280north / cappuccino

[[column dataView] setAlignment:CPRightTextAlignment];

is not respected, text remains left aligned.

(CPTableView from nib2cib or not…)

DOM generated: (see the text-align: left in the inner div)

I have investigate if there was not a side effect somewhere else resetting the text alignment, but CPLog.trace("dataView alignement %d", [[column dataView] alignment]) ;
shows up the alignment is 1 == CPRightTextAlignment everywhere.

Test case added in http://github.com/280north/cappuccino/commit/403277ed0641b98dfdddea45c2f2cda26e9140c8 .

For example:

[[undoManager prepareWithInvocationTarget:self] setValue:5 forKey:"something"];

The shortcut key label of a CPMenu on windows still displays the clover.

Clicking on the swatches (especially clicking rapidly) on both the color wheel mode and slider mode of a CPColorPanel produces the following error after the application hangs for several seconds (in Safari, not sure about other browsers at the moment, if it would even matter):

• [CPColor string] unrecognized selector sent to instance 0x0001de

I am invoking the CPColorPanel from a CPColorWell control which is added to a standard CPView, nothing fancy. Using a couple days old build of the jake branch.

• R

So is decodeHTMLComponent. I don't have a test case because I don't even use Objective-J. I was just browsing the code and saw that it would totally not work.

http://github.com/280north/cappuccino/blob/master/Objective-J/plist.js#L417-425

In JS, passing a string to replace() will only replace the first occurrence. Also, the ampersand (&) is replaced at the end, which would wipe out the ampersand from HTML Entities created by the previous calls to replace().

I have spend quite a long time to realize where the problem is (because I was working on something else, related to extending the Response with access to headers).

The CPURLConnection initWithRequest does not set the _isLocalFileConnection correctly. This is against AB-8 (working with Atlas!).

Adding some debug statements makes the problem visible:

CPLog.debug("CPURLConnection::URL=" + URL);

    CPLog.debug("URL::scheme=" + scheme);
    // Browsers use "file:", Titanium uses "app:"
    _isLocalFileConnection =    scheme === "file" ||
                                ((scheme !== "http" || scheme !== "https:") &&
                                window.location &&
                                (window.location.protocol === "file:" || window.location.protocol === "app:"));
    CPLog.debug("CPURLConnection::isLocalFileConnection=" + _isLocalFileConnection);

produces:

2009-11-07 18:31:08.271 Cappuccino [debug]:     CPURLConnection::URL=http://192.168.251.5/~wimpi/test/test.xml
2009-11-07 18:31:08.272 Cappuccino [debug]: URL::scheme=http
2009-11-07 18:31:08.273 Cappuccino [debug]: CPURLConnection::isLocalFileConnection=true

Consequently the CPURLConnection _readyStateDidChange will not call the delegate with the right type (will be CPURLResponse instead of CPHTTPURLResponse).

After setting texts to "" the DOM elements for the text and text shadows are removed. When setting the text to a non empty string later the DOM elements are created but the shadow color isn't being set.

This patch seems to fix the problem

diff --git a/AppKit/_CPImageAndTextView.j b/AppKit/ 
_CPImageAndTextView.j 
index d27d009..246f67d 100644 
--- a/AppKit/_CPImageAndTextView.j 
+++ b/AppKit/_CPImageAndTextView.j 
@@ -377,6 +377,8 @@ var HORIZONTAL_MARGIN   = 3.0, 
             shadowStyle.position = "absolute"; 
             shadowStyle.whiteSpace = textStyle.whiteSpace; 
+            shadowStyle.color = [_textShadowColor cssString]; 
+ 
             shadowStyle.zIndex = 150; 
             shadowStyle.textOverflow = textStyle.textOverflow; 

CPTableView

• (void)selectedRowIndexes { return _selectedRowIndexes; }

should be

• (CPIndexSet)selectedRowIndexes { return _selectedRowIndexes; }

Characters such as @ (alt + 2) aren't recognized anymore, I believe since the refactored key handling commits.

(Jake branch, up to date at the time of writing.)

Example App:
http://luddep.se/static/TextfieldAltBug/

I have the following CPPopUpButton in a CPView:

var sizeSelector = [[CPPopUpButton alloc] initWithFrame:CGRectMake(160,4,50,18) pullsDown:NO];
[sizeSelector addItemsWithTitles:["1","2","3","4","5","6","7"]]; [self addSubview:sizeSelector];

If I click the CPPopUpButton and then hover my mouse over an area outside of the CPMenu area that pops up I get an error:

index (-1) out of bounds (0 - 7)

I started seeing this when the index out of bounds exception was added a couple of days ago. Thanks.

