Anish,
Thanks for your review - new version
of the proposal attached with comments inline below:
Yours, Mike.
Strategist - Emerging Technologies, SCA & SDO.
Co Chair OASIS SCA Assembly TC.
IBM Hursley Park, Mail Point 146, Winchester, SO21 2JN, Great Britain.
Phone & FAX: +44-1962-818014 Mobile: +44-7802-467431
Email:
From:
Anish Karmarkar <>
To:
Mike Edwards/UK/IBM@IBMGB
Cc:
OASIS Assembly <>
Date:
25/11/2008 07:12
Subject:
Re: [sca-assembly] [ISSUE 33] Updated
Proposal for Issue 33
Thanks for sending the updated version.
Three nits:
1) There is a lowercase 'must' in "... namely that the binding must
be
able to treat the transmission ...".
s/must/MUST/
<mje> I did not want
to place a normative requirement on the bindings here (I have no
idea how to verify such
a requirement from the Assembly perspective) so I changed
"must" to "needs
to". </mje>
2) For the same sentence, I prefer to talk about connections rather than
treatment of transmissions. I would also like to change the requirement
from 'must be able to' to 'it MUST'. I.e., it is not the capability that
we are requiring, which may or may not be used at runtime, but that it
must be so at runtime.
So the new suggested sentence would be:
"The requirements on a binding to support the asyncInvocation intent
are
the same as those required to support services with bidirectional
interfaces - namely that the binding MUST [some-tag] send the response
message in a separate connection from the one used by the request
message, with an arbitrarily large time interval between the two
transmissions.
<mje> I'm not convinced
that changing to "connection" helps that much. I was
particularly thinking of
the scheme that is used to deal with asynchrony over HTTP,
with a client polling the
server. That may or may not involve a new connection,
but it certainly involves
a separate transmission. </mje>
3) In the "Implementation Type support", I would like to replace
the
"asynchronous client-side and asynchronous server-side mappings"
with
"non-blocking client-side and non-blocking server-side mappings/APIs".
Talk of async/sync at the API-level when there is async/sync going on at
the transport-level leads to too much confusion. Especially, when
blocking APIs can be used with async transport and non-blocking APIs can
be used with sync transport.
<mje> Hmm. Tough
one, The words we use usually are ;-)
Client side I followed
the terminology of JAX-WS, which describes the methods as
"Asynchronous"
(see section 2.3.4 of the JAX-WS 2.0 spec), since the Java impl
follows this spec for these
methods. I'd prefer not to use different terminology which
people would then have
to translate for themselves.
As for server side, if
you take a look at the Java implementation document for this
proposal, it describes
both a blocking and a non blocking approach to the implementation
of a "long lived"
service method - so I don't think that "non-blocking" is appropriate
there.
Personally, I'm quite happy
with using the same terms at the API level and in the transport
layer.
</mje>
-Anish
--
Mike Edwards wrote:
>
> Folks,
>
> Here is an updated version of the proposal for Issue 33 following
the
> discussion on the call yesterday:
>
>
>
>
> This is the version that I am sending to the other TCs.
>
> Note that the current Assembly specification does not define the
> requirements that bidirectional interfaces place on bindings, so the
> text in this proposal is the first place that this is defined for
the
> Assembly specification.
>
>
> Yours, Mike.
>
> Strategist - Emerging Technologies, SCA & SDO.
> Co Chair OASIS SCA Assembly TC.
> IBM Hursley Park, Mail Point 146, Winchester, SO21 2JN, Great Britain.
> Phone & FAX: +44-1962-818014 Mobile: +44-7802-467431
> Email:
>
>
> ------------------------------------------------------------------------
>
> /
> /
>
> /Unless stated otherwise above:
> IBM United Kingdom Limited - Registered in England and Wales with
number
> 741598.
> Registered office: PO Box 41, North Harbour, Portsmouth, Hampshire
PO6 3AU/
>
>
>
>
>
>
>
> ------------------------------------------------------------------------
>
> ---------------------------------------------------------------------
> 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
Unless stated otherwise above:
IBM United Kingdom Limited - Registered in England and Wales with number
741598.
Registered office: PO Box 41, North Harbour, Portsmouth, Hampshire PO6
3AU
2008-11-12 - SCA-Assembly Issue 33 - Proposal_3.odt
2008-11-12 - SCA-Assembly Issue 33 - Proposal_3.doc
