1. Google's ranking algorithm
2.How to write an article to increase your web site popularity
3. Take Exception to Errors

1.RANKING HIGH AT GOOGLE When Google went online in late 1997, it entered a market that was already filled by several other competitors. To mention a few, Altavista, HotBot, Excite, Infoseek and Lycos had all managed to claim their own share of the searching industry. Despite being late in the game, in four years Google has managed to secure a place for itself in the search engine world. While many of the search engines that used to dominate the field have now fallen or become shadows of their former selves, nothing seems to be able to stop the success of Google. After gaining the position of providing secondary results to Yahoo from Inktomi in 2000, Google broke into the big league. Claiming to serve over 70 million searches each day through its own site and its affiliates, Google can arguably be called the most important and powerful search engine of today. What has made Google grow so fast? Will they be able to continue their outstanding performance? Is it possible that they will be able to reach profitability without having to stuff the pages with advertisements like other search engines have done? Those are interesting questions and I could offer a theory or two to answer them, but I won't. What you and I are really interested in is how to grab our own small share of the enormous amount of traffic Google sends out to sites that are listed in its index. That is why the only thing I'll try to do in this article is to give you some insight on the ranking algorithm used by Google. Such knowledge is vital, because without it you will be unable to rank highly on the index and shall never receive the amount of traffic you had wished to. Random thoughts before we begin The most cautious (or paranoid) of you might have already started to wonder why I'm willing to share my views on the algorithm. After all, detailed information on the ranking methods used by different search engines can rarely be found on the web. Most search engines attempt to protect their secrets as well as they can, but occasionally someone spots a pattern and is able to "crack" the algorithm at least partially and is able to send his sites rocketing to the top. However, these persons don't usually share their secrets with others. Why should they? In these days, the Internet is not the hippie land of flowers and love anymore. If you know how to secure good positions in the search engines, your site will get hits. If you are able to get hits, you will be able to make a nice amount of money. You'd have to be a fool to tell anyone about your experiences with the algorithms of search engines, because if you did, you'd suddenly find out that someone used your own weapons against you and dropped your pages out of the top ten. So why the heck am I willing to give you some guidelines without charging you for it, like some of the other sites do? Well, general search engine optimization info is commonly available on several sites for free, but like I said, it is very difficult to find up-to-date information on the algorithms of specific engines. I'm kind of flattered with the thought of doing something fairly unique, offering information that very few others are willing to share. If my ideas about Google's algorithm are even remotely correct, I have achieved this goal, even while it probably means that I'll have to fight a more difficult battle to rank my pages highly at Google in the future as this information keeps spreading. On the other hand, if I'm wrong, I can always say that you got exactly what you paid for. In addition to the above, I have other reasons as well. Frankly, I'm sick and tired watching people spend tens or hundreds of dollars to buy books or subscriptions to web sites that promise to reveal all the ranking secrets you could imagine. Don't get me wrong, if the information is accurate, paying for it is a honest deal, but too often you notice that you paid for something you could have (or should have) been able to get for free. And what about those of us who just aren't able to pay? Throughout the history of the web, the search engine optimization game has become more and more difficult for the small guys as the Internet has grown and advanced. This article is my attempt to level the playing field a little. What makes me the expert on this subject? Who am I to stand up and paint myself as an authority? To tell you the truth, not much. I don't work for the search engines and I don't have any secret contacts at Google that would be willing to give me the details of the algo. But I have achieved Top 10 rankings on competitive 2-word keyphrases with around 500,000 returns, which isn't a bad achievement in my book, especially while some of these words are often targeted by those who do posess fairly strong knowledge about search engine optimization. In any case, I'm not attempting to say that I have the best or most detailed information about this subject. All I can offer is to share what I know and hope that it will be of use to you. OK, you've probably heard enough talk without any hard facts. Let us begin. Ranking high at Google - key number one While there are numerous things measured by the Google algorithm, one thing seems to outweigh every other aspect. I'm talking about listings in the Open Directory Project. Google seems to heavily favor sites and pages that are listed in there. At the very least, you will have to be able to get your root/index page into the ODP. Attempt to include your most important keywords in the title and in the description you submit to the ODP. Having these words in the name of the category you're submitting to or in the URL you submit are also things that might have a positive effect, but I am unsure whether they produce a significant benefit or not. After you have been