Issue 15877: Issue on Alignment of next OCL version and references UML 2.4/ MOF 2.4 (ocl2-rtf) Source: Fujitsu (Mr. Tom Rutt, tom(at)coastin.com) Nature: Uncategorized Issue Severity: Summary: Nature of problem: Informal references to UML 1.4.1 and UML 1.5 are included as part ofexplanatory text in the OCL 2.2 spec which refers to UML 1.x to explain differences of this new version of OCL.. The ISO/IEC 10151 (UML 1.4.1) needs to be added as an informative reference, for use in these explanations. UML 1.4.1 needs to remain in force, because so many UML models in may standards throughout the world are specified using UML 1.x notation, which is not backwards compatible with the new notation in UML 2.x. Given the normative content of OCL 2.3 (after RTF completes) is aligned technically with UML 2.4 and MOF 2.4, its normative references should be updated before publication of the RTF output, so that the OMG spec cross references will remain appropriate.. The references, and their uses in the OCL 2.3spec, need to be updated to reflect these latest UML/MOF versions. In addition, the Output of the OCL 2.3 RTF should be labeled as OCL 2.4, to avoid clarify the technical alignment of OMG’s latest versions of UML and MOF. Proposed Changes: Change version in title to OCL 2.4. Change all self references in the text from “OCL version 2.2” to “this OMG Specification”. Change all references from UML 2.0 and MOF 2.0 to UML 2.4 and MOF 2.4. In Section 1 ­ Scope Clause: Change: “ This specification defines the Object Constraint Language (OCL), version 2.3. OCL version 2.3 is the version of OCL that is aligned with UML 2.3 and MOF 2.0. “ to “ This specification defines the Object Constraint Language (OCL), version 2.4. OCL version 2.4 is the version of OCL that is aligned with UML 2.4 and MOF 2.4. “ Section 3 ­ Normative References Change: “ 3 Normative References The following normative documents contain provisions which, through reference in this text, constitute provisions of this specification. For dated references, subsequent amendments to, or revisions of, any of these publications do not apply. • UML 2.0 Superstructure Specification • UML 2.0 Infrastructure Specification • MOF 2.0 Core Specification • UNICODE 5.1 Standard: http://www.unicode.org/versions/Unicode5.1.0/ « To : « 3 References 3.1 Normative References The following normative documents contain provisions which, through reference in this text, constitute provisions of this specification. For dated references, subsequent amendments to, or revisions of, any of these publications do not apply. • UML 2.4 Superstructure Specification <omg spec Ref URL> • UML 2.4 Infrastructure Specification <omg spec Ref URL> • MOF 2.4 Core Specification <omg spec Ref URL> • UNICODE 5.1 Standard: http://www.unicode.org/versions/Unicode5.1.0/ 3.2 Informative References The following specification is reference in explanatory text, which describes differences between this specification and the version of OCL included in the existing standard. Its provisions do not constitute provisions of this specification. • ISO/IEC 19501:2005 Information technology -- Open Distributed Processing -- Unified Modeling Language (UML) Version 1.4.2 , also <omg Spec Ref URL> “ Change all uses of the reference in the text From “ UML 1.x” or “UML 1.4.x” “ To: “ ISO/IEC 19501:2005 “ In Section 6.1 “Changes to Adopted OMG Specifications” Replace: “ This specification replaces the specification of OCL given in UML 1.4.1 and UML 1.5. “ With: “ This specification replaces the specification of OCL given in OCL 2.2. The version of OCL specified in ISO/IEC 19505:2005 is intended for use in models based on UML 1.4.1 and UML 1.5. However, use of the OCL specified by ISO/IEC 19505:2005 is not prescribed by this specification. Resolution: Revised Text: Actions taken: December 7, 2010: received issue Discussion: End of Annotations:===== uthentication-Results: cm-omr3 smtp.user=tom@coastenterprises.com; auth=pass (PLAIN) X-Authenticated-UID: tom@coastenterprises.com Date: Tue, 07 Dec 2010 16:28:59 -0500 From: Tom Rutt Reply-To: tom@coastin.com User-Agent: Mozilla/5.0 (Windows; U; Windows NT 5.1; en-US; rv:1.9.2.12) Gecko/20101027 Thunderbird/3.1.6 To: issues@omg.org, ocl2-rtf@omg.org CC: Juergen Boldt , "ab@omg.org" , Tom Rutt Subject: NEW Issue on Alignment of next OCL version and references UML 2.4 / MOF 2.4 I believe it is important to resolve this issue before the close of the OCL 2.3 RTF. Juergen,: can you give this Issue a number and assign to the OCL 2 RTF New ISSUE xxxxx Base document OCL 2.2 Title: Alignement of OCL version and References with UML 2.4 UML and MOF 2.4: Source: Tom Rutt (Fujitsu) Nature of problem: Informal references to UML 1.4.1 and UML 1.5 are included as part ofexplanatory text in the OCL 2.2 spec which refers to UML 1.x to explain differences of this new version of OCL.. The ISO/IEC 10151 (UML 1.4.1) needs to be added as an informative reference, for use in these explanations. UML 1.4.1 needs to remain in force, because so many UML models in may standards throughout the world are specified using UML 1.x notation, which is not backwards compatible with the new notation in UML 2.x. Given the normative content of OCL 2.3 (after RTF completes) is aligned technically with UML 2.4 and MOF 2.4, its normative references should be updated before publication of the RTF output, so that the OMG spec cross references will remain appropriate.. The references, and their uses in the OCL 2.3spec, need to be updated to reflect these latest UML/MOF versions. In addition, the Output of the OCL 2.3 RTF should be labeled as OCL 2.4, to avoid clarify the technical alignment of OMG.s latest versions of UML and MOF. Proposed Changes: Change version in title to OCL 2.4. Change all self references in the text from .OCL version 2.2. to .this OMG Specification.. Change all references from UML 2.0 and MOF 2.0 to UML 2.4 and MOF 2.4. In Section 1 ­ Scope Clause: Change: . This specification defines the Object Constraint Language (OCL), version 2.3. OCL version 2.3 is the version of OCL that is aligned with UML 2.3 and MOF 2.0. . to . This specification defines the Object Constraint Language (OCL), version 2.4. OCL version 2.4 is the version of OCL that is aligned with UML 2.4 and MOF 2.4. . Section 3 ­ Normative References Change: . 3 Normative References The following normative documents contain provisions which, through reference in this text, constitute provisions of this specification. For dated references, subsequent amendments to, or revisions of, any of these publications do not apply. . UML 2.0 Superstructure Specification . UML 2.0 Infrastructure Specification . MOF 2.0 Core Specification . UNICODE 5.1 Standard: http://www.unicode.org/versions/Unicode5.1.0/ « To : « 3 References 3.1 Normative References The following normative documents contain provisions which, through reference in this text, constitute provisions of this specification. For dated references, subsequent amendments to, or revisions of, any of these publications do not apply. . UML 2.4 Superstructure Specification . UML 2.4 Infrastructure Specification . MOF 2.4 Core Specification . UNICODE 5.1 Standard: http://www.unicode.org/versions/Unicode5.1.0/ 3.2 Informative References The following specification is reference in explanatory text, which describes differences between this specification and the version of OCL included in the existing standard. Its provisions do not constitute provisions of this specification. . ISO/IEC 19501:2005 Information technology -- Open Distributed Processing -- Unified Modeling Language (UML) Version 1.4.2 , also . Change all uses of the reference in the text From . UML 1.x. or .UML 1.4.x. . To: . ISO/IEC 19501:2005 . In Section 6.1 .Changes to Adopted OMG Specifications. Replace: . This specification replaces the specification of OCL given in UML 1.4.1 and UML 1.5. . With: . This specification replaces the specification of OCL given in OCL 2.2. The version of OCL specified in ISO/IEC 19505:2005 is intended for use in models based on UML 1.4.1 and UML 1.5. However, use of the OCL specified by ISO/IEC 19505:2005 is not prescribed by this specification. . -- ---------------------------------------------------- Tom Rutt email: tom@coastin.com; trutt@us.fujitsu.com Tel: +1 732 801 5744 Fax: +1 732 774 5133 From: "Willink, Ed" To: "'tom@coastin.com'" , issues@omg.org, ocl2-rtf@omg.org Cc: Juergen Boldt , ab@omg.org Subject: RE: NEW Issue on Alignment of next OCL version and references UML 2.4 / MOF 2.4 Date: Wed, 8 Dec 2010 14:55:52 -0000 X-Mailer: Internet Mail Service (5.5.2657.72) Hi Tom I'm afraid I have to diagree quite strongly with the idea that 10-11-42 is OCL 2.4 that is aligned with UML 2.4. a) UML 2.4 is/was not publicly available b) the UML 2.4 RTF report of 10-12-02 was not available when 10-11-42 was prepared so no consideration of the consequences of any UML 2.4 change has been made and so any claim that 10-11-42 is UML 2.4 aligned is unfounded. c) AFAIK no RTF member has read any UML 2.x specification with a view to achieving alignment. d) The RTF is aware of a significant number of areas of misalignment with UML 2.0 so 10-11-42 is probably worse aligned. In 10-11-42 we have addressed a number of areas of under-specification and conflict. A new, perhaps private, issue is required so that the next issue can be the 'no-change' auto-generation, that will then allow us to move on the the voluminous textual anomalies. At this point we can be using UML 2.4/5 models and at least have syntactic consistency, and can start to review for semantic consistency. To take one specific example. Issue 14593 identifies 4 UML mis-alignments. The RTF has not had time to deal with any of them, so I'm not sure why a brand new UML alignment issue should get faster treatment. The critique of OCL's alignment could/should have been made for OCL 2.0, or perhaps OCL 2.2. The AB has now twice endorsed the unsatisfactory alignment and we are now making progress to resolve it. Changing to OCL 2.4 would create the misleading impression that significant progress had been made. The mish-mash of UML references accurately reflects the rigour of the specification. Please leave the version as 2.3 and we can endeavour to auto-generate an OCL 2.4 that has more than accidental correlation with UML 2.4. [Crunch test: Would I recommend a vendor to base an OCL product on 10-11-42 rather than say OCL 2.2? No. All OCL 2.x specifications are (pedantically) unimplementable. Any vendor must use their best endeavours to anticipate the issue resolutions that the RTF must make. A vendor with an OCL 2.x product should therefore review the RTF report and consider whether some changes are appropriate now. I doubt vendors will rush to support, for instance, oclLocale. They may choose to go for closure() rather sooner, and double check their null/invalid/OclVoid/OclInvalid behaviours.] Regards Ed Willink > -----Original Message----- > From: Tom Rutt [mailto:tom@coastin.com] > Sent: 07 December 2010 21:29 > To: issues@omg.org; ocl2-rtf@omg.org > Cc: Juergen Boldt; ab@omg.org; Tom Rutt > Subject: NEW Issue on Alignment of next OCL version and > references UML 2.4 / MOF 2.4 > > I believe it is important