http://www.example.com or http://example.com?

Which is better?

[spikedo55]

In my linking campaigns, I've asked people to link to my site using the format [example.com....] But lately, I'm seeing more people ask for links using this format: [example.com....]

Is one format better than the other? If so, why?

[jo1ene]

They're the same animal.

[dvduval]

No difference, but best to be 100% consistent (use either one or the other but not both).

[Receptional Andy]

Although convention dictates otherwise, [example.com...] and [example.com...] are two different sites, just as [something.example.com...] and [example.com...] are different.

However, for linking purposes, just pick one and stick with it. Using both runs the risk that some of your links won't be counted for the same site.

[trillianjedi]

Personally, I prefer with the WWW prefix. No idea why.

I do always 301 redirect the without-prefix version to it though.

No point serving a 404.

TJ

[martinibuster]

How about http*://www.example.com/

?

[edited by: martinibuster at 4:49 pm (utc) on June 7, 2004]

[Receptional Andy]

Good point martini (I assume you meant the trailing slash rather than the question mark)

I've got into the habit of always including the slash, although I missed it in this thread ;)

[spikedo55]

Thanks for the feedback.

Why include the slash?

[martinibuster]

Why include the slash?

Less server load.

[pete_m]

Are we sure that the trailing slash is required when requesting the root document?

I know that it matters when we request a sub-directory - the server sends a 301 redirect to the full slashed-address.

Have just tried this using Live HTTP Headers on Firefox, and I didn't receive a 301. Is Firefox just being clever and automatically inserting the slash? If so, do all browsers do this?

[pageoneresults]

It appears that the standard practice of www.example.com/ would be the preferred method. A simple 301 for example.com to www.example.com is suggested.

There are two reasons for requesting links without the www. One is for branding purposes, the other is for manipulation. ;)

I've not checked lately, but it used to be that some search engines would count non www links and www links differently. You could theoretically achieve two separate listings based on the linking strategies.

example.com is a sub-domain of www.example.com and as mentioned above, can have different content. I've seen many topics in the past that discussed PR values between the two and the problems that arised when using the non www method of linking.

Here is the problem with picking one and sticking to it. If you pick the non www version, be prepared to constantly track your backlinks as 9 out of 10 people are going to include the www in their link.

When choosing the www link method, you avoid the above issues of backtracking. For that 1 out of 10 that links without the www, your 301 will whisk the visitor to the correct URI. If the webmaster on the linking site is checking their outbounds, they should see the 301, investigate and update their references accordingly.

P.S. Yes, a 301 will address the issue of redirecting www requests to the non www version. But why go through all of that?

[EliteWeb]

www.hostname.com/ that is the standard method documented places.

[PatrickDeese]

I had some subdirectories on a site that my assistant forgot to include the trailing slash for and G listed both the www.example.com/dir/ and the url only www.example.com/dir - so I have made sure ever since to make sure that the trailing slash is always included.

[spikedo55]

Not sure if I understand the 301 part. Don't both the www and the non www addresses go to the same index page?

[pageoneresults]

Not to take this too far off topic, let's discuss this trailing slash issue. PatrickDeese brings up an interesting point in how the search engines are indexing URI paths.

My developer and I are working with the Google Web APIs and noticed that our on site searches were returning results with and without the trailing forward slashes. This was only relevant to any URIs that were rewritten through our ISAPI filter.

We decided to force the trailing forward slash on those URI paths. Our on site searches using the Google Web APIs now return results with just the trailing forward slash.

Typically the browser appends the trailing forward slash to root level directory pages. There are instances though where this does not occur. My testing shows that it normally occurs when you are rewriting URIs or there is something at the server level that is not reacting correctly.

More comments from a technical perspective would be appreciated. Maybe a new topic is in order?

P.S. We're finding some interesting things when using the Google Web APIs for on site searches. ;)

[andrew_m]

/ on the root level is enforced by all browsers.

The first line they all send to web servers is "GET / HTTP/1.x" -- that is space separated three-part string. There has to be something in the middle position and since domain name is not normally a part of it it can't be empty, thus it's always at least a single slash.

[pageoneresults]

Not sure if I understand the 301 part. Don't both the www and the non www addresses go to the same index page?

In most instances yes. In some instances no. Again, the non www version is a sub-domain of the www version. You can have different content residing at both.

The search engines are becoming smarter in this scenario and most are effectively catching that the two are the same. But there is a problem at hand that needs to be addressed. To be on the safe side, it is suggested that you 301 one to the other to prevent possible duplicate content issues.

In the past, the non www version and www version were returning different levels of PageRank. In most cases, this has been corrected (I think).

[Roel]

I had some issues with a site couple of months ago with the -www or +www issue. I agree that you should NEVER place links to both -www and +www versions, as GG (and others) will think that it's dup content.

I personally disable the non-www version from the start now. It somethimes takes a bit of hasseling your hosting company but eventuall they work it all out. A DNS record that points * to the non-www version is even worse. This way if someone links to abc.yoursite.com and to www.yoursite.com and to yoursite.com guess what GG will see? Dup content!

More info here:
[webmasterworld.com...]
[webmasterworld.com...]
[webmasterworld.com...]

[manwah]

in my linking strategy, 80% of my request will be www. and 20% will not use www.

Finally, I found that when I type both domainname.com and www.domainname.com have same PR. It's not for some sites