Nabble - qooxdoo

DateChooser Problem

2008-10-10T17:15:48Z

Hi!
Does anybody notice that the DateChooser component shows October 11th twice?
I've tested both 0.7.3 and 0.8 versions.

Regards,

Marcio Shimoda

From forum: qooxdoo-devel

Re: javascript anomaly with regexp literal?

2008-10-10T10:49:00Z

> > There's a simpler case which exhibits behaviour that is unexpected:
> >
>
> In your "simpler case", I think it's doing exactly the documented
> behavior, and demonstrates precisely the reason for the existence of
> lastIndex. When a regexp.test() returns true, it sets lastIndex to
> where it found the first match that follows the pre-call lastIndex
> position.

OK, that is actually documented in the Definitive Guide (Flanagan) under
RegExp.lastIndex. But confusingly not under RegExp.test() though it does
refer to lastIndex.

Hugh

-------------------------------------------------------------------------
This SF.Net email is sponsored by the Moblin Your Move Developer's challenge
Build the coolest Linux based applications with Moblin SDK & win great prizes
Grand prize is a trip for two to an Open Source event anywhere in the world
http://moblin-contest.org/redirect.php?banner_id=100&url=/
_______________________________________________
qooxdoo-devel mailing list
qooxdoo-devel@...
https://lists.sourceforge.net/lists/listinfo/qooxdoo-devel

From forum: qooxdoo-devel

Re: generate build fails

2008-10-10T10:48:26Z

This was because one of the contrib’s I was using needed the optimisation for renaming private members turned off; it would be great if the optimisation was specified in the Manifest.json of the individual module so that generator.py could either use this information for that module or at least give an error that the application’s optimisation is incompatible.

John

From: John Spackman [mailto:john.spackman@...]
Sent: 10 October 2008 13:04
To: 'qooxdoo Development'
Subject: [qooxdoo-devel] generate build fails

Hi,

I’m having problems generating a build version of my app; the generate.py runs fine without any errors but when I go to run the app under FF3 or IE7 I get this error:

this.__hR.__hV is undefined

this.__hR.__hV[m].customUndo=true;

I’ve tried this with current download and Tuesday’s SVN with the same result, each time deleting the cache and build directories.

The source version runs fine, although slowly.

Any ideas what’s causing this?

John

-------------------------------------------------------------------------
This SF.Net email is sponsored by the Moblin Your Move Developer's challenge
Build the coolest Linux based applications with Moblin SDK & win great prizes
Grand prize is a trip for two to an Open Source event anywhere in the world
http://moblin-contest.org/redirect.php?banner_id=100&url=/
_______________________________________________
qooxdoo-devel mailing list
qooxdoo-devel@...
https://lists.sourceforge.net/lists/listinfo/qooxdoo-devel

From forum: qooxdoo-devel

Re: Info Regarding Splitplane

2008-10-10T08:38:49Z

On Fri, Oct 10, 2008 at 11:04 AM, Gaurav Jauhri <gaurav.jauhri@...> wrote:

Thanks for your reply.
I am sorry that i forgot to mention that i use 0.7 version and here there are no composite types available.

No problem. In 0.7, instead of a a Composite, you can use the HorizontalBoxLayout and VerticalBoxLayout directly.

var page = new qx.ui.container.Composite(new qx.ui.layout.HBox(4));
var left = new qx.ui.container.Composite(new qx.ui.layout.VBox());
page.add(left, { flex : 1 });
var right = new qx.ui.container.Composite(new qx.ui.layout.VBox());
page.add(right, { flex : 1 });

From memory (it's been a while since I've done 0.7 work) that would map to something like this:

var page = new qx.ui.layout.HorizontalBoxLayout();
page.setSpacing(4);
var left = new qx.ui.layout.VerticalBoxLayout();
left.setWidth(100); // your small-width left-side vbox
page.add(left);
var right = new qx.ui.layout.VerticalBoxLayout();
right.setWidth("1*"); // let this one take up the remaining space
page.add(right);

Does that accomplish what you're looking for?

Derrell

From forum: qooxdoo-devel

Re: Info Regarding Splitplane

2008-10-10T08:04:44Z

Hi Derell,

Thanks for your reply.
I am sorry that i forgot to mention that i use 0.7 version and here
there are no composite types available.

Also I have my Docklayout and inside it i have Horizontal Splitplane. On
the left side I have a VericalBox containing Buttons whose Width is
small, and hence i need to fix it, so that I cannot drag over to the
widgets i have on the right hand side.

And on the right hand part of it, i further have vertical and horizontal
split planes.
Can you please tell me some property if applicable, so that i can get
the desired results.
I am attaching a Screenshot to give u a better idea.

Regards,
Gaurav

Derrell Lipman ha scritto:

