| From | Sent On | Attachments |
|---|---|---|
| Jacques R. Durand | Jun 23, 2009 10:54 am | |
| Bryan Aupperle | Jun 23, 2009 11:43 am | |
| Jacques R. Durand | Jun 23, 2009 2:29 pm | |
| Bryan Aupperle | Jul 16, 2009 9:44 am |
| Subject: | RE: [sca-c-cpp-comment] comments from Jacques D. | |
|---|---|---|
| From: | Jacques R. Durand (JDur...@us.fujitsu.com) | |
| Date: | Jun 23, 2009 2:29:39 pm | |
| List: | org.oasis-open.lists.sca-c-cpp-chair | |
Bryan:
I guess I am a little confused about how the distinction between Composite and Component plays in the C++ implementation model:. Do you consider "composite" as just an implementation concept only? (i.e. no dedicated class). I see composite mentioned as a COmponent implementation "scope", so it seems it does not have a construct or class on its own. If that is the case that should be more clearly stated, as in the Assembly mark-up it appears that a component is always used inside a Composite - and not by itself. So that would also address my question about the "promotion" concept in Assembly that relates the Services of a composite to the Service of a component inside.
Other comments:
- it is unclear what the notion of "SCA runtime" corresponds to in C++. Is there a particular framework or container (in C++) to manage components? For example, what entity is raising SCA Exceptions ? (as opposed to business exceptions).
- The ServiceProxy base class is empty... is it really needed?
- it looks like (in PDF) the table of contents is not uptodate: section 6.5 is announced as SCAExceptions, but it is actually 6.6.
Regards, Jacques
________________________________
From: Bryan Aupperle [mailto:aupp...@us.ibm.com] Sent: Tuesday, June 23, 2009 11:44 AM To: Jacques R. Durand Cc: sca-...@lists.oasis-open.org Subject: Re: [sca-c-cpp-comment] comments from Jacques D.
Thank-you for comment on the formatting. Definitely a PDF generation problem.
Before the TC can fully consider your comment on promotion, would you be able to elaborate on what you think may be needed? The promotion of services and reverences as defined in the Assembly specification is completely invisible to a component implementation and thus would not be reflected in an API or component/componentType definition.
Bryan Aupperle, Ph.D. STSM, WebSphere Enterprise Platform Software Solution Architect Chair OASIS SCA-C-CPP TC
Research Triangle Park, NC +1 919-254-7508 (T/L 444-7508) Internet Address: aupp...@us.ibm.com
"Jacques R. Durand" <JDur...@us.fujitsu.com>
06/23/2009 01:55 PM
To <sca-...@lists.oasis-open.org> cc Subject [sca-c-cpp-comment] comments from Jacques D.
The spec appears to cover a lot of ground for C/C++ developers. My only comments so far:
- in the PDF, title of section 6.4 seems to not be right (formatting issue?)
- the "promotion" mechanism described in Assembly specification, does not seem to be addressed here.
Jacques Durand Fujitsu





