[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: [Simple] New IM-specific feature
Deepanshu Gautam wrote:
> When i say "protocols" i only mean to define a new XML schema (to carry
> QA as instance of it) and a content-type (for labelling that QA), thats
> it. I don't see any thing ugly in that. It has been done for other two
> IM-specific feature already.
I hadn't realized you wanted to carry this info in a body part.
But then you would need to specify in which messages it is conveyed, and
what the semantics are of including it in those messages. What did you
have in mind? Were you thinking of using REGISTER and its response to
convey this? Or were you going to send this in a MESSAGE?
> I can agree with you on bit of a relation between QA and what you call
> "IM-mail" and would like to consider it further. However, i can think of
> a prominent difference i.e "IM-mail" will not allow user to choose from
> the avaliable options in real time (i.e when it has received an IM)
>
> About storage, i think that would not be on registrar. Can we consider
> storage of QA as implementation specific? We just provide a mechanism
> (XML schema/content-type) to deliver QA between two entities. How to
> store them will be implementation specific. Some IM application server
> (e.g OMA IM-XDMS) or provisioning server can do that. Will that make sense?
The problem is that you have to pass the data in some message, and have
some reason to believe it will be received by something that will
process it as you expect. That implies some level of standardization of
the behavior.
(The long history of trouble with INFO is testament to what happens if
you don't address such issues.)
If I am to send the info in REGISTER, then I am counting on some
behavior by the registrar to do something with it. Its even more
important if I expect the *response* to the REGISTER to give me such data!
I still think this sounds like a provisioning problem.
Thanks,
Paul
> Regards
> Deepanshu
>
> ----- Original Message ----- From: "Paul Kyzivat" <pkyzivat at cisco.com>
> To: "Deepanshu Gautam" <deepanshu at huawei.com>
> Cc: "Simple WG" <simple at ietf.org>
> Sent: Tuesday, August 11, 2009 8:57 PM
> Subject: Re: [Simple] New IM-specific feature
>
>
>>
>>
>> Deepanshu Gautam wrote:
>>> inline with [DG]
>>> ----- Original Message ----- From: "Paul Kyzivat" <pkyzivat at cisco.com>
>>> To: "Deepanshu Gautam" <deepanshu at huawei.com>
>>> Cc: "Simple WG" <simple at ietf.org>
>>> Sent: Tuesday, August 11, 2009 6:44 AM
>>> Subject: Re: [Simple] New IM-specific feature
>>>
>>>
>>>> Inline...
>>>>
>>>> Deepanshu Gautam wrote:
>>>>> Dear All,
>>>>>
>>>>> I would like to propose a new IM-specific feature for SIMPLE, which i
>>>>> would call as "Quick Answer(QA)". Allow me to explain about it
>>>>> briefly.
>>>>>
>>>>> In instant messaging system, it is useful to have some readily
>>>>> available
>>>>> IM (text, audio or video) which can be sent in case of the receiver is
>>>>> too busy to type/speak/record for a reply. User can create and
>>>>> store IM
>>>>> (QA) using a particular client device and can use them when required.
>>>>> Creating and then storing on a single device may not be very useful
>>>>> because user can log-in a particular IM system using different client
>>>>> devices at different point of time. In that case QA created using one
>>>>> client device will not be available to another client device.
>>>>>
>>>>> I would like to propose a draft to solve this problem and perform the
>>>>> needed standardization, which will further make this IM-feature
>>>>> possible. Before moving forward I would like to get some initial
>>>>> comment
>>>>> from the group about this idea. You comments would be valuable.
>>>>
>>>> Where do you imagine this new state would be stored?
>>>> And who/what would use it?
>>> [DG] I don't see this as a new state (if you are referring to IM user
>>> state
>>> like "ideal", "active" defined in 3994). Irrespective of state, the only
>>> concerns here is that "an IM is received and user would like to
>>> select from
>>> a list of messages (here QA) as reply to that IM". Problem to solve: how
>>> to create and make that list of messages avaliable on various UAs.
>>> I foresee these QA to be stored both on server and UA.[/DG]
>>
>> Which server? The registrar?
>>
>> IIUC you are then just proposing that the registrar provide storage for
>> attributes that the various UAs registering to an AOR would share?
>>
>> That is a very slippery slope, since there are many pieces of
>> state/configuration that UAs might like to store or share.
>>
>>>> Are you proposing a "database" that is shared by the various UAs for an
>>>> AOR, that they would each update and retrieve from?
>>> [DG]No, Actually i'm proposing protocol(s) to make that kind of database
>>> avaliable[/DG]
>>
>> This is especially ugly when it calls for specific protocol additions to
>> SIP to maintain each piece of such state that comes along.
>>
>> This sounds much more like provisioning information, to be maintained by
>> a provisioning server and made available to the controlled devices.
>>
>>>> Would the sending of the answer still depend on a UA receiving the
>>>> message? Or do you expect that the answer will be sent even if there is
>>>> no UA registered?
>>> [DG] According to me, till now, it depends on UA receiving the message.
>>> But,
>>> it may be other way round too, if we have feasible
>>> usecase/requirement.[/DG]
>>
>> If you go that way, it becomes much more like a voicemail system.
>> (In this case, a "IM-mail" system.) Namely it is a UA of last resort for
>> responding to incoming requests if no other UA does. Its not really part
>> of a registrar, though it *might* be collocated with one.
>>
>> This is not an unreasonable thing to do - its pretty common. But so far
>> there have been no proposals to standardize anything about this as part
>> of SIMPLE. (Nor have there been any to standardize voicemail systems.)
>>
>> Thanks,
>> Paul
>>
>>>> IMO this doesn't fit well with the simple model.
>>>>
>>>> Thanks,
>>>> Paul
>>>>
>>>>> Regards
>>>>> Deepanshu Gautam
>>>>>
>>>>>
>>>>> ------------------------------------------------------------------------
>>>>>
>>>>>
>>>>> _______________________________________________
>>>>> Simple mailing list
>>>>> Simple at ietf.org
>>>>> https://www.ietf.org/mailman/listinfo/simple
>>>
>>>
>
>