• Ryan

Random bug:

NSTabView in IB .xib file converted into a .cib with nib2cib

the execution generates
TypeError: Result of expression 'aSubview' [null] is not an object.

The stack crawl show up that
- (void)insertTabViewItem:atIndex: was walled with a null view and a index of -1

Sometimes just changing the selected tab item in InterfaceBuilder, then save and re-run nib2cib clears the problem…

Steps to reproduce :
1. create a collection view
2. set its item prototype using with _ initWithCibName_ :

[[CPCollectionViewItem alloc] initWithCibName:@"TheCibName" bundle:nil]

If there are 100 items in the collection view, the cib will be loaded 100 times using XHR.
(You'll see something like this in the Webkit console : "100 XHR finished loading: TheCibName.cib")

The cib should be loaded only once and cached.

If using AppKit scrollers in a CPWebView, it will always show scrollers, as it doesn't subtract the width of the vertical scroller and the height of the horizontal scroller from the size of the IFrame.

CPString is missing NSString's commonPrefixWithString method that finds the common prefix between two strings.

CPGeometry CPRectContainsRect calls CPUnionRect in stead of CPRectUnion.

Possible fix: 974524f
It's probably easier to change the call yourself though :)

I saw this the first time at almost.at. The page was still loading while the status in Safari tells the user "Canceled opening the page" which is simply wrong because the page is still loading.

So first I thought "Ok maybe theres some error in the app, let's load it again". All in all the status message indicates the user that there is an error which is simply not a good thing.

I think this could simply be overwritten by adding a

window.status = "Some positive message";

If you set the window style to a CPAlert to HUD the buttons will be added with the Aristo theme instead.

to fix add

   if(_windowStyle === CPHUDBackgroundWindowMask)
        [button setTheme:[CPTheme themeNamed: "Aristo-HUD"]];

inside of addButtonWithTitle:

(I'd commit it, but I'm on the jake branch and I don't have my jake branch on github)

If the CPTableView is first responder and you do something like "cmd+s" to save (a menu shortcut) the tableview doesn't handle the keyboard event correctly.

Theres a bug in ie7 and ie8 when using 1x1 pixel transparent images for backgrounds. This is happening with the HUD window style. All that needs to be done is change the 1x1 pixel file to be a 5x5 pixel file.

Test case:

var zero = [CPNumber numberWithInt:0];
var set = [CPSet setWithObject:zero];
CPLog.trace(@"containsObject: %d", [set containsObject:zero]);

Oops :-(

if (_contents[[anObject UID]] && [_contents[[anObject UID]] isEqual:anObject]) should be replaced with
if ([anObject UID] in _contents && [_contents[[anObject UID]] isEqual:anObject])

from IRC:

saikatc: tlrobinson: btw, i just realized that after running bootstrap.sh, restarting the shell won't work as expected if the user has a .bash_profile or .bash_login. it should probably add the export path to the correct profile based on the unix priority order - http://hayne.net/MacDev/Notes/unixFAQ.html under "Bash startup files"

I believe this only occurs when the Cappuccino application has menu items with key equivalents for copy and paste. With the new way CPTextField works, these key equivalents get caught in CPApplication, I believe, and never hit the CPTextField. Not sure what the proper solution is.

I placed a tableView in a scrollView and set the documentView.
When you turn of the header in a CPTableView with

[tableView setHeaderView:nil];

the table data itself is on the right position but the scrollbars are on the same position like there were headers. Take a look at the screenshot: http://i33.tinypic.com/kdl3rs.png

[button setDefaultButton:YES] seems to only change the appearance of the button. Pressing the enter key does not send the action, even if the button is the only element in the window.

Hi, here is my config on Snow Leopard 10.6.2:

$gcc -v Using built-in specs.
Target: i686-apple-darwin10
Configured with: /var/tmp/gcc/gcc-5646.1~2/src/configure --disable-checking --enable-
werror --prefix=/usr --mandir=/share/man --enable-languages=c,objc,c++,obj-c++ --
program-transform-name=/^[cg][^.-]*$/s/$/-4.2/ --with-slibdir=/usr/lib --build=i686-
apple-darwin10 --with-gxx-include-dir=/include/c++/4.2.1 --program-prefix=i686-apple-
darwin10- --host=x86_64-apple-darwin10 --target=i686-apple-darwin10
Thread model: posix
gcc version 4.2.1 (Apple Inc. build 5646) (dot 1)

