[USTF] con-call minutes

Present

Ayse Dilber
Frank McCabe
David Orchard
Igor Sedukin
Heather Kreger
Hao He
Mario Jeckle
Chris Ferris - chair/scribe
Roger Cutler
Sharad Garg
Mark Hapner

Summary:

More discussion on whether a Use Case or a Usage Scenario is the
high-level thing. Much back and forth (Use Case seems intuitively
like the lower level thing and Usage Scenario suggests the higher-level
thing). However, we pulled out many defs from Rational, OMG, etc.
and there seemed to be a pretty good case built for Use Case as the
higher-level concept. Chair decided that rather than loose more time
on this issue, that we use Use Case to mean the high-level and Usage
Scenario to mean a path, aspect or instance of a Use Case. The Use Case
document will provide a clear description of these terms so that the
reader is not confused. Should cite references such as those pulled
out on the call.

Hao and Igor presented their classification proposals and there was
more discussion resulting in some general agreement that it would be
a good thing(tm) to classify Use Cases (and scenarios) with the
conceptual and technical "classifiers". Mark Hapner suggested three
types of relationship; inclusion, generalization and extension which
seemed to have some traction. See Mark's email[1].

Roger presented where he was with his EDI Use Case(s) and indicated
that more analysis was forthcoming. He indicated that Hugo is trying to
contact someone who can take the HTML and turn it into xmlspec for
inclusion in the Use Case document.

Detailed IRC log follows below.

[1] http://lists.w3.org/Archives/Public/www-ws-arch/2002Jun/0038.html

Cheers,

Chris

[17:03] <chris> what are we going to call these things?
[17:05] <chris> e.g. rational has use cases (high level) and path through that is called usage scenario
[17:07] <chris> cf: can we pick some names and move on?
[17:08] <chris> mj: not in favor of calling security a use scenario
[17:10] * chris sez no problem
[17:10] * chris zakim, who's here?
[17:12] <chris> if we choose these names, need to be explained
[17:21] <chris> cf: chair sez that we will go with high-level=use case lower-level or 
instance/path=scenario
[17:21] *** dbooth has quit IRC
[17:22] <chris> per roger clearly in the use case document
[17:23] <frankmcca> The UML reference manual:
[17:23] <chris> http://lists.w3.org/Archives/Member/w3c-ws-arch/2002May/0246.html
[17:25] <chris> hao: thought that original doc too technical/detailed, need high-level classifications
[17:27] <chris> daveo: take use cases and put them in the use case doc, call it sect 0, when we get 
down to usage scenario (sec, reliability, messaging, etc.) we reference back to use case
[17:27] <chris> hao: yes, that sounds reasonable
[17:29] <chris> igor: says he put out email that added conceptual level/categorization e.g. 
intermediary which is neither technical nor business level
[17:30] *** dougb has quit IRC (User pushed the X - because it's Xtra, baby)
[17:30] <chris> http://lists.w3.org/Archives/Public/www-ws-arch/2002May/0455.html
[17:31] <chris> igor's proposal for business, conceptual and technical classification of use cases
[17:32] <chris> do: is the notion of concept an attribute?
[17:32] <chris> igor, would like to have classification
[17:34] <chris> do: stock quote is use case
[17:35] <chris> igor: want tree or list when click on and get list of use cases (or usage scenarios) 
that map to that
[17:35] <chris> do: let's keep that in mind as we go forward
[17:36] <chris> do: let's actually start working on these things
[17:36] <chris> do: definitely something there
[17:37] <chris> markh: typically 3 kinds of (in)formal relationships; incusion, generalization, 
extension
[17:37] <chris> extension is more formal, take use case and add elements to it
[17:37] <chris> mark: if we organize these things, use typical mechanisms
[17:38] <chris> igor: discussed including all wsd use cases, only way to do that is take their doc 
and dump them in. doc becomes 600 pages of who knows what
[17:42] <chris> do: scenarios doc already has ids
[17:43] <chris> do: chose numbers or strings, don't care
[17:45] <chris> igor: how do we assign these numbers?
[17:45] <chris> do: went by what structure was in the document
[17:46] <frankmcca> Sorry, gotto go
[17:46] *** frankmcca has quit IRC
[17:47] <chris> do: maybe bit of issue in terms of the ids
[18:00] <Zakim> WS_ArchWG()3:30PM has ended

Received on Thursday, 6 June 2002 18:42:34 UTC