Menu
▾
▴
actionstep-core — This is the list for managing the architecture and implementation of ActionStep
You can subscribe to this list here.
| 2005 |
Jan
|
Feb
|
Mar
|
Apr
(1) |
May
(72) |
Jun
(97) |
Jul
(194) |
Aug
(209) |
Sep
(57) |
Oct
(119) |
Nov
(58) |
Dec
(186) |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 2006 |
Jan
(227) |
Feb
(209) |
Mar
(90) |
Apr
(46) |
May
(20) |
Jun
(114) |
Jul
(25) |
Aug
(41) |
Sep
(42) |
Oct
(14) |
Nov
(1) |
Dec
|
| 2007 |
Jan
(4) |
Feb
(3) |
Mar
|
Apr
|
May
|
Jun
|
Jul
|
Aug
|
Sep
|
Oct
|
Nov
|
Dec
|
| 2011 |
Jan
|
Feb
|
Mar
|
Apr
|
May
|
Jun
|
Jul
|
Aug
|
Sep
|
Oct
(1) |
Nov
|
Dec
|
|
From: Ray C. <rc...@gm...> - 2005-05-30 10:27:13
|
Hi,
at this time, casting has problems. I think Flash has some sort of
fuzzy concept of pointers, so you can use it.
CODE>>>func: getRowColumnForPoint
var colWidth:Number =3D m_cellsize.width + m_cellspacing.width;
//....no need to assign aPoint
The lame part is at the return part, where since you were going
ternary, we could do:
CODE>>>func: getRowColumnForPoint
return (
(aPoint.x % colWidth) > m_cellspacing.width &&
(aPoint.y % rowHeight) > m_cellspacing.height &&
((aPoint.x =3D loc.x) || true) &&
((aPoint.y =3D loc.y) || true)
) ? true : false;
The last 2 part is always true; it's just there to ensure aPoint looks
like loc. You cannot assign and evaluate; seems it is always false.
The other part is:
CODE>>>func: mouseDown:
var loc:NSPoint =3D convertPointFromView(theEvent.mouseLocation, null);
if (getRowColumnForPoint(loc))
{//...
As in the Cocoa implementation, getRowColumnForPoint can return
boolean, and use it to indicate out-of-bounds.
|
|
From: Scott H. <Sc...@af...> - 2005-05-29 20:20:45
|
I put in my latest NSMatrixTest. Take a look. You'll see that I'm having = a problem. Now, I can obviously get around this fairly easily, but I = think that it would be best if we came up with a general solution for = cell destruction. May I suggest that we have cells implement some kind of dispose() method = for internal use by the control classes? This method would just perform = necessary=20 clean up (like destroying associated TextField instances). /Scott By the way, ScrollView looks great Rich. |
|
From: Richard K. <ri...@in...> - 2005-05-28 13:59:46
|
On May 28, 2005, at 9:08 AM, Jos wrote: > Richard Kilmer wrote: > >> See: http://www.infoether.com/~rich/ActionStep.swf >> > > Cool! I've got a bug/functionality report/question. > > First. When clicking in the scrollbar area, not on the scrollbar > itself (what _is_ that called?), the Its called the KnobSlot (not my name). > left/top of the bar tries to line up with the mouse. On windows, > for each click on the scrollbar area, it moves the bar one width/ > height towards the mouse. How do we want this to work? Yeah, they perform some weird calculation, I am going to advance a 'page' at a time. There are two settings on scrollers, one for line advance and one for page advance. Clicking the in the knobslot will cause the page up/down I think. > > Second. Using the h-scrollbar, move the bar so you are all the way > to the right. Then click close to the left edge, so you can now > see the buttons in the window. Without moving the mouse, click and > drag the h-scrollbar. The window does not update while scrolling, > or when you release the mouse. This also works for the v-scrollbar. Yeah, I think this is caused by the same strange computation I am performing when clicking in the knobslot. I will update this and see how it behaves. > > Wicked work so far guys - i can't believe how great this stuff is. > Thanks for all the hard work! > > j > > -- > jos yule > Digital Hyakugei > > mailto:jo...@th... > http://www.theorganization.net > > aim:josyule > msn:hya...@ho... > > > > ------------------------------------------------------- > This SF.Net email is sponsored by Yahoo. > Introducing Yahoo! Search Developer Network - Create apps using Yahoo! > Search APIs Find out how you can build Yahoo! directly into your own > Applications - visit http://developer.yahoo.net/?fr=offad-ysdn-ostg- > q22005 > _______________________________________________ > actionstep-core mailing list > act...@li... > https://lists.sourceforge.net/lists/listinfo/actionstep-core > > |