$java -v java version "1.6.0_17"
Java(TM) SE Runtime Environment (build 1.6.0_17-b04-237-10M3013)
Java HotSpot(TM) 64-Bit Server VM (build 14.3-b01-100, mixed mode)

$git --version git --version
git version 1.6.4.2

$rake --version rake, version 0.8.7

$echo $CAPP_BUILD /Library/Frameworks/cappuccino.framework

I clone cappuccino from github, I do a rake install and I get:

(in /Users/fred/dev/src/cappuccino.src) (in /Users/fred/dev/src/cappuccino.src) (in /Users/fred/dev/src/cappuccino.src/External) Submodule 'External/browserjs' (git://github.com/tlrobinson/browserjs.git) registered for path
'External/browserjs' Submodule 'External/jack' (git://github.com/tlrobinson/jack.git) registered for path
'External/jack' Submodule 'External/narwhal' (git://github.com/tlrobinson/narwhal.git) registered for path
'External/narwhal' Submodule 'External/ojunit' (git://github.com/280north/ojunit.git) registered for path
'External/ojunit' Initialized empty Git repository in /Users/fred/dev/src/cappuccino.src/External/browserjs/.git/
remote: Counting objects: 120, done.
remote: Compressing objects: 100% (103/103), done.
remote: Total 120 (delta 38), reused 0 (delta 0)
Receiving objects: 100% (120/120), 22.55 KiB, done.
Resolving deltas: 100% (38/38), done.
Submodule path 'External/browserjs': checked out
'98a71a4122bde9cacebe5b437e66c4b17208a195' Initialized empty Git repository in /Users/fred/dev/src/cappuccino.src/External/jack/.git/
remote: Counting objects: 2614, done.
remote: Compressing objects: 100% (1357/1357), done.
remote: Total 2614 (delta 1478), reused 2089 (delta 1179)
Receiving objects: 100% (2614/2614), 866.16 KiB | 226 KiB/s, done.
Resolving deltas: 100% (1478/1478), done.
Submodule path 'External/jack': checked out '3ee7b56812c97770318891a0fd21287cae9ccfc2'
Initialized empty Git repository in /Users/fred/dev/src/cappuccino.src/External/narwhal/.git/
remote: Counting objects: 7149, done.
remote: Compressing objects: 100% (2287/2287), done.
remote: Total 7149 (delta 4600), reused 7149 (delta 4600)
Receiving objects: 100% (7149/7149), 4.21 MiB | 229 KiB/s, done.
Resolving deltas: 100% (4600/4600), done.
Submodule path 'External/narwhal': checked out 'cd6e4b02ab7f039e2f106ef6f8edb3efbfab5b8a'
Initialized empty Git repository in /Users/fred/dev/src/cappuccino.src/External/ojunit/.git/
remote: Counting objects: 80, done.
remote: Compressing objects: 100% (74/74), done.
remote: Total 80 (delta 38), reused 0 (delta 0)
Receiving objects: 100% (80/80), 12.50 KiB, done.
Resolving deltas: 100% (38/38), done.
Submodule path 'External/ojunit': checked out
'352b7e8c618729a48d130098368ae0a774ce5f0d' mkdir -p /Library/Frameworks/cappuccino.framework/Debug
mkdir -p /Library/Frameworks/cappuccino.framework/Debug/env
rm -rf /Library/Frameworks/cappuccino.framework/Debug/env
(in /Users/fred/dev/src/cappuccino.src/Objective-J) mkdir -p /Library/Frameworks/cappuccino.framework/Debug/Objective-J
:576:26: warning: missing terminating " character
:840:26: warning: missing terminating " character
:841:22: warning: missing terminating " character
mkdir -p /Library/Frameworks/cappuccino.framework/Debug/Objective-J/rhino.platform
:576:26: warning: missing terminating " character
:840:26: warning: missing terminating " character
:841:22: warning: missing terminating " character
cp /Users/fred/dev/src/cappuccino.src/Tools/Rake/lib/licenses/LGPL-v2.1
/Library/Frameworks/cappuccino.framework/Debug/Objective-J/LICENSE (in /Users/fred/dev/src/cappuccino.src/Objective-J/Tools) rm -rf /Library/Frameworks/cappuccino.framework/Debug/env/packages/objj
cp -r objj /Library/Frameworks/cappuccino.framework/Debug/env/packages/objj
mkdir -p
/Library/Frameworks/cappuccino.framework/Debug/env/packages/objj/lib/Frameworks rm -rf
/Library/Frameworks/cappuccino.framework/Debug/env/packages/objj/lib/Frameworks/Objecti ve-J
cp -r /Library/Frameworks/cappuccino.framework/Debug/Objective-J
/Library/Frameworks/cappuccino.framework/Debug/env/packages/objj/lib/Frameworks/Objecti ve-J
(in /Users/fred/dev/src/cappuccino.src/Foundation) mkdir -p /Library/Frameworks/cappuccino.framework/Debug/Foundation/Resources
cp -r Resources/log.css
/Library/Frameworks/cappuccino.framework/Debug/Foundation/Resources/log.css Exception in thread "main" java.lang.NullPointerException

