Here’s the changelog:

• Properly added Marcus Sykes as a contributor
• Added a full calendar
• Added an #_EDITEVENT placeholder
• Added Brazialian Portuguese localization and some translatable strings
• Added a today scope for events lists
• Increased to 3 te default lists limit
• Categories are now displayed in the events table
• Now weeks starts according to WP settings
• Moved the hide page option up for better access
• Attributes column was not created when the plugin was upgraded, fixed
• Added comment field to the RSVP form and #_COMMENT placeholder in RSVP email templates
• Added customizable title to small calendar
• Removed php short tags
• Changed RVSP ‘no seats available message’ for better English
• Bugfix: there was a time bug
• Bugfix: event_time not taken into consideration in ordering events, fixed
• Bugfix: on calendar for days after 28 on the event calendar view
• Bugfix: for events in days with single digit
• Bugfix: events link in the calendar now work with permalink
• Bugfix: today in next mont was not matched in the calendar
• Bugfix: _RESPPHONE was not matched in emails
• Bugfix: fixed security vulnerability, which could lead to sql inject attacks
• Bugfix: locations with apostrophe were duplicated
• Bugfix: bloginfo(‘wpurl’) instead of bloginfo(‘url’)
• Bugfix: now loading textdomain in the new (not deprecated) way

Most of the bugfixes, and some of the features, were sponsored by Sara of Paperlantern, to whom I am most grateful.

Also, I just got a haircut:

Cheers,
Davide

function approve_guestbook_comments($comment_ID) {
	$comment = get_comment($comment_ID);    
	if ($comment->comment_post_ID == guestbook_page())
		wp_set_comment_status($comment_ID,'approve');
 
}
add_action('comment_post','approve_guestbook_comments');

As you see it’s a super-simple tweak, but it’s quite effective.

Enjoy,
Davide

	* The $userdata array can contain the following fields:
	* 'ID' - An integer that will be used for updating an existing user.
	* 'user_pass' - A string that contains the plain text password for the user.
	* 'user_login' - A string that contains the user's username for logging in.
	* 'user_nicename' - A string that contains a nicer looking name for the user.
	*		The default is the user's username.

The other keys are user_url,user_email, display_name, nickname, first_name, last_name, description, rich_editing, user_registered, role, jabber, aim, yim; for more detal consult the source itself.

My script goes like this:

<?php
include 'wp-blog-header.php'; 
include 'wp-includes/registration.php';   
include 'wp-includes/pluggable.php'; 
ini_set("memory_limit","1024M"); 
ini_set("max_execution_time", "240");
global $wpdb;
?>
 
<h1>Importazione utenti EC</h1>
 
<?php
$connection = mysql_connect("host", "user", "password") or die("Unable to connect to MySQL");
mysql_select_db("database", $connection) or die("Unable to connect to the database"); 
$result = mysql_query("SELECT * FROM source_users ;");
	while ($row = mysql_fetch_object($result)) {
		echo "<strong>ID:</strong>".$row->id." <strong>login:</strong>".$row->user_name." <strong>password:</strong> ".$row->password." <strong>e-mail:</strong>".$row->email_address." <strong>name:</strong> ".$row->name." <strong>surname:</strong> ".$row->surname."<br/>";  
		// Add the ID to trick WP
		$add_id = "INSERT INTO ".$wpdb->users." (id, user_login) VALUES (".$row->id.",'"."$row->user_name"."' ); ";
		 mysql_query($add_id) or die(mysql_error());
		// Add the rest
		$userdata = array(
		 'ID' => $row->id,  
		 'user_login' => $row->user_name,
		 'user_pass' => wp_hash_password($row->password),
		 'user_nicename' => $row->user_name,
		 'user_email' => $row->email_address,
		 'first_name'  => $row->name,
		 'last_name'  => $row->surname,
		 'role' => 'subscriber'
		);                          
		wp_insert_user($userdata) ;
	}
}
 
 
mysql_close($connection);

Some highlights:

• Import all the WP files you need
• Change PHP memory and time limits if you have many records
• Explicitly encrypt the password with wp_hash_password. I have no idea why this is necessary, the WP code should do this by itself; but trust me, it is.

Since we are using the wp_insert_user() function, WP will crate records both in the users and in the usermeta tables.

A last warning. This huge amount of users could conflict with your php settings. To get WP working properly I had to add this line to wp-config.php:

	define('WP_MEMORY_LIMIT', '64M');

Without it I could not get the users list in the admin session. I also filed a bug, but this last line should keep you on the safe side.
Enjoy!

