Issue 11628: SysML:Ports can't be blocks (sysml-rtf) Source: Lockheed Martin (Mr. Michael Jesse Chonoles, michael.j.chonoles@lmco.com mjchonoles@yahoo.com) Nature: Uncategorized Issue Severity: Summary: Apparently from 9.3.2.3 while you can type a flowport by a block, that block indicates the things that flow over the port. A more intuitive interpretation would be that the flowport is a block. Most flowports appear to be physical things that may convey blocks. Without the ability to indicate the physical thing that is the block, you lose the ability to specify it, reuse it, define it, etc. It’s much more intuitive to indicate that the flowport is a US-110voltACmale In addition, for example, in Figure B.19. There are flowpoints named Port:FuelTankFitting Port:ICEFuelFitting Based on section 9.3.2.3, these flowports convey Fittings, not Fuel. There needs to be a way, preferably graphically, that would indicate that the type of Flowport is a block, and in that block, allow for the definitiaojn of what flows. Resolution: Revised Text: Actions taken: October 4, 2007: received issue Discussion: Discussion by the RTF recognized that additional typing of a flowports may be needed, but no resolution was reached. Disposition: Deferred Discussion: This issue is deferred because no other proposed resolution was voted on during the schedule of the SysML 1.2 RTF. Following is the discussion from a previous deferred resolution by the SysML 1.1 RTF: Discussion by the RTF recognized that additional typing of a flowports may be needed, but no resolution was reached. Disposition: Deferred End of Annotations:===== te: Thu, 04 Oct 2007 20:06:39 -0400 From: "Chonoles, Michael J" Subject: SysML:Ports can't be blocks To: issues@omg.org, sysml-rtf@omg.org Thread-Topic: SysML:Ports can't be blocks Thread-Index: AcgG45eFPiyJEqh3RoqJVzrothKHPg== X-MS-Has-Attach: X-MS-TNEF-Correlator: X-OriginalArrivalTime: 05 Oct 2007 00:07:16.0442 (UTC) FILETIME=[B04083A0:01C806E3] Apparently from 9.3.2.3 while you can type a flowport by a block, that block indicates the things that flow over the port. A more intuitive interpretation would be that the flowport is a block. Most flowports appear to be physical things that may convey blocks. Without the ability to indicate the physical thing that is the block, you lose the ability to specify it, reuse it, define it, etc. It.s much more intuitive to indicate that the flowport is a US-110voltACmale In addition, for example, in Figure B.19. There are flowpoints named Port:FuelTankFitting Port:ICEFuelFitting Based on section 9.3.2.3, these flowports convey Fittings, not Fuel. There needs to be a way, preferably graphically, that would indicate that the type of Flowport is a block, and in that block, allow for the definitiaojn of what flows. Michael Chonoles Lockheed Martin Date: Thu, 04 Oct 2007 23:21:52 -0400 From: "Friedenthal, Sanford" Subject: RE: SysML:Ports can't be blocks To: "Chonoles, Michael J" , issues@omg.org, sysml-rtf@omg.org Thread-Topic: SysML:Ports can't be blocks Thread-Index: AcgG45eFPiyJEqh3RoqJVzrothKHPgAGg3nQ X-MS-Has-Attach: X-MS-TNEF-Correlator: X-OriginalArrivalTime: 05 Oct 2007 03:21:52.0393 (UTC) FILETIME=[DFA93F90:01C806FE] Michael We did consider early in the development of the flowport concept that it represent the physical thing that conveyed the thing that flows. However, we decided that the physical thing you are referring to is just another part. We used the port to represent a .specification. of the type of thing that can flow at that interaction point. Sandy Sanford Friedenthal Lockheed Martin (703) 293-5557 -------------------------------------------------------------------------------- From: Chonoles, Michael J Sent: Thursday, October 04, 2007 8:07 PM To: issues@omg.org; sysml-rtf@omg.org Subject: SysML:Ports can't be blocks Apparently from 9.3.2.3 while you can type a flowport by a block, that block indicates the things that flow over the port. A more intuitive interpretation would be that the flowport is a block. Most flowports appear to be physical things that may convey blocks. Without the ability to indicate the physical thing that is the block, you lose the ability to specify it, reuse it, define it, etc. It.s much more intuitive to indicate that the flowport is a US-110voltACmale In addition, for example, in Figure B.19. There are flowpoints named Port:FuelTankFitting Port:ICEFuelFitting Based on section 9.3.2.3, these flowports convey Fittings, not Fuel. There needs