5 minutes and a reboot to fix though. I can also deploy 96 ports of serial for less than $2k, I don't know any other commercial solution that can come close to that price point.
I've not seen a hardware/software combination in the past ~5 years that varied the initialization order. Now, the order in which the USB ports are initialized is often non-intuitive and non-documented, but just plug a flash drive into each available port and reboot the server to find the order they are initialized in.
Where I see most people go wrong is they don't figure out the order the ports are initialized, and thus don't plug the serial device into the first to be initialized. By plugging into some other port, then adding a device and rebooting they get reordered.
Does it take some care? Yes. However it's also an order of magnitude cheaper than most of the packaged solutions, and leaves you with standard unix TTY's which open up a world of scripting opportunities.
There's a little known, but very useful program called rtty. You can find it at ftp://ftp.isc.org/isc/rtty/rtty-4.0.shar.gz. Yes, it was last updated in 2003. Yes, there are package for major open source distributions.
Here's serial consoles on the cheap:
Buy multiport USB to Serial devices. They are a USB hub with a bunch of USB to Serial adapters hung off of them. Here's a 16-porter for an example: http://www.startech.com/Cards-Adapters/Serial-Cards-Adapters/~ICUSB23216F
Hang them off a low end box, I like half-depth Intel Atom servers with lots of USB ports.
Run rtty. It records each console to a log file 24x7, and allows multiple people to connect at the same time (including typing).
Tire Rack to the Rescue: http://www.tirerack.com/tires/tiretech/techpage.jsp?techid=35
"V" speed rated tires, rated to 149 MPH, are common on just about any "performance" car, like say a Mustang.
"W" (168 MPH) and "Y" (186 MPH) are in fact relatively common, and stocked in most performance sizes at your local warehouse.
The "Z" rating simply means "somewhere above that", and "go read the manual".
For instance the stock tires on a Z06 Corvette are: http://www.tirerack.com/tires/tires.jsp?tireMake=Goodyear&tireModel=Eagle+F1+Supercar+G%3A+2+RunOnFlat&leftTire=735YR8F1G2LROF&rightTire=735YR8F1G2RROF&vehicleSearch=false&fromCompare1=yes
$385 each, rated to 186+.
Now, if you want to push the boundaries of DOT legal, here are some auto-cross tires. They are in fact DOT legal, but not recommended for use on the street: http://www.tirerack.com/tires/tires.jsp?tireMake=Toyo&tireModel=Proxes+RA1&partnum=735ZR8RA1HC4/32&vehicleSearch=false&fromCompare1=yes
$265 each, Z rated. Once heat cycled they should easily be good into the low 200MPH range.
The insane Veyron tires become necessary somewhere north of 225MPH or so. Like with many things the forces get exponentially stronger with speed, and thus the cost gets exponentially higher.
This is someone who was ok with ASN.1, OID's, and "walking" tables that had no business in being walked, over an unreliable UDP protocol that initially had effectively zero security.
Someone stop him from developing a home automation protocol before his being "first" relegates that industry to 30 years of pain and suffering.
Absolutely. Here's another way to look at it. At $200/mil a year in spend, an (to make numbers easy) an average spend of $100k per employee (which for a salary+benefits+sg&a number is probably low), that's 2,000 people working on the web site. Facebook has 3,000 employees. And thats if it was all employee costs, no equipment.
Remember they have to interface with 50 different state systems, plus other government agencies. It adds up quick, 5 people per state to figure out the state requirements and interfaces is 250 people right off the top, for instance.
I'm afraid there are a lot of./ users who have build a web site that sees 10k users a day, which is not that hard on commodity hardware, but have no clue what it takes to handle 10 million while staying inside the lines of a bevy of government requirements.
Facebook had a COGS cost of 1.36 billion, and a R&D cost of 1.4 billion. Since they basically only operate a web site, that's all the cost of operating a web site. So to 2.76 billion dollars in a single year spent on Facebook.com.
Healthcare.gov spent 614 million over three years. At $200 million a year, that's roughly in line with Facebook's spending level back in 2009.
