Issue 5103: PrimitiveFunction shouldn't be restricted to datatypes (action-semantics-ftf) Source: NIST (Mr. Conrad Bock, conrad.bock@nist.gov conradb@nist.gov) Nature: Uncategorized Issue Severity: Summary: ApplyFunctionAction should work with any user-defined function, including those that operate on objects, and that have no outputs. Resolution: decline Revised Text: Actions taken: April 2, 2002: received issue December 11, 2002: closed issue Discussion: Use the new CallProcedureAction. Primitive functions is for math, string functions, etc, that are defined externally to the model. End of Annotations:===== Reply-To: From: "Conrad Bock" To: "ActionFTF" Subject: More action issues Date: Tue, 2 Apr 2002 10:23:55 -0800 X-Mailer: Microsoft Outlook IMO, Build 9.0.2416 (9.0.2910.0) Importance: Normal Juergen, More action issues, thanks. Conrad 3) Summary: PrimitiveFunction shouldn't be restricted to datatypes. Text: ApplyFunctionAction should work with any user-defined function, including those that operate on objects, and that have no outputs. From: "Rumbaugh, James" To: conrad.bock@kabira.com, ActionFTF Subject: RE: Draft fourth Action FTF ballot Date: Thu, 4 Apr 2002 10:02:52 -0800 X-Mailer: Internet Mail Service (5.5.2653.19) > -----Original Message----- > From: Conrad Bock [mailto:conrad.bock@kabira.com] > Sent: Wednesday, April 03, 2002 2:32 PM > To: ActionFTF > Subject: Draft fourth Action FTF ballot > > > > > VOTE (YES|NO|ABSTAIN): > > Issue 5103: PrimitiveFunction shouldn't be restricted to datatypes > http://cgi.omg.org/issues/issue5103.txt > > Resolution: Decline > > Discussion: > > Use the new CallProcedureAction. Primitive functions is for > math, string functions, etc, that are defined externally > to the model. > I would phrase the second sentence: Primitive functions are for math, string functions, etc, that are executed locally with absolutely no interaction with the rest of the system beyond accepting input values and producing output values. (This is what forces the values to be datatypes--they must be completely meaningful from only the information at hand. Allowing an object reference would require the ability to to access values within the object, which defeat the purpose of primitive functions as self-contained computations.) From: "Stephen J. Mellor" To: "Rumbaugh, James" , , "ActionFTF" Subject: RE: Draft fourth Action FTF ballot Date: Sun, 7 Apr 2002 14:02:43 -0700 X-Mailer: Microsoft Outlook IMO, Build 9.0.2416 (9.0.2911.0) Importance: Normal Don't we also have a TestIdentity action that requires references? | -----Original Message----- | From: Rumbaugh, James [mailto:rumbaugh@rational.com] | Sent: Thursday, 2002-04-04 11:03 | To: conrad.bock@kabira.com; ActionFTF | Subject: RE: Draft fourth Action FTF ballot | | | | | > -----Original Message----- | > From: Conrad Bock [mailto:conrad.bock@kabira.com] | > Sent: Wednesday, April 03, 2002 2:32 PM | > To: ActionFTF | > Subject: Draft fourth Action FTF ballot | > | > | > | > | > VOTE (YES|NO|ABSTAIN): | > | > Issue 5103: PrimitiveFunction shouldn't be restricted to datatypes | > http://cgi.omg.org/issues/issue5103.txt | > | > Resolution: Decline | > | > Discussion: | > | > Use the new CallProcedureAction. Primitive functions is for | > math, string functions, etc, that are defined externally | > to the model. | > | | | I would phrase the second sentence: | | Primitive functions are for math, string functions, etc, that are | executed locally with absolutely no interaction with the rest of | the system beyond accepting input values and producing output values. | | (This is what forces the values to be datatypes--they must be completely | meaningful from only the information at hand. Allowing an object reference | would require the ability to to access values within the object, which | defeat the purpose of primitive functions as self-contained | computations.)