Tuesday, September 16th, 2008
Do you remember my article from yesterday about the sudden drop in Google search ranking for my friend’s website? Well, I just can’t stop thinking about it.
From what I have been reading, it seems as though my conclusion may be correct. At least I am hoping it is. If I ever conclude anything semi-concrete while thinking about Google, it’s a good day for me.
Ok, I found this very helpful and thorough website that pretty much described the exact problem my friend is having. It’s titled “Google Proxy Hijacking” and tells the whole story.
Here is what struck me as I think about this some more.
- My friend’s website has been live since 2004.
- The site seemed to be in the Google sandbox for the entire 4 years.
- For his most competitive keywords, he was ranking past page 20 on Google.
- About two months ago, he made some changes to the homepage copy as well as an HTML overhaul.
- About a month after that, the site ranked number 3 for his most competitive keywords.
- The site ranked on page 1 of Google for about a month.
- The site now sits at page 25 for its most competitive keywords.
Here is my theory. I think the website has been proxy hijacked for a number of years. This is what caused the poor rankings for such a long time. When the homepage text and HTML changes were made about 2 months ago, Google visited the site and found it unique. Google ranked the site well, due to this new unique content. During the month, Google noticed the proxy website was now a duplicate of my friend’s website once again and dropped the website’s ranking.
Does that make sense? From what I read on the website I linked to, it does.
Here are the similarities with what we are experiencing and what the author wrote on the other website:
- My friend’s website has never been banned.
- We did a quoted Google search for supposedly unique content on my friend’s website and a proxy website showed in the search results.
- The proxy URL looked like this: proxysite.com/cgi-bin/pxy/nph-pxy.pl/000010A/http/www.friendssite.com/
- The proxy site was an exact duplicate of my friend’s website.
Now, I am not sure if this is what caused my friends ranking to drop, but all the factors are there. The keywords we are talking about are very competitive, but the fact that his site showed so well in the search results for a month shows me that the potential is there.
I would appreciate your thoughts on this.
Related posts
Tuesday, September 16th, 2008
There may be cases when you would like to see what your webpage HTTP headers look like. Why? Well, because they are kind of important. As Wikipedia states, the HTTP headers define what the returned data looks like.
Still you ask, “Why in the world do I care about that?” Ok, I’ll keep going. The main reason I look at the HTTP headers is to find out what the HTTP status code is. The reason the status code is important to me is because this is the code the search engines use for a multitude of things.
Let me give you a little example, and this related to my previous post regarding the sudden drop in Google rankings. As I was doing research into what the problem may be for this particular website, I came across an issue where someone had recently put custom “404 Not Found” error pages up on some of their websites. Everyone knows that custom “404 Not Found” error pages are cool, but what some people don’t know is that if those 404 error pages show a “200 OK” (successful HTTP requests) code, the site may be in big trouble, SEO-wise. The reason for this is because there are going to be many “404 Not Found” error pages on a dynamic website. If you have your custom “404 Not Found” error page showing a “200 OK” response code, the search engines will think that all the instances of this page are duplicate. You know as well as I do, that spells trouble.
What’s worse is if you set your homepage as your “404 Not Found” page. Your homepage is going to return a response code of “200 OK.” That’s not good, because now you have multiple instances of your homepage…all duplicate content.
It’s my opinion that the search engines are smart enough to figure this out. The page (such as your homepage) with the highest Pagerank will prevail. Still, I have some websites that I am working on that have multiple instances of the homepage and they all have Pagerank, which isn’t good, because the duplicates are taking the Pagerank from the real page. Now, again, that’s my opinion.
Here are two tips:
- How to check your HTTP headers – visit this website or just Google “Website header check”
- How to set a particular page as your “404 Not Found” error page in your .htaccess file – Just place this code in the file: “ErrorDocument 404 /404.php” without the quotes. The 404.php file is the actual error page in this case.
Related posts
Monday, September 15th, 2008
A colleague of mine gave me a call yesterday morning with some rather upsetting news. Apparently, one of his websites took a plunge in its Google Ranking. He wanted to know what could cause such a sudden drop in Google Ranking like this.
I really didn’t have an answer for him. The site has been alive (but in the Google Sandbox) for about four years. It always struck me as strange that the site was sandboxed for such a long time. It literally took four years to come from page 30 in the Google Rankings to page one. Suddenly, last month, the website appeared on page one for its most prime keywords. Now, this wasn’t a gradual change in ranking, it was a huge jump.
The website doesn’t appear to have anything wrong with it. I gave the entire site a once over. I checked the typical meta information and linking structure and found nothing wrong. The website really hasn’t changed in months, besides the content, so it led me to believe there are outside forces at work.
The question I have is, “Why would a website, with a poor ranking, suddenly rank number five on Google one month and then fall back to page 24 the next month?”
I tried to get some information out of my friend. The only thing major he did in the past few weeks is to add a custom 404 or Not Found error message. I checked the 404 page to make sure the headers were correct and not giving 200 results. They error 404 pages were fine.
