x
<<   <   Page 4 / 4
dljvjbsl 12/4/2012 | 11:13:07 PM
re: RBOC VOIP Coming in 2004
Architecturally H.323 and SIP are almost isomorphic. If you think SIP does not contain POTS semantics, then can you explain why SIP has overlap dialing (cell phones don't have them!)?


The resemblance between the H.323 gatekeeper and he SIP proxy is only superficial. The SIP proxy with its registration and location functions has an entirely different purpose than the H.323 GK. SIP has developed its event service to create features and services that are beyond what is achievable with H.323.

SIP is about establishing relationships among parties and allowing them to create and manipulate the services that are specific to their needs.

H.323 extended the POTS model beyond any user need. Thick standards document on how to create features such as Call Forward on Busy for example, indicated to me anyway that this standards group did not understand what multimedia services are and how they are useful to customers.

The failure of NetMeeting proved that to me. I was told once seriously by product managemnent types that there was no need to develop multimedia features since that had already been accomplished by NetMeeting. I knew that wasn't so even then becasue I observed that no one was using NetMeeting. The company had supplied eveyone with NetMeeting, headphones and cameras. No one used any of it. All the video cameras that I saw had their lens caps firmly in place. No one could discern any practical use for it.

In answer to the issue of overlapped outpusing in SIP, this is a necessary function if SIP is to be gatewayed to the PSTN. A demonstration that SIP is fully capable of creating all necessary gateway functions is essential to its acceptance in the network. SIP has accomplished this with some exceptions.

The E911 issue is difficult for IP telephony with its lack of location information and the unwillingness of some IETF IP zealots to understand the necessity of obeying necessary legal and safety standards also makes this work difficult.
aswath 12/4/2012 | 11:13:07 PM
re: RBOC VOIP Coming in 2004 dljvjbsl:

Can you please clarify some of the points you make in Msg.#31? Just to make it explicit, I am not reopening 323 vs. SIP debate. It is over and nothing to be gained in continuing the debate. But it will be instructive for me to understand some of the points I might have missed. Thanks in advance.

Aswath

1. The SIP proxy with its registration and location functions has an entirely different purpose than the H.323 GK. I thought 323 also has registration function and can handle the services based on location information. Is that not true?

2. SIP has developed its event service to create features and services that are beyond what is achievable with H.323. In a recent presentation to FCC Henning also mentioned the event services. Why can't we define these event services in H.323? In what way this is different than trigger points in SS7? I am not talking about ease of development; but focusing on logical capabilities.

3. The failure of NetMeeting proved that to me. Is it 323's fault or that at that time people didn't have a need for multimedia and hence the lens caps were shut? Now that Windows Messenger supports SIP, how widespread is its use?

4. In answer to the issue of overlapped outpusing in SIP, this is a necessary function if SIP is to be gatewayed to the PSTN. I am assuming that the gateways interconnect to PSTN using ISDN PRI or SS7 trunks. If that is the case, the GW can send the addressing information to PSTN in en bloc mode. In the reverse direction, even if the address information comes in overlap sending mode, couldn't the GW accumulated the digits and send it to its SIP peer in enbloc mode? I thought the reason to have overlap sending is the desire to maintain the same user interface in the SIP end device as the ordinary phone.
<<   <   Page 4 / 4
HOME
Sign In
SEARCH
CLOSE
MORE
CLOSE