I also noticed that articles with hundreds of comments were loading significantly slower than pages with less comments. This made me reduce the number of comments per page to 50 root comments (plus their answers).

Today I noticed that WordPress added another JavaScript to ever page. The script in wp-inlcudes/l10n.js that is related to the admin bar that the WordPress developers added to one of the recent versions of the blogging platform.

While it is a small file with a size of 233 bytes it is still a script that is loaded by anyone, not just the admin of the blog. This somehow does not make a lot of sense. The important thing here is that the blog needs to make the request to load the element, not the time it takes to load the 223 bytes.

WordPress administrators may want to remove the JavaScript element from being loaded by every blog visitor to speed up the blog’s page loading times. The easiest way to remove the l10n.js from being loaded with WordPress is to unload it in the functions.php file in the WordPress theme folder.

All you need to do for that is to add the following line of code to the end of the functions.php file.

wp_deregister_script('l10n');

Please note that this may disable part of the admin bars toolbar functionality. Users over at Stack Exchange have found a way to keep the JavaScript loaded for administrators of the blog.

if ( !is_admin() ) {
function my_init_method() {
wp_deregister_script( ‘l10n’ );
}
add_action(‘init’, ‘my_init_method’);
}

Just add this code instead to the functions.php file.

MobiTest is a free online service that webmasters can use to test the performance of websites on mobile devices. The service is completely free and requires no registration.

Webmasters find the options of the screenshot above on the frontpage of the service. All that it takes to run a test is to enter a website url into the form on the start page. The remaining options are completely optional. They include a menu to select a specific mobile device for the test (several iPhone and Android models mostly from Canada but also the Netherlands and United States), the number of runs, video capturing of the loading and whether the results should be made public or private.

Tests are then started with a click on the Run Performance Test button. The service will add the speed test to the end of the queue. Information about the test’s status are always displayed on the screen.

The developers suggest to bookmark the unique page address to come back later if a lot of websites are listed in the queue.

Each test result highlights several important information. Among them a screenshot of the website and how it is displayed on the selected mobile device, the page loading time and the total size of the page in Kilobytes. MobiTest grades each tested web page and displays a waterfall chart that highlights all connections that were made during the test.

Especially the ability to create a video snapshot of the page loading time on the selected mobile device is handy. As are the different locations and devices that webmasters can choose from. An option to run tests with all supported devices would have been handy.

Webmasters who want to test their website’s mobile performance can head over to the Blaze MobiTest website to do so.

Those information previously were only available after clicking on a site in Google Webmaster Tools which was not a issue for webmasters with one or two domains listed there, but could very well have been a deal-breaker for webmasters with dozens or even hundreds of verified domains. Imagine having to click on each domain to check their health status.

Google yesterday made the announcement that they have redesigned the Webmaster Tools start page. Instead of just displaying a list of domains there, the page is now showing site health for every domain right on the homepage.

Webmasters now see on first glance if something’s wrong with one of their websites. Sites with health problems are listed at the top of the list. Warnings may include that malware has been detected on the site, that important pages have been removed or that the robots.txt file is blocking important pages.

Each site is displayed with a thumbnail screenshot and link to manage the site in Webmaster Tools. A click on a problem opens additional information about the problem detected on the site.

Susan Moskwa, Google Webmaster Trends Analyst, notes that Google is using the display currently only for accounts with less than 100 sites listed. Webmasters with more than 100 sites will see site health at the top of the Dashboard for each site. While I’m not 100% certain it seems to suggest that accounts with more than 100 domains still have to use the old way of verifying that everything’s all right with their websites.

Google plans to expand the site health notifications in the future to include additional issues (currently only malware, removed urls, blocked urls are reported).

I personally like the new layout, as it speeds up the process of making sure that no potential issues are missed when using Webmaster Tools.

One would assume that search engines like Google or Bing had the tools to distinguish between original and copy and act accordingly. This unfortunately though is not always the case.

These sites often rank for long tail keywords which drive some traffic from the search engines to the sites. And since the majority of them runs Adsense ads, they make a pretty penny from that.