to resolve this issue before the > close of the > OCL 2.3 RTF. > > Juergen,: can you give this Issue a number and assign to the OCL 2 RTF > > > > New ISSUE xxxxx > Base document OCL 2.2 > Title: Alignement of OCL version and References with UML > 2.4 UML and MOF 2.4: > Source: Tom Rutt (Fujitsu) > > Nature of problem: > > Informal references to UML 1.4.1 and UML 1.5 are included > as part ofexplanatory text in the OCL 2.2 spec which refers > to UML 1.x to explain differences of this new version of > OCL.. The ISO/IEC 10151 (UML 1.4.1) needs to be added as > an informative reference, for use in these explanations. > > UML 1.4.1 needs to remain in force, because so many UML > models in may standards throughout the world are specified > using UML 1.x notation, which is not backwards compatible > with the new notation in UML 2.x. > > Given the normative content of OCL 2.3 (after RTF > completes) is aligned technically with UML 2.4 and MOF 2.4, > its normative references should be updated before > publication of the RTF output, so that the OMG spec cross > references will remain appropriate.. > > The references, and their uses in the OCL 2.3spec, need to > be updated to reflect these latest UML/MOF versions. > > In addition, the Output of the OCL 2.3 RTF should be > labeled as OCL 2.4, to avoid clarify the technical > alignment of OMG's latest versions of UML and MOF. > > Proposed Changes: > > Change version in title to OCL 2.4. > > Change all self references in the text from "OCL version > 2.2" to "this OMG Specification". > > Change all references from UML 2.0 and MOF 2.0 to UML 2.4 > and MOF 2.4. > > In Section 1 - Scope Clause: > > Change: > " > This specification defines the Object Constraint Language > (OCL), version 2.3. OCL version 2.3 is the version of OCL > that is aligned with UML 2.3 and MOF 2.0. > " > to > " > This specification defines the Object Constraint Language > (OCL), version 2.4. OCL version 2.4 is the version of OCL > that is aligned with UML 2.4 and MOF 2.4. > " > > Section 3 - Normative References > > Change: > " > 3 Normative References > The following normative documents contain provisions which, > through reference in this text, constitute provisions of > this specification. For dated references, subsequent > amendments to, or revisions of, any of these publications > do not apply. > * UML 2.0 Superstructure Specification > * UML 2.0 Infrastructure Specification > * MOF 2.0 Core Specification > * UNICODE 5.1 Standard: > http://www.unicode.org/versions/Unicode5.1.0/ > < > To : > < > 3 References > 3.1 Normative References > The following normative documents contain provisions which, > through reference in this text, constitute provisions of > this specification. For dated references, subsequent > amendments to, or revisions of, any of these publications > do not apply. > * UML 2.4 Superstructure Specification > * UML 2.4 Infrastructure Specification > * MOF 2.4 Core Specification > * UNICODE 5.1 Standard: > http://www.unicode.org/versions/Unicode5.1.0/ > 3.2 Informative References > The following specification is reference in explanatory > text, which describes differences between this > specification and the version of OCL included in the > existing standard. Its provisions do not constitute > provisions of this specification. > * ISO/IEC 19501:2005 Information technology -- Open > Distributed Processing -- Unified Modeling Language (UML) > Version 1.4.2 , also > " > > Change all uses of the reference in the text > From > " > UML 1.x" or "UML 1.4.x" > " > To: > " > ISO/IEC 19501:2005 > " > > In Section 6.1 "Changes to Adopted OMG Specifications" > > Replace: > " > This specification replaces the specification of OCL given > in UML 1.4.1 and UML 1.5. > " > With: > " > This specification replaces the specification of OCL given > in OCL 2.2. > > The version of OCL specified in ISO/IEC 19505:2005 is > intended for use in models based on UML 1.4.1 and UML 1.5. > However, use of the OCL specified by ISO/IEC 19505:2005 is > not prescribed by this specification. > " > > > > > -- > ---------------------------------------------------- > Tom Rutt email: tom@coastin.com; trutt@us.fujitsu.com > Tel: +1 732 801 5744 Fax: +1 732 774 5133 > > Please consider the environment before printing a hard copy of this e-mail. The information contained in this e-mail is confidential. It is intended only for the stated addressee(s) and access to it by any other person is unauthorised. If you are not an addressee, you must not disclose, copy, circulate or in any other way use or rely on the information contained in this e-mail. Such unauthorised use may be unlawful. If you have received this e-mail in error, please inform us immediately on +44 (0)118 986 8601 and delete it and all copies from your system. Thales Research and Technology (UK) Limited. A company registered in England and Wales. Registered Office: 2 Dashwood Lang Road, The Bourne Business Park, Addlestone, Weybridge, Surrey KT15 2NX. Registered Number: 774298 Thales UK Limited. A company registered in England and Wales. Registered Office: 2 Dashwood Lang Road, The Bourne Business Park, Addlestone, Weybridge, Surrey KT15 2NX. Registered Number: 868273 From: "Brucker, Achim" To: "'Willink, Ed'" , "'tom@coastin.com'" , "'issues@omg.org'" , "'ocl2-rtf@omg.org'" CC: "'Juergen Boldt'" , "'ab@omg.org'" Date: Wed, 8 Dec 2010 17:01:50 +0100 Subject: RE: NEW Issue on Alignment of next OCL version and references UML 2.4 / MOF 2.4 Thread-Topic: NEW Issue on Alignment of next OCL version and references UML 2.4 / MOF 2.4 Thread-Index: AcuW6NBHVM80guOYSM6FU9wFiRYwHQAB+4xA Accept-Language: de-DE X-MS-Has-Attach: X-MS-TNEF-Correlator: acceptlanguage: de-DE X-MIME-Autoconverted: from quoted-printable to 8bit by amethyst.omg.org id oB8FftQU023921 Hi, I totally agree with Ed's statement and second the conclusion that "claiming" an alignment between 10-11-42 UML 2.4 (10-12-02) should be avoided. Kind regards, Achim > -----Original Message----- > From: Willink, Ed [mailto:Ed.Willink@thalesgroup.com] > Sent: Wednesday, December 08, 2010 3:56 PM > To: 'tom@coastin.com'; issues@omg.org; ocl2-rtf@omg.org > Cc: Juergen Boldt; ab@omg.org > Subject: RE: NEW Issue on Alignment of next OCL version and references > UML 2.4 / MOF 2.4 > > Hi Tom > > I'm afraid I have to diagree quite strongly with the idea that 10-11-42 > is > OCL 2.4 that is aligned with UML 2.4. > > a) UML 2.4 is/was not publicly available > b) the UML 2.4 RTF report of 10-12-02 was not available when 10-11-42 > was > prepared so no consideration of the consequences of any UML 2.4 change > has > been made and so any claim that 10-11-42 is UML 2.4 aligned is > unfounded. > c) AFAIK no RTF member has read any UML 2.x specification with a view > to > achieving alignment. > d) The RTF is aware of a significant number of areas of misalignment > with > UML 2.0 so 10-11-42 is probably worse aligned. > > In 10-11-42 we have addressed a number of areas of under-specification > and > conflict. A new, perhaps private, issue is required so that the next > issue > can be the 'no-change' auto-generation, that will then allow us to move > on > the the voluminous textual anomalies. At this point we can be using UML > 2.4/5 models and at least have syntactic consistency, and can start to > review for semantic consistency. > > To take one specific example. Issue 14593 identifies 4 UML mis- > alignments. > The RTF has not had time to deal with any of them, so I'm not sure why > a > brand new UML alignment issue should get faster treatment. > > The critique of OCL's alignment could/should have been made for OCL 2.0, > or > perhaps OCL 2.2. The AB has now twice endorsed the unsatisfactory > alignment > and we are now making progress to resolve it. Changing to OCL 2.4 would > create the misleading impression that significant progress had been > made. > The mish-mash of UML references accurately reflects the rigour of the > specification. Please leave the version as 2.3 and we can endeavour to > auto-generate an OCL 2.4 that has more than accidental correlation with > UML > 2.4. > > [Crunch test: Would I recommend a vendor to base an OCL product on 10- > 11-42 > rather than say OCL 2.2? No. All OCL 2.x specifications are > (pedantically) > unimplementable. Any vendor must use their best endeavours to > anticipate the > issue resolutions that the RTF must make. A vendor with an OCL 2.x > product > should therefore review the RTF report and consider whether some > changes are > appropriate now. I doubt vendors will rush to support, for instance, > oclLocale. They may choose to go for closure() rather sooner, and > double > check their null/invalid/OclVoid/OclInvalid behaviours.] > > Regards > > Ed Willink > > > > > -----Original Message----- > > From: Tom Rutt [mailto:tom@coastin.com] > > Sent: 07 December 2010 21:29 > > To: issues@omg.org; ocl2-rtf@omg.org > > Cc: Juergen Boldt; ab@omg.org; Tom Rutt > > Subject: NEW Issue on Alignment of next OCL version and > > references UML 2.4 / MOF 2.4 > > > > I believe it is important to resolve this issue before the > > close of the > > OCL 2.3 RTF. > > > > Juergen,: can you give this Issue a number and assign to the OCL 2 > RTF > > > > > > > > New ISSUE xxxxx > > Base document OCL 2.2 > > Title: Alignement of OCL version and References with UML > > 2.4 UML and MOF 2.4: > > Source: Tom Rutt (Fujitsu) > > > > Nature of problem: > > > > Informal references to UML 1.4.1 and UML 1.5 are included > > as part ofexplanatory text in the OCL 2.2 spec which refers > > to UML 1.x to explain differences of this new version of > > OCL.. The ISO/IEC 10151 (UML 1.4.1) needs to be added as > > an informative reference, for use in these explanations. > > > > UML 1.4.1 needs to remain in force, because so many UML > > models in may standards throughout the world are specified > > using UML 1.x notation, which is not backwards compatible > > with the new notation in UML 2.x. > > > > Given the normative content of OCL 2.3 (after RTF > > completes) is aligned technically with UML 2.4 and MOF 2.4, > > its normative references should be updated before > > publication of the RTF output, so that the OMG spec cross > > references will remain appropriate.. > > > > The references, and their uses in the OCL 2.3spec, need to > > be updated to reflect these latest UML/MOF versions. > > > > In addition, the Output of the OCL 2.3 RTF should be > > labeled as OCL 2.4, to avoid clarify the technical > > alignment of OMG's latest versions of UML and MOF. > > > > Proposed Changes: > > > > Change version in title to OCL 2.4. > > > > Change all self references in the text from "OCL version > > 2.2" to "this OMG Specification". > > > > Change all references from UML 2.0 and MOF 2.0 to UML 2.4 > > and MOF 2.4. > > > > In Section 1 - Scope Clause: > > > > Change: > > " > > This specification defines the Object Constraint Language > > (OCL), version 2.3. OCL version 2.3 is the version of OCL > > that is aligned with UML 2.3 and MOF 2.0. > > " > > to > > " > > This specification defines the Object Constraint Language > > (OCL), version 2.4. OCL version 2.4 is the version of OCL > > that is aligned with UML 2.4 and MOF 2.4. > > " > > > > Section 3 - Normative References > > > > Change: > > " > > 3 Normative References > > The following normative documents contain provisions which, > > through reference in this text, constitute provisions of > > this specification. For dated references, subsequent > > amendments to, or revisions of, any of these publications > > do not apply. > > * UML 2.0 Superstructure Specification > > * UML 2.0 Infrastructure Specification > > * MOF 2.0 Core Specification > > * UNICODE 5.1 Standard: > > http://www.unicode.org/versions/Unicode5.1.0/ > > < > > To : > > < > > 3 References > > 3.1 Normative References > > The following normative documents contain provisions which, > > through reference in this text, constitute provisions of > > this specification. For dated references, subsequent > > amendments to, or revisions of, any of these publications > > do not apply. > > * UML 2.4 Superstructure Specification > > * UML 2.4 Infrastructure Specification > > * MOF 2.4 Core Specification > > * UNICODE 5.1 Standard: > > http://www.unicode.org/versions/Unicode5.1.0/ > > 3.2 Informative References > > The following specification is reference in explanatory > > text, which describes differences between this > > specification and the version of OCL included in the > > existing standard. Its provisions do not constitute > > provisions of this specification. > > * ISO/IEC 19501:2005 Information technology -- Open > > Distributed Processing -- Unified Modeling Language (UML) > > Version 1.4.2 , also > > " > > > > Change all uses of the reference in the text > > From > > " > > UML 1.x" or "UML 1.4.x" > > " > > To: > > " > > ISO/IEC 19501:2005 > > " > > > > In Section 6.1 "Changes to Adopted OMG Specifications" > > > > Replace: > > " > > This specification replaces the specification of OCL given > > in UML 1.4.1 and UML 1.5. > > " > > With: > > " > > This specification replaces the specification of OCL given > > in OCL 2.2. > > > > The version of OCL specified in ISO/IEC 19505:2005 is > > intended for use in models based on UML 1.4.1 and UML 1.5. > > However, use of the OCL specified by ISO/IEC 19505:2005 is > > not prescribed by this specification. > > " > > > > > > > > > > -- > > ---------------------------------------------------- > > Tom Rutt email: tom@coastin.com; trutt@us.fujitsu.com > > Tel: +1 732 801 5744 Fax: +1 732 774 5133 > > > > > > Please consider the environment before printing a hard copy of this > e-mail. > > The information contained in this e-mail is confidential. It is > intended > only for the stated addressee(s) and access to it by any other person > is > unauthorised. If you are not an addressee, you must not disclose, copy, > circulate or in any other way use or rely on the information contained > in > this e-mail. Such unauthorised use may be unlawful. If you have > received > this e-mail in error, please inform us immediately on +44 (0)118 986 > 8601 > and delete it and all copies from your system. > > Thales Research and Technology (UK) Limited. A company registered in > England and Wales. Registered Office: 2 Dashwood Lang Road, The Bourne > Business Park, Addlestone, Weybridge, Surrey KT15 2NX. Registered > Number: > 774298 > > Thales UK Limited. A company registered in England and Wales. > Registered > Office: 2 Dashwood Lang Road, The Bourne Business Park, Addlestone, > Weybridge, Surrey KT15 2NX. Registered Number: 868273 Subject: RE: NEW Issue on Alignment of next OCL version and references UML 2.4 / MOF 2.4 Date: Wed, 8 Dec 2010 17:16:19 +0100 X-MS-Has-Attach: X-MS-TNEF-Correlator: Thread-Topic: NEW Issue on Alignment of next OCL version and references UML 2.4 / MOF 2.4 Thread-Index: AcuW6LM1+K3RfzraRG6M+OYAAV+JewAAuPBA From: To: , , Cc: , X-OriginalArrivalTime: 08 Dec 2010 16:16:18.0662 (UTC) FILETIME=[3EE53860:01CB96F3] X-MIME-Autoconverted: from quoted-printable to 8bit by amethyst.omg.org id oB8Fu7gb026459 Hi Ed, Thanks for providing your opinion. I think we have two distinct issues here: - We want the OCL spec to be "editorially consistent" regarding references - We want the OCL spec to have formal content that is 100% "syntactically and semantically correct" We know that the second objective will still take some time to be (partially) reached, even if auto-generation work is progressing. Regarding the first one, if possible it would be good to try to solve it in the context of the actual RTF. Indeed, we need to decide whether we want to editorially refer to UML 2.3 or UML 2.4. In favor of UML 2.3 - No examination of impact of UML 2.4 has been done in the RTF. In favor of UML 2.4 and MOF 2.4 - Overall number consistency. There is no MOF 2.3 specification. - No evidence that 10-11-42 will be significantly less compatible to UML 2.4 than UML 2.3 (UML 2.4 has not changed fundamentally UML structure). We will discuss this with AB tomorrow. By the way, one option we could have could be to extend deadline of the RTF for three months so that we have time to examine in deep the impact of UML 2.4. Best Regards, Mariano -----Message d'origine----- De : Willink, Ed [mailto:Ed.Willink@thalesgroup.com] EnvoyĂ© mercredi 8 dĂ©mbre 2010 15:56 Ă€: 'tom@coastin.com'; issues@omg.org; ocl2-rtf@omg.org Cc : Juergen Boldt; ab@omg.org Objet : RE: NEW Issue on Alignment of next OCL version and references UML 2.4 / MOF 2.4 Hi Tom I'm afraid I have to diagree quite strongly with the idea that 10-11-42 is OCL 2.4 that is aligned with UML 2.4. a) UML 2.4 is/was not publicly available b) the UML 2.4 RTF report of 10-12-02 was not available when 10-11-42 was prepared so no consideration of the consequences of any UML 2.4 change has been made and so any claim that 10-11-42 is UML 2.4 aligned is unfounded. c) AFAIK no RTF member has read any UML 2.x specification with a view to achieving alignment. d) The RTF is aware of a significant number of areas of misalignment with UML 2.0 so 10-11-42 is probably worse aligned. In 10-11-42 we have addressed a number of areas of under-specification and conflict. A new, perhaps private, issue is required so that the next issue can be the 'no-change' auto-generation, that will then allow us to move on the the voluminous textual anomalies. At this point we can be using UML 2.4/5 models and at least have syntactic consistency, and can start to review for semantic consistency. To take one specific example. Issue 14593 identifies 4 UML mis-alignments. The RTF has not had time to deal with any of them, so I'm not sure why a brand new UML alignment issue should get faster treatment. The critique of OCL's alignment could/should have been made for OCL 2.0, or perhaps OCL 2.2. The AB has now twice endorsed the unsatisfactory alignment and we are now making progress to resolve it. Changing to OCL 2.4 would create the misleading impression that significant progress had been made. The mish-mash of UML references accurately reflects the rigour of the specification. Please leave the version as 2.3 and we can endeavour to auto-generate an OCL 2.4 that has more than accidental correlation with UML 2.4. [Crunch test: Would I recommend a vendor to base an OCL product on 10-11-42 rather than say OCL 2.2? No. All OCL 2.x specifications are (pedantically) unimplementable. Any vendor must use their best endeavours to anticipate the issue resolutions that the RTF must make. A vendor with an OCL 2.x product should therefore review the RTF report and consider whether some changes are appropriate now. I doubt vendors will rush to support, for instance, oclLocale. They may choose to go for closure() rather sooner, and double check their null/invalid/OclVoid/OclInvalid behaviours.] Regards Ed Willink > -----Original Message----- > From: Tom Rutt [mailto:tom@coastin.com] > Sent: 07 December 2010 21:29 > To: issues@omg.org; ocl2-rtf@omg.org > Cc: Juergen Boldt; ab@omg.org; Tom Rutt > Subject: NEW Issue on Alignment of next OCL version and references UML > 2.4 / MOF 2.4 > > I believe it is important to resolve this issue before the close of > the OCL 2.3 RTF. > > Juergen,: can you give this Issue a number and assign to the OCL 2 RTF > > > > New ISSUE xxxxx > Base document OCL 2.2 > Title: Alignement of OCL version and References with UML > 2.4 UML and MOF 2.4: > Source: Tom Rutt (Fujitsu) > > Nature of problem: > > Informal references to UML 1.4.1 and UML 1.5 are included as part > ofexplanatory text in the OCL 2.2 spec which refers to UML 1.x to > explain differences of this new version of OCL.. The ISO/IEC 10151 > (UML 1.4.1) needs to be added as an informative reference, for use in > these explanations. > > UML 1.4.1 needs to remain in force, because so many UML models in may > standards throughout the world are specified using UML 1.x notation, > which is not backwards compatible with the new notation in UML 2.x. > > Given the normative content of OCL 2.3 (after RTF > completes) is aligned technically with UML 2.4 and MOF 2.4, its > normative references should be updated before publication of the RTF > output, so that the OMG spec cross references will remain > appropriate.. > > The references, and their uses in the OCL 2.3spec, need to be updated > to reflect these latest UML/MOF versions. > > In addition, the Output of the OCL 2.3 RTF should be labeled as OCL > 2.4, to avoid clarify the technical alignment of OMG's latest versions > of UML and MOF. > > Proposed Changes: > > Change version in title to OCL 2.4. > > Change all self references in the text from "OCL version 2.2" to "this > OMG Specification". > > Change all references from UML 2.0 and MOF 2.0 to UML 2.4 and MOF 2.4. > > In Section 1 - Scope Clause: > > Change: > " > This specification defines the Object Constraint Language (OCL), > version 2.3. OCL version 2.3 is the version of OCL that is aligned > with UML 2.3 and MOF 2.0. > " > to > " > This specification defines the Object Constraint Language (OCL), > version 2.4. OCL version 2.4 is the version of OCL that is aligned > with UML 2.4 and MOF 2.4. > " > > Section 3 - Normative References > > Change: > " > 3 Normative References > The following normative documents contain provisions which, through > reference in this text, constitute provisions of this specification. > For dated references, subsequent amendments to, or revisions of, any > of these publications do not apply. > * UML 2.0 Superstructure Specification > * UML 2.0 Infrastructure Specification > * MOF 2.0 Core Specification > * UNICODE 5.1 Standard: > http://www.unicode.org/versions/Unicode5.1.0/ > < > To : > < > 3 References > 3.1 Normative References > The following normative documents contain provisions which, through > reference in this text, constitute provisions of this specification. > For dated references, subsequent amendments to, or revisions of, any > of these publications do not apply. > * UML 2.4 Superstructure Specification > * UML 2.4 Infrastructure Specification > * MOF 2.4 Core Specification > * UNICODE 5.1 Standard: > http://www.unicode.org/versions/Unicode5.1.0/ > 3.2 Informative References > The following specification is reference in explanatory text, which > describes differences between this specification and the version of > OCL included in the existing standard. Its provisions do not > constitute provisions of this specification. > * ISO/IEC 19501:2005 Information technology -- Open