Passing named parameter without the URL

[Alexander Back]

Hi Kenny,

Kenny Tilton wrote:
> Looks like we'll be adopting qooxdoo. First impressions are really over
> the top. Anywho...
Nice to hear that :-)

> Now that I am getting serious about qooxdoo I want to start getting my
> code base organized. One easy way to divide it up comes from the app
> itself being eally multiple browsers in one, with a tabview just under
> the page banner, one tab per browser.
First of all I must admit I do not get your point about creating tabs
for the different browsers since qooxdoo looks the same in every
browser. This is one of the big advantages: write it once and run it on
every major browser.

> In my code, I have this:
>
> var tabView = new qx.ui.tabview.TabView();
> var tab = new qx.ui.tabview.Page("XYZ Browser");
> tab.setLayout(new qx.ui.layout.Grow());
> tab.add(this.buildXYZ());
> tabView.add(tab);
>
> Now I guess what I do is subclass....tabView.Page?! Sounds wrong.
Exactly. You do not want to extend/change the tabView.Page widget
itself. All you want to do is to add content to this widget. So there is
no need the subclass the tabView.Page class since it is acting only as a
container for your widgets.

> What would you do? I could just make up a new class nova.xyz (where nova
> is my top-level application) and then:
>
> var tabView = new qx.ui.tabview.TabView();
> var xyz = new nova.xyz;
> var tab = new qx.ui.tabview.Page(xyz.view.title);
> tab.setLayout(new qx.ui.layout.Grow());
> tab.add(xyz.view);
> tabView.add(tab);
>
> That way the view does not know about being in a tabView.

You can use the method "getLayoutParent" to determine which parent your
view has. But I think it's more interesting with which layout the view
is rendered. As long you use the "Grow" layout (->
http://qooxdoo.org/documentation/0.8/layout/Grow ) the view will always
take as much space as it needs and the widgets of the view look the same
in a different container using the same layout.

> I'm not really asking about OO design -- I am pretty comfortable with
> that -- I guess I am just wondering if my usual thinking needs
> adjustment under the qooxdoo model.
I guess you're heading the right way :-)

cheers,
   Alex

-------------------------------------------------------------------------
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

[Kenny Tilton-2]

Alexander Back wrote:
> Hi Kenny,
>
> just for completion (and for all other users out there using the Remote
> Table Model): I wrote a basic tutorial on how to implement it (client-
> and server-side). And of course, Derrell's demos is mentioned as well.
>
> -> http://qooxdoo.org/documentation/0.8/remote_table_model

Ah, that was you! Thx, that write-up is what got me going. I owe you a beer.

>
> Any feedback welcome ;-)

In here I think you want to subsitute the values firstRow and lastRow
into the URL instead of hardcoding 0/100:

  _loadRowData : function(firstRow, lastRow)
     {
        // Call the backend service (example) - using XmlHttp
        var url  =
"http://localhost/services/getTableRowData.php?from=0&to=100";
        var req = qx.io.remote.Request(url, "GET", "application/json");

        // Add listener
        req.addListener("completed", this._onLoadRowDataCompleted,
this);

        // send request
        req.send();
     },

Also, I am not sure how to factor this into the tutorial since it also
involves the table (the view, not the model) but to get column sorts
working I had to have the remote model watching the metadata (where
column sorts get published):

                        m.addListener(qx.ui.table.ITableModel.EVENT_TYPE_META_DATA_CHANGED
    , function (e) {
        var order;
        if (this.__sortAscending)
                order="asc";
        else
                order = "dsc";
                                       
        var req = new qx.io.remote.Request("/supplierssort?ns=asp"
                 + "&key=" + this.getColumnId(this.__sortColumnIndex)
                + "&order=" + order
                , "GET", "application/json"); // app/json????
         req.addListener("completed"
                , function(e){
                      m.reloadData();
                 });
         req.send();
       });
                               
Regarding that app/json???: I am still learning this stuff and have not
figured out how to handle the case where I just want the server app to
do something for its side effect, such as resort the data, and I have no
intention of doing anything with the response. someone on c.l.javascript
suggested a POST.

btw, not sure this is good for pedagogy, but I think the code gets a
little simpler when an anonymous function is provided to listeners.

Thx again.

cheers, kenny

>
> cheers,
>    Alex
>
> Kenny Tilton wrote:
>
>>Woohoo, getting close: I have a subclass of the remote table model at
>>least getting its rowcount back from Lisp and being reported accurately
>>by the table view, but that's it no rows. And I do not see Lisp being
>>hit for the rowdata URL, so I guess no one is asking.
>>
>>The example:
>>
>>
>>http://demo.qooxdoo.org/current/demobrowser/#table~Table_Remote_Model.html
>>
>>...seems unfinished. It uses a Simple model and generates random data.
>>Is there a better example anywhere? I searched the entire qx source tree
>>for qx.ui.table.model.Remote and just found its definition.
>>
>>I'll start digging around in the source of Table now to see what makes
>>it ask for data, but any hints are welcome.
>>
>>kt
>
>
> -------------------------------------------------------------------------
> 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
>

--
http://www.theoryyalgebra.com/

-------------------------------------------------------------------------
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

[Alexander Back]

Hi Kenny,

Kenny Tilton wrote:

> Alexander Back wrote:
>> Hi Kenny,
>>
>> just for completion (and for all other users out there using the Remote
>> Table Model): I wrote a basic tutorial on how to implement it (client-
>> and server-side). And of course, Derrell's demos is mentioned as well.
>>
>> -> http://qooxdoo.org/documentation/0.8/remote_table_model
>
> Ah, that was you! Thx, that write-up is what got me going. I owe you a beer.

