OK, seems to be working now.
-Alex
On 11/16/15, 2:08 PM, "Alex Harui" wrote:
>
>On 11/16/15, 1:56 PM, "Alex Harui" wrote:
>
>>FWIW, something Infra said just clicked. The CMS is injecting the bad
>>CSS
>>so this issue probably only started when I pushed updates to the website
>>today.
>>
>>I
On 11/16/15, 2:54 PM, "Michael Schmalle" wrote:
>On Mon, Nov 16, 2015 at 5:44 PM, Alex Harui wrote:
>
>>
>>
>> On 11/16/15, 2:04 PM, "Michael Schmalle"
>> wrote:
>>
>> >Is there some one else working on the compiler code other than you
>>Alex?
>>
>> Josh has made a few changes, but lately I t
OK, I finally got it working.
Please use the installer to install Apache FlexJS 0.5.0. I just tried it
on Win7 and there was several jars in the lib folder.
If it fails, there should be an installer.log in the SDK folder. Please
open a JIRA issue at [1] and attach the log and maybe a dump of th
On Mon, Nov 16, 2015 at 5:44 PM, Alex Harui wrote:
>
>
> On 11/16/15, 2:04 PM, "Michael Schmalle"
> wrote:
>
> >Is there some one else working on the compiler code other than you Alex?
>
> Josh has made a few changes, but lately I think I’m the main person. The
> log will show you.
>
> -Alex
>
On 11/16/15, 2:04 PM, "Michael Schmalle" wrote:
>Is there some one else working on the compiler code other than you Alex?
Josh has made a few changes, but lately I think I’m the main person. The
log will show you.
-Alex
On 11/16/15, 1:56 PM, "Alex Harui" wrote:
>FWIW, something Infra said just clicked. The CMS is injecting the bad CSS
>so this issue probably only started when I pushed updates to the website
>today.
>
>I have switched the installer to not use the mirrors and successfully
>installed both 4.14.1
On Mon, Nov 16, 2015 at 4:20 PM, Alex Harui wrote:
> Looks like the work would be:
>
> 1) add the wrapping function
> 2) change all method and property output to not include the package.
>
> Is it worth it just to try to get better privacy? I think just adding an
> inner class would be less work
FWIW, something Infra said just clicked. The CMS is injecting the bad CSS
so this issue probably only started when I pushed updates to the website
today.
I have switched the installer to not use the mirrors and successfully
installed both 4.14.1 and 0.5.0 from archive.a.o.
-Alex
On 11/16/15, 1:
Yeah, an inner class sounds like an acceptable approach. I don't think we
absolutely need to enforce this privacy on the JavaScript side. As long as
the ActionScript compiler won't let you use the inner class elsewhere,
that's fine.
I would say that adding some kind of prefix or something to the i
Hi,
I tried the installer on the weekend (Monday your time?) and had no issues.
> And actually, that’s the workaround I will implement. I will point 4.14.1
> and 0.5.0 right to Archive. I think it can handle 100 downloads a day
> until they get this fixed.
Just so you know we’re likely to hit
Thanks, not quite sure what that means. Has anyone been successful in the
past hour, 2 hours, 4 hours? For 4.14.1 (or FlexJS 0.5.0) since those use
the mirrors. Older releases shouldn’t be affected.
And actually, that’s the workaround I will implement. I will point 4.14.1
and 0.5.0 right to Ar
Here are the Installer failures for the last 10 days:
Date
Pageviews
11/16/2015
70
11/11/2015
115
11/10/2015
119
11/5/2015
69
11/4/2015
55
10/29/2015
62
10/28/2015
80
10/27/2015
69
10/22/2015
68
10/21/2015
60
On Mon, Nov 16, 2015 at 12:58 PM, Alex Harui wrote:
>
>
> On 11/16/15, 12:56 PM, "Just
Looks like the work would be:
1) add the wrapping function
2) change all method and property output to not include the package.
