[ 
https://issues.apache.org/jira/browse/CB-4096?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13940907#comment-13940907
 ] 

Ian Clelland commented on CB-4096:
----------------------------------

[~mrbillau], are you referring to [this ML 
thread|http://callback.markmail.org/thread/lqadhj3uqqtmohka]? I thought that 
the problem of two competing formats was raised, and that the consensus at the 
end was to go with "*.origin.com" as the single standard. (And the parent of 
this issue, CB-4093, has the description 'Ditch the subdomains="true" 
attribute')

I didn't realize it still parsed it; I think that code should be removed, but 
was probably left in for a very weak sort of backwards compatibility -- just 
not breaking existing apps.

Digging back through the repo looking for history, support for the attribute 
has been there since the very first Android whitelist commit (80df4a8, Sept. 
2011) but doesn't appear to ever have been documented.

I'm not convinced that it makes sense to re-introduce support for this 
undocumented flag, 8 months later, in the name of backwards compatibility. 
Anybody who has used Cordova since v2.9 (when it was removed) has already had 
to deal with it not working, and would have switched to "*.origin.com" already. 
The only people who would get the backwards compatibility benefits now would be 
those who are upgrading directly from =<2.8 to 3.5.

Could we belatedly deprecate this, without changing current behaviour, by 
printing a notice if the subdomains attribute is used, on "cordova prepare", or 
in the Android log?

If we add it back now, I think we should immediately deprecate it, and then 
completely remove it in Cordova 3.8 or so.

> Android - Unify Whitelist Implemenations
> ----------------------------------------
>
>                 Key: CB-4096
>                 URL: https://issues.apache.org/jira/browse/CB-4096
>             Project: Apache Cordova
>          Issue Type: Sub-task
>          Components: Android
>    Affects Versions: 3.0.0
>            Reporter: Andrew Grieve
>            Assignee: Ian Clelland
>             Fix For: 3.1.0
>
>




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

Reply via email to