Wordpress 30min Install





Domain and Wordpress in 30 minutes

In this post I will describe what I just did: grab a domain name, install Wordpress and get blogging in 30 minutes. This all assumes you have access to your own (Ubuntu) Apache webserver and some basic skills. If you don't have that, Google's your friend. If you want me to write a tutorial on how to set up your own webserver, leave a comment.

1. Get your domain name

Go to Go Daddy to get your domain. I used the search bar to find an available address (not so easy these days) and skipped through all the extra offers they wanted me to take (I didn't fell for that) until I reached the checkout screen. In another tab window you can search for Go Daddy coupons! You can get .com domains for $6 at the moment. Enter the coupon code and finish the order.

2. Configure DNS and Apache

In Go Daddy, go to my account, domains, click launch and then below DNS manager click again launch. Change the first entry (under 'A (Host)', item '@') to the IP address of your server. Save.

Now on your server, create the file /etc/apache2/sites-available/yourdomain (you may name it differently but then mind the next step). And fill it like so:
<virtualhost *:80>
 ServerName www.yourdomain.com
 ServerAlias yourdomain.com *.yourdomain
 DocumentRoot /var/www/yourdomain
 
 <directory /var/www/yourdomain/>  
  Options FollowSymLinks MultiViews
  AllowOverride All
  Order allow,deny
  allow from all
 </directory>

 ErrorLog /var/log/apache2/error.log
 LogLevel warn
 CustomLog /var/log/apache2/access.log combined
</virtualhost>
Note that /var/www/yourdomain may be something different, but than you should stick to that in later steps as well!

Now go to /etc/apache2/sites-enabled and type sudo ln -s ../sites-available/yourdomain 009-yourdomain

Now restart apache sudo /etc/init.d/apache2 restart

3. Get Wordpress

  • Note: some commands might require sodu!
  • Go to /var/www and create a new folder: mkdir /var/www/yourdomain
  • Change group to www user: chgrp www-data /var/www/yourdomain
  • Go to /var/www/yourdomain.
  • Get Wordpress: wget http://wordpress.org/latest.tar.gz
  • Unpack: tar xvfz latest.tar.gz
  • Now I like to do: mv wordpress/* . and then: rm wordpress -r 
  • Change group to www user: chgrp www-data * -R

4. Make database and configure Wordpress

Open mysql prompt: mysql -u.... -p (you should know your username and password...often this boils down to root / root). Type: create database mydomain (other db name by your chosing is ok, mind later on though).

Open your webbrowser and go to www.yourdomain.com. Should work... Enter mysql database name, password, location (often localhost). Continue configuration. At some point you are stuck. Copy paste the configuration text into /var/www/yourdomain/wp-config.php and continue (in the browser). You should be able to finish the process now.

5. Install stuff without ftp

Finally, set the permissions on /var/www/yourdomain/wp-content:
  • cd /var/www/yourdomain/wp-content
  • chmod g+w .
  • chmod g+w * -R
And in my case, I also had to edit /var/www/yourdomain/wp-config.php:
  • cd /var/www/yourdomain/wp-content
  • vi wp-config.php
  • Add this line: define('FS_METHOD', 'direct');
Note that this is not safe if there is other websites on your server that you don't trust

Wrap up

You should now have your blog available. You can install plugins and themes right from the Wordpress admin console in your webbrowser, from a white beach in a sunny place. If you have questions or additions to the above, feel free to leave a comment!





Got Chrome?

Browser War
Browsing the web has become part of our lives. Everything favoring this activity has improved: computers, internet, number of websites, features of websites (from animated GIFs to petabytes of HD video), online services and applications, etc. Needless to say the market of webbrowsers is an interesting one to be in, to such an extend that browsers are offered for free.

This has resulted in 'browser wars' ever since Microsoft's Internet Explorer pushed away Netscape (remember all the fuzz when MS integrated IE in Windows XP). Nowadays, war continues, but new parties have entered the battlefield. The user benefits but developers are left in a forest of different standards (or different interpretations of the same standards, which to me seems as an contradictio in terminis).

