The url in the options table can be overridden in the wp-config file on my dev environment.

define('WP_SITEURL', 'http://dev/gh');

These browsers are easy to setup and only require a 16×16 image file (png, jpg, or gif) and a link tag in the head tag of your code.

<link rel="icon" type="image/png" href="/favicon.png" />

IE7:

IE 7 is a little more tricky. First, IE requires an actual icon file. Not just a jpg or png renamed with an ico extension. But no worries there are a few sites online that will convert your images into ico format for you.

IE automaticaly checks the root directory for a file called favicon.ico, so thats where I recommend placing your image. (/favicon.ico)

You don’t need any link code snippet to make this work, but if you wanted to force IE to find it and or change it for another page you can use the following link code in the head tag.

<link href="/favicon.ico" type="image/x-icon" rel="shortcut icon" />

I wrap this code with an IE conditional comment. Although the other browsers recognize this code, it’s not correct if your trying to produce accurate XHTML, because of the space in “shortcut icon”.

<!--[if IE]>
	<link href="/favicon.ico" type="image/x-icon" rel="shortcut icon" />
<![endif]-->

Troubleshooting IE:

If your favicon doesn’t show up immediatly it’s because you need to clear your temporary files as well as your browser history. That should force IE to update the favicon.

Also make sure that your not preventing IE from cacheing your favicon. You’ll need to check your headers, using a tool like fiddler. But most likely you won’t need to worry about this.

If you have any questions don’t hesitate to drop me an email.

overage considering I live in the middle of nowhere. I’m about 20 min from Columbus OH.

My broadband connection is wireless and it caps me at 185 kilobytes per second.

The app I used for the test was Speedtest[itunes].

If the pattern ends with a slash, it is removed for the purpose of the following description, but it would only find a match with a directory. In other words, foo/ will match a directory foo and paths underneath it, but will not match a regular file or a symbolic link foo (this is consistent with the way how pathspec works in general in git).

This never worked for me. The only way I could get it to work was to use something like “directory/*” in my .gitignore file. I could also just use “directory”, but that would match any file and directory with that name. It would also look for other matches further down the tree recursively.

So in summary to hide the build directory use:

build/*

I’m going to focus on the album/singles pages. My first thought when glancing at a sample page is information overload. I took a screenshot of coldplay’s Prospekt’s March EP page and proceeded to examine in detail what I think should be taken out.

In the screenshot, I mention that I would hide some information. The information should still be kept, just have it hidden till the user wants to see it. And then it can slide into view with some jQuery magic. It should be loaded when the page is first loaded; no ajax. That would just be a wasted call to the server.

Next I’d like to focus on where customers spend most of their time. At the MP3 box type thingy where you can preview and purchase individual songs.

1. There is no way to select a couple of songs and buy them all at once without going to a different page and back again.
   1. I noticed today that when you click on Buy MP3 it no longer takes you to another page but replaces the Buy MP3 button with “Purchased”. One down three more to go.
2. Amazon should also copy Itunes and show a small bar graph showing a songs popularity.
3. Why can’t we preview the entire track? If it’s because of Piracy, there are much easier ways to steal music, than ripping the audio from a streaming site.
4. I would like to see a fast forward type bar when listening to the previews.

Something else I’d like Amazon to add is a better recommendation engine. Better than “Customers who bought this, also bought this”. And no, Apple genius isn’t any better.

I hope you enjoyed my review.

So wouldn’t it be great if I could use vims syntax highlighting? Here are the steps I used.

• Make sure that your current terminal is capable of displaying 256 colors. Since I use a Mac I had to make the switch from terminal.app to Iterm. In Iterm’s profiles I changed the terminal type to “xterm-256color”. Then I ran a color test that verified that I was using 256 color mode.
• On my machine I also had to set my TERM variable to “xterm-256color”.
  • export TERM=xterm256color
• Then you just need to find a nice color scheme that only uses a 256 color palette. I choose summerfruit.
• And finally you have to turn on syntax highlighting in vim with (if it’s not already on):
  • :syntax on

“Tell me something I don’t know”, your saying. I have a handful of sites on the grid service, from wordpress to static html. They have one common trait. Extremely long latency. Typically 1 to 4 seconds on a good day.

