rewrite or 301 redirect?

[inertia]

As i think about this question i cant help but think im missing some fundamental point, but here goes.

rewrite or 301 redirect? Some situations need one and some need another but i am a little confused as to which is needed when? For example, I've just started work on a clients site and he has the old www versus non-www thing. Should i use rewrite or redirect? He has links that go to domain dot com and domain dot com/index etc. Should i fix the links and 301 /index to the root domain or rewrite?

Most of the standard coding for fixing these problems is rewrite but i always think, why not use redirect? Its a lot simpler?

What am i missing? As far as i can see rewrite should be used for unfriendly URLs but why use it for redirects?

[activeco]

If you have just a few static redirections, then using simple redirect is the way to go. For all 3xx status codes it doesn't really matter which method you use.
However when you deal with many redirected pages, it is much easier to make some rules using rewrite way, where you can dynamically match a large number of different conditions in a single or few lines.
Also, rewrite can make all internal changes non-visible to users, so providing opportunity to e.g. make SEO friendly url's, while maintaining or even changing internal structure (among many other purposes).

In your case it is perfectly OK to use simple Redirect 301

[wige]

I am not sure quite what you are asking. If you are asking when you should use an internal rewrite as opposed to a redirect, redirects should be used when something on your site is being changed - pages are being deleted or renamed, or you are converting from PHP to ASP, etc. Rewrites are invisible to the user, and should be used when you want to make a page URL easier to remember or link to. Or, if you are asking which method (mod_rewrite or the Redirect directive) should be used to perform redirects, the answer is a bit more complicated.

In general, you would want to avoid using mod_rewrite unless you will be using the mod to convert friendly urls, because when you enable to the mod_rewrite engine you can slow the performance of your server slightly. The rewrite engine adds (sometimes noticeable) overhead to the server as the server has to run complex regex analysis on requested URLs.

If your server only requires basic rewrites, use the Redirect and RedirectMatch directives. You can create redirects with the same level of complexity as you can using mod_rewrite using the RedirectMatch directive, and you can even perform some more complex functions easily. Need to make a directory completely inaccessible? Use:

RedirectMatch 404 /blocked/directory/.*

Folders can be redirected just as easily:

RedirectMatch 301 /oldfolder/(.*) http://server.tld/newfolder/$1

Generally, mod_rewrite should only be used for internal rewriting, where you are converting a URL that the user sees into a URL that your scripts can understand. This can make troubleshooting easier (since when you look at the configuration you will know that all mod_rewrites are rewrites and all RedirectMatches are redirects) and can simplify updates and possibly reduce server overhead.

Note, if you are using mod_rewrite to create internal redirects, I would actually recommend combining the two methods. Use the RedirectMatch 404 method above to block access to the folder containing the actual scripts, and use the mod_rewrite to translate the friendly URLs into URLs your scripts can understand. This will help protect against possible duplicate content issues.

[inertia]

One of my chief confusions comes from the advice given to people about canonicalization (cant we find an easier word than this?lol). People are often advised to sort out www versus non www problems with rewrite. What would be wrong with doing a 301 redirect:

redirect 301 domain.com www.domain.com

The same with the /index.htm issue:

redirect 301 domain.com/index.htm www.domain.com

Do these work?

[activeco]

I thought I answered that. Do you have me blocked or so?

[inertia]

Yeah, you did cheers. Just clarifying.

[dubbing]

that was a LOL of the year ?

[Jean-Luc]

Originally Posted by inertia

What would be wrong with doing a 301 redirect:

redirect 301 domain.com www.domain.com

The same with the /index.htm issue:

redirect 301 domain.com/index.htm www.domain.com

Do these work?

They do not work. The domain name cannot be included in the first URL of a "redirect 301".

Jean-Luc

P.S. I did not see any answer on this elsewhere in this topic.

[inertia]

Yeah! Good point! Erm, so those codes dont work. Thats sort of what i was asking. So redirects in this form dont work they have to be formulated relative to absolute.

So rewrite is the only way to solve these canonicalization issues?

Sorry for my lack of brain power today. It is explained in this thread http://www.webproworld.com/breakroom...thday-you.html

[activeco]

LOL indeed.

Well, my whole point was around the basic question "rewrite or 301 redirect?", which I understood as the general question about preferences in using it, without paying too much attention to your details or validity of your code.
And yes, Jan-Luc is correct. The first part of all mod_alias directives (including Redirects) refers to url-path only.

