[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]
Subject: Re: [ws-tx] RE: Issue 118 - Add Conformance section to WS-Coordination, WS-AT and WS-BA specifications
hi, Just to add my $.02: While it is certainly a goal to make sure that various copies of "things" all say the same thing, as mary and others have pointed out, "stuff happens". The (new) OASIS rules REQUIRE that TC's designate which of multiple copies are considered to be the authoritative, in the event of a discrepancy. IMO you don't have a choice on whether you'd like to or not. What you do have a choice about is deciding which copy is authoritative. My suggestion is that the TC do that or the "spec" will be authoritative, by default. I'd say that's a poor choice, since (as has been pointed out) it is the external files that people will try to use, but it is up to the TC to decide. Sure you can try to use the errata process to fix things, and eventually inconsistencies will be ironed out. The problem is what happens in the meantime. (We won't get into the fact that the "errata" process is essentially useless, since it says only non- substantive changes are allowed. :-) I disagree that having precedence rules are arbitrary - to my mind that is a red herring. Let's say you only had one copy published. It obviously would be authoritative. Let's say there was an error in it. It is still the approved specification and everyone has to live with it. If it says to do A, and "someone" may have meant it to say B, you still have to do A. It is the responsibility of the TC to precisely and unambiguously identify the authoritative description of what MUST be implemented. People don't get to pick and choose what they want to implement (and claim conformance). cheers, jeff On Mar 20, 2008, at 10:13 AM, Ian Robinson wrote: > > Just to clarify. > The TX TC does not have copies of WSDL and Schema embedded in the > spec - we discussed and agreed previously that "multiple copies are > bad" so we really do not have this more general concern. Due > diligence is performed on all our spec materials - WSDL, schema and > normative text. > > There have been TC member arguments made in favour and against > adding a precedence statement. The argument for is essentially to > show that we have thought about it. The argument against is > essentially that an inconsistency should be fixed rather than > having an arbitrary precedence applied. I have asked the TC members > to be prepared for a TC ballot on this issue next week. > > Regards, > Ian Robinson > > > > "Mary McRae" <mary.mcrae@oasis-open.org> > Sent by: Mary McRae <marypmcrae@gmail.com> > 20/03/2008 16:39 > Please respond to > <mary.mcrae@oasis-open.org> > > To > "'Martin Chapman'" <martin.chapman@oracle.com>, "'Ram Jeyaraman'" > <Ram.Jeyaraman@microsoft.com>, "'Peter Furniss'" > <peter.furniss@irisfinancialsolutions.com>, Ian Robinson/UK/IBM@IBMGB > cc > <ws-tx@lists.oasis-open.org> > Subject > RE: [ws-tx] RE: Issue 118 - Add Conformance section to WS- > Coordination, WS-AT and WS-BA specifications > > > > > > ...pardon a small interruption. > > I'd like to agree with both sides - there *shouldn't* be any > inconsistencies > between the various copies. Which means that each TC should take > very careful > measures to ensure that element identifiers are spelled > consistently throughout > the spec (including case), that descriptions are consistent > throughout the spec > and correspond to the schema, etc. But the reality is that stuff > happens. And if > one person states that they've implemented the specification and > used the > element identifiers in the specification document, and another > person states > that they've implemented the specification and used the schema > file, they will > not interoperate. One of them is incorrect. By declaring one to > take precedence > over the other, you are in fact certifying that you have done due > diligence to > ensure that there are no inaccuracies in a particular copy, and > that everyone > has used that copy for their testing/implementation. It should be > noted that > it's impossible