at org.mozilla.javascript.ScriptableObject.getTopLevelScope(ScriptableObject.java:1514)
at org.mozilla.javascript.ScriptRuntime.setObjectProtoAndParent(ScriptRuntime.java:3291)
at org.mozilla.javascript.NativeJavaPackage.getPkgProperty(NativeJavaPackage.java:166)
at org.mozilla.javascript.NativeJavaPackage.get(NativeJavaPackage.java:105)
at org.mozilla.javascript.ScriptableObject.getProperty(ScriptableObject.java:1575)
at org.mozilla.javascript.ScriptRuntime.getObjectProp(ScriptRuntime.java:1397)
at org.mozilla.javascript.ScriptRuntime.getObjectProp(ScriptRuntime.java:1383)
at org.mozilla.javascript.Interpreter.interpretLoop(Interpreter.java:3054)
at org.mozilla.javascript.Interpreter.interpret(Interpreter.java:2484)
at org.mozilla.javascript.InterpretedFunction.call(InterpretedFunction.java:162)
at org.mozilla.javascript.NativeArray.iterativeMethod(NativeArray.java:1565)
at org.mozilla.javascript.NativeArray.execIdCall(NativeArray.java:313)
at org.mozilla.javascript.IdFunctionObject.call(IdFunctionObject.java:127)
at org.mozilla.javascript.Interpreter.interpretLoop(Interpreter.java:3335)
at org.mozilla.javascript.Interpreter.interpret(Interpreter.java:2484)
at org.mozilla.javascript.InterpretedFunction.call(InterpretedFunction.java:162)
at org.mozilla.javascript.optimizer.OptRuntime.callName(OptRuntime.java:97)
at org.mozilla.javascript.gen.c1._c1(Unknown Source)
at org.mozilla.javascript.gen.c1.call(Unknown Source)
at org.mozilla.javascript.optimizer.OptRuntime.call2(OptRuntime.java:76)
at org.mozilla.javascript.gen.c1._c0(Unknown Source)
at org.mozilla.javascript.gen.c1.call(Unknown Source)
at org.mozilla.javascript.ContextFactory.doTopCall(ContextFactory.java:401)
at org.mozilla.javascript.ScriptRuntime.doTopCall(ScriptRuntime.java:3003)
at org.mozilla.javascript.gen.c1.call(Unknown Source)
at org.mozilla.javascript.gen.c1.exec(Unknown Source)
at org.mozilla.javascript.tools.shell.Main.evaluateScript(Main.java:526)
at org.mozilla.javascript.tools.shell.Main.processFileSecure(Main.java:448)
at org.mozilla.javascript.tools.shell.Main.processFile(Main.java:414)
at org.mozilla.javascript.tools.shell.Main.processSource(Main.java:405)
at org.mozilla.javascript.tools.shell.Main.processFiles(Main.java:179)
at org.mozilla.javascript.tools.shell.Main$IProxy.run(Main.java:100)
at org.mozilla.javascript.Context.call(Context.java:499)
at org.mozilla.javascript.ContextFactory.call(ContextFactory.java:511)
at org.mozilla.javascript.tools.shell.Main.exec(Main.java:162)
at org.mozilla.javascript.tools.shell.Main.main(Main.java:140)

Can anyone point me in the right direction?
Cheers, Fred.

please exclude .DS_Store files from copying to ./Build/Release/<project_name>

Firefox (and Opera) sent the keypress event to often which causes the textfield to sent it's action more than once when the enter key is pressed.

As an example, compare scrolling the sidebar in iTunes with a large number of playlists with scrolling the widgets library in Mockingbird (http://gomockingbird.com/mockingbird/). iTunes will speed up the scrolling when you perform small, quick scrolling motions, or longer, slower scrolling motions.

There is situation when hasSuffix returns wrong result:

var string = @"Test";

var suffix = @"Wrong";

var result = [string hasSuffix:suffix];

Returned result is equal to YES ...

Problem appears always when [string length] == [suffix length]-1

At least in Safari, hitting command+v, etc. results in a CPEvent with characters = "V" (should be "v").

