If you use twitter you'll be familiar with the concept of URL shortening. You want to tweet about that video where the dog thinks its own leg is trying to steal the bone but you've only got 140 characters to say what the video is and include the link to YouTube. URL shorteners allow you to change:

http://www.youtube.com/watch?v=tJgMueh-zLM&feature=youtu.be
to:
http://bit.ly/dfzFE6

Even if you don't use twitter URL shorteners can come in handy. For example, at the beginning and end of the ITauthor podcast I use some music by Amplifico and I like to put a link to their page on musically.com in the MP3 description that you can read on your iPod when you're listening to the podcast. It's much nicer to give the URL http://tinyurl.com/amplifico, rather than http://www.musicalley.com/music/listeners/artistdetails.php?BandHash=cdef1ecef0d12844ed816b922fcada5d.

Some popular URL shorteners

• tinyurl – This was the first URL shortener most of us will have come across - way back before twitter appeared and ramped up demand for short URLs, leading to a proliferation of shortening services.
• bit.ly – twitter supported use of bit.ly which made it a popular service. Recently there have been doubts raised about the wisdom of using a Libyan registered domain (.ly) as the Libyan government have said they will take down domains that contain immoral content.
• j.mp – This is just bit.ly but with 2 fewer characters. If you already have a bit.ly URL you can use the same shortened path, stick it on the end of the j.mp domain and save yourself those 2 precious characters. For example, the dog video gets shortened to http://j.mp/dfzFE6.
• goo.gl – Google are one of the many big companies that have now got into the URL shortening business.
• is.gd – just a nice simple Web page that produces nice short URLs.
• ... I could go on, but there's not a whole lot of difference between these services.

Your very own URL shortener

Shortening URLs isn't difficult to do and there are a selection of free URL shorteners that allow you to produce your own short URLs. All you need is your own Web site and your own domain name. So, for example, I own the domain name itauthor.com, so I can produce short URLs like http://itauthor.com/1 or (more descriptively) http://itauthor.com/podcast36.

The solution I'm using is called yourls. It's a series of PHP scripts with a MySQL database behind it. So if you're already running a Web site based on PHP and MySQL (for example, a WordPress blog) then you've already got everything you need. Just upload it and browse to the admin page. The yourls contains all the instructions you need.

The only problem I had was as a result of some changes not getting written to my .htaccess file in my root Web directory. I had to go and manually add the following at the start of the .htaccess file :

# BEGIN YOURLS
<IfModule mod_rewrite.c>
RewriteEngine On
RewriteBase /
RewriteCond %{REQUEST_FILENAME} !-f
RewriteCond %{REQUEST_FILENAME} !-d
RewriteCond %{HTTP_HOST} !^itauthor.com$
RewriteRule . - [S=3]
RewriteRule ^([0-9A-Za-z]+)/?$ /yourls-go.php?id=$1 [L]
RewriteRule ^([0-9A-Za-z]+)\+/?$ /yourls-infos.php?id=$1 [L]
RewriteRule ^([0-9A-Za-z]+)\+all/?$ /yourls-infos.php?id=$1&all=1 [L]
</IfModule>
# END YOURLS

You don't need the two lines highlighted in red if you're not running WordPress, or anything similar that relies on being able to rewrite URLs. The yourls documentation says, in this situation, you need to put all the yourls files and directories in a subdirectory of your root Web directory (e.g. in a directory called "u"). However, this means that you need to include the subdirectory in the YOURLS_SITE configuration setting and it'll then be part of the shortened URL (e.g. http://itauthor.com/u/123, which kind of defeats the purpose. So the two red lines get around this by diverting URLs without "www" to yourls.

The first of the red lines says "only apply the following rule if the URL doesn't begin http://itauthor.com". The second red line says "if the previous condition resolved as true then skip the following three rules".

This seems a bit like a double negative but it's necessary because RewriteCond only applies to the RewriteRule that immediately follows it, so we need the skip rule. The result is that, on my site, the three RewriteRules that divert page requests to the yourls PHP scripts are only applied to URLs beginning http://itauthor.com. The "[L]" means "last" - in other words, if this RewriteRule is applied don't go any further, so we never reach the rules that WordPress uses, which are further down the .htaccess file. If a URL begins http://www.itauthor.com then the yourls rules are skipped and the URL is processed using the WordPress rules.

This means that http://itauthor.com/2 is sent to yourls to retrieve the original, long URL from its database, whereas http://www.itauthor.com/podcasts is sent to WordPress to create a Web page using content from its database.

What's the point?

Well, okay, there's really no point other than a bit of personal domain name vanity. Why have your tweets full of bit.ly or goo.gl URLs when you could have your own domain name showing up – even if clicking the link doesn't take your tweet readers to your Web site.

And to finish, just because I find it very funny, here's that video of the back leg bone thief:

Writer River – the collaborative list of links to tech-writer-related online content – has changed slightly.

You can still post links to it directly, but you can now also syndicate content to the site from your blog. This means that you can produce you own list of things you’ve read and found interesting, publish it on your blog with an accompanying RSS feed, and have those items also appear, alongside everyone else’s, on Writer River.

Why would you want to do this?

There are a few reasons why this might be something you’d like to do.

