Issue 19335: UML 2.5 Overly strict restriction on message slope in seq diagrams (uml2-rtf) Source: Change Vision (Mr. Michael Jesse Chonoles, mjchonoles(at)yahoo.com) Nature: Uncategorized Issue Severity: Summary: (BTW, where should the issues found on UML 2.5 go, Is there a 2.6 mailing list?) In UML 2.5 Paragraph 17.4.4. Notation Message: “..A line must be such that every line fragment is either horizontal or downwards when traversed from send event to receive event.” While this restriction appears reasonable when first read, it is really overly strict. In the default, WEAK, interpretation of sequence diagrams, the time-wise ordering of occurrences within a lifeline is independent of occurrences on other lifelines, subject to cause/effect sequencing (message sending—>message reception). And, of course, the order of occurrences as depicted on the lifeline must be maintained. This is practically equivalent to saying that each lifeline has its own clock or timescale, and that ordering by that clock must be maintained and that causality across lifelines must be maintained. It is often thought that one could change the timescale (and not necessarily smoothly) on a lifeline without changing the interpretation or legality of the diagram. However, the restriction on messages not taking a non-negative slope, prevents otherwise legal changes in scale and makes invalid some diagrams that do not violate causality or the within-lifeline ordering. Please relax this restriction Resolution: Revised Text: Actions taken: April 13, 2014: received issue Discussion: End of Annotations:===== iler: QUALCOMM Windows Eudora Version 7.1.0.9 Date: Mon, 14 Apr 2014 08:03:06 -0400 To: issues@omg.org, uml2-rtf@omg.org From: Juergen Boldt Subject: issue 19335 -- UML 2 RTF issue X-Virus-Scanned: amavisd-new at omg.org X-Rocket-Received: from mjchonolesHP (mjchonoles@71.225.93.40 with plain [98.139.211.125]) by smtp225.mail.ne1.yahoo.com with SMTP; 13 Apr 2014 23:34:19 +0000 UTC From: "Michael Chonoles" To: , , Cc: , "'Jon Siegel'" , "'Østein Haugen'" Subject: UML 2.5 Overly strict restriction on message slope in seq diagrams Date: Sun, 13 Apr 2014 19:34:01 -0400 X-Mailer: Microsoft Outlook 15.0 Thread-Index: Ac9Xa91g27H+nnpJQDmSXjZzaNoZwA== X-Virus-Scanned: amavisd-new at omg.org (BTW, where should the issues found on UML 2.5 go, Is there a 2.6 mailing list?) In UML 2.5 Paragraph 17.4.4. Notation Message: ...A line must be such that every line fragment is either horizontal or downwards when traversed from send event to receive event.. While this restriction appears reasonable when first read, it is really overly strict. In the default, WEAK, interpretation of sequence diagrams, the time-wise ordering of occurrences within a lifeline is independent of occurrences on other lifelines, subject to cause/effect sequencing (message sending­>message reception). And, of course, the order of occurrences as depicted on the lifeline must be maintained. This is practically equivalent to saying that each lifeline has its own clock or timescale, and that ordering by that clock must be maintained and that causality across lifelines must be maintained. It is often thought that one could change the timescale (and not necessarily smoothly) on a lifeline without changing the interpretation or legality of the diagram. However, the restriction on messages not taking a non-negative slope, prevents otherwise legal changes in scale and makes invalid some diagrams that do not violate causality or the within-lifeline ordering. Please relax this restriction. Michael Jesse Chonoles [] Michael Jesse Chonoles OMG Analysis & Design Task Force Co-Chair Representing: Change-Vision Makers of the Astah line of modeling tools Telephones: Work 610 644-8404 Cell: 267-315-2410 Home 610 644-8404 Fax: 215-790-2976 Co-author: UML 2 For Dummies mjchonoles@yahoo.com Michael@ChonolesConsulting.Com Juergen Boldt Director, Member Services 109 Highland Ave Needham, MA 02494 USA Tel: 781 444 0404 x 132 fax: 781 444 0320 www.omg.org [] From: Østein Haugen To: Michael Chonoles , "issues@omg.org" , "uml2-rtf@omg.org" , "uml25-ftf@omg.org" CC: "marte-rtf@omg.org" , "'Jon Siegel'" Subject: RE: UML 2.5 Overly strict restriction on message slope in seq diagrams Thread-Topic: UML 2.5 Overly strict restriction on message slope in seq diagrams Thread-Index: Ac9Xa91g27H+nnpJQDmSXjZzaNoZwAATPGo3 Date: Mon, 14 Apr 2014 08:17:53 +0000 Accept-Language: en-US X-MS-Has-Attach: yes X-MS-TNEF-Correlator: x-originating-ip: [10.218.112.13] X-Virus-Scanned: amavisd-new at omg.org Dear Michael. This is one example where it would have been useful if the rationale of how things are, had been associated with the standard text :-) Even though your arguments are reasonable, you overlook two points which were the reasons for why the rule for restricting messages not to go upwards. Just to make this perfectly clear, I was the one including this restriction in the spec originally and my reasons were the ones quoted below: 1) The early dialects of sequence diagrams (e.g. Message Sequence Charts of ITU Z.120) all had this restriction, so that it was established practice. 2) More technically, we want to avoid messages defining a cycle of order dependencies, which of course would be illegal in a partial order. However, it is not so simple for a tool to check when InteractionUse as well as combined fragments add in principle unlimited complexity to the syntax. However, the notation rule that you consider to remove is a sufficient condition to keep the partial order free of cycles (in combination with a similar rule on gates on InteractionUse). Thus, I will very much advise NOT to remove this rule of messages going either horizontal or downwards. Regards, Østein -------------------------------------------------------------------------------- From: Michael Chonoles [mjchonoles@yahoo.com] Sent: Monday, April 14, 2014 1:34 AM To: issues@omg.org; uml2-rtf@omg.org; uml25-ftf@omg.org Cc: marte-rtf@omg.org; 'Jon Siegel'; Østein Haugen Subject: UML 2.5 Overly strict restriction on message slope in seq diagrams (BTW, where should the issues found on UML 2.5 go, Is there a 2.6 mailing list?) In UML 2.5 Paragraph 17.4.4. Notation Message: ...A line must be such that every line fragment is either horizontal or downwards when traversed from send event to receive event.. While this restriction appears reasonable when first read, it is really overly strict. In the default, WEAK, interpretation of sequence diagrams, the time-wise ordering of occurrences within a lifeline is independent of occurrences on other lifelines, subject to cause/effect sequencing (message sending.>message reception). And, of course, the order of occurrences as depicted on the lifeline must be maintained. This is practically equivalent to saying that each lifeline has its own clock or timescale, and that ordering by that clock must be maintained and that causality across lifelines must be maintained. It is often thought that one could change the timescale (and not necessarily smoothly) on a lifeline without changing the interpretation or legality of the diagram. However, the restriction on messages not taking a non-negative slope, prevents otherwise legal changes in scale and makes invalid some diagrams that do not violate causality or the within-lifeline ordering. Please relax this restriction. Michael Jesse Chonoles Michael Jesse Chonoles OMG Analysis & Design Task Force Co-Chair Representing: Change-Vision Makers of the Astah line of modeling tools Telephones: Work 610 644-8404 Cell: 267-315-2410 Home 610 644-8404 Fax: 215-790-2976 Co-author: UML 2 For Dummies mjchonoles@yahoo.com Michael@ChonolesConsulting.Com