[activeco]

But...
Since my statement was firm that using Redirect is possible even in that case, after all it gave me some food for thought about how to do it. Sure enough, I soon came up with the solution.
YES, it is possible to redirect www to / or the other way around by using Redirect directive only.
So you are witnessing the first outline of that technique ever. It's not the most elegant, but it works.

Assuming Apache with virtual domains (I guess a typical server) and access to the server.

As most of the www subdomains are aliases for the root domain it is necessary first to check/change it in DNS records.
Instead of "www CNAME domain.com." make it "www IN A xxx.xxx.xxx.xxx".

The second step is making additional virtual domain with a few lines in httpd.conf (or in include file):

<VirtualHost xxx.xxx.xxx.xxx:80>
    ServerName www.example.com
    DocumentRoot (the same as root domain)
<Directory />
Redirect 301 / http://example.com/
</Directory>
</VirtualHost>

And that's it. Restart the server and be happy redirecting.

#Be sure to comment out or delete ServerAlias www directive in / section:
### ServerAlias http://www.example.com

[sofomor]

by inertia: rewrite or 301 redirect?

Simply correct the links first (point to either /index or / ) and complete the "consistency cycle". This www vs non-www is mostly related to "home page" of a website. Re-direction is performed when "extensions" have changed or format has been modified. Rewrite is not a subject here

I suggest pay attention to details (internal links) and re-submit a Sitemap. I would suggest using website grader
to check "re-direction" issues and suggestions.

Use this code:

Redirect 301 /index.htm http://www.yr-domainexample.com/

Thank you

[wige]

Originally Posted by activeco

So you are witnessing the first outline of that technique ever. It's not the most elegant, but it works.

Bah! You beat me to it!

But I would suggest using RedirectMatch 301 /(.*) http://www.domain.tld/$1 instead. This will mimic the mod_rewrite method more effectively, as it will redirect the user to the correct page, rather than the homepage of the www site.

[activeco]

Originally Posted by wige

Bah! You beat me to it!

But I would suggest using RedirectMatch 301 /(.*) http://www.domain.tld/$1 instead. This will mimic the mod_rewrite method more effectively, as it will redirect the user to the correct page, rather than the homepage of the www site.

Ha, ha, OK, you get the Silver Medal!

I agree with the RedirectMatch and after a bit of further thinking I can't recall anymore why I suggested changing DNS records. It is not needed at all.
I have put the redirect into the <Directory> section just in case of some later additions, which would not be supported in <VirtualHost>, otherwise that too was not needed.
So, to summarize the shortest way:

#(From www to /)

<VirtualHost xxx.xxx.xxx.xxx:80>
    ServerName http://www.example.com
    DocumentRoot (the same as non-www domain)
RedirectMatch 301 /(.*)? http://example.com/$1
</VirtualHost>

I actually like it.

[wige]

Just to add to activeco's solution a bit more... If you put this virtual host before the virtual host entry for your www domain, it becomes the default for the entire server, basically redirecting every single possible subdomain or any domain name you have pointing to that IP address to your main site.

So, for example, if you have the domain "example.com", and want to use "www.example.com", you would create the following virtual hosts:

# Default catch all - redirects everything not covered in another VirtualHost section
<VirtualHost xxx.xxx.xxx.xxx:80>
   ServerName http://example.com
   DocumentRoot (the same as non-www domain)
   RedirectMatch 301 /(.*)? http://www.example.com/$1
</VirtualHost>
# Your main domain, the www. subdomain
<VirtualHost xxx.xxx.xxx.xxx:80>
   ServerName http://www.example.com
   DocumentRoot (the same as non-www domain)
   # Your directives here
</VirtualHost>

Now, lets say you buy "sample.com" for whatever reason, but don't have any content for that domain yet. Simply point the domain to the server's IP address, and traffic is automatically redirected immediately to your main site, giving your main site the full benefit of any pre-existing links to that domain. You don't need to change a single setting on the server. When you are ready to develop a site for sample.com, simply create a new VirtualHost section, with the ServerName sample.com and an appropriate DocumentRoot. (Note, sample.com will still redirect to example.com, but at this point you would replace the RedirectMatch with a RewriteRule that will handle non-www to www conversion for all domains on the server.)

[kevsta]

this is the code i generally use on the canonicalization issue and to lose the index.html in favour of domain.com/

RewriteEngine on
rewritecond %{http_host} ^domain.com [nc]
rewriterule ^(.*)$ http://www.domain.com/$1 [r=301,nc]

RewriteCond %{THE_REQUEST} ^[A-Z]{3,9}\ /.*index\.html\ HTTP/ 
RewriteRule ^(.*)index\.html$ http://www.domain.com/$1 [R=301,L]

is there a better way of doing this that doesnt tax the server so heavily?

