tamie message
[Date Prev]
| [Thread Prev]
| [Thread Next]
| [Date Next]
--
[Date Index]
| [Thread Index]
| [List Home]
Subject: Detailed points for discussion
- From: "Jacques R. Durand" <JDurand@us.fujitsu.com>
- To: <tamie@lists.oasis-open.org>
- Date: Tue, 24 Aug 2010 12:09:00 -0700
More precise items
to be discussed today:
1.
Specification:
- Should we worry
about event posting for V1? (this operation is not designed)
Or should it be
specified , just not in "core" lge.
- remove <nms-binding> from execution context because redundant?
(need to declare these namespaces in usual form
anyway)
-
review / complete the "function" design, examples.
- the
Event Model and Management section: how much is needed
here.
2. Use
Cases:
- establishing a
small set of canonical use cases, that cover the language capabilities, to be
developped in a "user guide".
- 2 of them already
in spec appendix.
- 3rd use case is an
extension of UC #2 in appendix, but with metrics (avg
time...)
- 2 other UC
candidates: (a) advanced test report parametrization (X-effect with inline
<eval> and more)
(b) a monitoring UC,
with loop kicking off a reporting scriplet e.g. every hour.
- is there other
valuable UCs we could add?
3.
Prototype:
What is missing - or
left out of implementation on purpose:
- <sleep>,
-
POST: not implemented - (do we want to? needs declaring output event
board)
-
<match min="" max=""> attributes?
-
bubbling of
exit.
Changes (some already done):
- @maxduration -> @tryfor
- identing of output.
- emam, etsm -->
xtemp
- @async --> @concurrent
Implementation question:
- all the
translator:cursor intermediate
outputs could be just not created
instead of removed afterward? (debugging
purpose?)
-jacques
[Date Prev]
| [Thread Prev]
| [Thread Next]
| [Date Next]
--
[Date Index]
| [Thread Index]
| [List Home]