> On Fri, Oct 10, 2008 at 10:11 AM, Gaurav Jauhri <gaurav.jauhri@...
> <mailto:gaurav.jauhri@...>> wrote:
>
> Hi,
>
> I have a splitpane (Horizontal) dividing the screen into left and
> right.
> I want that the screen be constant so that it cannot be resized.
> I tried doing *this.setLiveResize(false);
> *But it does not change, and still i can the left and right part
> on both
> the sides.
> Any suggestions would be of help.
>
>
> If you don't want it resizable then you probably don't need a
> splitpane. Instead just add to VBox-layout composites to an HBox:
>
> var page = new qx.ui.container.Composite(new qx.ui.layout.HBox(4));
> var left = new qx.ui.container.Composite(new qx.ui.layout.VBox());
> page.add(left, { flex : 1 });
> var right = new qx.ui.container.Composite(new qx.ui.layout.VBox());
> page.add(right, { flex : 1 });
>
> Now you can add stuff to left and right.
>
> You could also use other layouts, e.g. Dock, to accomplish the same thing.
>
> Derrell
>
>
> ------------------------------------------------------------------------
>
> -------------------------------------------------------------------------
> This SF.Net email is sponsored by the Moblin Your Move Developer's challenge
> Build the coolest Linux based applications with Moblin SDK & win great prizes
> Grand prize is a trip for two to an Open Source event anywhere in the world
> http://moblin-contest.org/redirect.php?banner_id=100&url=/
> ------------------------------------------------------------------------
>
> _______________________________________________
> qooxdoo-devel mailing list
> qooxdoo-devel@...
> https://lists.sourceforge.net/lists/listinfo/qooxdoo-devel
>

Splitplane.GIF (96K) Download Attachment

From forum: qooxdoo-devel

Re: Info Regarding Splitplane

2008-10-10T07:22:51Z

On Fri, Oct 10, 2008 at 10:11 AM, Gaurav Jauhri <gaurav.jauhri@...> wrote:

Hi,

I have a splitpane (Horizontal) dividing the screen into left and right.
I want that the screen be constant so that it cannot be resized.
I tried doing *this.setLiveResize(false);
*But it does not change, and still i can the left and right part on both
the sides.
Any suggestions would be of help.

If you don't want it resizable then you probably don't need a splitpane. Instead just add to VBox-layout composites to an HBox:

var page = new qx.ui.container.Composite(new qx.ui.layout.HBox(4));
var left = new qx.ui.container.Composite(new qx.ui.layout.VBox());
page.add(left, { flex : 1 });
var right = new qx.ui.container.Composite(new qx.ui.layout.VBox());
page.add(right, { flex : 1 });

Now you can add stuff to left and right.

You could also use other layouts, e.g. Dock, to accomplish the same thing.

Derrell

From forum: qooxdoo-devel

Disabled state forwarding to children

2008-10-10T07:21:16Z

As a follow-up to my discussion in:
http://bugzilla.qooxdoo.org/show_bug.cgi?id=1459

I'd like to know what is causing this behavior:

If I disable a container, its children are disabled too. I find it
pretty logical and very convenient this way, but I'd like to
understand why/how it happens. I've searched the code but couldn't
find anything which could be causing this. All I know is that it
apparently doesn't happen because of the _forwardStates mechanism.
Also, given the reactions in the above bug report, I'm beginning to
doubt whether this is an intended feature or just an accident?
Apparently this behavior works/is present most of the time, but not
always (apparently if the parent was not visible when it's disabled
can cause a problem), hence my discussion in the linked bug report.

Thanks for any enlightenment.
--
Gaëtan de Menten
http://openhex.org

-------------------------------------------------------------------------
This SF.Net email is sponsored by the Moblin Your Move Developer's challenge
Build the coolest Linux based applications with Moblin SDK & win great prizes
Grand prize is a trip for two to an Open Source event anywhere in the world
http://moblin-contest.org/redirect.php?banner_id=100&url=/
_______________________________________________
qooxdoo-devel mailing list
qooxdoo-devel@...
https://lists.sourceforge.net/lists/listinfo/qooxdoo-devel

From forum: qooxdoo-devel

Re: Info Regarding Splitplane

2008-10-10T07:20:09Z

Gaurav,

Why are you using splitpane if you need constant sizes? Try using
qx.ui.container.Composite with qx.ui.layout.Dock instead. Given that qx
0.8 allows separators in Dock layout - I think you'll be fine with that.

Thanks,
Oleksiy

Gaurav Jauhri wrote:

> Hi,
>
> I have a splitpane (Horizontal) dividing the screen into left and right.
> I want that the screen be constant so that it cannot be resized.
> I tried doing *this.setLiveResize(false);
> *But it does not change, and still i can the left and right part on both
> the sides.
> Any suggestions would be of help.
>
> Regards,
> Gaurav
>
> -------------------------------------------------------------------------
> This SF.Net email is sponsored by the Moblin Your Move Developer's challenge
> Build the coolest Linux based applications with Moblin SDK & win great prizes
> Grand prize is a trip for two to an Open Source event anywhere in the world
> http://moblin-contest.org/redirect.php?banner_id=100&url=/
> _______________________________________________
> qooxdoo-devel mailing list
> qooxdoo-devel@...
> https://lists.sourceforge.net/lists/listinfo/qooxdoo-devel
>

From forum: qooxdoo-devel

Info Regarding Splitplane

2008-10-10T07:11:10Z

Hi,

I have a splitpane (Horizontal) dividing the screen into left and right.
I want that the screen be constant so that it cannot be resized.
I tried doing *this.setLiveResize(false);
*But it does not change, and still i can the left and right part on both
the sides.
Any suggestions would be of help.