I’ve contacted support multiple times and it’s like speaking to a broken record. “Wordpress is the cause, Drupal is the cause, Your sucky programming is the cause” you get the picture. I guess I’m being hard on them.  How was I thinking they would respond?

Warning Extreme Sarcasm ahead.

“I’m sorry about our crappy network. We oversell accounts and so we are constantly having network issues. Let me recommend another more competent host for you. Of course we will refund your account and pay you for your unnecessary time optimizing the already optimized sites.”

If they should that kind of honesty, I might consider keeping them as a host. But as it stand now I will be switching over to slicehost.

Regardless of the situation at MT, I should be using host that gives me root access to my VPS.

P.S. Slicehost, PLEASE don’t let me down.

The trouble came when I decided to revert my code back to a previous instance, and after reviewing my site, I pushed the updates to my slice and turned on wp-super-cache. But after a few hours I noticed my site was continually going offline. Even though it never affected any of the other sites I hosted on that server.

The problem was that git doesn’t remember file permissions and I forgot to tell git to ignore the cache folder. So when I reverted back to a different version it also reverted the cache folder, its contents, and proceeded to change the permissions to what I assume was the default permissions when I first initiated the repository. So not being able to write to the cache folder made wp-super-cache crash the apache process and eventually all the apache processes took all the memory. 

So when using git don’t forget to reset your permissions, when reverting back to an earlier instance of code.

SPF stands for Sender Policy Framework. It allows domain owners to specify which mail servers they use to send mail from their domain. The policy is specified by adding a txt record containing the SPF policy to your domains nameservers. So for SPF to work the receiving mail server needs to be setup to check the sender’s domains policy.

Just because your domain doesn’t have an SPF policy doesn’t mean your email will automatically be labeled as spam, But it can help you get out of the spam folder. An emails header without a SPF policy would look like this: 

Received-SPF: neutral (google.com: 210.19.84.47 is neither permitted nor denied by best guess record for domain of www-data@gnix.us) client-ip=210.19.84.47;

Authentication-Results: mx.google.com; spf=neutral (google.com: 210.19.84.47 is neither permitted nor denied by best guess record for domain of greg[at]gnix.us) smtp.mail=greg[@]gnix.us

Once you add the policy to your nameservers, your headers will look like this: 

Received-SPF: pass (google.com: domain of greg[at]gnix.us designates 210.19.84.47 as permitted sender) client-ip= 210.19.84.47; Authentication-Results: mx.google.com; spf=pass (google.com: domain of greg[at]gnix.us designates 210.19.84.47 as permitted sender) smtp.mail=greg[at]gnix.us

Of course there are a lot of other factors that spam filters use to eliminate spam, but adding an SPF policy is an easy way to add trust between email senders and spam filters.

On SPF’s website, they have a wizard to generate your TXT record. Just input your domain and answer a few questions about which mail servers will be used to send emails from your domain.

SPF isn’t the only protocol that tries to validate the sender of an email. And out of the three big webmail companies (Google, Yahoo, Microsoft) only Google checks for SPF records. Here’s a summary on what the others use.

Microsoft

Sender ID - Is confusingly similar to SPF. they both validate email sender addresses and both use dns records to publish their policies. I won’t get into the differences, but they are different.

Yahoo

Domain Keys Identified Mail (DKIM) - Signs the email with a key, and then the recieving mail server verifies key from a record in the sender’s domain nameservers.

SPF - They don’t check for SPF records, but they do have a SPF policy in their dns records.

Google

SPF, DKIM

If your serious about your email getting through spam filters, you should look into DKIM. Unfortunately DKIM requires a lot more effort to get working, but is probably the most supported.

As far as content goes, and as long as I can stay on topic, 90% of my posts will be about web design. A more detailed list of topics are as follows:

• (x)html
• css
• javascript
• php
• mysql
• server settings
• photoshop
• illustrator