It is bad enough that those sites can copy and paste contents automatically on their blogs and earn money from it. Even worse is the fact that the process of setting up a new scraper site does not take more than ten minutes tops, and with automation even less.

So called auto blogs have been a trend in recent years in the Black Hat communities.

Some legit webmasters even experience something that they should not: A scraper site outranking the site where the article was published originally.

The search engines leave webmasters who experience the issue more or less alone. They basically ask the webmaster to fill out DCMA requests and send them to the scraper sites. Problem here is that many use proxy hosting or other forms of obfuscation so that it is not possible to contact the webmaster directly. Plus, webmasters usually deal with multiple scraper sites which leads to a never ending cat and mouse game, especially if you take the easy set up of new sites into account.

Webmasters have criticized Google in particular for this in the past, considering that Google could identify the majority of domain owners easily through their Adsense program as the majority of scraper sites uses Adsense for monetization.

Google recently asked webmasters to report scraper pages to them. The data will be used to test and improve algorithms that target those scraper sites.

Webmasters can submit scraper sites on this web form.

It is about time that Google puts an end to this practice, especially since the company’s recent drive to promoting “quality” sites in their search engines.

Here are several good resources for webmasters who want to do more than just reporting.

How to deal with content thieves
How to deal with content scrapers
Report Spam To Google

If you are a webmaster, what’s your experience with scraper sites?

It is thankfully not super complicated to configure your site to be compatible with the new author photo snippets that show up on Google Search. It is at this point in time however not guaranteed that your photo, or a photo of one of your authors, will appear in Google Search. It is rumored that Google has enabled that feature only for selected blogs and sites.

Nevertheless, it does not hurt to add the code to your site and Google Profile. And if your site gets picked by Google, you can expect a higher CTR (click-through rate) than sites without an author profile photo in Google.

Lets take a look at how it looks like in the search engine if it is configured correctly and enabled by Google:

Here are the steps that you need to undertake:

Creating An About Author page

You first need to create an about author page on your website. You then need to link to that page from all posts that this author writes. The link needs to use the rel=”author” parameter. How it is added depends largely on how the author information are posted on the post pages.

Here on Ghacks, the code looks like this:

<p class=”tags”>Author: <a href=”http://www.ghacks.net/author/martin/” title=”Posts by Martin Brinkmann” rel=”author”>Martin Brinkmann</a>

Adding the code can be an easy task on single author blogs, but a bit more complicated on multi-author blogs. The best way to add support on those blogs and sites is to install a plugin like Allow REL= and HTML in Author Bios that adds the support.

You are not yet finished though. The about author page needs to exist for every author on the site. The actual page can contain any kind of information. One important part though is that it needs to link to the Google profile of that author. The link to the Google profile needs to contain the parameter rel=”me”.

<a rel=”me” href=”http://profiles.google.com/ID”> Author Name </a>

Linking Back from Google Profile

In the last step, we need to link back from the Google profile to the about page on the website. Open your Google Profile, click the Profile icon at the top and then the Edit Profile button at the top. Locate links on the right side and select the Add custom link option.

You need to link to the author page on the website, and enable the “this page is especially about me”. You also need to make sure that a photo is added to the profile.

Testing the rich snippet

Open the rich snippet testing tool and enter an url from your website that is linked to a single author (a blog post for instance).

If you see the author photo next to the listing, you have configured the author profile photo support on your site correctly.

To summarize:

• Create an about author page on your site. Link from that page to the Google Profile of that author with the rel=”me” tag. Link the author’s blog posts on the website with the rel=”author” tags to the about author page on the blog.
• Add a photo and a link to the author’s about page on the Google Profile of that author. Make sure the “this page is especially about me” box is ticked.
• Check for correct implementation with the help of the Rich Snippet Tool.

Google today announced that they have modified site links. When you now search for Ghacks in the Google search engine, you will notice that the site links use up a lot of space. On my 1920×1200 resolution and a maximized window, that’s about two thirds of the screen.