This is a regression. This happens in both firefox and safari and on both release and debug frameworks. To test, create a secure text field and setStringValue to "test". Open the file and you will see only three bullets instead of four in the secure text field. Test gist - http://gist.github.com/223321

I believe this a regression. But, for example, if I set an action as the target of a text field, pressing enter on that text field will fire that action twice. Or, in another case, in the action of a text field I was setting another view as a first responder. Hitting enter in the text field was causing the action to fire, but then was also sending a keyDown with CPReturnKeyCode to the view that I was setting as first responder. This does not happen in Safari 4, but happens in Firefox 3 and Firefox 3.5.

looks like http://github.com/280north/cappuccino/blob/master/AppKit/CPTableView.j#L239 is a copy-paste typo from line 236

1. Create a new application
   
2. Make a custom CPWindowController subclass with a cib file
   
3. Override windowDidLoad and call [self window] from it
   
4. Load the window controller subclass from the cib in the applications delegate's applicationDidFinishLaunching
   
5. Call showWindow on the window controller
   
6. Load the application

The above steps will caused the windowDidLoad method to loop because there is no window outlet set up in the cib file. I think this behavior should be more like the iPhone where an exception is thrown if the view could not be loaded.

I've tried the starter application in Adobe AIR and I've got the following exception:
Error: Adobe AIR runtime security violation for JavaScript code in the application security sandbox (Function constructor)
objj_exception_report at app:/Frameworks/Objective-J/Objective-J.js : 1556
fragment_evaluate_code at app:/Frameworks/Objective-J/Objective-J.js : 2243
at app:/Frameworks/Objective-J/Objective-J.js : 2211 at app:/Frameworks/Objective-J/Objective-J.js : 2225

It looks like the "new Function()" is called after onLoad is trigger. In AIR all the evals should happen before onLoad.

Try the code here - http://gist.github.com/220843. Copy prints twice on each command-c.

Discussed with solution here - http://groups.google.com/group/objectivej-dev/browse_thread/thread/9f9a8764e4bade31

Here is a simple app showing the problem:
http://github.com/cacaodev/cappuccino/commit/4b1c27dde22284c870377d06142ce5b06b888059

Either there is a path in the context current path(*) or not, nothing appears on the view.
IE console outputs an error "Object does not support that property". The log included in this function is: "<cg_vml_:fill type="gradient" colors="0% rgb(0, 0, 0),100% rgb(255, 255, 255)" />"

(*) this is currently required for canvas browsers.

In linux (ubuntu 9.04) with firefox 3.5.2, operra 9.64 or midori 0.1.2 (webkit) the bottom of the button is not displed properly if you use a value miner of 24.0 in your button declaration for the height value. It's the same if I need a button with 50px * 50 px the displaying is very bad.

var button = [[CPButton alloc] initWithFrame: CGRectMake(
            CGRectGetWidth([contentView bounds])/2.0 - 40.0,
            CGRectGetMaxY([label frame]) + 10.0,
            80.0, 18.0
    )] ;

The menubar implementation does not support opening a dropdown menu, then hovering over another menu label to open that dropdown. An example of a proper implementation would be the menubar in Mac OS X.

The workaround is to put:

[tableView.headerView setNeedsDisplay:YES]; [tableView.headerView setNeedsLayout];

before:

[tableView reloadData];

CPPopUpButton has no disabled color.

Hi!

My port of NSOperation might be a good addition to the core: http://github.com/jfahrenkrug/CPOperation
Maybe, maybe not.

• Johannes

A native cursor support for elements like buttons, collectionViews etc. would be very important.

Of course, in native desktop apps there isn't the "pointer" cursor for example.

In the past I had some conversations with customers who didn't understand why there's no pointer-Cursor on a menu-item. And I understand the point.
This cursor got important when the web got bigger. It indicated that if you click on an element something would happen.

Because of cappuccino apps live in the browser it would be very important to supprt the feature to manually choose the cursor.

By the way at the moment you have to solve this problem by touching the DOM:

self._DOMElement.style.cursor = "pointer";

testItemPrototypeActuallyReturnsTheItemPrototype in the CPCollectionView test tests if the class CPCollectionViewItem are equal with each other. The test should test if the item prototype sent to the setter is identical to the one returned by the getter.

With last nights updates to CPAnimation to support curves (yay!) a critical bug seems to have slipped in.

Unless you set the curve directly using setAnimationCurve:, the _timingFunction ivar isn't set and everything goes bonkers.

initWithDuration:animationCurve: sets _animationCurve directly, simply replacing

_animationCurve = anAnimationCurve;

with

[self setAnimationCurve:anAnimationCurve];

in initWithDuration:animationCurve: seems to fix it.

