Dan's Web Tips | Auto-Refreshing

Dan's Web Tips:

Auto-Refreshing Pages

[<== Previous] | [Up] | [Next ==>]

See also Danish translation (done by others in their own site, with my permission)

TIP: Here's how to make your page automatically load another page after a given number of seconds. Now, do you really, really need to do this?

This message will self-destruct in 30 seconds!

Only kidding, but if I was using a "META refresh" tag on this page, it actually would. In spy movies, that's a pretty neat thing, but when you're trying to read a web page it could get annoying... that's why I don't really do it here!

You've probably seen web pages that automatically move the user on to another web page, sometimes repeatedly in slide-show fashion. Here's how to do it, but I'll also explain why it's probably a bad idea in most sites.

How To "META-Refresh"

Use this syntax, within the <HEAD> section of the document:

CONTENT="30; URL=http://www.example.net/some/place/">

The contents of the "CONTENT" attribute consist of the number of seconds until the page load takes place, followed by a semicolon and a space, then "URL=" followed by the URL of the site to load. Note that the "URL=" part is within the "CONTENT=" parameter, not a separate parameter.

Since this is an HTTP-EQUIV META tag, it is actually equivalent to sending a Refresh header in the server's actual HTTP headers. If you have control of this server-level stuff, you can do refreshes directly through the headers without having to insert a tag in your HTML documents. (Though, it's possible there are a handful of browsers that don't like it that way; many years ago I observed WebTV to fail to honor such headers, but that's ancient history.)

I should note, however, that the Refresh header doesn't appear in any of the standards documents regarding the HTTP protocol, so it is technically nonstandard (whether sent as a real header or a META tag), though widely supported.

Why Not To "META-Refresh"

OK, now you know how to do it. Now here's why you probably shouldn't:

  • Not all browsers necessarily support it. Somebody using an older browser, or a browser (like Opera) that allows users to disable features like "META refreshes", might never make it to the page they're supposed to be shunted to, and thus be unable to see your entire site.
  • Some browsers may have nasty bugs that cause a page to infinitely reload without giving the user a chance to read it... this page did exactly that for me in some versions of the Mozilla browser, though its META Refresh tag seems to be valid. Use of META tags with messed-up syntax, of course, increases the chance of such mishaps happening.
  • Many users find automatic page-shunting to be very annoying. They prefer to surf the web at their own speed, under their own control, and resent sites that take this control away from them by moving them on to a new page at a speed chosen by the page author rather than when the user follows a link. Maybe the user wants more time to read the text or view the graphics on the first page before going on to the next one, or maybe he or she is bored with the page and wants to move on right away, not wait 30 seconds for the refresh to move on.
  • If you've scrolled down in a long page, while reading it or while you've left the page in your browser while you went and did something else, it's annoying to find you've been placed back at the beginning of the page because it reloaded on you.
  • The most common use of such refreshes, to show a "splash screen" at the beginning of a site before bringing the user to the real home page, is particularly annoying to users who went to the site to find information, and would prefer to get the main menu right away rather than sit through the author's attempts to "entertain" the user. This will be especially annoying if the user's already been to the site a few times, and is forced to sit through the opening animation yet again. People will probably bypass this by bookmarking the real menu page rather than the original front URL, but this isn't very satisfactory because this URL is likely to be less memorable (http://www.example.net/docs/real-front-page.html instead of just http://www.example.net/), and is more prone to change in future site redesigns, breaking all such bookmarks.
  • Also, just try to use the "Back" button to back out of a site with "refresh" pages; especially if the time interval used is very short, you may find your browser immediately starts loading the next page in the sequence, keeping you from going back any further unless you press the button really rapidly. Maybe you'd like to purposely stop people from backing out of your site (especially if you're a marketing person and your site is trying to sell something), but users who get annoyed this way are likely to have a bad taste in their mouth about your site and avoid it in the future.