|
From: Jos <jo...@th...> - 2005-05-28 13:08:51
|
Richard Kilmer wrote: > See: http://www.infoether.com/~rich/ActionStep.swf Cool! I've got a bug/functionality report/question. First. When clicking in the scrollbar area, not on the scrollbar itself (what _is_ that called?), the left/top of the bar tries to line up with the mouse. On windows, for each click on the scrollbar area, it moves the bar one width/height towards the mouse. How do we want this to work? Second. Using the h-scrollbar, move the bar so you are all the way to the right. Then click close to the left edge, so you can now see the buttons in the window. Without moving the mouse, click and drag the h-scrollbar. The window does not update while scrolling, or when you release the mouse. This also works for the v-scrollbar. Wicked work so far guys - i can't believe how great this stuff is. Thanks for all the hard work! j -- jos yule Digital Hyakugei mailto:jo...@th... http://www.theorganization.net aim:josyule msn:hya...@ho... |
|
From: Richard K. <ri...@in...> - 2005-05-27 23:07:32
|
See: http://www.infoether.com/~rich/ActionStep.swf Code is: var app = NSApplication.sharedApplication(); var window1 = (new NSWindow()).initWithContentRect(new NSRect (0,0,500,500)); var view:ASTestView = new ASTestView(); view.initWithFrame(new NSRect(0, 0, 1024, 350)); view.setBorderColor(new NSColor(0xff0000)); var actionButton = (new NSButton()).initWithFrame(new NSRect (10,10,70,30)); actionButton.setTitle("Scroll 1!"); view.addSubview(actionButton); var actionButton2 = (new NSButton()).initWithFrame(new NSRect (200,200,70,30)); actionButton2.setTitle("Scroll 2!"); view.addSubview(actionButton2); var scrollView:NSScrollView = (new NSScrollView()).initWithFrame(new NSRect(0,0,250,250)); scrollView.setBorderType(NSBorderType.NSLineBorder); scrollView.setDocumentView(view); scrollView.setHasHorizontalScroller(true); scrollView.setHasVerticalScroller(true); window1.setContentView(scrollView); app.run(); Note: until you call app.run no movieclips are created. This is a cool feature because you can have all this code that is building components and hierarchies and then once 'app.run()' is run all the movieclips are faulted into existence. If you moved the view out of the window, (with view.removeFromSuperview() ) it would still exist, but the movieclips would be destroyed. If you then add it to another window it all the movieclips would be created again. Best, -rich |
|
From: Scott H. <Sc...@af...> - 2005-05-26 14:37:35
|
Yeah, I wrote one. It might take some debugging. It's can be found at http://www.sourceforge.net/projects/flashdebugtool/ You'll need to be able to compile C#. VS.NET is preferable. /Scott -----Original Message----- From: act...@li... [mailto:act...@li...] On Behalf Of Ray Chuan Sent: Wednesday, May 25, 2005 10:45 PM To: act...@li... Subject: [actionstep-core] XML Socket Server on win32 Hi, ActionStep uses some socket server with ASDebugger. Currently I'm using Eclipse/Flashout, so I'm losing out on information "levels". Anyone here can share a server that works on win32? ------------------------------------------------------- SF.Net email is sponsored by: GoToMeeting - the easiest way to collaborate online with coworkers and clients while avoiding the high cost of travel and communications. There is no equipment to buy and you can meet as often as you want. Try it = free.http://ads.osdn.com/?ad_idt02&alloc_id=16135&op=3Dick _______________________________________________ actionstep-core mailing list act...@li... https://lists.sourceforge.net/lists/listinfo/actionstep-core |
|
From: Scott H. <Sc...@af...> - 2005-05-26 14:30:33
|
I'm writing this, so I wouldn't worry about it. You can vet it when I'm
done.
