Maybe give more details about the Ajax implementation of your site. For example, does different content correspond to different URLs, or to the same URL? Have a look at your site using 'view source' in a browser with JavaScript disabled, also look at your site in a text browser like lynx.
In addition to what cristina mentioned and suggested, my experience with AJAX and search engines is that using AJAX to provide user interactivity is GREAT. Using AJAX for navigation is a BIG mistake.
But as cristina requested, more details on what your plans are would be useful. :-) Craig
> Maybe give more details about > the Ajax implementation of your site. > For example, does different content correspond > to different URLs, or to the same URL? > Have a look at your site > using 'view source' > in a browser with JavaScript disabled, > also look at your site > in a text browser like lynx.
> On Aug 2, 6:30 am, Webmaster1 wrote:
> > Hello, > > I've decided to swap from html to ajax, so can anyone help me to > > optimize ajax site.
> In addition to what cristina mentioned and suggested, my experience > with AJAX and search engines is that using AJAX to provide user > interactivity is GREAT. Using AJAX for navigation is a BIG mistake.
> But as cristina requested, more details on what your plans are would > be useful. :-) > Craig
> On Aug 2, 3:40 pm, cristina wrote:
> > Maybe give more details about > > the Ajax implementation of your site. > > For example, does different content correspond > > to different URLs, or to the same URL? > > Have a look at your site > > using 'view source' > > in a browser with JavaScript disabled, > > also look at your site > > in a text browser like lynx.
> > On Aug 2, 6:30 am, Webmaster1 wrote:
> > > Hello, > > > I've decided to swap from html to ajax, so can anyone help me to > > > optimize ajax site.
Because ajax'ed HTML is way cooler than plain HTML? Could be a huge mistake. Ajax is suitable to enhance pages, not a technique you should build sites with. Sebastian
Hi Adam, Thank you very much for your posting. It is great that new articles about AJAX will appear on the Google Blog, thank you.
Hi Webmaster1, What I meant was that you can use AJAX in various ways, as Craig and Sebastian already wrote. AJAX applications let you use JavaScript to contact the server to respond to user requests for more information by writing content to a page without the browser downloading the full page again. Google Maps and GMail use AJAX. Another example is that you could have an AJAX application for an international clock, you could select a time zone and initiate an HTTP request to display the time for that time zone.
In these examples you have HTML elements, like <div id="something"></div>, with no content pre-written in the HTML source, an event listener defined with a JavaScript function that can trigger a HTTP request to the server to write content to that div that it finds from its name given by the id attribute.
In this case, when only a part of your HTML document has content generated by the AJAX application, the HTML source of the document has enough content to appear in search results, like any other HTML document, also the HTML document has its own URL that will appear in the SERP if the document is indexed by Google.
Another way of using AJAX, that as far as I know has problems being indexed by all search engines, is to use the JavaScript generated HTTP request to dynamically present content for the whole page, or most of the page, so this content would not appear in the HTML source, you do not see much if you look at 'view source' from a browser. Some of those AJAX application use hidden iFrames to write content to a page and all you can see with 'view source' is some frames code, and sometimes multiple pages have the same URL, so they cannot be indexed with individual URLs by search engines, and it is these implementations that have problems with being indexed by search engines.
We've been playing around with trying to optimize a client's site which we used AJAX in pretty extensively (and for navigation also). So far there are over 1500 pages indexed in G. Much of them are duplicate and 'snippets' from the Javascript (Google was, believe it or not, crawling through some of them). We've been implementing 'noindex, nofollow' meta tags on the database pages, but many are still indexed. As I think Christina suggested to me before, a few lines in the robots.txt tag would probably be more efficient, but things keep changing so rapidly and there are so many other things that the programmers are constantly working on that there doesn't seem to be enough time to even bother asking them to move them to a new/ different folder. I think for now the noindex tags will start working. Anyway, at the same time we constructed a script which spits out sitemap files (in html for some search engines and an XML sitemap - http://www.sea2skyrentals.com/sitemap.php & http://www.sea2skyrentals.com/sitemap.xml). BTW, If anyone plays around on the site please don't give me #$!@ about the meta tags on the property pages as they are created by the property owners (members).