Why To META-Refresh

OK, enough negativity; people say I'm too negative, and all I do is keep shooting down ideas rather than propose anything constructive. Killing a bad idea can be a productive thing if it saves everybody from the problems the bad stuff causes, but I will get a bit more positive here by listing a few places where the use of these "refresh" pages might be a useful thing:

  • Interactive sites such as online games and real-time chats, which need to load and reload pages to show what's happening in a dynamic program. HTML is not really a very good language for such interactive content (that's not what it was designed for), but in a pinch you can do quite a bit with HTML, CGI scripts, and such features as the "refresh" META tag.
  • News and sports sites with continuous updates, where you might want the page to keep reloading every 30 or 60 seconds so a user can keep seeing the latest scores, etc. (If you omit the "URL" portion of the Refresh content, the same page will be reloaded, including its Refresh tag so that the new copy of the page will itself reload in the given number of seconds, infinitely long or at least until the user leaves the page manually.)
  • Temporary pages placed at obsolete URLs to redirect users to the new correct URL, if you don't have the low-level server access needed to do the "cleaner" solution of having the server automatically redirect the URLs. (The disadvantage of this, however, is that if such a move causes the obsolete URL to seem to function normally, users never do get any incentive to update their bookmarks and links to the new URL, and hence you never can get rid of these "temporary" redirects; it might be better in the long run to put up a warning page telling people to change their links, and using a normal link to the new URL instead of any sort of automated refresh or redirect.)

And there are probably a few other possibilities. No feature of HTML, or anything else, is totally evil; there are always some good uses. But some "features" are misused more often than they are properly used, so you should think carefully before using them.

Even if you do have what seems like a good reason to use an automatic refresh, you may still have to get rid of it... I was in that situation a long time ago. I had used a refreshed page to get around a technical difficulty in a site I worked on for my employer, but when they made a new marketing deal with a major company which involved them linking to that page, somebody at that company objected to the refresh on the grounds that it made it difficult for anyone to return to the original site via the BACK button. (I did use a 2-second pause on the refresh to allow time to go back, as opposed to some other refreshed sites that have no pause and pre-empt the BACK button altogether, but that wasn't enough, since most users don't figure out what's happening quick enough to hit BACK again within this short time.) So I was forced to come up with an alternative, non-refreshed solution really quickly. Let this be a lesson; if you're doing a site that's anything other than your personal home page, you may have a client, boss, affiliate, or other person who will someday demand that you get rid of that annoying refresh right now. It's a lot easier for you to develop the site without a refresh from the start than it is for you to figure out how to do away with it once it's embedded in your site's basic structure!

If you do use a refresh in your pages, at least provide a regular link to the next page in addition to the automated refresh, for the benefit of those with non-refresh-supporting browsers and those who don't want to wait for the automatic refresh to kick in.

And Now For Something Even More Annoying...

Worse than "META refreshes" are web pages that use embedded JavaScript code to load another page. I've encountered sites whose front pages consist entirely of JavaScript code that checks such things as the user's browser type, the time of day, the phase of the moon, or whatever, and then proceeds to launch different pages in each case. Probably the site author thought they were being pretty clever, but what this means is that any user with a non-JavaScript-capable browser, or who has disabled JavaScript (which many users do for security reasons, given the browser security scares that erupt every few months when another bug is found), ends up seeing a totally blank page. This is a really frustrating thing to inflict on your users. And don't forget that the search engine robots see the same thing when they're indexing your site!

One U.S. Senator had this sort of front page, and when a constituent complained by e-mail about his inability to enter the site using the MacWeb browser, he got a rude reply from a staffer that "90% of the users use Netscape or Internet Explorer, so that's all we care about." It made me wish I lived in that state so I could vote against this Senator. Rudely turning away 10% of the voters hardly seems like good campaign advice, not to mention the likely higher percentage of people with JavaScript disabled who will also be barred from the site. And don't forget that JavaScript is loaded with incompatibilities between the various implementations; do you really want to trust the whole navigation of your site to a script language that might blow up in a different browser version? Did you test it on all versions of all browsers?

I use a JavaScript-capable browser with JavaScript enabled, but I still found sites like those to be highly annoying; since they proceeded to load pages with no time delay (unlike the META-refresh), they made the use of the Back button nearly impossible.


  • AllStocks.Com AutoRefreshing Page Builder lets you set up a page that refreshes your favorite site, useful for monitoring stock quote, news, and sports-score pages. It seems to do this with some combination of frames and scripting (and claims to be seeking a patent on the technique they use). They also have a downloadable program to automatically refresh sites.


[<== Previous] | [Up] | [Next ==>]


This page was first created 16 Nov 1997, and was last modified 24 Mar 2012.
Copyright © 1997-2012 by Daniel R. Tobias. All rights reserved.