1. If you maintain a website, you might be interested in getting as many people to visit your site as possible. One way this can happen is through trackbacks (automatically generated links that appear on blog B when a post is published on blog A that includes a link to blog B). If you add items of interest directly into Writer River, the trackbacks will lead back to Writer River. If you list those interesting posts and articles on your own site, the trackbacks will lead people back to your site and boost your numbers.
2. You might just like to have your own list of the things that have interested you, all listed together on your own site. Now, because of the way Writer River is designed, you could already do this by going to the page on Writer River that lists just your posts (you can find this by clicking your name in the list of authors), and grabbing the RSS for this page (e.g. my old list is currently still there at: http://writerriver.com/author/alistair-christie/feed/) and then taking that and publishing it on your own site. RSS stands for Really Simple Syndication and that’s what it’s for: allowing content to be published in one place and syndicated out for publication in lots of other places. However, if you’re a blogger, it seems weird to post items somewhere else in order to get them to appear on your own site. Better to post to your own site and have that content crop up elsewhere (according to whoever wants to pick up your feeds and use them as they were intended to be used).
3. Doing things this way round means that when a Writer River reader, like myself, clicks an item link at Writer River they are transported to your blog, which gives them the context of who it was who found the thing interesting. The down-side to this is that there’s an extra step in there that people won’t be expecting. You’d expect, when you see something that looks interesting and click you click it, to be taken straight there to the original source, but instead you’re taken to the full version of the reference (e.g. a mini post on itauthor.com) and from there you can click through to the original article. This is a bit disorientating and confusing at first, but it does at least reveal something more about the poster, and it clicks up another hit on the poster’s website.

So how do you do this?

The very, very short and sweet version is:

• Produce an RSS feed containing your reading list
• Tell Tom Johnson (the creator/maintainer of Writer River) about the RSS feed. 
  Tom then adds it to the list of feeds that are aggregated into Writer River.

There are a number of ways you could create your RSS feed. The way I’m going to explain uses categories in a WordPress blog. You could equally well use WordPress tags. You could use another type of blog. You could even manually update a hand-crafted RSS file, if you felt so inclined.  

The basic steps for the WordPress method are as follows:

1. Within your WordPress blog, create a new blog category that you’ll give to all your reading list posts. You can call it anything you like (e.g. “What I’m Reading”) – all it’s used for is to collect together mini blog posts that are simply references to something interesting you’ve been reading. You need to be able to separate these out from your other blog posts so that you can display them separately as a reading list.
2. Alter your home page so that these reading list mini posts don’t appear there. They’re going to be listed separately so there’s no point putting them on the Home Page, and if you don’t exclude them you run the risk of alerting subscribers several times about the same reading list entry.  
         
   To exclude a category from your home page, first get the ID number of the category you want to exclude. You can find the category number by going to your Categories list in the WordPress admin pages, clicking the link for a category and looking in your browser’s address bar for the ID number. Make a note of this.
        
   Next, go to Appearance > Editor and open functions.php. Add the following function (changing 34 in this example to the number of the category you want to exclude):
        
   

   function exclude_category($query) {         if ( $query->is_home ) {              $query->set('cat', '-34');         }        return $query;     }     add_filter('pre_get_posts', 'exclude_category');

   (Source: http://blogmum.com/2009/04/how-to-exclude-categories-from-the-home-page-of-your-wordpress-blog/.)

3. Add a couple of posts about things you’ve read recently.
4. Go to the “What I’m Reading” category on your blog (assuming that’s what you called it) and check you’ve got an RSS feed that lists the items you posted. You do this in WordPress by adding /feed to the end of the URL for the category (for example, my feed is http://www.itauthor.com/category/what-i-am-reading/feed).
5. Now, with the web page version of the RSS feed displayed in your browser, right-click the page and choose to view the source code for the page. 
         
   This will show you the XML that produced this page. This is the RSS feed itself and, after a collection of details about the feed itself, it should contain a couple of item elements: one for each item on your list. Each item element should look something like this: 
          

   <item>
      <title>Fascinating Blog Post</title>
      <link>http://www.yourblog.com/path/to/this/particular/blog/entry/</link>
      <comments>http://www.yourblog.com/path/to/this/particular/blog/entry’s/#comments</comments>
   
      <pubDate>Sun, 28 Jun 2009 19:31:13 +0000</pubDate>
      <dc:creator>Your Name</dc:creator>
      <category><![CDATA[What I'm Reading]]></category>
   
      <guid isPermaLink="false">http://www.yourblog.com/?p=3926</guid>
      <description><![CDATA[Fascinating Blog Post. This is a fantastic read. You’d be mad not to
   have a look [...]]]></description>
      <content:encoded><![CDATA[<p><a href="http://www.fascinatingblog.com/path/to/source/article">
   
   Fascinating Blog Post</a>. This is a fantastic read. You’d be mad not to have a look.</p>
   ]]></content:encoded>
      <wfw:commentRss>http://www.yourblog.com/path/to/this/particular/blog/entry’s/feed/</wfw:commentRss>
   
      <slash:comments>0</slash:comments>
   </item>

            
   The bits that get used on Writer River are the title, which is linked to the URL in the link element (i.e. the individual blog post on your blog) and the the contents of the description element (which are an abbreviated, plain text, version of your post). 
        