Distributed > Processing -- Unified Modeling Language (UML) Version 1.4.2 , also > " > > Change all uses of the reference in the text From " > UML 1.x" or "UML 1.4.x" > " > To: > " > ISO/IEC 19501:2005 > " > > In Section 6.1 "Changes to Adopted OMG Specifications" > > Replace: > " > This specification replaces the specification of OCL given in UML > 1.4.1 and UML 1.5. > " > With: > " > This specification replaces the specification of OCL given in OCL 2.2. > > The version of OCL specified in ISO/IEC 19505:2005 is intended for use > in models based on UML 1.4.1 and UML 1.5. > However, use of the OCL specified by ISO/IEC 19505:2005 is not > prescribed by this specification. > " > > > > > -- > ---------------------------------------------------- > Tom Rutt email: tom@coastin.com; trutt@us.fujitsu.com > Tel: +1 732 801 5744 Fax: +1 732 774 5133 > > Please consider the environment before printing a hard copy of this e-mail. The information contained in this e-mail is confidential. It is intended only for the stated addressee(s) and access to it by any other person is unauthorised. If you are not an addressee, you must not disclose, copy, circulate or in any other way use or rely on the information contained in this e-mail. Such unauthorised use may be unlawful. If you have received this e-mail in error, please inform us immediately on +44 (0)118 986 8601 and delete it and all copies from your system. Thales Research and Technology (UK) Limited. A company registered in England and Wales. Registered Office: 2 Dashwood Lang Road, The Bourne Business Park, Addlestone, Weybridge, Surrey KT15 2NX. Registered Number: 774298 Thales UK Limited. A company registered in England and Wales. Registered Office: 2 Dashwood Lang Road, The Bourne Business Park, Addlestone, Weybridge, Surrey KT15 2NX. Registered Number: 868273 Authentication-Results: cm-omr1 smtp.user=tom@coastenterprises.com; auth=pass (PLAIN) X-Authenticated-UID: tom@coastenterprises.com Date: Wed, 08 Dec 2010 12:22:13 -0500 From: Tom Rutt Reply-To: tom@coastin.com User-Agent: Mozilla/5.0 (Windows; U; Windows NT 5.1; en-US; rv:1.9.2.12) Gecko/20101027 Thunderbird/3.1.6 To: "Willink, Ed" CC: ocl2-rtf@omg.org, ab@omg.org Subject: Re: NEW Issue on Alignment of next OCL version and references UML 2.4 / MOF 2.4 On 12/8/2010 9:55 AM, Willink, Ed wrote: Hi Tom I'm afraid I have to diagree quite strongly with the idea that 10-11-42 is OCL 2.4 that is aligned with UML 2.4. Thanks for the clarification. I guess I misunderstood the degree of alignment. I understand now that the OCL version lifecycle is currently independent of the UML version lifecycle. The alignment of UML 2.4, MOF 2.4, and XMI 2.4 was carefully orchestrated, and that is why it is a good idea to synchonize the versions of MOF and XMI with the version of UML. So I now agree that the version of OCL out of the RTF should remain as 2.3. However, I do not understand why OCL references UML 2.0, rather than the latest one for reference purposes. I still think it important that OCL 2.3 should change its UML reference to be UML 2.3, the version which was available at the time the work of the OCL RTF was being carried out. Tom a) UML 2.4 is/was not publicly available b) the UML 2.4 RTF report of 10-12-02 was not available when 10-11-42 was prepared so no consideration of the consequences of any UML 2.4 change has been made and so any claim that 10-11-42 is UML 2.4 aligned is unfounded. c) AFAIK no RTF member has read any UML 2.x specification with a view to achieving alignment. d) The RTF is aware of a significant number of areas of misalignment with UML 2.0 so 10-11-42 is probably worse aligned. In 10-11-42 we have addressed a number of areas of under-specification and conflict. A new, perhaps private, issue is required so that the next issue can be the 'no-change' auto-generation, that will then allow us to move on the the voluminous textual anomalies. At this point we can be using UML 2.4/5 models and at least have syntactic consistency, and can start to review for semantic consistency. To take one specific example. Issue 14593 identifies 4 UML mis-alignments. The RTF has not had time to deal with any of them, so I'm not sure why a brand new UML alignment issue should get faster treatment. The critique of OCL's alignment could/should have been made for OCL 2.0, or perhaps OCL 2.2. The AB has now twice endorsed the unsatisfactory alignment and we are now making progress to resolve it. Changing to OCL 2.4 would create the misleading impression that significant progress had been made. The mish-mash of UML references accurately reflects the rigour of the specification. Please leave the version as 2.3 and we can endeavour to auto-generate an OCL 2.4 that has more than accidental correlation with UML 2.4. [Crunch test: Would I recommend a vendor to base an OCL product on 10-11-42 rather than say OCL 2.2? No. All OCL 2.x specifications are (pedantically) unimplementable. Any vendor must use their best endeavours to anticipate the issue resolutions that the RTF must make. A vendor with an OCL 2.x product should therefore review the RTF report and consider whether some changes are appropriate now. I doubt vendors will rush to support, for instance, oclLocale. They may choose to go for closure() rather sooner, and double check their null/invalid/OclVoid/OclInvalid behaviours.] Regards Ed Willink -----Original Message----- From: Tom Rutt [mailto:tom@coastin.com] Sent: 07 December 2010 21:29 To: issues@omg.org; ocl2-rtf@omg.org Cc: Juergen Boldt; ab@omg.org; Tom Rutt Subject: NEW Issue on Alignment of next OCL version and references UML 2.4 / MOF 2.4 I believe it is important to resolve this issue before the close of the OCL 2.3 RTF. Juergen,: can you give this Issue a number and assign to the OCL 2 RTF New ISSUE xxxxx Base document OCL 2.2 Title: Alignement of OCL version and References with UML 2.4 UML and MOF 2.4: Source: Tom Rutt (Fujitsu) Nature of problem: Informal references to UML 1.4.1 and UML 1.5 are included as part ofexplanatory text in the OCL 2.2 spec which refers to UML 1.x to explain differences of this new version of OCL.. The ISO/IEC 10151 (UML 1.4.1) needs to be added as an informative reference, for use in these explanations. UML 1.4.1 needs to remain in force, because so many UML models in may standards throughout the world are specified using UML 1.x notation, which is not backwards compatible with the new notation in UML 2.x. Given the normative content of OCL 2.3 (after RTF completes) is aligned technically with UML 2.4 and MOF 2.4, its normative references should be updated before publication of the RTF output, so that the OMG spec cross references will remain appropriate.. The references, and their uses in the OCL 2.3spec, need to be updated to reflect these latest UML/MOF versions. In addition, the Output of the OCL 2.3 RTF should be labeled as OCL 2.4, to avoid clarify the technical alignment of OMG's latest versions of UML and MOF. Proposed Changes: Change version in title to OCL 2.4. Change all self references in the text from "OCL version 2.2" to "this OMG Specification". Change all references from UML 2.0 and MOF 2.0 to UML 2.4 and MOF 2.4. In Section 1 - Scope Clause: Change: " This specification defines the Object Constraint Language (OCL), version 2.3. OCL version 2.3 is the version of OCL that is aligned with UML 2.3 and MOF 2.0. " to " This specification defines the Object Constraint Language (OCL), version 2.4. OCL version 2.4 is the version of OCL that is aligned with UML 2.4 and MOF 2.4. " Section 3 - Normative References Change: " 3 Normative References The following normative documents contain provisions which, through reference in this text, constitute provisions of this specification. For dated references, subsequent amendments to, or revisions of, any of these publications do not apply. * UML 2.0 Superstructure Specification * UML 2.0 Infrastructure Specification * MOF 2.0 Core Specification * UNICODE 5.1 Standard: http://www.unicode.org/versions/Unicode5.1.0/ < To : < 3 References 3.1 Normative References The following normative documents contain provisions which, through reference in this text, constitute provisions of this specification. For dated references, subsequent amendments to, or revisions of, any of these publications do not apply. * UML 2.4 Superstructure Specification * UML 2.4 Infrastructure Specification * MOF 2.4 Core Specification * UNICODE 5.1 Standard: http://www.unicode.org/versions/Unicode5.1.0/ 3.2 Informative References The following specification is reference in explanatory text, which describes differences between this specification and the version of OCL included in the existing standard. Its provisions do not constitute provisions of this specification. * ISO/IEC 19501:2005 Information technology -- Open Distributed Processing -- Unified Modeling Language (UML) Version 1.4.2 , also " Change all uses of the reference in the text From " UML 1.x" or "UML 1.4.x" " To: " ISO/IEC 19501:2005 " In Section 6.1 "Changes to Adopted OMG Specifications" Replace: " This specification replaces the specification of OCL given in UML 1.4.1 and UML 1.5. " With: " This specification replaces the specification of OCL given in OCL 2.2. The version of OCL specified in ISO/IEC 19505:2005 is intended for use in models based on UML 1.4.1 and UML 1.5. However, use of the OCL specified by ISO/IEC 19505:2005 is not prescribed by this specification. " -- ---------------------------------------------------- Tom Rutt email: tom@coastin.com; trutt@us.fujitsu.com Tel: +1 732 801 5744 Fax: +1 732 774 5133 Please consider the environment before printing a hard copy of this e-mail. The information contained in this e-mail is confidential. It is intended only for the stated addressee(s) and access to it by any other person is unauthorised. If you are not an addressee, you must not disclose, copy, circulate or in any other way use or rely on the information contained in this e-mail. Such unauthorised use may be unlawful. If you have received this e-mail in error, please inform us immediately on +44 (0)118 986 8601 and delete it and all copies from your system. Thales Research and Technology (UK) Limited. A company registered in England and Wales. Registered Office: 2 Dashwood Lang Road, The Bourne Business Park, Addlestone, Weybridge, Surrey KT15 2NX. Registered Number: 774298 Thales UK Limited. A company registered in England and Wales. Registered Office: 2 Dashwood Lang Road, The Bourne Business Park, Addlestone, Weybridge, Surrey KT15 2NX. Registered Number: 868273 -- ---------------------------------------------------- Tom Rutt email: tom@coastin.com; trutt@us.fujitsu.com Tel: +1 732 801 5744 Fax: +1 732 774 5133 Authentication-Results: cm-omr1 smtp.user=tom@coastenterprises.com; auth=pass (PLAIN) X-Authenticated-UID: tom@coastenterprises.com Date: Wed, 08 Dec 2010 12:22:13 -0500 From: Tom Rutt Reply-To: tom@coastin.com User-Agent: Mozilla/5.0 (Windows; U; Windows NT 5.1; en-US; rv:1.9.2.12) Gecko/20101027 Thunderbird/3.1.6 To: "Willink, Ed" CC: ocl2-rtf@omg.org, ab@omg.org Subject: Re: NEW Issue on Alignment of next OCL version and references UML 2.4 / MOF 2.4 On 12/8/2010 9:55 AM, Willink, Ed wrote: Hi Tom I'm afraid I