Regards,
Gaurav

-------------------------------------------------------------------------
This SF.Net email is sponsored by the Moblin Your Move Developer's challenge
Build the coolest Linux based applications with Moblin SDK & win great prizes
Grand prize is a trip for two to an Open Source event anywhere in the world
http://moblin-contest.org/redirect.php?banner_id=100&url=/
_______________________________________________
qooxdoo-devel mailing list
qooxdoo-devel@...
https://lists.sourceforge.net/lists/listinfo/qooxdoo-devel

From forum: qooxdoo-devel

Re: Rounded borders

2008-10-10T07:04:38Z

I'd like to read that "for dummies" appearance article too! :-) .
-Chris

> Hi all,
>
> I'm trying to create round-bordered widget. I'm investigating qooxdoo's
> appearance declaration (in 'window'-related part) although it looks a
> bit like magic to me (yes, I've read appearance-related and image
> clipping/combining qooxdoo documentation:)). The problem is that
> documentation doesn't describe more or less complicated cases (like
> rounded borders or gradients).
>
> Could anyone, please, provide detailed instruction (sort of "Rounded
> borders in qooxdoo for dummies"), how I could achieve that?
>
> Thanks,
> Oleksiy
>
>
> -------------------------------------------------------------------------
> This SF.Net email is sponsored by the Moblin Your Move Developer's challenge
> Build the coolest Linux based applications with Moblin SDK & win great prizes
> Grand prize is a trip for two to an Open Source event anywhere in the world
> http://moblin-contest.org/redirect.php?banner_id=100&url=/
> _______________________________________________
> qooxdoo-devel mailing list
> qooxdoo-devel@...
> https://lists.sourceforge.net/lists/listinfo/qooxdoo-devel
>
>
>

From forum: qooxdoo-devel

Rounded borders

2008-10-10T06:53:30Z

Hi all,

I'm trying to create round-bordered widget. I'm investigating qooxdoo's
appearance declaration (in 'window'-related part) although it looks a
bit like magic to me (yes, I've read appearance-related and image
clipping/combining qooxdoo documentation:)). The problem is that
documentation doesn't describe more or less complicated cases (like
rounded borders or gradients).

Could anyone, please, provide detailed instruction (sort of "Rounded
borders in qooxdoo for dummies"), how I could achieve that?

Thanks,
Oleksiy

-------------------------------------------------------------------------
This SF.Net email is sponsored by the Moblin Your Move Developer's challenge
Build the coolest Linux based applications with Moblin SDK & win great prizes
Grand prize is a trip for two to an Open Source event anywhere in the world
http://moblin-contest.org/redirect.php?banner_id=100&url=/
_______________________________________________
qooxdoo-devel mailing list
qooxdoo-devel@...
https://lists.sourceforge.net/lists/listinfo/qooxdoo-devel

From forum: qooxdoo-devel

Re: javascript anomaly with regexp literal?

2008-10-10T05:35:31Z

On Fri, Oct 10, 2008 at 3:12 AM, Hugh Gibson <hgibson@...> wrote:

> *Subject:* [qooxdoo-devel] javascript anomaly with regexp literal?
> *From:* "Derrell Lipman" <derrell.lipman@...>
> *To:* "qooxdoo Development" <qooxdoo-devel@...>
> *Date:* Thu, 9 Oct 2008 12:36:13 -0400

>
> This seems odd and I can't find any documentation stating why it's
> so...
>
> This application has a function which tests a string for containing
> only digits. It uses a regular expression literal assigned to a
> local variable in a function. That regular expression, however,
> appears to be persistent across calls to the function.

Hi Hugh,

Thanks for looking at this and especially for the links showing that my expectations aren't out of whack. :-)

There's a simpler case which exhibits behaviour that is unexpected:

In your "simpler case", I think it's doing exactly the documented behavior, and demonstrates precisely the reason for the existence of lastIndex. When a regexp.test() returns true, it sets lastIndex to where it found the first match that follows the pre-call lastIndex position. Initially, with lastIndex at 0, it finds the very first match. Using a string which contains multiple matches, though, you can call regexp.test() again with the same string and obtain the next match, as shown here:

/*
#require(qx.log.appender.Native)
#require(qx.log.appender.Console)
*/

qx.Class.define("custom.Application",
{
extend : qx.application.Standalone,
members :
{
    main: function()
    {
      this.base(arguments);

      var re1 = /\?/g;
      this.debug("lastIndex=" + re1.lastIndex);
      this.debug("test=" + re1.test("a?a?a?"));
      this.debug("lastIndex=" + re1.lastIndex);
      this.debug("test=" + re1.test("a?a?a?"));
      this.debug("lastIndex=" + re1.lastIndex);
      this.debug("test=" + re1.test("a?a?a?"));
      this.debug("lastIndex=" + re1.lastIndex);
      this.debug("test=" + re1.test("a?a?a?"));
      this.debug("lastIndex=" + re1.lastIndex);
    }
});

which yields this:

3323ms custom.Application[d]: lastIndex=0

3330ms custom.Application[d]: test=true

3337ms custom.Application[d]: lastIndex=2

3343ms custom.Application[d]: test=true

3349ms custom.Application[d]: lastIndex=4

3355ms custom.Application[d]: test=true

3361ms custom.Application[d]: lastIndex=6

3368ms custom.Application[d]: test=false

3383ms custom.Application[d]: lastIndex=0

Thanks again,

Derrell

From forum: qooxdoo-devel

Re: HtmlArea for 0.8 still not usable

2008-10-10T05:07:59Z

Hi guys,

I've uploaded my sample and the code to
http://www.zenesis.com/qx/dev/myapp/source/ - there's a readme at
http://www.zenesis.com/qx/dev/README.txt and if you want to take a look at
the code it's http://www.zenesis.com/qx/dev.zip.

John

-----Original Message-----
From: John Spackman [mailto:john.spackman@...]
Sent: 09 October 2008 12:32
To: 'qooxdoo Development'
Subject: Re: [qooxdoo-devel] HtmlArea for 0.8 still not usable

Hi Petr,

I'm very happy to collaborate on this - although I suspect I'll be more a
user of HtmlArea itself than a major contributor to it. I think that there
is a reasonable amount of work to add rich controls to the editor and make
it a very accessible and easy to use component, especially for the smarter
features like image pickers and other dialogs etc so it would be great to
work together on this. Is it possible for you to put your RichEditor on the
web somewhere? I need to finish some things off today but upload it later
on for you to look at.

