cordova-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Joe Bowser <bows...@gmail.com>
Subject Re: Found a regression in cordova-android@3.7.0
Date Wed, 04 Feb 2015 01:26:08 GMT
BTW: The issue was that the patch made a change and CordovaActivity longer
implemented the CordovaInterface, but still passed itself.  I fixed it and
re-added the code with some comments. I think that we really should have
unit tests to make sure that we have a CordovaInterfaceImpl, since that
seems like a spot where the code could break, especially when dealing with
non-CTS AOSP builds where just about anything can happen.

On Tue Feb 03 2015 at 5:08:16 PM Steven Gill <stevengill97@gmail.com> wrote:

> Cool. Thanks for looking into this Joe. I will get the vote thread for
> 3.7.1 started tomorrow.
>
> -Steve
>
> On Tue, Feb 3, 2015 at 5:04 PM, Joe Bowser <bowserj@gmail.com> wrote:
>
> > BTW: This works fine on 3.7.x, the issue was on master. I reverted
> > CordovaActivity.java only so that master now builds again, but master and
> > 4.0.x have now merged.  We can proceed with doing a 3.7.1 release.
> >
> > On Tue Feb 03 2015 at 4:55:18 PM Joe Bowser <bowserj@gmail.com> wrote:
> >
> > > OK, I see what happened. I think we outsmarted ourselves.  I'm going to
> > > revert the API change that Andrew and I talked about this morning,
> since
> > it
> > > seems that's currently what's breaking on master.  This is why we need
> to
> > > make sure we run a sanity check every time we do a commit, since we
> could
> > > shoot ourselves in the foot, like what happened here.
> > >
> > > On Tue Feb 03 2015 at 4:41:32 PM Joe Bowser <bowserj@gmail.com> wrote:
> > >
> > >> Is this an issue on master, or on 3.7.x.  If it's only master, it
> > >> shouldn't hold up the .1 release.
> > >>
> > >> Also, Andrew, you didn't even build before committing this?
> > >>
> > >> On Tue Feb 03 2015 at 3:38:25 PM Steven Gill <stevengill97@gmail.com>
> > >> wrote:
> > >>
> > >>> Nice catch. I will patiently wait until this is ready before I move
> on
> > >>> this
> > >>> release (release issue: https://issues.apache.org/
> jira/browse/CB-8415)
> > >>>
> > >>> On Tue, Feb 3, 2015 at 3:31 PM, Murat Sutunc <muratsu@microsoft.com>
> > >>> wrote:
> > >>>
> > >>> > Looks like this fix broke `cordova build` for windows:
> > >>> >
> > >>> > :CordovaLib:compileDebugJava
> > >>> > C:\dev\test\platforms\android\CordovaLib\src\org\apache\cord
> > >>> ova\CordovaActivity.java:151:
> > >>> > error: method init in interface CordovaWebView cannot be applied
to
> > >>> given
> > >>> > types;
> > >>> >         appView.init(this, pluginEntries, internalWhitelist,
> > >>> > externalWhitelist, preferences);
> > >>> >                ^
> > >>> >   required:
> > >>> > CordovaInterface,List<PluginEntry>,Whitelist,Whitelist,Cordo
> > >>> vaPreferences
> > >>> >   found:
> > >>> > CordovaActivity,ArrayList<PluginEntry>,Whitelist,Whitelist,C
> > >>> ordovaPreferences
> > >>> >   reason: actual argument CordovaActivity cannot be converted
to
> > >>> > CordovaInterface by method invocation conversion
> > >>> > C:\dev\test\platforms\android\CordovaLib\src\org\apache\cord
> > >>> ova\CordovaActivity.java:217:
> > >>> > error: cannot find symbol
> > >>> >         return ret;
> > >>> >                ^
> > >>> >   symbol:   variable ret
> > >>> >   location: class CordovaActivity