to test a schema that is included in a document > without first > extracting it into a standalone file. > > Multiple copies of anything are bad. They introduce errors that > oftentimes go > unnoticed. By telling implementers that they should use a > particular copy you > are letting them know that that was the particular copy that was > tested against > and that statements of use were issued against. > > But you're absolutely right that it could be the version that is > declared as > taking precedence that has the error. Depending on the scenario, an > errata may > not suffice. Errata is defined in the TC Process as: > "Errata" means a set of changes or proposed changes to a > specification that are > not Substantive Changes > > Following the thread, Substantive Change is defined as: > "Substantive Change" is a change to a specification that would > require a > compliant application or implementation to be modified or rewritten > in order to > remain compliant. > > Even the smallest of changes can result in a substantive change. > > ... now returning to your regularly scheduled programming. > > Mary > > > -----Original Message----- > > From: Martin Chapman [mailto:martin.chapman@oracle.com] > > Sent: Thursday, March 20, 2008 10:55 AM > > To: 'Ram Jeyaraman'; 'Peter Furniss'; 'Ian Robinson' > > Cc: ws-tx@lists.oasis-open.org > > Subject: RE: [ws-tx] RE: Issue 118 - Add Conformance section to WS- > > Coordination, WS-AT and WS-BA specifications > > > > I have to strongly dissagree about precedence/authoritative as it > really > > may affect interop, regardless of whether it's a genuine > > mistake or not. > > Reporting an inconsistency does not solve any interop issue. By > saying one > > copy (the external Files) is authoritative removes any > > ambuguity as to what vendors and users must conform to! > > > > Martin. > > > > >-----Original Message----- > > >From: Ram Jeyaraman [mailto:Ram.Jeyaraman@microsoft.com] > > >Sent: Monday, March 17, 2008 8:15 PM > > >To: Peter Furniss; Ian Robinson > > >Cc: Martin Chapman; ws-tx@lists.oasis-open.org > > >Subject: RE: [ws-tx] RE: Issue 118 - Add Conformance section > > >to WS-Coordination, WS-AT and WS-BA specifications > > > > > > > > >On conformance: > > > > > >I have suggested below some modifications to the conformance > > >text to indicate that a non-conformant implementation must not > > >use element and attributes of the TX namespace within a SOAP > Envelope: > > > > > >"Conformance > > > An implementation is not conformant with this specification > > >if it fails to satisfy one or more of the MUST or REQUIRED > > >level requirements defined herein. A SOAP Node MUST NOT use > > >elements and attributes of the declared XML Namespace (listed > > >on the title page) for this specification within SOAP > > >Envelopes unless it is conformant with this specification." > > > > > >I hope this addresses Peter's point about use of TX namespace > > >within a SOAP Envelope. > > > > > >On precedence: > > > > > >Since precedence rules really do not help address > > >inconsistencies, I suggest that we do not include any > > >statement about precedence rules. > > > > > >Thanks. > > > > > >-----Original Message----- > > >From: Peter Furniss [mailto:peter.furniss@ebusinessware.com] > > >On Behalf Of Peter Furniss > > >Sent: Wednesday, March 12, 2008 9:09 AM > > >To: Ian Robinson > > >Cc: Ram Jeyaraman; Martin Chapman; ws-tx@lists.oasis-open.org > > >Subject: Re: [ws-tx] RE: Issue 118 - Add Conformance section > > >to WS-Coordination, WS-AT and WS-BA specifications > > > > > >On "precedence" statement: > > > > > >Some spe > > >cifications that have taken the approach Ram proposes have > > >included a paragraph on the lines: > > > > > >"If inconsistency are found between any of the normative text > > >within this specification, the normative outlines, the XML Schema > > >*_[XML-Schema1]_* <#XMLSchema1> *_[XML-Schema2]_* > > ><#XMLSchema2>* *and the WSDL *_[WSDL]_* <#WSDL>* > > >*descriptions, the reader is requested to report this defect > > >to the OASIS WS-Tx TC (if it is still active) or