Thanks for the feedbacks.
Jerome Dochez wrote:
> EmbeddedInfo is not generic enough, it asks for http ports but what
> about if you need to add more listener, or even what about if you
> don't want to open any port (like an embedded EJB Container).
I think the intent was to make it simple for typical use cases. People
can use CommandExecutor for more complicated configuration changes.
> We need one consistent way of defining 0 to many http listeners and ports.
>
Do you have a proposal for EmbeddedInfo? We can adjust it.
> EmbeddedDeployer is also not generic enough, the scatteredWar concept
> needs to be extended to an ScatteredArchive concept (which would map
> to jar, war, rar and maybe even ear) and the deployScattered() methods
> need to use that ScatteredArchive as a parameter. In fact, I am not
> sure why we need a deployScattered() method all together, I think
> deploy() should be enough and we should be smart enough to figure out
> what was passed to us.
Agreed. We should use ScatteredArchive and deploy method if possible.
>
> EmbeddedMain talks about using the uber jar, it should in fact be
> moved into the glassfish.jar so that people can do the same thing
> using a normal distribution (as they can already do today for most of
> the features of that class, today I can do java -jar glassfish.jar
> foo.war).
There is no glassfish.jar in this case. Are you asking for supporting
"java -jar glassfish.jar foo.war" use case also? Please note that for
Prelude, we are not supporting the implanted use case yet. When we
support "implanted" we can certainly use glassfish.jar.
>
> Server : why do we return the cataline Engine object ? I suppose this
> would be removed once we integrate the new web container embedded
> APIs, right ? If that's the case, we should not have here.
>
This is a bug. Byron/Jennifer: Could you please remove this?
> So as a conclusion, this APIs is very targeted to war file deployment,
> and we cannot hope to be able to extend it to support full V3 styles
> of deployment and maintain backward compatibility.
>
> Either we spend the time to make it right for embedded release or we
> declare it as a non stable (with knowledge that it will extensively
> change for v3 release).
Lets see what can be done. This is not throw away work.
--
Nazrul Islam - (408) 276-6468 - Sun Microsystems, Inc.
> Jerome
>
> On Mar 4, 2009, at 12:04 PM, Nazrul Islam wrote:
>
>> We are planning to wrap-up the development activities for GFv3
>> Prelude and move to GFv3 trunk. We would like your feedbacks on the
>> current APIs
>> <https://embedded-glassfish.dev.java.net/nonav/gf-embedded-api/apidocs/>.
>>
>> _Background on Current APIs_:
>> The APIs has gone through several iterations already. Thanks for
>> filing all the issues
>> <https://embedded-glassfish.dev.java.net/issues/buglist.cgi?Submit+query=Submit+query&issue_type=DEFECT&issue_type=ENHANCEMENT&issue_type=FEATURE&issue_type=TASK&issue_type=PATCH&component=embedded-glassfish&subcomponent=API&issue_status=RESOLVED&issue_status=VERIFIED&issue_status=CLOSED&version=current&email1=&emailtype1=exact&emailassigned_to1=1&email2=&emailtype2=exact&emailreporter2=1&issueidtype=include&issue_id=&changedin=&votes=&chfieldfrom=&chfieldto=Now&chfieldvalue=&short_desc=&short_desc_type=fulltext&long_desc=&long_desc_type=fulltext&issue_file_loc=&issue_file_loc_type=fulltext&status_whiteboard=&status_whiteboard_type=fulltext&field0-0-0=noop&type0-0-0=noop&value0-0-0=&cmdtype=doit&order=Issue+Number>.
>> The APIs has been reviewed internally also few times lately ([1
>> <https://embedded-glassfish.dev.java.net/issues/buglist.cgi?Submit+query=Submit+query&issue_type=DEFECT&issue_type=ENHANCEMENT&issue_type=FEATURE&issue_type=TASK&issue_type=PATCH&component=embedded-glassfish&subcomponent=API&issue_status=RESOLVED&issue_status=VERIFIED&issue_status=CLOSED&version=current&email1=&emailtype1=exact&emailassigned_to1=1&email2=ai109478&emailtype2=exact&emailreporter2=1&issueidtype=include&issue_id=&changedin=&votes=&chfieldfrom=&chfieldto=Now&chfieldvalue=&short_desc=&short_desc_type=fulltext&long_desc=&long_desc_type=fulltext&issue_file_loc=&issue_file_loc_type=fulltext&status_whiteboard=&status_whiteboard_type=fulltext&field0-0-0=noop&type0-0-0=noop&value0-0-0=&cmdtype=doit&order=Issue+Number>],
>> [2
>> <https://embedded-glassfish.dev.java.net/issues/show_bug.cgi?id=67>],
>> [3
>> <https://embedded-glassfish.dev.java.net/issues/show_bug.cgi?id=86>])
>> and is scheduled for a formal review at GlassFish architecture forum
>> on 3/16. Since we will be doing the webtier APIs on the trunk, for
>> GFv3 Prelude we focused on basics first such as start/stop,
>> configuration management, etc. Refer to attached presentation for a
>> brief overview of the APIs.
>>
>> Jerome: You mentioned that you may have some feedback. Please let us
>> know.
>>
>> Thanks much.
>>
>> _References_:
>> Javadocs:
>> https://embedded-glassfish.dev.java.net/nonav/gf-embedded-api/apidocs/
>> --
>> Nazrul Islam - (408) 276-6468 - Sun Microsystems, Inc.
>>
>> <Embedded_GlassFish_API.odp>
>