have to diagree quite strongly with the idea that 10-11-42 is OCL 2.4 that is aligned with UML 2.4. Thanks for the clarification. I guess I misunderstood the degree of alignment. I understand now that the OCL version lifecycle is currently independent of the UML version lifecycle. The alignment of UML 2.4, MOF 2.4, and XMI 2.4 was carefully orchestrated, and that is why it is a good idea to synchonize the versions of MOF and XMI with the version of UML. So I now agree that the version of OCL out of the RTF should remain as 2.3. However, I do not understand why OCL references UML 2.0, rather than the latest one for reference purposes. I still think it important that OCL 2.3 should change its UML reference to be UML 2.3, the version which was available at the time the work of the OCL RTF was being carried out. Tom a) UML 2.4 is/was not publicly available b) the UML 2.4 RTF report of 10-12-02 was not available when 10-11-42 was prepared so no consideration of the consequences of any UML 2.4 change has been made and so any claim that 10-11-42 is UML 2.4 aligned is unfounded. c) AFAIK no RTF member has read any UML 2.x specification with a view to achieving alignment. d) The RTF is aware of a significant number of areas of misalignment with UML 2.0 so 10-11-42 is probably worse aligned. In 10-11-42 we have addressed a number of areas of under-specification and conflict. A new, perhaps private, issue is required so that the next issue can be the 'no-change' auto-generation, that will then allow us to move on the the voluminous textual anomalies. At this point we can be using UML 2.4/5 models and at least have syntactic consistency, and can start to review for semantic consistency. To take one specific example. Issue 14593 identifies 4 UML mis-alignments. The RTF has not had time to deal with any of them, so I'm not sure why a brand new UML alignment issue should get faster treatment. The critique of OCL's alignment could/should have been made for OCL 2.0, or perhaps OCL 2.2. The AB has now twice endorsed the unsatisfactory alignment and we are now making progress to resolve it. Changing to OCL 2.4 would create the misleading impression that significant progress had been made. The mish-mash of UML references accurately reflects the rigour of the specification. Please leave the version as 2.3 and we can endeavour to auto-generate an OCL 2.4 that has more than accidental correlation with UML 2.4. [Crunch test: Would I recommend a vendor to base an OCL product on 10-11-42 rather than say OCL 2.2? No. All OCL 2.x specifications are (pedantically) unimplementable. Any vendor must use their best endeavours to anticipate the issue resolutions that the RTF must make. A vendor with an OCL 2.x product should therefore review the RTF report and consider whether some changes are appropriate now. I doubt vendors will rush to support, for instance, oclLocale. They may choose to go for closure() rather sooner, and double check their null/invalid/OclVoid/OclInvalid behaviours.] Regards Ed Willink -----Original Message----- From: Tom Rutt [mailto:tom@coastin.com] Sent: 07 December 2010 21:29 To: issues@omg.org; ocl2-rtf@omg.org Cc: Juergen Boldt; ab@omg.org; Tom Rutt Subject: NEW Issue on Alignment of next OCL version and references UML 2.4 / MOF 2.4 I believe it is important to resolve this issue before the close of the OCL 2.3 RTF. Juergen,: can you give this Issue a number and assign to the OCL 2 RTF New ISSUE xxxxx Base document OCL 2.2 Title: Alignement of OCL version and References with UML 2.4 UML and MOF 2.4: Source: Tom Rutt (Fujitsu) Nature of problem: Informal references to UML 1.4.1 and UML 1.5 are included as part ofexplanatory text in the OCL 2.2 spec which refers to UML 1.x to explain differences of this new version of OCL.. The ISO/IEC 10151 (UML 1.4.1) needs to be added as an informative reference, for use in these explanations. UML 1.4.1 needs to remain in force, because so many UML models in may standards throughout the world are specified using UML 1.x notation, which is not backwards compatible with the new notation in UML 2.x. Given the normative content of OCL 2.3 (after RTF completes) is aligned technically with UML 2.4 and MOF 2.4, its normative references should be updated before publication of the RTF output, so that the OMG spec cross references will remain appropriate.. The references, and their uses in the OCL 2.3spec, need to be updated to reflect these latest UML/MOF versions. In addition, the Output of the OCL 2.3 RTF should be labeled as OCL 2.4, to avoid clarify the technical alignment of OMG's latest versions of UML and MOF. Proposed Changes: Change version in title to OCL 2.4. Change all self references in the text from "OCL version 2.2" to "this OMG Specification". Change all references from UML 2.0 and MOF 2.0 to UML 2.4 and MOF 2.4. In Section 1 - Scope Clause: Change: " This specification defines the Object Constraint Language (OCL), version 2.3. OCL version 2.3 is the version of OCL that is aligned with UML 2.3 and MOF 2.0. " to " This specification defines the Object Constraint Language (OCL), version 2.4. OCL version 2.4 is the version of OCL that is aligned with UML 2.4 and MOF 2.4. " Section 3 - Normative References Change: " 3 Normative References The following normative documents contain provisions which, through reference in this text, constitute provisions of this specification. For dated references, subsequent amendments to, or revisions of, any of these publications do not apply. * UML 2.0 Superstructure Specification * UML 2.0 Infrastructure Specification * MOF 2.0 Core Specification * UNICODE 5.1 Standard: http://www.unicode.org/versions/Unicode5.1.0/ < To : < 3 References 3.1 Normative References The following normative documents contain provisions which, through reference in this text, constitute provisions of this specification. For dated references, subsequent amendments to, or revisions of, any of these publications do not apply. * UML 2.4 Superstructure Specification * UML 2.4 Infrastructure Specification * MOF 2.4 Core Specification * UNICODE 5.1 Standard: http://www.unicode.org/versions/Unicode5.1.0/ 3.2 Informative References The following specification is reference in explanatory text, which describes differences between this specification and the version of OCL included in the existing standard. Its provisions do not constitute provisions of this specification. * ISO/IEC 19501:2005 Information technology -- Open Distributed Processing -- Unified Modeling Language (UML) Version 1.4.2 , also " Change all uses of the reference in the text From " UML 1.x" or "UML 1.4.x" " To: " ISO/IEC 19501:2005 " In Section 6.1 "Changes to Adopted OMG Specifications" Replace: " This specification replaces the specification of OCL given in UML 1.4.1 and UML 1.5. " With: " This specification replaces the specification of OCL given in OCL 2.2. The version of OCL specified in ISO/IEC 19505:2005 is intended for use in models based on UML 1.4.1 and UML 1.5. However, use of the OCL specified by ISO/IEC 19505:2005 is not prescribed by this specification. " -- ---------------------------------------------------- Tom Rutt email: tom@coastin.com; trutt@us.fujitsu.com Tel: +1 732 801 5744 Fax: +1 732 774 5133 Please consider the environment before printing a hard copy of this e-mail. The information contained in this e-mail is confidential. It is intended only for the stated addressee(s) and access to it by any other person is unauthorised. If you are not an addressee, you must not disclose, copy, circulate or in any other way use or rely on the information contained in this e-mail. Such unauthorised use may be unlawful. If you have received this e-mail in error, please inform us immediately on +44 (0)118 986 8601 and delete it and all copies from your system. Thales Research and Technology (UK) Limited. A company registered in England and Wales. Registered Office: 2 Dashwood Lang Road, The Bourne Business Park, Addlestone, Weybridge, Surrey KT15 2NX. Registered Number: 774298 Thales UK Limited. A company registered in England and Wales. Registered Office: 2 Dashwood Lang Road, The Bourne Business Park, Addlestone, Weybridge, Surrey KT15 2NX. Registered Number: 868273 -- ---------------------------------------------------- Tom Rutt email: tom@coastin.com; trutt@us.fujitsu.com Tel: +1 732 801 5744 Fax: +1 732 774 5133 X-IronPort-Anti-Spam-Filtered: true X-IronPort-Anti-Spam-Result: AvsEALNO/0zUnw4S/2dsb2JhbACjX3iuQJF2AoMKgj0EhTo Date: Wed, 08 Dec 2010 17:26:00 +0000 From: Ed Willink User-Agent: Mozilla/5.0 (Windows; U; Windows NT 6.0; en-GB; rv:1.9.2.12) Gecko/20101027 Thunderbird/3.1.6 To: mariano.belaunde@orange-ftgroup.com CC: Ed.Willink@thalesgroup.com, tom@coastin.com, ocl2-rtf@omg.org, juergen@omg.org, ab@omg.org Subject: Re: NEW Issue on Alignment of next OCL version and references UML 2.4 / MOF 2.4 Hi Mariano The trouble is we have no editorial consistency. I agree that it is likely that 10-11-42 is equally coherent when read with respect to UML 2.4 or UML 2.3, but that is only because no OCL 2.x is coherent with any UML 2.x. The references to UML 1.4 in Clause 12 should be preserved because they accurately reflect the way in which ExpressionInOcl and other concepts must be understood with respect to ValueSpecification. The current text shows the intent and allows a reader to eventaually derive a plausible re-interpreation. If we change the reference to UML2.x we confuse the reader by giving the impression that what is described is aligned with UML 2.x. I am not sure we are even talking about the correct details. Tom, originally wrote: In Section 6.1 "Changes to Adopted OMG Specifications" Replace: " This specification replaces the specification of OCL given in UML 1.4.1 and UML 1.5. With: " This specification replaces the specification of OCL given in OCL 2.2. The version of OCL specified in ISO/IEC 19505:2005 is intended for use in models based on UML 1.4.1 and UML 1.5. However, use of the OCL specified by ISO/IEC 19505:2005 is not prescribed by this specification. " But 10-11-42 has "This specification replaces the specification of OCL given in OCL 2.2." so Tom cannot have been reviewing 10-11-42. 