I'll forward to it ;-)

>> Any feedback welcome ;-)
>
> In here I think you want to subsitute the values firstRow and lastRow
> into the URL instead of hardcoding 0/100:
You're absolutely right. I've just corrected this.

> Also, I am not sure how to factor this into the tutorial since it also
> involves the table (the view, not the model) but to get column sorts
> working I had to have the remote model watching the metadata (where
> column sorts get published):
>
> m.addListener(qx.ui.table.ITableModel.EVENT_TYPE_META_DATA_CHANGED
>     , function (e) {
> var order;
> if (this.__sortAscending)
> order="asc";
> else
> order = "dsc";
>
> var req = new qx.io.remote.Request("/supplierssort?ns=asp"
>                  + "&key=" + this.getColumnId(this.__sortColumnIndex)
> + "&order=" + order
> , "GET", "application/json"); // app/json????
>          req.addListener("completed"
> , function(e){
>                       m.reloadData();
>                  });
>          req.send();
>        });

Again right. I aimed a basic introduction into the Remote Table Model
stuff with this tutorial. So it's far from covering all the additional
work to get all features up and running. I guess implementing such
features should reside in an own wiki article and to keep the tutorial
small and basic.
Maybe you can add an article to the wiki? Documentation and tutorials
are always welcome :-)

> Regarding that app/json???: I am still learning this stuff and have not
> figured out how to handle the case where I just want the server app to
> do something for its side effect, such as resort the data, and I have no
> intention of doing anything with the response. someone on c.l.javascript
> suggested a POST.
Yepp, a POST request would fit better here. Regarding the
"application/json" mimetype just take a look at the http://www.json.org 
page, there you get all your answers.
Quote from json.org:
"JSON (JavaScript Object Notation) is a lightweight data-interchange
format. It is easy for humans to read and write. It is easy for machines
to parse and generate. It is based on a subset of the JavaScript
Programming Language, Standard ECMA-262 3rd Edition - December 1999.
JSON is a text format that is completely language independent but uses
conventions that are familiar to programmers of the C-family of
languages, including C, C++, C#, Java, JavaScript, Perl, Python, and
many others. These properties make JSON an ideal data-interchange language."

> btw, not sure this is good for pedagogy, but I think the code gets a
> little simpler when an anonymous function is provided to listeners.
Yes maybe simpler, but avoiding inline (=anonymous) functions is intended.
-> http://qooxdoo.org/documentation/0.8/antipatterns#inline_functions

cheers,
   Alex

>> Kenny Tilton wrote:
>>
>>> Woohoo, getting close: I have a subclass of the remote table model at
>>> least getting its rowcount back from Lisp and being reported accurately
>>> by the table view, but that's it no rows. And I do not see Lisp being
>>> hit for the rowdata URL, so I guess no one is asking.
>>>
>>> The example:
>>>
>>>
>>> http://demo.qooxdoo.org/current/demobrowser/#table~Table_Remote_Model.html
>>>
>>> ...seems unfinished. It uses a Simple model and generates random data.
>>> Is there a better example anywhere? I searched the entire qx source tree
>>> for qx.ui.table.model.Remote and just found its definition.
>>>
>>> I'll start digging around in the source of Table now to see what makes
>>> it ask for data, but any hints are welcome.
>>>
>>> kt
>>
>> -------------------------------------------------------------------------
>> 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
>>
>
>

--
Alexander Back
Core Development::Webtechnologies
alexander.back@...
fon [phone]
http://www.1und1.de

1&1 Internet AG
Ernst-Frey-Straße 9
76135 Karlsruhe

Vorstände: Henning Ahlert, Ralph Dommermuth, Matthias Ehrlich, Thomas
Gottschlich, Robert Hoffmann, Markus Huhn, Hans-Henning Kettler, Dr.
Oliver Mauss, Jan Oetjen
Aufsichtsratsvorsitzender: Michael Scheeren
Amtsgericht Montabaur / HRB 6484

-------------------------------------------------------------------------
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

[Kenny Tilton-2]

Alexander Back wrote:

> Hi Kenny,
>
> Kenny Tilton wrote:
>
>>Looks like we'll be adopting qooxdoo. First impressions are really over
>>the top. Anywho...
>
> Nice to hear that :-)
>
>
>>Now that I am getting serious about qooxdoo I want to start getting my
>>code base organized. One easy way to divide it up comes from the app
>>itself being eally multiple browsers in one, with a tabview just under
>>the page banner, one tab per browser.
>
> First of all I must admit I do not get your point about creating tabs
> for the different browsers since qooxdoo looks the same in every
> browser. This is one of the big advantages: write it once and run it on
> every major browser.

Oops, I used the wrong word, but that is because I think of the overall
application of consisting of different, um, "explorer" modules for
different aspects of the clients business.

