Techniques to use and techniques to avoid

Don't use "refresh" to redirect

If you want http://www.example.org/foo to actually display what's at http://www.example.org/bar you should not use "refresh" techniques like :

<META HTTP-EQUIV=REFRESH CONTENT="1; URL=http://www.example.org/bar">.

Why? because it could break the "back" button. Imagine that the user presses the "back" button, the refresh would work again, and the user would bounce forward. The user will most likely get very annoyed, and close the window, which is probably not what you, as the author of this page, want.

Use HTTP redirects instead

When using the "refresh" meta-tag to create a redirection, what we have is a specific instruction within the document. The User Agent (be it a browser or the Markup Validator) is expected to download the page, look at its contents, find the "refresh" instruction, wait the specified amount of time (which could just happen to be "0" seconds for an "immediate" refresh, but really could be anything), and then proceed to the new address.

A "HTTP Redirect" on the other hand acts much more directly because it is done within another layer. When the User Agent (i.e. a browser or the validator) first contacts the server and requests the document, the _server_ itself, having been configured to redirect the document to another address, replies to the user-agent that it should instead look at the new address.

A "HTTP Redirect" is also a richer way to redirect because it gives the User Agent more information than just the new address: the server also gives some information about the purpose and type of redirection, which allows the User Agent to behave differently depending on the type of redirect. The different types of HTTP redirects are as follow (with the relevant HTTP status code sent by the server) are the Permanent Redirect (HTTP 301), the Temporary Redirect (307), and the undefined redirect (302).

Vía: www.w3.org/QA/Tips/reback »