able to squeeze your index page into the directory, try to do the same to as many of your subpages as you can. ODP's rules state that in most cases, they will only list one page per site, but I've seen plenty of sites that have at least five subpages listed. Be careful while doing this, because excessive submitting can in extreme cases result in all of your pages being dropped from ODP and your site banned for life. The minimum requirement is to make sure that each page has plenty of useful, unique content that is relevant to the category you are submitting to. It might also pay off to keep a brief "cooling off" period in between submissions. Never, ever even attempt to get all of the pages on your 200-page site into ODP. Again, include your most important keywords both in the title and the description you submit to the ODP. For example, if you sell cars in your online store called "Auto Shop", have a subpage about Ferrari Testarossa and you want it to rank highly for those words, the title and description you submit to ODP should be something like: Title: "Auto Shop's Ferrari Testarossa page" Description: "Read about the history of Ferrari Testarossa, learn about its driving characteristics, visit a gallery of pictures or buy the thing!" Got it? For each page, select one unique keyphrase, get it into the title and the description and submit. Choose the keyphrase carefully, because once you've submitted, it can be difficult to change the information you have entered. Repeat this process as many times as you dare, selecting content-rich pages from your site and submitting them into different categories. This is a case of greed versus fear - if you're too frightened to try, you'll never get anything. But if you let your greed push your brains into the background.. you'll lose everything you already had. Should you want to get further details on submitting your website to the ODP, simply read my article about the subject. Ranking high at Google - key number two At this point, you hopefully have at least one, but preferably a couple listings at ODP with perfect descriptions and titles. The next part is to optimize the HTML code of the pages to match Google's algorithm as well as you can. While I believe that the ODP listings are the most important factor in the ranking, a completely unoptimized page that is listed in ODP can certainly be beat by a well-optimized page that is not in it. Of course, the best combination is a page that is both optimized and listed in the directory, what is exactly you should shoot for. OK, let's take a look at the various areas of page optimization for Google: Title: The keyword or phrase should be included in the title of the page. However, it is probably best to include other words in addition to the keyword as well. For single keywords or two word phrases, I'd consider a title of 3-5 words in length to be the best choice. Headings: Placing the keyphrase in a H1 or H2 heading at the very beginning of the page seems to work well. I have seen pages that rank high without headings, but it would seem to me that a good heading makes the job a bit easier. For the heading, I generally use just the keyword or keyphrase without adding any other words into it. If the page in question is a very long one, using a H3 heading with the keyword in it every now and then to retain the focus doesn't seem to hurt. Density: Google doesn't seem to be too picky about keyword density, just as long as the keyphrase is found often on the page. I myself tend to aim at around 9-10% as measured with this tool, but a slight deviation from it either upwards or downwards shouldn't cause great problems. In many cases, Google seems to tolerate and even like very high keyword densities. The page should be somewhat "front-heavy", meaning that you should work the first instance of the keyword somewhere very near to the beginning of the page and make it appear once or twice fairly close to this first keyword, scattering the rest across the page. Special words: Including the keyword in link text or in bold text does seem to give a slight advantage, but is not mandatory in my opinion. If I would have to choose between the two, I'd see using the keyword in link text as more important than using it in bold. Meta tags: Not useful with Google, but you won't get into trouble for using the standard keyword and description tags either. Include them or leave them out, your choice. Link popularity: As said, ODP links are gold, but links from other respected sources, especially Yahoo, can be very valuable as well. Links from normal pages, if you have a large number of them pointing at the page you're optimizing, will provide a good edge against the competition. Themes: Haven't seem them playing a part in the Google algorithm, except if you count relevant link popularity as a part of themes. Click popularity: Not in use. There you have it, the outlines of the Google algorithm as seen by me. Not very complicated, is it? I hope that the information you've read has been detailed enough to give you some ideas on how to improve your ranking and get more traffic to your site. The next project I'm planning is to examine Altavista's and FAST's algorithms closer, but I'm currently a bit in the dark as to what comes to those. If I get them nailed one day and if the feedback from this article is positive, you might get to read an article about them sometime in the future. Only time will tell. 2. HOW TO WRITE AN ARTICLE TO INCREASE YOUR WEB SITE POPULARITY Quite frankly, it's hard to quickly increase website traffic. When you submit a site to a search engine or a directory, it takes a while before it gets listed and the effects start to show. However, if you happen to be a webmaster of a new site, you know that it is difficult to wait for the traffic to start growing. There's nothing that would sound better than receiving a few