Regards,
John

-----Original Message-----
From: Petr Kobalíček [mailto:kobalicek.petr@...]
Sent: 09 October 2008 11:57
To: qooxdoo Development
Subject: Re: [qooxdoo-devel] HtmlArea for 0.8 still not usable

Hi Guys,

I'm happy for too much comments here :-)

First

a) about [HTMLSpanElement] bug:

I'm corrected it, but unfortunally there is something wrong with my
code (see latest commit), because it not behaves good (I don't know
how to document it, there are growing number if <p> tags and I don't
know).

b) about anything else :)

I have sometimes time to contribute to HtmlArea, I have read early all
code, but I have sometimes problems to understand it at all (it's not
badly written, but there seems be too much bugs in browsers and Html
editing is COMPLEX to implement).

Problem for me is that the "firefox focus bug" is still there, try this
code:

qx.Class.define("bfly.admin.Application",
{
extend : qx.application.Standalone,

members :
{
// [Entry point]
main: function()
{
// [SuperClass]
this.base(arguments);

// [Main Container]
this._container = new qx.ui.container.Composite(
new qx.ui.layout.VBox().set({
spacing: 1
})
);
this.getRoot().add(this._container, {edge:0});

// [MenuBar]
this._tabView = new qx.ui.tabview.TabView();
this._container.add(this._tabView, {flex: 1});

for (var i = 0; i < 3; i++)
{
var page = new qx.ui.tabview.Page("Page " + i);
var area = new htmlarea.HtmlArea();

page.setLayout(new qx.ui.layout.Canvas);
page.add(area, {edge: 0});

this._tabView.add(page);
}
}
}
});

When you use TABs, you can't focus to previously hidden HtmlAreas in
Firefox.

c) And for John,

John, I'm doing something very similar to your work. I called it
RichEditor and it's HtmlArea with toolbars with some customizing way.
My code depends to our system, but I try to make it generic. We can
colaborate on this if you want.

Cheers
- Petr

2008/10/9 Fabian Jakobs <fabian.jakobs@...>:
> Andreas Ecker schrieb:
>> Hi John!
>>
>>
>>> I realise Alexander is on holiday right now so I'm leaving this as a
note
>>> for when he returns as much as anything.
>>>
>>> I'd like to add my vote to this - a rich text editor is important and
we've

>>> abandoned the FCKeditor integration because it's too slow.
>>>
>>
>> Would be great we could collaborate on improving HtmlArea. Such a widget
>> requires quite some community effort to iron out immanent cross-browser
>> issues.
>>
>>
>>> I've nearly finished a wrapper around HtmlArea which adds toolbars for
>>> styling, formatting, images, links, etc in one Widget; I'd like to
>>> contribute it as part of the HtmlArea package but obviously don't want

to go
>>> steaming in until Alexander is back. I'll post it here if anyone wants
to

>>> take a look in the mean time.
>>>
>>
>> Yes, that could become a fine contribution to qooxdoo-contrib for people
>> interested in an out-of-the-box solution for an HTML editor. Actually,
>> we never wanted to package up such a combined solution (other than the
>> sample included with HtmlArea), as one would need to support so many
>> variants that it easily becomes (again) a heavy-weight configuration
>> thingy. Attaching toolbars and other elements to HtmlArea should be
>> fairly easy in qooxdoo to satisfy individual requirements. But if you
>> think you've come up with a solution that could be widely used, that's
>> great.
>>
>> Anyway, as I suggested it should rather be a separate contribution
>> (maybe called "HtmlEditor") next to the existing HtmlArea. The area
>> itself should stay a plain html editing widget without custom chrome
>> (toolbars, etc.). Well, we'll figure out the details, also when Alex is
>> back.
>>
> I absolutely agree that the core HtmlArea widget should only focus on
> plain Html editing but why not include it into the current HtmlArea as
> an example application. It is not required to have just one example
> application in a contrib project. This way example editor is always in
> sync with the editing component and serves as a good example.
>> What do you think? In the meantime you could have a look at
>> http://contrib.qooxdoo.org for the contribution details and then email
>> me directly to answer all your questions and getting you started,
>>
>>
> Best Fabian
>
>
> --
> Fabian Jakobs
> JavaScript Framework Developer
>
> 1&1 Internet AG
> Brauerstraße 48
> 76135 Karlsruhe
>
> Amtsgericht Montabaur HRB 6484
>
> Vorstand: Henning Ahlert, Ralph Dommermuth, Matthias Ehrlich, Thomas

Gottschlich, Matthias Greve, Robert Hoffmann, Markus Huhn, Oliver Mauss,
Achim Weiss
> Aufsichtsratsvorsitzender: Michael Scheeren
>
>
> -------------------------------------------------------------------------
> This SF.Net email is sponsored by the Moblin Your Move Developer's
challenge
> Build the coolest Linux based applications with Moblin SDK & win great
prizes
> Grand prize is a trip for two to an Open Source event anywhere in the
world
> http://moblin-contest.org/redirect.php?banner_id=100&url=/
> _______________________________________________
> qooxdoo-devel mailing list
> qooxdoo-devel@...
> https://lists.sourceforge.net/lists/listinfo/qooxdoo-devel
>