to be a way, preferably graphically, that would indicate that the type of Flowport is a block, and in that block, allow for the definitiaojn of what flows. Michael Chonoles Lockheed Martin Date: Fri, 05 Oct 2007 10:18:33 -0400 From: "Chonoles, Michael J" Subject: RE: SysML:Ports can't be blocks To: "Friedenthal, Sanford" , issues@omg.org, sysml-rtf@omg.org Thread-Topic: SysML:Ports can't be blocks Thread-Index: AcgG45eFPiyJEqh3RoqJVzrothKHPgAGg3nQABZ8BkA= X-MS-Has-Attach: X-MS-TNEF-Correlator: X-OriginalArrivalTime: 05 Oct 2007 14:18:40.0672 (UTC) FILETIME=[A0D36A00:01C8075A] If so, then at least some of the diagrams in the spec are wrong, such as B-19. However, the problem with the decision is that makes diagramming that much more difficult. Imagine modeling a laptop. Do I really want to have one additional part for every port. If I could specify both -- what the port is and what it conveys, I would be much happier. Having the port.s block be what it conveys is also non-intuitive and violates some of the basic patterns in UML/SysML. Is this the only case where something is typed by a block, but isn.t that block? Michael From: Friedenthal, Sanford Sent: Thursday, October 04, 2007 11:22 PM To: Chonoles, Michael J; issues@omg.org; sysml-rtf@omg.org Subject: RE: SysML:Ports can't be blocks Michael We did consider early in the development of the flowport concept that it represent the physical thing that conveyed the thing that flows. However, we decided that the physical thing you are referring to is just another part. We used the port to represent a .specification. of the type of thing that can flow at that interaction point. Sandy Sanford Friedenthal Lockheed Martin (703) 293-5557 -------------------------------------------------------------------------------- From: Chonoles, Michael J Sent: Thursday, October 04, 2007 8:07 PM To: issues@omg.org; sysml-rtf@omg.org Subject: SysML:Ports can't be blocks Apparently from 9.3.2.3 while you can type a flowport by a block, that block indicates the things that flow over the port. A more intuitive interpretation would be that the flowport is a block. Most flowports appear to be physical things that may convey blocks. Without the ability to indicate the physical thing that is the block, you lose the ability to specify it, reuse it, define it, etc. It.s much more intuitive to indicate that the flowport is a US-110voltACmale In addition, for example, in Figure B.19. There are flowpoints named Port:FuelTankFitting Port:ICEFuelFitting Based on section 9.3.2.3, these flowports convey Fittings, not Fuel. There needs to be a way, preferably graphically, that would indicate that the type of Flowport is a block, and in that block, allow for the definition of what flows. Michael Chonoles Lockheed Martin Date: Fri, 05 Oct 2007 10:22:09 -0400 From: "Friedenthal, Sanford" Subject: RE: SysML:Ports can't be blocks To: "Chonoles, Michael J" , issues@omg.org, sysml-rtf@omg.org Thread-Topic: SysML:Ports can't be blocks Thread-Index: AcgG45eFPiyJEqh3RoqJVzrothKHPgAGg3nQABZ8BkAAAM3kkA== X-MS-Has-Attach: X-MS-TNEF-Correlator: X-OriginalArrivalTime: 05 Oct 2007 14:22:10.0058 (UTC) FILETIME=[1DA12EA0:01C8075B] In a sense, it is consistent with the standard ports in that the port represents a .specification. of the interaction. I do agree that it is more intuitive for the port to represent the physical port, but again, I believe there may be issues by treating a .part. at the interface as a port. Others would need to weigh in. Eldad was the originator of the flow port concept so perhaps he has an input. Sanford Friedenthal Lockheed Martin (703) 293-5557 -------------------------------------------------------------------------------- From: Chonoles, Michael J Sent: Friday, October 05, 2007 10:19 AM To: Friedenthal, Sanford; 'issues@omg.org'; 'sysml-rtf@omg.org' Subject: RE: SysML:Ports can't be blocks If so, then at least some of the diagrams in the spec are wrong, such as B-19. However, the problem with the decision is that makes diagramming that much more difficult. Imagine modeling a laptop. Do I really want to have one additional part for every port. If I could specify both -- what the port is and what it conveys, I would be much happier. Having the port.s block be what it conveys is also non-intuitive and violates some of the basic patterns in UML/SysML. Is this the only case where something is typed by a block, but isn.t that block? Michael From: Friedenthal, Sanford Sent: Thursday, October 