-----Original Message-----
From: act...@li...
[mailto:act...@li...] On Behalf Of Ray
Chuan
Sent: Wednesday, May 25, 2005 10:36 PM
To: act...@li...
Subject: [actionstep-core] NSMatrix.mouseDown: code for radio mode
Hi,
here's the first case switch for mouseDown:
--
switch (cell.type()) {
case NSCellType.NSTextCellType:
if(m_mode=3D=3DNSMatrixMode.NSRadioModeMatrix) {
var x =3D m_sel.objectAtIndex(0)
m_cell.setHighlighted(false);
drawCellAtRowColumn(x.row, x.column);
m_sel.clear();
selectCellAtRowColumn(loc.row, loc.column);
}
break;
/*...*/
}
--
ASTestMatrix works fine with this code.
-------------------------------------------------------
SF.Net email is sponsored by: GoToMeeting - the easiest way to
collaborate
online with coworkers and clients while avoiding the high cost of travel
and
communications. There is no equipment to buy and you can meet as often
as
you want. Try it =
free.http://ads.osdn.com/?ad_idt02&alloc_id=16135&op=3Dick
_______________________________________________
actionstep-core mailing list
act...@li...
https://lists.sourceforge.net/lists/listinfo/actionstep-core
|
|
From: Scott H. <Sc...@af...> - 2005-05-26 14:29:27
|
Mouse down is a bit more complicated then that.
-----Original Message-----
From: act...@li...
[mailto:act...@li...] On Behalf Of Ray
Chuan
Sent: Wednesday, May 25, 2005 10:36 PM
To: act...@li...
Subject: [actionstep-core] NSMatrix.mouseDown: code for radio mode
Hi,
here's the first case switch for mouseDown:
--
switch (cell.type()) {
case NSCellType.NSTextCellType:
if(m_mode=3D=3DNSMatrixMode.NSRadioModeMatrix) {
var x =3D m_sel.objectAtIndex(0)
m_cell.setHighlighted(false);
drawCellAtRowColumn(x.row, x.column);
m_sel.clear();
selectCellAtRowColumn(loc.row, loc.column);
}
break;
/*...*/
}
--
ASTestMatrix works fine with this code.
-------------------------------------------------------
SF.Net email is sponsored by: GoToMeeting - the easiest way to
collaborate
online with coworkers and clients while avoiding the high cost of travel
and
communications. There is no equipment to buy and you can meet as often
as
you want. Try it =
free.http://ads.osdn.com/?ad_idt02&alloc_id=16135&op=3Dick
_______________________________________________
actionstep-core mailing list
act...@li...