Reality doesn't always make sense
It's not necessarily browser speed, user friendliness or aesthetics that determines market shares. Currently there's a bunch of browsers (versions) lagging behind that are still being used quite frequently, and there's browsers that are quite up to par but not equally popular.

Chrome
Chrome was developed by Google (based on existing stuff) and is aiming at speed, realizing browsers have to act more and more as 'platforms' for quite sophisticated software (Chrome OS even aims at a computer solely based on a webbrowser). This has urged other vendors to do the same (optimize the hell out of their JS engine). Opera is doing well, Firefox is catching up, Safari is as well and Internet Explorer... well IE itself might be catching up, but the users are much less so because they are not always installing the new browsers. This will become even worse now IE9 will not even be available for Windows XP or Vista, while especially XP will be around for quite some time (is my guess, especially in less developed countries).

Chrome
Looking at usage, Chrome is the one in the rise (see picture). Its growth curve even seems exponential. Internet Explorer on the other hand, is going down down down. That's reality and it seems to make sense this time. Opera on the other hand has less market share than it deserves based on the specs.



Versioning
One problem for web developers, apart from browser vendors, is browser versions. The rendering of a page might differ between versions from the same vendor just as much as between two versions from different vendors. Chrome, however, has a very sophisticated update system. Small patches (containing only the differences between the old and new version, taking into account that it are binary files containing executable or library code) are pushed to the client as they come out. When I looked at my website stats I first didn't believe this, because I saw multiple Chrome versions with high usage percentages pop up. But when I took into account the release dates of these versions, it turned out that almost all Chrome users use the 'current' version.

Google Chrome Frame (IE Plugin)
Google came up with something brave, clever and slightly backstabberish: the Google Chrome Frame, "[..] an open source plug-in that seamlessly brings Google Chrome's open web technologies and speedy JavaScript engine to Internet Explorer". From the developer perspective, it works like flash: if the user has it installed it works transparently, otherwise the user is directed to a download page through which he/she download the plugin very easily.

Conclusions
If you are still using IE and you are not using Windows 7: go Chrome! In other cases: at least consider Chrome! It's available for Windows, Mac and Linux (32 and 64 bit). Those who want to be exclusive: consider Opera.

Stat Stuff & Meta Update

No new is good news
No news is good news, or, as the English King James I wrote in 1616 'No newis is bettir than evill newis.' Anyway, I'm doing fine thanks! I hope the same holds true for Murlu (murlu.com). I was following his blog already for a while, for reasons I blogged about before. After leaving his job, cranking up the posting speed, and then leaving for Japan, all went silent: his main blog since march, his traveling blog since april. His Twitter feed continues, but all tweets since April appear automated. That's odd isn't it? And Spooky.

Referral spam killed the fun
What is odd as well, is referral spam. I blogged about it before, mainly focusing on one player that went pretty far and made things worth diving into. Then at some point my Blogger statistics went crazy with tons of weird sites. Here's my all-time top list of referrers:

getrxpills.com
edprescriptiondrugs.com
antibiotics-shop.com
basicpills.com
god-sersy.livejournal.com
9pillsonline.com
condition.dn.ua
deluxetravelua.com
amoxil-pharm.com
e-pharmacy-online.com

I don't recommend visiting any of those. It sure killed the fun of tracking Blogger's statistics. Google's Analytics however is not disturbed by all this (funny fact: they're both owned by the same company). So I've slowly switched to tracking that and webmaster tools, and reactions and follower counts both on Blogger and Feedburner. As long as Blogger doesn't do anything to defend against this, Blogger stats are plain useless, and a waste of programmer's time!

Back in town
Basically, Google Analytics showed an increase in visitors over time, even without new posts arriving. This encouraged me to return to the front line. Both my own activities and things happening in the outside world give rise to lots of stuff to talk about. A lot of my time recently went into the development of a Facebook application. Developing an app for Facebook, and learning more about this humongous beast of a website along the way, was surely an interesting process. Now there's the question: will it blend? Will our ideas, our technical implementations and - most important of all - our user's interests and demands mix into something sustainable? More about that later!

