Minutes
Opening
Agenda - accepted as posted
MikeE:
Notes that agenda is expanding and we really need to get some of the issues resolved
No comments or objections
Resolution: Minutes of 2010-10-26 accepted w/o
Action Items
id=2010-09-22-3 status=done owner="Anish" Produce concrete proposal for ASSEMBLY-241 based on directional resolution in F2F
minutes
id=2010-09-22-8 status=pending owner="EricJ" produce new proposal for ASSEMBLY-227
EricJ:
Please provide comments wrt acceptability of new proposal
id=2010-09-22-9 status=done owner="MikeE" Raise issue WRT cardinality of channels
Administrivia
From Microsoft (Michael Champion) thanking TC for work on Implementation Type & Test Suite Adaption doc's
MikeE:
Notes that MS is expection these to be approved as Committee Specification
JeffM:
How can these be linked to existing docs?
MikeE:
Point to existing implementation types (as example)
JeffM:
We could/should ask MS if they have used these docs (e.g. for C#) and if they would share with TC
Action: owner=JeffM Draft response to MS asking for implementation experience with docs
New Issues
MikeE:
Lot of discussion on E-mail list - not all messages linked in minutes
...Please concnetrate on OPENING issues - not proposals
...from internal discussion - notion of "reply" to event messages
...Producer can add "replyTo" to messages - either to producer or third party
...reply is not necessarliy required and producer has no control over consumer
AnishK:
Similar to "wiredByImpl" allow binding & policy but don't want to explicity wire
...but event's don't allow bindings so doesn't seem appropriate
...Also greatly complicates application of Policy
PeterN:
Adds/transforms model into message flow model - quite different from existing model
...may be useful at application level, but not appropriate in Assembly spec
...Maybe for version 1.3?
<Peter Niblett>
don't we have oneway operations precisely to handle less loosely defined message exchanges?
EricJ:
Some of our customers are using similar model so thinks it would be good to formalize in spec
...otherwise TIBCO will have to add extension to support this model
MartinC:
Use case OK, but can't understand how it fits in SCA
...requires producers/consumer are aware of each other and this is explicitly NOT part of the SCA model
MikeE:
Use case can already be supported by service/reference using one-way and callback
...callback can have any number of messages and can use many different protocols
...Just because JMS can support this doesn't mean SCA should
...Sort of imposes obligation on consumer to produce response and that is not part of SCA model
DannyV:
Want to "bind" "EPR" of subsequent producer - ie consumer who receives original "replyTo" message
<MartinC>
the two wsdl meps that are ignore are ignored for a reason, no one could define how it works
DannyV:
Also SCA ignores some WSDL message patterns
<MartinC>
eventing isnt really about message patterns its about a different style of application interactions
PeterN:
Echos MartinC's comment - WS-I explicitly profiled them out due to being unable to figure them out
<MartinC>
it would be nice to get to a motion;-)
...WSDL "replyTo" was introduced to allow pub/sub on a point-to-point message network
<Mike Edwards>
In services/references, sequences of one-way messages between 2 components, in both directions, organised via callbacks, are
very close to event processing EXCEPT that the parties involved can be closely controlled AND the requirement for a response
is defined by the actual presence of the callback interface
Note there's approx 10 minutes left
<Mike Edwards>
Thanks Eric ;-)
AnishK:
Note sure use case justifies
DannyV:
Question of who (producer/consumer) gets to define "replyTo"
...in pub/sub subscribers can really be allowed to define "return address"
DannyV:
Gives purchasing example illustration need for producer to set "replyTo"
Chair notes that there's been a lot of discussion and would like to see a motion
Motion: m=EricJ s=DannyV Open ASSEMBLY-249
AnishK:
If understood you want to be able to dynamically state where messages are sent
...this makes application of policy really difficult
<MartinC>
especially from an assemblers point of view
MartinC:
Issues not adequately described in ASSEMBLY-249 - needs rewriting
PeterN:
+1 to Martin - not explicit enough
EricJ:
Would be reticent to rewrite without help from others or will just end up going round in circles
Chair notes virtually out of time - motion allowed to expire
MikeE:
Encourages collaboration to refine ASSEMBLY-249
AOB
JeffM, GrahamC, DannyV, ScottV
MartinC:
Probably will not be able to make next week
MikeE:
Also at a conference but should be able to run call
...don't want to cancel unless absolutely necessary
Meeting adjourned 9:03AM PDT
<Mike Edwards>
Many thanks Eric - that was not an easy meeting to scribe
Schreiber diagnostics output
[Delete this section before publishing the minutes]
final validation: Date not specified, the date '2010-11-02' was assumed
final validation: Title not specified, default title 'Oasis SCA-Assembly Teleconference...' was assumed
final validation: Chair not specified, default chair was assumed
statistics: Schreiber found 85 input lines
edits: Schreiber found the following text-edit commands:
edits: Line 33: s/From/... from/
command-scribe: Line 2: Scribe 'Eric Wells' is recognized by use of the nick 'EricW'
command-scribe: Line 2: EricW's nick 'EricW' has been selected
citation-detection-scribed: Line 19: Check for possible unrecognized nick 'a) Public Review of Test Suite Adaptation and Implementation
Type Documentation Requirements for SCA Assembly Model v1.1
- ended 30 October 2010
One comment received'
citation-detection-scribed: Line 30: Check for possible unrecognized nick 'ASSEMBLY-249'
edit-substitute: command on line 33 succeeded, changed line 32 from 'From' to '... from'
edit-delete: Line 33 was deleted
citation-detection-scribed: Line 120: Check for possible unrecognized nick 'Meeting adjourned 9'
system: Transformer: SAXON 9.2.1.2
[End of Schreiber diagnostic output]