Archive for November, 2009

Using Rel HTML attribute to your Advantage

Tuesday, November 17th, 2009

Since the introduction of HTML, the link element “a” is one of the oldest elements most commonly used in web development. Included in HTML 2 initially, it still exists in HTML 5. However, even though this tag was enhanced multiple times with different attributes over the years, most webmasters still stick to what they’ve learned, not realizing there were many more options available, or preferring to avoid new attributes like “rel” altogether.

The rel attribute had its fifteen minutes of fame in 2005 when Google announced that they would be ignoring links with the rel=”nofollow” attribute when ranking websites in search results. Unfortunately, 2005 was the “golden age” in the rise of SEO marketing companies and webmasters took that announcement to heart – avoiding the rel attribute altogether or thinking it was only for “nofollow”.

Also, Google’s “nofollow” announcement was greeted with applause by famous blog software makers because it helped to fight spammers who used to abuse blog comments and create site trackbacks using the rel attribute.

However “nofollow” behavior has changed lately.

So, what is really wrong with the rel attribute? Can we use it without “nofollow” effectively?

A hyperlink connects two HTML pages; the rel attribute is supposed to identify the type, relationship or value of the link itself. Since the introduction of microformats and the XHTML Friends Network (XFN) many more rel values have appeared in the web developer’s toolbox. Multiple values can be used on the same link tag, like this: <a rel="home me contents" href="http://example.org">.

A few of the most popular/appropriate uses for the rel attribute are included here. But compare that list with what is available in HTML 5. So much more than “nofollow”.

What about rel’s cousin – the “rev” attribute?

Rev is less popular than rel and is about to die a slow death since it is being phased out in HTML 5. Since the rel attribute defines the link target’s relationship to the current page, rev’s purpose is to define the current document’s relationship to the link target (the reverse). For example, <a rev="help" href="example.org">. This example indicates that the current document is the help manual for the site/page specified in the href attribute.

These types of HTML features have existed for many years, yet webmasters are very slow to adopt and use them with confidence. Taking a more informed approach to HTML can have two main benefits: it would help many search engines to identify what’s really important and the end user would benefit with new and exciting features (that would not be blocked out automatically). For example, tags like rel=”payment” can be used in RSS feeds. If you are media/text publisher, you can ask your readers for a comment or tip without asking them to click away and visit your page. What features could you introduce by gaining control over your rel attributes?


Breaking up is hard to do: Will your links come back?

Tuesday, November 10th, 2009

Often, broken links can be organized into two categories: gone “permanently” and gone “temporarily”. Why is there this distinction at all? A broken link is a broken link – isn’t it?

A “permanent” broken link occurs when a domain name is completely inaccessible or has expired. The domain wasn’t renewed by the webmaster – abandoned only to be bought by spammers who now have the advantage of the site’s former Internet credibility. Other sites that have a link to that domain now have sent their users to the spammer’s site.

Obviously, it’s good practice to know when certain domains expire, especially your own. One of the more popular services is Go Daddy, who will monitor expiry date and domain information for a fee in return. This is good and bad: you might be a responsible and proactive webmaster, who renews in plenty of time, or you could be a spammer waiting to pounce.

If you find a domain inaccessible error, you should look for alternatives or cached content – refer to this LinkAider blog post for more information about how to deal with broken links.

“Temporary” broken links occur when sites go down for maintenance from time to time. Other sites experience network or software errors. In these cases, you’ll probably see a custom error message explaining what has happened. Twitter is the most obvious example of constant downtimes. Its infamous downtime has become the topic of Internet folklore and is a part of many jokes already. Furthermore, Twitter was the inspiration for funny projects like: istwitterdown.com (now defunct itself – how ironic!) August 2009 was rather tough for Twitter according to the chart from the pingdom monitoring service. Growing pains or maintenance can take a toll – but downtime on a social network is not as serious as downtime for your income-generating site. Although these days you may argue that Twitter now provides this type of service and should be making uptime guarantees like other hosting providers.

Hosting providers build contracts around guarantees for uptime. Typically, they are expressed using a percentage. Percentages put a context around the amount of time that a server/hardware/network is continually available (its availability). Percentages make the hosting contract look great at first sight. However, let’s do some math. If a hosting provider says:

  • 90% uptime, it means the server may be down 876 hours a year
  • 95% = 438 hours down
  • 99% = 87 hours, 36 minutes
  • 99.9% = 8 hours, 45 minutes, 36 seconds
  • 99.99% = 52 minutes, 33.6 seconds
  • 99.999% = 5 minutes, 15.36 seconds
  • 99.9999% = 31.68 seconds

Of course, these are just network guarantees. Webmasters can also experience some downtime (on top of that) due to script mis-configuration, bad coding or human errors.

Also, when double checking and/or fixing your broken links manually, keep in mind that some websites may be inaccessible within a certain network location or country. So if you have some doubts, check services like these: downforeveryoneorjustme.com and dingitsup.com.

down for everyone?

Downforeveryoneorjustme.com is a nice test to do when your users tell you that a site is down. You can check it online and then diagnose the problem – it might be their machine, not the server or the network.

 

ding its up

Dingitsup.com sends a text message to let you know when a site is back up or has gone down. Be the first one to know if a site goes down before your end-users or customers start calling. The last thing you want to send out is a large email campaign and have a flood of complaints about a site that is unavailable.

 

In summary, if you have any doubts about a domain link, double-checking is always the safest route. If you’ve already signed up with LinkAider, you know that you can re-run your broken link report with a single click. If you haven’t signed up yet, consider doing it now. It’s free.