(in simple language please, i didn't write it myself )

[activeco]

Originally Posted by wige

RedirectMatch 301 /(.*)? http://example.com/$1

Another Speedo
I knew you meant: RedirectMatch 301 /(.*)? http://www.example.com/$1

[wige]

Yup, thanks. Fixed in my previous post.

[activeco]

Originally Posted by wige

But I would suggest using RedirectMatch 301 /(.*) http://www.domain.tld/$1 instead. This will mimic the mod_rewrite method more effectively, as it will redirect the user to the correct page, rather than the homepage of the www site.

Sorry for coming back to this.
I agreed with the suggestion, but as I remembered doing something similar in the past with Redirect only, it looked like this case did not need RedirectMatch.
Redirect still transfers the whole request to the new destination.
So, to save even more typing and processing power "Redirect 301 / http://example.com/" should suffice.

[Szanyijr]

When you want to fix Canonicalization you have to use 301 redirect. It's neccesary 'cause the search engines handle yoursite.com, www.yoursite.com and yoursite.com/index.html like three different pages and rank them different as well. To correct this you have to do 301 redirects what the search robots understands.

Example:

RewriteCond %{HTTP_HOST} ^yoursite\.com [NC]

RewriteRule ^(.*) http://www.yoursite.com/$1 [R=301,L]

RewriteRule ^/index.php http://www.yoursite.com/ [R=301,L]

Cheers,

[inertia]

Read the thread dude!

[slobjones]

Originally Posted by Szanyijr

When you want to fix Canonicalization you have to use 301 redirect. It's neccesary 'cause the search engines handle yoursite.com, www.yoursite.com and yoursite.com/index.html like three different pages and rank them different as well. To correct this you have to do 301 redirects what the search robots understands.

Example:

RewriteCond %{HTTP_HOST} ^yoursite\.com [NC]

RewriteRule ^(.*) http://www.yoursite.com/$1 [R=301,L]

RewriteRule ^/index.php http://www.yoursite.com/ [R=301,L]

Cheers,

Szanyijr, this isn't working for me. I've tried removing other rules from my htaccess file that could potentially conflict, but index.php still goes to index.php.

Could something be missing from this rule set?

Thanks.

[MuNKyonline]

Originally Posted by slobjones

Szanyijr, this isn't working for me. I've tried removing other rules from my htaccess file that could potentially conflict, but index.php still goes to index.php.

Could something be missing from this rule set?

Thanks.

I think what you want is this which also accounts for the trailing slash version of a URL ...

########## Require the www to avoid cannonicalization issues ###
RewriteCond %{HTTP_HOST} ^domain.com [NC]
RewriteRule ^(.*)$ http://www.domain.com/$1 [L,R=301]

########## Redirect index.php to / ##########
RewriteCond %{THE_REQUEST} ^[A-Z]{3,9}\ /.*index\.php?\ HTTP/
RewriteRule ^(.*)index\.php?$ http://www.domain.com/$1 [R=301,L]

########## Require to add trailing slash if not present to avoid cannonicalization issues ###
RewriteCond %{HTTP_HOST}   !^www\.domain\.com [NC]
RewriteCond %{HTTP_HOST}   !^$
RewriteRule ^/(.*)         http://www.domain.com/$1 [L,R]

[slobjones]

Originally Posted by MuNKyonline

I think what you want is this which also accounts for the trailing slash version of a URL ...

########## Redirect index.php to / ##########
RewriteCond %{THE_REQUEST} ^[A-Z]{3,9}\ /.*index\.php?\ HTTP/
RewriteRule ^(.*)index\.php?$ http://www.domain.com/$1 [R=301,L]

Here's the solution I found yesterday while nosing around the mod rewrite forum. It varies a bit from MuNKyonline's code.

RewriteCond %{THE_REQUEST} \ /(.+/)?index\.php(\?.*)?\  [NC]
RewriteRule ^(.+/)?index\.php$ http://www.domain.com/%1 [R=301,L]

[slobjones]

The redirect is working. But now I find that my sitemap generator produces both the old urls and the redirected urls. Google doesn't like this.

For example, I now have all variations of "mysite.com" and "mysite.com/index.php" in my sitemap, where, pre-redirect, I had only the latter.

I've been advised to change all my old internal links to reflect the redirect. But this seems to defeat the purpose of the redirect.

How should I proceed?