Is it worth it just to try to get better privacy? I think just adding an
inner class would be less work for me. Some other volunteer could make
your proposed change la
All members assigned to the prototype would automatically be available.
You'd use pretty much the same syntax, but you wouldn't need to reference
the full package name for each member:
some.package.SomeClass = (function()
{
var SomeClass = function(){};
SomeClass.prototype.someMethod = fun
On 11/16/15, 12:56 PM, "Justin Mclean" wrote:
>Hi,
>
>You probably know this but it's broken for FlexJS 0.5 as well.
Yeah, that’s how I first found it. Is our analytics behind by several
hours so we don’t know when the breakage started? I’m sort of surprised
it is working at all for anybody
Hi,
You probably know this but it's broken for FlexJS 0.5 as well.
Justin
HI,
> I’m trying to fashion a workaround in the meantime. We should be seeing
> 100% failures on 4.14.1 since whatever this change was went in.
I’m still seeing successful downloads - perhaps it’s mirror dependant OR the
change has been in the last 24 hours.
Thanks,
Justin
On Mon, Nov 16, 2015 at 11:28 AM, Alex Harui wrote:
>
>
> On 11/16/15, 2:38 AM, "Michael Schmalle"
> wrote:
>
> >It's important. You need to figure out how to do it in JS first, then bend
> >the compiler to your will. :)
> >
> >In a SWF decompile, these are labeled private class.
>
> OK. Well A
I filed
On 11/16/15, 12:29 PM, "Justin Mclean" wrote:
>Hi,
>
>> Hey, can someone else try installing Apache Flex SDK 4.14.1?
>
>Broken for me as well.
>
>> I think the mirror code is broken again.
>
>Look like we have some extra CSS in the output that’s causing the
>installer to barf?
Yep. I
Hi,
> Hey, can someone else try installing Apache Flex SDK 4.14.1?
Broken for me as well.
> I think the mirror code is broken again.
Look like we have some extra CSS in the output that’s causing the installer to
barf?
/* The following code is added by mdx_elementid.py
It was originally li
On 11/16/15, 10:36 AM, "Josh Tynjala" wrote:
>The Immediately-Invoked Function Expression (IIFE) pattern could be used
>to
>create private classes.
>
>some.package.SomeClass = (function()
>{
>var SomeHelperClass = function(){};
>
>var SomeClass = function(){};
>
>return SomeClass;
>
On 11/16/15, 11:43 AM, "Harbs" wrote:
>
>On Nov 16, 2015, at 9:40 PM, Harbs wrote:
>
>>
>> On Nov 16, 2015, at 9:14 PM, Alex Harui wrote:
>>
>>> Hmm. I wonder what other operators work. Like “-=“.
>>
>> It looks like -= does not work.
>>
>>> And does xmlList1 + xmlList2 concatenate them
On Nov 16, 2015, at 9:40 PM, Harbs wrote:
>
> On Nov 16, 2015, at 9:14 PM, Alex Harui wrote:
>
>> Hmm. I wonder what other operators work. Like “-=“.
>
> It looks like -= does not work.
>
>> And does xmlList1 + xmlList2 concatenate them?
>
> Yes.
+= also works for this.
On Nov 16, 2015, at 9:14 PM, Alex Harui wrote:
> Hmm. I wonder what other operators work. Like “-=“.
It looks like -= does not work.
> And does xmlList1 + xmlList2 concatenate them?
Yes.
Creating a new setChild method should be fine. How would we differentiate
between + and +=? Two new methods? Or something else?
On Nov 16, 2015, at 9:14 PM, Alex Harui wrote:
>
>
> On 11/16/15, 10:54 AM, "Harbs" wrote:
>
>>
>> On Nov 16, 2015, at 8:03 PM, Alex Harui wrote:
>>
>>>
>>>
On 11/16/15, 10:54 AM, "Harbs" wrote:
>
>On Nov 16, 2015, at 8:03 PM, Alex Harui wrote:
>
>>
>>
>> On 11/16/15, 9:54 AM, "Harbs" wrote:
>>
2) detect read and write on XML/XMLList and convert read to child().