>
>
>>In my code, I have this:
>>
>> var tabView = new qx.ui.tabview.TabView();
>> var tab = new qx.ui.tabview.Page("XYZ Browser");
>> tab.setLayout(new qx.ui.layout.Grow());
>> tab.add(this.buildXYZ());
>> tabView.add(tab);
>>
>>Now I guess what I do is subclass....tabView.Page?! Sounds wrong.
>
> Exactly. You do not want to extend/change the tabView.Page widget
> itself. All you want to do is to add content to this widget. So there is
> no need the subclass the tabView.Page class since it is acting only as a
> container for your widgets.
>
>
>>What would you do? I could just make up a new class nova.xyz (where nova
>>is my top-level application) and then:
>>
>> var tabView = new qx.ui.tabview.TabView();
>> var xyz = new nova.xyz;
>> var tab = new qx.ui.tabview.Page(xyz.view.title);
>> tab.setLayout(new qx.ui.layout.Grow());
>> tab.add(xyz.view);
>> tabView.add(tab);
>>
>>That way the view does not know about being in a tabView.
>
> You can use the method "getLayoutParent" to determine which parent your
> view has. But I think it's more interesting with which layout the view
> is rendered. As long you use the "Grow" layout (->
> http://qooxdoo.org/documentation/0.8/layout/Grow ) the view will always
> take as much space as it needs and the widgets of the view look the same
> in a different container using the same layout.

Ah, yes, it took a day of experimentation but I finally got to where I
can get pretty much what I want with qooxdoo's layout capabilities.
Really excellent, those. I was amazed at how well the table widget
(allowed to flex in a vbox, IIRC) responded to more/less real estate
being available.

But this reminds me...does the table widget need similar elegance for
the handling of column widths? I have a simple two-column table where I
would like one column with a fixed size and the other column to get the
rest. And of course there would be other needs. Maybe something like the
"flex" logic for column widths?

Jes thinkin out loud. :)

cheers, kenny

>
>
>>I'm not really asking about OO design -- I am pretty comfortable with
>>that -- I guess I am just wondering if my usual thinking needs
>>adjustment under the qooxdoo model.
>
> I guess you're heading the right way :-)
>
> cheers,
>    Alex
>
> -------------------------------------------------------------------------
> 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
>

--
http://www.theoryyalgebra.com/

-------------------------------------------------------------------------
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

[Derrell Lipman]

On Thu, Nov 20, 2008 at 8:38 AM, Kenny Tilton <kennytilton@...> wrote:

But this reminds me...does the table widget need similar elegance for
the handling of column widths? I have a simple two-column table where I
would like one column with a fixed size and the other column to get the
rest. And of course there would be other needs. Maybe something like the
"flex" logic for column widths?

Jes thinkin out loud. :)

Hi Kenny, you're right!  That's needed functionality!  And we have it. :-)

You're looking for Table's Resize column model, qx.ui.table.columnmodel.Resize.  There's a demo in demobrowser table/Table_Resize_Columns) that shows how to use it.

Cheers,

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

[Kenny Tilton-2]

Derrell Lipman wrote:

> On Thu, Nov 20, 2008 at 8:38 AM, Kenny Tilton <kennytilton@...
> <mailto:kennytilton@...>> wrote:
>
>
>     But this reminds me...does the table widget need similar elegance for
>     the handling of column widths? I have a simple two-column table where I
>     would like one column with a fixed size and the other column to get the
>     rest. And of course there would be other needs. Maybe something like the
>     "flex" logic for column widths?
>
>     Jes thinkin out loud. :)
>
>
> Hi Kenny, you're right!  That's needed functionality!  And we have it. :-)
>
> You're looking for Table's Resize column model,
> qx.ui.table.columnmodel.Resize.  There's a demo in demobrowser
> table/Table_Resize_Columns) that shows how to use it.

Ah, I kept looking at that example wondering what the point was. :)

Good timing, I just broke out my first tab as a distinct class and when
it Just Worked I noticed the table had a horizontal scrollbar because I
was off by a couple of pixels somewhere.

Thx,

kenny

--
http://www.theoryyalgebra.com/

-------------------------------------------------------------------------
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

[Kenny Tilton-2]

Alexander Back wrote:

> Hi Kenny,
>
> Kenny Tilton wrote:
>
>>Alexander Back wrote:
>>
>>>Hi Kenny,
>>>
>>>just for completion (and for all other users out there using the Remote
>>>Table Model): I wrote a basic tutorial on how to implement it (client-
>>>and server-side). And of course, Derrell's demos is mentioned as well.
>>>
>>>-> http://qooxdoo.org/documentation/0.8/remote_table_model
>>
>>Ah, that was you! Thx, that write-up is what got me going. I owe you a beer.
>
> I'll forward to it ;-)
>
>
>>>Any feedback welcome ;-)
>>
>>In here I think you want to subsitute the values firstRow and lastRow
>>into the URL instead of hardcoding 0/100:
>
> You're absolutely right. I've just corrected this.
>
>
>>Also, I am not sure how to factor this into the tutorial since it also
>>involves the table (the view, not the model) but to get column sorts
>>working I had to have the remote model watching the metadata (where
>>column sorts get published):
>>
>> m.addListener(qx.ui.table.ITableModel.EVENT_TYPE_META_DATA_CHANGED
>>    , function (e) {
>> var order;
>> if (this.__sortAscending)
>> order="asc";
>> else
>> order = "dsc";
>>
>> var req = new qx.io.remote.Request("/supplierssort?ns=asp"
>>                 + "&key=" + this.getColumnId(this.__sortColumnIndex)
>> + "&order=" + order
>> , "GET", "application/json"); // app/json????
>>         req.addListener("completed"
>> , function(e){
>>                     m.reloadData();
>>                 });
>>         req.send();
>>       });
>
> Again right. I aimed a basic introduction into the Remote Table Model
> stuff with this tutorial. So it's far from covering all the additional
> work to get all features up and running. I guess implementing such
> features should reside in an own wiki article and to keep the tutorial
> small and basic.
> Maybe you can add an article to the wiki? Documentation and tutorials
> are always welcome :-)

Yes, I might do that once I am sure I know what I am doing. First I have
to find out if I am being riffed because I just killed three weeks
assessing JS frameworks, if so I'll have more time on my hands. But I
want to blog about qooxdoo anyway, maybe I'll document my grid work as a
concrete example.