-------------------------------------------------------------------------
This SF.Net email is sponsored by the Moblin Your Move Developer's challenge
Build the coolest Linux based applications with Moblin SDK & win great
prizes
Grand prize is a trip for two to an Open Source event anywhere in the world
http://moblin-contest.org/redirect.php?banner_id=100&url=/
_______________________________________________
qooxdoo-devel mailing list
qooxdoo-devel@...
https://lists.sourceforge.net/lists/listinfo/qooxdoo-devel

-------------------------------------------------------------------------
This SF.Net email is sponsored by the Moblin Your Move Developer's challenge
Build the coolest Linux based applications with Moblin SDK & win great
prizes
Grand prize is a trip for two to an Open Source event anywhere in the world
http://moblin-contest.org/redirect.php?banner_id=100&url=/
_______________________________________________
qooxdoo-devel mailing list
qooxdoo-devel@...
https://lists.sourceforge.net/lists/listinfo/qooxdoo-devel

-------------------------------------------------------------------------
This SF.Net email is sponsored by the Moblin Your Move Developer's challenge
Build the coolest Linux based applications with Moblin SDK & win great prizes
Grand prize is a trip for two to an Open Source event anywhere in the world
http://moblin-contest.org/redirect.php?banner_id=100&url=/
_______________________________________________
qooxdoo-devel mailing list
qooxdoo-devel@...
https://lists.sourceforge.net/lists/listinfo/qooxdoo-devel

From forum: qooxdoo-devel

generate build fails

2008-10-10T05:03:44Z

Hi,

I’m having problems generating a build version of my app; the generate.py runs fine without any errors but when I go to run the app under FF3 or IE7 I get this error:

this.__hR.__hV is undefined

this.__hR.__hV[m].customUndo=true;

I’ve tried this with current download and Tuesday’s SVN with the same result, each time deleting the cache and build directories.

The source version runs fine, although slowly.

Any ideas what’s causing this?

John

From forum: qooxdoo-devel

validators gone in 0.8

2008-10-10T04:54:18Z

Folks,

Every day I am learning new things while working with qooxdoo 0.8.
Today I tried to migrate some old code I found that it seems that
the 'setValidator' method has disapeared in 0.8. Is there a new
'official' way for adding validators or is this part of the
framework 'gone' ?

cheers
tobi
--
Tobi Oetiker, OETIKER+PARTNER AG, Aarweg 15 CH-4600 Olten, Switzerland
http://it.oetiker.ch tobi@... ++41 62 775 9902 / sb: -9900

-------------------------------------------------------------------------
This SF.Net email is sponsored by the Moblin Your Move Developer's challenge
Build the coolest Linux based applications with Moblin SDK & win great prizes
Grand prize is a trip for two to an Open Source event anywhere in the world
http://moblin-contest.org/redirect.php?banner_id=100&url=/
_______________________________________________
qooxdoo-devel mailing list
qooxdoo-devel@...
https://lists.sourceforge.net/lists/listinfo/qooxdoo-devel

From forum: qooxdoo-devel

Custom job: "buildsrc"

2008-10-10T04:37:40Z

hi,

I'd like to have an intermediate job, between "generate source" and
"generate build": it's not the source job because I need all the classes
used to be in the target directory (the script runs with a reference to
qooxdoo sdk), and it's not the build job because I need the target code
to be as in the source (readable). Basically, it's like the source, but
standalone.

I'm not sure if it is possible (I'm afraid it's not), but at least I'd
like to get a readable target code; there are some options
http://qooxdoo.org/documentation/0.8/generator_config_ref#compile-dist
for that purpose, but I didn't yet understand how to overwrite
"optimize" : []

So far, my config.json file looks like this:

"build" :
{
"extend" : ["include-contrib", "appconf::build"],
"run" :
[
"appconf::build",
"build-files"
]
},
("build-files" refers to a custom operation)

How to write a new job "buildsrc" that extends this one, and that set
"optimize" : [] for the task compile-dist ?

"buildsrc" :
{
"extend" : ["build"],
...and now ?

Thanks

--
Cordialement,

///
(. .)
--------ooO--(_)--Ooo--------
| Philippe Poulard |
-----------------------------
http://reflex.gforge.inria.fr/
Have the RefleX !

-------------------------------------------------------------------------
This SF.Net email is sponsored by the Moblin Your Move Developer's challenge
Build the coolest Linux based applications with Moblin SDK & win great prizes
Grand prize is a trip for two to an Open Source event anywhere in the world
http://moblin-contest.org/redirect.php?banner_id=100&url=/
_______________________________________________
qooxdoo-devel mailing list
qooxdoo-devel@...
https://lists.sourceforge.net/lists/listinfo/qooxdoo-devel

From forum: qooxdoo-devel

Re: Qooxdoo python version required == 2.4?

2008-10-10T03:24:44Z

Gaetan de Menten schrieb:

> Hi there,
>
> I've read in the doc that python requirements is 2.4+, though in the
> code, I've seen many occurrences of pre-2.4 code style, for example
> "dict.has_key(key)" instead of "key in dict", do you confirm that the
> requirements are indeed 2.4+ and that I can change old code to use 2.4
> style?
>
> Thanks,
>

I think it would be good style to use 2.4 features where appropriate.
However I would like to hear Thomas' view on this before taking action.
He is back in office on Monday.