What
about write()?
>>>
>>> Good question. I think we should map x
On 11/16/15, 1:30 AM, "Christofer Dutz" wrote:
>
>2. The Problem with a dual approach of ANT and Maven is that Maven uses
>convention over configuration, this means that it is possible to setup a
>maven build for the current project structure, but it's quite a
>configuration effort. If however w
It was just based on someone else's comment about cleaning up and restructuring
some. I'm sure it only really affects those of us who using synching. Having
a few thousand less files per sdk to check/synch can make a speed difference.
-Mark
-Original Message-
From: Alex Harui [mailt
On Nov 16, 2015, at 8:03 PM, Alex Harui wrote:
>
>
> On 11/16/15, 9:54 AM, "Harbs" wrote:
>
>>> 2) detect read and write on XML/XMLList and convert read to child().
>>> What
>>> about write()?
>>
>> Good question. I think we should map xml.foo = to
>> xml.replace(“foo”,new XML(“”));
>
> W
On 11/16/15, 9:36 AM, "Kessler CTR Mark J"
wrote:
>Well for one of my dev machines, we've removed all the source, build
>files, property files, samples, and templates out of the folders.
>Multiple SDK's stored in a shared parent folder. The sdk folders are
>labeled things like "Flex 4.14.1
The Immediately-Invoked Function Expression (IIFE) pattern could be used to
create private classes.
some.package.SomeClass = (function()
{
var SomeHelperClass = function(){};
var SomeClass = function(){};
return SomeClass;
})();
SomeHelperClass will only be accessible inside the (fu
Hey, can someone else try installing Apache Flex SDK 4.14.1? I think the
mirror code is broken again.
Thanks,
-Alex
On 11/16/15, 9:54 AM, "Harbs" wrote:
>>2) detect read and write on XML/XMLList and convert read to child().
>>What
>> about write()?
>
>Good question. I think we should map xml.foo = to
>xml.replace(“foo”,new XML(“”));
What does the above actually do if there is only one foo child and if
ther
On Nov 16, 2015, at 6:06 PM, Alex Harui wrote:
>
>
> On 11/16/15, 12:51 AM, "Harbs" wrote:
>
>> This seems like a partial solution. The thing with dot notation, is that
>> you can use anything — even elements that do not exist. You have a
>> mismatch between your xml and your code. In your e
Well for one of my dev machines, we've removed all the source, build files,
property files, samples, and templates out of the folders. Multiple SDK's
stored in a shared parent folder. The sdk folders are labeled things like
"Flex 4.14.1" / "Flex 4.13.0" / "Flex 4.6.0". I have them setup i
ok, i'll wait for the new release, thank you!!
2015-11-16 17:36 GMT+01:00 Alex Harui :
> Hi Alan,
>
> I’m sorry you are having so many problems getting FlexJS installed. Today
> I’m bringing the new FlexJS 0.5.0 release live. Once that is completed,
> please try installing 0.5.0 and then we ca
Hi Alan,
I’m sorry you are having so many problems getting FlexJS installed. Today
I’m bringing the new FlexJS 0.5.0 release live. Once that is completed,
please try installing 0.5.0 and then we can compare results. I’ll post on
this thread again when the release is live.
Thanks,
Alex
On 11/1
On 11/16/15, 2:38 AM, "Michael Schmalle" wrote:
>It's important. You need to figure out how to do it in JS first, then bend
>the compiler to your will. :)
>
>In a SWF decompile, these are labeled private class.
OK. Well AFAIK, there is no such thing as private classes in JS. One
thought I ha
On 11/16/15, 3:22 AM, "Michael Schmalle" wrote:
>On Mon, Nov 16, 2015 at 6:07 AM, Harbs wrote:
>
>> It would actually be very nice if there was some way to (optionally)
>> prevent null being passed into constructors and function calls. There
>>has
>> been more than once that I would have liked
Could someone please check this issue:
https://issues.apache.org/jira/browse/FLEX-34954
Fix is provided in the ticket.
