[Council] JEP-0050: 4-Part vs 2-Part

Joe Hildebrand JHildebrand at jabber.com
Tue Apr 29 18:44:30 CDT 2003


For the simple case, it's still a 2-part process.
c:      iq-get; command/@node
s:      iq-result; command/@status='completed'

For the non-simple case, it's one more get, so that you can separate
intermediate results from the next form.

otherwise, you *actually* need this:

s: iq-result; x:data "form", x:data "result";command/@status='executing'

that's actually my main argument.  the "intention" argument is actully just
an interesting side-effect.

-- 
Joe Hildebrand


> -----Original Message-----
> From: Matthew A. Miller [mailto:linuxwolf at outer-planes.no-ip.com]
> Sent: Tuesday, April 29, 2003 4:43 PM
> To: council at jabber.org
> Cc: stpeter at jabber.org
> Subject: [Council] JEP-0050: 4-Part vs 2-Part
> 
> 
> It had been suggested that JEP-0050 would be "better behaved" by
> changing to a 4-part process for each stage of a command execution:
> 
> c:      iq-get; command/@node
> s:      iq-result; x:data "form";command/@status='executing'
> c:      iq-set; x:data "submit";command/@status='executing'
> s:      iq-result; x:data "result";command/@status='executing'
> 
> c:      iq-get; command/@node;command/@status='executing'
> s:      iq-result; x:data "form";command/@status='executing'
> c:      iq-set; x:data "submit";command/@status='executing'
> s:      iq-result; x:data "result";command/@status='executing'
> 
> ...
> 
> c:      iq-get; command/@node;command/@status='executing'
> s:      iq-result; x:data "form";command/@status='executing'
> c:      iq-set; x:data "submit";command/@status='executing'
> s:      iq-result; x:data "result";command/@status='completed'
> 
> as opposed to the currently documented 2-part process:
> 
> c:      iq-set; command/@node
> s:      iq-result; x:data "form";command/@status='executing'
> 
> c:      iq-set; x:data "submit";command/@status='executing'
> s:      iq-result; x:data "form";command/@status='executing'
> 
> ...
> 
> c:      iq-set; x:data "submit";command/@status='executing'
> s:      iq-result; x:data "result";command/@status='completed'
> 
> The primary reasoning that seems to have been used is that the latter
> "is not in the spirit of x:data", while the former is. 
> 
> In my opinion, the former overly complicates the entire process for no
> appreciable gain. The former implies the command could be complete by
> this stage, which is may or may not be true.  The former also 
> breaks up
> the flow into too fine of a granularity, adding additional layers and
> traffic.
> 
> Looking for user scenarios where the former is necessary, I 
> cannot find
> them.  In every user scenario I can come up with, the results are not
> available until the command is completed.
> 
> If a viable reason why the former process must be done, I'll acquiesce
> on this point.  Until then, I see no point to do this.
> 
> -- 
> 
> Matt "linuxwolf" Miller
> JID:	linuxwolf at outer-planes.net
> E-MAIL:	linuxwolf at outer-planes.net
> 
> - Got "JABBER"? (http://www.jabber.org/)
> 
> _______________________________________________
> Council mailing list
> Council at jabber.org
> http://mailman.jabber.org/listinfo/council
> 



More information about the Council mailing list