https://lists.sourceforge.net/lists/listinfo/actionstep-core
|
|
From: Scott H. <Sc...@af...> - 2005-05-26 14:29:02
|
I've fixed that in my latest. -----Original Message----- From: act...@li... [mailto:act...@li...] On Behalf Of Ray Chuan Sent: Wednesday, May 25, 2005 10:22 PM To: act...@li... Subject: [actionstep-core] NSMatrix: undeclared variable m_cell Hi, used in line 1563. I assume it refers to current cell selected by matrix. ------------------------------------------------------- SF.Net email is sponsored by: GoToMeeting - the easiest way to collaborate online with coworkers and clients while avoiding the high cost of travel and communications. There is no equipment to buy and you can meet as often as you want. Try it = free.http://ads.osdn.com/?ad_idt02&alloc_id=16135&op=3Dick _______________________________________________ actionstep-core mailing list act...@li... https://lists.sourceforge.net/lists/listinfo/actionstep-core |
|
From: Richard K. <ri...@in...> - 2005-05-26 03:16:34
|
m_cell is actually declared in the superclass NSControl. It is the =20 current cell and used for mouse tracking. Scott is the author of =20 NSMatrix though, and I cannot speak for what he is doing there... -rich On May 25, 2005, at 10:21 PM, Ray Chuan wrote: > Hi, > used in line 1563. I assume it refers to current cell selected by =20 > matrix. > > > ------------------------------------------------------- > SF.Net email is sponsored by: GoToMeeting - the easiest way to =20 > collaborate > online with coworkers and clients while avoiding the high cost of =20 > travel and > communications. There is no equipment to buy and you can meet as =20 > often as > you want. Try it free.http://ads.osdn.com/?=20 > ad_idt02&alloc_id=16135&op=3Dclick > _______________________________________________ > actionstep-core mailing list > act...@li... > https://lists.sourceforge.net/lists/listinfo/actionstep-core > > |
|
From: Richard K. <ri...@in...> - 2005-05-26 03:13:24
|
You can either download and install Ruby and then ruby the 'debug' =20 file located in the root of the actionstep CVS module or Scott has =20 written a C# debugger (but I don't know if he released it as an EXE =20 or not). The Ruby one is console based and you run it (once you =20 install Ruby) with: > ruby debug Then just start up ActionStep and you should get TRACE messages. To =20 install Ruby on win32 you can download the installer from: http://rubyforge.org/frs/?group_id=3D167 Best, Rich On May 25, 2005, at 10:44 PM, Ray Chuan wrote: > Hi, > ActionStep uses some socket server with ASDebugger. Currently I'm > using Eclipse/Flashout, so I'm losing out on information "levels". > > Anyone here can share a server that works on win32? > > > ------------------------------------------------------- > SF.Net email is sponsored by: GoToMeeting - the easiest way to =20 > collaborate > online with coworkers and clients while avoiding the high cost of =20 > travel and > communications. There is no equipment to buy and you can meet as =20 > often as > you want. Try it free.http://ads.osdn.com/?=20 > ad_idt02&alloc_id=16135&op=3Dclick > _______________________________________________ > actionstep-core mailing list > act...@li... > https://lists.sourceforge.net/lists/listinfo/actionstep-core > > |
|
From: Ray C. <rc...@gm...> - 2005-05-26 02:45:03
|
Hi, ActionStep uses some socket server with ASDebugger. Currently I'm using Eclipse/Flashout, so I'm losing out on information "levels". Anyone here can share a server that works on win32? |
|
From: Ray C. <rc...@gm...> - 2005-05-26 02:36:31
|
Hi,
here's the first case switch for mouseDown:
--
switch (cell.type()) {
case NSCellType.NSTextCellType:
if(m_mode=3D=3DNSMatrixMode.NSRadioModeMatrix) {
=09var x =3D m_sel.objectAtIndex(0)
=09m_cell.setHighlighted(false);
=09drawCellAtRowColumn(x.row, x.column);
=09m_sel.clear();
=09selectCellAtRowColumn(loc.row, loc.column);
}
break;
/*...*/
}
--
ASTestMatrix works fine with this code.
|
|
From: Ray C. <rc...@gm...> - 2005-05-26 02:21:50
|
Hi, used in line 1563. I assume it refers to current cell selected by matrix. |
|
From: Ray C. <rc...@gm...> - 2005-05-26 01:35:18
|
Hi,
I think the loc object should have row and column swapped ie:
--
/* Change this:
=09=09var loc:Object =3D {};
=09=09loc.row =3D Math.floor(aPoint.x / colWidth);
=09=09loc.column =3D Math.floor(aPoint.y / rowHeight);
* to:*/
=09=09var loc:Object =3D {};
=09=09loc.row =3D Math.floor(aPoint.y / rowHeight);
=09=09loc.column =3D Math.floor(aPoint.x / colWidth);
rows are horizontal, so uses y, and columns are vertical, so uses x
|
|
From: Richard K. <ri...@in...> - 2005-05-25 16:28:12
|
Ray,
Thanks for all your posts. As a rule, the classes in
org.actionstep.test are just that...test classes...so don't worry too
much about them.
