Paul,
From B.2.2.2/H.245v7 again:
simultaneousCapabilities is a set of AlternativeCapabilitySet. It is used to list the simultaneous capabilities of the terminal. An AlternativeCapabilitySet is a sequence of CapabilityTableEntryNumbers. Only those CapabilityTableEntrys that have been defined shall be present in an AlternativeCapabilitySet, although it is possible to define CapabilityTableEntrys and refer to them in the same TerminalCapabilitySet. If a terminal has a preference for the mode it would like to transmit or receive, and wishes to express this when transmitting its capabilities, it may do so by listing elements in AlternativeCapabilitySets in order of decreasing preference.
So I think you mean to refer to the order in AlternativeCapabilitySet.
The question remains: how to compare capabilities that are listed in different AlternativeCapabilitySets. These AlternativeCapabilitySets may occur in different elements of simultaneousCapabilities (unlikely) or in different CapabilityDescriptors (more likely).
Would you agree that the preference between two capabilities is structured as follows?
1. compare descriptor numbers - lowest is preferred 2. else compare simultaneousCapability element index - lowest is preferred 3. else compare AlternativeCapabilitySet element index - lowest is preferred
There is text in B.2.2.2 to support #1 and #3 of the above. Nothing is said about #2. And since other permutations of logic are possible, some clarification would help.
- TLyons@sonusnet.com +1 732 625-3003 ext 212
-----Original Message----- From: Paul Long [mailto:plong@IPDIALOG.COM] Sent: Thursday, November 01, 2001 3:04 PM To: ITU-SG16@mailbag.cps.INTEL.COM Subject: Re: [imtch323implementors] Conflict resolution for symmetrical OLCs - which is the preferred mode?
Terry,
B.2.2.2 refers to capabilityDescriptors, which are numbered according to preference, whereas C.4.1.3 refers to simultaneousCapabilities, which are arranged in order of preference. There is no conflict.
Paul Long ipDialog, Inc.
-----Original Message----- From: Lyons, Terry [mailto:TLyons@sonusnet.com] Sent: Thursday, November 01, 2001 10:41 AM To: ITU-SG16@mailbag.cps.INTEL.COM Cc: 'imtch323implementors@mail.imtc.org'; Lyons, Terry Subject: [imtch323implementors] Conflict resolution for symmetrical OLCs - which is the preferred mode?
Please consider these two extracts from RECOMMENDATION H.245 - VERSION 7
B.2.2.2 Capability Descriptors
If a terminal has a preference for the mode it would like to transmit or receive, and wishes to express this when transmitting its capabilities, it may do so by giving CapabilityDescriptors that relate to its preferred mode or modes small values of capabilityDescriptorNumber.
C.4.1.3 Conflict resolution
When the master and the slave have indicated choices of receive capabilities for a particular media type, the slave should attempt to open a logical channel for the master's most preferred capability for which it has capability, as given by the order the master has expressed its capabilities; and the master should attempt to open a logical channel for its most preferred capability for which the slave has capability, as given by the order it has expressed its capabilities.
---------------------------------------
The new wording "the order the master has expressed its capabilities" is insufficiently defined to act on and seems to conflict with the old definition of prefered mode.
As explained in section B.2.1, a terminal's total capabilities may be expressed over an extended period of time, not in a single message (Terminals may dynamically add capabilities during a communication session by issuing additional CapabilityDescriptor structures, or remove capabilities by sending revised CapabilityDescriptor structures). And the capability table entries by themselves are incomplete without accompanying descriptors (The terminal's total capabilities are described by a set of CapabilityDescriptor structures). Both capabililty table entries and capability descriptors are already numbered. Is the Order of Expression an additional attribute that needs to be inferred when processing the pieces of a received Terminal Capability Set?
It seems that some clarification is needed. Preferrably, C.4.1.3 should refer back to B.2.2.2 for an operational definition.
- TLyons@sonusnet.com +1 732 625-3003 ext 212
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ For help on this mail list, send "HELP ITU-SG16" in a message to listserv@mailbag.intel.com
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ For help on this mail list, send "HELP ITU-SG16" in a message to listserv@mailbag.intel.com