Re: [FalconJX] yet another one

I pushed changes that should fix this. On 3/7/17, 2:08 PM, "Harbs" wrote: >The other one is the problem: > >super.replaceChildren.apply(this, applyParams); > >Becomes: >.replaceChildren.apply(this, applyParams); > >The “super” just drops out… > >My workaround attempts were in the AS3, nit the JS

Re: [FlexJS] Working to fix a compiler bug - problem with the Problems

OK. I'll take care of it. On 3/7/17, 10:11 PM, "Greg Dove" wrote: >Yes, I think they are errors with the tests. > >iirc it was > > > > > >or something like that. So I think the new code is picking up the right >things :) > >I am struggling for time right now, if you have time that would be >fan

Re: [FlexJS] Working to fix a compiler bug - problem with the Problems

Yes, I think they are errors with the tests. iirc it was or something like that. So I think the new code is picking up the right things :) I am struggling for time right now, if you have time that would be fantastic, otherwise I will try to get to it later tonight. On Wed, Mar 8, 2017

Re: [FlexJS] Working to fix a compiler bug - problem with the Problems

Greg, I just realized that these might be legitimate errors. I can fix them if you don't have time. -Alex On 3/7/17, 1:32 PM, "Alex Harui" wrote: > > >On 3/7/17, 1:04 PM, "Greg Dove" wrote: > >>I had never run those tests specifically - I was not even aware of those >>yet, the unit tests run

[4.16.0] # Apache Flex SDK 4.16.0 nightly build 252: Successful

flex-sdk_release-candidate - Build #252 - Successful Changes since last build: No changes For more information, check the console output at http://apacheflexbuild.cloudapp.net:8080/job/flex-sdk_release-candidate/252/.

Re: [FalconJX] yet another one

The other one is the problem: super.replaceChildren.apply(this, applyParams); Becomes: .replaceChildren.apply(this, applyParams); The “super” just drops out… My workaround attempts were in the AS3, nit the JS output. > On Mar 7, 2017, at 11:37 PM, Alex Harui wrote: > > The output JS looks ok

Re: [FlexJS] Working to fix a compiler bug - problem with the Problems

On 3/7/17, 1:04 PM, "Greg Dove" wrote: >I had never run those tests specifically - I was not even aware of those >yet, the unit tests run during ant all build had proceeded without >problems: There are three sets of unit tests. Because the compiler is needed to build the AS framework, we can'

Re: [FalconJX] yet another one

The output JS looks ok to me. AS: super.replaceChildren(termIdx, termIdx+1); JS: org.apache.flex.textLayout.elements.ParagraphElement.superClass_.replaceChi ldren.apply(this, [ termIdx, termIdx + 1] ); I could be wrong, but I thought there is no "super" in ES5. Does it not work correctly? Ar

[FalconJX] yet another one

Most of the other ones are “nice to fix”, but have been not too bad to work around. This one seems more serious to me. (i.e. I’m not sure how to work around it.) The “super” dropped off in the Javascript, and I don’t know why. https://paste.apache.org/FpNl Harbs

Re: [FalconJX] yet another one

I tried the following two workarounds and neither one worked: super["replaceChildren"].apply(this, applyParams); var func:Function = super.replaceChildren; func.apply(this, applyParams); > On Mar 7, 2017, at 11:13 PM, Harbs wrote: > > Most of the other ones are “nice to fix”, but have been not

Re: [FlexJS] Working to fix a compiler bug - problem with the Problems

I had never run those tests specifically - I was not even aware of those yet, the unit tests run during ant all build had proceeded without problems: I am seeing this: [junit] Running org.apache.flex.compiler.internal.codegen.js.flexjs.TestFlexJSFile [junit] Unable to parse D:\FLEXSDKS\_a

Re: [compc] another output issue

Bugs relating to internal classes wouldn't surprise me either. -Alex On 3/7/17, 12:59 PM, "Harbs" wrote: >There’s a class which has the following: >https://paste.apache.org/KEw5 > > >This is an internal class outside the file package. > >In Javascript, it outputs

Re: [compc] issue #4

On 3/7/17, 12:56 PM, "Harbs" wrote: >Yeah. Is the fact that it’s in static vars something that could break it? That wouldn't surprise me too much. -Alex

[compc] another output issue

There’s a class which has the following: https://paste.apache.org/KEw5 This is an internal class outside the file package. In Javascript, it outputs the following: org.apache.flex.textLayout.edit.ModelEdit.DeleteTextMemento.base(this, 'constructor', textFlow);

Re: [FalconJX] RegExp flags

On 3/7/17, 12:36 PM, "Harbs" wrote: >ActionScript supports different RegExp flags than Javascript suports. > >ActionScript has: g, i, s, m and x >Javascript has: g, i, m, u and y > >So we have a mismatch between “s” and “x” on one side and “u” and “y” on >the other. > >Currently, ActionScript f