10-11-42 is clearly flawed. All the suggested improvements seem to make it at best differently flawed and potentially more confusingly flawed. I see little point in tinkering with 10-11-42 when the time could be much better spent progressing towards OCL 2.4 (perhaps in 6 months) and 2.5 that may be fitting counterparts to UML 2.4 and 2.5. Regards Ed Willink On 08/12/2010 16:16, mariano.belaunde@orange-ftgroup.com wrote: Hi Ed, Thanks for providing your opinion. I think we have two distinct issues here: - We want the OCL spec to be "editorially consistent" regarding references - We want the OCL spec to have formal content that is 100% "syntactically and semantically correct" We know that the second objective will still take some time to be (partially) reached, even if auto-generation work is progressing. Regarding the first one, if possible it would be good to try to solve it in the context of the actual RTF. Indeed, we need to decide whether we want to editorially refer to UML 2.3 or UML 2.4. In favor of UML 2.3 - No examination of impact of UML 2.4 has been done in the RTF. In favor of UML 2.4 and MOF 2.4 - Overall number consistency. There is no MOF 2.3 specification. - No evidence that 10-11-42 will be significantly less compatible to UML 2.4 than UML 2.3 (UML 2.4 has not changed fundamentally UML structure). We will discuss this with AB tomorrow. By the way, one option we could have could be to extend deadline of the RTF for three months so that we have time to examine in deep the impact of UML 2.4. Best Regards, Mariano -----Message d'origine----- De : Willink, Ed [mailto:Ed.Willink@thalesgroup.com] EnvoyĂ© mercredi 8 dĂ©mbre 2010 15:56 Ă€: 'tom@coastin.com'; issues@omg.org; ocl2-rtf@omg.org Cc : Juergen Boldt; ab@omg.org Objet : RE: NEW Issue on Alignment of next OCL version and references UML 2.4 / MOF 2.4 Hi Tom I'm afraid I have to diagree quite strongly with the idea that 10-11-42 is OCL 2.4 that is aligned with UML 2.4. a) UML 2.4 is/was not publicly available b) the UML 2.4 RTF report of 10-12-02 was not available when 10-11-42 was prepared so no consideration of the consequences of any UML 2.4 change has been made and so any claim that 10-11-42 is UML 2.4 aligned is unfounded. c) AFAIK no RTF member has read any UML 2.x specification with a view to achieving alignment. d) The RTF is aware of a significant number of areas of misalignment with UML 2.0 so 10-11-42 is probably worse aligned. In 10-11-42 we have addressed a number of areas of under-specification and conflict. A new, perhaps private, issue is required so that the next issue can be the 'no-change' auto-generation, that will then allow us to move on the the voluminous textual anomalies. At this point we can be using UML 2.4/5 models and at least have syntactic consistency, and can start to review for semantic consistency. To take one specific example. Issue 14593 identifies 4 UML mis-alignments. The RTF has not had time to deal with any of them, so I'm not sure why a brand new UML alignment issue should get faster treatment. The critique of OCL's alignment could/should have been made for OCL 2.0, or perhaps OCL 2.2. The AB has now twice endorsed the unsatisfactory alignment and we are now making progress to resolve it. Changing to OCL 2.4 would create the misleading impression that significant progress had been made. The mish-mash of UML references accurately reflects the rigour of the specification. Please leave the version as 2.3 and we can endeavour to auto-generate an OCL 2.4 that has more than accidental correlation with UML 2.4. [Crunch test: Would I recommend a vendor to base an OCL product on 10-11-42 rather than say OCL 2.2? No. All OCL 2.x specifications are (pedantically) unimplementable. Any vendor must use their best endeavours to anticipate the issue resolutions that the RTF must make. A vendor with an OCL 2.x product should therefore review the RTF report and consider whether some changes are appropriate now. I doubt vendors will rush to support, for instance, oclLocale. They may choose to go for closure() rather sooner, and double check their null/invalid/OclVoid/OclInvalid behaviours.] Regards Ed Willink -----Original Message----- From: Tom Rutt [mailto:tom@coastin.com] Sent: 07 December 2010 21:29 To: issues@omg.org; ocl2-rtf@omg.org Cc: Juergen Boldt; ab@omg.org; Tom Rutt Subject: NEW Issue on Alignment of next OCL version and references UML 2.4 / MOF 2.4 I believe it is important to resolve this issue before the close of the OCL 2.3 RTF. Juergen,: can you give this Issue a number and assign to the OCL 2 RTF New ISSUE xxxxx Base document OCL 2.2 Title: Alignement of OCL version and References with UML 2.4 UML and MOF 2.4: Source: Tom Rutt (Fujitsu) Nature of problem: Informal references to UML 1.4.1 and UML 1.5 are included as part ofexplanatory text in the OCL 2.2 spec which refers to UML 1.x to explain differences of this new version of OCL.. The ISO/IEC 10151 (UML 1.4.1) needs to be added as an informative reference, for use in these explanations. UML 1.4.1 needs to remain in force, because so many UML models in may standards throughout the world are specified using UML 1.x notation, which is not backwards compatible with the new notation in UML 2.x. Given the normative content of OCL 2.3 (after RTF completes) is aligned technically with UML 2.4 and MOF 2.4, its normative references should be updated before publication of the RTF output, so that the OMG spec cross references will remain appropriate.. The references, and their uses in the OCL 2.3spec, need to be updated to reflect these latest UML/MOF versions. In addition, the Output of the OCL 2.3 RTF should be labeled as OCL 2.4, to avoid clarify the technical alignment of OMG's latest versions of UML and MOF. Proposed Changes: Change version in title to OCL 2.4. Change all self references in the text from "OCL version 2.2" to "this OMG Specification". Change all references from UML 2.0 and MOF 2.0 to UML 2.4 and MOF 2.4. In Section 1 - Scope Clause: Change: " This specification defines the Object Constraint Language (OCL), version 2.3. OCL version 2.3 is the version of OCL that is aligned with UML 2.3 and MOF 2.0. " to " This specification defines the Object Constraint Language (OCL), version 2.4. OCL version 2.4 is the version of OCL that is aligned with UML 2.4 and MOF 2.4. " Section 3 - Normative References Change: " 3 Normative References The following normative documents contain provisions which, through reference in this text, constitute provisions of this specification. For dated references, subsequent amendments to, or revisions of, any of these publications do not apply. * UML 2.0 Superstructure Specification * UML 2.0 Infrastructure Specification * MOF 2.0 Core Specification * UNICODE 5.1 Standard: http://www.unicode.org/versions/Unicode5.1.0/ < To : < 3 References 3.1 Normative References The following normative documents contain provisions which, through reference in this text, constitute provisions of this specification. For dated references, subsequent amendments to, or revisions of, any of these publications do not apply. * UML 2.4 Superstructure Specification * UML 2.4 Infrastructure Specification * MOF 2.4 Core Specification * UNICODE 5.1 Standard: http://www.unicode.org/versions/Unicode5.1.0/ 3.2 Informative References The following specification is reference in explanatory text, which describes differences between this specification and the version of OCL included in the existing standard. Its provisions do not constitute provisions of this specification. * ISO/IEC 19501:2005 Information technology -- Open Distributed Processing -- Unified Modeling Language (UML) Version 1.4.2 , also " Change all uses of the reference in the text From " UML 1.x" or "UML 1.4.x" " To: " ISO/IEC 19501:2005 " In Section 6.1 "Changes to Adopted OMG Specifications" Replace: " This specification replaces the specification of OCL given in UML 1.4.1 and UML 1.5. " With: " This specification replaces the specification of OCL given in OCL 2.2. The version of OCL specified in ISO/IEC 19505:2005 is intended for use in models based on UML 1.4.1 and UML 1.5. However, use of the OCL specified by ISO/IEC 19505:2005 is not prescribed by this specification. " -- ---------------------------------------------------- Tom Rutt email: tom@coastin.com; trutt@us.fujitsu.com Tel: +1 732 801 5744 Fax: +1 732 774 5133 Please consider the environment before printing a hard copy of this e-mail. The information contained in this e-mail is confidential. It is intended only for the stated addressee(s) and access to it by any other person is unauthorised. If you are not an addressee, you must not disclose, copy, circulate or in any other way use or rely on the information contained in this e-mail. Such unauthorised use may be unlawful. If you have received this e-mail in error, please inform us immediately on +44 (0)118 986 8601 and delete it and all copies from your system. Thales Research and Technology (UK) Limited. A company registered in England and Wales. Registered Office: 2 Dashwood Lang Road, The Bourne Business Park, Addlestone, Weybridge, Surrey KT15 2NX. Registered Number: 774298 Thales UK Limited. A company registered in England and Wales. Registered Office: 2 Dashwood Lang Road, The Bourne Business Park, Addlestone, Weybridge, Surrey KT15 2NX. Registered Number: 868273 ----- No virus found in this message. Checked by AVG - www.avg.com Version: 10.0.1170 / Virus Database: 426/3303 - Release Date: 12/07/10 Authentication-Results: cm-omr12 smtp.user=tom@coastenterprises.com; auth=pass (PLAIN) X-Authenticated-UID: tom@coastenterprises.com Date: Wed, 08 Dec 2010 12:37:52 -0500 From: Tom Rutt Reply-To: tom@coastin.com User-Agent: Mozilla/5.0 (Windows; U; Windows NT 5.1; en-US; rv:1.9.2.12) Gecko/20101027 Thunderbird/3.1.6 To: Ed Willink CC: mariano.belaunde@orange-ftgroup.com, Ed.Willink@thalesgroup.com, ocl2-rtf@omg.org, juergen@omg.org, ab@omg.org Subject: Re: NEW Issue on Alignment of next OCL version and references UML 2.4 / MOF 2.4 On 12/8/2010 12:26 PM, Ed Willink wrote: Clarification. My issue resolution is against the OCL 2.2, not the text in the output of the OCL 2.3 Task force. The replacement text is what is important. I did this issue because the UK national body has asked for a clarification on how the OCL we are progressing now is relating to the version of OCL in the existing UML 1.4 international standard. UML 2 does not replace UML 1.4, since the language is different, and not model instance compatible with UML 1.4. UK has asked whether the OCL in the UML 1.4 is being superceded for use in UML 1.4 models. So I think the paragraph needs to be changed somewhat along the lines of my suggestion. Tom Hi Mariano The trouble is we have no editorial consistency. I agree that it is likely that 10-11-42 is equally coherent when read with respect to UML 2.4 or UML 2.3, but that is only because no OCL 2.x is coherent with any UML 2.x. The references to UML 1.4 in Clause 12 should be preserved because they accurately reflect the way in which ExpressionInOcl and other concepts must be understood with respect to ValueSpecification. The current text shows the intent and allows a reader to eventaually derive a plausible re-interpreation. If we change the reference to UML2.x we confuse the reader by giving the impression that what is described is aligned with UML 2.x. I am not sure we are even talking about the correct details. Tom, originally wrote: In Section 6.1 "Changes to Adopted OMG Specifications" Replace: " This specification replaces the specification of OCL given in UML 1.4.1 and UML 1.5. With: " This specification replaces the specification of OCL given in OCL 2.2. The version of OCL specified in ISO/IEC 19505:2005 is intended for use in models based on UML 1.4.1 and UML 1.5. However, use of the OCL specified by ISO/IEC 19505:2005 is not prescribed by this specification. " But 10-11-42 has "This specification replaces the specification of OCL given in OCL 2.2." so Tom cannot have been reviewing 10-11-42. 