>
>
>>Regarding that app/json???: I am still learning this stuff and have not
>>figured out how to handle the case where I just want the server app to
>>do something for its side effect, such as resort the data, and I have no
>>intention of doing anything with the response. someone on c.l.javascript
>>suggested a POST.
>
> Yepp, a POST request would fit better here. Regarding the
> "application/json" mimetype just take a look at the http://www.json.org 
> page, there you get all your answers.
> Quote from json.org:
> "JSON (JavaScript Object Notation) is a lightweight data-interchange
> format. It is easy for humans to read and write. It is easy for machines
> to parse and generate. It is based on a subset of the JavaScript
> Programming Language, Standard ECMA-262 3rd Edition - December 1999.
> JSON is a text format that is completely language independent but uses
> conventions that are familiar to programmers of the C-family of
> languages, including C, C++, C#, Java, JavaScript, Perl, Python, and
> many others. These properties make JSON an ideal data-interchange language."
>
>
>>btw, not sure this is good for pedagogy, but I think the code gets a
>>little simpler when an anonymous function is provided to listeners.
>
> Yes maybe simpler, but avoiding inline (=anonymous) functions is intended.
> -> http://qooxdoo.org/documentation/0.8/antipatterns#inline_functions

Ummm..er..OK, that is just plain wrong! :)

I am a Lisper so we have a different take on anonymous functions. Ruby
and Smalltalk people would tell you the same. OTOH, I know GvR is firmly
on your side--he would have taken lambda out of Python recently but it
was too late.

Even before I ran into Lisp I alawys hated it in C when I had to create
a standalone function just so I could pass a pointer to it to sort or
one of my own functions that took another function as a parameter. This
gets to the points in that link about not getting separate doc
generated--it should not be, the handler is part of the request sender's
functionality. Likewise populating the function space with a name that
will not get used anywhere else is misleading to the code reader -- an
anon func makes it clear that this bit of code serves only this purpose.

Anyway, I have a resizer table to code up. :)

cheers, kenny


--
http://www.theoryyalgebra.com/

-------------------------------------------------------------------------
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

[thron7]

>> Yes maybe simpler, but avoiding inline (=anonymous) functions is intended.
>> -> http://qooxdoo.org/documentation/0.8/antipatterns#inline_functions
>>    
>
> Ummm..er..OK, that is just plain wrong! :)
>  

Yes ... let's start a flame :).

> I am a Lisper so we have a different take on anonymous functions. Ruby
> and Smalltalk people would tell you the same. OTOH, I know GvR is firmly
> on your side--he would have taken lambda out of Python recently but it
> was too late.
>
> Even before I ran into Lisp I alawys hated it in C when I had to create
> a standalone function just so I could pass a pointer to it to sort or
> one of my own functions that took another function as a parameter. This
> gets to the points in that link about not getting separate doc
> generated--it should not be, the handler is part of the request sender's
> functionality. Likewise populating the function space with a name that
> will not get used anywhere else is misleading to the code reader -- an
> anon func makes it clear that this bit of code serves only this purpose.
>
>  

I tend to agree. I think there are places where anonymous functions are
the better choice for exactly the reasons they are wrong in other
places. From the arguments given on the Web page against anonymous
functions,

   1.
      they are harder to debug
   2.
      API Viewer will not list them, therefore you can not comment them
      correctly
   3.
      other programmers will not find them at first glance
   4.
      they might not appear in an IDE’s outline view

2. and 4. might be just what you want (yes, there is merrit in
information hiding, which any OO person should agree with). 3. is
fallacious since the closure appears in exactly the spot where it is used.

What I don't like with them is when they disrupt the enclosing function
call (which you can amend with proper spacing), and spill the flow of
the "main" application code. A real argument for me is 1., but hey, in
qooxdoo we generate dynamic functions all over the place :).

In the end, it's a lot about taste and what suits a programmer's brain.
(Meaning: If you get into the habit of thinking in closures, you easily
get accustomed to them).

Just my 2ct.

Thomas



-------------------------------------------------------------------------
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

[Kenny Tilton-2]

thron7 wrote:

>
>
>
>>> Yes maybe simpler, but avoiding inline (=anonymous) functions is
>>> intended.
>>> -> http://qooxdoo.org/documentation/0.8/antipatterns#inline_functions
>>>    
>>
>>
>> Ummm..er..OK, that is just plain wrong! :)
>>  
>
>
> Yes ... let's start a flame :).

<g> I could not contain myself.

>
>> I am a Lisper so we have a different take on anonymous functions. Ruby
>> and Smalltalk people would tell you the same. OTOH, I know GvR is
>> firmly on your side--he would have taken lambda out of Python recently
>> but it was too late.
>>
>> Even before I ran into Lisp I alawys hated it in C when I had to
>> create a standalone function just so I could pass a pointer to it to
>> sort or one of my own functions that took another function as a
>> parameter. This gets to the points in that link about not getting
>> separate doc generated--it should not be, the handler is part of the
>> request sender's functionality. Likewise populating the function space
>> with a name that will not get used anywhere else is misleading to the
>> code reader -- an anon func makes it clear that this bit of code
>> serves only this purpose.
>>
>>  
>
>
> I tend to agree. I think there are places where anonymous functions are
> the better choice for exactly the reasons they are wrong in other
> places. From the arguments given on the Web page against anonymous
> functions,
>
>   1.
>      they are harder to debug
>   2.
>      API Viewer will not list them, therefore you can not comment them
>      correctly
>   3.
>      other programmers will not find them at first glance
>   4.
>      they might not appear in an IDE’s outline view
>
> 2. and 4. might be just what you want (yes, there is merrit in
> information hiding, which any OO person should agree with). 3. is
> fallacious since the closure appears in exactly the spot where it is used.