Here is a screenshot of Google’s old sitelinks layout for comparison.

Why has Google made the decision to change the sitelinks layout? The Webmaster Central blog lists four reasons: Visibility, Flexibility, Clarity and Quality.

• Visibility: The font size of the link text has been increased. Both the url and the first free words of the description added.
• Flexibility: Sitelinks are now query dependent, and the maximum number of sitelinks has been increased from eight to twelve.
• Clarity: Sitelinks and standard search results are now separated, which means that a sitelink entry will not appear in the general search results and vice verse.
• Quality: Google with the update has “combined the signals” used for “site links and ranking”.

I personally think that Google went a bit overboard with the new sitelinks design. It takes an extreme amount of space. Both the description and url information are more or less useless, considering that you won’t usually see the full url or more than the first five or six words of the description on the search results page.

I’d rather like to see options to customize the layout. For my site, I’d remove the descriptions completely and start the urls after the domain name, as it should be clear that results come from the same domain (with the exception if a subdomain is listed of course).

Another thing that I think is missing is the ability to pick your own links for the sitelinks. Google picks them automatically, and your only option of influencing the listing is by demoting urls in the Webmaster Tools. First, that is no guarantee that the link won’t show up as a sitelink anymore, and second it does not guarantee that a better link is picked by Google’s algorithm.

For my site, I’d really like to add more categories and get rid of the articles that are displayed. That’s however not really possible.

You find more information about Google’s new sitelinks at the Google Webmaster Central blog. Have you encountered the new sitelinks in Google’s search engine yet? If so, what’s your opinion?

Sometimes though you wait and wonder why the darn page is not in the index right now. This can also be a problem for an established site, for instance if you have changed content on a popular post or on your site in general. You’d like to see the new contents indexed in the search engines which usually does not happen right after you have hit the save button.

Google has now announced that they have added an option for webmasters to submit their urls to the search engine in Google Webmaster Tools.

The existing Fetch as Googlebot feature in Webmaster Tools has been improved with a submit to index option which basically allows you to submit the fetched url to Google for evaluation and inclusion in the index.

Here is how this works. Open Google Webmaster Tools and follow the Diagnostics > Fetch As Googlebot option in the left sidebar. You have to select the right domain first, of course. You need to add it if it is not already listed in Webmaster Tools.

You need to enter the url that you want to crawl. This can be a website’s homepage, a subpage or any other page that is publicly accessible on the Internet. Click Fetch to retrieve the site as Googlebot. This process may take between a few seconds to a few minutes. Once done you get a status report on the same page and an option to submit to index.

You get a prompt first where you can select to submit only the single url or the url and all pages that it links to.

Google currently has a limit of 50 individual page submissions per week, and 10 page with all linked pages submissions per week. The numbers are shown on the Fetch as Googlebot page.

Google has revamped the public Crawl URL form as well. The core difference besides the captcha is that webmasters and users do not have to verify ownership of the page or site to submit it to the index.

The new Submit to Index feature is handy for webmasters who have troubles getting their website or a specific part of that website indexed in Google. (via)

The Page Speed Service is a new addition to the Page Speed family. It basically acts as a proxy, or content delivery network for the website. The benefits according to Google lie in the 25% to 60% improvement range.

Google Page Speed Service is not as complicated to setup as a content delivery network. It does however require a change of nameservers at the domain registrar or web server. The DNS entry needs to point to Google. Here is what the service does:

Page Speed Service fetches content from your servers, rewrites your pages by applying web performance best practices, and serves them to end users via Google’s servers across the globe. Your users will continue to access your site just as they did before, only with faster load times. Now you don’t have to worry about concatenating CSS, compressing images, caching, gzipping resources or other web performance best practices.

Probably the biggest advantage over a CDN and local optimizations is that it is less complicated and time consuming. All that it takes is to change the DNS server to make use of recommended speed optimization settings and features such as compression, caching or merging of CSS files.