   The bit you’ll display on your blog are the contents of the content element, which contain the full text of your post, including any links, images, etc.

6. Copy your feed URL (e.g. http://www.itauthor.com/category/what-i-am-reading/feed) and email it to Tom Johnson asking him to add it to the Yahoo Pipes syndication aggregation he’s got set up for Writer River.

Tips

• Make sure the style of your reading list content doesn’t stray too far from the sort of thing that’s already on Writer River. 
• Don’t post too many items. Writer River works best as a “Best Of” site – so stick to posting just the most interesting things you’ve read. Please don’t just post everything you’ve read!

Other things you can do

• Create a feed on Feedburner.

        
  This isn’t necessary, but it adds a useful element of indirection. This might be useful down the line if your feed URL changes because if you promote the Feedburner version of the feed, rather than the original, then you just need to change the URL details at Feedburner and your syndications won’t break.

        
  It also provides a way of gathering statistics about how your feed is used.

• Create a category template for your “What I’m Reading” category.

     
  Again, why would you want to do this. Well if you don’t then your reading list, with it’s mini blog posts, will have to look like all your other categories. However (following Tom Johnson’s lead on this) I wanted to list my reading list without titles. The reason for this is that the title is linked to the individual page and I don’t want people going from the reading list to an individual reading list post: that would be really annoying for readers. What I wanted was that, if someone is interested in an item, they click the link and go to the web page I’m referring to. So, to do this, you need to be able to use a different template to display just this template.

         
  To create a category template, create a PHP file in your theme’s directory called category-n.php where n is the number of your “What I’m Reading” category. As noted above, you find the category number by going to your Categories page, clicking the link for a category and looking in your browser’s address bar for the ID number.

• Make a template that’s used to display single pages where the post belongs to a specific category.

           
  If you’ve got a category-specific template for your reading list you’ll probably want a category-specific single-post template, so that, when someone clicks a link on Writer River and arrives at your site, the post is displayed in a similar manner to the way it looks on the reading list. In my case this means it gets displayed without the title. And (another Tom Johnson tip) you can also put a note at the top of the page to explain that this is one item from a whole reading list of other links, and include a prominent link to that list (i.e. to your “What I’m Reading” category).

          
  To create a single-post template for a specific category, first add the following to your functions.php file, anywhere between the opening and closing PHP tags:

        

  add_filter('single_template', create_function('$t', 'foreach( (array) get_the_category() as $cat ) {             if ( file_exists(TEMPLATEPATH . "/single-{$cat->term_id}.php") )            return TEMPLATEPATH . "/single-{$cat->term_id}.php";     }    return $t;' ));

  (Source: http://www.nathanrice.net/blog/wordpress-single-post-templates/.)

  Now you can create a post template by creating a PHP file in your theme’s directory called single-n.php where n is the number of your “What I’m Reading” category. For example, the /itauthor.com/wp-content/themes/blueprint directory on my server now contains the templates: category-34.php and single-34.php.

        
  Once you’ve created these template files, you can edit them from the Edit Themes page of your WordPress admin pages. 

• Modify the Press This bookmarklet for your blog
  
        
  The Press This bookmarklet is a browser bookmark with added functionality provided by some nifty JavaScript. It makes it very easy to create a post relating to a web page you’re currently looking at – so it’s perfect for adding items to your reading list. 
         
  

         
  You get a Press This bookmark by going to the Tools page for your WordPress blog and dragging the Press This link into your bookmarks list.

        
  Note: The bookmarklet uses the inbuilt post editor for WordPress, so if you’ve installed a plugin that swaps the normal post editing buttons for something funkier (e.g. I was using Dean's FCKEditor For WordPress) you’ll have to disable this plugin or you’ll just get a blank text box in the little Press This window. I also had to disable the Disable wpautop plugin, which disables WordPress’s irritating automatic paragraph formatting, because with this enabled any line breaks in the Press This editor disappeared when I published the post.

        
  For me, posting to my reading list is the only thing I use my Press This bookmarklet for, so I want my “What I’m Reading” category to be selected by default.

         
   
        

  To add a default category (i.e. already checked) at the top of the Categories list in the Press This window, you need to edit the wp-admin/press-this.php file. Unfortunately you’re going to have to redo this change every time you upgrade WordPress. The stuff you need to add is in red. Change the category number and name as appropriate for your blog.

        

  <div id="categorydiv" class="stuffbox">       <h2><?php _e('Categories') ?></h2>     <div class="inside">              <div id="categories-all" class="ui-tabs-panel">          <ul id="categorychecklist" class="list:category categorychecklist form-no-clear">           <!-- Edit this line to suit your setup - my default category number is "34" and its name is "What I'm   Reading" –>           <li id='category-34'><label class="selectit"><input value="34" type="checkbox" checked    name="post_category[]" id="in-category-34"/>What I'm Reading</label></li>             <?php wp_category_checklist($post->ID, false, false, $popular_ids) ?>            </ul>          </div>     

  Kudos to johnke at the WordPress.org forums for suggesting this (http://wordpress.org/support/topic/216525).

• Create a twitterfeed for your reading list.

         
  Twitterfeed is an online resource that allows you to push the contents of an RSS feed (like a feed from a blog) into your twitter account, so that your blog posts (or the first few words plus an abbreviated URL) get sent out to your twitter followers. If you also syndicate your tweets to Facebook it means that your new reading list items will also show up as Facebook status messages.