Yeah, I was wondering about that one. :)

>
> What I don't like with them is when they disrupt the enclosing function
> call (which you can amend with proper spacing), and spill the flow of
> the "main" application code.

It is a nasty problem. Lisp editors have nice "conserve indentation"
modes one can turn on if like most of us we hate seeing code sail off to
the right. My solution has been a compromise on the indentation
releative to the function keyowrd:

                        this.supplierTypeChoice.addListener("changeValue", function(e){
        this.debug("ChangeValue: " + e.getData());
        var req = new qx.io.remote.Request("/suppliertypeset?ns=asp" +
                        "&type=" + e.getData()
                        , "GET", "application/json");
        req.addListener("completed", function(e){
                this.debug("typeset completed ");
                m.reloadData();
        });
        req.send();
});

I think elsewhere I have:

this.supplierTypeChoice.addListener("changeValue"
        , function(e){
                this.debug("ChangeValue: " + e.getData());
                var req = new qx.io.remote.Request("/suppyset?ns=asp"
                                 + "&type=" + e.getData()
                                , "GET", "application/json");
                req.addListener("completed"
                  , function(e){
                        this.debug("typeset completed ");
                        m.reloadData();
                });
                req.send();
        });

What is needed, btw, is a four-character tab.

> A real argument for me is 1.,

I debug with print statements anyway, but it can be an issue seeing a
lamba in the call stack on a backtrace.

> but hey, in
> qooxdoo we generate dynamic functions all over the place :).
>
> In the end, it's a lot about taste and what suits a programmer's brain.
> (Meaning: If you get into the habit of thinking in closures, you easily
> get accustomed to them).

Closures rock for certain.

cheers, ken


-------------------------------------------------------------------------
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

[Jean-Baptiste BRIAUD -- Novlog]

Hi Derrell,

I managed to build my "qooxdoo-all-in-one-file.js" from SVN HEAD, thanks to Petr for his help.

So, the result :

1. The log.

request that contained form parameters continue not to be logged in firebug.

This happen when I use setFormParameter method to the request.

2. The parameters

using the new last param :

using false : it is well encoded in the URL, seen by the backend as form value.

using true : can't see the param in the URL but the backend doesn't recognize it as a nammed parameter and in fact, doesn't recognize it at all.

It might be the way it is encoded in the data is not the equivalent as "nammed parameter passed in the request body".

I have to admit I never goes under the API level of HTTP request and I don't know how all that encoding works.

Let me ask a stupid question : are HTTP request nammed parameter different than form field value ?

If yes, how can it be distinguished is both are encoded in the URL ?

If no, why do we have different method in API (not only in qooxdoo) for setting parameter and form field ?

On 19 Nov 2008, at 16:02, Derrell Lipman wrote:

On Tue, Nov 18, 2008 at 8:31 AM, Jean-Baptiste BRIAUD -- Novlog <j-b.briaud@...> wrote:

On 18 Nov 2008, at 14:23, Derrell Lipman wrote:

On Tue, Nov 18, 2008 at 8:14 AM, thron7 <thron7@...> wrote:

>> Using setParameter() specifically tells it to encode the parameters in the
>> URL.  I believe you're looking for setFormField().
>>

> If you have form fields, then a different transport will be used.

This is true, but I strongly object against the XmlHttp transport being
restricted to GET and URL parameters. If this is really the case,
somebody has to open up a bug for it. XmlHttp transport has to support
POST and setParameters() working on the request body!

XmlHttp transport _does_ support POST.  That's not the issue.  The issue which has been discussed in depth in the past is that even with POST, it must be possible to add parameters to the URL.  You (and the reset of those in this discussion) are right that there *should* be a way to add parameters to the POST request body.  There's just no way to explicitly specify that currently.   This is just one of many issues that have been addressed in the past, that should be addressed in a rewrite of qx.io.remote.*

I didn't know it was discussed before, sorry ...

So, I do agree with you : parameter to the POST request body is a need.

Jean-Baptiste, would you please test the attached patch (or pull from SVN; I've checked it in).  Note the new parameter available to setParameter() which specifies whether you want the parameter to be in the URL or as data.  This should do exactly what you're looking for.  Let me know whether this works properly for you.

Cheers,

Derrell

<y.patch>-------------------------------------------------------------------------
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

[Petr Kobalíček]

Hi Jean,

The logging problems are probably caused by qxbuild, I'm expecting
something similar and working to fix it. In my case, if I turn debug
off, then I see logging and vice versa (probably trivial error in
config).