to OASIS. No > > >general precedence rule is defined as the inconsistency, if > > >genuine, is a mistake." > > > > > >Such a paragraph is really only needed if other specifications > > >have A > B > C precedence statements, just to show the TC > > >thought about and gave this answer. Since other OASIS > > >specifications will be adding precedence statements, we > > >probably need to add this null one too. > > > > > >On the conformance statement itself: > > > > > >The sentence "A SOAP Node MUST NOT use the declared XML > > >Namespace (listed on the title page) for this specification > > >within SOAP Envelopes unless it is conformant with this > > >specification" would seem to be too broad ((Apologies if this > > >wording has been beaten up on already) . That would seem to > > >disallow, for example, an error reporting service to quote > > >erroneous soap headers - or even to send a fault containing > > >"namespace <ws-tx ns> is not recognised". Obviously that's not > > >our intention, but a lot depends on "use" if we are to claim > > >we aren't saying it. Does it work to have something like "A > > >SOAP Node whose implementation is not conformant with this > > >specification MUST NOT use the declared XML Namespace (listed > > >on the title page) for this specification to identify SOAP > > >Headers." I don't think that's quite right either (probably > > >over-restrictive). > > > > > > > > >Peter > > > > > > > > >Ian Robinson wrote: > > >> > > >> I agree with Ram's rationale that it is better to fix any > > >> inconsistencies that may be found rather than defining an > order of > > >> authority. I suggest the conformance text be simply as follows: > > >> > > >> "Conformance > > >> An implementation is not conformant with this specification if it > > >> fails to satisfy one or more of the MUST or REQUIRED level > > >> requirements defined herein. A SOAP Node MUST NOT use the > > >declared XML > > >> Namespace (listed on the title page) for this specification > within > > >> SOAP Envelopes unless it is conformant with this specification. > > >> > > >> > > >> Regards, > > >> Ian Robinson > > >> > > >> > > >> > > >> *Ram Jeyaraman <Ram.Jeyaraman@microsoft.com>* > > >> > > >> 11/03/2008 23:01 > > >> > > >> > > >> To > > >> Martin Chapman <martin.chapman@oracle.com>, Ian > > >> Robinson/UK/IBM@IBMGB cc > > >> "ws-tx@lists.oasis-open.org" <ws-tx@lists.oasis-open.org> > > >> Subject > > >> RE: [ws-tx] RE: Issue 118 - Add Conformance section to > > >> WS-Coordination, WS-AT and WS-BA specifications > > >> > > >> > > >> > > >> > > >> > > >> > > >> > > >> > > >> > > >> Ideally, there should not be any inconsistencies between the > > >> specification and its associated artifacts. > > >> > > >> In case of an inconsistency between the specification and one > of its > > >> artifacts, there are a few possibilities: 1. The specification is > > >> incorrect 2. The artifact is incorrect > > >> 3. Both are incorrect > > >> > > >> The obvious solution to this problem is to resolve the > inconsistency > > >> by publishing an errata in order to bring the specification > > >and/or the > > >> artifact back to a consistent state. > > >> > > >> Setting a precedence rule, such as the assigning the artifact > or the > > >> specification an overriding position, carries the risk of > > >perpetuating > > >> an error. For example, if we say the artifact has a precedence > over > > >> the specification, and the artifact turns out to be erroneous, > then > > >> unfortunately the precedence rule would make the erroneous > artifact > > >> correct by default. > > >> > > >> I observe that any overriding precedence hierarchy does not > actually > > >> correct the underlying cause of the problem and it carries > > >the danger > > >> of making an erroneous publication correct. Hence, I do not see a > > >> value in prescribing a precedence rule. Thanks. > > >> > > >> *From:* Martin Chapman [mailto:martin.chapman@oracle.com] * > > >> Sent:* Tuesday, March 11, 2008 11:08 AM* > > >> To:* 'Ian Robinson'* > > >> Cc:* Ram Jeyaraman; ws-tx@lists.oasis-open.org* > > >> Subject:* RE: [ws-tx] RE: Issue 118 - Add Conformance section to > > >> WS-Coordination, WS-AT and WS-BA specifications > > >> > > >> Yes we do need it as "I am not aware" doesn't mean to say > > >there aren't > > >> any! Typically these are editorial errors/typos, which do > happen. So > > >> lets just cover our backs in case. -----Original Message-----* > > >> From:* Ian Robinson [mailto:ian_robinson@uk.ibm.com] * > > >> Sent:* Tuesday, March 11, 2008 12:37 PM* > > >> To:* Martin Chapman* > > >> Cc:* 'Ram Jeyaraman'; ws-tx@lists.oasis-open.org* > > >> Subject:* RE: [ws-tx] RE: Issue 118 - Add Conformance section to > > >> WS-Coordination, WS-AT and WS-BA specifications > > >> > > >> I knew we'd talked about this in TX in the past but it took > > >me a while > > >> to dig out where. It was in the context of issue 26 a long > time ago. > > >> [1] At that time we were deciding whether to have integrated or > > >> standalone WSDL and schema but precedence was discussed. It > seems we > > >> never stated our decision in the specs but our decision at > that time > > >> was the following precedence (from highest to lowest): 1. > Normative > > >> text within the specification. 2. WSDL & schema > > >> 3. Outlines/snippets within the specification > > >> > > >> Having said the above, I'm not aware of our specification > materials > > >> having any ambiguity or contradiction that requires this > > >statement at > > >> all. Do we actually need it? > > >> > > >> [1] > > >> > > >http://www.oasis-open.org/committees/download.php/17464/WS- > TX_Minutes_ > > >> 2006_03_14-15.htm > > >> > > >> > > >> Regards, > > >> Ian Robinson > > >> > > >> *"Martin Chapman" <martin.chapman@oracle.com>* > > >> > > >> 10/03/2008 20:24 > > >> > > >> > > >> To > > >> "'Ram Jeyaraman'" <Ram.Jeyaraman@microsoft.com>, Ian > > >> Robinson/UK/IBM@IBMGB cc > > >> <ws-tx@lists.oasis-open.org>, <mary.mcrae@oasis-open.org> > > >> Subject > > >> RE: [ws-tx] RE: Issue 118 - Add Conformance section to > > >> WS-Coordination, WS-AT and WS-BA specifications > > >> > > >> > > >> > > >> > > >> > > >> > > >> > > >> > > >> > > >> > > >> > > >> Ram, > > >> > > >> Sorry just catching up on travel backlog. > > >> > > >> Mary and myself have just been talking about the precedence > > >issue for > > >> another TC. Considering the external files are the ones that > > >should be > > >> verified for correctness, and that they will most probably > > >be the ones > > >> downloaded and used in projects, I recommend we make the external > > >> files the authoritative ones (highest precedence). > > >> > > >> How about: > > >> "The XML Schema *_[XML-Schema1]_* <outbind://24/#XMLSchema1> > > >> *_[XML-Schema2]_* <outbind://24/#XMLSchema2>* *and WSDL *_ > [WSDL]_* > > >> <outbind://24/#WSDL>* *descriptions are authoritative and take > > >> precedence over Normative text within this specification, > which in > > >> turn take precedence over normative outlines ." > > >> > > >> Martin. > > >> > > >> -----Original Message-----* > > >> From:* Ram Jeyaraman [mailto:Ram.Jeyaraman@microsoft.com] * > > >> Sent:* Friday, March 07, 2008 8:18 PM* > > >> To:* Ian Robinson* > > >> Cc:* Martin Chapman; ws-tx@lists.oasis-open.org* > > >> Subject:* RE: [ws-tx] RE: Issue 118 - Add Conformance section to > > >> WS-Coordination, WS-AT and WS-BA specifications > > >> > > >> Thanks Ian, > > >> > > >> I have suggested text below that includes your change. > > >> > > >> In the text, I have replaced "take precedence over the XML Schema > > >> *_[XML-Schema1]_* <#XMLSchema1> *_[XML-Schema2]_* <#XMLSchema2>* > > >> *descriptions" with "take precedence over the XML Schema > > >> *_[XML-Schema1]_* <#XMLSchema1> *_[XML-Schema2]_* > > ><#XMLSchema2>* *and > > >> WSDL *_[WSDL]_* <#WSDL>* *descriptions". > > >> > > >> The insertion point for the conformance section seems fine. > > >> "Conformance An implementation is not conformant with this > > >> specification if it fails to satisfy one or more of the MUST or > > >> REQUIRED level requirements defined herein. A SOAP Node MUST > NOT use > > >> the declared XML Namespace for