[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]
Subject: RE: [uddi-spec] Groups - uddi-spec-tc-prop023-keyedReferenceGroup-20040209.doc uploaded
I really can't see the problem as we are talking about two different
ranges in scenario 3 (i.e. I do appear to be on a different page)
In mathematical terms we are saying
1) inverse of (10 <= x <= 20)
2) not (10 <= x <= 20)
3) ( x < 10 ) OR ( 20 < x )
I can't really see anyone getting confused over 3
Two comments
a) Even if you really dislike 3, I can't see what (1) offers over (2),
so don't feel that the inverted attribute is needed for that alone.
b) (3) is still a valid query surely, even if you don't like its format
and want to offer an alternative. Unless we prohibit the use of OR with
ranges (in which case how would you say
(10 < x < 20) OR (40 < x < 50)
apart from
inverse of (10 <= x <= 20) AND inverse of (40 <= x <= 50)
Which looks even worse and still has the lowerBound of one range above
the upper bound of another.
Matthew
> -----Original Message-----
> From: Tom Bellwood [mailto:bellwood@us.ibm.com]
> Sent: 09 March 2004 04:32
> To: Rogers, Tony
> Cc: Matthew J. Dovey; uddi-spec@lists.oasis-open.org
> Subject: RE: [uddi-spec] Groups -
> uddi-spec-tc-prop023-keyedReferenceGroup-20040209.doc uploaded
>
> Yes, I believe I understood the intent, I just disagree that
> scenario 3 is a clear enough way to specify that. I do not
> feel that allowing a lowerEndpoint to be higher than an
> upperEndpoint make good sense. it's much more intuitive IMO
> to have lowerendpoint=10, upperEndpoint=20 and
> inversion=TRUE, which means: x is in the set of (x<10, x>20),
> which effectively creates two unibounded ranges. If we're
> still not on the same page please let me know.
>
> Thanks,
> Tom Bellwood Phone: (512) 838-9957 (external); TL: 678/9957
> (internal) IBM Digital Media Standards Project Office Lead
> Co-Chair, OASIS UDDI Specification TC STSM - Emerging
> Technologies IBM Corporation
>
>
> To: Tom Bellwood/Austin/IBM@IBMUS, "Matthew J. Dovey"
> <matthew.dovey@oucs.ox.ac.uk>
> cc: <uddi-spec@lists.oasis-open.org>
> Subject: RE: [uddi-spec] Groups -
> uddi-spec-tc-prop023-keyedReferenceGroup-20040209.doc uploaded
>
>
>
>
> Tom,
>
> scenario 3 has TWO ranges - one that runs up to 10 (no lower
> bound), and one that runs down to 20 (no upper bound),
> combined with OR.
> -----Original Message-----
> From: Tom Bellwood [mailto:bellwood@us.ibm.com]
> Sent: Mon 08-Mar-04 7:49
> To: Matthew J. Dovey
> Cc: uddi-spec@lists.oasis-open.org
> Subject: RE: [uddi-spec] Groups -
> uddi-spec-tc-prop023-keyedReferenceGroup-20040209.doc uploaded
>
>
>
> Hi Matthew,
>
> I agree that there is redundancy between the first two
> scenarios you list, but I'm not sure I like getting rid of it
> entirely if that means we leave things simpler to use. I can
> probably be convinced by strong opinions here though.
> Regarding scenario 3, ("Using a range with upperEndpoint=10
> and another range with lowerEndpoint=20 in a conjunctive
> OR"), this doesn't make sense to me. I define a
> lowerEndpoint to always be <= upperEndpoint, at least that
> was the intention, hence the need to allow inversion of the
> criteria (by some means), which would yield a criteria
> outside these bounds. Allowing the transposition of lower
> and upper bounds does not seem intuitive behavior to me, so
> I would submit that this scenario should be considered
> invalid. Do you agree?
>
> Thanks,
> Tom Bellwood Phone: (512) 838-9957 (external); TL: 678/9957 (internal)
> IBM Digital Media Standards Project Office Lead
> Co-Chair, OASIS UDDI Specification TC
> STSM - Emerging Technologies
> IBM Corporation
>
> To: "Matthew J. Dovey" <matthew.dovey@oucs.ox.ac.uk>, Tom
> Bellwood/Austin/IBM@IBMUS
> cc: <uddi-spec@lists.oasis-open.org>
> Subject: RE: [uddi-spec] Groups -
> uddi-spec-tc-prop023-keyedReferenceGroup-20040209.doc uploaded
>
>
>
> Its been pointed out that I didn't elaborate the point behind my
> question.
>
> Whilst I don't see any problem in alternative ways of expressing the
> same thing, there isn't any reason to add new ways unnecessarily. So
> whilst it is fine to have the latter two as alternates (since we use
> NotOperands elsewhere), I can't see the purpose of also
> having the first
> one. i.e. unless I've missed some subtlety, I can't see the
> need for the
> inverted attribute.