         
  You can set Twitterfeed to check the RSS feed for your reading list once an hour. You can use the Advanced settings to add a prefix to these posts (I use “I've been reading: “), and I’d recommend setting it to include the “title only” rather than “title & description” because once you’ve added the prefix and the link to your post there isn’t much room for any more than two or three words from your post content.

         
   

Respect

I’ve already mentioned him a few times and, at the risk of embarrassing him, I must doff my cap once again to Tom Johnson who created Writer River and came up with most of the ideas I’ve fleshed out in this post.

Want to find out more? Go read these posts on Tom’s blog:

• "What I'm Reading": A New Feature on My Site and a Tweak of Writer River
• I Need Your Human Aggregated Content

Respect

I’ve already mentioned him a few times and, at the risk of embarrassing him, I must doff my cap once again to Tom Johnson who created Writer River and came up with most of the ideas I’ve fleshed out in this post.

Want to find out more? Go read these posts on Tom’s blog:

• "What I'm Reading": A New Feature on My Site and a Tweak of Writer River
• I Need Your Human Aggregated Content

http://www.itauthor.com/articles/recently-posted-by-other-bloggers/

I realise this kind of entices people to leave my site, but so what? It’s kind of dumb to expect people to hang around your web site for very long. For example, I regularly go to the BBC web site. I never stay there very long. I don’t browse through lots of pages. I’ll look at two or three pages and then I’m off somewhere else. But because it’s well designed and has good content I go back again.

But I’ve been thinking about my web site and how I’ve pretty much just blindly gone along with the standard WordPress format, with a panel down one side with links to category and month archives. I sometimes click a month link – typically back a month or two but rarely further. But I never click a category link and I suspect nobody else does either. I don’t use tagging, but if I did I think it would be the same thing. Why would anyone browse through a tag or category archive?

The navigation that I think is useful is the collection of potentially similar posts at the end of each post. Sometimes this throws up something that I’d forgotten I wrote about and I’ll click to have a look. I think other people reading my posts might also click to have a look at an older “similar” post if it sounds interesting.

So I think I’m going to take the category links out of the side bar. I’ll probably leave the calendar links in there for now, but I’m tempted to get rid of that side bar altogether and go for a really minimalist look. On the other hand I was also thinking that the site looks quite austere. So I’m also a little tempted to try and make it look less utilitarian.

But I’d be interested to know if anyone finds tag/category lists useful on a blog. Aside from never using them on my own blog, I can’t remember ever having used them when I’m reading other people’s blogs either.

But these days blogs have long since evolved away from something that most closely resembled an online diary, into something that’s more like a multifunctional Web site. Nevertheless, blogs still make brilliant diaries. The trouble is, of course, that by default everyone gets to see what you write in your blog. One way round this is to download WordPress and install it locally on a PC, so that only someone logged onto that machine can access it. But that involves using something like XAMPP to install Apache Web server and MySQL.

A much easier way is just to sign up for a free blog at wordpress.com and then make particular posts, or the whole thing, private – as described in this WordPress video:

This happened while I had the WordPress Akismet plugin turned on. Akismet certainly picks up most comment spam, but a lot still gets through. After I turned on moderation I then had to do a weekly cleanup of all the spam sitting waiting for approval.

But since installing the WP-SpamFree plugin about six months ago, I’ve had no comment spam at all. None. Not one little bit of it.

So now, but with a certain amount of hesitation, I’m turning off moderation again. This means that anything you comment here, or on any other post, will be immediately visible to anyone who visits the site.

So go on. Give it a go. Leave me a comment. Go on!

Awh, go on, go on, go on, go on!

I’ve never found an altogether satisfactory way to include nicely formatted and syntax highlighted code in WordPress. I’ve tried various methods but always ended up settling on something so-so after spending a while on it and eventually deciding I was wasting my time.

But this time round I tried again and found Chris Cagle’s Definitive Guide on WordPress Syntax Highligher Plugins.

In the spirit of “if  it’s good enough for him, it’s good enough for me” I decided to use the same plugin Chris settled on: Chili. You can see the results for yourself. However, although I’m fairly pleased with the end product, the process for getting there (using Live Writer) isn’t ideal. I’m documenting it here because, fortunately, I don’t often use code snippets, so this will serve to remind me how I did it:

1. Copy the code you want to use.
2. Paste it into Postable (an online processor for making code browser friendly):
   http://www.elliotswan.com/postable/
3. Click make it friendly.
4. Copy the browser-friendly code.
5. In Live Writer, go to Source view.

6. Create nested elements as follows:

   <p><pre><code class="html"></code></pre></p>

7. Change "html" for "php" or "c#" or "mysql" or "javascript" or "css". There are a few other supported languages listed on the Chili site. 

