This wiki has undergone a migration to Confluence found Here
Difference between revisions of "Proposal: create a query mechanism which specifies templates for the response"
Jump to navigation
Jump to search
Rene spronk (talk | contribs) (New page: ==Summary== *If the receiver of a query is some sort of archive (e.g. an EHR, or a Datamart) which could potentially respond to a query using a response structure that is extremely rich in...) |
Rene spronk (talk | contribs) |
||
| Line 1: | Line 1: | ||
| + | <div class="messagebox cleanup metadata"> | ||
| + | <div style="float: left;">[[Image:OpenHotTopic.GIF|35px| ]]</div> | ||
| + | <div style="background:#F0F0F0"> | ||
| + | This page documents an [[:category:MnM Open Hot Topic|Open]] [[:category:MnM Hot Topic|MnM Hot Topic]], i.e. a methodology issue requiring a resolution from Modeling and Methodology. | ||
| + | </div> | ||
| + | </div> | ||
| + | [[Category:MnM Hot Topic]] | ||
| + | [[Category:MnM Open Hot Topic]] | ||
| + | |||
==Summary== | ==Summary== | ||
*If the receiver of a query is some sort of archive (e.g. an EHR, or a Datamart) which could potentially respond to a query using a response structure that is extremely rich in detail: how does one [the party that sends the query] "limit" the response to a certain more minimalistic subset? | *If the receiver of a query is some sort of archive (e.g. an EHR, or a Datamart) which could potentially respond to a query using a response structure that is extremely rich in detail: how does one [the party that sends the query] "limit" the response to a certain more minimalistic subset? | ||
| + | **The use-case was raised in a SOA environment using Query-By-Example | ||
*This proposal seeks to add a new attribute (see discussion section for the class in which it should be added) of type SET<II> to allow the query placer to specify a list of templateIds the response model should be compliant with. | *This proposal seeks to add a new attribute (see discussion section for the class in which it should be added) of type SET<II> to allow the query placer to specify a list of templateIds the response model should be compliant with. | ||
==Rejected Alternatives== | ==Rejected Alternatives== | ||
| − | *The current dynamic model requires that each and every Trigger Event (and hence: Interaction) have exactly one static response model. Thus, if we had 7 level-of-richness templates for a response model, we'd have to define 7 query interactions [all with the exact same static model] and 7 response interactions. | + | *Different interaction for each response-template: The current dynamic model requires that each and every Trigger Event (and hence: Interaction) have exactly one static response model. Thus, if we had 7 level-of-richness templates for a response model, we'd have to define 7 query interactions [all with the exact same static model] and 7 response interactions. |
| + | *Query parameter to specify templates: | ||
| + | **See [[Nature of ParameterItem]] - it clashes with the nature of PatremeterItem | ||
| + | **It doesn't offer a solution for Query-By-Example (QBE) scenarios, where the query model is comprised of Acts with isCriterionInd set to true. QBE doesn't have query paremeters. | ||
==Discussion== | ==Discussion== | ||
*See [[Harmonization: add QuerySpec.responseTemplateId attribute]], this effectively reopens that old discussion. | *See [[Harmonization: add QuerySpec.responseTemplateId attribute]], this effectively reopens that old discussion. | ||
Revision as of 12:56, 15 August 2010
This page documents an Open MnM Hot Topic, i.e. a methodology issue requiring a resolution from Modeling and Methodology.
Summary
- If the receiver of a query is some sort of archive (e.g. an EHR, or a Datamart) which could potentially respond to a query using a response structure that is extremely rich in detail: how does one [the party that sends the query] "limit" the response to a certain more minimalistic subset?
- The use-case was raised in a SOA environment using Query-By-Example
- This proposal seeks to add a new attribute (see discussion section for the class in which it should be added) of type SET<II> to allow the query placer to specify a list of templateIds the response model should be compliant with.
Rejected Alternatives
- Different interaction for each response-template: The current dynamic model requires that each and every Trigger Event (and hence: Interaction) have exactly one static response model. Thus, if we had 7 level-of-richness templates for a response model, we'd have to define 7 query interactions [all with the exact same static model] and 7 response interactions.
- Query parameter to specify templates:
- See Nature of ParameterItem - it clashes with the nature of PatremeterItem
- It doesn't offer a solution for Query-By-Example (QBE) scenarios, where the query model is comprised of Acts with isCriterionInd set to true. QBE doesn't have query paremeters.
Discussion
- See Harmonization: add QuerySpec.responseTemplateId attribute, this effectively reopens that old discussion.
