This is pretty long; but if you want to read it, it gives basic background of the telecom system, and places where you can research the nitty-gritty details that keep the 100 y/o phone network running. All this information is easily found online, and there are dozens of support sites that network operaters also frequent that deal with managing and working within, and around, the inherent limitations in the hardware and control systems. While it's nice to be able to arm-chair speculate, the Internet has made it much easier to get insight into what goes into operating many of the systems that run the world. The following is generalized from prior experiences implementing and working with systems; and may not represent 100% what Comcast may be doing, but is common practise. The maximum blocking limit is pretty standard in the US because of the PTSN (Telephone Switching Network) hardware. The switches themselves have extremely limited memory for additional switching features after all the routing information is loaded (which is more important, otherwise your calls won't reach their destination), and you get the "Your call could not be completed as dialed" message. Because of this, memory management is of greatest importance, and every last bit is managed. When states went through from 7 to 10 digit dialing, this increased the table size, and decreased the amount of space for secondary features. Now, this is for legacy systems that use the physical copper network; which is faster, and more reliable. As telecom operators switch to Internet Relay systems, this involves updating hardware, network infrastructure, management, and interfacing with the PTSN network. -IF- your local connection, and all associated hardware between your nodes have been upgraded, then there is a possibility of increasing this; as the newer devices have greater CPU power, and more memory, as they have to interact with SIP session protocols. However the limitation will still be in the switching hardware; which are commodity items that are purchases, and fairly standard (these restrictions will usually apply to all carriers, based on which features they enable or disable at the switch; out of the dozens of switching features available, each operator has to manage the fixed space, and how much data they can give to each feature, so that the switch doesn't run out of hardware memory and begin to drop calls. As for firefly5's comment on not possess the software to counter the spoof numbers: Loaded question. Systems generally don't work like this; and implementing such a feature would degrade the performance of all calls significantly, and severely decrase reliability. This is because in order to get the call to a software processor, one must convert the call from PTNS (Copper/SS7) to SIP (VoIP) and route the call to the processor. This processor then "answers" the call, and runs whatever logic they can against all the data that is passed in (which varies from every vendor of switch and intermediary router between call originator and destination, more data added as the call gets passed onwards; one of these is all the billing information. Each telecom operator must pay for all the data they send to other telecom networks, and charge for calls passing through their routers; this pricing information is available, and depending on how much you pay for your service, they will route to more expensive routes (faster, more direct, or reliable), or the discount routes (less reliable, longer route means more delay between speaking and hearing). This is a delicate balance. BUT; this information is in the headers, and you can only trust that the number (originating caller, billing caller, intermediatary billing callers, etc) and once the call is connected, the charges are concretely incurred. Now, as for not possessing the software; if this were to be used for all callers, there would be a load added to every incoming or outgoing call; and in addition, heuristics will need to be used to determine what constititues a "spoof number". From the call data standpoint, they look identical to authentic data; otherwise hey would be trivially blocked. And as for thinking that you can block someone simply, there is also the problem of each switch that processes the call routing along the path, may update or modify data before it is relayed forward depending on brand, date of purchase, etc. Thus, a person calling me may have different caller id's depending on time of day they called, or current network congestion. This is particularly true for "no caller id" users, where depending on the switch, the CNAM record will be populated as "6140000000", "CBUS W OH" "COLUMBU OH" "614" "6146661212"(the actual number duplicated), "CELL PHONE OH" "UNKNOWN" "OUT OF AREA" "614 CALLER" etc; These are set up by the individual switching operators; and are effectively meaningless, while it's required to pass this data onto the end-user unmodified in receiving the call; many places will filter out the more common items in billing and online histories ( ### number; CELL PHONE <state>, OUT OF AREA, etc) And presumably the third issue is the legal liability of doing it automatically. The US is a very litigatious society, and it has been proven over time, that even if a user consents to a service or action, if something unexpected happens, they can still sue for damages incurred by using the service they opted into using; not all risks are known, and it's folly to expect to convey all the unknowns to a user. Other countries use "personal responibility" to shield companies, not so in the US. Here, one is not allowed to say "By using this service ther are the following risks: <list>" Any unknown risks, extreme uncommon risks, anything that possiblily could happen, but never has -- are generally not listed; BUT if something happens, in the US, huge lawsuits will ensue. The rest of the world, one is expected to use critical thought, and think about what could happen if they use a service that changes the status quo; and up to them wheter the unknowns are worth the benefit.. It is currently law to complete every call into the network as dialled. Period. This was passed due to the former practise in the 80s to drop calls that come in on expensive route swtiches. Once this practise was outlawed, dropped calls ceased, but it also opened up the new process of creating cheaper, indirect, more-latency routes, which telecoms could use; but also make for annoying calls where there's a more noticable delay between speaking and when the other person begins hearing. Now, if Comcast were to block calls automatically, it would not only violate this law; it would also open itself up to potential lawsuits should an important call be mis-blocked. Taking the safe route, they will only block calls that are explictly blocked at the command of the end-user. Which is sensible. THey seem to be willing to offset that risk by enabling routing of your call (forwardngi) to multiple parties (like nomorobo) which you a) intentionally elect to do; and b) accept the risk that they will block important calls; and c) is not comcast, thus liabality is placed nearly exclusively on the third-party providing these blocking services. Common complaints have included not receiving any calls at all. Third party blocking services have incorrectly blocked calls from completing; answering calls without a call-transfer; incorrectly blocked all incoming international calls (being unable to handle 11+ digits) or simply gone down, leaving the subscriber with no incoming phone service. I can see why Comcast would not want this excessive liablity and risk on their shoulders; especially with technologies and services that are very nacient (< 10 years old) vs the services and tools which have been proven with known risks and error rates over decades across many providers.
... View more