8. Paste the code into the middle, between the code tags. For example:

   <p><pre>
     <code class="html">&lt;p&gt;&lt;pre&gt;&lt;code class=&quot;html&quot;&gt;&lt;/code&gt;&lt;/pre&gt;&lt;/p&gt;</code>
   </pre></p>

9. Publish your post.

The problem

The feed for the ITauthor Podcast (which is just a feed for a WordPress category called "Podcast") was coming out as ITauthor >> Podcast. It annoyed me that the podcast showed up in iTunes with those two greater than signs, or double angle brackets, chevrons, whatever you want to call them. In fact, although it looks like two characters it's actually the double right angle quote character:
»   which is typically written in HTML using the character code » and in XML using the code »

My solution

The way I got rid of this may not be the easiest or best way. I search the WordPress Codex without finding out how to fix this, so in the end this is just the way I figured out how to do it. It may not be good. It may not be nice. It's certainly not pretty. All I'm saying is that it works for me.

So what I wanted to do was have a feed for my podcast category page that had the feed title "The ITauthor Podcast".

There were two tasks:

1. Remove the angle quote completely in the cast of the feed for my "podcast" category page, but for other feeds change this character to a colon so that, for other categories, instead of "ITauthor >> Category", I'd have "ITauthor: Category".
2. For the "podcast" category page, add "The" at the start of the feed title.

Here's what you do:

1. Go into your wp-includes directory and take a backup copy of feed.php and feed-rss2.php.

   Now I've moved to Dreamhost, I have the luxury of an ssh connection and a command line, which makes this very quick and easy (provided you're happy working on the command line).

2. Open the feed.php file (e.g. using vi or some other plain text editor).

3. Change:

   function wp_title_rss($sep = '»') {
           echo apply_filters('wp_title_rss', get_wp_title_rss($sep));
   }

   to:

   function wp_title_rss($sep = '»') {
        if (in_category('26')) {
             echo ' ' . trim( get_wp_title_rss('') );
        } elseif ( is_archive() | is_category() | is_tag() | is_date() ) {
             echo ': ' . trim( get_wp_title_rss('') );
        }
   }

   What this does is as follows.

   For category 26 (which happens to be the number of my podcast category), it gets the category title. The empty string for an argument means that it prepends a space followed by nothing (because I don’t want a separator character), followed by a space. The second space is added because the function expects a separator character in the middle. The trim function trims away the two spaces, but we add back a single space (that's the first thing that gets echoed) otherwise the result would be "ITauthorPodcast". 

   For all other categories, and various other types of pages we do exactly the same, but prepend a colon followed by a space, rather than just a space. This gives us "ITauthor: Category".

   For anything else (e.g. the Home Page) this function does nothing, which results in a feed title of just "ITauthor".

4. Save your changes to this file.
5. Open feed-rss2.php.

   This is the file that creates the RSS2 feed (the default format for feeds). It's here that the category title (as generated by wp_title_rss() in the feed.php file) is appended to the blog title. So it's here we can prepend "The " to give us "The ITauthor Podcast.

6. Insert an if clause into the part of the script that prints the title, so that if this is the feed for the podcast category (ID 26 on my system) it adds "The " in front of the title generated by bloginfo_rss('name') and wp_title_rss().

7. Change:

   <title><?php bloginfo_rss('name'); wp_title_rss(); ?></title>

   to:

   <title><?php if (in_category('26')) echo 'The '; bloginfo_rss('name'); wp_title_rss(); ?></title>

8. Save the file.

All I wanted to do was change the title of one of the RSS feeds for this blog. Sounds simple enough and, to be honest it wasn’t rocket science, but it did involve digging around among the foundations of my WordPress blog. And this made me think: WordPress is a bit like Venice. What you see (gliding towards the Lido on a water taxi or arriving at the home page of a WordPress blog) is usually attractive. There are things that are not so great (crumbling masonry or a cumbersome list of categories in a side panel), but on the whole it’s a good experience.

However, WordPress, like Venice, is not built on strong foundations. It’s built on the software version of rotting wood.

WordPress is a great example of a software application that has grown organically and has ended up one great big mess. Granted it’s a mess that people (millions of them) manage to do wonderful things with. But if you sat down to design a highly configurable, mass market blogging application you would not build it like WordPress. For starters (sorry PHP lovers out there), you would not use PHP. WordPress’s collection of PHP on PHP on PHP, distributed through a collection of files that is only loosely architected (I’m trying to be kind here), is the equivalent of Venice’s wooden posts driven into the mud of a lagoon, long, long ago. It seemed like a good way to build a town at the time. It made perfect sense back then. But the years went by and Venice grew: outwards and upwards. And down the ages the architects of Venice have wished they had something better to support their beautiful, lovingly designed constructions than lots and lots of rotting posts.

But to answer my own question: Is WordPress the Venice of blogging software? No, it’s not. And the reason it’s not is that, unlike WordPress, Venice will outlive us all. It’s too important and we love it too much. The posts are being replaced. Huge amounts of effort and money have gone into solving the problem of sinking buildings. Venice’s future is secure (except maybe for the effects of climate change – but that’s another story).

WordPress on the other hand, I believe, has trouble ahead. It’s massively popular right now. But because of the mess at its core, there’s an opportunity for someone to come along with something that provides everything WordPress does but is much easier to use, is easier to configure into a wider variety of looks and styles, and is easier to create plugins for. And because all your data is in a MySQL database, it wouldn’t be difficult to make it very easy to migrate blogs over to a new system.