10-11-42 is clearly flawed. All the suggested improvements seem to make it at best differently flawed and potentially more confusingly flawed. I see little point in tinkering with 10-11-42 when the time could be much better spent progressing towards OCL 2.4 (perhaps in 6 months) and 2.5 that may be fitting counterparts to UML 2.4 and 2.5. Regards Ed Willink On 08/12/2010 16:16, mariano.belaunde@orange-ftgroup.com wrote: Hi Ed, Thanks for providing your opinion. I think we have two distinct issues here: - We want the OCL spec to be "editorially consistent" regarding references - We want the OCL spec to have formal content that is 100% "syntactically and semantically correct" We know that the second objective will still take some time to be (partially) reached, even if auto-generation work is progressing. Regarding the first one, if possible it would be good to try to solve it in the context of the actual RTF. Indeed, we need to decide whether we want to editorially refer to UML 2.3 or UML 2.4. In favor of UML 2.3 - No examination of impact of UML 2.4 has been done in the RTF. In favor of UML 2.4 and MOF 2.4 - Overall number consistency. There is no MOF 2.3 specification. - No evidence that 10-11-42 will be significantly less compatible to UML 2.4 than UML 2.3 (UML 2.4 has not changed fundamentally UML structure). We will discuss this with AB tomorrow. By the way, one option we could have could be to extend deadline of the RTF for three months so that we have time to examine in deep the impact of UML 2.4. Best Regards, Mariano -----Message d'origine----- De : Willink, Ed [mailto:Ed.Willink@thalesgroup.com] EnvoyĂ© mercredi 8 dĂ©mbre 2010 15:56 Ă€: 'tom@coastin.com'; issues@omg.org; ocl2-rtf@omg.org Cc : Juergen Boldt; ab@omg.org Objet : RE: NEW Issue on Alignment of next OCL version and references UML 2.4 / MOF 2.4 Hi Tom I'm afraid I have to diagree quite strongly with the idea that 10-11-42 is OCL 2.4 that is aligned with UML 2.4. a) UML 2.4 is/was not publicly available b) the UML 2.4 RTF report of 10-12-02 was not available when 10-11-42 was prepared so no consideration of the consequences of any UML 2.4 change has been made and so any claim that 10-11-42 is UML 2.4 aligned is unfounded. c) AFAIK no RTF member has read any UML 2.x specification with a view to achieving alignment. d) The RTF is aware of a significant number of areas of misalignment with UML 2.0 so 10-11-42 is probably worse aligned. In 10-11-42 we have addressed a number of areas of under-specification and conflict. A new, perhaps private, issue is required so that the next issue can be the 'no-change' auto-generation, that will then allow us to move on the the voluminous textual anomalies. At this point we can be using UML 2.4/5 models and at least have syntactic consistency, and can start to review for semantic consistency. To take one specific example. Issue 14593 identifies 4 UML mis-alignments. The RTF has not had time to deal with any of them, so I'm not sure why a brand new UML alignment issue should get faster treatment. The critique of OCL's alignment could/should have been made for OCL 2.0, or perhaps OCL 2.2. The AB has now twice endorsed the unsatisfactory alignment and we are now making progress to resolve it. Changing to OCL 2.4 would create the misleading impression that significant progress had been made. The mish-mash of UML references accurately reflects the rigour of the specification. Please leave the version as 2.3 and we can endeavour to auto-generate an OCL 2.4 that has more than accidental correlation with UML 2.4. [Crunch test: Would I recommend a vendor to base an OCL product on 10-11-42 rather than say OCL 2.2? No. All OCL 2.x specifications are (pedantically) unimplementable. Any vendor must use their best endeavours to anticipate the issue resolutions that the RTF must make. A vendor with an OCL 2.x product should therefore review the RTF report and consider whether some changes are appropriate now. I doubt vendors will rush to support, for instance, oclLocale. They may choose to go for closure() rather sooner, and double check their null/invalid/OclVoid/OclInvalid behaviours.] Regards Ed Willink -----Original Message----- From: Tom Rutt [mailto:tom@coastin.com] Sent: 07 December 2010 21:29 To: issues@omg.org; ocl2-rtf@omg.org Cc: Juergen Boldt; ab@omg.org; Tom Rutt Subject: NEW Issue on Alignment of next OCL version and references UML 2.4 / MOF 2.4 I believe it is important to resolve this issue before the close of the OCL 2.3 RTF. Juergen,: can you give this Issue a number and assign to the OCL 2 RTF New ISSUE xxxxx Base document OCL 2.2 Title: Alignement of OCL version and References with UML 2.4 UML and MOF 2.4: Source: Tom Rutt (Fujitsu) Nature of problem: Informal references to UML 1.4.1 and UML 1.5 are included as part ofexplanatory text in the OCL 2.2 spec which refers to UML 1.x to explain differences of this new version of OCL.. The ISO/IEC 10151 (UML 1.4.1) needs to be added as an informative reference, for use in these explanations. UML 1.4.1 needs to remain in force, because so many UML models in may standards throughout the world are specified using UML 1.x notation, which is not backwards compatible with the new notation in UML 2.x. Given the normative content of OCL 2.3 (after RTF completes) is aligned technically with UML 2.4 and MOF 2.4, its normative references should be updated before publication of the RTF output, so that the OMG spec cross references will remain appropriate.. The references, and their uses in the OCL 2.3spec, need to be updated to reflect these latest UML/MOF versions. In addition, the Output of the OCL 2.3 RTF should be labeled as OCL 2.4, to avoid clarify the technical alignment of OMG's latest versions of UML and MOF. Proposed Changes: Change version in title to OCL 2.4. Change all self references in the text from "OCL version 2.2" to "this OMG Specification". Change all references from UML 2.0 and MOF 2.0 to UML 2.4 and MOF 2.4. In Section 1 - Scope Clause: Change: " This specification defines the Object Constraint Language (OCL), version 2.3. OCL version 2.3 is the version of OCL that is aligned with UML 2.3 and MOF 2.0. " to " This specification defines the Object Constraint Language (OCL), version 2.4. OCL version 2.4 is the version of OCL that is aligned with UML 2.4 and MOF 2.4. " Section 3 - Normative References Change: " 3 Normative References The following normative documents contain provisions which, through reference in this text, constitute provisions of this specification. For dated references, subsequent amendments to, or revisions of, any of these publications do not apply. * UML 2.0 Superstructure Specification * UML 2.0 Infrastructure Specification * MOF 2.0 Core Specification * UNICODE 5.1 Standard: http://www.unicode.org/versions/Unicode5.1.0/ < To : < 3 References 3.1 Normative References The following normative documents contain provisions which, through reference in this text, constitute provisions of this specification. For dated references, subsequent amendments to, or revisions of, any of these publications do not apply. * UML 2.4 Superstructure Specification * UML 2.4 Infrastructure Specification * MOF 2.4 Core Specification * UNICODE 5.1 Standard: http://www.unicode.org/versions/Unicode5.1.0/ 3.2 Informative References The following specification is reference in explanatory text, which describes differences between this specification and the version of OCL included in the existing standard. Its provisions do not constitute provisions of this specification. * ISO/IEC 19501:2005 Information technology -- Open Distributed Processing -- Unified Modeling Language (UML) Version 1.4.2 , also " Change all uses of the reference in the text From " UML 1.x" or "UML 1.4.x" " To: " ISO/IEC 19501:2005 " In Section 6.1 "Changes to Adopted OMG Specifications" Replace: " This specification replaces the specification of OCL given in UML 1.4.1 and UML 1.5. " With: " This specification replaces the specification of OCL given in OCL 2.2. The version of OCL specified in ISO/IEC 19505:2005 is intended for use in models based on UML 1.4.1 and UML 1.5. However, use of the OCL specified by ISO/IEC 19505:2005 is not prescribed by this specification. " -- ---------------------------------------------------- Tom Rutt email: tom@coastin.com; trutt@us.fujitsu.com Tel: +1 732 801 5744 Fax: +1 732 774 5133 Please consider the environment before printing a hard copy of this e-mail. The information contained in this e-mail is confidential. It is intended only for the stated addressee(s) and access to it by any other person is unauthorised. If you are not an addressee, you must not disclose, copy, circulate or in any other way use or rely on the information contained in this e-mail. Such unauthorised use may be unlawful. If you have received this e-mail in error, please inform us immediately on +44 (0)118 986 8601 and delete it and all copies from your system. Thales Research and Technology (UK) Limited. A company registered in England and Wales. Registered Office: 2 Dashwood Lang Road, The Bourne Business Park, Addlestone, Weybridge, Surrey KT15 2NX. Registered Number: 774298 Thales UK Limited. A company registered in England and Wales. Registered Office: 2 Dashwood Lang Road, The Bourne Business Park, Addlestone, Weybridge, Surrey KT15 2NX. Registered Number: 868273 ----- No virus found in this message. Checked by AVG - www.avg.com Version: 10.0.1170 / Virus Database: 426/3303 - Release Date: 12/07/10 -- ---------------------------------------------------- Tom Rutt email: tom@coastin.com; trutt@us.fujitsu.com Tel: +1 732 801 5744 Fax: +1 732 774 5133 X-IronPort-Anti-Spam-Filtered: true X-IronPort-Anti-Spam-Result: AlUFAEts/0zUnw4S/2dsb2JhbACVPo4keMAUhUkE Date: Wed, 08 Dec 2010 19:34:53 +0000 From: Ed Willink User-Agent: Mozilla/5.0 (Windows; U; Windows NT 6.0; en-GB; rv:1.9.2.12) Gecko/20101027 Thunderbird/3.1.6 To: tom@coastin.com CC: mariano.belaunde@orange-ftgroup.com, Ed.Willink@thalesgroup.com, ocl2-rtf@omg.org, juergen@omg.org, ab@omg.org Subject: Re: NEW Issue on Alignment of next OCL version and references UML 2.4 / MOF 2.4 Hi Tom Glad we're converging. So perhaps we could add: "The version of OCL specified in ISO/IEC 19505:2005 is intended for use in models based on UML 1.4.1 and UML 1.5. However, use of the OCL specified by ISO/IEC 19505:2005 is not prescribed by this specification. " But surely the import of the first sentence is already contained in ISO/IEC 19505:2005? I did not understand "prescribed" in the second sentence. After consulting a dictionary presumably it means "However, use of the OCL specified by ISO/IEC 19505:2005 is not required by this specification." which is surely self evident since 10-11-42 is one OCL specification and SO/IEC 19505:2005 is another. "UK has asked whether the OCL in the UML 1.4 is being superceded for use in UML 1.4 models." perhaps needs a sentence to the effect that "The version of OCL specified in this specification is not directly applicable to UML 1.4.1 or UML 1.5 models or to ISO/IEC 19505:2005 which is based on UML 1.4.1 and UML 1.5." However the 10-11-42 text in 6.2 . The OCL Language Description clause gives an informal description of OCL in the style that has been used in the UML versions 1.1 through 1.4. This clause is not normative, but meant to be explanatory. . Clause 8 (.Abstract Syntax.) describes the abstract syntax of OCL using a MOF 2.0 compliant metamodel. This is the same