Cheers
- Petr

2008/11/21 Jean-Baptiste BRIAUD -- Novlog <j-b.briaud@...>:

> Hi Derrell,
> I managed to build my "qooxdoo-all-in-one-file.js" from SVN HEAD, thanks to
> Petr for his help.
> So, the result :
> 1. The log.
> request that contained form parameters continue not to be logged in firebug.
> This happen when I use setFormParameter method to the request.
> 2. The parameters
> using the new last param :
> using false : it is well encoded in the URL, seen by the backend as form
> value.
> using true : can't see the param in the URL but the backend doesn't
> recognize it as a nammed parameter and in fact, doesn't recognize it at all.
> It might be the way it is encoded in the data is not the equivalent as
> "nammed parameter passed in the request body".
> I have to admit I never goes under the API level of HTTP request and I don't
> know how all that encoding works.
> Let me ask a stupid question : are HTTP request nammed parameter different
> than form field value ?
> If yes, how can it be distinguished is both are encoded in the URL ?
> If no, why do we have different method in API (not only in qooxdoo) for
> setting parameter and form field ?
>
> On 19 Nov 2008, at 16:02, Derrell Lipman wrote:
>
> On Tue, Nov 18, 2008 at 8:31 AM, Jean-Baptiste BRIAUD -- Novlog
> <j-b.briaud@...> wrote:
>>
>> On 18 Nov 2008, at 14:23, Derrell Lipman wrote:
>>
>> On Tue, Nov 18, 2008 at 8:14 AM, thron7 <thron7@...>
>> wrote:
>>>
>>> >> Using setParameter() specifically tells it to encode the parameters in
>>> >> the
>>> >> URL.  I believe you're looking for setFormField().
>>> >>
>>> > If you have form fields, then a different transport will be used.
>>>
>>> This is true, but I strongly object against the XmlHttp transport being
>>> restricted to GET and URL parameters. If this is really the case,
>>> somebody has to open up a bug for it. XmlHttp transport has to support
>>> POST and setParameters() working on the request body!
>>
>> XmlHttp transport _does_ support POST.  That's not the issue.  The issue
>> which has been discussed in depth in the past is that even with POST, it
>> must be possible to add parameters to the URL.  You (and the reset of those
>> in this discussion) are right that there *should* be a way to add parameters
>> to the POST request body.  There's just no way to explicitly specify that
>> currently.   This is just one of many issues that have been addressed in the
>> past, that should be addressed in a rewrite of qx.io.remote.*
>>
>> I didn't know it was discussed before, sorry ...
>> So, I do agree with you : parameter to the POST request body is a need.
>
> Jean-Baptiste, would you please test the attached patch (or pull from SVN;
> I've checked it in).  Note the new parameter available to setParameter()
> which specifies whether you want the parameter to be in the URL or as data.
> This should do exactly what you're looking for.  Let me know whether this
> works properly for you.
>
> Cheers,
>
> Derrell
>
> <y.patch>-------------------------------------------------------------------------
> 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

[Jean-Baptiste BRIAUD -- Novlog]

On 21 Nov 2008, at 15:41, Petr Kobalíček wrote:

> Hi Jean,
>
> The logging problems are probably caused by qxbuild, I'm expecting
> something similar and working to fix it. In my case, if I turn debug
> off, then I see logging and vice versa (probably trivial error in
> config).
>
How do you see the difference in firebug between log and debug ?
It all ends with line in the console, isn't it ?

Also, not all request fall in that case. Only those with formParameter  
are not in firebug.
All the others are in firebug.

What I mean by "be in firebug" is that special line that can be  
unfolded and that allow to check the server response.

> Cheers
> - Petr
>
> 2008/11/21 Jean-Baptiste BRIAUD -- Novlog <j-b.briaud@...>:
>> Hi Derrell,
>> I managed to build my "qooxdoo-all-in-one-file.js" from SVN HEAD,  
>> thanks to
>> Petr for his help.
>> So, the result :
>> 1. The log.
>> request that contained form parameters continue not to be logged in  
>> firebug.
>> This happen when I use setFormParameter method to the request.
>> [CUT]

-------------------------------------------------------------------------
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

[thron7]

> How do you see the difference in firebug between log and debug ?
> It all ends with line in the console, isn't it ?
>
> Also, not all request fall in that case. Only those with formParameter  
> are not in firebug.
> All the others are in firebug.
>
> What I mean by "be in firebug" is that special line that can be  
> unfolded and that allow to check the server response.
>
>  

Have you ever looked into the "Net" tab of firebug?

-------------------------------------------------------------------------
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

[Jean-Baptiste BRIAUD -- Novlog]

On 21 Nov 2008, at 16:35, thron7 wrote:

>
>
>
>> How do you see the difference in firebug between log and debug ?
>> It all ends with line in the console, isn't it ?
>>
>> Also, not all request fall in that case. Only those with  
>> formParameter
>> are not in firebug.
>> All the others are in firebug.
>>
>> What I mean by "be in firebug" is that special line that can be
>> unfolded and that allow to check the server response.
>>
>>
>
> Have you ever looked into the "Net" tab of firebug?

No :-o
I was only talking about Console.
Amazingly, in the console, I can't see request witch have request  
parameter.

Now you can be sure I'll use the net tab, it's really cool.
I'm just not sure about the last column. The longer the blue bar are,  
the shorter the time is ?!
Also, what are the differences between gay and blue bar ?

It not clearly documented here but maybe there are other good places :
http://www.evotech.net/blog/2007/06/introduction-to-firebug/#fb_speed


-------------------------------------------------------------------------
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

[Kenny Tilton-2]

Alexander Back wrote:
> Hi Kenny,
>
> just for completion (and for all other users out there using the Remote
> Table Model): I wrote a basic tutorial on how to implement it (client-
> and server-side). And of course, Derrell's demos is mentioned as well.
>
> -> http://qooxdoo.org/documentation/0.8/remote_table_model

Thx agaon for that lifesaver. I have generalized it(and probably cocked
it up), code trailing below.

