It might be wise to decide how the term 'use case' will be used in the primer. What GGF usually has seems to me to be "examples of use". It sounds like the suggestion here is along the lines of "rationale for use". In the context I'm most familiar with, one coming from software development methodologies, use case is closer to "what is the workflow associated with this item" and "what objects does this workflow disclose". From my (somewhat persnickity) perspective, it's only the latter two meanings that lead to identify how one would actually use a tool and identify what protocols it would need to follow. Developing such use cases is also a lot more work, since there's a need to identify the environment of objects within which the scenarios associated with the use case work. Bruce R. Barkstrom Robert E. McGrath wrote:
Greetings,
As many of you know, I have been tasked to compile a revised "primer" for DFDL. Based on the draft spec. and other information, I have made some good progress, but there are a number of topics that I don't have enough information on.
This note summarizes areas where I need substantial contributions from you all.
This may not be things to tackle via the mail list, I don't know. Perhaps we can discuss these issues at the F2F.
1. Use cases
Actually, there should be a "Use Cases" tracker and a separate UC document, IMO.
By "Use Case" I mean statements of how DFDL is intended to be used. This is different from the current example suites which illustrate the language but do not discuss why you need DFDL or what it helps you do.
More on this one later. I'm developing a note with some interesting use cases I'd like to discuss.
2. Overview of the language
I would like the primer to say "the DFDL is the following XSD elements, with the following restrictions. DFDL defines the following tags and attributes". Just a summary, circa 2 pages.
I'm sure this informtion is implicit in the current spec and demo software, but I need help to write a clear summary.
3. Processing model
I would like the primer to explain how DFDL is used, i.e.,
Parse DFDL, which tells how to parse data, ...
Again, this is probably implicit in current work, but I need a short and clear summary.
4. Explanation of configurations, layers etc.
Related to above. Explain the non-trivial processing of DFDL.
Some of this awaits the outcome of trackers.