Re: [compc] issue #4

Yeah. Is the fact that it’s in static vars something that could break it? > On Mar 7, 2017, at 10:49 PM, Alex Harui wrote: > > > > On 3/7/17, 12:16 PM, "Harbs" wrote: > >> I might have been mistaken. >> >> Right now, I’m seeing the goo.requires in the original js files, but they >> seem to

Re: [FlexJS] Working to fix a compiler bug - problem with the Problems

On 3/7/17, 9:41 AM, "Greg Dove" wrote: >Anything else I should be doing? If you are using Ant to build flex-falcon, please run ant flexjs.dependent.tests This may require setting up an ASJS_HOME environment variable or setting in env.properties to point to the flex-asjs repo that has prev

Re: [compc] issue #4

On 3/7/17, 12:16 PM, "Harbs" wrote: >I might have been mistaken. > >Right now, I’m seeing the goo.requires in the original js files, but they >seem to be missing from the js files in my test app. I’m not sure why >they are being stripped out — possibly it’s happening from >remove-circulars? Re

[FalconJX] RegExp flags

ActionScript supports different RegExp flags than Javascript suports. ActionScript has: g, i, s, m and x Javascript has: g, i, m, u and y So we have a mismatch between “s” and “x” on one side and “u” and “y” on the other. Currently, ActionScript flags compile to Javascript, but the google compi

Re: [compc] issue #4

I might have been mistaken. Right now, I’m seeing the goo.requires in the original js files, but they seem to be missing from the js files in my test app. I’m not sure why they are being stripped out — possibly it’s happening from remove-circulars? I’m also getting errors from google compiler l

Re: [compc] issue #4

On 3/7/17, 10:49 AM, "Harbs" wrote: >There’s a lot of [ExcludeClass] tags in TLF. It seems like this prevents >the goog.requires to be created. > >I’m guessing that [ExcludeClass] was used to prevent the classes from >winding up in the ASDocs? Is there some way to do that which does not >break

Re: git commit: [flex-asjs] [refs/heads/develop] - Added Shader files (with compiler argument)

On 3/7/17, 10:26 AM, "Harbs" wrote: >I took the files from the flex-sdk repo. How did the files get there? The pbj files are not under source control in flex-sdk. They are downloaded from the Apache Flex Pixel Bender release. > >Can you fix this issue, because you obviously understand how an

[compc] issue #4

There’s a lot of [ExcludeClass] tags in TLF. It seems like this prevents the goog.requires to be created. I’m guessing that [ExcludeClass] was used to prevent the classes from winding up in the ASDocs? Is there some way to do that which does not break the JS output? Harbs

Re: git commit: [flex-asjs] [refs/heads/develop] - Added Shader files (with compiler argument)

I took the files from the flex-sdk repo. How did the files get there? Can you fix this issue, because you obviously understand how and I haven’t a clue… Oh. Are you saying that we should reference the files in the flex-sdk directory? Wouldn’t that require building the flex-sdk as well? > On M

Re: git commit: [flex-asjs] [refs/heads/develop] - Added Shader files (with compiler argument)

Talking about pbj and pbk, does anyone have that oilpaint effect? 2017-03-07 18:01 GMT+00:00 Alex Harui : > > > On 3/7/17, 9:29 AM, "Harbs" wrote: > > >I don’t understand what you mean. > > > >These files were just copied from the flex-sdk. > > > >Huh! The pbj files were not committed. It looks

Re: git commit: [flex-asjs] [refs/heads/develop] - Added Shader files (with compiler argument)

On 3/7/17, 9:29 AM, "Harbs" wrote: >I don’t understand what you mean. > >These files were just copied from the flex-sdk. > >Huh! The pbj files were not committed. It looks like they were excluded >in the .gitignore. > >I just fixed that. Funny! That was exactly the opposite reaction that we n

Re: [FlexJS] Working to fix a compiler bug - problem with the Problems

Anything else I should be doing? One thing I did not check (and generally have not checked specifically with anything I did so far in FlexJS) is anything to do with inline (fx:Component) mxml. I *think* there are some unit tests for these, I would need to double check but if you are using tha

Re: git commit: [flex-asjs] [refs/heads/develop] - Added Shader files (with compiler argument)

I don’t understand what you mean. These files were just copied from the flex-sdk. Huh! The pbj files were not committed. It looks like they were excluded in the .gitignore. I just fixed that. > On Mar 7, 2017, at 6:02 PM, Alex Harui wrote: > > Hi, > > Is it possible to use the upstream rele

Re: [FlexJS] Group container

In theory, the API for Container should abstract away how many internal pieces there are. It should probably be a div in JS. Adding a ScrollingContainerView should be the PAYG way of turning on overflow:auto. On the SWF side, the ScrollingContainerView has a lot more work to do. But it should al

Re: git commit: [flex-asjs] [refs/heads/develop] - Added Shader files (with compiler argument)

Hi, Is it possible to use the upstream release instead? Even the flex-sdk uses pbj files from the upstream. IIRC, we did that because not everyone wanted to deal with installing the PB compiler. Thanks, -Alex On 3/7/17, 2:25 AM, "ha...@apache.org" wrote: >Repository: flex-asjs >Updated Branc