One other thing... I think our sitemap.php file might have a few too many links on it to get thoroughly crawled, but not sure - the pages are indexing somehow.
Hi Tim, One important thing is to have valid (X)HTML. I looked now at http://www.sea2skyrentals.com/ with the W3C validator http://validator.w3.org/ and it has some HTML errors, including some about <script> elements for the JavaScript code. Errors like these might make parsing of JavaScript somehow unpredictable. Also your home page uses HTML 4.01, this is OK, but I am wondering if XHTML might not be easier to parse by bots, because it makes possible use of better parsing mechanisms associated with XML. In XHTML, if correctly marked, there is better separation of the JavaScript code from the HTML content, for example <script type="text/javascript"> <![CDATA[ <!-- ... script content ... // --> ]]> </script>
> We've been playing around with trying to optimize a client's site > which we used AJAX in pretty extensively (and for navigation also). > So far there are over 1500 pages indexed in G. Much of them are > duplicate and 'snippets' from the Javascript (Google was, believe it > or not, crawling through some of them). We've been implementing > 'noindex, nofollow' meta tags on the database pages, but many are > still indexed. As I think Christina suggested to me before, a few > lines in the robots.txt tag would probably be more efficient, but > things keep changing so rapidly and there are so many other things > that the programmers are constantly working on that there doesn't seem > to be enough time to even bother asking them to move them to a new/ > different folder. I think for now the noindex tags will start > working. Anyway, at the same time we constructed a script which spits > out sitemap files (in html for some search engines and an XML sitemap > -http://www.sea2skyrentals.com/sitemap.php&http://www.sea2skyrentals.c...). > BTW, If anyone plays around on the site please don't give me #$!@ > about the meta tags on the property pages as they are created by the > property owners (members).
> One other thing... I think our sitemap.php file might have a few too > many links on it to get thoroughly crawled, but not sure - the pages > are indexing somehow.
Thanks, Christina. I can't see the errors right now because W3C says 'service temporarily unavailable', but I've brought this up before and the general consensus around here seems to be that validation isn't a big concern since we've obtained very good rankings with quite a few sites which have quite a few errors. I think it depends on the type of error (it looks like you might agree with that theory). I will mention this to the team again though.
I wouldn't mind hearing some input on that subject (validation errors) from the Google team. In fact, the last time I checked, Google had quite a few errors on their own site.
____________________
Ok... W3C is working now. Wow... over 300 errors. Just to re-check it Google still does have 50 errors (including no doctype), Yahoo has over 30, nytimes.com has over 300, & myspace has over 200. Just a few random places I've checked. I can't help but look at W3C validation as the same as critiquing a piece of artwork in a way.
Google staff are saying validation doesn't matter because they are not thinking about how badly broken some websites are. You will need to be a very good coder and very conversant with html to be able to figure out which validaiton errors are really bad, which are rather bad and which are safe to be ignored as they don't break anything. If you are able to figure those out , yeah, you can pick and chosoe what to fix. Most people can't tell one kind of error from another. They end up either ignoring them all or worrying about them all.
> Thanks, Christina. I can't see the errors right now because W3C says > 'service temporarily unavailable', but I've brought this up before and > the general consensus around here seems to be that validation isn't a > big concern since we've obtained very good rankings with quite a few > sites which have quite a few errors. I think it depends on the type > of error (it looks like you might agree with that theory). I will > mention this to the team again though.
> I wouldn't mind hearing some input on that subject (validation errors) > from the Google team. In fact, the last time I checked, Google had > quite a few errors on their own site.
> ____________________
> Ok... W3C is working now. Wow... over 300 errors. Just to re-check > it Google still does have 50 errors (including no doctype), Yahoo has > over 30, nytimes.com has over 300, & myspace has over 200. Just a few > random places I've checked. I can't help but look at W3C validation > as the same as critiquing a piece of artwork in a way.
Hi Tim, It would be difficult to quantify the exact effect of each and all (X)HTML errors in a page on the way a search engines crawls and extracts content from that page. What I meant is that if you see in the search results something puzzling, like bits of JavaScript code in some search results, the first step is to check if your pages have correct (X)HTML mark-up, and if there are HTML errrors in the <script> and the JavaScript bit, then try to correct them.