Google has created an online tool to test the page loading time of the website and the optimized website. A form is displayed on the Web Page Test site. Webmasters need to enter the url of a website that they want to test into the form. The process will take several minutes to complete.

The results page displays the page load time, start render time of the page, repeat view page load time and repeat view render time of both the normal and optimized website.

A test on Ghacks revealed that the optimized service would take longer to load and render. That’s likely because Ghacks is highly optimized for fast page loading times. It could also be because of different kinds of ads that have been displayed during the optimized page’s loading.

The very same page displays a visual comparison of both page loading times. This could be interesting to analyze both states in detail.

A comparison for another website turned out to be more favorably.

This highlights that it is important to check individual websites to make sure that they might benefit from Google’s Page Speed Service.

A click on View Test displays detailed information about the conducted tests. The information display median values and detailed information from Google’s Page Speed tool. Webmasters can use the information to analyze unoptimized areas of their website.

Google plans to offer the service with “competitive” pricing. For now, it is open to a limited set of webmasters free of charge. Webmasters and developers who are interested in testing Google’s Page Speed Service can apply by filling out a web form over at Google Docs.

The form requires an email address of a Google Account, a website url that domain ownership needs to be verified if accepted into the trial and the approximate page views per day that the website gets. There is no guarantee of inclusion. Google has not yet revealed the pricing, it will be interesting to see how “competitive” it will be in comparison to content deliver networks.

The official announcement can be accessed here.

If you’d ask me I’d say ehow, answers.com, mahalo and the like, plus all article directories like ezinearticles, single-page sites like hubpages or Squidoo and Wikipedia like sites (including Wikipeda).

It is to early to say how the algorithmic change affects search. It is currently rolled out in the US with other countries and regions to follow later this year. If you look at webmaster forums like Webmaster World you notice that several posters are not happy at all with the changes. While it is not possible to determine the quality of their sites, many report that their “legit” sites saw a traffic drop of 30-60% while low-quality sites remained unaffected. The drops happened on February 23 for most webmasters.

It may take weeks before the search engine settles down which means that the results that are currently showing are likely not the results you will be seeing in March. Google confirms that the algorithmic improvement – that’s what they call it – impacts 11.8% of all search queries, a high percentage. The update did not rely on feedback from users who have been using the Personal Blocklist extension for Chrome. The algorithmic change did however address 84% of the “top several dozen or so” most blocked domains from the Chrome extension which Google sees as confirmation that the update is indeed improving search quality.

The problem with that assessment is that it does not necessarily mean that quality sites have filled the positions. Some webmasters who have voiced their opinion in the forum thread linked above have reported that low quality sites have taken over the positions, and that the change affected legit sites that they own or monitor.

Again, it may be to early to tell at this point if the algorithmic change has a positive effect on search quality. Do you see changes in the Google US search engine? If you do, would you say the search quality has improved, or declined? (via)

Website validators like the W3C Markup Validation Service check a web page for validation errors. Webmasters use these services to correct validation errors on their websites.

Here is an example why validation is important. About a year ago I noticed that the RSS feed of Ghacks was throwing an XML error in Internet Explorer 8. This was caused by a single character in one of the latest articles that was not escaped (meaning instead of writing the character & directly one would have to use & instead.

That single character caused problems in the feed delivery, and I had to work longer than usual to get it fixed. Consequences of invalid code are not always that drastic, but it exemplifies the impact that it can have on a website.

Validome is another website validation service. It is a bit stricter when it comes to validating pages. It can happen that a page passes as a XHTML 1.0 Transitional document in the W3C validation, and fails in the Validome validation.

Another difference between the two validation engines is that Validome supports verifications of local and remote documents. Users can either enter an url in the form on the site, or upload a document from their local computer system.

Extended settings are offered to check the document with a specific user agent or language.

A click on Validate performs the validation of the document. Errors are displayed directly.

The service lists the line and column of the validation error, the error position and an explanation of the error. This is often enough to fix the issue in the document.