And Facebook has never gone down, right? It's never had a load issue, right? Yeah, didn't think so.
I was under the impression that people who signed up before March 31, a one time delayed deadline, did not have to pay the 2014 fine. The deadline for subsequent years would be December 15th.
And the first year fine, $95 or 1% of income, whichever is larger.
I've seen several people say "the cloud is the answer". I have one simple counter example:
Reddit. Any time they have a flash mob, and seemingly randomly almost every day they fail.
The clould does some things better and some things worse, and scales in different ways than a more traditional layout. Both can work if properly implemented, and one or the other may be faster/cheaper/better depending on specific site requirements.
Turns out millions of users who start using something on the same day often don't follow the expected and tested for behavior.
Anyone who launches a service like this should expect to spend the first week in triage mode, and the first month making adjustments. I'd like to say proper planning would mean that never occurs, but the only way to insure that would be to spend 10x what is really needed. People would hate the government even worse if they did that.
This is not news, yet. It will be news in a month if it is still fubared.
You do realize apple controls how many are made, where they are released, and knows how many are sold the first month for each release?
No doubt there is a bean counter who figures out that they will do 20M 5s/5c the first month, so next year if they add 2 more countries they can do 11M day one, "breaking the record" on what will be 22M sales with growth.
These things aren't left up to chance. You'll know something is wrong when iPhone next is not 120% of iPhone previous.
Note that even if the consumer buys at the subsidized price Apple still receives revenue closer to the unlocked price, as the carriers pay the difference to attract the customer.
iPhone 5s "T-Moble Contract Free" prices, are $649, $749, and $849, depending on the amount of storage. See iPhone 5s. The iPhone 5C prices are $549, $649. See iPhone 5c.
Using an average price of $649, and 9 million units sold, that's $5.84 billion in revenue. That doesn't could any accessories (cases, car chargers, etc) or Apple Care sales.
GTA V made a relatively puny $1 billion. You know, chump change.
Correct. This is sort of like doing a review of an ordinary home door lock, smashing it with a police battering ram, and then declaring it useless.
The TouchID sensor is there to make things convenient enough that those who do not use a pass code now will use one. That's actually a huge leap forward, since it means a casual thief can't quickly get into your e-mail or contacts before ditching the phone. All it really has to do is slow someone down enough that you have time to get to a computer, invoke find my iphone, and remote lock it or even remote wipe it so people don't get your data.
You are correct that the disk will become fragmented, but the implication is fragmentation is a problem and that's simply not true. One of the prime causes of the misunderstanding is that fragmentation in Unix file systems is night and day different than fragmentation in a FAT file system, where most people are used to defragging windows drives. Unix file systems use much better algorithms to control fragmentation, so there is (generally) a lot less on a per file basis. They also automatically defragment, there are cases where when a fragmented file is written to the file system will defragment part of that file and rewrite it.
The Berkeley FFS was the first to "solve" this problem, reserving 10% of the disk space primarily to avoid fragmentation. Decades of experience show that for all but the most corner of corner cases, that is enough, causing no significant amount of fragmentation, or performance degradation.
The result is that for most applications fragmentation is a complete non-issue. After 25 years of playing with various file systems I've only seen it be an issue once, on an NNTP server that reached 20% fragmentation. Most user desktops and general purpose servers have under 1% fragmentation at all times. Generally, if you have a fragmentation problem it's because the storage is too full, and you need to add storage anyway (the aforementioned NNTP server was a good example). Adding the storage makes the problem go away.
Most users of Unix file systems will never need to give fragmentation a second thought.
The Civil Air Patrol is actually under the direction of the US Air Force, they are civilians in the sense that they are not members of the military, but they are military in the sense that they routinely train with the air force and national guard specifically to do SAR missions. The CAP is not "bubba jumping in an airplane to help", it's civilians who have invested hundreds of hours of training and coordination to be able to help in these situations.
http://www.gocivilairpatrol.com/about/
The national guard would be just fine with CAP planes as they have trained with them, and know that the CAP know all of the SAR procedures and their standard coordination to avoid mid-air collisions.