At the moment WordPress is king of the hill. I, like millions of others, enjoy using it from day to day to write up our thoughts (like these) and inflict them on the world (or some small part of it). Mind you, I think the things I like most about posting to my WordPress blog are mostly features of Microsoft’s Live Writer application, which I use for writing posts, rather than features of WordPress itself.

So this king better watch out. I wouldn’t be at all surprised if, before too long, WordPress ends up the ruler of a small and crumbling city state whose days of influence are long behind it.

Bill Rawlinson’s feedList plugin for WordPress allows you to do precisely that.

As a simple example of what it can do. Here’s my personal Writer River feed and my twitter feed grabbed and stuck together on a blog page:

http://www.itauthor.com/syndicated-posts/

Little did I know that all that while my Web site was unreachable. It wasn't down as such, it was just that before going away I'd changed the home option in the wp_options table in phpMyAdmin to www.itauthor.com instead of www.itauthor.com. This didn't have any bad effect for me immediately - presumably because of some page caching going on either in WordPress or, more likely, my browser - but this change caused www.itauthor.com to be redirected to www.itauthor.com which is already redirected to www.itauthor.com, with the effect that nothing ever got resolved and a page was never displayed.

The itauthor.eu domain is just a temporary home until I rehome my .com domain. Anyway, sorry if you tried to visit the site last week. Mind you the number of emails I received about the site being unavailable (none) suggests that nobody minded too much (or noticed!) that the site took a holiday when I did.

The only thing I'd say is that I'm not sure I trust the backups the plugin makes. I had a look in the zip file and it didn't have a wp-content directory. So I'll continue to do manual backups, as follows:

1. Using FileZilla, download the whole WordPress directory (in my case this is the Web root directory, called web), with the exception of a couple of things that live within the wp-content directory:

• advanced-cache.php
  This seems to be a link that FileZilla doesn't understand and treats like a directory, but drops the connection if you try to download it.
• cache
  This directory won't reveal its contents and, again, the connection drops if you try to download it.

2. Backup your data using phpMyAdmin (see this tutorial, or this one) and save the SQL somewhere safe.

That said, I'm very impressed with this upgrade plugin. I'd been resisting upgrading because I was sure it would cause me pain. However, everything seems to be working, even PodPress.

The following video (there are 2 parts to it) is very slow-paced but it does go through the process very thoroughly. It also shows how to use the WordPress Database Backup plugin, which I'm just about to try out.

I don't like to put myself through the pain of upgrading every time there's a new WordPress version, but I'm quite far behind now and there's a particular plugin I want to use which won't work under 2.2.

So, before upgrading live, I thought I'd migrate everything onto my laptop, do an upgrade there and see how I get on. If it just doesn't work, or breaks PodPress, for example, then I won't bother trying it on my live site.

Theorectically, migrating should just be a case of:

• Download all the files using FTP.
• Export the data from MySQL.
• Put Apache and MySQL on my laptop using XAMPP.
• Put all the downloaded files into the web directory hierarchy.
• Create a new database on the laptop with the same name and user as on my web host.
• Import the data into this new database.

That's basically what you do but, as ever with this kind of thing, it's the fiddly little extra things you have to do to make it work that always trip you up and take forever to figure out.

The best directions (though neither of them complete) for doing this task were:

http://mark-kirby.co.uk/2008/how-to-create-a-local-copy-of-a-wordpress-site/

and

http://www.aniquito.com/2008/08/12/how-to-move-wordpress-from-hosting-site-to-local-computer-or-to-a-new-domain/

The latter is good because it has some search/replace commands that you can run in phpMyAdmin to replace full URLs with relative paths.

There's also:

http://weblogtoolscollection.com/archives/2008/01/03/install-wordpress-locally-part-2-of-2/

Which doesn't add much, but contains some information about large SQL files and phpMyAdmin.

I mainly used the first of those pages. However, additional points to note before following these instructions:

1. Disable all plugins before exporting the data and downloading the files (you can turn them back on later, once everything else is working).
2. Make sure you've change the siteurl and home options in phpMyAdmin. After changing them search again and make sure they really have got the localhost value in them now.
3. Comment out the WP_CACHE line in the wp-config.php file:

   <?php
   // ** MySQL settings ** //
   //define('WP_CACHE', true); //Added by WP-Cache Manager
   define('DB_NAME', 'mydb');    // The name of the database
   define('DB_USER', 'mydbuser');     // Your MySQL username
   define('DB_PASSWORD', 'fiendishlycomplicatedpassword'); // ...and password
   define('DB_HOST', 'localhost');    // 99% chance you won't need to change this value
   define('DB_CHARSET', 'utf8');
   define('DB_COLLATE', '');

   For a long, long while, when I browsed to the local site I got redirected to the Web site home page. I puzzled with this for hours and eventually after trying lots of different things, it went away. I'm not quite sure how I solved it. I think it was something to do with caching, so make sure you comment out the line above if you use the wp-cache plugin.

4. Make sure the DB_HOST in this file is localhost. It should be anyway, but mine wasn't.
5. If you want to use permalinks, you will need to make a change inside Apache's httpd.conf file.

   In my case this is located within this directory:
   D:\programs\xampp\apache\conf

   Open that in a text editor. Use the search facility in the editor to find "rewrite". The line you need looks like this:

   #LoadModule rewrite_module modules/mod_rewrite.so

   You need to take away the hash sign so it looks like this:

   LoadModule rewrite_module modules/mod_rewrite.so

   Save the file.

