Create a Category Page for one Specific Category and Exclude this Category from the Main Page in WordPress

This blog serves as my digital notebook for more than eight years and I use to to collect all sorts of things, that I think are worth storing and sharing. Mainly, I blog about tiny technical bits, but recently I also started to write about my life here in Innsbruck, where I try to discover what this small city and its surroundings has to offer. The technical articles are written in English, as naturally the majority of visitors understands this language. The local posts are in German for the same reason. My intention was to separate this two topics in the blog and lot let the posts create any clutter between languages.

Child Themes

When tinkering with the code of your WordPress blog, it is strongly recommended to deploy and use a child theme. This allows to reverse changes easily and more importantly, allows to update the theme without having to re-implement your adaptions after each update. Creating a child theme is very easy and described here. In addition I would recommend using some sort of code versioning tool, such as Git.

Excluding a Category from the Main Page

WordPress offers user defined categories out of the box and category pages for each category. This model does not fit well for my blog, where I have static pages and a time series of blog posts on the main page. In order to prevent that the posts about Innsbruck show up at the main page, the category ‘Innsbruck’ needs to be excluded. We can create or modify the file functions.php in the child theme folder and add the following code.

This adds a filter which gets executed before the posts are collected. We omit all posts of the category Innsbruck, by adding a minus as prefix of the category ID. Of course you could also lookup the category ID in the administration dashboard, by hovering with your mouse over the category name and save one database query.

A Custom Page Specific for one Category

In the second step, create a new page in the dashboard. This page will contain all posts of the Innsbruck category that we will publish. Create the file page.php in your child theme folder and use the following code:

In this code snippet, we define a set of arguments, which are used for filtering the posts of the desired category. In this example, I used the id of the Innsbruck category (91) directly. We define that we want to display posts only, 5 per page. An important aspect is the pagination. When we only display posts of one category, we need to make sure that Worpress counts the pages correctly. Otherwise the page would always display the same posts, regardless how often the user clicks on the next page button. The reason is that this button uses the global paged variable, which is set correctly in the example above.

The if conditional makes sure that only the pages from the Innsbruck category are displayed. The while loop then iterates over all posts and displays them. At the bottom we can see the navigational buttons for older and newer posts of the Innsbruck category.

Continue reading


Build a SMS notification service

Automated event notification is useful for many applications. In a recent blog post, I presented a solution how to use Twitter for notifying about server events. In this post you learn how to achieve the same by using an SMS API. The rationale behind an SMS service is its robust availability.

There exist plenty of different services, I chose ClockworkSMS, as it provides a simple interface for several programming languages, plugins and a prepaid billing service. Once you registered and bought some credit, you can start using the API. Integrating the service into your applications is easy, as there does not only exist libraries for several languages (C#, Java, Node.js, PHP, Python, Ruby), but also HTTP requests are supported. Thus you can also send SMS text messages from command line. When you execute the following command in bash, a SMS message will be sent to the number provided containing the text, no magic involved.


curl https://api.clockworksms.com/http/send.aspx -d "key=YOUR-API-KEY&to=4361122334455&content=Hello+World"

Simply insert your personal API key, provide the receiver’s phone number and of course the content. The text of the message needs to be URL encoded. In order to advance the notification service, you can provide a call back URL which is called whenever a SMS message was sent.

The Delivery Receipts

The Delivery Receipts

Clockworksms provides HTTP GET, form posts and XML (bulk) POST requests which are issued against the URL provided. The service appends message metadata such as an id, a status code an the number of the receiver to the request. This information can be processed in custom applications which react on the data. In the easiest scenario, you can setup a virtual host and process the log files. Each sent SMS message will result to a request from the Clockworksms servers to your server and thus create a log record with the metadata of the sms message.


89.111.58.18 - - [30/Oct/2014:18:21:32 +0100] "GET /smsnotification.htm?msg_id=VE_282415671&status=DELIVRD&detail=0&to=4361122334455 HTTP/1.1" 200 330

The ClockworkSMS service also provides some plugins for WordPress, Nagios and other applications.

Continue reading


Configuring a catch-all VirtualHost for Apache2

Recently I noticed an interesting behavior of my Apache2 setup when I introduced a new VirtualHost and enabled the site. Usually I store the configuration for each host in an individual file, because it is much easier to maintain. The main page that I serve is this WordPress blog, having the ServerName blog.stefanproell.at and several aliases. The configuration files have the same name as the ServerName of the VirtualHost they configure. When a request comes in, Apache2 passes it to the first VirtualHost having a server name or alias fitting to the request. If no atch can be found, the request is handled by the first VirtualHost that can be found. This is obviously the one having the config file with the alphabetically first file name. Without noticing, the first config file to be read was exactly this blog. Other services and hosts have their configuration files names starting with letter being later in the alphabet than the blog, hence they were read later. So every unknown request was handed over to the blog, which is nice as I want to have it like this anyway.

But when I introduced a new host having a filename say 123.stefanproell.at for its config file, each request not to be answered by any host was handled by the new host 123.stefanproell.at. So far, this is no surprise at all, but then it got weird. As suggested by a lot of different sources and the official documentation, I declared a default page, having a filename with an alphabetically even lower letter: 000-default.

Also the directive especially designed for this porpose of having a wildcard VirtualHost did not work as expected. This directive

<VirtualHost _default_:*>
DocumentRoot /www/default
</VirtualHost> 

should according to the documentation

prevent any request going to the main server. A default vhost never serves a request that was sent to an address/port that is used for name-based vhosts. If the request contained an unknown or no Host: header it is always served from the primary name-based vhost (the vhost for that address/port appearing first in the configuration file).

But instead of serving an empty default page for every missed request, Apache2 stated to serve every request to the default page, although the other VirtualHosts have been defined. The same behavioralso occurred when the specific

 _default_ 

directive was not applied.

What I ended up with was a solution where I explicitly defined a dummy,catch all host having the lowest file name for its configuration. So now I have a fully configured VirtualHost having the configuration file name 000-dummy.stefanproell.at and the content

ServerAdmin webmaster@localhost
ServerName dummy.stefanproell.at
DocumentRoot /var/www/default

Do not forget to provide a DNS entry pointing to dummy.stefanproell.at as well.

Continue reading


Getting static – with WordPress

As I moved to a new server recently, I decided to convert my blog which I had during my semester abroad into static HTML, as the blog itself won’t change anyway. The easiest way to do so was to download the original blog with HTTrack completely to my local system. This tool converts dynamic links into static, relative ones, which is needed in order to move the blog to a new location. HTTrack preserves the page’s structure completely and includes all files such as images. After the download was complete, I just copied the whole, static website to its new destination. Great 🙂

Continue reading