Solution
First of all, I include the lowpro library in my layout to think unobstrusive about javascript:

<%= javascript_include_tag 'lowpro' %>

Here is my controller:

class SearchController <  ApplicationController
  def index
    @profiles = Profile.search params[:query] :page => (params[:profile_page] || 1)
    @videos = Video.search params[:query], :page => (params[:video_page] || 1)

    respond_to do |format|
      format.html
      format.js do
        render :update do |page|
          if params[:profile_page]
            page.replace_html 'profiles', :partial => "profiles"
          elsif params[:video_page]
            page.replace_html 'videos', :partial => "videos"
          end
        end
      end
    end
  end
end

Here is my view:

<div id="profiles">
	<%= render(:partial => 'profiles') %>
</div>
<div id="videos">
	<%= render(:partial => 'videos') %>
</div>
<script type="text/javascript">
//<![CDATA[
Event.addBehavior.reassignAfterAjax = true;
Event.addBehavior({
	'#profiles_pagination a' : Remote.Link
})
Event.addBehavior({
	'#videos_pagination a' : Remote.Link
})
//]]>
</script>

And last, one of my partial:

<h3>Profiles</h3>
<% for profile in @profiles -%>
<%= link_to profile.name, show_profile_path(profile.name) %>
<% end %>
<%= will_paginate @profiles, :param_name => 'profile_page', :id => true %>

Done
But... don't hesitate to comment, question or suggest improvements.

Thanks,
Guillaume

Si vous avez de l'aide à apporter pour répondre à ces données, manifestez vous, via les commentaires ou en me contactant  : sarah (point) haim (at) anaska (point) com.

PHP

• Pourquoi avoir choisi __construct() et ne pas avoir gardé la manière PHP4 ?
• Vaut-il mieux utiliser __FILE__ ou la variable d’environnemet ($_SERVER) ?
• Désactiver var_dump pour gérer des contextes
• Est-ce que je devrais ré-écrire tous mes scripts pour passer à PHP6 ?
• Puis-je utiliser une classe abstraite, pour des méthodes statiques ?
• Flux et sockets : dois-je refaire mon wrapper pour utiliser les sockets ?

Talend

• Faire un diff sur des fichiers
• Transformer des fichiers : plusieurs lignes à partir d’une seule

Javascript

• Est-ce que continue sert dans les boucles labellisées ?
• Peut-on utiliser AND au lieu de && et OR au lieu de || en Javascript ?
• Eclipse existe-t-il en français ?

OpenOffice

• Comment paramétrer OpenOffice pour qu’il ressemble à MS Word ?
• Comment voir le contrôle de repérage dans OpenOffice ?
• Comment organiser ses modèles dans OpenOffice ?

Did I forget to mention that we encapsulated our Add/Edit/View operations into a User Control consisting of a FormView control, and that these user controls were displayed in ModalPopupExtensions? So here's how it all went down....

We have our UC with a FormView containing all the necessary fields for adding/editing/viewing an entity in the database, as well as all the code for saving the record. We had the same experiences that many other people had using Validators inside a ModalPopup (in no particular order):

• Validators fire when the panel is initially loaded (not good at all when you have blanks in the "add new" case)
• Nothing happens when the "save" button is clicked (even when all validation rules are met)
• Cancel button doesn't work until all validation rules are met
• A few other behaviors that don't come to mind right now but were annoying or disastrous...

After trying the tips and workarounds I could find on Google with varying levels of near-success, I thought that the problem might have been the FormView control since that was the one element I was using that the posts were not. This made sense to me since the FormView was binding fields to a data source on load, whereas the other examples were just displaying various controls unbound. Just as I was about to rip out refactor all of our user controls to use plain panels and controls, inspiration struck.

Since the problem was that the validators were firing when I didn't want them to (on load when the FormView was bound to a new record or on Cancel with invalid fields) and not firing when I did, and I was already running code-behind when the user clicked the Save button (to Trim() TextBox text, validating data constraints prior to saves, etc), I figured why not take control of the validation process?