On May 25, 2005, at 6:26 AM, Ray Chuan wrote:
> Hi,
> calls to movie clip creation methods use hard-coded depths. I think a
> "mother" mc should be created, and getNextHighestDepth should be used.
> More developer friendly, I think.
We actually want to constrain a bit how folks create movieclips in
NSView objects so whether we use our own depths or
getNextHighestDepth should not be visible to the user/developer. If
you are in an NSView subclass and want to create a movieclip on that
NSView's m_mcBounds clip you call the method: createBoundsMovieClip*
and the same for TextFields with: createBoundsTextField. Those
methods handle levels and naming. In this framework the movieclips
are added and removed on the fly as NSViews are moved between
NSViews. Because of this, it would be unwise to have folks depend on
depths.
* Note: createBoundsMovieClip is a recent rename of
createSubviewMovieClip
>
> Also, some code uses "with" (supported in mtasc >1.02) but funcs like
> updateFrameMovieClipSize doesn't. Below is some ready-to-copy code:
Insofar as the use of with, as a rule I don't use it (except when I
am being lazy ;-). In the end, its just syntactic sugar for calling
multiple methods on the same object. I think its more understandable
if you just call those methods directly. But again, thanks for the
feedback on it.
> --
> private function updateFrameMovieClipSize():Void {
> if (m_mcFrame == null) {
> return;
> }
> with(m_mcFrame) {
> beginFill(0x000000, 0);
> moveTo(0,0);
> lineTo(m_frame.size.width, 0);
> lineTo(m_frame.size.width, m_frame.size.height);
> lineTo(0, m_frame.size.height);
> endFill();
> lineTo(0, 0);
> _x = m_frame.origin.x;
> _y = m_frame.origin.y;
> }
> with(m_mcFrameMask) {
> beginFill(0x000000, 100);
> moveTo(0,0);
> lineTo(m_frame.size.width, 0);
> lineTo(m_frame.size.width, m_frame.size.height);
> lineTo(0, m_frame.size.height);
> endFill();
> lineTo(0, 0);
> }
> }
> --
> I've also checked that m_frame is accessible from that scope.
>
>
> -------------------------------------------------------
> This SF.Net email is sponsored by Yahoo.
> Introducing Yahoo! Search Developer Network - Create apps using Yahoo!
> Search APIs Find out how you can build Yahoo! directly into your own
> Applications - visit http://developer.yahoo.net/?fr=offad-ysdn-ostg-
> q22005
> _______________________________________________
> actionstep-core mailing list
> act...@li...
> https://lists.sourceforge.net/lists/listinfo/actionstep-core
>
>
|
|
From: Richard K. <ri...@in...> - 2005-05-25 16:27:55
|
Done...thanks! On May 25, 2005, at 5:18 AM, Ray Chuan wrote: > Hi, > might want to change that > > > ------------------------------------------------------- > This SF.Net email is sponsored by Yahoo. > Introducing Yahoo! Search Developer Network - Create apps using Yahoo! > Search APIs Find out how you can build Yahoo! directly into your own > Applications - visit http://developer.yahoo.net/?fr=offad-ysdn-ostg- > q22005 > _______________________________________________ > actionstep-core mailing list > act...@li... > https://lists.sourceforge.net/lists/listinfo/actionstep-core > > |
|
From: Richard K. <ri...@in...> - 2005-05-25 16:27:53
|
Fixed...thanks! On May 25, 2005, at 5:53 AM, Ray Chuan wrote: > Hi, > I think line 121 should be > -- > m_notificationCenter = NSNotification.defaultCenter(); > -- > > > ------------------------------------------------------- > This SF.Net email is sponsored by Yahoo. > Introducing Yahoo! Search Developer Network - Create apps using Yahoo! > Search APIs Find out how you can build Yahoo! directly into your own > Applications - visit http://developer.yahoo.net/?fr=offad-ysdn-ostg- > q22005 > _______________________________________________ > actionstep-core mailing list > act...@li... > https://lists.sourceforge.net/lists/listinfo/actionstep-core > > |
|
From: Darron J. S. <da...@da...> - 2005-05-25 12:05:50
|
> lineTo(m_frame.size.width, 0); > lineTo(m_frame.size.width, m_frame.size.height); > lineTo(0, m_frame.size.height); > > Just a word of advice here - use local variables as much as possible for faster code execution. var w = m_frame.size.width; var h = m_frame.size.height; Local variables are stored in registers in the player and require much less processing than looking up m_frame, looking up size, looking up width, and then using that value. If you do this technique once you won't see that much of a difference, but if you do this all over the place you should notice snappier performance overall. -d |
|
From: Ray C. <rc...@gm...> - 2005-05-25 10:26:58
|
Hi,
calls to movie clip creation methods use hard-coded depths. I think a
"mother" mc should be created, and getNextHighestDepth should be used.