Validome is an excellent alternative to the W3C validation service. It is usually enough to use Validome since it appears to be stricter in the validation. Most webmasters on the other hand will probably use both services anyway.

Webmasters have lots of options to reduce the page loading time of their websites. This includes removing unnecessary scripts, using compression, minifying HTML, CSS and JavaScript files, merging files where possible but also to optimize images that are hosted on the server.

The difference between an optimized image and an unoptimized image can make a big difference in page loading times. Think of it this way: If you can halve the size of each image hosted on your web server, without reducing the visible image quality, then you have cut the image loading time by halve as well (well halve is not entirely correct but lets use that figure for the sake of the argument).

The two main image formats used on the web are png and jpg. Jpg images are usually well compressed and there is little to gain by reducing their quality further. The image quality drops significantly at a point.

Png images on the other hand offer lots of room for improvement, if they have been saved as true color png images. Lets take a look at the following two images for the sake of this argument.

Do you see a difference in image quality? The second image’s size is 64 Kilobytes, that is 102 Kilobytes less than the size of the first image.

Using Riot to optimize images

You can use lots of different programs to optimize images. They all come with the capabilities but differ highly in their batch optimization capabilities. Riot is a free portable software that can process images in batch. (see Image Resizer And Optimizer Riot)

The program interface looks like this on startup. I have already made the relevant changes to the lower half. In particular, you need to switch to the PNG tab, select Optimal 256 Colors Palette, best compression (slow), NeuQuant neural-net (slow) and PNGout Xtreme (very slow) for the output files.

You then click the Batch icon at the top which opens an overlay window. Click on Add Images (or the small arrow next to it) to load images directly or by folder. Please note that you should only load png images. It does not help to convert jpg images to png, considering that they are still linked as jpg images on the web.

Make sure you select a second folder for the output images. A click on start optimizes all images loaded in the window.

Webmasters can then upload the optimized images to their web server.

Please note that the reduction to 256 colors may not work for all image types. It works well for screenshots and other images that we publish here at Ghacks.

The best option for WordPress based websites was to process one image folder at a time. WordPress saves image uploads in monthly folders. The whole process per folder was to copy all png images from a folder to the local system, add those images to Riot, process them in Riot and reupload them to the server in the same directory after comparing some of the input and output images.

Webmasters with larger projects may want to use a software that can perform these validations on all pages of a website. Open Validator, the web conformance tool, can do that, and more.

The Open Source software is available for Linux, Mac and Windows PCs. A wizard is displayed on startup, that aids the user in the selection of a local or remote website. Additional parameters can be defined, including the depth of recursion, the conformance tests that should be run, the document type, whether a proxy server should be used for the connection, the browser user agent for emulation purposes and pages that should be ignored.

Open Validator then downloads and scans all pages that are part of the selection. This process is speedy, it usually takes less than 15 seconds for the results to appear. Large web projects may take longer on the other hand.

Each page is listed at the top, with its page title and url. Selecting a page displays a snapshot of it as a thumbnail image, the page’s source code and an error report in the lower half of the interface.

open validator

The three error and warning categories are sorted in expanding tables, that provide detailed information for the webmaster. The outline column lists validation categories like design practices, server, HTTP and site considerations or Body information. Error numbers are listed in each row to give the webmaster a first overview of the number of errors and warnings on the page.

Webmasters can directly edit the source code of the page, a click on Revalidate validates the page anew. This works similar to the Firefox add-on Firebug, with the difference that the changes need to be revalidated before the results are shown.