04, 2007 11:22 PM To: Chonoles, Michael J; issues@omg.org; sysml-rtf@omg.org Subject: RE: SysML:Ports can't be blocks Michael We did consider early in the development of the flowport concept that it represent the physical thing that conveyed the thing that flows. However, we decided that the physical thing you are referring to is just another part. We used the port to represent a .specification. of the type of thing that can flow at that interaction point. Sandy Sanford Friedenthal Lockheed Martin (703) 293-5557 -------------------------------------------------------------------------------- From: Chonoles, Michael J Sent: Thursday, October 04, 2007 8:07 PM To: issues@omg.org; sysml-rtf@omg.org Subject: SysML:Ports can't be blocks Apparently from 9.3.2.3 while you can type a flowport by a block, that block indicates the things that flow over the port. A more intuitive interpretation would be that the flowport is a block. Most flowports appear to be physical things that may convey blocks. Without the ability to indicate the physical thing that is the block, you lose the ability to specify it, reuse it, define it, etc. It.s much more intuitive to indicate that the flowport is a US-110voltACmale In addition, for example, in Figure B.19. There are flowpoints named Port:FuelTankFitting Port:ICEFuelFitting Based on section 9.3.2.3, these flowports convey Fittings, not Fuel. There needs to be a way, preferably graphically, that would indicate that the type of Flowport is a block, and in that block, allow for the definition of what flows. Michael Chonoles Lockheed Martin Date: Fri, 05 Oct 2007 10:22:51 -0400 From: "Friedenthal, Sanford" Subject: RE: SysML:Ports can't be blocks To: "Chonoles, Michael J" , issues@omg.org, sysml-rtf@omg.org Thread-Topic: SysML:Ports can't be blocks Thread-Index: AcgG45eFPiyJEqh3RoqJVzrothKHPgAGg3nQABZ8BkAAAM3kkAAAGq7Q X-MS-Has-Attach: X-MS-TNEF-Correlator: X-OriginalArrivalTime: 05 Oct 2007 14:22:52.0214 (UTC) FILETIME=[36C1AD60:01C8075B] Actually, Eran and Eldad were the originator of this concept, so perhaps they can comment. Sanford Friedenthal Lockheed Martin (703) 293-5557 -------------------------------------------------------------------------------- From: Friedenthal, Sanford Sent: Friday, October 05, 2007 10:22 AM To: Chonoles, Michael J; 'issues@omg.org'; 'sysml-rtf@omg.org' Subject: RE: SysML:Ports can't be blocks In a sense, it is consistent with the standard ports in that the port represents a .specification. of the interaction. I do agree that it is more intuitive for the port to represent the physical port, but again, I believe there may be issues by treating a .part. at the interface as a port. Others would need to weigh in. Eldad was the originator of the flow port concept so perhaps he has an input. Sanford Friedenthal Lockheed Martin (703) 293-5557 -------------------------------------------------------------------------------- From: Chonoles, Michael J Sent: Friday, October 05, 2007 10:19 AM To: Friedenthal, Sanford; 'issues@omg.org'; 'sysml-rtf@omg.org' Subject: RE: SysML:Ports can't be blocks If so, then at least some of the diagrams in the spec are wrong, such as B-19. However, the problem with the decision is that makes diagramming that much more difficult. Imagine modeling a laptop. Do I really want to have one additional part for every port. If I could specify both -- what the port is and what it conveys, I would be much happier. Having the port.s block be what it conveys is also non-intuitive and violates some of the basic patterns in UML/SysML. Is this the only case where something is typed by a block, but isn.t that block? Michael From: Friedenthal, Sanford Sent: Thursday, October 04, 2007 11:22 PM To: Chonoles, Michael J; issues@omg.org; sysml-rtf@omg.org Subject: RE: SysML:Ports can't be blocks Michael We did consider early in the development of the flowport concept that it represent the physical thing that conveyed the thing that flows. However, we decided that the physical thing you are referring to is just another part. We used the port to represent a .specification. of the type of thing that can flow at that interaction point. Sandy Sanford Friedenthal Lockheed Martin (703) 293-5557 -------------------------------------------------------------------------------- From: Chonoles, Michael J Sent: Thursday, October 04, 2007 8:07 PM To: issues@omg.org; sysml-rtf@omg.org Subject: SysML:Ports can't be blocks Apparently from 9.3.2.3 while you can type a flowport