More developer friendly, I think.
Also, some code uses "with" (supported in mtasc >1.02) but funcs like
updateFrameMovieClipSize doesn't. Below is some ready-to-copy code:
--
private function updateFrameMovieClipSize():Void {
if (m_mcFrame =3D=3D null) {
return;
}
=09with(m_mcFrame) {
=09 beginFill(0x000000, 0);
moveTo(0,0);
lineTo(m_frame.size.width, 0);
lineTo(m_frame.size.width, m_frame.size.height);
lineTo(0, m_frame.size.height);
endFill();
lineTo(0, 0);
_x =3D m_frame.origin.x;
_y =3D m_frame.origin.y;
=09}
with(m_mcFrameMask) {
=09 beginFill(0x000000, 100);
moveTo(0,0);
lineTo(m_frame.size.width, 0);
lineTo(m_frame.size.width, m_frame.size.height);
lineTo(0, m_frame.size.height);
endFill();
lineTo(0, 0);
=09}
}
--
I've also checked that m_frame is accessible from that scope.
|
|
From: Ray C. <rc...@gm...> - 2005-05-25 09:53:20
|
Hi,
I think line 121 should be
--
m_notificationCenter =3D NSNotification.defaultCenter();
--
|
|
From: Ray C. <rc...@gm...> - 2005-05-25 09:18:14
|
Hi, might want to change that |
|
From: Ray C. <rc...@gm...> - 2005-05-23 12:13:30
|
Hi,
as the subject suggests, the border is cropped when it is drawn using
drawRect. I've used red below:
CODE:>>
public function drawRect(rect:NSRect) {
var x =3D rect.size.width-1;
var y =3D rect.size.height-1;
with(m_mcBounds) {
lineStyle(0, 0x000000, 70);
beginFill(m_backgroundColor.value, 100);
moveTo(0,0);
lineTo(x, 0);
lineTo(x, y);
lineTo(0, y);
lineTo(0, 0);
endFill();
}
}
>>
This is not a bug, just informing future users...it happens in Flash too.
|
|
From: Ray C. <rc...@gm...> - 2005-05-17 10:19:25
|
Hi,
From what I understand this class acts as the SEL type (correct me if
I'm wrong).
I was wondering if you guys would consider searching for the instance
in intern:Function, instead of leaving the work for later, ie
extern:Function.
Also, use the object as an array for performance reasons, and the keys
are numbers.
CODE:
class org.actionstep.ASUtil {=09//no s
=09private static var g_internedObject =3D new Object();
=09private static var g_internedIndex =3D 0;
//=09private static var i;
=09public static function intern(string:String):Number {
=09=09var o =3D g_internedObject;
=09=09for(var i in o) {
=09=09=09if(o[i]=3D=3Dstring)=09return parseInt(i);
=09=09}
=09=09o[g_internedIndex]=3Dstring
=09=09return g_internedIndex++;
=09}
=09
=09public static function extern(number:Number):String {
=09=09var i =3D g_internedObject[number]
=09=09return ((i=3D=3Dundefined) ? null : i);
=09}
}
--
intern has low chances of clashing, thus for each string it is called
probably once. extern could be called many times, thus perf should be
in extern's favour.