Open Validator offers an impressive feature set:

1) Verification of a single page or entire web site at your option;
2) HTML Validation against W3C DTDs and WHATWG (X)HTML5 pre-release schemas;
3) Automatic checking and correction of many aspects of conformance to international best practices;
4) Visualization of a snapshot with a given character encoding (the user agent is also chosen in order to present the specific version originally intended to be seen in a given web browser);
5) Complete verification of links.
6) Automatic correction of missing widths and heights of images to enable better responsiveness of web pages;
7) Automatic correction of missing meta tags elements like default language and metadata;
8) Support for web pages in Chinese, Japanese, Korean and many other languages.
9) Performance optimization for multi-core processors.
10) Validation during navigation.
11) Application of standards during revalidation.
12) Memory revalidation with no need to make all corrections of a web page at once (users may correct small portions of the code and see the results pressing F5).
13) HTML Tidy accessibility check.
14) Balloon tooltips for contextual help.
15) Recursive validation for frames.
16) Proxy support.
17) Integrated web inspector with javascript debugger.
18) Site map detection in 51 languages.
19) Exportation of reports to HTML
20) Bypass selected paths when using recursive options
21) Exportation of reports to ODF
22) Memorization of login in protected parts of sites
23) Command line tool for unassisted report generation
24) SSL support

Not all errors or warnings that Open Validator detects will throw an error at the W3C Validator. The software goes beyond just checking the markup, which becomes obvious for instance in the design practices category where suggestions for well-engineered websites are included in the report. The findings of the tool are very thorough, which may turn away some webmasters who may feel overwhelmed by the reports.

Professional webmasters on the hand find in Open Validator a tool that can be used as a quality conformance test for their web properties.

Webmasters who want to try out Open Validator can download a version for their operating system from the project website.

Tools have been created to test the page loading time of a website, for instance Yslow or Google Page Speed which both show in detail the loading time and the elements that are responsible for it.

These tools are also used by webmasters to test the loading speed of related website. Those times give the webmaster a good indication if more optimization needs to be done. If a web page loads at least as fast as the fastest competing website then it is usually not necessary to optimize further, if it is slower there might be still room for improvement.

Which Loads Faster is an online service that compares the page load time of two websites by racing them against each other.

which loads faster

The website offers a few preset speed comparisons and the option to create a custom comparison.

To race two pages, a user needs to add two page urls into the form at the top and click the go button. Which Loads Faster will instantly start loading and displaying the websites in two columns on the same page.

Page loading times of both websites are displayed in milliseconds, and the faster one is indicated with an arrow and a percentage value that indicates how much faster it is than the other web page.

The Race link at the top navigation offers another interesting variant. Here it is possible to add up multiple urls to each side to see which loads faster than the other. This could be an interesting option to compare the homepage and several subpages of websites with each other.

The settings contain an option to switch the page loading mode from parallel to serial, helpful if the two sides may interfere with each other.

Which Loads Faster is a free online service that can be very beneficial to webmasters. Additional information about the page loading times would be helpful, for instance a display of the page elements and their loading times or the hardware and software that is powering the different sites. (via To The PC)

To make matters worse there seem to be two – possibly unrelated – issues that webmasters experience. One is a malware attack that is spreading malware on hacked blogs while the other is making use of cloaking techniques to serve a different version of the blog to search engine spiders than to regular visitors and admins.

The cloaking hack appeared on radars in March when bloggers and hosters reported about compromised sites. Media Temple for instance stated on March 2nd that “a number of customer sites [..] have been compromised”.

They identified several patterns the attackers used, one of which placed random-string names in the document root of the blog.

But Media Temple hosted WordPress blogs were not the only ones hit by the attack. Reports from webmasters hosted by Godaddy, Network Solutions or VPS.net indicated that the attack was not web hoster specific.

Fast forward to April 6. Christopher Penn discovered that his blog had been compromised. He found out that the hack on his site injected a new option name into the wp_options table that was using encoded Javascript. The option name always started with rss_.

Deleting that name from the database table stopped the cloaking issues he was experiencing. The key did however appear again which suggested that his blog was still open for the attack.

The vulnerability itself has not been discovered yet. Chris suggested that it has either been the TimThumb plugin or an outdated version of Magpie that WordPress ships with. Both have not yet been confirmed to be the entry points.

There has been no response yet from the WordPress developers regarding this issue.

To make matters worse a second attack has hit WordPress blogs, this time to spread malware. It is not yet clear if the two attacks are related but it is likely that they are.

Frank Gruber posted information about that second attack on his blog which ironically seems to have been successfully compromised as well.

