Together with Chris Martin of Coldplay Coldpie fame, we have founded a new company called Bright Machine. If you need some kick-ass mobile app or web development, then please email me here.

Some of the commands below may require additional software that doesn’t come with your default server installation.

Disclaimer: if you’re stupid enough to do any of the below, don’t blame me.

General Things

Firewall / iptables

From netfilter.org:

iptables is the userspace command line program used to configure the Linux 2.4.x and 2.6.x IPv4 packet filtering ruleset

Basic info:

• Config file: /etc/sysconfig/iptables
• List rules: iptables -L
• Open a port: iptables -I INPUT 1 -p tcp --dport 80 -j ACCEPT

Adding an IP address to an existing network adaptor

Stolen from here.

cd /etc/sysconfig/network-scripts
ls ifcfg-* # check for existing interfaces
cp ifcfg-eth0 ifcfg-eth0:0

Then edit the ifcfg-eth0:0 file and change the DEVICE and IPADDR settings, eg.

DEVICE=eth0:0
IPADDR=999.888.777.195 # whatever your IP address is

…then…

ifup eth0:0

Adding an additional loopback address in OS-X

ifconfig lo0 alias 127.0.0.2 up

Setting up SSH so you don’t need to enter a password

Stolen from here.

If you manage more than one or two hosts, you likely have to type the same password too often. This can get quite annoying. SSH allows you to setup a public and private keypair. Using these keys, you can connect to any host which has the public key, from any host which has the private key, typing your password only once.

So, the process goes like this:

1. Generate a keypair
2. Add your public key to the authorized_keys file on the remote host
3. … errr that’s it

NB: make sure the .ssh directory exists in the user home directory on both your local and remote machine.

ssh-keygen -t rsa
cd ~/.ssh
cat id_rsa.pub > authorized_keys; chmod 600 authorized_keys # adds your public key to your localhost
chmod 600 authorized_keys
cat id_rsa.pub | ssh YourRemoteHost 'cd .ssh; cat >> authorized_keys; chmod 600 authorized_keys'

Alternatively, you can do that last bit manually by scp‘ing the public key up to the remote host and cat‘ing it onto the authorized_keys.

Command Line Mojo

Change to the Previous Working Directory

cd $OLDPWD # old way
cd - # new way

- on its own is the same as $OLDPWD… nice!

Which version of Linux?

lsb_release -a

You may need to add lsb support, just yum search lsb to find the package for your OS.

Run the last command

!!
sudo !! # when you forget to run the command as root

This is great is your coding up a command line script, and debugging, and continually running the same command one after the other. Using !! will run the last command but not pollute your history.

Open ports

nmap -sT -O localhost
netstat -tlnp

One of those commands usually gives me the info I’m looking for.

Create a local mirror of a site using wget

wget -m -k --adjust-extension http://url-of-site.com

Disk Usage

df -h # show disk usage on all partitions
du -sh /var/log/maillog # size of 1 directory / file
du -sh /var/* # size of all files / folders within a folder

Truncate a file

Useful emptying a log file. There are 2 situations when I want to do this.

Emptying a file from within VI

vi some_file.log
[esc]
:1 # go to top of the file
dG # delete all contents until the bottom

From the command line

cat /dev/null > some_file.log

Clearing the terminal

Any of the following

CTRL+l # lowercase L
clear
reset # resets the shell if it starts acting a bit weird

Setting your date & time using an NTP server

ntpdate time.apple.com

I created traffic to take the pain out of mapping incoming requests to your site to code that does whatever you need it to do. Most existing solutions, like the atrocious Zend_Controller_Router_Rewrite whatever assume that you want to map requests to method on a class, with enough magic mixed in to be a pain in the arse.

With Traffic, I’ve taken the approach that you should be in charge of how you organise your code. All Traffic does is the matching, and leaves the rest up to you. if you’re a framework monkey with an incessant need for hand holding, then move on, coz I’m not gonna tell you how to structure your app or anything like that.

Traffic is inspired by Sinatra & Sammy JS without pretending that it’s not PHP.

Here’s a snippet:

use Fu\Traffic as t;

// relative_to is an alias to rel, in case you prefer to be more verbose
t::relative_to('/user/:username', function () {
  t::get(function ($params) {
    echo 'Viewing user:'.$params['username'];
  });

  t::post(function ($params) {
    echo 'Creating user:'.$params['username'];
  });

  t::put(function ($params) {
    echo 'Updating user:'.$params['username'];
  });

  t::delete(function ($params) {
   echo 'Deleting user:'.$params['username'];
  });
});

Go check out the code and a smidgen of documentation.

When CSS and images are included in your HTML, they are loaded asynchronously. This is not the case with JavaScript. When a JavaScript file is loaded, no other resources will be downloaded.

In terms of performance, most developers right now follow this rule: CSS at the top, JavaScript at the bottom. This allows all CSS & images to be loaded asynchronously, before it hits the roadblock of JavaScript, but at least by that point you have all your resources, so as soon as JavaScript is finished your page will be displayed.

However, there are alternatives available that allow JavaScript to be loaded asynchronously. These methods are documented here by Steve Soulders, but the main problem with this is it means a fair amount of work to implement… a developer must architect his or her code specifically to allow it to be loaded in this manner.

Over the last few weeks, there has been some progress in bringing one of the techniques (Script DOM Element) to the masses.

Analyze This!

First off, and probably the easiest way for a developer to get into asynchronous JavaScript loading, Google launched asynchronous tracking code for Google Analytics.

As Google Analytics uses JavaScript to collect usage data, it usually adds more load time to each page it is installed on, and can potentially cause pages to hang, if there are issues with Google’s servers. But with the following code, this issue goes away:

<script type="text/javascript">
var _gaq = _gaq || [];
  _gaq.push(['_setAccount', 'UA-XXXXX-X']);
  _gaq.push(['_trackPageview']);

(function() {
    var ga = document.createElement('script');
    ga.src = ('https:' == document.location.protocol ? 'https://ssl' :
        'http://www') + '.google-analytics.com/ga.js';
    ga.setAttribute('async', 'true');
    document.documentElement.firstChild.appendChild(ga);
  })();
</script>

Just replace the UA-XXXXX-X with your own identfier.

To see an example of this, just have a look at the source code of this site.

Straight From The Lab

Another project that has been emerging to tackle this same issue is LABjs.

From the LABjs documentation, the following code shows how one might use it:

$LAB
.script("framework.js").wait()
.script("plugin.framework.js")
.script("myplugin.framework.js").wait()
.script("init.js");

The wait() bit would be where you want the browser to block, everything else would be loaded asynchronously.

I’m sure the LABjs project will evolve a lot over the next few months, but it’s an interesting start… one to keep your eye on.

Related articles by Zemanta

• Yahoo! YSlow for Firebug (stereonaut.net)
• Google on Speed (bruceclay.com.au)

Firstly, there’s the new addition to the family that is coming along right around the corner. To make it work, I’m just gonna have to have complete control over how & when I work, more than I would be willing to ask of any employer.

Secondly, I’ve been wanting to run a business for a few years now. Although I didn’t envisage it being in a freelancing/consulting capacity, it made sense when I thought about it. Why jump in trying to do something new? Surely it would be sensible to do the thing that I’ve been doing for…

10 years. That’s the third reason, or at least part of the third reason. As well as my second child, 2010 will bring with it my 30th birthday (oy vey) and my 10th year in full-time employment as a web developer. The long and short of it is, it’s time for me to do this for myself.

So, there you have it. 2010 here I come. It’s gonna be HUGE.