Node Build and Configuration Notebook - page 5 of 55

First PagePrevious PageNext PageLast PageTable of ContentsSearch

Date and Author(s)

Abstract component plan

Here is the basic set of steps that I think we will be following for each of our components.  
 1. Limit the scope of the component  
 2. get straw man XML schema proposed  
    This only needs to be very preliminary as the next step may   
    substantially change it.  
 3. Analyze implicit information service usage in inter-component   
    messages. (we need to follow whatever explicit usage policy  
    we set for users of the information service, see info serv   
    specific issues for more complete summary)  
 4. Dummy implementation(s) - This implementation will only process  
    and generate XML. (no real component-specific functionality)  
    This will help to coarsely refine the xml schema and draw   
    attention to initial (probably large) shortcomings in it.  
    dunno if multiples of this will be required. This step can  
    be pursued in parallel with step 4; this will allow the XML  
    schema refinement to no block on functionality implementation,   
    as several components will be substantial new development   
 5. Reference implementations - fit the xml interfaces into   
    components actually providing the functionality. The hope is   
    that step 3 will help to provide most of the necessary  
    schema refinement, while this step will mainly produce working  
    code. These final two steps will probably be repeated several  
    times interleaved with several generations of information service   
    usage and schema redesign.  
 6. Inter-component usage testing -> Doug's domain?