I need an official fix confirm via a commit for this issue in GIT trunk,
because of internal company politics -.-
Thank you!
--
Konstantin Elstner
Sent with Airmail
Good catch. I will remove “initial (alpha)” from the first line. I think
we want folks to think of this as a beta.
-Alex
On 11/16/15, 6:41 AM, "Peter Ent" wrote:
>The opening of the announcement states this is "the initial (alpha)
>release", but a few paragraphs later says, "It is Œbeta¹ qual
On 11/16/15, 5:23 AM, "Kessler CTR Mark J"
wrote:
>
>One my ideas for improvement would be to change the way we do binary
>packages. The strip out everything from the binary package except the
>bare bones of what is required. So just keep it the libraries, asdocs,
>licenses, notices and su
On 11/16/15, 12:51 AM, "Harbs" wrote:
>This seems like a partial solution. The thing with dot notation, is that
>you can use anything — even elements that do not exist. You have a
>mismatch between your xml and your code. In your example, xml.committer
>will return an XMLList with two elements.
The opening of the announcement states this is "the initial (alpha)
release", but a few paragraphs later says, "It is Œbeta¹ quality". I think
it is confusing to use both "alpha" and "beta". IMO, "alpha" is still fine
for this release.
‹peter
On 11/16/15, 12:51 AM, "Alex Harui" wrote:
>Draft an
I use the environmental variables. Especially when I synch sdk/air/global
files between development machines. My main work development machine is on a
private network. Each of my machines are setup with different paths and have
slightly different environments. Using scripts to setup t
Hi,
> 5. You don't "have to" take a course to use Maven ... I bet 99% of the Maven
> users have never taken a course, it's just an offer from my side to help
> others get up to speed.
That’s 100% correct I’ve made binding votes on lots (50+?) of releases just
knowing the basics. i.e. “mvn comp
On Mon, Nov 16, 2015 at 6:07 AM, Harbs wrote:
> It would actually be very nice if there was some way to (optionally)
> prevent null being passed into constructors and function calls. There has
> been more than once that I would have liked to have that feature.
>
> I wonder if that’s something we
It would actually be very nice if there was some way to (optionally) prevent
null being passed into constructors and function calls. There has been more
than once that I would have liked to have that feature.
I wonder if that’s something we can add to Falcon?
On Nov 16, 2015, at 12:43 PM, Micha
On Mon, Nov 16, 2015 at 5:40 AM, OmPrakash Muppirala
wrote:
> Nope, you can still call new Singleton(null) and the compiler won't say
> anything :-)
> The 'enforcer' is required because of this loophole.
>
>
Yeah, you're right, to early in the morning.. :)
> Thanks,
> Om
>
> On Mon, Nov 16, 201
Nope, you can still call new Singleton(null) and the compiler won't say
anything :-)
The 'enforcer' is required because of this loophole.
Thanks,
Om
On Mon, Nov 16, 2015 at 2:36 AM, Michael Schmalle wrote:
> On Mon, Nov 16, 2015 at 4:14 AM, OmPrakash Muppirala >
> wrote:
>
> > Purely pedantic,
It's important. You need to figure out how to do it in JS first, then bend
the compiler to your will. :)
In a SWF decompile, these are labeled private class.
Mike
On Mon, Nov 16, 2015 at 2:53 AM, Alex Harui wrote:
> Hi,
>
> Every once in a while, the framework code uses an “internal class”. A
On Mon, Nov 16, 2015 at 4:14 AM, OmPrakash Muppirala
wrote:
> Purely pedantic, but the code does not do anything to 'enforce' the
> singleton-ness of the class.
>
The compiler won't even let you compile this code, so the != is just sauce
that never gets eaten. Since you have to pass an instance
Of course, my suggestion of using event.target.myXML.child(“foo”) will only
work if we do not prepend _as3_.
Actually, prepending will break any use of XML methods for non-typed XML.
