Issue 1504: Chapter 23.2.7 of CORBA 2.2, IDL to Ada Mapping ( pragma Subsystem) (ada-rtf) Source: (, ) Nature: Revision Severity: Significant Summary: Summary: Chapter 23.2.7 of the CORBA v2.2 Mapping of OMG IDL to Ada mentions: " Files (actually inclusion streams) create a package to contain the "bare" definitions defined in IDL"s global scope. The package name is formed from the concatenation of the file name and _IDL_File. " In the absence of a pragma Subsystem, this solution is felt to be unsatisfying. Resolution: Revised Text: Actions taken: June 4, 1998: received issue June 4, 1998: issue moved from orb_revision to the ada-rtf December 2, 2002: closed issue Discussion: End of Annotations:===== Return-Path: Date: Thu, 4 Jun 1998 11:26:22 -0400 From: www To: juergen@omg.org, web-incoming@omg.org Subject: WWW Form output Name: Oliver M. Kellogg Company: Daimler-Benz Aerospace AG Email: Oliver.Kellogg@vs.dasa.de Notification: Yes Specification: CORBA 2.2 Section: Formal #: Version: Revision_Date: Page: Nature: Clarification Severity: Significant full_desc: Subject: Server side mapping of IDL "attribute" References: Mapping of OMG IDL to C++ (chapter 20), Mapping of OMG IDL to Ada (chapter 23) Complaint: The CORBA 2.2 spec does not mention how IDL attributes should be mapped on the server side. Also, no justification or motivation is given concerning the lack of server side attribute mapping rules. Motivation: One vendor of an IDL to Ada translator decided to map the attribute to a single, package global, predefined instance of an Object. In other words, the Set_ and Get_ methods internally manipulate only one single, global object. This is in variance with the user's freedom to create more than one Object in his application. At least some elementary mapping rules should be given to preclude such an implementation for CORBA conforming products. Oliver M. Kellogg Daimler-Benz Aerospace AG Sensor Systems, Dept. VE3E3 D-89070 Ulm, Germany phone: (+49) 731 392-7138 fax: (+49) 731 392-4255 e-mail: Oliver.Kellogg@vs.dasa.de submit: Submit Issue Report Return-Path: Sender: jon@floorboard.com Date: Thu, 04 Jun 1998 09:36:54 -0700 From: Jonathan Biggar To: Juergen Boldt CC: issues@omg.org, orb_revision@omg.org Subject: Re: issue 1504 -- Core Revision References: <3.0.32.19980604114634.00a14a2c@emerald.omg.org> Juergen Boldt wrote: > > This is issue # 1504 > > Server side mapping of IDL "attribute" > > Complaint: The CORBA 2.2 spec does not mention how IDL attributes > should be mapped on the server side. Also, no justification or > motivation > is given concerning the lack of server side attribute mapping rules. > > Motivation: One vendor of an IDL to Ada translator decided to map > the > attribute to a single, package global, predefined instance of an > Object. > In other words, the Set_ and Get_ methods internally manipulate only > one single, global object. This is in variance with the user's > freedom > to create more than one Object in his application. At least some > elementary mapping rules should be given to preclude such an > implementation for CORBA conforming products. This sounds like an incredible case of vendor stupidity. I guess it goes to show that the CORBA spec isn't idiot-proof. I think there is plenty of text in the CORBA 2.2 spec that makes it clear that attributes are in a sense just a shorthand for paired get and set operations. -- Jon Biggar Floorboard Software jon@floorboard.com jon@biggar.org Return-Path: Date: Thu, 4 Jun 1998 19:32:10 -0400 (EDT) From: Bill Beckwith X-Sender: beckwb@gamma To: Jonathan Biggar cc: Juergen Boldt , issues@omg.org, orb_revision@omg.org Subject: Re: issue 1504 -- Core Revision On Thu, 4 Jun 1998, Jonathan Biggar wrote: > Juergen Boldt wrote: > > > > This is issue # 1504 > > > > Server side mapping of IDL "attribute" > > > > Complaint: The CORBA 2.2 spec does not mention how IDL attributes > > should be mapped on the server side. Also, no justification or > motivation > > is given concerning the lack of server side attribute mapping > rules. > > > > Motivation: One vendor of an IDL to Ada translator decided to map > the > > attribute to a single, package global, predefined instance of an > Object. > > In other words, the Set_ and Get_ methods internally manipulate > only > > one single, global object. This is in variance with the user's > freedom > > to create more than one Object in his application. At least some > > elementary mapping rules should be given to preclude such an > > implementation for CORBA conforming products. > > This sounds like an incredible case of vendor stupidity. I guess it > goes > to show that the CORBA spec isn't idiot-proof. I agree that the unnamed vendor did a poor job providing the user a useful default implementation of the set/get routines. But I assume the generated set/get routines are just that: a default implementation. > I think there is plenty of text in the CORBA 2.2 spec that makes it > clear that attributes are in a sense just a shorthand for paired get > and > set operations. The vendor in question, whoever they are, did a legal implementation of the IDL mapping for attributes since the _content_ of the set/get pair is (appropriately) not specified by the CORBA 2.2 spec. The user of this vendor's IDL to Ada translator would certainly be able to provide an alternative, more useful implementation of the attribute set/get routines. -- Bill