CPCollectionView doesn't have a getter for isSelectable however it does have one for isSelected which doesn't exist.

The isSelectable property is read from the xib file but it's never written to the cib file or read from it for that matter.

Summary

When one subclass CPTextField.
Ant then in the textDidBeginEditing method
we try to set the placeHolderStringValue of the textField to an empty string "".
But the placeHolderStringValue for the Field does not get updated at least not visually.

The Case Scenario

We try to emulate an field where we toggle between the defaultValue and empty string "" when the element regains and loses focus.
Setting the value to back todefaultValue if it's an empty string
in textDidEndEditing works without any problems.

The code is the following.

  - (void) textDidBeginEditing: (CPNotification) aNote
  {
    if(([self stringValue] == defaultValue)){
        [self setPlaceholderString: ""];
    }
  }

An possible solution is the following that does the trick, is to make the element
first responder twice.

  - (void) textDidBeginEditing: (CPNotification) aNote
  {
    if(([self stringValue] == defaultValue)){
                [self setStringValue: ""];
            [textField becomeFirstResponder];
    }
  }

Or the easiest but not ideal solution.

  - (void) textDidBeginEditing: (CPNotification) aNote
  {
    [self _inputElement].value = ""; 
  }

Problem and Solution

This is due that the [textField _inputElement].value does not get updated
when the the element has become the first responder.

So the ideal solution would be for the CPTextField, to update the

[textField _inputElement].value = value;

if it knows that it has become the first responder.

Sample code

I am listening for the CPWindowWillCloseNotification notification and compare the object value with my CPPanel to detect whether it is my window that closed or not. When using the uncompiled code in my project it works as expected. When compiling it through rake, this no longer works but delivers "null" as the object value.

In addition, programmatically closing the panel using [myPanel close] has no effect. Manually closing closes the panel but, as said, does not send the correct notification. In the uncompiled code, it works as expected.

To recreate, open a CPPanel, then listen for CPWindowWillCloseNotification when you close the panel.

Since Safari 4 has been released, pressing the arrow keys no longer causes interpretKeyEvents: to be called.

To see the problem try this demo:

http://timetableapp.com/PR/datepicker/

In Firefox when of the date picker cells has focus, you can change the value using the up/down keys, and tab from one cell to another. Not so in Safari 4.

You can add a log statement in DatePicker.j (or presumable other CPResponder subclasses)

• (void)interpretKeyEvents:(id)anEvent { var key = [anEvent keyCode];

  console.log('here in DataPicker interpretKeyEvents key is ' +key);

  ......
  

  }

Notice that the log is visible in Firefox, not in Safari 4.

Ran 'press test pressed' on a project generated with: 'capp gen test'
commit cb416e0
Date: Sat Sep 26 18:49:08 2009 -0700
identical results on Ubuntu Juanty Jackalope and Mac OSX Leopard
Rake 0.8.7 on both systems
ran 'rake clobber-all install' for both installs.
Trace follows:

2009-08-29 16:27:55.214 Cappuccino [info]: Application root:    /Users/anthony/work/cappuccino/temp/test
2009-08-29 16:27:55.220 Cappuccino [info]: Output directory:    pressed
2009-08-29 16:27:55.244 Cappuccino [info]: Main file:           /Users/anthony/work/cappuccino/temp/test/main.j
2009-08-29 16:27:55.275 Cappuccino [info]: Frameworks:          /Users/anthony/work/cappuccino/temp/test/Frameworks
2009-08-29 16:27:55.381 Cappuccino [error]: PHASE 1: Loading application...
2009-08-29 16:27:55.542 Cappuccino [debug]: Evaling /Users/anthony/work/cappuccino/temp/test/main.j / /Users/anthony/work/cappuccino/temp/test/Info.plist
/Users/anthony/work/cappuccino/temp/test/main.j
ReferenceError: "system" is not defined.
js: "Objective-J.js", line 1734: exception from uncaught JavaScript throw: ReferenceError: "system" is not defined.
    at Objective-J.js:1734 (objj_exception_report)
    at Objective-J.js:2460 (fragment_evaluate_code)
    at /Users/anthony/work/cappuccino/objj_home/share/narwhal/packages/objj/lib/press/objj-analysis-tools.j:152
    at Objective-J.js:2422
    at Objective-J.js:2439
    at bridge.js:161 (serviceTimeouts)
    at <cmd>:5
    at /Users/anthony/work/cappuccino/objj_home/share/narwhal/packages/objj/lib/press/objj-analysis-tools.j:202
    at /Users/anthony/work/cappuccino/objj_home/share/narwhal/packages/objj/lib/press/objj-analysis-tools.j:158
    at /Users/anthony/work/cappuccino/objj_home/share/narwhal/packages/objj/lib/press/main.j:85
    at /Users/anthony/work/cappuccino/objj_home/share/narwhal/packages/objj/lib-js/objj/objj.js:96
    at Objective-J.js:2430
    at Objective-J.js:2438
    at /Users/anthony/work/cappuccino/objj_home/share/narwhal/packages/browserjs/lib/browser/timeout.js:103
    at /Users/anthony/work/cappuccino/objj_home/share/narwhal/packages/objj/lib-js/objj/objj.js:120
    at /Users/anthony/work/cappuccino/objj_home/share/narwhal/packages/objj/bin/objj:3
    at /USers/anthony/work/cappuccino/objj_home/share/narwhal/lib/sandbox.js:250
    at narwhal.js:209