Your question is legitimate, and in point of fact we don't know yet.
I suspect so far what's happened is FEMA found out there was a flying object (the drone) in an area they had helicopters and planes in, so they shut it down. They don't know if it is safe, useful, or practical at that point, and their tired and true methods, and more importantly humans in those other machines take priority.
They may well talk to them, and find the tech useful, and be able to find windows for them to fly when other operations aren't in the area. Heck, they might be back in the air tomorrow. However badmouthing FEMA or the FAA to the press makes this a less likely outcome.
"Altitude: From the surface up to and including 13000 feet MSL"
I actually got the wrong NOTAM, which is why the date is wrong. The right one is http://tfr.faa.gov/save_pages/detail_3_4333.html and was issued back on September 14th. It says "Altitude: From the surface up to and including 11500 feet MSL"
I wasn't aware any counties had delegated authority from the FAA, I've only ever seen that given to other federal agencies (like FEMA), or sometimes state level agencies. Did Boulder County actually have the ability to approve such a thing in the first place?
I don't think that's what the article says, but you might have to know some things about FAA regs for them to make sense.
"It has public safety flight approvals from the Federal Aviation Administration (FAA) to fly in some parts of Colorado."
They most likely have a conditional certificate to fly only in particular areas of low traffic/population for an experimental drone. That's similar to having a pilots license (approval to fly a plane), or even a drivers license (approval to drive a car).
NOTAM's, or NOTice to AirMen are temporary restrictions on ALL flight operators. Think of them as a construction detour in your car, or a bridge washed out barricade. A common NOTAM might be that a runway is closed for resurfacing, or that a chuck of airspace is blocked off for an air show.
So while they may have general approval to fly, the NOTAM cancels that for the specific area covered. Most likely the FAA has delegated to FEMA the ability to control all flights in this box as they coordinate SAR, Search and Rescue operations.
So to extend my car analogy, it's like there's a washed out bridge from a flood, and they put up a barricade across the road while they tried to recuse someone from the flood waters and these people simply drove around the barricade and said "we're here to help!". The answer was get back on the other side of the barricade, or be arrested.
Slashdot Mirror
Tax revenues are always a percentage of GDP. Because of that, GDP is absolutely relevant.
In a nation that is not tied to a monitory standard (like gold), the government controls the inflation rate. Thus the rate of inflation and the rate of GDP growth are not the same.
Reducing packet loss is not always a good thing. Packet loss is mechanism that an IP network uses to indicate a lack of capacity somewhere in the system. BufferBloat is one attempt to eliminate packet loss with very bad consequences, never throw packets away by queueing them for a very long time. Pacing can be the opposite side of that coin, send packets so slowly loss never occurs, but that also means the transfer happens very slowly.
When many TCP connections are multiplexed onto a single link the maximum aggregate throughput is achieved with 3-5% packet loss. Less and there are idle points on the line, more and there is self synchronizing congestion collapse.
What's really amusing here is the notion that pacing is the fix for BufferBloat. That's sort of the two wrongs make a right theory; break per link queueing with BufferBloat and then break senders by making them all painfully slow.
This is not the answer.
Please try again. This time plot revenue and spending as a percentage of GDP. I'll save you some time, go here to see it.
You are correct that spending is up, even as a percentage of GDP. The budget should be reviewed, as some of the causes are cyclical (the recession) and will "self solve" as the economy improves, while others are structural issues, like devoting an ever larger chunk of the budget to military and war expenditures over the past decade.
But it's just as important to realize that as a percentage of GDP revenue is down. Those tax cuts mean the government is taking in a smaller percentage of economic output. So when inflation drives up the cost of guns/tanks/healthcare/office space/contractors for the government there isn't a corresponding increase in revenue to off set it, because we've chosen to end taxes on a number of things that get inflated (like the wealthiest 1%'s salaries).