Best Fabian

--
Fabian Jakobs
JavaScript Framework Developer

1&1 Internet AG
Brauerstraße 48
76135 Karlsruhe

Amtsgericht Montabaur HRB 6484

Vorstand: Henning Ahlert, Ralph Dommermuth, Matthias Ehrlich, Thomas Gottschlich, Matthias Greve, Robert Hoffmann, Markus Huhn, Oliver Mauss, Achim Weiss
Aufsichtsratsvorsitzender: Michael Scheeren

-------------------------------------------------------------------------
This SF.Net email is sponsored by the Moblin Your Move Developer's challenge
Build the coolest Linux based applications with Moblin SDK & win great prizes
Grand prize is a trip for two to an Open Source event anywhere in the world
http://moblin-contest.org/redirect.php?banner_id=100&url=/
_______________________________________________
qooxdoo-devel mailing list
qooxdoo-devel@...
https://lists.sourceforge.net/lists/listinfo/qooxdoo-devel

From forum: qooxdoo-devel

Re: Properties Window on clicking a node on a Tree

2008-10-10T01:07:52Z

Hi,

I am sorry for missing out the details (was pretty frustrated as I am unable to get what i want ;-) )

Hugh Gibson wrote:

I am facing a couple of problems in my Tree widget.


You don't say which version you are using, or which tree.

I am using v0.7.3 and testing on FF3 and IE7. And, I am referring to qx.ui.tree.Tree Widget.

1. *Context Menu on Node :*
I want to get a context menu for a node when it is clicked on. So, 
I have included the "click" event listener. But then, even if I 
click outside the tree (Not on a node but outside, I still get the 
context menu) which I don't want.


I'm using 0.7.x qx.ui.tree.Tree. To get a context menu I add a listener
on each node for the "contextmenu" event.

Thanks for the suggestion. I had tried it earlier but I guess I had missed on something. I will try again and let you know.

2. *Setting Current element to Null
*In my tree widget constructor, I call the setSelectedElement(null) 
and set the current element to null. And, this works because when I 
call alert(this.getSelectedElement()), I get the result as 
undefined.

I have also done the same inside a function called after my tree is 
created and after I have select some element in tree. But then it 
doesn't set the current element to null.


No idea on this one. Derrell?

Hugh

Here is the code snippet :

/* ------------ myTree class ------------------------*/
qx.Class.define("myTree", { extend: qx.ui.tree.Tree,

    //Constructor start
    construct : function(config)
          {
              var trs = qx.ui.tree.TreeRowStructure.getInstance().standard("Root");
             this.base(arguments, trs);

             this.setSelectedElement(null);
              alert(this.getSelectedElement());    //shows undefined which is correct
          },

    members: {

         onClickMenu: function(e){

        var contextMenu = new qx.ui.menu.Menu;
             var insertButton = new qx.ui.menu.Button("Insert",null,myCommand1);
             var deleteButton = new qx.ui.menu.Button("Delete",null,myCommand2);

        contextMenu.addToDocument();
         //contextMenu.show();

        if (contextMenu.isSeeable())        //code which never gets executed .. Not sure why
        {
        alert('hi');
        contextMenu.hide();
        this.setSelectedElement(null);
        alert(this.getSelectedElement);
        }
        else
        {
        var ele = this.getSelectedElement().getElement();
        contextMenu.setLeft(qx.html.Location.getPageBoxLeft(ele));
        contextMenu.setTop(qx.html.Location.getPageBoxBottom(ele));
        contextMenu.show();

          // alert('hi');
          // contextMenu.setDisplay(true);
        }

        this.setSelectedElement(null);
        alert(this.getSelectedElement());    //shows TreeFolder or TreeFile but not undefined as it should ... incorrect
        }

    }

});

I call this code from my other file as follows:
var tree = new myTree();
tree.addEventListener("click",tree.onClickMenu,tree);

Also, when I remove the line contextMenu.show(); from the else part and put below contextMenu.addToDocument(); (as commented now), I don't see the context menu on node click.
But when I add an alert and setDispaly() in the else part, I see the context Menu for a brief moment before the pop-up box comes.
Am i doing something wrong here ?

Thanks for the replies. I am happy for all the help. Hope I learn how to really work on Qooxdoo and understand it (little by little).

qooxdoo-devel mailing list
qooxdoo-devel@...



https://lists.sourceforge.net/lists/listinfo/qooxdoo-devel

Best regards 

Amit Rana
SpagoBI Developer

From forum: qooxdoo-devel

Re: javascript anomaly with regexp literal?

2008-10-10T00:12:00Z

> *Subject:* [qooxdoo-devel] javascript anomaly with regexp literal?
> *From:* "Derrell Lipman" <derrell.lipman@...>
> *To:* "qooxdoo Development" <qooxdoo-devel@...>
> *Date:* Thu, 9 Oct 2008 12:36:13 -0400
>
> This seems odd and I can't find any documentation stating why it's
> so...
>
> This application has a function which tests a string for containing
> only digits. It uses a regular expression literal assigned to a
> local variable in a function. That regular expression, however,
> appears to be persistent across calls to the function.

There's a simpler case which exhibits behaviour that is unexpected:

var re1 = /\?/g;
re1.lastIndex;
0

re1.test("a?");
true

re1.lastIndex;
2

re1.test("a?aaa");
false

re1.lastIndex;
0

re1.test("a?");
true

re1.lastIndex;
2

"a?aa".replace(re1, "b");
"abaa"

re1.lastIndex;
0

This was collated from various tests in Firebug console. So calling test()
using a regular expression changes lastIndex to after the match position
(or resets it to 0 if no match), but calling replace() resets it to 0
before and after the replace action.

So you have to be careful using test() quite apart from the fact that the
same regular expression object is used when a regex literal is declared.