I've created a CPTabView and am amble to add CPTabViewItem's but when I remove them the Tabs don't seem to be correctly redrawn. The old tab is drawn on top of the others. Below are 2 screenshots showing the bug.

Multiple tabs before removing:
http://tietema.net/wp-content/tabs_good.png

After removing:
http://tietema.net/wp-content/tabs_bad.png

Frameworks/AppKit/Resources/CPWindow/Standard/CPWindowStandardCLoseButtonHighlighted.png
has a typo - the "L" in "Close" is capitalized when it should not be. This causes the file not to be found on case-sensitive web servers.

This is a regression bug (just tried with 0.7.1 and the behavior works). If I add a separator CPMenuItem, anything I put after that separator does not move with the window's width and just gets cut off. Try the code here - http://gist.github.com/190119 .

https://gist.github.com/f04708b1e98c28f5de87

Trying to set header for tableview:
header = [[CPTableHeaderView alloc] initWithFrame: CGRectMake(10, 10, 1000, 10)];
[table setHeaderView: header];

Got error: ReferenceError: Can't find variable: aSize
Looks like problem here:

CPTableView.j:
705: [headerView setFrameSize:CGSizeMake(aSize.width, [_headerView frame].size.height)];

Using loadHTMLString to set the contents of the webView calls [self _setScrollMode:CPWebViewScrollAppKit];
I don't know why. And loadHTMLString should, imho, not touch the scrollMode.

Cappuccino should be able to do a better job reporting such errors (or include a server-side validation tool or some form). Currently, a typo in an Objective-J file is reported as a ParseError with no additional clue as to what was being parsed or how it is wrong.

For example, I had some code:

- (CPButton)makeDigit:(int)value at:(CGRect)frame
{
  CPButton button = [self makeButton:value at:frame 
                     action:@selector(digit:)];

  [button setTag:value];

  return button;
}

It took me a long time to realize that in Objective-J, local variables have no type (always var) and it would have been nice if Cappuccino had identified the file and preferably the method (and line) that was at the root of the problem, as well as what I was doing wrong.

I was looking for documentation for the setAlignment: method and it was nowhere to be found; perusing the source I saw it was created via a macro:

#define BRIDGE(UPPERCASE, LOWERCASE, ATTRIBUTENAME) \
/*! Sets the value for ATTRIBUTENAME */\
- (void)set##UPPERCASE:(id)aValue\
{\
[self setValue:aValue forThemeAttribute:ATTRIBUTENAME];\
}\
/*! Returns the current value for ATTRIBUTENAME */\
- (id)LOWERCASE\
{\
return [self valueForThemeAttribute:ATTRIBUTENAME];\
}

BRIDGE(Alignment, alignment, "alignment")
BRIDGE(VerticalAlignment, verticalAlignment, "vertical-alignment")
BRIDGE(LineBreakMode, lineBreakMode, "line-break-mode")
BRIDGE(TextColor, textColor, "text-color")
BRIDGE(Font, font, "font")
BRIDGE(TextShadowColor, textShadowColor, "text-shadow-color")
BRIDGE(TextShadowOffset, textShadowOffset, "text-shadow-offset")
BRIDGE(ImagePosition, imagePosition, "image-position")
BRIDGE(ImageScaling, imageScaling, "image-scaling")

Look at

http://cappuccino.org/learn/documentation/group__appkit.html

It's in a pretty random order; it should be in alphabetical order.

When autorecalculatesKeyViewLoop is set to NO the keyViewLoopIsDirty boolean is not set to NO as well. This results in the key view loop still being automatically recalculated.