Then, I went over to Copyscape to see if there were any copies of his homepage. I have heard this can cause a sudden drop in Google rankings. I did find a proxy website that had almost his entire website cached and was trying to pull it off as its own. This wasn’t a typical proxy server trying to speed up the internet. This was something else…more like an intercept proxy.
I looked in the log files to find the IP address of this proxy website. I found it and blocked the IP address in his .htaccess file and then checked the proxy website again. His website no longer showed and was replaced by the Red Hat error page instead.
We will have to give this a few weeks to see if anything changes. I am now thinking that is something does change (for the better), this may have been what was causing the extremely long Google Sandbox issue as well.
If you have any further suggestions, please let me know via comment.
Related posts
Thursday, July 17th, 2008
For some time, I have been looking for a versatile CMS (content management system) that would allow my mind to wander. So far, I have set up some pretty nice systems. They are meeting my goals, but I wanted to see how things would be outside the realm of business. I mean, a new website would certainly be set up under the business name, but it would be sort of a hobby site in the beginning and we would see what it looks like after a year or two.
There is a blurry line between CMS applications and blog applications now. I understand blogs are becoming actual CMS applications, so who knows what will happen with that. I can see WordPress staying my favorite for a while, but I feel like learning something new.
Here are some of the top open source CMS applications I looked at:
- Joomla
- e107
- Drupal
- Plone
- Mambo
- PHP-Fusion
- PHP-Nuke
- XOOPS
After going over a bunch of these, I decided upon Joomla (along with about a million other people). Just to let you know, there is a really great website that lists many top content management systems, along with blogs, wikis, etc… All listed systems are open source and the website has a link to their homepage as well as a demo of the front page and the admin. This helped tremendously during my research.
I really don’t know what I am going to do with this website yet, but I am sure that once I go through all the plug-ins, I will start forming a plan. For now, I am just going to do the install and document the process here. After that, I will document all the exciting things I come across.
Related posts
Sunday, July 22nd, 2007
I decided to pull the plug on my wiki. I really never got into it past setting it up and making a few configurations. I thought I would let that site sit for a while and age like a fine wine. Well, aparently some religious fellows had a different idea. They would post things to their heart’s content. Since I can’t babysit a site like this and since my interest wasn’t really with it, I said good bye.
If you would like to buy the domain name, I will sell it for $1,000,000 negotiable. If you can’t afford that, just wait a year until it becomes available and you can get it for $8.99 through GoDaddy.
Related posts
Friday, June 8th, 2007
Paul sent me an email this morning about the Math Function he added to his Wiki. Here is an entry that shows the result.
Paul did some lengthy and frustrating research trying to get this thing straight. He says, “It is called mimetex, and if you have a shared server, it may be the only way you can use a TeX markup to show your math formulas. Your server needs to be able to run a CGI script, and you need to create a cgi-bin/ director in the root of your webpage. All the installation directions are located here. There is also a wordpress plugin, which is really cool. You can find the user’s manual here.”
Well, thank you Paul. See, I told you he was a good researcher.
Related posts
Wednesday, June 6th, 2007
I came across this very important discovery this morning. I was attempting to edit the “About DotWiki.US” page of my Wiki when I thought a link would be nice.
Wiki syntax, Wikitext language or wiki markup is a bit different than regular HTML. I am not sure why they did it this way…perhaps because the general public is going to be the ones editing Wikis and Wiki syntax is easier and more logical (it seems).
Here is the line I put into the page editor…
DotWiki.US is a personal project aimed at learning how to use and promote a Wiki. DotWiki.US uses the same software as the widely known, [http://www.wikipedia.org Wikipedia].
Notice at the end of the second sentence, there is a link. Basically, an external link is enclosed in [ ]. The link above is an external link directed at Wikipedia. Inside the square brackets, the first thing you place is the actual URL. Right after that, place a space and then your link title. It’s as easy as that.
Now, for an internal link, it’s a bit different. Take a look at the paragraph in edit mode below…
While in theory any animal might be a pet, in practice only a small number of species of [[mammal|mammals]] (especially [[dog|dogs]] and [[cat|cats]]) and other small animals, such as [[bird|birds]], [[fish]], or [[lizard|lizards]], are practical. One reason for this is that large animals are not able to fit inside small dwellings.
You’ll notice a few different types of links in there. If you want to make an internal link in your Wiki to another page in your Wiki, simply find out the name of the page. Here, we will look at the “fish” link (end of first sentence). The name of the page this person wants to link to is called, “fish.” Therefore, a double [[ ]] is used with the name right in between. That’s pretty straightforward.
Where it gets a little more tricky is when you want an internal link to a page with a different page name than your link title. Lets look at the “mammals” link above (first link). The page name is “mammal” but the link text in the sentence is “mammals.” Basically, to get this result, you need to again use the [[ ]] with the page name you are linking to typed in first, with the link text you want to appear in your paragraph second.
Does that make sense?
Related posts
Wednesday, June 6th, 2007
This is a pretty easy one. I have been playing around with my Wiki a bit here and there and have discovered a few things.
One of the most recent is how to create an article, or page (well, that’s what it turns into anyway). Basically, you can either do this as a member of the website or not. Anyone has the ability to create an article. If you register as a member of the Wiki, you have many more options, such as protecting, deleting, moving and watching an article. Also, I think it gives it more authority, but I may be wrong.