hundred or thousand visitors within the next day or two, but how are you going to make that happen? While there are many possible solutions, the best way to increase website traffic for a brief period of time is to write newsletter articles. All that is required is an idea, decent writing skills, a couple hours of time and someone who would be interested in publishing your work. Because there is no money involved, this promotion technique is especially suited for small business owners who don't have a large marketing budget. The resource box is the key Newsletter editors are always looking for good content. Especially when their newsletter is published daily or weekly, it's hard for them to produce all the content they need on their own. However, while they admit that they need the content, most of them can't afford or simply don't want to pay for it. Because too many people aren't willing to write articles without any compensation, someone invented a clever arrangement: In exchange for allowing his content to be published for free, the author is allowed to place one small advertisement in the newsletter without charge. These advertisements are placed immediately below the article and are called "resource boxes" or "bylines". A typical resource box contains some information about the author and a link to his site. To give an example, here's one of my own: But is there any point in writing a full article just to get one 4-6 line ad? There probably wouldn't, if you were limited to publishing it in just one newsletter. However, usually you'll only give newsletters a right to use your article instead of transferring its ownership. This makes it possible to write a single piece and then have it published by several ezines. A good article may be featured in more than ten newsletters, thus allowing your 4-6 line ad reach tens or hundreds of thousands of subscribers and increase your website traffic substantially. As this one ad is the only compensation you'll get from your article, it has to be a good one. No matter how good your articles are, they won't produce too much traffic if your resource box is unattractive. To improve the effectiveness of your resource box, make sure that it Mentions the address of your website, preferably in the form http://www.mysite.com/. Unlike simply using your domain name "mysite.com", the longer "http://www.mysite.com/" allows users of Outlook Express and several other E-mail programs to click on the link rather than forcing them to open a new browser window and cut & paste the address. Doesn't have any affiliate URL's, ie. http://www.affiliateprogram.com/index.htm?affiliate_id=affiliate . The people who have been on the net for a while generally tend to avoid clicking at long URL's that look like affiliate links. If you really want to use such a link, create a page, for example "program.html" on your site that is nothing more than a JavaScript redirect to the affiliate program's address. Then simply use http://www.mysite.com/program.html in your resource box. Has been tested. If you write often for the same newsletters, it wouldn't be a bad idea to create several different resource boxes and compare their performance. Well-planned resource box testing enables you to get as many clickthroughs from your articles as possible. Some marketing required When you decide to start writing articles, the first step is to find as many newsletters as possible that are dedicated to the topic you want to write about. There are several methods you can use to accomplish this. Start out by visiting the leading sites about your topic and see whether some of them publish newsletters. Then use the search engines and ezine directories to find more of them. While searching for newsletters that might be willing to publish your article is not much fun, it's a very important job. Try to find 10-20 medium/large newsletters and absolutely refuse to quit until you have done so. The reason why this is vital for your success is that most of the work involved in this promotion method comes from writing the article itself. Once that is done, sending the article to a hundred editors instead of five will not take much more time, but it will explode the amount of visitors you'll receive. After you've found enough newsletters that accept articles from readers, arrange them into a list based on how many subscribers they have. Many newsletters publish the current number of subscribers in each issue, but if some of them don't, try to make an educated guess based on how professional they seem. When you're done, send the article to the first newsletter on your list, then wait a couple of days for their response before moving on to the next one. Granted, it would be easier to simply send your article to every newsletter at the same time, but it isn't wise for two reasons: Large newsletters tend to only accept articles that haven't already been widely published, while most smaller ones are not as picky. If you'd send your article to everyone simultaneously, two or more large ezines might publish it at the same time. This makes the editors of said ezines look bad, which in turn can result in them forwarding your next article submission straight to the trashcan. When you've sent to your article to all of the newsletters/ezines on your lists, it's beginning to approach the end of its lifespan. The last step is to send it to various lists and sites that distribute free articles, where it can be picked up by owners of smaller ezines that you don't know about. There are a large amount of these lists, but a few of the most important ones that accept nearly all sorts of articles are Article Announce: http://www.web-source.net/articlesub.htm IdeaMarketers: http://www.ideamarketers.com/ Publisher Network: http://groups.yahoo.com/group/publisher_network Does it increase my traffic enough to pay off? It's hard to answer