[standards-jig] Re: [Foundation] Motion for Last Call - Chat State Notifications (JEP-0085)

Alexander Gnauck gnauck at myjabber.net
Thu Dec 4 17:16:15 UTC 2003


Hi,

i see no use for JEP-0085. The most existing clients already support
JEP-0022(message events). And that works very well so far. Im not
interested to change our client and add JEP-0085. Instead of this we
should work and move forward with JEP-0022.

Alex

Sebastiaan Deckers wrote:
> Hello all,
>
> I recently implemented both jabber:x:event and chatstates while
> working
> on the conversation dialog of my client.
> Here is a list with pros and cons of each protocol, as seen by your
> humble servant.
>
> Chatstates
> + Straightforward design
> + Ridiculously easy to implement (~60 minutes in my case)
> + Introduces new conversation modes besides composing: active,
> inactive, gone
> +/- Works complementary to jabber:x:events offline/displayed/delivered
> - Can be tricky to coexist with jabber:x:events for composing.  The
> JEP could use some implementation notes on how to do this properly.
> - Why does this use <message/> instead of <presence/> or even <iq/>?
> (Note: I did not look at the thread handling as my client does not use
> threads in its display.)
>
> jabber:x:events
> + De facto standard
> + Widely supported (I implemented it how many years ago??)
> - Composing does not "belong" in the same namespace as delivered and
> displayed
> - Yucky yucky usage of id attribute which has lead to incompatible
> implementations
> - Limited to composing as the only conversation mode
>
> Overall I would say that Chatstates is a clear improvement.  However
> jabber:x:events is still required to handle displayed and delivered
> events as Chatstates has no provisions for those (but should not have
> those anyway).
>
> I hope this may help the decision of the Membership and Council.
>
> -Sebastiaan
>
>
> Julian Missig wrote:
>
>> Ok, I understand that I suck at writing up JEPs when I should and
>> that
>> I said I would a long time ago, but I still feel that the Chat State
>> JEP is more complicated than we really need... I'm also rather
>> disappointed that we have a standard which almost everyone is using
>> (jabber:x:event) that we're deprecating just as people are finishing
>> up implementing it. I do think that we could get the things we need
>> by modifying jabber:x:event rather than creating a whole new
>> standard and abandoning jabber:x:event.
>>
>> Julian
>>
>> On 3 Dec, 2003, at 12:20, Peter Saint-Andre wrote:
>>
>>> On Wed, Dec 03, 2003 at 02:16:15PM +1100, Robert Norris wrote:
>>>
>>>> I move that a Last Call be issued for JEP-0085, Chat State
>>>> Notifications. Seconds?
>>>
>>>
>>> Just a note (I'm one of the JEP authors, so I won't second it):
>>>
>>> As discussed in yesterday's Council meeting, several Council members
>>> have mixed feelings about moving forward with this JEP and would
>>> like to receive feedback from client developers about this, since
>>> it is intended to replace jabber:x:event (JEP-0022). However, the
>>> Council figured that the only way to close the circle here would be
>>> to hold a Last Call on JEP-0085 so that we can have a definitive
>>> discussion about this on the Standards-JIG list.
>>>
>>> So if you want to have the discussion and reach consensus on
>>> whether we will move forward to chat state notifications (JEP-0085)
>>> or use the message events protocol (JEP-0022) forever, please
>>> second the motion. Let's get this figured out and move on.
>>>
>>> Peter






More information about the Standards mailing list