The easiest way to create an article is to do a search for the topic on what you want to write about. If a similar article exists, you can edit that one or you can create a new one. If one is not there, you will see a line at the top of the page body that says, “There is no page titled “XXXXXXX”. You can create this page.” Just click the “create this page” link and go to town.
In the white box, place your text and whatever else you want (to be discussed later). Then, at the bottom of the page body, write a short summary of what your article is about. This will help things later on when other people are adding and editing.
Related posts
Saturday, June 2nd, 2007
Yeah, this took me a little while to figure out. I started last night and noticed that the clock was ticking rather quickly. I had to move away from the computer…MOVE AWAY FROM THE COMPUTER.
One of the most important things you can do for any website is to create pretty URLs for your pages, or what MediaWiki calls, “Short URLs.” The technique basically turns…
http://www.dotwiki.us/wiki/index.php?title=Main_Page
into something more like…
http://www.dotwiki.us/wiki/Main_Page
You get the idea of how short urls get rid of all the ugly ?, = and &. URLs like that are hard to remember and search engines just don’t like them. Google recently mentioned that they treat both ugly and pretty URLs the same, but I haven’t found any evidence of that.
So, how to get pretty URLs (or short URLs) in your Wiki? I will tell you that I tried each and every one of MediaWiki’s suggestions and none of them worked for me. It’s strange, because it almost seems like there is no continuity between the developers of this software. There are so many suggestions for so many things. I suppose that might be because of the various server configurations of so many users.
I searched in many web forums out there and came up with a solution that worked for me. I found it here and this is it…
My wiki is installed in /public_html/wiki and the clean url for my main page now comes up as http://www.mysite.com/wiki/Main_Page.
I added the following to .htaccess file located in /public_html.
RewriteEngine on
RewriteCond %{REQUEST_FILENAME} !-f
RewriteCond %{REQUEST_FILENAME} !-d
RewriteRule ^wiki/(.+)$ /wiki/index.php?title=$1 [L,QSA]
I also changed (or added) the following in LocalSettings.php located in /public_html/wiki.
##### ADDED FOR SHORT URLS #####
$wgScript = "$wgScriptPath/index.php";
$wgRedirectScript = "$wgScriptPath/redirect.php";
$wgArticlePath = "$wgScriptPath/$1";
##### ADDED FOR SHORT URLS #####
Thanks to trailville for this one. Just remember, when using this method, be sure to only have one .htaccess file, and keep it located in the root directory.
Related posts
Saturday, June 2nd, 2007
After I got the MediaWiki software all installed on the DotWiki.US domain and looking semi-like Wikipedia, my first task was to change the logo in the upper left hand corner.
I didn’t think this was going to be much of a challenge for a seasoned web developer, such as myself. Well, I must say that it took a little time to figure out the system.
Ok, here it is…after you install the Wiki, a file called “LocalSettings.php” is created on the server. You need to download that to your local computer in the proper directory (in this case, the “wiki” directory). You will want to do that to work on it, as well as for backup. MediaWiki has a good line on its configuration page…
This (page) is an index of all supported configuration options based on the DefaultSettings.php file. Never edit DefaultSettings.php; copy appropriate lines to LocalSettings.php instead and amend them as appropriate.
The reason you don’t want to edit DefaultSettings.php is because you will write over all your edits when you upgrade the software. Also, it would be quite amateur to edit such a file when there is another one created just for that purpose that won’t be overwritten. Any variable that you copy and modify from DefaultSettings.php and paste into LocalSettings.php will be given preference in LocalSettings.php, so don’t worry about leaving the variable hanging out there in DefaultSettings.php. Just copy it and ignore it.
So, back to how to change the logo. Well, the MediaWiki instructions say this…
The logo that appears in the top left of each page is determined by the $wgLogo configuration setting in the LocalSettings.php file. To change this you simply need to change the value of $wgLogo to point to the URL of your own logo image. You can upload a file via the wiki and use that address (which allows it to be replaced easily, so you may want to protect the page if you use this method) or use an image uploaded to your server via other means. Caution: It is possible to simply overwrite the default logo installed with MediaWiki, but this is strongly advised against, as an upgrade may end up overwriting it or change the default location of this file. Tip: The logo image should be 135 pixels square.
Here is the first problem for the few of us who may have a hangover from the night before…there is no $wgLogo variable yet in the LocalSettings.php file. You need to go into DefaultSettings.php and find that variable. Then, copy the whole line into a commented area of LocalSettings.php. Also, what exactly are you supposed to change? I knew I was supposed to upload a new image, but where and was I supposed to create a new path? I fiddled with it for a while and came up with this…
### ADDED TO CHANGE WIKI LOGO ###
$wgLogo = "http://www.dotwiki.us/wiki/wiki_logo.png";
### ADDED TO CHANGE WIKI LOGO ###
PS – I changed the actual logo directory above from the real directory.
That’s all there is to it. You need to take out some of the other stuff in the original line and modify it like above. I like to comment to hell out of things like this to keep them very visible.
Related posts
