So what I’ve been doing is setting up a new action using add_action for the following two action types:

• wp_ajax_<myaction>
• wp_ajax_nopriv_<myaction>

According to the documentation that should generate a hook to support both registered users (wp_ajax_<myaction>) and unregistered users (wp_ajax_nopriv_<myaction>). Which sounds great, since my theme should work for both type of users .

On the front-end I’ve created a Javascript method that should call the ajax command using the ‘wp-admin/admin-ajax.php’. With as action parameter the name of my ajax method. So in theory I should be setup correctly to handle ajax, but this is where the theory and actual situation start to differ.

To test my new theme I’ve been using the preview mode of WordPress, after all I don’t want to kill any existing feature or activate the theme before it is done. But for some reason my nicely setup Ajax hooks aren’t working as advertised.

Now I’m not going to be boring you with all kinds of tutorials on how to develop on Android, or give you any other clues on this subject. I just don’t know anything about Android development, well not yet anyhow. But some good reading material to get you on the right path might be:

• The Android development guide, all you ever will need to know about Android development (bit technical though)
• Writing your first Android app, good tutorial to get started in Android

The only thing I have learned so far is that I don’t really like the way of designing a UI for the Android. It’s just way to basic with XML files describing how the UI should look like. Feels like going back to 1990 when designing a UI was done in code rather then with a UI design tool.

So I started looking into a way to let my web application determine what browser was visiting and what capabilities it had. Since this was for the new WordPress theme I’m developing it had to be in PHP and be sorta light-weight.

First thing was to figure out the basic browser that was visiting. This can be done by doing a simple strpos call in PHP on the user agent property. Having a list of the various user agents helped, I found just such a list at UserAgentString. A brief summary for some common browsers is below.

  $user_agent = $_SERVER['HTTP_USER_AGENT'];
  $IE        = strpos($this->_user_agent, 'MSIE') > -1;
  $IPad      = strpos($this->_user_agent, 'iPad') > -1;
  $Firefox   = strpos($this->_user_agent, 'Firefox/') > -1;
  $Chrome    = strpos($this->_user_agent, 'Chrome/') > -1;
  $Opera     = strpos($this->_user_agent, 'Opera/') > -1;
  $Android = strpos($this->_user_agent, 'Android') > -1;

As you can see determining the basic browser type is pretty simple. But to be able to determine if CSS3 is supported you will also need to locate the version of the browser. Keeping in mind that IE9+, Firefox 4+, Opera 10+ and Chrome are the only versions that support CSS3. The check for this is slightly different for each version. But you can do it easily with a regular expresion.

 // For all non IE browsers
 $locator;
 if ($Firefox) $locator = "Firefox\/";
 if ($Android) $locator = "Android\s";
 if ($Chrome) $locator = "Chrome\/";
 if ($Opera) $locator = "Version\/";
 preg_match('/'.$locator.'([0-9]{1,4}(.[0-9]{1,3})+)/', $user_agent, $version);

 // For IE
 preg_match('/MSIE\s([0-9]{1,2}.[0-9]{1,2})/', $user_agent, $version);
 if (count($version) > 1) $version = $version[1];
 else $version = 0;

Having the version and the type of browser allows us to do all kinds of fancy browser specific stuff. Mostly introduce hacks for IE7 (yes many are needed), or add CSS3 support.

We are using various IDE’s for editing our Java projects, meanly because different developers like different IDE’s. Personally I’m using Netbeans, which has perfect Subversion, Git, CVS and Perforce support. Some other developers use IntelliJ from Netbrains. Which is sorta expensive and the version we are using has no support for TFS at all.

So when I want to edit a simple file I first locate it in my IDE, which takes me about 5 seconds. Then I need to look up its location on disk, followed by opening the completely useless Team Exporer. Then I again need to look up the file in TE, followed by a check-out. The whole process to the point that the file is checked out and ready to go takes probably about 1 minute, each and every time.

On an average day I probably waste up to an hour checking out files, and commiting them back into the version control system.

