Issue 3067: Relationship to PMF (pids-rtf2) Source: Level Seven Visualizations (Mr. Jon Farmer, jon(at)level7vis.com) Nature: Uncategorized Issue Severity: Summary: Is there, or can there be made, a formal relationship between PMF properties and PIDS traits? Partial Proposed Resolution: The PMF uses CosPropertyService interfaces for property manipulation. These properties are not explicitly namespace-qualified. typedef string PropertyName; struct Property { PropertyName property_name; any property_value; }; Resolution: Revised Text: Actions taken: November 23, 1999: received issue Discussion: End of Annotations:===== Reply-To: "Jon Farmer" From: "Jon Farmer" To: "pids-rtf2" , "Carol Burt" , "Ron Zahavi" , Subject: relationship to PMF Date: Tue, 23 Nov 1999 19:17:44 -0500 Organization: Care Data Systems MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Priority: 3 X-MSMail-Priority: Normal X-Mailer: Microsoft Outlook Express 5.00.2314.1300 X-MimeOLE: Produced By Microsoft MimeOLE V5.00.2314.1300 Content-Type: text/plain; charset="iso-8859-1" X-UIDL: gh)!!2mM!!6&Oe9ICmd9 I have included below the issue page for the "relationship to PMF" issue. This issue is especially relevant to Carol, Rob, Ron - Carol because I would like a prediction of "would it fly at the the AB" on this issue, and Rob and Ron because of their closeness to the PMF spec. In particular, what do you all think of the "how useful or risky" question in the last paragraph? Summary: Is there, or can there be made, a formal relationship between PMF properties and PIDS traits? Partial Proposed Resolution: The PMF uses CosPropertyService interfaces for property manipulation. These properties are not explicitly namespace-qualified. typedef string PropertyName; struct Property { PropertyName property_name; any property_value; }; Therefore even if we qualify the PIDS HL7 2.3. trait names under a naming root, we will not be able to use TQS (LQS).to map between PIDS traits and PMF properties. How useful or risky is it to specify in an appendix of PIDS and/or PMF, that PMF property names that match the PIDS HL7 trait names are assumed to correspond? Note that since each of the predefined HL7 PIDs trait names begin with "HL7/", I think there is very little risk of name-colliding with locally conceived names in an invalid way; That is, where they match, they are supposed to match! From: "David W. Forslund" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit Date: Wed, 24 Nov 1999 08:20:44 -0700 (MST) To: "Jon Farmer" Cc: "pids-rtf2" , "Carol Burt" , "Ron Zahavi" , Subject: Re: relationship to PMF In-Reply-To: <000701bf3611$56260760$731e85cd@toledolink.com> References: <000701bf3611$56260760$731e85cd@toledolink.com> X-Mailer: VM 6.43 under 20.4 "Emerald" XEmacs Lucid Message-ID: <14395.65530.567134.873474@tbp.acl.lanl.gov> Reply-To: "David W. Forslund" Content-Type: text/plain; charset=us-ascii X-UIDL: `>Ke9IgHe9Q]>!!aG5e9 The other connection between PIDS traits and PMF properties could be done through the new get_supported_properties call we have proposed to add in a separate issue (2670). At least, the properties could be made consistent between them. The issue below is a bigger one involved with connecting CosProperties with TQS. It seems to me this is fairly straightforward if the PropertyName maps to QualifiedCodeStr in TQS. Then Property is simply a QualifiedCodeInfo (with an added type which defines read_only etc). However, harmonizing CosProperties with TQS would seem to me be out of scope for this RTF, although we could describe the mechanism, perhaps. What do you think? Dave Jon Farmer writes: > I have included below the issue page for the "relationship to PMF" issue. > > This issue is especially relevant to Carol, Rob, Ron - Carol because I would > like a prediction of "would it fly at the the AB" on this issue, and Rob and > Ron because of their closeness to the PMF spec. > > In particular, what do you all think of the "how useful or risky" question > in the last paragraph? > > > Summary: > Is there, or can there be made, a formal relationship