> Just quickly, I noticed the gist contains a different set of namespaces,
> therefore excluding all of our new components, and still had a theme
> pointing at spark.css.  The flex-config.xml in a FlexJS SDK has a
> different set of manifest.xml files and nothing in the theme tag.

Not sure at all it would help, my guess is that it has its own rule to write 
its config.xml, the only way I guess would be a plugin to override its 
behaviour.

> And for the theme, maybe an empty Spark.css will work.

I can try that

Frédéric THOMAS

> From: aha...@adobe.com
> To: dev@flex.apache.org
> Subject: Re: AW: AW: AW: [FlexJS] IntelliJ Integration
> Date: Thu, 28 May 2015 12:58:45 +0000
> 
> Hi Fred,
> 
> Just quickly, I noticed the gist contains a different set of namespaces,
> therefore excluding all of our new components, and still had a theme
> pointing at spark.css.  The flex-config.xml in a FlexJS SDK has a
> different set of manifest.xml files and nothing in the theme tag.
> 
> I wonder if leaving a basic-manifest.xml in the SDK folder will convince
> IJ to use that, or if we have to copy basic-manifest into mxml-manifest.
> 
> And for the theme, maybe an empty Spark.css will work.
> 
> Thanks for digging into this,
> -Alex
> 
> On 5/28/15, 3:57 AM, "Frédéric THOMAS" <webdoubl...@hotmail.com> wrote:
> 
> >Actually, it is more complicated.
> >
> >This is what it is generated by IJ:
> >
> >https://gist.github.com/doublefx/ae135f1886cc2800c08d
> >
> >And I added the flex-config.xml from the framework, this one overrides
> >the one of IJ apparently
> >
> >it results:
> >
> >Information:28/05/2015 11:16 - Compilation completed with 7 errors and 11
> >warnings
> >U:\sources\flexJsProjects\DataBindingTest\src\DataBindingTest.mxml
> >Error:(26, 19) [DataBindingTest]: Call to a possibly undefined method
> >MyModel.
> >Warning:(30, 15) [DataBindingTest]: Definition models.MyModel could not
> >be found.
> >Error:(40, 8) [DataBindingTest]: This tag could not be resolved to an
> >ActionScript class. It will be ignored.
> >U:\sources\flexJsProjects\DataBindingTest\src\MyInitialView.mxml
> >Warning:(26, 10) [DataBindingTest]: Definition models.MyModel could not
> >be found.
> >Error:(53, 24) [DataBindingTest]: Call to a possibly undefined method
> >MyModel.
> >Error:(54, 27) [DataBindingTest]: Call to a possibly undefined method
> >MyModel.
> >Warning:(119, 50) [DataBindingTest]: Data binding will not be able to
> >detect assignments to 'MyModel'.
> >Warning:(119, 76) [DataBindingTest]: Data binding will not be able to
> >detect assignments to 'stockSymbol'.
> >Warning:(123, 63) [DataBindingTest]: Data binding will not be able to
> >detect assignments to 'MyModel'.
> >Warning:(123, 89) [DataBindingTest]: Data binding will not be able to
> >detect assignments to 'responseText'.
> >Warning:(131, 50) [DataBindingTest]: Data binding will not be able to
> >detect assignments to 'MyModel'.
> >Warning:(131, 76) [DataBindingTest]: Data binding will not be able to
> >detect assignments to 'strings'.
> >Warning:(143, 30) [DataBindingTest]: Data binding will not be able to
> >detect assignments to 'MyModel'.
> >Warning:(143, 56) [DataBindingTest]: Data binding will not be able to
> >detect assignments to 'allData'.
> >U:\sources\flexJsProjects\DataBindingTest\src\controllers\MyController.as
> >Warning:(26, 11) [DataBindingTest]: Definition models.MyModel could not
> >be found.
> >Error:(61, 3) [DataBindingTest]: Call to a possibly undefined method
> >MyModel.
> >Error:(66, 3) [DataBindingTest]: Call to a possibly undefined method
> >MyModel.
> >Error:(71, 12) [DataBindingTest]: Call to a possibly undefined method
> >MyModel.
> >
> >But if I pass directly to Mxmlc :
> >
> >-load-config="D:\SDKs\FlexJS\0.0.3_FP16.0_AIR16.0_en_US\frameworks\flex-co
> >nfig.xml"
> >-output=U:\sources\flexJsProjects\DataBindingTest\bin\DataBindingTest.swf
> >-library-path+=U:/sources/flexJsProjects/DataBindingTest/libs/libTest.swc
> >U:\sources\flexJsProjects\DataBindingTest\src\DataBindingTest.mxml
> >
> >It builds.
> >
> >And if I ommit the flex-config.xml from the framework, it results in