Not sure about the bit that avoids repeatedly asking the server for the
count (I saw about twenty requests from the model before I could get an
answer back).

Instances are created with two URLs, one to get the count, one the data.

Sample usage:

  new nuvia.data.TablePagerJSON("/getdatacount?store=netitems&ns=asp"
                              , "/getitems?ns=asp", 10);

I know this is a wreck, it is more a suggestion for ways to enhance the
RemoteModel.

cheers, kenny

--------------- Code starts here ----------------------------

qx.Class.define("nuvia.data.TablePagerJSON", {
     extend: qx.ui.table.model.Remote,
     construct: function(countURL, rowURL, blocksize){
         this.base(arguments);
                this.debug("making new pager "+countURL);
         this.countU = countURL;
         this.rowU = rowURL;
                this.setBlockSize(blocksize);
     },
     members: {
         countU: null,
         rowU: null,
                countPending : false,
         _loadRowCount: function(){
                        self = this;
                       
                        if (!this.countPending) {
                                var req = new qx.io.remote.Request(this.countU, "GET",
"application/json");
                                this.debug("asking for count ");
                                req.addListener("completed", function(response){
                                        var result = response.getContent();
                                        self.countPending = false;
                                        if (result != null) {
                                                this.debug("got count! "+ result);
                                                this._onRowCountLoaded(result);
                                        }
                                }, this);
                                this.countPending = true;
                                req.send();
                        }
         },
         _loadRowData: function(firstRow, lastRow){
             var fullU = this.rowU;
             if (fullU.indexOf("?", 0) > -1)
                 fullU = fullU + "&";
             else
                 fullU = fullU + "?";
             fullU = fullU + "from=" + firstRow + "&to=" + lastRow;
             this.debug("JSON pager hits " + fullU);
             var req = new qx.io.remote.Request(fullU, "GET",
"application/json");
             req.addListener("completed", function(response){
                 var result = response.getContent();
                 //this.debug("json pager gets " + result);
                 if (result != null)
                     this._onRowDataLoaded(result);
             }, this);
             req.send();
         }
     }
});

-------------------------------------------------------------------------
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

[Kenny Tilton-2]

Kenny Tilton wrote:

> Alexander Back wrote:
>
>>Hi Kenny,
>>
>>just for completion (and for all other users out there using the Remote
>>Table Model): I wrote a basic tutorial on how to implement it (client-
>>and server-side). And of course, Derrell's demos is mentioned as well.
>>
>>-> http://qooxdoo.org/documentation/0.8/remote_table_model
>
>
> Thx agaon for that lifesaver. I have generalized it(and probably cocked
> it up), code trailing below.
>
> Not sure about the bit that avoids repeatedly asking the server for the
> count (I saw about twenty requests from the model before I could get an
> answer back).
>
> Instances are created with two URLs, one to get the count, one the data.
>
> Sample usage:
>
>   new nuvia.data.TablePagerJSON("/getdatacount?store=netitems&ns=asp"
>                               , "/getitems?ns=asp", 10);
>
> I know this is a wreck, it is more a suggestion for ways to enhance the
> RemoteModel.
>
> cheers, kenny
>
> --------------- Code starts here ----------------------------
>
> qx.Class.define("nuvia.data.TablePagerJSON", {
>      extend: qx.ui.table.model.Remote,
>      construct: function(countURL, rowURL, blocksize){
>          this.base(arguments);
> this.debug("making new pager "+countURL);
>          this.countU = countURL;
>          this.rowU = rowURL;
> this.setBlockSize(blocksize);
>      },
>      members: {
>          countU: null,
>          rowU: null,
> countPending : false,
>          _loadRowCount: function(){
> self = this;
>
> if (!this.countPending) {
> var req = new qx.io.remote.Request(this.countU, "GET",
> "application/json");
> this.debug("asking for count ");
> req.addListener("completed", function(response){
> var result = response.getContent();
> self.countPending = false;
> if (result != null) {
> this.debug("got count! "+ result);
> this._onRowCountLoaded(result);
> }
> }, this);
> this.countPending = true;

This trick of not kicking off redundant requests seems to break
something. I have disabled that in my code.

FYI.

cheers, ken

> req.send();
> }
>          },
>          _loadRowData: function(firstRow, lastRow){
>              var fullU = this.rowU;
>              if (fullU.indexOf("?", 0) > -1)
>                  fullU = fullU + "&";
>              else
>                  fullU = fullU + "?";
>              fullU = fullU + "from=" + firstRow + "&to=" + lastRow;
>              this.debug("JSON pager hits " + fullU);
>              var req = new qx.io.remote.Request(fullU, "GET",
> "application/json");
>              req.addListener("completed", function(response){
>                  var result = response.getContent();
>                  //this.debug("json pager gets " + result);
>                  if (result != null)
>                      this._onRowDataLoaded(result);
>              }, this);
>              req.send();
>          }
>      }
> });
>
> -------------------------------------------------------------------------
> 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
>

--
http://www.theoryyalgebra.com/

-------------------------------------------------------------------------
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

[Fabian Jakobs]

Nathan Hadley schrieb:
> Ah was looking to see if it had already been reported,
>
>
>  
Hi Nathan,

the fix was very simple. You can easily patch your qooxdoo if you don't
want to wait for 0.8.1:

Modified Paths:
--------------
    trunk/qooxdoo/framework/source/class/qx/ui/form/SplitButton.js

