One thing that seems to slide back and forth with no real differentiation is exactly what someone means when they use the term VoIP. This always presents a somewhat interesting challenge to figure out in what capacity someone means VoIP.
I think that most of you here mean VoIP as in SIP across the public internet. The bigger problem here is not completely VoIP but more that the network that this method of VoIP is riding over is not capable of providing the QoS and SLA needed. Everyone knows these reasons and i'm not going to beat a dead horse here.
VoIP in a controlled network environment (just like the PSTN network) is just as reliable. VoIP and TDM are only as good as the switch they connect to. VoIP and TDM are only as good as the network they travel. VoIP and TDM are only as good as the software and hardware they run on. Logically we are not comparing apples to oranges but comparing an Orange to a Tangerine. Unfortunately there are more IT then Phone guys and therefore people with little to no experience with cabling systems or network design get designated to do so and only have the marketing they read telling them what they want. Silly consumers.
Telco switches do not have a consumer market, and therefore do not suffer from consumer woes. If no-name chinese offshoots made PSTN switches how stable do you think that network would be? Obviously not very. Unfortunately (or fortunately depending on how you want to look at it) the VoIP arena does have a consumer market, and hence, consumer woe's. It is being flooded with hack-job hardware and software in a "First to Market" attempt.
Has anyone looked at the VoIP angle from the manufacturer's and telco's point of view?
Now i'm sure someone will correct me, but is a Internet-Based VoIP or PSTN-Based service easier for the telco to provide? All they need to provide the customer is a data jack (read: DSL). No porting (termination done in some magic POP somewhere), no need to roll a truck to turn up an extra line or two, Changes and features are instantaneous, No need to run seperate copper for voice or data, and the internet (which you have assets in) already exists.
The Telco's are already doing something similar with SS7 by using D-Channels to dynamically determine switch circuit routing paths instead of static paths. DNS operates on the same principle.
I've already stated that I think it's cheaper for the commercial vendors to develop IP PBX's because of the vast majority of development tools out there for PC's. Are there more updates for an IP-PBX because it's that much less stable? Or because you dont need hardware changes or large production investments to change the system (Like older TDM systems). How many bugs have you found in your PBX Hardware that wouldn't be fixed because not enough people complained? If you say none because the hand of god came down and touched your hardware then either you dont know it's there or are lying
Now all that being said, VoIP software is still lacking stability and robustness. But then again, it's also the youngest thing out there. TDM Phone systems do the same thing with a V.1.0 but the primary difference is the vendor's keep it at V.1.0 (short of a major bug) for a few years making everyone get used to it's quirk and recouping costs before releasing an updated version as a completely different product. An old AT&T Partner and a Lucent ACS are essentially the same design, but with different updates, making one not necessarily work with the other. Other then the corporate buy-out, did it really take them that long to develop, test, etc the hardware design from a Partner to an ACS?
VoIP will be prime-time soon enough. The telco's are providing better provisioning, the commercial vendor's are scaling it up and TDM down, the marketting is going after it, and it truly does have some promise if you can look past today.
If I argue anything I argue for ambiguity. I still say the Internet and PSTN will merge to a point that it becomes the "PSTiNet" and no one really knows what they ride over.
Ohh yeah, I made post #100. I win!