Re: [FlexJS] Dual branch

On 3/7/17, 4:36 AM, "Christofer Dutz" wrote: >Hi guys and especially Alex, > >I would like to ask you to give a short summary at what you changed in >the dual branch. I don’t quite know as to what I should be looking for to >make Maven support to match. > >Could you please summarize what new co

Re: [FlexJS] Group container

As I go through FlexJS, I am wondering if we need to continue the idea of chrome. Chrome is not something built into HTML/CSS/JS, it is artificial. We use it for the title bar and control bar in a Panel, but a Panel can be composed of nesting Groups and applying the correct styles. In Flex, scroll

Re: [compc] error #3

Have you tried xml["@class"] instead? - Josh On Tue, Mar 7, 2017 at 3:19 AM, Harbs wrote: > Related: > > This produces an error: Syntax error: ‘class’ is not allowed here. > xml.@class = elem.styleName; > > and the following gets a null pointer: > xml.@["class"] = elem.styleName; > > That doesn

Re: [FlexJS] Embed Workaround

On 3/7/17, 2:46 AM, "yishayw" wrote: >I guess nightly wasn't up2date. After rebuilding falcon I'm not getting >anymore compile errors with embed. Thanks. > >I added > >org.apache.flex.core.ByteArrayAssetbed-class> >to >Graphics\src\main\config\compile-as-config.xml > >to make the ant build work

Re: [FlexJS] Group container

Thanks for the feedback! As I go through the examples, I see that we (mostly me) created a number of nested elements, such as a for an item renderer. That makes it more difficult to lay out since the content of the div is not always set to fill the div's space. I think each of the components and

Re: [FlexJS] Group container

Hi Peter, I think this awesome. get rid of the hardcoded styles in component classes is such important thing and first point not only in your effort of create a good layout strategy, but an important previous step if we want to implement theming in FlexJS. Great! :D Carlos 2017-03-07 14:23 GMT+

Re: [FlexJS][FalconJX] Dual output Compilation

The second time, it worked. > On Mar 7, 2017, at 12:40 PM, Harbs wrote: > > Bah. I’m getting the following when I try to install dual: > > Installer version 3.2.0 (mac) > Using Locale: en_US > Fetched the SDK download mirror URL from the CGI. > SDK version Apache FlexJS Dual > AIR version 3.1 >

Re: [FlexJS] Group container

This is the theory, yes. A way to do your own thing using AS and MXML to construct the app which then generates the right amount of HTML structure, making it easier to style. Or use pre-built constructions and layouts as templates that also generate the right amount of HTML structure. I'm thinking

Re: git commit: [flex-asjs] [refs/heads/develop] - Introducing the Group container.

Will do on the my check-in. ‹peter On 3/6/17, 2:28 PM, "Piotr Zarzycki" wrote: >Hi Peter, > >Nice job! Could you please add @productversion FlexJS 0.8 ? I think it can >be useful in future. > >Thanks, >Piotr > >2017-03-06 19:15 GMT+01:00 : > >> Repository: flex-asjs >> Updated Branches: >> ref

[FlexJS] Dual branch

Hi guys and especially Alex, I would like to ask you to give a short summary at what you changed in the dual branch. I don’t quite know as to what I should be looking for to make Maven support to match. Could you please summarize what new compiler options/arguments, which new/changed/obsolete

Re: [compc] error #3

Related: This produces an error: Syntax error: ‘class’ is not allowed here. xml.@class = elem.styleName; and the following gets a null pointer: xml.@["class"] = elem.styleName; That doesn’t leave many options for assigning a “class” attribute to XML.. ;-) > On Mar 7, 2017, at 1:09 PM, Harbs wr

Re: [compc] error #3

Bingo. This is causing an error: var version:String = xmlToParse.@["version”]; Changing it to the following works: var version:String = xmlToParse.@version; > On Mar 6, 2017, at 9:31 PM, Alex Harui wrote: > > Ah, after, digging deeper, it is the @ operator in XML. And some bugs > around that

RE: [FlexJS] Embed Workaround

I guess nightly wasn't up2date. After rebuilding falcon I'm not getting anymore compile errors with embed. Thanks. I added org.apache.flex.core.ByteArrayAsset to Graphics\src\main\config\compile-as-config.xml to make the ant build work, however my Maven build is failing. Where is the standard pl

Re: [FlexJS][FalconJX] Dual output Compilation

Bah. I’m getting the following when I try to install dual: Installer version 3.2.0 (mac) Using Locale: en_US Fetched the SDK download mirror URL from the CGI. SDK version Apache FlexJS Dual AIR version 3.1 Flash Player version 11.1 Creating Apache FlexJS home Creating temporary directory Downloadi

RE: [FlexJS] Embed Workaround

I’m getting command line Error: unknown configuration variable 'compiler.byte-array-embed-class'. After downloading the nightly. From: Alex Harui Sent: Monday, March 6, 2017 9:35 PM To: dev@flex.apache.org Subject: Re: [FlexJS] Embed Workarou