this question, because everything depends on how good your article is and how well you are able to market it. In the best possible scenario your article will be published by several ezines that have 100,000+ subscribers each, resulting in >5000 visitors to your site. On the flip side of the coin, it's possible that nobody will publish your article and you won't get a single hit. My personal experience is that a reasonably well-written article that is actively marketed usually produces around 500-2500 unique visitors. If you decide to try your hand at writing articles, remember that the results tend to improve as you become more experienced. Should your first pieces be rejected by most newsletters and only produce a handful of visitors, don't get depressed. It takes a bit of practice, but it is worth it in the long run. 3. TAKE EXCEPTION TO ERRORS Know the difference between an error and an exception-and what to do when you catch one. by Josh Street Many programmers don't realize there's a difference between an error and an exception, and the distinction can have important implications for how your code behaves when things go wrong . As Mary Campione writes in The Java Tutorial, "An exception is an event that occurs during the execution of a program that disrupts the normal flow of instructions." According to the American Heritage Dictionary, an error is "The act or an instance of deviating from an accepted code of behavior." So, deviation, disruption, what's the difference? Think of it this way: If you're driving down the road and someone cuts you off, that's a disruption. If the car doesn't start, that's a deviation (unless it's my car, in which case, that's what we call normal). What does this have to do with Java? Lots. Java has a fairly interesting hierarchy of errors and exceptions Yes, that's right: All those snippets of code using try {} catch(Exception e) {} find only half of your errors. But whether you should try to catch Throwable depends on what you plan to do with it once you catch it. A quick perusal of Error's children shows classes with names like VirtualMachineError, ThreadDeath, and LinkageError. Before you try to catch these guys, make sure you're actually handling them, because these are serious problems, and therefore errors. But isn't a ClassCastException an error? Not really. A ClassCastException-or an exception of any kind-is just the VM's way of letting you know that you, the developer, have made a mistake and now have a chance to correct it. On the other hand, an error is the fault of the VM (typically, although it could be any system-level service). To quote the JavaDoc entry for Error: "An Error is a subclass of Throwable that indicates serious problems that a reasonable application should not try to catch. Most such are abnormal conditions." So, errors are pretty powerful, and the subtleties involved in handling them are beyond the understanding of the average developer (not you, of course). And what if you're dealing with a fairly low-level piece of work that you feel could cause an event worthy of being called an error? Take Exception to Errors (Continued) First, remember that errors are thrown just like exceptions, with one difference. A method that throws an error doesn't need to declare that it does so (in other words, the exception is unchecked): public void myFirstMethod() throws Exception //Since it's an exception, I have to declare //it in the throws clause { throw new Exception(); } public void mySecondMethod() //Because errors aren't supposed to occur, you //don't have to declare them. { throw new Error(); } Note that several exceptions are unchecked, and as such, behave like errors: NullPointerException, ClassCastException, and IndexOutOfBoundsException are all subclasses of RuntimeException. RuntimeException and all of its children are always unchecked. So what should you do about all these nasty unchecked exceptions? You could catch the exceptions in the method where they might occur, but this is a haphazard solution. Doing so solves one problem, but it leaves the rest of the code open to disruption by other unchecked exceptions. There's a better way, thanks to a nifty little method provided by the ThreadGroup class: public class ApplicationLoader extends ThreadGroup { private ApplicationLoader() { super("ApplicationLoader"); } public static void main(String[] args) { Runnable appStarter = new Runnable() { public void run() { //invoke your application (i.e. MySystem.main(args)} } new Thread(new ApplicationLoader(), appStarter).start(); } //We overload this method from our parent //ThreadGroup , which will make sure that it //gets called when it needs to be. This is //where the magic occurs. public void uncaughtException(Thread thread, Throwable exception) { //Handle the error/exception. //Typical operations might be displaying a //useful dialog, writing to an event log, etc. } This technique rocks. Think of all those times you perform an operation in your GUI, and an unchecked exception happens. Often, you leave the GUI in an unusual state (dialogs still open, buttons disabled, cursors in the wrong state), but with this technique, you could return the GUI to its natural state, inform the user of an error, and feel good about yourself because you wrote a high-quality application. But this technique isn't just for GUIs. Server applications prone to excessive resource usage can use this technique to free resources at a global level, often preventing the VM from entering an unstable state. Catching errors early and often, and dealing with them smartly, can make the difference between a great programmer and an average one. And by virtue of your having read this far, we know which one you want to be. About the Author Josh Street is an architect at Bank of America, where he is responsible for developing e-commerce solutions. Reach him at rjstreet@computer.org.