6. Make sure the .htaccess file in the root of your wordpress directory (in my case this directory is called itauthor) contains the following:

   # BEGIN WordPress
   <IfModule mod_rewrite.c>
   RewriteEngine On
   RewriteBase /
   RewriteCond %{REQUEST_FILENAME} !-f
   RewriteCond %{REQUEST_FILENAME} !-d
   RewriteRule . /itauthor/index.php [L]
   </IfModule>

   Note the name of the wordpress directory in the RewriteRule. You may need to change:

   RewriteRule . /index.php [L]

   to:

   RewriteRule . /yourwordpressdirectory/index.php [L]

   If the WordPress files aren't in the root of your web server.

7. Now restart the apache2 service in the Windows services dialog box.
8. Live Writer likes to use URLs, so if you use it to post to your blog you'll probably have to spend a while carefully searching/replacing URLs to relative paths. For me, I wasn't too bothered as I'm just going to use this as a sandbox for checking the effects of running through the upgrade process.

All I wanted to do was to make comments, and the comments form, show up at the bottom of each post, no matter what type of page it appeared on. By default, comments in WordPress only show up if you display a post on a page of its own. I know a lot of people who visit this site click the Podcasts tab. This displays posts in a category view, which means the comments form is not displayed.

WordPress, like MovableType, demands an extremely high degree of patience of anyone who wants to configure their own site. WordPress isn't as bad as MT in this respect (hence my decision a few years ago to make the switch across to WordPress), but it can still sap hour upon hour of your free time to make what you might expect would be the simplest of changes.

Given the popularity of blogging, there is real scope for someone to come in here and steal a market away from WordPress. I wonder what Google and Microsoft have up their sleeves? Microsoft already have the best tool for writing blog posts (Live Writer), I wouldn't be surprised if they had a WordPress killer among their plans for Live.

So for now, unless you're reading this on the front page or on a single-post page, if you want to comment on this post you're going to have to click on the heading above to display it singly and then scroll down to the comments form.

One annoying thing about this editor is that it removes your line breaks. For example, suppose you're in a list and you do a couple of shift-returns to add a new paragraph within the same point in the list (i.e. an indented paragraph that's not numbered). WordPress will remove the <br/> tags and close up the space. This might not happen straight away but it will happen if you reopen the post in the editor.

Even if you go into code view and manually type in <br/>, WordPress will take them out. It seems to think it knows better. So you have to prevent it recognising the <br/> tags by entering them manually in code view as:

<br align="left" />

At around the same time some of my domain names came up for renewal and when I was renewing itauthor.eu with Euro-Reg.com I decided, on the spur of the moment, to buy a hosting package and move everything onto their servers. You may have noticed that, at the time of writing at any rate, the URL for the site is .eu rather than the usual .com.

Things may change because I subsequently realised that Euro-Reg only allows me 500MB of space, which isn't enough to put everything up that was originally on my site (so you may notice some missing pics etc.). I think I might move everything over to Dreamhost, but for the time being it's working OK where it is and it means I can save a bit of electricity by turning off my Web server.

If you've visited before you may notice that the site runs a little faster than before - not a huge amount though, which is a little disappointing given that I got a tiny upload speed from Sky.

The downside of moving things to Euro-Reg is that I have no command line access and I have to move files via FTP. I'd forgotten how slow and arduous FTP is. I'd got so used to using WinSCP for putting stuff on my Web server. I'd kind of consigned FTP to history and thought I'd never use it again. Funny how things turn around.

However, the big gain from moving my WordPress blog onto someone else's server is that - for reasons unknown - the "Next/Previous" links in WordPress that had never appeared before now work just fine. I'd spent ages in WordPress forums trying to get these links to show up, without success, so it's nice that the site is now much more accessible and you can get more than 5 search results now.

For example, say you have a table containing Web links to a server on your network and the server name gets changed or, more likely, the Web pages get moved. You might have links such as http://oldwebserver/importantstuff/index.html that should now be http://newwebserver/importantstuff/index.html.  To correct these with an SQL statement, use:

UPDATE table_name SET field_name = REPLACE (
field_name,
"old text",
"new text");

For example:

UPDATE mytable SET url = REPLACE (
url,
"http://oldwebserver/importantstuff/",
"http://newwebserver/importantstuff/");

Note: The reason I didn't just replace "oldwebserver" with "newwebserver" is as a safeguard against changing things I didn't want changed. You've got to be pretty careful about what you do, and I would strongly advise that you always back up the database before making any changes like this.

WordPress expert Lorelle has some good advice about doing this kind of thing on your WordPress tables at:

http://lorelle.wordpress.com/2005/12/01/search-and-replace-in-wordpress-mysql-database/

Live Writer large image problem

Live Writer large image problem #2

Basically, if I tried including an image above about 30KB I got the following error:

I got the problem again, this evening, trying to upload my previous post, which included several decent-sized images, and I was determined to get to the bottom of the problem. It turns out the problem is with the PHP memory limit. The fix is to adjust the memory limit for PHP for your blog. To do this, simply add the following line to the .htaccess file in the root directory of your blog (in my case this directory is /var/www/html/wordpress on my Web server):

php_value memory_limit 16M

You don't need to restart Apache or anything. Just change the .htaccess file and click Publish in Live Writer and it should work fine now. I can't promise it'll work in all cases, but it worked for me.

