cordova-issues mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "ASF GitHub Bot (JIRA)" <j...@apache.org>
Subject [jira] [Commented] (CB-7067) run jasmine tests individually
Date Thu, 03 Jul 2014 13:52:24 GMT

    [ https://issues.apache.org/jira/browse/CB-7067?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14051449#comment-14051449
] 

ASF GitHub Bot commented on CB-7067:
------------------------------------

Github user kamrik commented on the pull request:

    https://github.com/apache/cordova-lib/pull/51#issuecomment-47931920
  
    I really like the results it produces!
    But I'm somewhat surprised by the amount of code. The diff with jasmine cli.js looks pretty
big, which probably means that much of the code _is_ new, or did it come from some place else?
    
    We should write a big comment first thing in that file, that it should be spun off as
an independent package soon.
    
    How does it go with license compatibility, can we incorporate MIT licensed files or parts
of files?
    
    The other option os to just publish it as an independent package right away.
    
    Here is an independent superspawn package [1], we may choose to depend on it or just bundle
the required version of superspawn.js file. 
    
    [1] https://www.npmjs.org/package/superspawn


> run jasmine tests individually
> ------------------------------
>
>                 Key: CB-7067
>                 URL: https://issues.apache.org/jira/browse/CB-7067
>             Project: Apache Cordova
>          Issue Type: Bug
>          Components: CordovaLib
>    Affects Versions: 3.5.0
>            Reporter: Josh Soref
>            Assignee: Josh Soref
>
> The way we run tests right now has a couple of painful points:
> 1. We can't skip only some tests in a single spec file because jasmine is treating all
of our tests as a single set run together, and so by saying "I want to prefer test X in file
A", it's interpreting that as "I don't want to run any other tests in any other file unless
they are also preferred". That's pretty annoying.
> 2. Spys and other manipulations we do in a spec can have adverse impact on unrelated
tests. Trying to get tests to work thus involves debugging all other tests and hunting for
common problems.
> Neither of these things pain ports are necessary afaict.



--
This message was sent by Atlassian JIRA
(v6.2#6252)

Mime
View raw message