>
> Matthew
>
> > -----Original Message-----
> > From: Matthew J. Dovey
> > [mailto:matthew.dovey@computing-services.oxford.ac.uk
> <mailto:matthew.dovey@computing-services.oxford.ac.uk> ]
> > Sent: 26 February 2004 09:51
> > To: Tom Bellwood
> > Cc: uddi-spec@lists.oasis-open.org
> > Subject: RE: [uddi-spec] Groups -
> > uddi-spec-tc-prop023-keyedReferenceGroup-20040209.doc uploaded
> >
> > Reviewing this, can you clarify what the difference is between
> >
> > Using a range of lowerEndpoint=10 and upperEndpoint=20 with
> > the inverted attribute=true
> >
> > Using a range of lowerEndpoint=10
> > and upperEndpoint=20 with the inverted attribute=false using
> > find_tModelNotOperands
> >
> > Using a range with upperEndpoint=10
> > and another range with lowerEndpoint=20 in a conjunctive OR
> >
> > (Of course the inclusive/exclusive attribute would need to be set
> > accordingly)
> >
> > Matthew
> >
> > > -----Original Message-----
> > > From: Tom Bellwood [mailto:bellwood@us.ibm.com
> <mailto:bellwood@us.ibm.com> ]
> > > Sent: 11 February 2004 18:57
> > > To: Matthew J. Dovey
> > > Cc: uddi-spec@lists.oasis-open.org
> > > Subject: RE: [uddi-spec] Groups -
> > > uddi-spec-tc-prop023-keyedReferenceGroup-20040209.doc uploaded
> > >
> > > Hi Matthew,
> > >
> > > We've adopted your suggestion during today's FTF discussion.
> > > The next version posted will have that update.
> > >
> > >
> > > Thanks,
> > > Tom Bellwood Phone: (512) 838-9957 (external); TL: 678/9957
> > > (internal) IBM Digital Media Standards Project Office Lead
> > Co-Chair,
> > > OASIS UDDI Specification TC STSM - Emerging Technologies IBM
> > > Corporation
> > >
> > >
> > > To: Tom Bellwood/Austin/IBM@IBMUS
> > > cc: <uddi-spec@lists.oasis-open.org>
> > > Subject: RE: [uddi-spec] Groups -
> > > uddi-spec-tc-prop023-keyedReferenceGroup-20040209.doc uploaded
> > >
> > >
> > >
> > > Tom,
> > >
> > > Thanks for picking this up - I'm afraid my attention has
> > been diverted
> > > at present!
> > >
> > > "Note that a rangeMatch element must include either an
> > upperEndpoint,
> > > lowerEndpoint, or both" - do we need to say this. A range with no
> > > endpoints would match everything (an inverted range with no
> > endpoints
> > > would match nothing).
> > >
> > > Rather than have an upperEndpointOperator (and
> > lowerEndpointOperator),
> > > how about a boolean inclusive=true/false (or
> exclusive=true/false).
> > > That way we can have the same type (endpoint) for both
> > upper and lower
> > > endpoints.
> > >
> > > Matthew
> > >
> > >
> > >
> > > > -----Original Message-----
> > > > From: bellwood@us.ibm.com [mailto:bellwood@us.ibm.com
> <mailto:bellwood@us.ibm.com> ]
> > > > Sent: 10 February 2004 15:22
> > > > To: uddi-spec@lists.oasis-open.org
> > > > Subject: [uddi-spec] Groups -
> > > > uddi-spec-tc-prop023-keyedReferenceGroup-20040209.doc uploaded
> > > >
> > > > The document
> > > > uddi-spec-tc-prop023-keyedReferenceGroup-20040209.doc has been
> > > > submitted by Tom Bellwood (bellwood@us.ibm.com) to the
> OASIS UDDI
> > > > Specification TC document repository.
> > > >
> > > > Document Description:
> > > > I've been trying to post this doc for a couple days
> w/o success.
> > > > Sorry. It's a first crack at solving Req 23.
> > > > Comments please.
> > > >
> > > > Download Document:
> > > >
> http://www.oasis-open.org/apps/org/workgroup/uddi-spec/downloa
> <http://www.oasis-open.org/apps/org/workgroup/uddi-spec/downloa>
> > > d.php/5395/uddi-spec-tc-prop023-keyedReferenceGroup-20040209.doc
> > > >
> > > > View Document Details:
> > > >
> http://www.oasis-open.org/apps/org/workgroup/uddi-spec/documen
> <http://www.oasis-open.org/apps/org/workgroup/uddi-spec/documen>
> > > t.php?document_id=5395
> > > >
> > > >
> > > > PLEASE NOTE: If the above links do not work for you,
> your email
> > > > application may be breaking the link into two pieces.
> > > > You may be able to copy and paste the entire link address
> > into the
> > > > address field of your web browser.
> > > >
> > > >
> > > >
> > > > To unsubscribe from this mailing list (and be removed from the
> > > > roster of the OASIS TC), go to
> > > >
> http://www.oasis-open.org/apps/org/workgroup/uddi-spec/members
> <http://www.oasis-open.org/apps/org/workgroup/uddi-spec/members>
> > > /leave_workgroup.php.
> > > >
> > > >
> > > >
> > >
> > >
> >
> > To unsubscribe from this mailing list (and be removed from
> > the roster of the OASIS TC), go to
> >
> http://www.oasis-open.org/apps/org/workgroup/uddi-spec/members
> <http://www.oasis-open.org/apps/org/workgroup/uddi-spec/members>
> /leave_workgroup.php.
> >
> >
>
> To unsubscribe from this mailing list (and be removed from
> the roster of the OASIS TC), go to
> http://www.oasis-open.org/apps/org/workgroup/uddi-spec/members
> /leave_workgroup.php
> <http://www.oasis-open.org/apps/org/workgroup/uddi-spec/member
> s/leave_workgroup.php> .
>
>
>
>
[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]