Issue 3445: MOF 1.3 IDL template for clustering uses wrong name (mof-rtf) Source: Escape Velocity (Mr. Don Baisley, donbaisley(at)live.com) Nature: Uncategorized Issue Severity: Summary: In the MOF 1.3 specification, section 5.8.4, the use of "<clustered_package_name>_ref" (both in the template and as a subheading) is incorrect. It should be "<import_name>_ref". The Import name provides the alias for a clustered package within a clustering package. The Package name is used to identify the type of the M1 package object, not the IDL attribute name. Recommendation: change "<clustered_package_name>_ref" to "<import_name>_ref". Resolution: Revised Text: Actions taken: March 22, 2000: received issue Discussion: End of Annotations:===== rom: "Baisley, Donald E" To: issues@omg.org Subject: MOF 1.3 IDL template for clustering uses wrong name Date: Wed, 22 Mar 2000 17:27:55 -0600 MIME-Version: 1.0 X-Mailer: Internet Mail Service (5.5.2650.21) Content-Type: text/plain; charset="iso-8859-1" X-UIDL: +B9e9WhVd94'T!!aA[!! In the MOF 1.3 specification, section 5.8.4, the use of "_ref" (both in the template and as a subheading) is incorrect. It should be "_ref". The Import name provides the alias for a clustered package within a clustering package. The Package name is used to identify the type of the M1 package object, not the IDL attribute name. Recommendation: change "_ref" to "_ref". Don Baisley Unisys X-Mailer: exmh version 2.1.0 09/18/1999 To: "Baisley, Donald E" cc: crawley@dstc.edu.au, mof-rtf@omg.org Subject: Re: MOF 1.3 IDL template for clustering uses wrong name In-reply-to: Your message of "Wed, 22 Mar 2000 17:27:55 CST." Mime-Version: 1.0 Date: Thu, 30 Mar 2000 14:47:39 +1000 From: Stephen Crawley Content-Type: text/plain; charset=us-ascii X-UIDL: +K]d9aAD!!kA;!!OD@e9 > In the MOF 1.3 specification, section 5.8.4, the use of > "_ref" (both in the template and as a > subheading) is > incorrect. It should be "_ref". The Import name > provides the > alias for a clustered package within a clustering package. The > Package name > is used to identify the type of the M1 package object, not the IDL > attribute > name. > > Recommendation: change "_ref" to > "_ref". As far as I can see, this is purely editorial, harmless and should be done. -- Steve X-Mailer: exmh version 2.2 06/23/2000 with nmh-1.0.4 To: mof-rtf@omg.org Subject: Re: Issue 3445: MOF 1.3 IDL template for clustering uses wrong name Mime-Version: 1.0 Date: Wed, 13 Jun 2001 16:32:12 +1000 From: Stephen Crawley X-Scanned-By: MIMEDefang 1.0 (http://www.roaringpenguin.com/mimedefang/) Content-Type: text/plain; charset=us-ascii X-UIDL: d0=e9'Ged9&9[d9@T[d9 > Source: Unisys (Mr. Don Baisley, Donald.Baisley@unisys.com) > Nature: Uncategorized Issue > Severity: > Summary: In the MOF 1.3 specification, section 5.8.4, the use of > "_ref" (both in the template and as a > subheading) is incorrect. It should be "_ref". The > Import > name provides the alias for a clustered package within a clustering > package. The Package name is used to identify the type of the M1 > package > object, not the IDL attribute name. > > Recommendation: change "_ref" to > "_ref". Proposed Resolution: Close with no action. Discussion: Don has a valid point. Technically, an clustered Package is expressed using an Import ... at least given the way we currently meta-model the construct. However, if we were to change the templates and descriptions as Don recommends, we'd run the risk that readers failed to notice that the template only applies to Imports when isClustered == true. On the whole, the current templates / text are more clear as they stand. It should also be noted that the IDL template notation was designed for maximum human readability. Formality, precision and completeness were secondary goals. From: "Baisley, Donald E" To: mof-rtf@omg.org Subject: RE: Issue 3445: MOF 1.3 IDL template for clustering uses wrong na me Date: Wed, 13 Jun 2001 18:51:38 -0500 MIME-Version: 1.0 X-Mailer: Internet Mail Service (5.5.2653.19) Content-Type: text/plain; charset="iso-8859-1" X-UIDL: >Zfd9/0,e9Gp1!!n'Zd9 Status: RO The proposed resolution (no action) leaves an obvious bug in the IDL mapping. The relationship from an Import to a Namespace is called "Aliases" because the Import provides an alias for imported Namespace within the importing package. The alias capability is important for metamodels that cluster other metamodels that have the same names. This can happen when a metamodel is used to define mappings across versions of some other metamodel (e.g. a map between OMG's UML and OIM's UML -- One might give an alias of OmgUml to one UML and an alias of OimUml to the other). The IDL will not compile if the aliasing is ignored. > Formality, precision and completeness were secondary goals. Are the templates nonnormative? Shall we ignore a template where it is incomplete or incorrect? There is no readability issue here. This is just a straightforward bug fix. Regards, Don -----Original Message----- From: Stephen Crawley [mailto:crawley@dstc.edu.au] Sent: Tuesday, June 12, 2001 11:32 PM To: mof-rtf@omg.org Subject: Re: Issue 3445: MOF 1.3 IDL template for clustering uses wrong name > Source: Unisys (Mr. Don Baisley, Donald.Baisley@unisys.com) > Nature: Uncategorized Issue > Severity: > Summary: In the MOF 1.3 specification, section 5.8.4, the use of > "_ref" (both in the template and as a > subheading) is incorrect. It should be "_ref". The > Import > name provides the alias for a clustered package within a clustering > package. The Package name is used to identify the type of the M1 > package > object, not the IDL attribute name. > > Recommendation: change "_ref" to "_ref". Proposed Resolution: Close with no action. Discussion: Don has a valid point. Technically, an clustered Package is expressed using an Import ... at least given the way we currently meta-model the construct. However, if we were to change the templates and descriptions as Don recommends, we'd run the risk that readers failed to notice that the template only applies to Imports when isClustered == true. On the whole, the current templates / text are more clear as they stand. It should also be noted that the IDL template notation was designed for maximum human readability. Formality, precision and completeness were secondary goals. X-Mailer: exmh version 2.2 06/23/2000 with nmh-1.0.4 To: "Baisley, Donald E" cc: mof-rtf@omg.org, crawley@dstc.edu.au Subject: Re: Issue 3445: MOF 1.3 IDL template for clustering uses wrong na me In-Reply-To: Message from "Baisley, Donald E" of "Wed, 13 Jun 2001 18:51:38 EST." Mime-Version: 1.0 Date: Thu, 14 Jun 2001 10:27:05 +1000 From: Stephen Crawley X-Scanned-By: MIMEDefang 1.0 (http://www.roaringpenguin.com/mimedefang/) Content-Type: text/plain; charset=us-ascii X-UIDL: The proposed resolution (no action) leaves an obvious bug in the IDL > mapping. The relationship from an Import to a Namespace is called > "Aliases" > because the Import provides an alias for imported Namespace within > the > importing package. The alias capability is important for metamodels > that > cluster other metamodels that have the same names. This can happen > when a > metamodel is used to define mappings across versions of some other > metamodel > (e.g. a map between OMG's UML and OIM's UML -- One might give an > alias of > OmgUml to one UML and an alias of OimUml to the other). The IDL > will not > compile if the aliasing is ignored. Umm ... maybe you are correct. I've never come across this scenario. Anyway, I propose that we DO NOT put the resolution for 3445 to a vote tomorrow. > Are the templates nonnormative? They are normative. They are intended to be sufficiently complete and correct to allow different vendors to produce generators that will produce compatible IDL. If they are not, that's a bug. > Shall we ignore a template where it is incomplete or