Profile

dlots
dlots
Member Since
September 22, 2011

Search Members

  

dlots

23
@

Badges

This user hasn't earned any badges yet.

Posts

Viewing 1 to 20 (22 Total)

Re: No nme.system.ApplicationDomain?

You're entirely correct. Apart from perhaps having an API facilitating accessing static properties of default and basic types (though I think there are much higher priority tasks than even the concept of such a module API), there does not appear to be much use for this as the class definition cannot be transferred between parent and child in a reasonable way. An interesting thing to solve for a scripting language perhaps.

Posted on May 22, 2013 at 2:28 AM

No nme.system.ApplicationDomain?

Hello, I am curious if it is possible to implement something like ApplicationDomain for the hxcpp target with regards to the ability to access objects and static variables of classes in different applicationDomain.

Posted on May 21, 2013 at 1:20 PM

Rectangle/Point need to serialize cross platform.

At the moment, there is no serialization strategy on how to serialize and then deserialize classes from one target to the next. In general, I think there needs to be a patch to haxe serialize that specifies a class map, specifying that a particular class should be serialized into a different class.

Posted on October 29, 2012 at 4:30 AM

Re: ByteArray needs to be compatible with haxe.io.Bytes

And the other way:

#if flash
return data.getData().readObject();
#else
return ByteArray.fromBytes(data).readObject();
#end

Posted on June 18, 2012 at 10:04 PM

ByteArray needs to be compatible with haxe.io.Bytes

Typedefing ByteArray to flash.utils.ByteArray makes it incompatible with haxe.io.Bytes. This needs to be resolved. This forces the following code:

#if flash
return Bytes.ofData(bytes);
#else
return bytes;
#end

Which isn't so bad in my case as its only in one place hopefully, but still.

So the solution would be, for #if flash nme.util.ByteArray to extend Bytes, and wrap the calls to flash.util.ByteArray methods. I doubt this is gong to be implemented but worth considering.

Posted on June 18, 2012 at 10:00 PM

Re: Is there a display list engine using Tilesheet (like Starling)

Also, when is the expected timeline for (presumably drawTiles to support) Stage3d?

Posted on May 28, 2012 at 5:38 PM

Is there a display list engine using Tilesheet (like Starling)

Has any one done any work on writing a display list like engine like Starling using Tilesheet?

Posted on May 28, 2012 at 5:37 PM

Re: Link nmml documentation/developer documentation from the documentation main menu

Also, the following questions are relevant to nmml files:

What are the required tags?
What are the required ndlls?

Posted on April 30, 2012 at 1:42 AM

Link nmml documentation/developer documentation from the documentation main menu

The website only links to relevant documentation from the very bottom ofhttp://www.haxenme.org/developers/get-started/...

http://www.haxenme.org/developers/documentation/nmml-project-files/...

Also, I think the forum software needs to be changed. There is no support for code presumably, no post preview, and each thread takes up way too much space in the list of threads in a forum category, namely because of the user's icon. I would recommend removing the user's icon as it serves no purpose.

Posted on April 30, 2012 at 1:40 AM

Re: Does ByteArray.writeObject the same as output as AMF3?

... There is no writeObject method sad

Posted on April 27, 2012 at 5:16 PM

Does ByteArray.writeObject the same as output as AMF3?

If so, it appears you have written the only well-supported known C++ implementation of AMF3 that I found. Congrats, you might even want to publish it.

If not, what is the output? Haxe serialized object?

Posted on April 27, 2012 at 2:22 PM

Dynamic on TextFormat color and size?

Why does the AS3 TextFormat and Nme TextFormat use Dynamic for color and size? Why not Int? Is it safe to typecast to Int these properties?

Posted on April 26, 2012 at 6:26 PM

Detecting if using a Touch device

The reason I am asking is the following scenario.

I want to bind actions to an over event or a move event if it's on touch devices. Ie, I want the user to be able to grab items on over/move depending on the device or slice the screen ie fruit ninja.

As over events do not occur starling and presumably do not occur with NME, I need a way to detect for the javascript target if it is a touch device. For Flash/C++ I can do it manually depending on the target, but for Javascript, when manually firing the Over event on a Move event, if there is no way to detect to if it's a touch device, I would end up firing the Over event twice if the user is executing the javascript on a computer, assuming the Over event is fire if the mouse is down, or if its not the behavior would not be the intending behavior for a computer. Ie, it would be firing over for a Move event.

Now from the last sentence it may appear that it should not matter that the behavior on Javascript vs Flash is changed, as if Over is fired even when the mouse is down, you would get that behavior by default. That is however, not the case as one can adjust for that behavior and take action only when the mouse is not down. Without a touch check it would still fire a Over on move manually. The adjustment is at an abstract display list interface for Starling and NME. Adding the adjustment to the actual handler is not really reasonable.

Anyway the point, theoretically I dont want an Over fired when on mouse devices on Move and I want it fired if it is not fired on Touch devices. In order to faciliate this regardless of behavior one needs to know if it is a touch device or not. I have contemplated all behaviors.

The other behavior is if Over is fired on Move regardless, for both Touch and Mouse, ignoring an Over with buttonDown does not work unless you know it's a touch device. Otherwise you would eliminate all inputs.

tldr, I'd like to know if its possible to detect if its a touch device. Yes I know I can do it manually by detecting on the first click, listening for both touch event and mouse event and comparing.

Posted on April 26, 2012 at 4:21 PM

Re: ScrollRect

lol, my bad there is scrollRect. Cool!

Posted on April 23, 2012 at 3:48 PM

Re: ScrollRect

I see that DisplayObject has mask. Any objections to me implementing scrollRect?

Posted on April 23, 2012 at 3:46 PM

ScrollRect

It doesn't appear that nme.display.Sprite has scrollRect. Is scrollRect implemented in any way. If it is not, it should be a critical feature as it isn't possible to have any sort of scrolling or UI without scrollRect or masking.

Posted on April 23, 2012 at 2:49 AM

Re: Are all of the non-package Haxe apis compilable to C++?

I see that the API doc specified which targets a class is available.

Posted on April 13, 2012 at 5:39 AM

Are all of the non-package Haxe apis compilable to C++?

Are the non-package Haxe API all useable in a C++ target application? Namely the Reflect class. What C++ API does the Reflect class use in the C++ target?

Is the haxe package classes useable in a C++ target application? Is the haxe package is not compilable to a C++ application, what are the available targets for the haxe target?

Posted on April 13, 2012 at 5:30 AM

Re: NMEPreloader & haxe relative paths

Also, from what is the Assets.hx generated? In samples/02-Text it references icon.png which seems like it is generated from the svg specified in the icon tag.

Posted on September 25, 2011 at 11:20 AM

NMEPreloader & haxe relative paths

It appears that compiling with nme generates a preloader for the swf. Is there a way to disable this or implement a custom preloader?

Also, relating to haxe:

It appears that directives in hxml files are written in relation from where haxe is being executed as opposed to location of the hxml file. For example, the generated release.hxml file specifies:

-cp .
-cp bin/flash/haxe

Posted on September 25, 2011 at 11:17 AM
« Previous12Next »