Is there a safe way to do a page refresh?

That does not in get trouble with the SE's......

[trillianjedi]

I've got a new page on site that updates automatically every 5 minutes (showing live data).

Is there anyway I can do a META REFRESH tag to:-

a) refresh the page (i.e. not jump to another one)
b) that the search engines will be reasonably happy with?

Thanks,

TJ

[dmorison]

If you're worried about a penalty for using refresh, can you not block the page via robots.txt.

If you're updating every 5 minutes then presumably you're not very indexable anyway;

Alternatively; if you do need certain information on that page spiderable you could split the content that you want indexed and the refreshed content onto separate pages, and either display them together using <ducks>frames</ducks> or some other mechanism; with the live page still blocked via robots.txt.

You could also do a JavaScript refresh, but usual user-agent dependency warnings apply:

<script type='text/javascript'>

function doreload()
{
window.location.href = 'http://www.example.com/';
}

window.setTimeout("doreload()",300000); // 5 minutes

</script>

[StanBo]

I don't believe that SEs will be happy with the proposed script once they discovered it.

A couple of checks are to be added to any script to justify refresh in the eyes of anyone who counts your pageviews - the first one is to determine that the page in question is in the topmost window. If not - you shouldn't refresh.
The second one - to find-out whether the computer is active (doesn't run screensaver, not in a sleep mode etc.)
Without such checks, it's almost impossible to prove the information update had any chance to be noticed and benefited from. But with them implemented, you can theoretically address SE support team and discuss the possibility of such a code usage.

[killroy]

also, use the appropriate functions (like reload() / replace() ) to preserve the history order.

SN