OASIS Mailing List ArchivesView the OASIS mailing list archive below
or browse/search using MarkMail.

 


Help: OASIS Mailing Lists Help | MarkMail Help

sca-bindings message

[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]


Subject: Re: [sca-bindings] Re: Groups - oasis - Ballot "SCA service wsdlElementpointing to a WSDL service" has closed


Anish Karmarkar wrote:
> I had similar concerns but since this was a strawpoll didn't ask for an 
> STV.
> 
> My further complain about OASIS KAVI is:
> The option 'other' didn't get counted.
> It truncates the responses so I can tell what people voted for (three of 
> the options start with 'We should allow this case'). Fortunately, only 
> one of those three got votes -- so by process of elimination I can 
> figure out what it was.
> Looks like SimonN voted for 'other' with an explanation. But I can't see 
> the explanation. So potentially, Simon could have said something that 
> changed peoples mind and altered the result.
> 
I didn't know about these problems when I cast my vote for "Other".
Actually "Other" was my top choice, though Kavi didn't allow me to
indicate this.  Here is the full text of my nomination for "Other":
  "We should allow this case and say the runtime SHOULD provide all
   of the matching ports (as in the issue 54 proposal)"

My reasons for allowing it are as follows:

1. It is consistent with allowing it for references, where it does
    have a clear use case.

2. In the case of a reference directly targeting a service+binding,
    not allowing on the service side would mean it can't be used by
    the reference either.  So if the service uses the alternative
    approach of defining a binding for each port, and the reference
    uses SCA wiring to connect to the service, this forces the
    reference to select (hard-wire) one of the available ports when
    it is being wired.  However, if the SCA reference is conecting to
    a non-SCA service using a binding on the reference, SCA would
    allow the specific port decision can be made at runtime.  This
    results in the extremely strange situation of SCA->SCA connections
    providing less flexibility than SCA->non-SCA connections.

3. The reference and service could both refer to a WSDL file from
    some shared registry/repository.  This WSDL file could include
    a WSDL service with multiple ports.  Allowing wsdl.service for
    SCA services would enable new ports to be added on both the
    reference and service side by updating the WSDL file in the
    registry, without the need to change any SCA configuration.
    Disallowing wsdl.service for SCA services means that both the
    external WSDL file and the SCA service configuration would
    need to be updated.

   Simon

> -Anish
> -- 
> 
> Eric Johnson wrote:
>> Anyone know who should be informed that the ballot mailer cannot count?
>>
>> workgroup_mailer@lists.oasis-open.org wrote:
>>> OASIS Service Component Architecture / Bindings (SCA-Bindings) TC 
>>> member,
>>>
>>> A ballot presented to OASIS Service Component Architecture / Bindings 
>>> (SCA-Bindings) TC has closed.
>>> The text of this closed ballot is as follows:
>>> ---
>>> "SCA service wsdlElement pointing to a WSDL service"
>>> Do you think we should define the behaviour of wsdlElement on an SCA 
>>> service pointing to a WSDL service, and if so, what should the 
>>> behaviour be?
>>>
>>> - We should disallow this case
>>> - We should allow this case but say nothing
>>> - We should allow this case and require the runtime to provide at 
>>> least one of the ports, possibly more
>>> - We should allow this case and require the runtime to provide all of 
>>> the matching ports
>>> - Other
>>>
>>> ---
>>>
>>> Quick Summary of Voting Results:  - We should disallow this case 
>>> received 9 Votes
>>>  - We should allow this case but say nothing received 0 Votes
>>>  - We should allow this case and require the runtime to provide at 
>>> least one of the ports, possibly more received 0 Votes
>>>  - We should allow this case and require the runtime to provide all 
>>> of the matching ports received 6 Votes
>>>
>>>  15 of 17 eligible voters cast their vote before the deadline.
>>>   
>> If you actually look at the ballot page, only ten people voted:
>> http://www.oasis-open.org/apps/org/workgroup/sca-bindings/ballot.php?id=1674 
>>
>>
>> Since we were allowed to vote for two options, that apparently
>> hopelessly confuses the balloting system.
>>
>> Of the ten people who voted, nine voted for "we should disallow this 
>> case."
>>
>> I'm not sure if the "may vote for two" option was intended as "instant
>> runoff", but if it were, then that suggests that people like myself who
>> only voted for one item "lost" a vote.  Arguably, if I only voted for
>> one option, does that mean that my vote should count twice?  If so, you
>> could tabulate it as:
>> 14 votes for "we should disallow".
>> 6 votes for "allow, but match all ports"
>>
>> Seems like another demonstration that voting systems are unavoidably 
>> biased.
>>
>> -Eric.
>>> Voting results for all closed ballots are available on the 
>>> sca-bindings eVote Archive at:
>>> http://www.oasis-open.org/apps/org/workgroup/sca-bindings/ballot_archive.php 
>>>
>>>
>>> Thank you,
>>> OASIS Open Administration
>>
>> ---------------------------------------------------------------------
>> To unsubscribe from this mail list, you must leave the OASIS TC that
>> generates this mail.  Follow this link to all your TCs in OASIS at:
>> https://www.oasis-open.org/apps/org/workgroup/portal/my_workgroups.php
> 
> ---------------------------------------------------------------------
> To unsubscribe from this mail list, you must leave the OASIS TC that
> generates this mail.  Follow this link to all your TCs in OASIS at:
> https://www.oasis-open.org/apps/org/workgroup/portal/my_workgroups.php
> 
> 




[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]