I was able to do this by setting the Enabled property of the validators to False and then, in the code behind, performing my data checks and displaying the appropriate validator through code. Since I was using the ValidatorCalloutExtensions, I was able to take advantage of the fact that the IsValid property of the target control of the VCE is what triggers the visibility of the VCE - regardless of whether or not the Validator is enabled! The very nice side effect of this approach is that making the validator IsValid property equal false while the validator is not enabled, the Page itself is not set to an invalid state and postbacks can still occur!

So here's a quick code block showing this:

MyControl.aspx

<%@ Control Language="vb" AutoEventWireup="false" CodeBehind="MyControl.ascx.vb"
	Inherits="MyControl" %>
<%@ Register Assembly="AjaxControlToolkit" Namespace="AjaxControlToolkit" TagPrefix="cc1" %>
<asp:HiddenField ID="currentID" runat="server" value="0" />
<asp:FormView ID="FormView1" runat="server" BorderColor="Tan" BorderWidth="1px" DataKeyNames="GoalID"
	ForeColor="Black" DataSourceID="uxGoalDataSource">
	<InsertItemTemplate>
		<asp:Label ID="Label9" runat="server" Text="Abbreviation:"></asp:Label>
		<asp:TextBox ID="uxAbbreviationTextBox" runat="server" Text='<%# Bind("Abbreviation") %>' Width="30px" MaxLength="2" />
		<asp:RequiredFieldValidator ID="uxAbbreviationTextBox_RequiredFieldValidator"
			runat="server"
			ErrorMessage="Goal Abbreviation is required."
			SetFocusOnError="true"
			Display="None"
			Enabled="False"
			ControlToValidate="uxAbbreviationTextBox" />
		<cc1:ValidatorCalloutExtender ID="uxAbbreviationTextBox_RequiredFieldValidator_ValidatorCalloutExtender"
			runat="server"
			Enabled="True"
			TargetControlID="uxAbbreviationTextBox_RequiredFieldValidator" />
		<asp:Button ID="uxSaveButton" runat="server" OnClick="uxSaveButton_Click" Text="Save"/>
		<asp:Button ID="uxCancelButton" runat="server" OnClick="uxCancelButton_Click" Text="Cancel"/>
	</InsertItemTemplate>

</asp:FormView>

MyControl.aspx.vb

Partial Public Class MyControl
	Inherits System.Web.UI.UserControl

	Public Delegate Sub SavedHandler(ByVal sender As Object, ByVal e As EventArgs)
	<ComponentModel.Browsable(True)> _
	Public Event Saved As SavedHandler

	Public Sub Save(ByVal sender As Object, ByVal e As EventArgs)
		Dim vo As New dataVO()
		Dim result As Boolean
		Dim value As String
		Integer.TryParse(currentId.Value, vo.Id)
		vo.Abbreviation = DirectCast(FormView1.FindControl("uxAbbreviationTextBox"), TextBox).Text.Trim()
		If (String.IsNullOrEmpty(vo.Abbreviation)) Then
			CType(FormView1.FindControl("uxAbbreviationTextBox_RequiredFieldValidator"), IValidator).IsValid = False
			Return
		End If
		RaiseEvent FilterButtonClicked(Me, EventArgs.Empty)
	End Sub
End Class

With this user control, all I need to do is include it in another page and add an event handler for the Saved event to hide the modal popup - or do whatever else I would want to do when the user saves the record.

JSON response should be like this.

bookobj=
[
"book1":{
“bookname”:”ajax”,
“price”:600,
“chapters”:{”chapter1″:”Ajax Basics”, “chapter2″:”Ajax Examples”, “chapter3″:”Using JSON in Ajax”},
“authors”:["orielly","nicholas"]
},
"book2":{
“bookname”:”json”,
“price”:800,
“chapters”:{”chapter1″:”JSON Basics”, “chapter2″:”JSON DataTypes”, “chapter3″:”JSON Examples”},
“authors”:["orielly","douglas crockford"]
}
]

Code to retrieve the data from the above response.