To ease the burden a bit I’ve tried the SvnBridge tool to link my IDE’s Subversion system with TFS, but this tool has many flaws. Just a few of which are:

• No longer appears to be developed
• A merge in TE causes all updates to fail, you need to do a clean checkout again
• Random update failures due to commits of other developers

So for now I’m stuck. Really, really, really stuck.

Developing a website with JQuery / Ajax and advanced JavaScript prototypes isn’t anything new to me. But having all of these features and still make the site indexable is! So far I’m stuck at the simple point that crawlers don’t support JavaScript, so no go for that part. It looks like I will have to create a module for the WordPress plugin to detect what type of user is visiting the website. Serving the search engines a version that will not use any of the advanced features that normal browsers will support.

Besides the issue with the indexing there is still the task of getting all the pages on the website to be tracked with Google Analytics. Currently that is easy, though the website is dynamic each page is a new request to the server. But using advanced features like Ajax stops that in its tracks. Currently I don’t even know if it is possible to track pages when the actual URL doesn’t change, just the content of the page. If it is possible to do this I would have to build a module to do so.

So after just ranting and brain storming for a few seconds I already found to nice features that would have to be build. Another person might call them serious issues, but I prefer good challenges. At least for the moment I do. Not even mentioning the fact that I have to build a complete new template for WordPress too.

I started out by creating a custom component, which inherits from the Panel. This would give me pretty much a blank object that could be added to the GUI. The painting part on its own is not that difficult either. Done that hundreds of times. The problem is in adding the various objects to paint on top of the panel to the panel before drawing commences.

Ok, I’ll try to explain a little better. I keep track of what to draw in an array inside the custom panel. This because these objects contain more information then just how to draw. But they are not part of the component list of the panel. I need to add them dynamically just before drawing and remove them again after drawing. But I’ve yet to find the correct way of doing this.

Tried adding them in the onPaint method, but this doesn’t trigger them to be painted. So I’ll probably have to intercept the WM_PAINT message sent to the panel and attach the

So far for the good idea, but when Ubuntu asked me what partition to delete I wasn’t sure which was the spare one and which one contained all the source code of all applications and websites I developed. And you probably guessed it, I erased the wrong one.

Gone is a couple of days of work. Serves me right for not checking it in every day

public static final int ANIMAL_DOG    = 1;
public static final int ANIMAL_FISH   = 2;
public static final int ANIMAL_CAT    = 3;

Into something a bit more object oriented:

public enum Animal {
  Dog,
  Fish,
  Cat
}

But when you have the need to serialize the enum values then you may have to use a bit complicator code. It is supported just not out of the box. So lets say I want to predefine the ordinal number and add a name for each animal in my enum. Then I would have to add something similar to this:

public enum Animal {
  Dog(10, 'Swoop doggie'),
  Fish(20, 'Blub'),
  Cat(30, 'Garfunkel');

  int iOrdinal;
  String iName;

  private Animal(int aOrdinal, String aName) {
    iOrdinal = aOrdinal;
    iName = aName;
  }
}

Now each one of the enums values consists out of a name and a custom ordinal which can be used to save the value in a database or represent it in a form. Off course you would have to add methods to obtain the name and ordinal like:

public String getName()    { return iName; }
public int    getOrdinal() { return iOrdinal; }

That introduces a second problem, if you represent the enum Animal by its ordinal on for example a webpage you will need to be able to regenerate the correct enum value by its ordinal representation. To do this you could add a method to lookup the correct Animal for any ordinal like:

public static Animal getAnimal(int aOrdinal) {
 for(Animal lAnimal : values()) {
   if (lAnimal.getOrdinal() == aOrdinal) {
     return lAnimal;
   }
 }
 return Animal.Dog; // default value required
}

As this plugin will require a bit more work then the combobox I have no idea as to when I can release a first usable version. But as soon as I do I will publish it here .

As it turns out the autoloader functionality is called when PHP is unable to find the class you are using automatically itself as a last effort. Though that wasn't the issue I encountered. It took me about 3 hours of debugging the function to find out there was a problem with the logic of the autoloader function I wrote myself, doh!

Something with breaking out of a loop to early . After fixing this issue my autoloader was working beautifully. So I am happy to say that it is possible to write a modular PHP web application after all.