Today I have had to test a specifical IP based feature. Simply put, I get the user IP and store it into the database.

The simple rails way to get a visitor IP address is

request.remote_ip

If you do this while developing you’ll usually get a predictable

127.0.0.1

How do I test that my controller puts a correct IP in the database?
At first I thought about creating a mock request object. But this way I would be testing only the saving portion of the method, not the part detecting the IP. In other words, I needed to simulate an IP address at a lower layer.

Webrat to the rescue! Thanks to the nice folks at the webrat irc channel I got an answer. Simply:

header('REMOTE_ADDR', ip_address)

The Webrat headermethod lets you set all the header variable of you session. Handy, isn’t it?

Marcus contacted me a few weeks ago, sending a modded version of Events Manager. As soon as I gave it a try I realised that Marcus had done a wonderful job patching and integrating new features into EM.

Thanks to Marcus’s work now we have a number of frequently requested features:

• WP 2.8 support
• Events Categories
• WYSIWYG editor for events description (and you can add pictures too!)
• multi-day events, kinda

And much more.¹

Since Marcus is expanding the plugin for some of his gigs, I invited him to co-author Events Manager, and I am glad he accepted.

I just tagged EM 2.0rc2 with Marcus’s integration, hope you’ll enjoy it as much as I did.

Download EM 2.0rc2

Davide

Notes

1. Since these features are not yet fully documented, I suggest you take a look at the change log in the readme for instructions.

ActionMailer::Base.smtp_settings = {
  :enable_starttls_auto => true,
  :address => "smtp.gmail.com",
  :port => 587,
  :domain => "your-google-apps-domain",
  :authentication => :plain,
  :user_name => "username",
  :password => "password"
}

The most important line is enable_starttls_auto => true.

Unlike pre 2.2 Rails, no plugin is required. Not sure this setup works with Rails 2.2 though, never tried it.

Via this forum discussion.

Blueprint is probably the most widespread CSS framework out there; I ignore whether Josh Clayton, blueprint’s dad, was the first to think about CSS frameworks, but for certain blueprint was the first framework to get good coverage in the geek news.

Since I had to deal with it anyway, I said “Why shouldn’t I use this gig as an occasion to learn something new?” And I did indeed. Not sure the last line is graceful to the ear, but let’s roll on.

How I blueprinted at first

The style-sheet I was working on wasn’t exactly tidy. To tidy it up, I had to look into the infamous grid.css and try to understand what those weird “span-n” classes were doing. Luckily, all of blueprint source files [[I know what you're thinking, source files, what now, do I have to use gcc for a css style-sheet? Not quite, but we'll get there.]] are well commented and clear.

The basic concept is that your layout is divided i 24 columns. You can put div whose width is determined by the aforementioned .span-n class. Just remember:

• put all you columns in the .container class
• to make sure you reach “24″ as a sum
• you include .last in the last column of each row.

You know the headaches of creating a well-clearing three-columns layout? Here’s how you do that with blueprint:

<div class='container'>
	<div class='span-6'>
		left column
	</div>
	<div class='span-12'>
		center column
	</div>
	<div class='span-6 last'>
		right column
	</div>
</div>

And this is it; but not all of it. You can add .prepend-n classes to prepend empty column, and .append-nto append them. You also have other nifty classes for hr, vertical lines and so on. In addition, and maybe more importantly, blueprint gives good typography [[Ever heard about vertical rhythm?]] and a CSS reset to make sure your website is cross-browser compatible. And it’s free.
So it is good, why shouldn’t I use it?

Ehm, what about semantic mark-up?

Exactly, that is the problem. Just when we managed to get rid of the tables and their meaningless cell classes, here comes a CSS framework which binds you to class names that are just referred to measures, not semantic at all. Also, someone with no awareness of the concept of semantic could (mis)use blueprint as a pseudo-table; .span-n could be seen as a regurgitation of the colspan attribute of td cells. And sure we don’t want that, do we?

Luckily there’s more

Earlier I mentioned the sourcefiles, and I am sure some of you startled and thought that they had to learn c to use blueprint. Luckily that’s not the case, although there’s some ruby behind blueprint’s magic. The stylesheet you are supposed to attach to your html files (screen.css,print.css,ie.css) originally result from the compilation of the source style-sheets (grid.css,typography.css ans so an, all in the src directory) according to certain criteria. You can choose to ignore all of this, or you can decide you want to be a good, semantic coder and become a blueprint superuser. You don’t have much of a choice, do you?