bookobj.book1.bookname - return the value of bookname "ajax"
bookobj.book1.chapters.chapter1 - return value of chapter1 in book1 "Ajax Basics"
bookobj.book2.authors[1] - return the value of book2 author "douglas crockford"
bookobj.book2.price - this will return the price of book2. "800"

Example for execute the scripts in the JSON data.
<html>
<body>
<div id="titleid" style="display:none;">Book Name</div>
<div id="bookname"></div>
</body>
<script>
function showTitle()
{
document.getElementById('titleid').style.display = '';
}
function showData(param)
{
document.getElementById('bookname').innerHTML = param;
}
</script>
</html>

JSON response.

obj=
{
"callmethod1":"showTitle()",
"callmethod2":"showData('Ajax')"
}

execute the scripts using javascript eval function.

eval(obj.callmethod1); - this will execute the javascript showTitle method.
eval(obj.callmethod2); - this will execute the javascript showData method.

In de jaren'60 en '70 werden diverse programma's van Ajax, het Nederlands Elftal, KNVB-Beker finales en andere grote wedstrijden voorzien van tekeningen van Dik Bruynestein. Voor het duel met Atletico Madrid op 28 april 1971 werden de trainers van beide teams (Michels en Domingo) vereeuwigd als stier en stierenvechter, met de stier als meest optimistische van de twee. Terecht bleek achteraf, want Ajax plaatste zich voor de tweede keer in de historie voor de Europa Cup I finale en zou deze voor het eerst gaan winnen tegen Panathinaikos.

Het programma is uitgegeven in A5, 36 pagina's, kostte 75 cent en heeft als nr. 36. Ir Ad van Emmenes schrijft een uitgebreide opening. Veel foto's van de verschillende spelers van beide ploegen sieren het boekje op. Een prima uitgave.

Some quick testing proved they building an index clientside was too slow, so we're building the r-tree serverside and serializing to Json.

On a 6500 feature dataset we're getting .01s (FF) to .03s (IE) search times -- fast enough for on-mouse-move GUI response.

The only caveat is that JSON nodes consisting of integer IDs and double-precision bounding-boxes run ~85 bytes per feature. Above 2500 features you must consider wire-size. Including attribute data would quickly bring this over the top.

However, optimizing the leaf node structure for point data and g-zipping the r-trees should run <3 bytes per feature -- making clientside indexing of 100,000+ features within reason.

ASP.NET

• Buenas prácticas para crear sitios ASP.NET con IIS 6.0: Omar Al Zabir, autor del libro Creando un portal web 2.0 con ASP.NET 3.5, tiene un artículo que detalla las buenas prácticas que hay que seguir cuando configuramos un sitio con IIS 6.0. Es una lectura y compra obligada.
• Videos de ASP.NET Dynamic Data con VB: Bill burrows ha unido en una serie de vídeos un tutorial que nos enseña a usar el nuevo soporte de ASP.NET Dynamic Data de .NET 3.5 SP1. Podéis encontrar más enlaces a tutoriales de ASP.NET Dynamic Data en el anterior post sobre links aquí.
• Explorando la cache en ASP.NET: Abhijit Jana nos muestra las opciones de cacheado con ASP.NET. Si estáis interesados en otra gran técnica (pero no muy conocida), querréis leer un post que escribí aquí.
• Routing con WebForms: Wally McClure tiene un podcast donde nos muestra cómo usar la nueva infraestructura de rutado de ASP.NET en .NET 3.5 SP1 con páginas basadas en formularios. Un montón de gente piensa incorrectamente que esta característica sólo funciona en aplicaciones ASP.NET MVC - cuando en realidad también funciona con formularios web (en realidad todos los sitios con APS.NET Dynamic Data lo usan.
• Integración continua con ASP.NET y despliegue usando CruiseControl.NET, Subversion, MSBuild y Robocopy: Omar Al Zabir tiene otro artículo - esta vez implementando la integración continua con ASP.NET.