benchmarking>>
CODE:
=09=09var str:String =3D "NSApplicationDidUnhideNotification", n=3D0, i=3D1=
000, x, f
=09=09n=3DgetTimer();
=09=09while(i--) {
=09=09=09x =3D ASUtil.intern(str+i);
=09=09=09f=3DASUtil.extern(i);
=09=09}
=09=09n=3DgetTimer()-n;
=09=09TRACE(n+": "+f);
--
note the lack of "s" as in "ASUtil"
readings>>>
old: 4688, 4728, 4686
obj: 3279, 3280, 3277 (above)
arr: 3357, 3306, 3375
array version uses push, faster implementations might be available.
Good luck!
|
|
From: Richard K. <ri...@in...> - 2005-05-17 03:11:51
|
Yep, we are going to ultimately target colors, fonts, graphics for =20 buttons (check buttons, etc), etc all from the this central class =20 which is now going to be called ASTheme rather than ASDrawer since =20 there is an NSDrawer class in OpenStep/Cocoa which we may implement =20 some day and not want to get the names confused. Best, Rich On May 16, 2005, at 9:45 PM, Ron Wheeler wrote: > It sounds like you are taking the right approach and thinking about =20= > the future in a very constructive way. > > If I am understanding the context correctly, I think that we will =20 > be happy that you designed it to be easy to customize without =20 > getting into a big mess when we have to use a client's color scheme =20= > in a project. > > Ron > > Scott Hyndman wrote: > > >> I have just added ASDrawerProtocol, which is a class that must be >> implemented by Drawers, which are classes that provide primitive-=20 >> level >> drawing functions to controls. Replacing the current Drawer is =20 >> done by >> setting the ASDrawer.current property to a new class that implements >> this protocol. >> >> To any developers that are writing controls. PLEASE do not placing =20= >> any >> movieclip drawing API functions directly in your control / cell code. >> Instead, use the ASDrawer.current drawer. If the protocol does not =20= >> offer >> to some method you need, please extend it, as well as our default =20 >> drawer >> (ASDrawer). >> >> If we don't start doing these things early, it will be a bitch to =20 >> clean >> up later. >> >> I also have some gripes about colours (the NSColor object =20 >> specifically). >> We have not implemented any of the static methods that return colour >> constants to be used by controls (see section beginning with =20 >> "Creating a >> system color" at >> http://developer.apple.com/documentation/Cocoa/Reference/=20 >> ApplicationKit/ >> ObjC_classic/Classes/NSColor.html#//apple_ref/doc/uid/20000353-=20 >> BCIHAFHF >> ). It is EXTREMELY important that we do this, and exclusively use =20 >> these >> colour values in our controls, for this reason: >> >> I propose that we provide setters for this colours, resulting in =20 >> easy to >> change, global colour settings. This will be the most common avenue >> (read: simplest) for an application developer to change the look of >> controls globally. >> >> How's all that sound? >> >> /Scott >> >> >> >> ------------------------------------------------------- >> This SF.Net email is sponsored by Oracle Space Sweepstakes >> Want to be the first software developer in space? >> Enter now for the Oracle Space Sweepstakes! >> http://ads.osdn.com/?ad_idt12&alloc_id=16344&op=CCk >> _______________________________________________ >> actionstep-core mailing list >> act...@li... >> https://lists.sourceforge.net/lists/listinfo/actionstep-core >> >> >> >> >> > > > > > ------------------------------------------------------- > This SF.Net email is sponsored by Oracle Space Sweepstakes > Want to be the first software developer in space? > Enter now for the Oracle Space Sweepstakes! > http://ads.osdn.com/?ad_idt12&alloc_id=16344&op=3Dclick > _______________________________________________ > actionstep-core mailing list > act...@li... > https://lists.sourceforge.net/lists/listinfo/actionstep-core > > |
3 messages has been excluded from this view by a project administrator.