Your bottom line is wrong. Revenue is up in dollar amount, but down as a percentage of the economy. Spending is up by both measures. Revenue has not kept pace with economic growth. To solve the debt and deficits we must both lower spending and raise tax revenue, ideally by closing loopholes and credits, rather than raising the marginal rates.
I think you misunderstood what I said.
The OS (at least, Linux and FreeBSD) enumerate the motherboard ports in order. So if your motherboard has 8 ports you can discover the order and number them 1-8.
If you plug in a device to port 1, and then later plug the next device into port 2, there is no chance the first device will be renumbered.
However, if you plug into port 2, and later add a device on port 1, it will get renumbered.
So when I get a new motherboard, I take a pile of old 16M (yes, Meg) flash drives I have, and name them A-G or whatever, then plug them into every port and reboot the box. By checking which one is on which device, I can label the USB ports 1-N. All that has to be done then is always plug the new device into the next lowest port and there is no renumbering ever.
Yes, if one comes unplugged AND you reboot, the ones past it will renumber. However, plug it back in and reboot and it all numbers back just fine, no config work necessary.
I've deployed a few hundred of these, and maybe once every 2 years or so have a minor issue where a colo tech moves cables wrong. It's never taken more than 5 minutes and a reboot to fix though. I can also deploy 96 ports of serial for less than $2k, I don't know any other commercial solution that can come close to that price point.
I've not seen a hardware/software combination in the past ~5 years that varied the initialization order. Now, the order in which the USB ports are initialized is often non-intuitive and non-documented, but just plug a flash drive into each available port and reboot the server to find the order they are initialized in.
Where I see most people go wrong is they don't figure out the order the ports are initialized, and thus don't plug the serial device into the first to be initialized. By plugging into some other port, then adding a device and rebooting they get reordered.
Does it take some care? Yes. However it's also an order of magnitude cheaper than most of the packaged solutions, and leaves you with standard unix TTY's which open up a world of scripting opportunities.
There's a little known, but very useful program called rtty. You can find it at ftp://ftp.isc.org/isc/rtty/rtty-4.0.shar.gz. Yes, it was last updated in 2003. Yes, there are package for major open source distributions.
Here's serial consoles on the cheap:
Buy multiport USB to Serial devices. They are a USB hub with a bunch of USB to Serial adapters hung off of them. Here's a 16-porter for an example: http://www.startech.com/Cards-Adapters/Serial-Cards-Adapters/~ICUSB23216F
Hang them off a low end box, I like half-depth Intel Atom servers with lots of USB ports.
Run rtty. It records each console to a log file 24x7, and allows multiple people to connect at the same time (including typing).
Tire Rack to the Rescue: http://www.tirerack.com/tires/tiretech/techpage.jsp?techid=35
"V" speed rated tires, rated to 149 MPH, are common on just about any "performance" car, like say a Mustang.
"W" (168 MPH) and "Y" (186 MPH) are in fact relatively common, and stocked in most performance sizes at your local warehouse.
The "Z" rating simply means "somewhere above that", and "go read the manual".
For instance the stock tires on a Z06 Corvette are: http://www.tirerack.com/tires/tires.jsp?tireMake=Goodyear&tireModel=Eagle+F1+Supercar+G%3A+2+RunOnFlat&leftTire=735YR8F1G2LROF&rightTire=735YR8F1G2RROF&vehicleSearch=false&fromCompare1=yes
$385 each, rated to 186+.
Now, if you want to push the boundaries of DOT legal, here are some auto-cross tires. They are in fact DOT legal, but not recommended for use on the street: http://www.tirerack.com/tires/tires.jsp?tireMake=Toyo&tireModel=Proxes+RA1&partnum=735ZR8RA1HC4/32&vehicleSearch=false&fromCompare1=yes
$265 each, Z rated. Once heat cycled they should easily be good into the low 200MPH range.
The insane Veyron tires become necessary somewhere north of 225MPH or so. Like with many things the forces get exponentially stronger with speed, and thus the cost gets exponentially higher.
This is someone who was ok with ASN.1, OID's, and "walking" tables that had no business in being walked, over an unreliable UDP protocol that initially had effectively zero security.