> > >>> > Note: Some input files use or override a deprecated API.
> > >>> > Note: Recompile with -Xlint:deprecation for details.
> > >>> > 2 errors
> > >>> > :CordovaLib:compileDebugJava FAILED
> > >>> >
> > >>> > FAILURE: Build failed with an exception.
> > >>> >
> > >>> > * What went wrong:
> > >>> > Execution failed for task ':CordovaLib:compileDebugJava'.
> > >>> > > Compilation failed; see the compiler error output for details.
> > >>> >
> > >>> > * Try:
> > >>> > Run with --stacktrace option to get the stack trace. Run with
> --info
> > or
> > >>> > --debug option to get more log output.
> > >>> >
> > >>> > BUILD FAILED
> > >>> >
> > >>> > Total time: 3.117 secs
> > >>> >
> > >>> > C:\dev\test\platforms\android\cordova\node_modules\q\q.js:126
> > >>> >                     throw e;
> > >>> >                           ^
> > >>> > Error code 1 for command: cmd with args: /s /c
> > >>> > "C:\dev\test\platforms\android\gradlew cdvBuildDebug -b
> > >>> > C:\dev\test\platforms\android\build.gradle
> -Dorg.gradle.daemon=true"
> > >>> > ERROR building one of the platforms: Error:
> > >>> > C:\dev\test\platforms\android\cordova\build.bat: Command failed
> with
> > >>> exit
> > >>> > code 8
> > >>> > You may not have the required environment or OS to build this
> project
> > >>> >
> > >>> > -----Original Message-----
> > >>> > From: Joe Bowser [mailto:bowserj@gmail.com]
> > >>> > Sent: Tuesday, February 3, 2015 1:41 PM
> > >>> > To: dev@cordova.apache.org
> > >>> > Subject: Re: Found a regression in cordova-android@3.7.0
> > >>> >
> > >>> > Was the plugin using public APIs? I'm fine with a 3.7.1, but I
am
> > less
> > >>> OK
> > >>> > with calling this a regression if we never supported this.
> > >>> >
> > >>> > On Tue, 3 Feb 2015 1:38 pm Steven Gill <stevengill97@gmail.com>
> > wrote:
> > >>> >
> > >>> > > Okay, lets get android 3.7.1 vote started, update blog post
for
> > 3.7.1
> > >>> > > and publish it when vote is successful. Do tools release
> following
> > >>> > > completion of 3.7.1 vote with 3.7.1 pinned.
> > >>> > >
> > >>> > > On Tue, Feb 3, 2015 at 1:09 PM, Andrew Grieve <
> > agrieve@chromium.org>
> > >>> > > wrote:
> > >>> > >
> > >>> > > > Regression: https://issues.apache.org/jira/browse/CB-8411
> > >>> > > >
> > >>> > > > Already applied the fix to the 3.7.x branch (and to
master).
> > >>> > > >
> > >>> > > > Question is whether we should skip pinning of 3.7.0,
and go
> > >>> straight
> > >>> > > > to 3.7.1? Wouldn't take much extra effort, but we should
hold
> off
> > >>> > > > the blog post though.
> > >>> > > >
> > >>> > > > The bug breaks Telerik's NativePageTransitions plugin
(
> > >>> > > >
> > https://github.com/Telerik-Verified-Plugins/NativePageTransitions)
> > >>> > > > and potentially others?
> > >>> > > >
> > >>> > > > Thoughts?
> > >>> > > >
> > >>> > > >
> > >>> > > > The good news is that I found the bug by adding a
> > >>> > > > --thirdpartyplugins
> > >>> > > flag
> > >>> > > > to createmobilespec.js, which just adds in a bunch of
other
> > popular
> > >>> > > plugins
> > >>> > > > to the project. So, finding these sorts of breakages
in the
> > future
> > >>> > > > will
> > >>> > > be
> > >>> > > > easier.
> > >>> > > >
> > >>> > >
> > >>> >
> > >>>
> > >>
> >
>

Mime
  • Unnamed multipart/alternative (inline, None, 0 bytes)
View raw message