ASP.NET AJAX y jQuery

• Introducción a jQuery (Parte 1): Rick Strahl ha posteado un artículo de introducción a jQuery, y habla sobre cómo aprovechar sus características en páginas ASP.NET.
• Nuevo soporte AJAX para aplicaciones guiadas por datos: Bertrand Le Roy ha escrito un artículo para MSDN que muestra las nuevas características disponibles en ASP.NET AJAX en la preview de hoy. Leed también sus post aquí y aquí para aprender más sobre el soporte de templates de lado del cliente.
• Usando jQuery para mejorar el indicador de progreso de ASP.NET AJAX: Dave Ward tiene un artículo que describe cómo integrar la funcionalidad de jQuery con el control UpdatePanel de ASP.NET AJAX para mejorar el estado de progreso.
• ASP.NET AJAX: Habilitar marcadores en el botón de volver del navegador: Scott Mitchell continúa con su serie sobre ASP.NET AJAX y muestra cómo añadir puntos en el historial en paginas con AJAX para que los visitantes puedan marcarlas, así como habilitar la navegación atras/siguiente en el navegador. Esta es una nueva característica que se ha añadido a ASP.NET  en .NET 3.5 SP1.
• 46 tutoriales para el ASP.NET AJAX Control Toolkit: Christian Wenz ha publicado 46 tutoriales tanto en VB como en C# en el que nos muestra escenarios típicos con el ASP.NET AJAX Control Toolkit.

Microsoft Web Platform

• Instalación de plataformas web: Hacer un instalador para despliegue web: Scott Hanselman tiene un post donde nos muestra el nuevo "Microsoft Web Platform Installer" que estamos creanto para permitir una forma rápida de instalar cada componente web de Microsoft - y tener rápidamente una máquina lista para el desarrollo web.

Espero que sirva.

Scott.

Traducido por: Juan María Laó Ramos.

Artículo original.

Derrière se nom un peut ‘barbare’ se cache une interface utilisateur qui combine un Vista-like avec un menu "Démarrer" et un OS X avec un dock (très fluide d’ailleurs).  On y trouve par exemple des widgets du système Google Widgets.

En parlant de Google, l’environnement de développement est "GwtPHP" qui tire les avantages de GWT et  les déploie à PHP  pour tenter de résoudre le problème de Java qui est un peut limité et moins convivial.

Rappelons que GWT (Google Web Toolkit) est un framework développé par google, permettant de créer des pages web dynamiques en utilisant la technologie AJAX. C'est un logiciel libre distribué selon les termes de la licence Apache 2.0.

GWT  renvoie  massivement à des appels distants asynchrones qui gère le bouton « Suivant » et « Précédent » d’une grande majorité de navigateurs. Au delà du simple Framework, cette solution apporte une nouvelle approche du développement d'application de type client léger. En effet, le développeur fait ici abstraction quasi complète de la complexité habituelle liée à javascript, css et html et utilise une bibliothèque graphique fondée sur le Modèle-Vue-Contrôleur.

Enfin bon, revenons à nos mouton, malheureusement le service PostAffiliateXpress,   n'est pas disponible pour l’instant, mais se webos seras ouvert très probablement au début du mois de Novembre.

Globalement, c'est plutôt bien mis en place notamment pour les fenêtrages en Ajax.
Pour la licence les développeurs ont prit l’initiative d’utiliser une double licence : "Gratuit pour les amateurs, payant pour les usages commerciaux" (il faut bien un business model derrière tout ça !).

Sans plus tarder je vous laisse tester PostAffiliateXpress et en juger par vous même !

]]> http://matthiaswessendorf.wordpress.com/?p=134 Sat, 11 Oct 2008 15:18:58 +0000 matthiaswessendorf http://matthiaswessendorf.ar.wordpress.com/2008/10/11/nice-charts-with-trinidad/ Apache MyFaces Trinidad provides a lot of cool JSF components; a really nice one it the chart component. It offers several possibilities to visualize your structured data. Like this pie chart:

To actual provide such a nice looking chart graphic, you just need to implement an abstract class, provided by Trinidad's API. This may sound a bit complicated... but actually it is straightforward:

[sourcecode language="java"]
public class CustomerDevelopment extends ChartModel
{
  public List getGroupLabels()
  {
    List years = new ArrayList();
    years.add("2006");
    years.add("2007");
    years.add("2008");
    years.add("2008");
    return years;
  }
  public List getSeriesLabels()
  {
    List products = new ArrayList();
    products.add("Product A");
    products.add("Product B");
    return products;
  }
  public List> getYValues()
  {
    List> chartValues = new ArrayList>();
    // iterate over the groups (years...)
    for(int i = 0; i < getGroupLabels().size(); i++)
    {
      List numbers = new ArrayList();

      // now, just read the series, per group (product/year)
      for(int j = 0; j < getSeriesLabels().size(); j++)
      {
        // yes... some random stuff here, in the demo ...
        numbers.add(10000 * Math.random());
      }
      chartValues.add(numbers);
    }
    return chartValues;
  }

}
[/sourcecode]
Actually, the only "tricky" part is the getYValues() method. Here you do two loops to get the correct data. When you are already in the business of providing chart data... this is somewhat familiar.

The JSPX (or Facelet) is also very simple:

[sourcecode language="html"]

XML

• XML parsing is generic.
• XML is a markup language and has semantics.
• XML is most easily readable format for humans.
• Easily validate the data using schema.

JSON

• JSON data is faster than XML over the Http.
• Scripts for JSON data are simpler than XML.
• For i18n,  JSON seems to be an easy one.
• Its very hard to read humans.

To conclude, based on your requirements you can choose the format either XML or JSON.

What Is jQuery?
jQuery is yet another JavaScript library to join the previously crowded space that includes Prototype, Scriptaculous, Rico, Moo.Fx and more than a dozen others. To use it, simply attach the .js file in the head of your page: magically, you have access to lots of pre-built functions and gizmos.

Why jQuery?
The true beauty of jQuery is what it can present you within the first 10 minutes of your using it. The key feature of jQuery is its simplicity. Few lines of jQuery code can replace a dozen lines of normal JavaScript, yet it remains very elemental and flexible. Let me illustrate this point with an example. Two years ago, we used the following script to fix the web page horizontal rule:

function ourRules() {
if (!document.getElementsByTagName) return;
var hr = document.getElementsByTagName("hr");
for (var i=0; i
var newhr = hr[i];
var wrapdiv = document.createElement('div');
wrapdiv.className = 'line';
newhr.parentNode.replaceChild(wrapdiv, newhr);
wrapdiv.appendChild(newhr);
}
}

window.onload = ourRules;

Result of the code, the browser waits for the page to finish loading before rifling through the DOM to find each occurrence of hr. Each time it finds one, it creates a new div, gives it the class name "line", inserts it where the hr was, and pops the old hr inside the new div, to achieve the markup required to apply this particular effect. The end result of this script was that we were able to get the desired result without having to change hundreds of pages.

But, we'd achieve the same result using jQuery.

$(document).ready(function(){
$("hr").wrap("<div class='fline'> </div>");
});

To break it down:

$(document).ready(function(){
...
});

The first and third lines are jQuery's load event, and they replace the old window.onload from above. Any activity that we wish to complete during the page load can be dropped inside these curly braces. This is a great upgrade on the old onload method, because rather than waiting until everything has finished loading, jQuery's function watches everything that comes in, and starts working as soon as it has all the parts it needs. It's really very neat.

Surprisingly, the second line is even simpler:

$("hr").wrap("<div class='fline'></div>");

The "dollar object" $("hr") is all we need to tell jQuery to grab every horizontal rule on this page, and wrap is what we will be doing to those hr elements.

jQuery's built-in wrap function takes in whatever HTML we give it (in this case "<div class='fline'> </div>") and wraps it around each hr in our page - no loops or tests required.

We've used a div here, but we could just as easily been modifying or wrapping a b, span, or a element.

And although we've used a very simple selection rule here (all hrs), we could have easily been much more specific with what we targeted. Using familiar old CSS syntax, we could have used any of the following:

$("hr.separate") - Get the hr elements with the class name "separate ".
$("li:only-child") - Get list items that are by themselves.
$("ul > li") - Get only list items with unordered parent lists.