approach as used in the UML, v1.4 and other UML 2.0 specifications. The metamodel is MOF 2.0 compliant in the sense that it only uses constructs that are defined in the MOF 2.0. is dated in terms of the numeric references but only uses them for descriptive purposes. It is only the final sentence that is explicitly 2.x and therefore clearly not UML 1.x or SO/IEC 19505:2005, so I don't think we need a change. We can store up the clarifying sentence for the next version. Regards Ed Willink Date: Wed, 08 Dec 2010 17:19:31 -0500 From: Tom Rutt Reply-To: tom@coastin.com User-Agent: Mozilla/5.0 (Windows; U; Windows NT 5.1; en-US; rv:1.9.2.12) Gecko/20101027 Thunderbird/3.1.6 To: Ed Willink CC: mariano.belaunde@orange-ftgroup.com, Ed.Willink@thalesgroup.com, ocl2-rtf@omg.org, juergen@omg.org, ab@omg.org Subject: Re: NEW Issue on Alignment of next OCL version and references UML 2.4 / MOF 2.4 On 12/8/2010 2:34 PM, Ed Willink wrote: Hi Tom Glad we're converging. At this point, I agree that the clarifications below are not crucial, and can be put in during the next RTF cycle. There will probably be other changes resulting from the National Body comments on the ballot for 19507, and we can just bundle these changes into a next RTF output. However, I think we should still consider changing the UML ref to UML 2.3 now, since it was/is the latest spec available at the time of the OCL 2.3 RTF work. Is there any reason why the UML ref in OCL 2.3 should remain as UML 2.0?? Tom So perhaps we could add: "The version of OCL specified in ISO/IEC 19505:2005 is intended for use in models based on UML 1.4.1 and UML 1.5. However, use of the OCL specified by ISO/IEC 19505:2005 is not prescribed by this specification. " But surely the import of the first sentence is already contained in ISO/IEC 19505:2005? I did not understand "prescribed" in the second sentence. After consulting a dictionary presumably it means "However, use of the OCL specified by ISO/IEC 19505:2005 is not required by this specification." which is surely self evident since 10-11-42 is one OCL specification and SO/IEC 19505:2005 is another. "UK has asked whether the OCL in the UML 1.4 is being superceded for use in UML 1.4 models." perhaps needs a sentence to the effect that "The version of OCL specified in this specification is not directly applicable to UML 1.4.1 or UML 1.5 models or to ISO/IEC 19505:2005 which is based on UML 1.4.1 and UML 1.5." However the 10-11-42 text in 6.2 . The OCL Language Description clause gives an informal description of OCL in the style that has been used in the UML versions 1.1 through 1.4. This clause is not normative, but meant to be explanatory. . Clause 8 (.Abstract Syntax.) describes the abstract syntax of OCL using a MOF 2.0 compliant metamodel. This is the same approach as used in the UML, v1.4 and other UML 2.0 specifications. The metamodel is MOF 2.0 compliant in the sense that it only uses constructs that are defined in the MOF 2.0. is dated in terms of the numeric references but only uses them for descriptive purposes. It is only the final sentence that is explicitly 2.x and therefore clearly not UML 1.x or SO/IEC 19505:2005, so I don't think we need a change. We can store up the clarifying sentence for the next version. Regards Ed Willink -- ---------------------------------------------------- Tom Rutt email: tom@coastin.com; trutt@us.fujitsu.com Tel: +1 732 801 5744 Fax: +1 732 774 5133 Subject: RE: NEW Issue on Alignment of next OCL version and references UML 2.4 / MOF 2.4 Date: Thu, 9 Dec 2010 00:12:38 +0100 X-MS-Has-Attach: X-MS-TNEF-Correlator: Thread-Topic: NEW Issue on Alignment of next OCL version and references UML 2.4 / MOF 2.4 Thread-Index: AcuXJgB0YllYR5z4TIuGL2iSPx6JkAABs9zg From: To: , Cc: , , , X-OriginalArrivalTime: 08 Dec 2010 23:12:42.0961 (UTC) FILETIME=[6AB2B410:01CB972D] X-MIME-Autoconverted: from quoted-printable to 8bit by amethyst.omg.org id oB8MqX5i024732 >>> Is there any reason why the UML ref in OCL 2.3 should remain as UML 2.0?? No. There is no other reason than editorial: the sentence was included when UML 2.0 was the latest UML version. So, I guess this can be fixed editorially. -Mariano -----Message d'origine----- De : Tom Rutt [mailto:tom@coastin.com] EnvoyĂ© mercredi 8 dĂ©mbre 2010 23:20 Ă€: Ed Willink Cc : BELAUNDE Mariano RD-MAPS-LAN; Ed.Willink@thalesgroup.com; ocl2-rtf@omg.org; juergen@omg.org; ab@omg.org Objet : Re: NEW Issue on Alignment of next OCL version and references UML 2.4 / MOF 2.4 On 12/8/2010 2:34 PM, Ed Willink wrote: > Hi Tom > > Glad we're converging. At this point, I agree that the clarifications below are not crucial, and can be put in during the next RTF cycle. There will probably be other changes resulting from the National Body comments on the ballot for 19507, and we can just bundle these changes into a next RTF output. However, I think we should still consider changing the UML ref to UML 2.3 now, since it was/is the latest spec available at the time of the OCL 2.3 RTF work. Is there any reason why the UML ref in OCL 2.3 should remain as UML 2.0?? Tom > > So perhaps we could add: > > "The version of OCL specified in ISO/IEC 19505:2005 is intended for > use in models based on UML 1.4.1 and UML 1.5. > However, use of the OCL specified by ISO/IEC 19505:2005 is not > prescribed by this specification. > " > > But surely the import of the first sentence is already contained in > ISO/IEC 19505:2005? > > I did not understand "prescribed" in the second sentence. After > consulting a dictionary presumably it means > > "However, use of the OCL specified by ISO/IEC 19505:2005 is not > required by this specification." > > which is surely self evident since 10-11-42 is one OCL specification > and SO/IEC 19505:2005 is another. > > "UK has asked whether the OCL in the UML 1.4 is being superceded for > use in UML 1.4 models." > > perhaps needs a sentence to the effect that > > "The version of OCL specified in this specification is not directly > applicable to UML 1.4.1 or UML 1.5 models or to ISO/IEC 19505:2005 > which is based on UML 1.4.1 and UML 1.5." > > However the 10-11-42 text in 6.2 > > * The OCL Language Description clause gives an informal description of > OCL in the style that has been used in the UML versions 1.1 through > 1.4. This clause is not normative, but meant to be explanatory. > * Clause 8 ("Abstract Syntax") describes the abstract syntax of OCL > using a MOF 2.0 compliant metamodel. This is the same approach as used > in the UML, v1.4 and other UML 2.0 specifications. The metamodel is > MOF 2.0 compliant in the sense that it only uses constructs that are > defined in the MOF 2.0. > > is dated in terms of the numeric references but only uses them for > descriptive purposes. It is only the final sentence that is explicitly > 2.x and therefore clearly not UML 1.x or SO/IEC 19505:2005, so I don't > think we need a change. We can store up the clarifying sentence for > the next version. > > Regards > > Ed Willink > -- ---------------------------------------------------- Tom Rutt email: tom@coastin.com; trutt@us.fujitsu.com Tel: +1 732 801 5744 Fax: +1 732 774 5133 Date: Wed, 08 Dec 2010 18:31:57 -0500 From: Tom Rutt Reply-To: tom@coastin.com User-Agent: Mozilla/5.0 (Windows; U; Windows NT 5.1; en-US; rv:1.9.2.12) Gecko/20101027 Thunderbird/3.1.6 To: mariano.belaunde@orange-ftgroup.com CC: ed@willink.me.uk, Ed.Willink@thalesgroup.com, ocl2-rtf@omg.org, juergen@omg.org, ab@omg.org, Tom Rutt Subject: Re: NEW Issue on Alignment of next OCL version and references UML 2.4 / MOF 2.4 On 12/8/2010 6:12 PM, mariano.belaunde@orange-ftgroup.com wrote: Is there any reason why the UML ref in OCL 2.3 should remain as UML 2.0?? No. There is no other reason than editorial: the sentence was included when UML 2.0 was the latest UML version. So, I guess this can be fixed editorially. I talked to Andrew Watson, and he said that such a change would require an issue and vote. So I recommend now that we can just keep the UML ref to 2.0 in the OCL 2.3 RTF output. We should just keep this new issue in the database and assign it for resolution by the next RTF, which can introduce an appropriate UML reference change and also to add the clarifications regarding 19501:2005. Tom -Mariano -----Message d'origine----- De : Tom Rutt [mailto:tom@coastin.com] EnvoyĂ© mercredi 8 dĂ©mbre 2010 23:20 Ă€: Ed Willink Cc : BELAUNDE Mariano RD-MAPS-LAN; Ed.Willink@thalesgroup.com; ocl2-rtf@omg.org; juergen@omg.org; ab@omg.org Objet : Re: NEW Issue on Alignment of next OCL version and references UML 2.4 / MOF 2.4 On 12/8/2010 2:34 PM, Ed Willink wrote: Hi Tom Glad we're converging. At this point, I agree that the clarifications below are not crucial, and can be put in during the next RTF cycle. There will probably be other changes resulting from the National Body comments on the ballot for 19507, and we can just bundle these changes into a next RTF output. However, I think we should still consider changing the UML ref to UML 2.3 now, since it was/is the latest spec available at the time of the OCL 2.3 RTF work. Is there any reason why the UML ref in OCL 2.3 should remain as UML 2.0?? Tom So perhaps we could add: "The version of OCL specified in ISO/IEC 19505:2005 is intended for use in models based on UML 1.4.1 and UML 1.5. However, use of the OCL specified by ISO/IEC 19505:2005 is not prescribed by this specification. " But surely the import of the first sentence is already contained in ISO/IEC 19505:2005? I did not understand "prescribed" in the second sentence. After consulting a dictionary presumably it means "However, use of the OCL specified by ISO/IEC 19505:2005 is not required by this specification." which is surely self evident since 10-11-42 is one OCL specification and SO/IEC 19505:2005 is another. "UK has asked whether the OCL in the UML 1.4 is being superceded for use in UML 1.4 models." perhaps needs a sentence to the effect that "The version of OCL specified in this specification is not directly applicable to UML 1.4.1 or UML 1.5 models or to ISO/IEC 19505:2005 which is based on UML 1.4.1 and UML 1.5." However the 10-11-42 text in 6.2 * The OCL Language Description clause gives an informal description of OCL in the style that has been used in the UML versions 1.1 through 1.4. This clause is not normative, but meant to be explanatory. * Clause 8 ("Abstract Syntax") describes the abstract syntax of OCL using a MOF 2.0 compliant metamodel. This is the same approach as used in the UML, v1.4 and other UML 2.0 specifications. The metamodel is MOF 2.0 compliant in the sense that it only uses constructs that are defined in the MOF 2.0. is dated in terms of the numeric references but only uses them for descriptive purposes. It is only the final sentence that is explicitly 2.x and therefore clearly not UML 1.x or SO/IEC 19505:2005, so I don't think we need a change. We can store up the clarifying sentence for the next version. Regards Ed Willink -- ---------------------------------------------------- Tom Rutt email: tom@coastin.com; trutt@us.fujitsu.com Tel: +1 732 801 5744 Fax: +1 732 774 5133 -- ---------------------------------------------------- Tom Rutt email: tom@coastin.com; trutt@us.fujitsu.com Tel: +1 732 801 5744 Fax: +1 732 774 5133