I suspect they will be doing PTT over something like SIP and RTP. This is known as Push to Talk over Cellular (iDEN is not classified as a cellular network for regulatory purposes in the US) and is abbreviated PoC. It was standardized by the Open Mobile Alliance for both 3GPP (GSM-family) and 3GPP2 (CDMA-family) networks.
I'm curious as to what makes COBOL the right tool for data processing tasks.
I was under the impression that much of the reason it was still around is generally because there are existing large projects already written in it, and it is generally deemed to expensive to try to convert to some more modern language. You make it sound like there is more to it than just that (although surely it plays a part).
What makes it a better language than say Java or Python for data processing tasks? If one chooses to use those languages in a more purely procedural style (rather than an object oriented style) would they not produce similarly straightforward code, but with the advantage of having a much larger pool of developers?
Early on, cable lines broadcast exactly the same signal to everybody in a city. These days that is no longer true. Cable internet basically requires that the city be broken up into multiple signal domains, perhaps as small as one per neighborhood. This is also used to provide targeted commercials, and on demand content.
Now that we have targeted areas, it is possible in theory to only send the channels in use in that area, and letting the system reuse the space for unviewed channels as DOCSIS channels. Indeed this technology has existed for a while. Yet, correct me if I am wrong, I believe this system is not in active use.
This is true, to an extent.
Targeted area's are really only as accurate as the provider makes them, and its filtered more by the physical line that they're on vs the IP address that they have.
True, but the fewer people on each cable, the more useful being able to able to broadcast the digital channels only 'on demand', letting them become data channels when not in use. If you have one area for the whole city then most of the channels will be in use most of the time, but the fewer in each area, the fewer distinct tv channels are likely to be being watched at any given time, and thus more channels available for data.
For example if CMTS 1 Services Central PHX and CMTS 2 Services East PHX, you can know what area's a node on each is going to affect down to the street addresses if you have an outage.
The problem is Analog broadcasting. The FCC says that if you aren't transmitting for older TV's on your lines, you have to provide an Analog converter. In many smaller systems its cheaper to supply a digital converter and do away with analog entirely since the equipment costs for side by side broadcast are more than just putting out a couple hundred converters (that the government gives a tax credit on).
There's the final part of the problem. The internet switches (nodes) only control the access so long as the equipment exists in three places. The office, the node and the modem at the user. In order to broadcast digitally in the same manner that the internet works, every TV for every customer must be compatible. That means the big, expensive converters the government doesn't subsidize. You know how you pay 5$ a month for them right now? If they threw that switch, there's a good chance the FCC could interpret the rules of the digital cut over to provide those for free, since now they're 'necessary' to have any TV connected. By keeping it simpler its easier to charge more money. *
Could they not avoid this whole mess by always broadcasting the non-encrypted analog channels, and only do the 'channels on demand, DOCSIS when not demanded' for the channels that already require a set-top box or CableCard? That sounds pretty easy. Existing cable boxes could be upgraded in place with a firmware update, so that they communicate the channels they are watching in the same way they communicate VOD requests. Upgrading the CableCards might be harder depending on a few implementation details, but my understanding was that relatively few subscribers opt to use CableCards anyway, so even if they had to all be replaced, the cost should be relatively low.
*Note: I never said I -agree- with any of the practices in place. However, show me a for-profit business that isn't out for money and I'll show you a lie.
Upgrading systems costs an insane amount of money. That more than anything is the reason that cable monopolies exist, the cost of entry prohibits competition. To install a new plant in an town of 50k takes something to the tune of 2-3 million dollars, with zero guarantee on how long it will take to recover that cost, if ever.
That might be part of the reason for cable monopolies, but the bigger reason is the local laws in most cities that explicitly grant one company a monopoly.
Cable lines have reached their limit unless someone comes up with a new way of multiplexing, and if its that significant a step up you'll see it deployed very rapidly.
Early on, cable lines broadcast exactly the same signal to everybody in a city. These days that is no longer true. Cable internet basically requires that the city be broken up into multiple signal domains, perhaps as small as one per neighborhood. This is also used to provide targeted commercials, and on demand content.
Now that we have targeted areas, it is possible in theory to only send the channels in use in that area, and letting the system reuse the space for unviewed channels as DOCSIS channels. Indeed this technology has existed for a while. Yet, correct me if I am wrong, I believe this system is not in active use.
That would be true, except that in the claims themselves they do not use the term broadcast media, only in the disclosure. In the disclosure one may freely use the general meaning of terms even when they have a distinct legal meaning.
In the general meaning of the terms, cable is definitely broadcast media, since it is not unicast, anycast, or even multicast. Besides they specify in the vdisclosure that what they label a broadcast source "includes, for example, a satellite, an antenna, and/or a cable network such as, for instance, fiber optics, analog-to-digital conversion, and/or other types of cable networks."