Someone stop him from developing a home automation protocol before his being "first" relegates that industry to 30 years of pain and suffering.
Absolutely. Here's another way to look at it. At $200/mil a year in spend, an (to make numbers easy) an average spend of $100k per employee (which for a salary+benefits+sg&a number is probably low), that's 2,000 people working on the web site. Facebook has 3,000 employees. And thats if it was all employee costs, no equipment.
Remember they have to interface with 50 different state systems, plus other government agencies. It adds up quick, 5 people per state to figure out the state requirements and interfaces is 250 people right off the top, for instance.
I'm afraid there are a lot of ./ users who have build a web site that sees 10k users a day, which is not that hard on commodity hardware, but have no clue what it takes to handle 10 million while staying inside the lines of a bevy of government requirements.
I see this comparison a lot, so let's dive into it.
http://www.marketwatch.com/investing/stock/fb/financials
Facebook had a COGS cost of 1.36 billion, and a R&D cost of 1.4 billion. Since they basically only operate a web site, that's all the cost of operating a web site. So to 2.76 billion dollars in a single year spent on Facebook.com.
Healthcare.gov spent 614 million over three years. At $200 million a year, that's roughly in line with Facebook's spending level back in 2009.
And Facebook has never gone down, right? It's never had a load issue, right? Yeah, didn't think so.
I was under the impression that people who signed up before March 31, a one time delayed deadline, did not have to pay the 2014 fine. The deadline for subsequent years would be December 15th.
And the first year fine, $95 or 1% of income, whichever is larger.
I've seen several people say "the cloud is the answer". I have one simple counter example:
Reddit. Any time they have a flash mob, and seemingly randomly almost every day they fail.
The clould does some things better and some things worse, and scales in different ways than a more traditional layout. Both can work if properly implemented, and one or the other may be faster/cheaper/better depending on specific site requirements.
GTA V? Sim City? Final Fantasy? Battlefield?
Turns out millions of users who start using something on the same day often don't follow the expected and tested for behavior.
Anyone who launches a service like this should expect to spend the first week in triage mode, and the first month making adjustments. I'd like to say proper planning would mean that never occurs, but the only way to insure that would be to spend 10x what is really needed. People would hate the government even worse if they did that.
This is not news, yet. It will be news in a month if it is still fubared.
You do realize apple controls how many are made, where they are released, and knows how many are sold the first month for each release?
No doubt there is a bean counter who figures out that they will do 20M 5s/5c the first month, so next year if they add 2 more countries they can do 11M day one, "breaking the record" on what will be 22M sales with growth.
These things aren't left up to chance. You'll know something is wrong when iPhone next is not 120% of iPhone previous.
Note that even if the consumer buys at the subsidized price Apple still receives revenue closer to the unlocked price, as the carriers pay the difference to attract the customer.
iPhone 5s "T-Moble Contract Free" prices, are $649, $749, and $849, depending on the amount of storage. See iPhone 5s. The iPhone 5C prices are $549, $649. See iPhone 5c.
Using an average price of $649, and 9 million units sold, that's $5.84 billion in revenue. That doesn't could any accessories (cases, car chargers, etc) or Apple Care sales.
GTA V made a relatively puny $1 billion. You know, chump change.
Correct. This is sort of like doing a review of an ordinary home door lock, smashing it with a police battering ram, and then declaring it useless.
The TouchID sensor is there to make things convenient enough that those who do not use a pass code now will use one. That's actually a huge leap forward, since it means a casual thief can't quickly get into your e-mail or contacts before ditching the phone. All it really has to do is slow someone down enough that you have time to get to a computer, invoke find my iphone, and remote lock it or even remote wipe it so people don't get your data.
I bet most people unlock with a thumb, but use an index finger on the screen.
