Things I learned from personal website: setup

This entry is part 1 of 3 in the Things I learned from personal website series:

  • Part I, setup
  • Part II, seo (coming soon)
  • Part III, speed (coming soon)


What can be better than creating a small personal website after years of working with complex server-side driven systems. What I mean is that when setting up a simple static website from scratch I had more chances to get known with all the aspects of a well baked website without being diverted by server-side issues.

In this entry I will share my experience with some nice tools and services I used while setting up my personal website:

  • Jekyll to handle website generation
  • Markdown format to write blog posts
  • Disqus for commenting
  • Git to keep track of changes and to deploy
  • Github to host the website

If you're unfamiliar with one of those tools, I highly recommend you checking them out and in this entry I expect that you have done that alrady. After you completed your basic setup use this article as a helper guide for more advanced customization.


Jekyll reads configuration data from _config.yml file in site's root directory.

You are not limited to options recognized by Jekyll. You can store any data and retrieve it in any page/layout that contains YAML form matter block. For example, I can store my personal info:

  name: Sergey Lukin
  about: About me text goes here
    - name: Github
    - name: Twitter

and retrieve it in a page using Liquid syntax:

<p>My name is {{ }}</p>

<p>Social links:</p>
{% for profile in %}
    <a href="{{ profile.url }}">{{ }}</a>
{% endfor %}

The key here is site object - in this case we use it to access variables set in _config.yml

Navigation can be rendered by looping through site.pages. Additionally, you can also set a variable, let's say, navigation in every page you want to be in navigation menu and filter it in the loop. So, here is the YAML front matter for a page that should show up in navigation:

title: Page title
navigation: true

..then in navigation template loop over all pages and only print ones that have navigation variable set to true (also we can check the URLs and add additional markup if necessary):

  {% for node in site.pages %}
    {% if node.navigation == true %}
      {% assign attr = nil %}
      {% if page.url == node.url %}
        {% capture attr %}class="active"{% endcapture %}
      {% endif %}
        <a href="{{ node.url }}"{{ attr }}>{{ node.title }}</a>
    {% endif %}
  {% endfor %}

In this example a comparison is done beween page.url and node.url and if no difference found, then variable attr gets a value of class="active" which allows us to apply special styles to menu item.

#List posts

Printing a list of items inside _posts directory is as easy as:

  {% for item in site.posts limit:5 %}
      <time datetime="{{ | date: "%Y-%m-%d" }}" pubdate>{{ | date: "%B %d, %Y" }}</time>
      <a href="{{ item.url }}">{{ item.title }}</a>
  {% endfor %}

#Creating a New Post

To create a New Post I run newdraft script:

./_bin/newdraft HERE GOES POST TITLE

#GIT and Draft posts

Usually it takes a while untill I finish writing a blog post. In the meanwhile I can do changes to other parts of the website and I don't want GIT to recognize a draft post as a new object. To avoid this I added following line in .gitignore file:


So that any file in _posts directory that ends with or draft.markdown is ignored and only when I finish writing the post I remove the draft from the filename.

#Syntax highlighting

In almost every blog post of mine I insert code snippets so it's extremely important to have syntax highlighting feature included. Jekyll uses Pygments to handle this. Here is how I use it:

{% highlight html %}
..some code here..
{% endhighlight %}

Wonder how I succeeded to print Liquid syntax without it being parsed? I placed it between {% raw %} and {% endraw %} tags. For more Liquid stuff consider reading Liquid for Designers.

Please note that Liquid v2.3.0 and higher supports tag {% raw %} instead of {% literal %} as it was in v2.2.2. Besides that, I would highly recommend installing locally Jekyll and Liquid of exactly same versions as Github pages server is running (you can always check what software they are running here) and generating a website locally before deploying. That way you will definitely identify any potential problem as early as possible and it will save your time.


My commenting platform of choice is Disqus. After signing up add a little bit of configuration to _config.yml:

  enabled: true

and insert this piece of HTML/JavaScript wherever you want to enable comments:

{% if site.disqus.enabled %}
<!-- Disqus -->
<div id="disqus_thread"></div>
<noscript>Please enable JavaScript to view the </noscript>
<a href="" class="dsq-brlink">comments powered by <span class="logo-disqus">Disqus</span></a>
<script type="text/javascript">
  var disqus_shortname = '{{ site.disqus.shortname }}',
      disqus_identifier = '{{ }}';
  (function() {
    var dsq = document.createElement('script');
    dsq.type = 'text/javascript';
    dsq.async = true;
    dsq.src = 'http://' + disqus_shortname + '';
    (document.getElementsByTagName('head')[0] ||
{% endif %}

Please note that you can place this code in a separate file, let's say disqus.html inside _includes directory and only add one line wherever you want to include it:

{% include disqus.html %}

#Twitter button

For official integration instructions refer to their documentation. As an option you can add following to _config.yml:

like_button_twitter: true

create _includes/like_button_twitter.html with following contents:

{% if site.like_button_twitter %}
<!-- Tweet btn -->
<a href="" class="twitter-share-button">Tweet</a>
  var js,fjs=d.getElementsByTagName(s)[0];
  if(!d.getElementById(id)) {
{% endif %}

and finally include it in template, like so:

{% include like_button_twitter.html %}

#404 Page

GitHub pages service makes it extremely easy to set custom 404 page - just place, let's say, file in the root directory. Also, I'm not huge fan of that, but I add file to directories that don't have index page (like /js/, /css/ etc.) and include content's of 404 page. As long as I have those directories in the list of directories to ignore in robots.txt file, it's not a big deal.


Don't forget to place favicon.ico file in the root directory. Don't have one? Check out

Add meta tag in page's head:

<link rel="shortcut icon" href="/favicon.ico" />


What is humans.txt?

It's an initiative for knowing the people behind a website. It's a TXT file that contains information about the different people who have contributed to building the website.

Why humans.txt?

Because it's something simple and fast to create

So why not:) Place it in the root directory of your website with some message to humans and optionally add author tag to the <head> tag:

<link rel="author" href="/humans.txt" />



And what tools/services are you using for your static websites?

Update on 30/01/2013

Replace {% literal %} with {% raw %} tags as Github upgraded running version of Liquid.

Hi, I'm Sergey, 30yo, father of 2, currently based in Tel Aviv, Israel.

I'm mostly passionate about #music, #programming, #sport, #ui, #ux alphabetical order :)

Read more about me in my Résumé