[Standards] MUC History clearing

Joe Hildebrand hildjj at gmail.com
Wed May 14 16:48:17 UTC 2008


I would propose that we write a new XEP for the new functionality.  In my
opinion, adding new features to an existing, non-experimental XEP should be
done only as an absolute last resort.  It makes the life of a software
developer that does QA *very* difficult.

In this case, I would add one new XEP that shows how to do ad-hoc commands
in MUC.  If particular commands need to be standardized, they can have their
own XEPs.  If we wanted to, we could bundle one or more of these
standardized commands in with the base MUC/Commands XEP.

The only reason you would need to standardize particular commands is if you
wanted to be able to process them semantically (like via a bot), since
human-facing clients can just render the x:data form to the user.  In my
opinion, it should not be required that we standardize all, or even many of
the commands, unless we wanted to start re-implementing the base use cases
of XEP-45 like kick, ban, and edit affiliation list.


On 5/14/08 7:36 AM, "Boyd Fletcher" <boyd.fletcher at je.jfcom.mil> wrote:

> so how are we going to publish the new ad-hoc commands? separate XEPs are
> modify the  MUC XEP?
> 
> 
> On 5/13/08 11:09 PM, "Peter Saint-Andre" <stpeter at stpeter.im> wrote:
> 
>> On 05/13/2008 6:04 PM, Joe Hildebrand wrote:
>>> Any change we make to support clearing history is "the new way".  It has no
>>> bearing on testing or interoperability.
>> 
>> Right.
>> 
>> If we were defining XEP-0045 right now, we'd use ad-hoc commands for
>> features like kick and ban. As people want to define new features, it
>> might make sense to use ad-hoc commands, because there is no way we can
>> foresee all the features people may want to add over time.
>> 
>> Peter
> 





More information about the Standards mailing list