this specification (listed in > section > > >> 1.x) within SOAP Envelopes unless it is conformant with this > > >> specification. Normative text within this specification takes > > >> precedence over normative outlines, which in turn take > > >precedence over > > >> the XML Schema > > >> *_[XML-Schema1]_* <#XMLSchema1> *_[XML-Schema2]_* > > ><#XMLSchema2>* *and > > >> WSDL *_[WSDL]_* <#WSDL>* *descriptions." > > >> > > >> * > > >> From:* Ian Robinson [mailto:ian_robinson@uk.ibm.com] * > > >> Sent:* Friday, March 07, 2008 2:50 AM* > > >> To:* Ram Jeyaraman* > > >> Cc:* Martin Chapman; ws-tx@lists.oasis-open.org* > > >> Subject:* Re: [ws-tx] RE: Issue 118 - Add Conformance section to > > >> WS-Coordination, WS-AT and WS-BA specifications > > >> > > >> > > >> My comments on the proposed new section: > > >> > > >> * "the XML namespace identifier for this specification > (listed in > > >> section 1.4)" might be better as "the declared XML > > >Namespace for > > >> this specification" since this is part of the front page > > >> material for each spec. > > >> * The position of the new sections in each TX spec should be > > >> between "Protocol Elements" and the References section. > > >> > > >> > > >> Regards, > > >> Ian Robinson > > >> STSM, WebSphere Transactions Architect > > >> IBM Hursley Lab, UK > > >> > > >> *Ram Jeyaraman <Ram.Jeyaraman@microsoft.com>* > > >> > > >> 06/03/2008 18:43 > > >> > > >> > > >> > > >> > > >> To > > >> Martin Chapman <martin.chapman@oracle.com> > > >> cc > > >> "ws-tx@lists.oasis-open.org" <ws-tx@lists.oasis-open.org> > > >> Subject > > >> [ws-tx] RE: Issue 118 - Add Conformance section to > > >> WS-Coordination, WS-AT and WS-BA specifications > > >> > > >> > > >> > > >> > > >> > > >> > > >> > > >> > > >> > > >> > > >> > > >> > > >> Hi Martin, > > >> > > >> Do you see any modifications to the conformance text > > >proposed below in > > >> the case of WS-Coordination, WS-AT and WS-BA? > > >> > > >> Thank you. > > >> * > > >> From:* Ram Jeyaraman [mailto:Ram.Jeyaraman@microsoft.com] * > > >> Sent:* Thursday, March 06, 2008 9:51 AM* > > >> To:* ws-tx@lists.oasis-open.org* > > >> Subject:* [ws-tx] Issue 118 - Add Conformance section to > > >> WS-Coordination, WS-AT and WS-BA specifications > > >> > > >> This issue was raised by Martin Chapman (Oracle) during the > > >March 06, > > >> 2008 TX TC call. > > >> > > >> Description: > > >> > > >> The WS-Coordination, WS-AT and WS-BA specifications currently > do not > > >> have a conformance section. > > >> > > >> For example, RX specifications use the following conformance > text: > > >> > > >> "1.5 Conformance > > >> An implementation is not conformant with this specification if it > > >> fails to satisfy one or more of the MUST or REQUIRED level > > >> requirements defined herein. A SOAP Node MUST NOT use the XML > > >> namespace identifier for this specification (listed in section > 1.4) > > >> within SOAP Envelopes unless it is conformant with this > > >specification. > > >> Normative text within this specification takes precedence over > > >> normative outlines, which in turn take precedence over the > > >XML Schema > > >> [XML Schema Part 1, Part 2] descriptions." > > >> > > >> A similar conformance section should be added to the TX > > >> specifications. > > >> > > >> > > >> > > >> > > > >--------------------------------------------------------------------- > - > > >> -- > > >> > > >> > > >> > > >> > > > > > > > > > >--------------------------------------------------------------------- > > >To unsubscribe from this mail list, you must leave the OASIS > > >TC that generates this mail. You may a link to this group and > > >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. You may a link to this group and 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 > > > > > > -- Jeff Mischkinsky jeff.mischkinsky@oracle.com Director, Oracle Fusion Middleware and Web Services Standards +1(650) 506-1975 Consulting Member Technical Staff 500 Oracle Parkway, M/ S 2OP9 Oracle Redwood Shores, CA 94065
[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]