by a block, that block indicates the things that flow over the port. A more intuitive interpretation would be that the flowport is a block. Most flowports appear to be physical things that may convey blocks. Without the ability to indicate the physical thing that is the block, you lose the ability to specify it, reuse it, define it, etc. It.s much more intuitive to indicate that the flowport is a US-110voltACmale In addition, for example, in Figure B.19. There are flowpoints named Port:FuelTankFitting Port:ICEFuelFitting Based on section 9.3.2.3, these flowports convey Fittings, not Fuel. There needs to be a way, preferably graphically, that would indicate that the type of Flowport is a block, and in that block, allow for the definition of what flows. Michael Chonoles Lockheed Martin So, should I consider this email thread "discussion" or should I assign an issue number? -Juergen At 11:21 PM 10/4/2007, Friedenthal, Sanford wrote: Michael We did consider early in the development of the flowport concept that it represent the physical thing that conveyed the thing that flows. However, we decided that the physical thing you are referring to is just another part. We used the port to represent a .specification. of the type of thing that can flow at that interaction point. Sandy Sanford Friedenthal Lockheed Martin (703) 293-5557 From: Chonoles, Michael J Sent: Thursday, October 04, 2007 8:07 PM To: issues@omg.org; sysml-rtf@omg.org Subject: SysML:Ports can't be blocks Apparently from 9.3.2.3 while you can type a flowport by a block, that block indicates the things that flow over the port. A more intuitive interpretation would be that the flowport is a block. Most flowports appear to be physical things that may convey blocks. Without the ability to indicate the physical thing that is the block, you lose the ability to specify it, reuse it, define it, etc. It.s much more intuitive to indicate that the flowport is a US-110voltACmale In addition, for example, in Figure B.19. There are flowpoints named Port:FuelTankFitting Port:ICEFuelFitting Based on section 9.3.2.3, these flowports convey Fittings, not Fuel. There needs to be a way, preferably graphically, that would indicate that the type of Flowport is a block, and in that block, allow for the definitiaojn of what flows. Michael Chonoles Lockheed Martin "Chonoles, Michael J" Hi Michael, is this clarified or should I file as an issue? -Juergen At 08:06 PM 10/4/2007, you wrote: Apparently from 9.3.2.3 while you can type a flowport by a block, that block indicates the things that flow over the port. A more intuitive interpretation would be that the flowport is a block. Most flowports appear to be physical things that may convey blocks. Without the ability to indicate the physical thing that is the block, you lose the ability to specify it, reuse it, define it, etc. It.s much more intuitive to indicate that the flowport is a US-110voltACmale In addition, for example, in Figure B.19. There are flowpoints named Port:FuelTankFitting Port:ICEFuelFitting Based on section 9.3.2.3, these flowports convey Fittings, not Fuel. There needs to be a way, preferably graphically, that would indicate that the type of Flowport is a block, and in that block, allow for the definitiaojn of what flows. Michael Chonoles Lockheed Martin Date: Wed, 10 Oct 2007 22:40:57 -0400 From: "Friedenthal, Sanford" Subject: RE: SysML:Ports can't be blocks To: Juergen Boldt , "Chonoles, Michael J" Thread-Topic: SysML:Ports can't be blocks Thread-Index: AcgLbsopjdAC/9HiSW2bEAcY0974JgAQVhfw X-MS-Has-Attach: X-MS-TNEF-Correlator: X-OriginalArrivalTime: 11 Oct 2007 02:40:57.0785 (UTC) FILETIME=[27144A90:01C80BB0] Up to Michael. Sanford Friedenthal Lockheed Martin (703) 293-5557 -------------------------------------------------------------------------------- From: Juergen Boldt [mailto:juergen@omg.org] Sent: Wednesday, October 10, 2007 2:53 PM To: Friedenthal, Sanford; Chonoles, Michael J Subject: RE: SysML:Ports can't be blocks So, should I consider this email thread "discussion" or should I assign an issue number? -Juergen At 11:21 PM 10/4/2007, Friedenthal, Sanford wrote: Michael We did consider early in the development of the flowport concept that it represent the physical thing that conveyed the thing that flows. However, we decided that the physical thing you are referring to is just another part. We used the port to represent a specification of the type of thing that can flow at that interaction point. Sandy