Modified: trunk/qooxdoo/framework/source/class/qx/ui/form/SplitButton.js
===================================================================
--- trunk/qooxdoo/framework/source/class/qx/ui/form/SplitButton.js 2008-10-02 13:37:07 UTC (rev 16447)
+++ trunk/qooxdoo/framework/source/class/qx/ui/form/SplitButton.js 2008-10-02 13:56:23 UTC (rev 16448)
@@ -194,7 +194,7 @@
           control = new qx.ui.form.Button;
           control.addListener("execute", this._onButtonExecute, this);
           control.setFocusable(false);
-          this._addAt(control, 0);
+          this._addAt(control, 0, {flex: 1});
           break;
 
         case "arrow":


Best Fabian

> --
> Nathan Hadley
> www.celcat.com
> Tel: [phone]
> CELCAT is the trading name of Corbett Engineering Ltd,
> a limited company registered in England and Wales with
> company number 1448722. VAT number: 418696219.
> Registered office: 21-23 Mercia Business Village,
> Torwood Close, Coventry, CV4 8HX, United Kingdom
>  
> -----Original Message-----
> From: Christian Schmidt [mailto:schmidt.christian@...]
> Sent: 19 November 2008 16:47
> To: qooxdoo Development
> Subject: Re: [qooxdoo-devel] Split button not taking up set Width
>
> Hi Nathan,
>
> yes, this is a bug. This bug is fixed in the SVN trunk.
> (http://bugzilla.qooxdoo.org/show_bug.cgi?id=1441)
>
> Regards
> Chris
>
> Nathan Hadley schrieb:
>  
>> Hello all,
>>
>> Bit of a basic question, I'm trying to set the width a SplitButton and
>> I'm finding that although the space is given, the button itself
>> doesn't fill it up.
>>
>> In this example I wanted the button to take up all of the room under
>> the group box;
>>
>> Is this a bug or intentional?
>>
>> --
>> **Nathan Hadley**
>> www.celcat.com
>> Tel: [phone]
>>
>> CELCAT is the trading name of Corbett Engineering Ltd,
>> a limited company registered in England and Wales with
>> company number 1448722. VAT number: 418696219.
>> Registered office: 21-23 Mercia Business Village,
>> Torwood Close, Coventry, CV4 8HX, United Kingdom
>>
>> ------------------------------------------------------------------------
>>
>> *From:* Jean-Baptiste BRIAUD -- Novlog [mailto:j-b.briaud@...]
>> *Sent:* 19 November 2008 15:44
>> *To:* qooxdoo Development
>> *Subject:* Re: [qooxdoo-devel] Passing named parameter without the URL
>>
>> Hi Derrell,
>>
>> Thanks a lot, that's very kind !
>>
>> Once again, I'd like to hightlight another good point of qooxdoo : the
>> community.
>>
>> Now, about the patch itself, could you confirm it patch the
>> io.remote.Request class ?
>>
>> I never have to deal with qooxdoo patch, but what should I do then ?
>>
>> I'm afraid I need to be educated on that point. This is not lost since
>> I'll be able to help on testing other patches after that :-)
>>
>> I think the simplest thing will be to :
>>
>> * checkout the HEAD from SVN,
>>
>> * use the qooxdoo build to produce that qooxdoo in one file with Petr
>> json config file
>>
>> * test against my code.
>>
>> I have a big company meeting now, but I'll check all that tomorow.
>>
>> JBB.
>>
>> On 19 Nov 2008, at 16:02, Derrell Lipman wrote:
>>
>>     On Tue, Nov 18, 2008 at 8:31 AM, Jean-Baptiste BRIAUD -- Novlog
>>     <j-b.briaud@... <mailto:j-b.briaud@...>> wrote:
>>
>>     On 18 Nov 2008, at 14:23, Derrell Lipman wrote:
>>
>>         On Tue, Nov 18, 2008 at 8:14 AM, thron7
>>         <thron7@...
>>         <mailto:thron7@...>> wrote:
>>
>>
>>         >> Using setParameter() specifically tells it to encode the
>>         parameters in the
>>         >> URL. I believe you're looking for setFormField().
>>         >>
>>
>>         > If you have form fields, then a different transport will be used.
>>
>>         This is true, but I strongly object against the XmlHttp
>>         transport being
>>         restricted to GET and URL parameters. If this is really the case,
>>         somebody has to open up a bug for it. XmlHttp transport has to
>>         support
>>         POST and setParameters() working on the request body!
>>
>>
>>         XmlHttp transport _does_ support POST. That's not the issue.
>>         The issue which has been discussed in depth in the past is
>>         that even with POST, it must be possible to add parameters to
>>         the URL. You (and the reset of those in this discussion) are
>>         right that there *should* be a way to add parameters to the
>>         POST request body. There's just no way to explicitly specify
>>         that currently. This is just one of many issues that have been
>>         addressed in the past, that should be addressed in a rewrite
>>         of qx.io.remote.*
>>
>>     I didn't know it was discussed before, sorry ...
>>
>>     So, I do agree with you : parameter to the POST request body is a
>>     need.
>>
>>
>>     Jean-Baptiste, would you please test the attached patch (or pull
>>     from SVN; I've checked it in). Note the new parameter available to
>>     setParameter() which specifies whether you want the parameter to
>>     be in the URL or as data. This should do exactly what you're
>>     looking for. Let me know whether this works properly for you.
>>
>>     Cheers,
>>
>>     Derrell
>>
>>     <y.patch>-------------------------------------------------------------------------
>>     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=/_______________________________________________
>>     <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
>>  
>>    
>
>
>  

--
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