The virus somehow infiltrates WordPress and adds a new file in your scripts directory called jquery.js and then inserts that file into the header or footer files of your site. It also inserts an iFrame that calls a 3rd party site which is known for malware or other malicious activities.

The Trend Micro blog is sharing additional information about the virus that is being spread using this attack. The attack “leads into an infection chain that leads to various malware, including a rogue antivirus[..]“.

To sum it up:

• Several WordPress blogs running the latest official version are currently successfully compromised.
• Attackers either manipulate the blog to spread malware (more recently) or to cloak links that are only visible to search engines
• It is currently not clear how the attacks are carried out.
• Some pointers are given on how to disinfect a blog

WordPress webmasters should check their blogs immediately to make sure that it has not been compromised yet. A wordpress plugin like Antivirus might also help in preventing a successful attack.

One of the online services that can be used for that is Web Resizer. The picture optimizer supports images up to a size of five Megabytes that can be loaded directly on the service’s homepage. Web Resizer will automatically process the uploaded image and display an optimized version of it. This new picture is displayed to the original one so that both can be compared on the screen.

The service will furthermore display the new size of the picture and the image reduction percentage. The controls at the bottom can be used to further optimized the result. It is here for instance possible to change the output image quality, sharpen the image or change the color saturation and contrast. Web Resizer will then reprocess the original image and display a new processed picture on the page.

One interesting option is the ability to compare even more image variants by enabling the “compare image sizes” checkbox. This displays the image in various picture quality settings on the screen. The results are fine for most purposes but the most professional environments. Webmasters who want to optimize the pictures on their websites can access Web Resizer here.

This can result in two problems for the webmaster. Problem one is indexation. It can happen that a tag page gets indexed instead of the actual post. There are plugins that can deal with that problem by adding canonical tags to those links so that the search engines know that the post page should be indexed with priority for the keyword and not the tags page.

The second problem is that the tag page can irritate search engine users as they only see excerpts and not a single article. This can look spammy to some and might have a negative impact on a visitor’s perception of the page and the website in general.

We have posted an option to create custom tag pages in WordPress back in April of last year and that method is still valid and working.

This method basically creates a new template file for each tag that the webmaster wants to customize. It can for instance be used to display an introduction to visitors so that they know what the page is about. This method allows the posting of media as well as text on the tag pages.

Posting those additional information can be beneficial to the search engine indexation of the tag page as well since it adds unique content to those pages.

The disadvantage of that method is that a new template file needs to be created for each tag that the webmaster wants to customize. This is impractical for websites with thousands of tags.

Another method was posted at the Likemind blog. This new method has been introduced in WordPress 2.8. It makes use of a new tag variable called description which can be added to the tag’s template so that it will be displayed on the tag pages of the website.

Some WordPress themes come without a tag.php file. If that is the case it needs to be created and modified to include the description. The easiest way to do so is to copy the index.php file and rename the copy to tag.php. The following code needs to be added before the loop in tag.php

<!–Start Code–>
<?php $tag_desc = tag_description(); ?>
<?php if($tag_desc) : ?>
<div class="tag-description">
<?php print $tag_desc ;?>
</div>
<?php endif; ?>
<!–End Code–>

Each tag page will from then on display the tag’s description if available. Tag descriptions can be edited in the Posts > Post Tags menu in the WordPress admin interface.

The benefit of this approach is that only one template needs to be created and that everything else can be handled from within the WordPress admin area. That tag description can also be used as the meta description. There are some disadvantages though as it is difficulty to add formatting and media to the description (if at all possible).

The service will display transparent information on top of the website that has been loaded visualizing how much of the website visitors see without having to scroll. This is done by displaying percentages on the screen and using colors to differentiate the different groups from each other.

A 98% rating for example means that 98% of all visitors of the website have their browser window open to the size or larger. It does also mean that 2% use a smaller size than the one displayed.

The service can provide the webmaster with important information about the site’s visitors as it can aid in optimizing specific elements on the website to increase their exposure to all visitors without them having to scroll to locate it.