On Nov 16, 2015, at 10:51 AM, Harbs wrote:
> This seems like a partial solution. The thing with dot notation, i
Not sure if it’s useful yet, but apparently Windows 10 has “OneGet” which is
supposed to fill the same need. There’s currently Chocolatey as well. I’m not
sure how either one measures up. Some more reading:[3]
[1]https://github.com/oneget/oneget
[2]https://chocolatey.org/
[3]http://www.hanselman
Ok ... let me do a summary response too :-)
1. I have to admit that I have never, ever in my life installed Maven ... for
me downloading a zip and unpacking it, isn't installing ... you don't have to
setup environment variables or paths if you use it from inside an IDE ... on
the command-line t
Purely pedantic, but the code does not do anything to 'enforce' the
singleton-ness of the class.
It should actually be:
public function Singleton(e:SingletonEnforcer){
if(e != null)
{
trace("new instance of singleton created");
}
}
On Mon, Nov 16, 2015 at 12:54 AM, yishayw wrote:
> A
Actually the comment above 'class SingletonEnforcer' explains it.
package{
public class Singleton{
private static var _instance:Singleton=null;
public function Singleton(e:SingletonEnforcer){
trace("new instance of singleton created");
}
public stati
Ah okay. PC guy here :-)
On Mon, Nov 16, 2015 at 1:05 AM, Harbs wrote:
> “brew” is the command for Homebrew which is a Mac package manager similar
> to apt-get on Linux. The other popular one is MacPorts. I have both
> installed, but I tend to use Homebrew.
>
> Here’s some interesting reading o
It ensures you can’t instantiate your singleton outside the singleton class.
On Nov 16, 2015, at 10:50 AM, yishayw wrote:
> Not sure why...
I've seen them used in Singleton implementations. Not sure why...
See here
http://stackoverflow.com/questions/10233868/why-there-are-no-private-constructors-in-as3-version-of-singleton
--
View this message in context:
http://apache-flex-development.247.n4.nabble.com/FlexJS-FalconJX-Inter
“brew” is the command for Homebrew which is a Mac package manager similar to
apt-get on Linux. The other popular one is MacPorts. I have both installed, but
I tend to use Homebrew.
Here’s some interesting reading on them.[1]
Homebrew has had recipes for every package I’ve needed.
[1]http://app
Yes. That’s bad for two reasons:
1. You could end up with conflicts between two different packages. (unless I’m
misunderstanding you)
2. Polluting the global object is really frowned on. We should have as few
globals as possible.
I’m not sure what the internal classes should become though.
On N
This seems like a partial solution. The thing with dot notation, is that you
can use anything — even elements that do not exist. You have a mismatch between
your xml and your code. In your example, xml.committer will return an XMLList
with two elements. xml.employee will return an XMLList with a
Hi Alex,
In my opinion internal classes are important and compiler should understand
them correctly.
One thing which always drive me crazy is that those classes are usually
placed in the same file with some other class. (That how it looks like in
ActionScript world).
I think SomeClass.as and SomeH
also, the following folder are empty:
lib, runtimes, sample, scripts and templates, is it correct??
Thanks...
2015-11-16 9:44 GMT+01:00 Buranello Alan :
> well, i used the installer e i installed apache flex js 0.5.0, but when i
> tryed to add the flex sdk in my eclipse, i encountered in this p
well, i used the installer e i installed apache flex js 0.5.0, but when i
tryed to add the flex sdk in my eclipse, i encountered in this problem:
"mxmlc.jar" not found in classpath /flex-js-0.5.0
is it right to copy this file from the flex-js-0.0.2 in lib folder??
2015-11-13 18:35 GMT+01:00
Hi,
> Do you see any conflicts with ASF policy if our Maven scripts bring down
> Adobe artifacts?
No issues there I believe. If they not bundled then we’re not distributing them
so there's no issue. It’s more to do with Adobe policy rather than Apache.
> Apparently, I don’t have brew installed.
67 matches
Mail list logo
