[arquillian-issues] [JBoss JIRA] Commented: (ARQ-126) rename RunModeType.REMOTE to RunModeType.IN_CONTAINER

Fri May 7 14:32:05 EDT 2010

    [ https://jira.jboss.org/jira/browse/ARQ-126?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12529664#action_12529664 ] 

Andrew Lee Rubinger commented on ARQ-126:
-----------------------------------------

(02:25:04 PM) aslak: ALR, comments on ARQ-126 ?
(02:25:05 PM) jbossbot: [ARQ-126] rename RunModeType.REMOTE to RunModeType.IN_CONTAINER [Open, Major, Unassigned] http://jira.jboss.org/jira/browse/ARQ-126
(02:25:48 PM) ALR at Freenode: aslak: Does that affect packaging?
(02:26:14 PM) aslak: RunMode?
(02:26:22 PM) ALR at Freenode: Or the proposed @Run, yes
(02:26:28 PM) aslak: yes
(02:26:43 PM) ALR at Freenode: aslak: Then IMO it should be atop @Deployment
(02:26:54 PM) ALR at Freenode: Where the deployable archive is declared
(02:27:10 PM) ALR at Freenode: Else I find it unclear what's gonna happen under the covers.
(02:27:14 PM) aslak: RunMode exists now, but the same arquillian packaging happens either way you run it, just that its executed locally
(02:27:20 PM) ALR at Freenode: Right.
(02:27:30 PM) ALR at Freenode: Your last comments work best IMO from a grammar perspective.
(02:27:32 PM) aslak: ARQ_134
(02:27:36 PM) ALR at Freenode: @Run(IN_CONTAINER) 
(02:27:36 PM) aslak: ARQ-134
(02:27:36 PM) jbossbot: [ARQ-134] RunMode LOCAL changes deployment. [Open, Major, Aslak Knutsen] http://jira.jboss.org/jira/browse/ARQ-134
(02:27:42 PM) ALR at Freenode: @Run(AS_CLIENT)
(02:27:46 PM) ALR at Freenode: But it affects packaging
(02:28:00 PM) ALR at Freenode: And I've always felt that becomes a property of the @Deployment
(02:28:09 PM) ALR at Freenode: Because in effect we change the deployment. :)
(02:28:10 PM) aslak: hmmm
(02:28:38 PM) aslak: i see your point
(02:28:50 PM) aslak: it effects more then just packaging tho, but
(02:28:55 PM) ALR at Freenode: Yeah, I know. :)
(02:29:11 PM) ALR at Freenode: Again, I love the in-container feature.
(02:29:29 PM) ALR at Freenode: Would love even more to find a way to make it plain what we're doing to enable it though
(02:29:46 PM) ALR at Freenode: Because of the lack of EAR support, changing JNDI names, etc... everything that results from wrapping.

> rename RunModeType.REMOTE to RunModeType.IN_CONTAINER
> -----------------------------------------------------
>
>                 Key: ARQ-126
>                 URL: https://jira.jboss.org/jira/browse/ARQ-126
>             Project: Arquillian
>          Issue Type: Bug
>    Affects Versions: 1.0.0.Alpha2
>            Reporter: Dan Allen
>
> I think that the combination of the terms LOCAL and REMOTE in the RunModeType enum is confusing. RunModeType is attempting to describe where the test runs in relationship to the test runner. In that case, the terms are defensible.
> LOCAL = test is local to the test runner (same JVM)
> REMOTE = test is remote from the test runner (different JVM)
> In the case of LOCAL, the test must use remote communication to invoke the code under test. This is where we overuse the term "remote". The term "remote" implies remote communication, but it's not clear what is remote from the standpoint of the test writer. We are describing internal behavior. And besides, REMOTE isn't entirely accurate since in embedded containers, no remote communication is used.
> It's more natural to describe where the test is in relation to the code under test. In that case, local means that the test case stays behind. The other term would be in-container, which may or may not be remote. It's about the packaging. The test is with the test code executing in the container.
> I'd like to propose these constants instead:
> public enum RunModeType
> {
>    /**
>     * In LOCAL mode, the @Deployment is processed and deployed to the Container, 
>     * but the test is not executed inside the container. It's local to the test runner.
>     */
>    LOCAL,
>    
>    /**
>     * In IN_CONTAINER mode, the @Deployment is processed and deployed to the container along side the test
>     * case and the test case is executed inside the container.<br/>
>     * This is the default mode when none specified. 
>     */
>    IN_CONTAINER
> }
> Perhaps I'm just thinking about it wrong, but I just can't get my mind to believe that @RunMode(REMOTE) means that the test case is running in the container. I get how it is technically correct, but human tendency leads me to believe we are going to have to constantly clarify this. @RunMode(IN_CONTAINER) makes it clear that the test is getting relocated to the container.
> (I'm even open to changing LOCAL to NORMAL if you think it helps the terminology)

-- 
This message is automatically generated by JIRA.
-
If you think it was sent incorrectly contact one of the administrators: https://jira.jboss.org/jira/secure/Administrators.jspa
-
For more information on JIRA, see: http://www.atlassian.com/software/jira