between PMF properties > and PIDS traits? > Partial Proposed Resolution: > The PMF uses CosPropertyService interfaces for property manipulation. These > properties are not explicitly namespace-qualified. > > typedef string PropertyName; > struct Property { > PropertyName property_name; > any property_value; > }; > > Therefore even if we qualify the PIDS HL7 2.3. trait names under a naming > root, we will not be able to use TQS (LQS).to map between PIDS traits and > PMF properties. > > How useful or risky is it to specify in an appendix of PIDS and/or PMF, that > PMF property names that match the PIDS HL7 trait names are assumed to > correspond? Note that since each of the predefined HL7 PIDs trait names > begin with "HL7/", I think there is very little risk of name-colliding with > locally conceived names in an invalid way; That is, where they match, they > are supposed to match! > > Reply-To: "Jon Farmer" From: "Jon Farmer" To: "David W. Forslund" , "pids-rtf2" References: <000701bf3611$56260760$731e85cd@toledolink.com><14395.65530.567134.873474@tbp.acl.lanl.gov><00c601bf369b$9ab9c580$a81e85cd@toledolink.com> <14396.7444.547765.345802@tbp.acl.lanl.gov> Subject: Re: relationship to PMF Date: Wed, 24 Nov 1999 14:34:53 -0500 Organization: Care Data Systems MIME-Version: 1.0 Content-Transfer-Encoding: 8bit X-Priority: 3 X-MSMail-Priority: Normal X-Mailer: Microsoft Outlook Express 5.00.2314.1300 X-MimeOLE: Produced By Microsoft MimeOLE V5.00.2314.1300 Content-Type: text/plain; charset="iso-8859-1" X-UIDL: (i`d9S^W!!W>/!!Nk~e9 Here is what I propose for the revised issue page. Please QA my material regarding the specific mapping of struct elements. Also, I think we need to specify the qualifiedCodes for the HL7 an vCard traits. Any ideas there? Volunteers? Summary: Is there, or can there be made, a formal relationship between PMF properties and PIDS traits? If we were to directly map the predefined trait names to identically-named PMF properties, we would run the common risks of "rootless namespaces" and misinterpreted traits. Proposed Resolution: The PMF uses CosPropertyService interfaces for property manipulation. These properties are not explicitly namespace-qualified. typedef string PropertyName; struct Property { PropertyName property_name; any property_value; }; Our resolution to issue 2670 introduces properties for traits, through the new get_supported_properties operation. The PIDS and PMF properties, then, are a perfect basis for a symmetric trait-level relationship between the two specs. Furthermore, by mapping the combination of the property name and value to the TQS scheme and local code respectively, we have elegantly integrated not only PIDS, PMF, and TQS, but also any other present and future specs that need to interrelated unambiguously by properties even across disparate coding schemes. Specifically in relation to PIDS traits, this solution makes it easy for clients to not only directly compare traits across PIDS servers, but also to map them through an LQS /TQS if necessary. The property-to-TQS mapping is as follows: The Property.property_name maps to the TQS QualifiedCodeStr, making the property itself unambiguous in its definition; The Property.property.property_value maps to the TQS local code. Revised Text: Just following the narrative for get_supported_properties() as described in issue 2670, add the paragraph: PIDS clients and servers may compare trait definitions based on the combined property name and value. Trait definitions may also be mapped through and LQS/TQS as follows: 7 The Property.property_name maps to the TQS QualifiedCodeStr, making the property itself unambiguous in its definition; 7 The Property.property.property_value maps to the TQS local code. The property names and values for the predefined traits (HL7 2.3 and vCard elements) are standardized in their respectives modules of this specification. ----- Original Message ----- From: David W. Forslund To: Jon Farmer Cc: David W. Forslund Sent: Wednesday, November 24, 1999 12:20 PM Subject: Re: relationship to PMF > Jon Farmer writes: > > see inline > > ----- Original Message ----- > > From: David W. Forslund > > To: Jon Farmer > > Cc: pids-rtf2 ; Carol Burt > ; Ron > > Zahavi ; > > Sent: Wednesday, November 24, 1999 10:20 AM > > Subject: Re: relationship to PMF > > > > > > > The other connection between PIDS traits and PMF properties > could be > > > done through the new get_supported_properties call we have > proposed to > > > add in a separate issue (2670). > > > > Yes I think this is key. To use the (existing HL7/...) trait > names > > themselves we run a small but nonzero risk > > of rootless namespaces. By using the new properties we can > employ > > fully-qualified names. > > > > Exactly. There are actually some words in 2670 resolution to this > effect already. Perhaps we just beef them up a little. > > > > > At least, the properties could be made > > > consistent between them. > > > > > The issue below is a bigger one involved with > > > connecting CosProperties with TQS. It seems to me this is > fairly > > > straightforward if the PropertyName maps to QualifiedCodeStr in > > > TQS. Then Property is simply a QualifiedCodeInfo (with an added > type > > > which defines read_only etc). However, harmonizing > CosProperties with > > > TQS would seem to me be out of scope for this RTF, although we > could > > > describe the mechanism, perhaps. > > > > Given that properties are name-value pairs, we could have > > > > 1) property VALUES that are qualifiedCodeStrs > > This would require that we standardize the property name that > will carry the > > values. > > > > 2) combinations of NAME AND VALUE that map to the TQS scheme and > local code > > respectively. > > This would be far-reaching precedent but I think it is the most elegant. > > > I guess I like version 2, since a coded concept is the only thing > that > makes sense to return in a Property, with its PropertyName coming > from > TQS. > > I think the Name is the QualifiedCodeStr and the value is the local > code, > corresponding to TQS. I don't think this is any real problem, is > elegant, won't break anything and pulls the specifications (PIDS, > PMF, > and TQS) together quite nicely. The QualfiedCodeStr could be the HL7 > naming scheme indicated by COAS. I'm not sure we need to flesh out > the > supported codes at this time, but we could if we get ambitious in > the > next few weeks. > > Dave > > > > > > > What do you think? > > > > > > Dave > > > > > > > > > Jon Farmer writes: > > > > I have included below the issue page for the "relationship > to PMF" > > issue. > > > > > > > > This issue is especially relevant to Carol, Rob, Ron - Carol because I > > would > > > > like a prediction of "would it fly at the the AB" on this > issue, and > > Rob and > > > > Ron because of their closeness to the PMF spec. > > > > > > > > In particular, what do you all think of the "how useful or > risky" > > question > > > > in the last paragraph? > > > > > > > > > > > > Summary: > > > > Is there, or can there be made, a formal relationship > between PMF > > properties > > > > and PIDS traits? > > > > Partial Proposed Resolution: > > > > The PMF uses CosPropertyService interfaces for property manipulation. > > These > > > > properties are not explicitly namespace-qualified. > > > > > > > > typedef string PropertyName; > > > > struct Property { > > > > PropertyName property_name; > > > > any property_value; > > > > }; > > > > > > > > Therefore even if we qualify the PIDS HL7 2.3. trait names > under a > > naming > > > > root, we will not be able to use TQS (LQS).to map between > PIDS traits > > and > > > > PMF properties. > > > > > > > > How useful or risky is it to specify in an appendix of PIDS > and/or PMF, > > that > > > > PMF property names that match the PIDS HL7 trait names are > assumed to > > > > correspond? Note that since each of the predefined HL7 PIDs > trait names > > > > begin with "HL7/", I think there is very little risk of name-colliding > > with > > > > locally conceived names in an invalid way; That is, where > they match, > > they > > > > are supposed to match! > > > > > > > > > > > > > >