So, when I do find some time (hopefully some weekend soon), I'm going to upgrade WordPress and choose an off-the-shelf theme for 2.3 that looks like it might be supported through future upgrades and stick with that, so that at least everything works, even if it doesn't work, or look, like I'd like it to.

At the moment I'm on WordPress 2.1.3 so I'm just hoping the upgrade to 2.3 is fairly painless. We'll see.

In the list of Latest Essays on Joshua's site I noticed one called A Guide to Writing Well, which isn't aimed at technical writing but, nevertheless, lists lots of useful principles that we should keep in mind while writing technical documents. Here's a few:

• Omit needless words. Write simply and without clutter. Don’t add words for “style.”
• Avoid fancy words.
  “Never use a long word where a short one will do.” (George Orwell)
  “Look for all fancy wordings and get rid of them.” (Jacques Barzun)
  Examples: Assistance (help), facilitate (ease), implement (do), referred to as (called).
• After every sentence, ask yourself what the reader wants to know next.
• Use orthodox spelling, punctuation, and capitalization.
• Use active verbs. Example: “He was seen by Joe” should be “Joe saw him.”
• Keep sentences short.
  “There’s not much to be said about the period except that most writers don’t reach it soon enough.” (Zinsser)
• Remove laborious phrases. Why use “at the present time” instead of “now”?
• Use contractions when they sound natural.
• Write like a person and not like a scientist.

I've just installed a plugin that makes it much easier to find posts in your blog. The plugin is very easy to install, just follow the instructions here:

http://www.zirona.com/software/wordpress-advanced-search

1. If you're in the middle of a redesign, where you're changing the presentation features of the site without changing the content, you have to clear the cache to see the effects of changes.

   To clear the cache, go to the wp-content/cache directory and do:
   rm -rf *
   

2. If, like me, you have some fairly static pages that contain dynamic content in a sidebar (e.g. an About page that has a list of recent posts in the sidebar), the sidebar content will not change because wp-cache makes sure the page from the cache gets served up unless something actually changes in the content.

   Effectively this means doing a manual cache clear every time you post if you want your static pages to be regenerated.

Turning off wp-cache

WordPress leads you to believe you can turn off wp-cache by deactivating it in the Plugin Management page of the admin interface. Unfortunately it's not as easy as that.

What you need to do is go into your wp-config.php file and delete or comment out (by adding // at the start of the line) the following line:

define('WP_CACHE', true)

If you want to completely remove wp-cache, rather than just temporarily disabling it while you do some redesign work:

1. Delete the wp-cache directory within the wp-content/plugins directory.
2. Delete the line from the wp-config.php file, as above.
3. Delete the wp-cache-config.php and advanced-cache.php files from the wp-content directory.

Where wp-cache comes into its own is if you have a standard blog that gets a lot of hits. wp-cache stores ready-made HTML pages that can be served straight up to your visitors without having to use PHP to generate SQL queries to the MySQL database and build HTML dynamically.

http://www.tamba2.org.uk/wordpress/backup/

The tables are backed up as one SQL file, which you can save locally.

http://codex.wordpress.org/Backing_Up_Your_Database

and

http://codex.wordpress.org/WordPress_Backups#Backing_Up_Your_WordPress_Site

from the WordPress Codex, from which the first link I posted took the information.

http://blog.elinc.ca/rod/

Yes, that's what it looks like!

http://mnm.uib.es/gallir/wp-cache-2

Installing this plug-in has taken my time-to-load from 25+ seconds to just 3”“4 seconds, which compares very nicely with the performance I was getting from Plone. It does this by caching pages on the server so that preprepared HTML pages are served up to your browser, rather than dynamically generating pages on request, using PHP. You can check that you're looking at a cached page by viewing the source of the page. If it's a cached page, the last line of the source will be:

!-- Cached page served by WP-Cache --

Two installation methods are documented. I chose the manual method, just because that's the one everyone seems to talk about, so I reckonned I'd stick with tried-and-tested. Installation is a little more difficult than installing most WordPress plug-ins. But if you're used to ploytering around on a UNIX command line on your Web server you'll find it easy enough.

There are 2 little problems with this plug-in that you must watch out for. Both of these cause your site to serve up completely blank pages after you install the wp-cache plug-in.

1. If you're server uses PHP 5 you need to edit /wp-content/plugins/wp-cache/wp-cache-phase2.php, locate ob_end_clean(); in this file and replace it with ob_end_flush();That is, you need to change "clean" to "flush".
2. Don't use the documented command for creating the symbolic link. I copied and pasted it onto the command line and it created a duff symlink. The correct command, from the wp-content directory is:ln -s plugins/wp-cache/wp-cache-phase1.php advanced-cache.php

Once you've sorted this out and added the define statement to your wp-config.php, you can go to the Options section of the WordPress admin and you'll see a WP-Cache tab. Go in there and click the button to enable the cache. Having done that everything should work fine and your pages will fly.

Incidentally, I just did a php -v and realised I'm still on PHP 4.3, so I guess I should go and change back that ob_end_clean(); function, but it seems to work fine as it is, so maybe I won't bother.