You are correct that the disk will become fragmented, but the implication is fragmentation is a problem and that's simply not true. One of the prime causes of the misunderstanding is that fragmentation in Unix file systems is night and day different than fragmentation in a FAT file system, where most people are used to defragging windows drives. Unix file systems use much better algorithms to control fragmentation, so there is (generally) a lot less on a per file basis. They also automatically defragment, there are cases where when a fragmented file is written to the file system will defragment part of that file and rewrite it.
The Berkeley FFS was the first to "solve" this problem, reserving 10% of the disk space primarily to avoid fragmentation. Decades of experience show that for all but the most corner of corner cases, that is enough, causing no significant amount of fragmentation, or performance degradation.
* http://www.eecs.harvard.edu/~keith/research/tr94.html
* http://www.cs.berkeley.edu/~brewer/cs262/FFS.pdf
* http://www.cs.rutgers.edu/~pxk/416/notes/12-fs-studies.html
* http://pages.cs.wisc.edu/~remzi/OSTEP/file-ffs.pdf
The result is that for most applications fragmentation is a complete non-issue. After 25 years of playing with various file systems I've only seen it be an issue once, on an NNTP server that reached 20% fragmentation. Most user desktops and general purpose servers have under 1% fragmentation at all times. Generally, if you have a fragmentation problem it's because the storage is too full, and you need to add storage anyway (the aforementioned NNTP server was a good example). Adding the storage makes the problem go away.
Most users of Unix file systems will never need to give fragmentation a second thought.
The Civil Air Patrol is actually under the direction of the US Air Force, they are civilians in the sense that they are not members of the military, but they are military in the sense that they routinely train with the air force and national guard specifically to do SAR missions. The CAP is not "bubba jumping in an airplane to help", it's civilians who have invested hundreds of hours of training and coordination to be able to help in these situations.
http://www.gocivilairpatrol.com/about/
The national guard would be just fine with CAP planes as they have trained with them, and know that the CAP know all of the SAR procedures and their standard coordination to avoid mid-air collisions.
Your question is legitimate, and in point of fact we don't know yet.
I suspect so far what's happened is FEMA found out there was a flying object (the drone) in an area they had helicopters and planes in, so they shut it down. They don't know if it is safe, useful, or practical at that point, and their tired and true methods, and more importantly humans in those other machines take priority.
They may well talk to them, and find the tech useful, and be able to find windows for them to fly when other operations aren't in the area. Heck, they might be back in the air tomorrow. However badmouthing FEMA or the FAA to the press makes this a less likely outcome.
I hope you get modded funny.
Apparently you didn't even read the NOTAM.
"Altitude: From the surface up to and including 13000 feet MSL"
I actually got the wrong NOTAM, which is why the date is wrong. The right one is http://tfr.faa.gov/save_pages/detail_3_4333.html and was issued back on September 14th. It says "Altitude: From the surface up to and including 11500 feet MSL"
Ah, thanks for the geographic correction.
I wasn't aware any counties had delegated authority from the FAA, I've only ever seen that given to other federal agencies (like FEMA), or sometimes state level agencies. Did Boulder County actually have the ability to approve such a thing in the first place?
I don't think that's what the article says, but you might have to know some things about FAA regs for them to make sense.
"It has public safety flight approvals from the Federal Aviation Administration (FAA) to fly in some parts of Colorado."
They most likely have a conditional certificate to fly only in particular areas of low traffic/population for an experimental drone. That's similar to having a pilots license (approval to fly a plane), or even a drivers license (approval to drive a car).
NOTAM's, or NOTice to AirMen are temporary restrictions on ALL flight operators. Think of them as a construction detour in your car, or a bridge washed out barricade. A common NOTAM might be that a runway is closed for resurfacing, or that a chuck of airspace is blocked off for an air show.
So while they may have general approval to fly, the NOTAM cancels that for the specific area covered. Most likely the FAA has delegated to FEMA the ability to control all flights in this box as they coordinate SAR, Search and Rescue operations.
So to extend my car analogy, it's like there's a washed out bridge from a flood, and they put up a barricade across the road while they tried to recuse someone from the flood waters and these people simply drove around the barricade and said "we're here to help!". The answer was get back on the other side of the barricade, or be arrested.