There is a lot more in a PSAP than a typical call center. First of all, they are also paying a portion of the networking costs that get the calls to the call center. Then the skill level for a telecommunicator is a lot higher than a call center, and the manpower costs are probably twice or three times your number. Then they also have to overstaff it - you don't get a "we are experiencing a large volume of calls..." from 9-1-1. They have telecommunitors waiting for calls, not callers waiting for telecommunicators. Then they have dispatch systems, which have accompanying record management systems.

And all of it has to be reliable, and it's specialized for 9-1-1, which is a limited market, and the net result is that the systems are expensive. If there is not a lot of diversion, the surcharge pay about half of the cost of running the system (not including responders).

If you doubt it, DOT paid for a study a few years ago. Booz Allen did it. I'm sure you can find it if you want to look. Total cost of the system at that time was $2.3B. Has lots of detail on where the money comes from and where it goes.

Turns out this is a horrible idea. It sounded great, but it causes many, many more problems and it roughly never helps. When you call 9-1-1 from a phone that is not active (it's called an "uninitialized phone"), they don't know who you are (there is no phone number, and no subscriber for that phone number) or where you are (the location stuff needs the phone number to work). 9-1-1 gets this untraceable call from an unknown location.

Every 9-1-1 call center (PSAP) gets thousands, sometimes tens of thousands of calls from these devices every year. Most PSAPs never get a "good" call - they just don't happen. There are kids playing with them, there are flea markets where, since the only call a phone without a plan can make is 9-1-1, they call 9-1-1 to show the phone works before a buyer pays for it. There are people trying to trick 9-1-1 to sending a SWAT team. It goes on and on.

What if speaking isn't a possibility? Like, say, you are deaf. Or maybe you are in danger and letting someone know you are calling 9-1-1 would not be good for you. Of you are young, you are stressed, and it just makes more sense to you to text rather than call.

Normal SMS does not send location. When you send a text to 9-1-1, they get your location and forward it to 9-1-1.

Yeah, at least in most places, there is a "surcharge" on your phone bill that pays for part of the 9-1-1 costs. HOWEVER, it's not enough to pay the actual cost AND in many, many states, the money is siphoned off for other uses. This diversion of surcharges is a huge problem, but states love dipping in to that piggy bank, and sometimes even local governments decide to use the money for a new police cruiser rather than give it to the 9-1-1 system. The rest of the money to operate the system comes from local revenue, except in some states like Vermont, where it comes from general state revenue.

This is not true. Calls from cell phones, as well as SMS is routed to the nearest 9-1-1 center (PSAP) based on the location of the cell tower and sector serving the call or text. In California, and a couple other places, the state police call center answers the call, but they get your location. Then they try to get your exact location using the GPS in the phone, or by triangulating from cell towers. This doesn't always work, and even when it does, it may not be very accurate. So, they always ask you where you are. They know about where you are, but not exactly. Of course, if you are indoors GPS may not work.

No different I think. Yahoo would pay for peering if it's bandwidth was asymmetric, but it would probably pay less to peer directly than through a transit network. Before peering, they would pay an ISP for access. After peering, they would pay the peering partner for the asymmetry, but that would be less than what they were paying for transit. If Yahoo users generated more traffic to a peer than they consumed, the peer would pay Yahoo. That probably didn't happen.

Yeah, you are paying for it, and they should deliver it to you.

But both ends pay. Netflix, or whomever, pays their ISP, you pay your ISP. Netflix doesn't get a free ride.

When people talk about net neutrality, they worry that Neflix has to pay twice, once to their ISP, and once to your ISP. We don't want that.

But Level(3), one of Netflix's ISPs, may have to pay Comcast if Level(3) sends more traffic to Comcast than Comcast sends to Level(3).

Then again, Comcast better handle that traffic equally well, and better have the capacity to exchange the traffic fairly.

You can't sweep the problem of real cost under a claim that we have lots of asymmetry. If I generate about as much traffic as I get from a peer, then our costs to deal with the traffic are roughly equal. If he sends me 10X what I send him, my costs are higher than his. If the costs are different, at some point, the price is different.

Since the beginning of peering, the rules have always been that if you have roughly the same amount of traffic inbound and outbound, peering has no charge. If one direction generates more traffic than the other, the source pays for the asymmetry. If you give me 200 GB per minute average, and I give you 100 KB per minute average, you have to pay me for the traffic you are giving to me to deliver to my customers.