Five, ten or even twenty percent of all visitors is a huge number on some sites. That’s a big number considering that some visitors tend to leave a website without scrolling the page they landed on.

To use Google Browser Size for a specific website enter the url of the website into the text form at the top. Please note that the service will not automatically align the website at the left browser border which might be necessary for sites that do not do that automatically to visualize the real percentages. A user with a 1920 monitor resolution will for instance see a different result for the same website than a user with a 1024 monitor resolution. It is therefor important to align the content to the left side of the screen.

Google Analytics users who can access it will find a display of their website’s traffic graph along with another bar below that is showing the alerts for the corresponding days. Alerts are color coded with automatic alerts using a light green and custom alerts a light blue color. There will only be automatic alerts in the beginning.

The automatic alerts will give a webmaster immediate information about traffic jumps. This includes major traffic changes to certain pages on the website, a change in a page’s bounce rate, pageviews or a higher or lower than usual visitor stream from a specific country, state or region.

This gives the webmaster a quick overview of traffic developments on a specific web project or website. There are certain possibilities where Google Analytics Intelligence module can help. Webmasters could for instance investigate traffic increases or decreases more effectively as it will for instance display a decrease in traffic to a very popular page on the website. The webmaster could then analyze the issue further.

It is furthermore possible to create custom alerts and this could be very interesting to some webmasters. Custom alerts are created from a list of available alert conditions. The webmaster selects a traffic type to be monitored. This can be all traffic, traffic from a campaign, keyword, medium and landing or exit page. This traffic type is then combined with a metric that is triggering the alert. This again can be pageviews, total visitors, revenue or pre-defined goals. The selected metric is then given a threshold value. An alert is then automatically created whenever that value is crossed.

Another interesting option is to look at weekly or monthly reports which even gives a better overview as day to day variations are something that every webmaster has to live with. Google Analytics Intelligence is available in the left sidebar menu of the Google Analytics account.

There are various options to test a website in different versions of Internet Explorer. Some, like IETester, make it possible to install multiple versions of Internet Explorer on the same computer system. Others, like Browsershots, are online tools that can test a website in various web browsers as well.

A third possibility is now offered by Microsoft directly in the form of virtual PC images. Internet Explorer 6, Internet Explorer 7 and Internet Explorer 8 are provided as virtual PC images. It is possible to download a virtual PC image for each web browser running under Windows XP or Internet Explorer 7 or Internet Explorer 8 images running under Windows Vista.

The Windows XP images have a size of around 600 Megabytes each while the Windows Vista images triple that size. The only limitation is an expiration date for the virtual PC images. The Windows XP images will expire in January 2010 while the Vista images will expire 120 days after their first execution. These expiration dates make the virtual PC images a less than favorable solution but its probably better than not testing the various editions of Internet Explorer at all.

Users who are interested in the virtual PC images can download them directly from the Microsoft website.

Wampserver is an Apache web server for the Windows operating system. It provides in many regards the same functionality of other web server solutions like XAMPP or Home Web Server which we have reviewed in the past. The latest version of the local web server package will install Apache 2.2.11, MySQL 5.1.36 and PHP 5.3.0 on the local computer system.

One interesting unique feature is the ability to add different versions of Apache, MySQL and PHP to the installation in the form of add-ons. This can be very handy to test a website locally under multiple different Apache, MySQL or PHP release versions.

The installation of Wampserver will always install the latest versions of Apache, MySQL and PHP. One interesting security feature is the ability to limit access to the web server to localhost. It is required to click on the “put online” option to make it available to other computer systems as well.

Additional Apache, MySQL and PHP versions can be installed from the Wamp add-ons page. All add-ons are provided as executables that can be installed easily on the operating system.

Webmasters who need to test code offline can use Wampserver to do so. It is especially useful to test code against multiple versions of Apache, MySQL and PHP. Something that the other home server solutions do not provide.

Wampserver is Open Source and can be downloaded from the developer’s website.