How to not load a file that doesn't exist?

July 28, 2009 3:03 PM

How can I get a specific site to stop trying to load a specific file? Specifically, how can I get NotAlwaysRight to quit trying to load an image from Clustrmaps?

I'm using Firefox, and am running RequestPolicy, NoScript and Adblock Plus.

Whenever I visit NotAlwaysRight, the page never loads fully. By using LiveHTTPHeaders, I've managed to narrow the problem down to an image residing on a Clustrmaps server. Until this image loads (which it never does, even when Firefox is in Safe Mode), the tab is shown as loading. I have to manually click stop to get it to quit attempting.

So far, I have tried:

1] Using Adblock Plus to block the image. This just caused about 5,000 requests in the space of a few minutes. I could see the number of blocked attempts shooting up in the Adblock filter list.
2] Allowing/disallowing NAR to request information across the sites with RequestPolicy. This makes no difference.
3] Adding http://www2.clustrmaps.com and http://www3.clustrmaps.com to my HOSTS file, and redirecting them to localhost.
4] Viewing the site in Internet Explorer. The image still doesn't load, according to the status bar at the bottom.

Any ideas? How can I get the page to stop trying to load a file that apparently doesn't exist? Or even, how can I get the page to quit loading after x number of seconds?

posted by Solomon to computers & internet (6 comments total)

Heh--I came here to suggest #3 on your list. But my next suggestion is to rewrite the page with greasemonkey to remove the link.
posted by Obscure Reference at 3:14 PM on July 28 [1 favorite]

Well I suppose you could use their comment form to bring the issue to their attention. If it were my site I'd be glad of the info.
posted by le morte de bea arthur at 3:17 PM on July 28

#1 and #3 were what I came to suggest, and I'm go to resuggest them with twists.

#1b) I believe adblock plus has an option to not even attempt to load the image, rather than just hide it but still load it. Make sure this is turned on.

#3b) The HOSTS file takes domain names, not full urls, so try just www2.clustermaps.com without the http://

and #5) There's definitely a way to solve this with a GreaseMonkey userscript. Assuming you aren't a javascript developer, Platypus will generate a script for you that permanently removes certain parts of pages.
posted by comwiz at 3:36 PM on July 28

Echoing 3b above, but if you want the file to load successfully, you're going to need to run a webserver on localhost to give it a dummy when it asks.
posted by fightorflight at 3:48 PM on July 28

Assuming you aren't a javascript developer, Platypus will generate a script for you that permanently removes certain parts of pages.

Sadly, it appears Platypus is not supported anymore (at least not on Firefox 3.5). The most recent update in the changelog notes compatibility with 2.0.
posted by entropic at 3:51 PM on July 28

With #1), the page has Javascript in it that makes it load a "not available" image when it has an error loading the desired image. Unfortunately that image is also hosted on clustrmaps, which causes it to fail, which fires off the javascript, which causes it to fail, which ...

If you use the Element Hiding Helper add-on to Adblock Plus, you can block the whole <a href ...> statement around the image, including that Javascript. The Adblock Plus rule it just generated for me is :

notalwaysright.com##A#clustrMapsLink
posted by Pinback at 10:44 PM on July 28