Haven't tested it but while reading through the code, I noticed the for loop was using index but incrementing i. Here's an untested patch.

index 566eb57..ecc1b47 100755
--- a/Foundation/CPArray.j
+++ b/Foundation/CPArray.j
@@ -209,13 +209,13 @@
         var index = 0,
             count = [anArray count];

-        for(; index < count; ++i)
+        for(; index < count; ++index)
         {
-            if (anArray[i].isa)
-                self[i] = [anArray copy];
+            if (anArray[index].isa)
+                self[index] = [anArray copy];
             // Do a deep/shallow copy?
              else
-                self[i] = anArray;
+                self[index] = anArray;
         }
     }

See http://github.com/280north/cappuccino/blob/master/AppKit/CPTextField.j#L658

Notification should be CPControlTextDidEndEditingNotification but is CPControlTextDidBeginEditingNotification.

If the a dimenions of the document view's size is less than that dimension in the containing CPScrollView's size but more than that dimension in the size of the CPScrollView minus a scroller, an incorrect disabled scroller appears. For example, if I have a document that is 100x100, and my scroll view is 90 x 105, then a horizontal scroller appears as expected. However, since this horizontal scroller makes the effective height of the scroll view 105 - the height of the horizontal scroller (about 17 pixels right now), the CPScrollView's size becomes 90 x 88. This makes the height of the CPScrollView smaller than the height of the document view, and causes a vertical scroller to also appear - however this vertical scroller appears disabled and, so, I am not able to scroll to clipped parts of my document. The same happens in the other dimension.

Trying to build from source I get the following error and warnings:

cp -r Templates/NibApplication/Resources/MainMenu.xib /Users/mattmini/dev/src/cappuccino/Build/Release/capp/Resources/Templates/NibApplication/Resources/MainMenu.xib
2009-07-22 15:03:43.176 Cappuccino [fatal]: Auto-detected Cocoa Nib or Xib File
i-nit form coder
2009-07-22 15:03:54.450 Cappuccino [warn]: Mapping NSApplication to CPApplication
:::CPApplication 2009-07-22 15:03:54.730 Cappuccino [debug]: => isBordered=false, isBezeled=true, bezelStyle=0(, null)
i-nit form coder
2009-07-22 15:03:54.995 Cappuccino [warn]: Mapping NSApplication to CPApplication
:::CPApplication i-nit form coder
:::AppController 2009-07-22 15:03:55.147 Cappuccino [warn]: Mapping NSWindow to CPWindow

I got this code:

var request = [CPURLRequest requestWithURL:@"http://myJSONPsite.com"];
[CPJSONPConnection connectionWithRequest:request callback:@"callback" delegate:self];

If there is data, the didReceiveData delegate gets called. In case of an error (wrong url, etc.) the didFailOnError delegate won't get called.
If there's data received, the didFinishLoading delegate also won't get called.

So it seems there's actually no way to respond to connection errors. (Version 0.7.1)

Enabling and Setting focus on a CPTextField and assigning it as the first responder adds an odd behaviour. Whenever in a view collectionView with more than one CPTextField in each CollectionView the value of the last edited textfield will pop as the new value for the new selected textfield.

An workaround is to have the textfield enabled and have an overlay CPView with transparency.
Then hide and show the CPView overlay over the TextField.

Attached are the following files.

The code http://pyced.com/capp/scrollview.zip
Initial Edit State http://pyced.com/capp/Picture.png
Editing and Jump forth an back between the fields http://pyced.com/capp/PictureBug.png

When creating a CPTableView, the frame passed in via initWithFrame is not respected. Instead, the tableView expands to fill it's container's width.

A CPTextField with its target and action set might not always return the full contents of itself when you send stringValue to it in the action method body.

Example:

someTextField = [CPTextField textFieldWithStringValue:nil
                             placeholder:@"blah"
                             width:200];
[someTextField setTarget:self];
[someTextField setAction:@selector(echoInputFrom:)];
....
- (IBAction)echoInputFrom:(id)sender
{
    CPLog.debug([sender stringValue]);
}

Type some stuff in the text field and hit enter quickly after typing or while still in the middle of editing.

asdfasdfasfdasdf
-> asdfasdfasf
1234567
-> 123456

Solution: right now I just wrap the body of the action method in a setTimeout(..., 0). I'm assuming the real way to do this would be to correctly insert something into the runloop?

var someButton = [CPButton buttonWithTitle:@""];
[someView addSubview:someButton]; [someButton setTitle:@"Messed up"];

And the button text will look like it's bold and slightly off.