While I've found wrap to be of the most instantly helpful jQuery functions, it's just one of many, including hide, show, fadeOut("slow") and slideUp("fast"), just to name a few. You can perhaps guess what each one of these functions does. The jQuery starter's tutorial on the jQuery site is quite a gentle beginner's guide, and takes you through some of the most regular functions. But perhaps jQuery's single neatest feature is its ability to "chain" functions together. Like, if I wanted to add a second div to our hr elements for some foolish reason, I could simply add another call to the wrap function to the end of my code, like this:

$("hr").wrap("<div class='fline'></div>").wrap("<div class='sline'></div>");

It's so easy, it's wild. Wild like a fox!

While designing a web page, you want to add a small icon to the bottom corner of each thumbnail. This required each img element to be wrapped in a container div, and another div showing the icon to be positioned in the container div. Again, the jQuery code is just one line.

$("#thumbnails li img")
.wrap("<div class='mywrap'></div>")
.before("<div class='mythumb'></div>");

In simple English, this code just asks jQuery to locate all the images in li elements that are inside #thumbnails, Wrap these images in a div called "wrap", Squeeze another div (the one with the icon graphic) in my "wrap" div just before my image.

Now that we have the structure, CSS does the rest. If JavaScript is turned off, the thumbnails link directly to the raw image files, and there's no need for the icons. Now that's what I call graceful degradation. Like most other JavaScript libraries, jQuery is competent of some very high-end actions, but the main attraction for me was its ability to solve the little problems quickly and with a minimum of fuss.

I hope you'll find it helpful too.

The first stream was devoted to discussing the hot topic of providing rich customer experiences - and you can read a session review here. It studied the technical, marketing and operational issues related to rich media.

In this stream, Salmon customer Oliver Bishop, the e-Business Technical Architect from AkzoNobel Decorative Coatings, presented on "Making product choices simpler using RIA technology."  The latest relaunch of dulux.co.uk has taken advantage of Rich Internet Application (RIA) technology to help customers to choose the colours they love and products that are right for the job.   This is achieved using functionality like Ajax, drag and drop and DHTML extensively within the site to create an interactive scrapbook, moodboard and the ability to upload and decorate photos of their own rooms.  Oliver's presentation covered how Web 2.0 technologies can enhance the experience for your customers.

Stream two looked at second mover advantage on the web - and focused on the approach taken by retailers who have launched transactional websites only recently; and how they have taken advantage of their second mover position to succeed.

In this 2nd stream, Wendy Derbyshire from Scotts & Stow told delegates how the company had been running nine sites on four different platforms and faced a huge range of legacy issues and a major challenge to consolidate everything into one technical base that could underpin the whole business.  Scotts decided to bite the bullet and embark on a major project (with Salmon and SAFE™ - read the case study) to put in place one infrastructure for the whole business.

And then there was Stream three, which examined the latest trends in multiple channel retailing, i.e. Moving beyond multichannel to true, integrated "cross channel" solutions (is it me or does this makes our 2005 seminar series "eCommere Junkies: Cross Channel Retail" look impressively visionary?).

In this final stream Argos' presentation looked at "Customer centric multi-channel development".  There is no doubt Argos is recognised as pioneering new approaches in multi-channel high street retailing, introducing concepts (and delivering them with Salmon's help) such as in-store reservations via web, phone and text, in-store stock-check and quick-pay kiosks, all tied into the overarching online experience and offline catalogue.

It's funny.  You can have the biggest booth in the exhibition hall and learn to talk a good game.  But nothing compares to the experience you get from actually living, breathing and delivering state-of-the-art eCommerce.

Link para as apostilas:http://blog.caelum.com.br/apostilas/

O site da Caelum para quem quer saber sobre os cursos é www.caelum.com.br

Caelum, parabéns pelo material. Os cursos devem seguir o mesmo nível das apostilas!

Ótima qualidade!

Até a próxima!

A great view after hiking down Little Annie's Road, off the back of Aspen Mountain, less than a month after I moved to Aspen.  The mountain wasn't opened for skiing yet, but the gondola was open, so I went exploring... Good exercise, before eating and drinking waayyy too much.  Taken 11/22/2007.