Hugh

-------------------------------------------------------------------------
This SF.Net email is sponsored by the Moblin Your Move Developer's challenge
Build the coolest Linux based applications with Moblin SDK & win great prizes
Grand prize is a trip for two to an Open Source event anywhere in the world
http://moblin-contest.org/redirect.php?banner_id=100&url=/
_______________________________________________
qooxdoo-devel mailing list
qooxdoo-devel@...
https://lists.sourceforge.net/lists/listinfo/qooxdoo-devel

From forum: qooxdoo-devel

Re: javascript anomaly with regexp literal?

2008-10-09T23:24:00Z

> This seems odd and I can't find any documentation stating why it's
> so...

Check out http://blog.stevenlevithan.com/archives/es3-regexes-broken and
then https://bugzilla.mozilla.org/show_bug.cgi?id=98409

I'm not sure of the timeframe for a fix though as valid discussion in the
bug report finished a year ago.

Hugh

-------------------------------------------------------------------------
This SF.Net email is sponsored by the Moblin Your Move Developer's challenge
Build the coolest Linux based applications with Moblin SDK & win great prizes
Grand prize is a trip for two to an Open Source event anywhere in the world
http://moblin-contest.org/redirect.php?banner_id=100&url=/
_______________________________________________
qooxdoo-devel mailing list
qooxdoo-devel@...
https://lists.sourceforge.net/lists/listinfo/qooxdoo-devel

From forum: qooxdoo-devel

Re: javascript anomaly with regexp literal?

2008-10-09T19:32:43Z

Now I get what you mean (should have read your first mail better...).

Indeed appears like this regular expression is in a global context, thus
not being created anew while you call doTest.

Would you be surprised if I told you that the behaviour in Opera 9.52 is
the same as in FF? What about IE6... well, here it differs... Makes any
"12b13b" illegal!

You can get that same result in FF2 and Opera when you change the regex
instantiation to:

var illegalChars = new RegExp("[^0-9]", "g");

Now they all seem to do the same:-)

My interpretation is that
var illegalChars = /[^0-9]/g;

sees the regex /[^0-9]/g as a non-primitive type which just gets
referenced any time you call the doTest. Using the 'new RegExp' syntax
you'd instead force the creation of a new object which in turn caused
the expected behaviour. Well, just a wild guess... maybe you'll have to
check with the developers of the javascript engine.

Anyway I consider the deviating behaviour of the browsers dangerous.

> > PS: as a side effect of your example, I learned about a problem
I had
> > with this.debug() and was able to solve it by dealing with
> > qx.log.appender. I'm not sure if I missed something, but maybe the
> > documentation at
> > http://qooxdoo.org/documentation/0.8/debugging
> > is not complete regarding this.debug().

> Are you talking about the fact that the appenders are not included
> unless you reference them, or something else?

Yes, I didn't find that detail on the mentioned page. I can't remember
if I had to reference anything prior to using this.debug() in qx0.6.6 or
qx0.7.3, I believe that was all triggered by Makefile settings these
days. Nevertheless it works for now.

I'm curious if someone finds out details about these regular expression
objects...

Bye,
Stefan

Derrell Lipman wrote:

> On Thu, Oct 9, 2008 at 9:01 PM, Derrell Lipman
> <derrell.lipman@...
> <mailto:derrell.lipman@...>> wrote:
>
>
> Hi Stefan. What you describe is exactly the "problem" that I don't
> understand. With a regular expression literal, as with an object
> literal in 'var testobj={x:2}' or an array literal in 'var
> testarr=[23,42], If someplace in that method I say 'testobj.x=3;'
> the next time into that method, it'll reinitialize testobj to an
> object in which x has the value 2. Similarly, I would expect a new
> instantiation of a regular expression on each call into the method.
> It's not assigned to a static or global variable; it's a literal
> assigned to a local variable. Do you see any rationale for the
> regular expression retaining its state on subsequent calls to
> doTest()? Shouldn't we get a brand new, just-initialized regular
> expression object assigned to illegalChars on each call to doTest()?
>
>
> Here's a minor change to the application to demonstrate what I'm talking
> about. We can see that on each call, testobj.x and testarr[0] contain
> the values initialized by the object and array literals, respectively,
> but illegalChars.lastIndex is clearly referencing the original regular
> expressing object, not a newly-instantiated one...
>
> /*
> #require(qx.log.appender.Native)
> #require(qx.log.appender.Console)
> */
>
> qx.Class.define("custom.Application",
> {
> extend : qx.application.Standalone,
> members :
> {
> main: function()
> {
> this.base(arguments);
>
> this.doTest("123123", true);
> this.doTest("12b13b", true);
> this.doTest("12b13b", true);
> this.doTest("12b13b", true);
> this.doTest("12b13b", true);
>
> this.doTest("123123", false);
> this.doTest("12b13b", false);
> this.doTest("12b13b", false);
> this.doTest("12b13b", false);
> this.doTest("12b13b", false);
> this.doTest("12b13b", false);
> this.doTest("12b13b", false);
> this.doTest("12b13b", false);
> this.doTest("12b13b", false);
> this.doTest("12b13b", false);
> },
>
> doTest : function(s, reset)
> {
> var illegalChars = /[^0-9]/g;
> var testobj = { x : 2 };
> var testarr = [ 23, 42 ];
> var resetText = "without";
> this.debug("lastIndex=" + illegalChars.lastIndex + ", " +
> "testobj.x=" + testobj.x + ", " +
> "testarr[0]=" + testarr[0]);
>
> if (reset)
> {
> illegalChars.lastIndex = 0;
> resetText = "with";
> }
>
> var isIllegal = illegalChars.test(s);
> this.debug(s + " " + resetText +
> " reset: " + (isIllegal ? "illegal" : "legal"));
>
> testobj.x = 3;
> testarr[0] = 223;
> }
> }
> });
>
>
>
> ------------------------------------------------------------------------
>
> -------------------------------------------------------------------------
> This SF.Net email is sponsored by the Moblin Your Move Developer's challenge
> Build the coolest Linux based applications with Moblin SDK & win great prizes
> Grand prize is a trip for two to an Open Source event anywhere in the world
> http://moblin-contest.org/redirect.php?banner_id=100&url=/
>
>
> ------------------------------------------------------------------------
>
> _______________________________________________
> qooxdoo-devel mailing list
> qooxdoo-devel@...
> https://lists.sourceforge.net/lists/listinfo/qooxdoo-devel