> >compilation error because it can't find mxml-2009-manifest.xml neither
> >the bundle resources, actually, the locale folder in my FlexJS SDK is
> >empty, it I fill it with the locales of my Flex SDK, the compilation
> >fails because the 
> >it can't find the components from the
> >D:\SDKs\FlexJS\0.0.3_FP16.0_AIR16.0_en_US\frameworks\themes\Spark\spark.cs
> >s 
> >
> >Frédéric THOMAS
> >
> >> From: webdoubl...@hotmail.com
> >> To: dev@flex.apache.org
> >> Subject: RE: AW: AW: AW: [FlexJS] IntelliJ Integration
> >> Date: Thu, 28 May 2015 08:47:26 +0100
> >> 
> >> Hi,
> >> 
> >> Adding support for Comp C / JSC, I've been able to compile swc and js
> >>files from IntelliJ.
> >> 
> >> In IntelliJ, I still have to add the FlexJS as a Lib to have things
> >>green though, also, despite, it passes a correct config.xml, the MXMLC
> >>compiler can't find the classe in the lib (the class is in the lib and
> >>the lib is well referenced, I tried with both include path and external
> >>and had the same result), did someone already tried this kind of config ?
> >> 
> >> Thanks,
> >> Frédéric THOMAS
> >> 
> >> > From: webdoubl...@hotmail.com
> >> > To: dev@flex.apache.org
> >> > Subject: RE: AW: AW: AW: [FlexJS] IntelliJ Integration
> >> > Date: Wed, 27 May 2015 01:12:22 +0100
> >> > 
> >> > > If you modify the Configuration class in flex-oem-compiler does it
> >>get
> >> > > used or does IJ have its own version?  If you can use a modified
> >> > > Configuration, you might be able to change getCompilerConfiguration
> >>to
> >> > > call OEMUtil, or maybe some earlier call can call OEMUtil and store
> >>the
> >> > > config someplace.
> >> > 
> >> > Yeah, I was thinking to stub most of the legacy config and still make
> >>IJ happy.
> >> > 
> >> > > Just so I understand: does this mean the FlexJS SDK can do these
> >>things in
> >> > > IJ without overlaying a regular Flex SDK or being a library within a
> >> > > regular Flex SDK?
> >> > 
> >> > For some reasons I didn't get yet (didn't dug into either, we still
> >>need to add the FlexJS Lib  + FlexJS SDK
> >> > 
> >> > Frédéric THOMAS
> >> > 
> >> > > From: aha...@adobe.com
> >> > > To: dev@flex.apache.org
> >> > > Subject: Re: AW: AW: AW: [FlexJS] IntelliJ Integration
> >> > > Date: Tue, 26 May 2015 23:58:16 +0000
> >> > > 
> >> > > 
> >> > > 
> >> > > On 5/26/15, 4:41 PM, "Frédéric THOMAS" <webdoubl...@hotmail.com>
> >>wrote:
> >> > > 
> >> > > >> I may not be understanding the issue, but if you look at
> >> > > >> flex2/tools/oem/internal/OEMUtil.java I believe there is code in
> >>there
> >> > > >> that calls Falcon’s MXMLC to get Falcon’s config, then populates
> >> > > >> CompilerConfiguration from the Falcon config.
> >> > > >
> >> > > >IJ calls the proccessConfiguration which returns a legacy
> >>Configuration
> >> > > >nd from there IJ getCompilerConfiguration(), so, IFAIK, I can't use
> >> > > >OEMUtil to give it directly.
> >> > > 
> >> > > If you modify the Configuration class in flex-oem-compiler does it
> >>get
> >> > > used or does IJ have its own version?  If you can use a modified
> >> > > Configuration, you might be able to change getCompilerConfiguration
> >>to
> >> > > call OEMUtil, or maybe some earlier call can call OEMUtil and store
> >>the
> >> > > config someplace.
> >> > > 
> >> > > >
> >> > > >Anyway, good news, I can now make / build / compile from IntelliJ
> >>to .swf
> >> > > >/ .js
> >> > > 
> >> > > Just so I understand: does this mean the FlexJS SDK can do these
> >>things in
> >> > > IJ without overlaying a regular Flex SDK or being a library within a
> >> > > regular Flex SDK?
> >> > > 
> >> > > >
> >> > > >It is too late for me to try to adapt and make it build with Ant
> >>tonight
> >> > > >though, I'll keep you in touch.
> >> > > 
> >> > > Hey, this is great progress.  I’m going to be a while on the FDB for
> >> > > Falcon so no hurry.
> >> > > 
> >> > > -Alex                          
> >> > > 
> >> >                                            
> >>                                      
> >                                       
> 
                                          

Reply via email to