Post Scriptum
In other news, my first offspring blog wasn't really going anywhere. And my priorities have been such that a second one has not aired (yet). My company's doing fine although the Facebook activities are actually out of it's scope and hopefully will result in another one launched soon.

NoFollow

Having links to your blog or website is important on the web. Not just because people may actually click on them and end up on your site. The mere presence of a backlink to your site by itself may increase your search result ranking (this is not the only parameter though!). And this simple fact is like pandorra's box...

Link spam
Some time ago I had created a website for a friend, very basic, very local, custom made. I had added a guest book (old school) and even though it was handcrafted it got picked up by spambots at some point in time, posting pages and pages of links and garbage. A simple captcha-like solution was enough to stop the madness. Captcha's indeed have conquered the web as a response to spambots and the like. But nowadays spamfighters have another weapon at their disposal..

rel=nofollow
By adding rel='nowfollow, to a link, as attribute inside the anchor tag (a-tag), a webmaster can prevent 'PageRank' attribution to this external link. The behavior is quite established amongst all major search engines. Whether the searchbot actually follows the link depends on the search engine. Note: Google nowadays does subtract the would-be pagerank from the total amount you can give away (ie other links on your site do get less as if there was no nofollow, but the nofollow link doesn't actually get it).

Why is this important?
Apart from being effective against spam(bots), the nofollow attribute is also important when you are seeking ways to (legitimately I presume) promote your own site. Since version 1.5 WordPress adds nofollow to all links in user-submitted content (ie comments headings and contents) by default. This may be relevant if you want to acquire links to your site by posting comments on other blogs (as is suggested in numerous videos and blogs elsewhere). Using the tool SEO Quake (which has plenty more features you may want to turn off for your own surfing convenience) you can quickly identify which links on a site actually have the nofollow attribute.

The future of the web?
It is interesting to think about where this is going. More and more sites and platforms are making use of nofollow, as backlink seekers' strategies evolve. Ultimately, this may at some point degrade the informative value of ranking algorithms because 'true' links are also removed from the equation. It is therefore not surprising that the amount of 'PageRank' is not the sole determinant of search rankings anymore. As more intelligent systems emerge, this ratio will most likely keep decreasing over time. What to do then, if you want to promote your website? Content, my friend, is King. Social media, my friend, the Queen.

PS:: this was, in hindsight, part three in an otherwise unintended series of html tags that play a role in search engine optimization :-D

Canonicalization

So how big do you think your site is? 100 products and some general pages, say 150 pages? Think again. Are both www.yoursite.com and yoursite.com useable and displaying the same content? And both index.php and '/'? And shoes.php, shoes.php?sort=price, and shoes.php?sort=date? Or even
www.youroldsite.com and www.yournewsite.com? And do you care about SEO and your search engine ranking? You might want to continue reading...

Duplicate content
To Google (or any other search engine) the above pages are in principle all different. It does however, often recognize this fact and display only one in the search results. Still, you could do better.

PageRank diffusion
If sites A and B link to yours, that's great. You got some backlinks. Depending on their PageRank, you will get some PageRank juice as well! Twice! Nice! But wait, closer inspection reveals page A links to www.yoursite.com/index.php, site B to yoursite.com/index.php. To make matters worse, you find a third site that links to yoursite.com/ (no index.php). Every individual page in that list will start with zero PageRank and gets a little but only from either page A, B OR C! So, you might have three backlinks, but the PageRank is divided over the three 'different' pages. If only there was a way of Google to stop this madness! Luckily, there is...

Link rel=canonical
With the canonical tag, a page can say to Google 'I am insignificant, please ignore me and see me as duplicate of ... (page in the href part of the tag)'. You can even do this for pages that are not 100% identical (but still are very similar). And it even works cross-domain. Remember you are always 'giving away' your PageRank (but if it is to a page of your own, that can be very beneficial), so the tag should be in the 'inferior' pages. Google will also take the 'canonical' (or 'master') page as the one actually shown in the index. Remember this is just a guideline for Google...it can always decide to ignore your effords, if big'ol search bot feels like it ;-D

Here's the tag:
<link rel="canonical" href="http://www.someplace.com/heresthemotherofallsimilarpages"/>