incorrect? An incomplete template should not be ignored. Rather, the vendor should fill in the details that are missing. (For example the templates do not spell out how to map TypeCodes to IDL data types. This is left to the implementor to suss out ... on the basis of his / her knowledge of the syntax of CORBA IDL.) The only case where incompleteness is a problem is when it allows different interpretations to give incompatible IDL. A significantly incorrect template would be a real problem. There are (in general) precedents for ignoring incorrect OMG spec in order to produce working implementations. But this makes claims of compliance unconvincing, and is often a problem for interoperability. > There is no readability issue here. This is just a straightforward > bug fix. I hadn't realised this ... sorry. -- Steve X-Mailer: exmh version 2.2 06/23/2000 with nmh-1.0.4 To: Donald.Baisley@unisys.com Cc: mof-rtf@omg.org Subject: Re: Issue 3445: MOF 1.3 IDL template for clustering uses wrong name Mime-Version: 1.0 Date: Thu, 25 Oct 2001 16:27:54 +1000 From: Stephen Crawley X-Scanned-By: MIMEDefang 1.0 (http://www.roaringpenguin.com/mimedefang/) Content-Type: text/plain; charset=us-ascii X-UIDL: =Nhd9Do,!!1LQ!!3TZd9 > Source: Unisys (Mr. Don Baisley, Donald.Baisley@unisys.com) > Summary: In the MOF 1.3 specification, section 5.8.4, the use of > "_ref" (both in the template and as a > subheading) > is incorrect. It should be "_ref". The Import name > provides > the alias for a clustered package within a clustering package. The > Package > name is used to identify the type of the M1 package object, not the > IDL > attribute name. Recommendation: change > "_ref" to > "_ref". Don, There is currently an IDL mapping precondition (section 5.5, bullet point #9) that says the following: * After name substitution (5.6.2.1), the name of an Import must equal the name of its "importedNamespace". The result of this is that the template / heading are consistent with the metamodel spec (constrained by the preconditions). Presumably you actually want precondition #9 be removed. If so, can you say why you want this, and why it is important to Unisys? I can guess, but for completeness ... [If not, your issue is purely editorial ... and, IMO, makes the template less clear.] -- Steve From: "Martin Matula" To: , "Stephen Crawley" Cc: References: <200110250627.f9P6Rnc11701@piglet.dstc.edu.au> Subject: Re: Issue 3445: MOF 1.3 IDL template for clustering uses wrong name Date: Thu, 25 Oct 2001 10:58:18 +0200 MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Priority: 3 X-MSMail-Priority: Normal X-Mailer: Microsoft Outlook Express 5.00.2919.6700 X-MimeOLE: Produced By Microsoft MimeOLE V5.00.2919.6700 Content-Type: text/plain; charset="iso-8859-1" X-UIDL: gM\!!($>e9[+1e9p4J!! Hi Steve, I agree with Don. As Don says, Import name provides alias for the clustered package name. Import name should be used instead of imported namespace's name, because for import's name it is always guaranteed that it is unique in its namespace, however the name of the imported package is not necessarily unique within the namespace that imports it. My personal feeling is that the approach that Don proposes is more straightforward than the current one with substitute names. Martin ----- Original Message ----- From: "Stephen Crawley" To: Cc: Sent: Thursday, October 25, 2001 8:27 AM Subject: Re: Issue 3445: MOF 1.3 IDL template for clustering uses wrong name > > Source: Unisys (Mr. Don Baisley, Donald.Baisley@unisys.com) > > Summary: In the MOF 1.3 specification, section 5.8.4, the use of > > "_ref" (both in the template and as a subheading) > > is incorrect. It should be "_ref". The Import name provides > > the alias for a clustered package within a clustering package. The Package > > name is used to identify the type of the M1 package object, not > > the IDL > > attribute name. Recommendation: change > > "_ref" to > > "_ref". > > Don, > > There is currently an IDL mapping precondition (section 5.5, bullet > point #9) that says the following: > > * After name substitution (5.6.2.1), the name of an Import must > > equal > the name of its "importedNamespace". > > The result of this is that the template / heading are consistent > > with > the metamodel spec (constrained by the preconditions). > > Presumably you actually want precondition #9 be removed. If so, can > > you > say why you want this, and why it is important to Unisys? I can > > guess, > but for completeness ... > > [If not, your issue is purely editorial ... and, IMO, makes the > > template > less clear.] > > -- Steve > > > X-Mailer: exmh version 2.2 06/23/2000 with nmh-1.0.4 To: "Martin Matula" cc: Donald.Baisley@unisys.com, "Stephen Crawley" , mof-rtf@omg.org, crawley@dstc.edu.au Subject: Re: Issue 3445: MOF 1.3 IDL template for clustering uses wrong name In-Reply-To: Message from "Martin Matula" of "Thu, 25 Oct 2001 10:58:18 +0200." <15f401c15d33$3d434aa0$844b9c81@matula> Mime-Version: 1.0 Date: Fri, 26 Oct 2001 00:38:47 +1000 From: Stephen Crawley X-Scanned-By: MIMEDefang 1.0 (http://www.roaringpenguin.com/mimedefang/) Content-Type: text/plain; charset=us-ascii X-UIDL: 3??!!hWVd9-?4e9\]B!! Martin, My point is that Don's issue (as stated) doesn't solve any problem. The current spec explicitly says that you can't do import/cluster aliasing with the IDL mapping, and Don's proposed change doesn't alter this. If we want to change this, we need to remove the IDL mapping precondition that precludes aliasing. We also need to recognise that in the IDL mapping aliasing only "works" for clusters. For imports, aliasing can have no effect. Finally, since this is a substantative change (enabling something that was previously explicitly disabled), the RTF needs to know why it should make the change ... for the record. [I've never encountered a metamodel where cluster aliasing was necessary. Nor apparently have any of DSTC's customers. And I'm not keen on supporting metamodel constructs that nobody seems to need.] -- Steve From: "Baisley, Donald E" To: Stephen Crawley Cc: mof-rtf@omg.org Subject: RE: Issue 3445: MOF 1.3 IDL template for clustering uses wrong na me Date: Mon, 29 Oct 2001 11:23:53 -0600 MIME-Version: 1.0 X-Mailer: Internet Mail Service (5.5.2653.19) Content-Type: text/plain; charset="iso-8859-1" X-UIDL: 8%h!!:$k!!=cF!!7<[d9 The IDL mapping precondition simply makes some metamodels unmappable to IDL without special tags. My only reason for removing the precondition is that it is pointless -- extra complexity for no purpose. Don -----Original Message----- From: Stephen Crawley [mailto:crawley@dstc.edu.au] Sent: Wednesday, October 24, 2001 11:28 PM To: Donald.Baisley@unisys.com Cc: mof-rtf@omg.org Subject: Re: Issue 3445: MOF 1.3 IDL template for clustering uses wrong name Importance: High > Source: Unisys (Mr. Don Baisley, Donald.Baisley@unisys.com) > Summary: In the MOF 1.3 specification, section 5.8.4, the use of > "_ref" (both in the template and as a > subheading) > is incorrect. It should be "_ref". The Import name > provides > the alias for a clustered package within a clustering package. The > Package > name is used to identify the type of the M1 package object, not the IDL > attribute name. Recommendation: change "_ref" to > "_ref". Don, There is currently an IDL mapping precondition (section 5.5, bullet point #9) that says the following: * After name substitution (5.6.2.1), the name of an Import must equal the name of its "importedNamespace". The result of this is that the template / heading are consistent with the metamodel spec (constrained by the preconditions). Presumably you actually want precondition #9 be removed. If so, can you say why you want this, and why it is important to Unisys? I can guess, but for completeness ... [If not, your issue is purely editorial ... and, IMO, makes the template less clear.] -- Steve