Welcome settings.yml

settings.yml is your best friend. Really, it is. This nice guy allows you to customise any aspect of the blueprint stylesheet. You might want more or less columns, wider ones, or more space between them. josh has a nice page illustrating how you can customise blueprint.

I have a settings.yml file in ~/developing/bluprint/lib, containing all the settings for each single project I am working on.
When I start a new blueprint-based project, I just add a new entry to this file.

What I want to stress is the importance of leaving the span-n class alone (unless occasionally and during wireframing) and use the semantic classes. These classes (or ids) are defined at the end of each project entry. What they do is associating a semantic name to a geometric blueprint class. Look at the entry corresponding to my current project:

personalproject:
	path: ~/blahblah/personalproject/
	custom_css:
    	ie.css:
    		- custom-ie.css
    	screen.css:
    		- custom-screen.css
  	custom_layout:
    	column_count: 24
    	column_width: 30
    	gutter_width: 10
  	semantic_classes:
    	"#featured": ".span-15, div.span-15"

In the last section, #featured is defined as equivalent to .span-15, div.span-15. In this way my layout will contain this:

<div id='featured'>
	blah
</div>

instead of this

<div class='span-15'>
	blah
</div>

The difference is obvious.

And there’s more. Think about how easy it is to manage your layout this way. You don’t have to think about paddings and margins, you just think about columns, their number and their width. If I decide that I want to give a column more to my #featured article, I just alter a line in settings.yml: [[Of course you must also contract the following column, the sum must always be the same. ]]

	"#featured": ".span-16, div.span-16"

and recompress through this terminal command:

	ruby compress.rb -p personalproject

If you decide your whole website is too wide, just narrow the columns and recompile. And so on.
Of course your workflow has to adjust a little, but the advantages in terms of flexibility are enormous.

Blueprint is not a ghetto

Blueprint is perfectly suited for building an entire layout, but if you decide this is not your piece of cake, there’s no problem at all. All you have to do is insert one or more .containers into you css layout. I generally prefer to use blueprint for the main textual area, and leave the header (and sometimes the footer) alone. This way I can use wider divs for the background, and avoid using empty columns at the sides.

<html>
<head>[...]</head>
<body>
	<div id='wrapper'>
		<!-- This div has a nice background with shadows ad stuff -->
		<!-- Here I often put my header -->
		<div class='container'>
			<!-- Here goes the blueprint stuff -->
		</div>
	</div>
</body>
</html>

How I blueprint now

Remember that first super-fast layout? Now we can remake it semantically. We just have to play a bit with settings.yml and compress.rb.

Here’s our settings.yml

semanticlayout:
	path: ~/blahblah/semanticlayout/
	custom_css:
  		ie.css:
   			- custom-ie.css
   		screen.css:
   			- custom-screen.css
	custom_layout:
  		column_count: 24
   		column_width: 30
   		gutter_width: 10
	semantic_classes:
   		"#left-sidebar": ".span-6, div.span-6"
   		"#content": ".span-12, div.span-12"
   		"#right-sidebar": ".span-6, div.span-6, .last"

And here’s a clear-as-crystal layout:

<div class='container'>
	<div id='left-sidebar'>
		left column
	</div>
	<div id='content'>
		center column
	</div>
	<div class='right-sidebar'>
		right column
	</div>
</div>

Ain’t it nice? Fast, reliable, typographical and semantic.

If you play music and have a Mac you must get Capo by Chris Liscio of SuperMegaUltraGroovy.

Capo is the slow downer I had been waiting for for years. You might know that there’s another one, supposed to be amazing, but trust me, the amazement is all in Capo.

You import tunes simply by dragging them over the icon or Capo windows, you adjust the rate and pitch, and you’re done. If you’re obsessive-compulsive about hearing the exact not you can drop markers on the timeline, they’lll get saved with the capo document.

Capo is not available yet, but you can sign in for a sneak peek here; moreover, the beta is very stable, and I guess that app will be released in a matter of days.

By the way, Andy McGann is also amazing.

Davide

If you have any question about Events Manager 2.0, after checking out the docs, please register on the forum and post your questions there.

As you see there are 3 areas so far, bugs, how-to and feature requests. Enjoy!
Davide