The patent is explicitly about purchasing recommendations influenced by the broadcast media the user is currently consuming (i.e. recomendations based on the TV show/commercial/movie/infomercial you are watching *right at this minute*), and what other people purchased while consuming the same program, combined with data about items being shown in said program.
The description gives the example of a button on your cable box/sat receiver remote that you can push while watching TV, which will add a border next to the show allowing you to purchase what is being shown, and offering recommendations for other similar products.
Did you set the options correctly?
If I set the cracking hardware to be "an average GPU" and the same password that would take 2 days when protected by Microsoft Windows System (1 round md4) would take 54 centuries using bcrypt.
Admittedly the software on the website is only set up as a demonstration. It grossly underestimates the speed of GPU based cracking at the moment (it multiplies the speed of CPU by the number of stream processors), and lacks many types like crypt-md5.
But the underlying concept (determine the pattern used, assume the cracker knows the pattern used, calculate the number of passwords that fit this pattern, divide by crackers check rate) is sound.
Correct. Both issues did come up, but it is likely that some form of compromise could have been found. The real problem though was that the firefox logo was under a non-free copyright license, in addition to the trademark license. Debian does not allow any software under a non-free license (except for license texts themselves) in main, and they take that very seriously.
Debian wanted to use the empty globe logo, but Mozilla declined to permit that.
Debian would have been willing to use a redrawn version of the logo that had a free copyright license, but Mozilla declined to permit that.
Thus the "fork" was required if Debian wanted to ship a Firefox compatible browser in main.
You also gain the bonus feature that with a single config line change, you can put one of your private "NATed" machines out in your DMZ and don't have to reconfigure anything else but one entry on the firewall
To people who care about security and know their stuff that is a bug not a feature. Think about what happens if one day someone fat-fingers the firewall config. The DMZ servers would be hardened so they might survive the exposure. The other machines on your private network are unlikely to be safe when accidentally exposed to the world. In many real world corporations there are usually servers that can't be locked down that tightly.
Really? That's your argument?
If you are using a many-to-many NAT setup (as many reasonably sized companies would require), you are able to place up to one machine in the DMZ per external IP. So the mistake in question is already possible without
Furthermore many large companies have never used NAT, and they don't have these problems. They have only ever used public IP addresses, and a stateful firewall. They avoid issues like you are talking about by being careful, and having security in depth. For example having multiple firewalls, can prevent accidentally placing a machine in the DMZ with a single mistake. You could make it such that an IP address must be explicitly listed in the edge firewall to be in the DMZ. If you also have the inner firewall configured to require stateful connections for all machines, then the only way to accidentally expose a machine is to make two mistakes. The mistakes could be placing an internal machine in the DMZ vlan and also adding its IP address to the edge firewall, or managing to mess up the configuration of both firewalls simultaneously.
You can avoid your first issue with Google by using verbatim search mode.
To activate it, add '&tbs=li:1' (without the quotes, of course) to the url. In the alternative, it can be manually activated by clicking the link on the left side of the results page labeled "more search tools", which will cause a list of search modes to appear. You can then choose verbatim.
DHS is not a law enforcement agency. Rather is is a United States federal department. It does contain more federal law enforcement officers than any other branch, but that does not make it a LEA.
The NCIC database is another example. They normally only give access to employees of actual law enforcement agencies (the employes are, however, not required to be law enforcement officers). Thus for the TSA to access it, would require that they used employes of the Federal Air Marshal Service, or get an exception to the usual policy.
Did you see the price tag on it??
I never understood why they didn't tie in the TSA checkpoint with state DMV and ICE.
Does the photo on the ID handed to you look like the photo on the screen? Yes/No
Does the name on the ID handed to you match the name on the screen? Yes/No
Does the name on the boarding pass match the name on the ID? Yes/No
Does the airline ticketing system information match the boarding pass as provided? Yes/No
If any questions were answered with a "No", separate the person for further evaluation.
Nice, but you forgot one important step there. Namely:
Does the photo on the ID match the person presenting it? Yes/No
That and the fact they they had the ID in their possession are the only things that tie the person to the claimed identity.
But yes, that is a far more sane idea for improving security than almost anything the TSA has done.
The real reason that this is not being done might be that the TSA is not a law enforcement agency. Thus it is possible if not plausible that one or more states would not permit them electronic access to DMV (or BMV, as the case may be) records. State law may reserve that ability for law-enforcement (and the DMV itself), leaving only printouts or quarterly data dumps.
Why would end-to-end be broken?
Just because your IP address might change every week or so does not prevent a remote computer from connecting to yours. And you can always use dynamic DNS to host a server.
Actually the value being widely recommended is a
Let's assume your provider gives you a
It would be easy enough for the provider to offer dynamic IP block service, where your modem/router gets assigned a different
Then all you need to do is turn on privacy SLAAC addresses, which prevents using just the local part to identify you.
I fully expect that by default ISPs will provide dynamic blocks to most consumer clients, if for no other reason than to discourage hosting servers.
We all like praise, but a hike in our pay is the best kind of ways.