Streaming video has this problem - it's all one way. Peering should cost video streaming sources. The RATE charged has to be reasonable, but they don't get free peering.

They pushed the update out on Jan 7. By Feb 22, it hadn't been completed. Something is not right with this explanation. Doesn't matter how low a priority it was, it should have been pushed out within in what? Two weeks?, a month?

Your kids/grandkids/greatgrandkids (as appropriate) will not know what it feels like to be lost. Think about that. The notion of being lost is a universal human feeling not quite like any other feeling. We've all been lost, regardless of our navigation skills. The next generation won't ever be lost. They won't experience that feeling. They may be in unfamiliar territory, but they will know where they are.

There are a limited number of sources for the data that is "what street address is at what latitude/longitude?" which is technically "reverse geocoding". They are:
a) The government
b) Private companies who spend lots of $$ gathering the data

In the U.S, the government sources are:
a) The TIGER database - this is not good enough for the task, but it's free
b) Local city/township and county governments - this is the very best data when it exists, but it doesn't exist in lots of places, and it's hard to get in many places where it does exist
c) The 9-1-1 system often has their own source of address data which is used to figure out where you are when you call from a mobile phone

In the U.S. the private sources are:
a) Navteq
b) Tele Atlas

All of the other places that seem to have data actually get it from the above sources one way or another. Sometimes, they have auxiliary data like satellite images or street level images, but the database that links street addresses to geocoordinates comes from one of the above sources. Note that Navteq and Tele Atlas try to get the local city/county data when they can. When they can't they "drive" streets with a GPS equipped vehicle, clicking on houses and other buildings as they go. The 9-1-1 system does the same. The city/county data is actual map data, with polygons for streets, parcels, etc. It's often hard to get address data from it without additional work because the city/county data is developed for land use planning and tax revenue and not reverse geocoding.

The local data probably ought to be freely available, and it's the most accurate, although often somewhat incomplete source of data. Trying to get free access to TeleAtlas and Navteq data is not going to work, which means getting it from Google, Twitter, etc is not going to work.

Other countries have different situations. As noted above, the U.K. mapping data is available, and is excellent quality.

One good story deserves another, from several years ago

There was this medical device manufacturer. It had an older product, pre-microprocessor. One day, the FDA came for an inspection. When they do that, they usually send at least one person with clue, but they cross train other people and send them too. On this inspection, one of the inspector's regular job was inspecting galleys in ships (another FDA function you may not know about). This guy had been cross trained.

So, they are walking down the manufacturing line, and the employee shows them the board from the product. One of the chips has a label on on. The inspector says "PROM"? Meaning, is that chip a programmable read only memory (like today's flash, but usually one time programmable and a lot smaller). The employee says "Yes, that's a PROM". The inspector says "Checksum?" and the employee says "yes, the checksum is on the label". The inspector says "Verify?" and the employee takes the board, pulls the chip, goes over to the programmer, plugs it in and verifies that the checksum is valid.

The inspector says "Source Code?". The employee is a bit stumped. He goes away to ask some engineers who were around for a while, then goes to the manufacturing engineering guys and finally goes back to the inspector and asks them to accompany him to a storage room.

In the storage room, there are a number of 4 drawer file cabinets. The employee searches around, and finally finds the right file.

The file has the right build data on the cover. He opens the file and triumphantly removes the floppy disk with the source code on it.

An 8" floppy disk.

You know what's coming right?








No 8" drive left in the company.

Increasingly, there is good data from local government. Both TeleAtlas and Navteq try to get this data. Not all governments make it available on reasonable terms. In some states, there is an organized effort to create good maps of the entire state. There is also some effort to coalesce mapping collection within government. Often there are four or five independently developed maps. A county may have a GIS department, your local town or city may have one, your local 9-1-1 PSAP has one, and often there is a state map. While today, they all are independent, with different "base maps", we do see some changes where there is sharing of map data among the government entities. The ideal is that local government has a single, accurate, up to date map, which feeds both state-wide maps, and is made available to the commercial companyies who depend on good map data. I work on the 9-1-1 system, and I can tell you that, for example, if the local utilities used the same base map as the PSAP, things would be A LOT better, and the utility crews could probably provide another great source of error checking, updates and additional information that would benefit other map users. It could be win-win: local government provides the base map and a set of public layers, which is given at low cost to commercial enterprises so long as they contribute errors, updates and layers appropriate for the government to have.