Originally Posted by Operon
Quite right about session management and the like. If they used some other protocol other than IP, then finding an ex-Cisco solution would be a nightmare. Now, however, presumambly, a third party might reverse engineer a solution, (less likely), or Moxi, might reasonably easily and with minimum cost accomodate TA issues in future models. Notwithstanding the total number of Moxi, Tivo and Ceton/HTPCs, its a relatively small share of the STB user base, and unless the FCC makes some sort of broad declaration vis-a-vie, third party devices speaking to CC's head ends, this Cisco box is all we'll see for quite some time.
The situation is actually worse than you surmise due to the cartel organization known as CableLabs. Basically, it's literally going to take an act of Congress to fix this (I'm not sure that the FCC currently has the legal power to mandate what is needed, though they are certainly welcome to try).
Reverse engineering a TA is a near-impossibility due to the fact that CableLabs would never certify a resulting device, thus rendering it useless for the market. It's doubtful that Cisco or Motorola would license their tech since that would mean less revenue for them and they aren't having any issues meeting current demand. Even if, in some dream world, Moxi/Arris *did* manage to get permission to implement TA functionality into their offerings...we would still be stuck with an external dongle/box most likely since the TA tech for Cisco head-ends is not compatible with with Motorola headends, and vice-versa. Even within the same vendor..there are individual market differences in the software revisions that the TAs/head ends run for the SDV communication that would have to be taken into consideration and kept updated market-by-market similar to guide data. I just don't see Arris taking that on even *if* they were legally and technically able to do so...especially given the ramifications of a screw up.
So again, we circle back to the SDV-over-IP thing whereby each consumer device communicates directly with either the head end or some proxy/gateway device that is provided by the cable company (eg a single stand-alone "TA" that your gear would speak to over your home LAN rather than being connected via USB). In the first case (least likely IMHO due to security/stability concerns over 3rd party communication stacks) the existing TA's could be re-flashed with a gateway software and re-used as basically special-purpose cable modems to provide an IP-based out-of-band signalling channel for consumer devices like Moxi and Tivo. Even the Cisco TA boxes that we all know and love currently would be fine in this case (with the addition of a usb-based ethernet and/or moca adapter for LAN communication) as the channel guide and session info would be off-loaded to the consumer gear, and the TA's would only be providing a communication conduit so session/stream limitations would not be an issue for the TA.
In the second case (more likely, but likely more expensive and with a much longer timeline for rollout) a beefier gateway device would be placed into each home that needed SDV functionality and it would function basically as today's TA's do...though it would only communicate with our consumer devices over ethernet/Moca rather than USB and there would only need to be a single one per household rather one per consumer device needing SDV access. I think this is the "middle ground" that CableLabs and the FCC/Congress will ultimately settle on..as it addresses the CableLabs concerns about letting end-user gear talk directly to their head-end equipment...and it gets rid of the per-device TA fiasco that we are currently languishing in. The Motorola TA's might actually be strong enough to use as-is for the purpose with new software (again, with the addition of a usb NIC of some sort)...but Cisco would definitely need to bring something new to the table hardware-wise to pull this off. The reason I keep harping on the re-use thing is that the solution might be more palatable to the vendors if they don't have to trash their investment they've already made in TA adapters...thus speeding up deployment of the new system.
The third option (favored by the FCC though sure to meet with very stiff resistance) is a universal gateway that would replace all cable boxes, TA's, satellite receivers, etc...and abstract the provider from the consumer by providing a standardized interface for consumer electronic devices (such as our beloved Moxi) to pull down channel information, request channel changes, and participate in 2-way services such as on-demand content. The home gateway would disperse content in standardized DLNA-based streams over the customer's LAN which the gateway would have transcoded on-the-fly as required. That's the FCC's wet dream....and I agree it would be nice to have all consumer equipment such as Moxi and Tivo be able to work with any video content provider similar to how your landline phone works with all telcos....but I just don't see it happening in the reasonably short-term future. But I think "option 2", as I said previously, is do-able and probably something that might happen in the next 3-5 years.