From forum: qooxdoo-devel

Re: javascript anomaly with regexp literal?

2008-10-09T18:07:31Z

On Thu, Oct 9, 2008 at 9:01 PM, Derrell Lipman <derrell.lipman@...> wrote:

Hi Stefan. What you describe is exactly the "problem" that I don't understand. With a regular expression literal, as with an object literal in 'var testobj={x:2}' or an array literal in 'var testarr=[23,42], If someplace in that method I say 'testobj.x=3;' the next time into that method, it'll reinitialize testobj to an object in which x has the value 2. Similarly, I would expect a new instantiation of a regular expression on each call into the method. It's not assigned to a static or global variable; it's a literal assigned to a local variable. Do you see any rationale for the regular expression retaining its state on subsequent calls to doTest()? Shouldn't we get a brand new, just-initialized regular expression object assigned to illegalChars on each call to doTest()?

Here's a minor change to the application to demonstrate what I'm talking about. We can see that on each call, testobj.x and testarr[0] contain the values initialized by the object and array literals, respectively, but illegalChars.lastIndex is clearly referencing the original regular expressing object, not a newly-instantiated one...

/*
#require(qx.log.appender.Native)
#require(qx.log.appender.Console)
*/

qx.Class.define("custom.Application",
{
extend : qx.application.Standalone,
members :
{
    main: function()
    {
      this.base(arguments);

      this.doTest("123123", true);
      this.doTest("12b13b", true);
      this.doTest("12b13b", true);
      this.doTest("12b13b", true);
      this.doTest("12b13b", true);

      this.doTest("123123", false);
      this.doTest("12b13b", false);
      this.doTest("12b13b", false);
      this.doTest("12b13b", false);
      this.doTest("12b13b", false);
      this.doTest("12b13b", false);
      this.doTest("12b13b", false);
      this.doTest("12b13b", false);
      this.doTest("12b13b", false);
      this.doTest("12b13b", false);
    },

    doTest : function(s, reset)
    {
      var illegalChars = /[^0-9]/g;
      var testobj = { x : 2 };
      var testarr = [ 23, 42 ];
      var resetText = "without";
      this.debug("lastIndex=" + illegalChars.lastIndex + ", " +
                 "testobj.x=" + testobj.x + ", " +
                 "testarr[0]=" + testarr[0]);

      if (reset)
      {
        illegalChars.lastIndex = 0;
        resetText = "with";
      }

      var isIllegal = illegalChars.test(s);
      this.debug(s + " " + resetText +
                 " reset: " + (isIllegal ? "illegal" : "legal"));

      testobj.x = 3;
      testarr[0] = 223;
    }
}
});

From forum: qooxdoo-devel

Re: javascript anomaly with regexp literal?

2008-10-09T18:01:49Z

On Thu, Oct 9, 2008 at 8:05 PM, Stefan Volbers <volbers@...> wrote:

The missing of the illegalChars.lastIndex = 0; operation in the second
part of the doTest() series makes the illegal.test(s) iterate through
the string in question and, with the '123123' string, positively fails
to find an illegal char, resulting in a 'not found' output.
The following illegalChars.test('12b13b') evaluates correctly as true.
But now the lastIndex of the regex is not set to 0 and so the index
remains after the '12b' part of the string. The next function call
evaluates the '13b' of it and finds the illegal letter, too. No index
reset following, the rest of the test string is empty and finds
_nothing_ left, so no illegal letter is found, which means that the
test() fails (finally evaluating to the output of 'not found'). This
failure implies that the index is automatically reset, and so the test
will continually find twice and fail once after.

Hi Stefan. What you describe is exactly the "problem" that I don't understand. With a regular expression literal, as with an object literal in 'var testobj={x:2}' or an array literal in 'var testarr=[23,42], If someplace in that method I say 'testobj.x=3;' the next time into that method, it'll reinitialize testobj to an object in which x has the value 2. Similarly, I would expect a new instantiation of a regular expression on each call into the method. It's not assigned to a static or global variable; it's a literal assigned to a local variable. Do you see any rationale for the regular expression retaining its state on subsequent calls to doTest()? Shouldn't we get a brand new, just-initialized regular expression object assigned to illegalChars on each call to doTest()?

PS: as a side effect of your example, I learned about a problem I had
with this.debug() and was able to solve it by dealing with
qx.log.appender. I'm not sure if I missed something, but maybe the
documentation at
http://qooxdoo.org/documentation/0.8/debugging
is not complete regarding this.debug().

Are you talking about the fact that the appenders are not included unless you reference them, or something else?

Thanks,

Derrell

From forum: qooxdoo-devel