editing html before it hits my browser

[PhiJ]

There's a bug on the network I'm under, so when I download a page from nlab, it inserts invalid code into the page and firefox/uzbl can't view it.

I'd like to write a script which will remove the bad code from any webpages from ncatlab.org before they reach the browser, and was wondering if anybody could advise me as to how I might do that.

I've been told that greasemonkey might work in firefox, but I'd like to do something that is further 'back' ('then it will work in uzbl and wget' as well as 'that sounds like the better way to do it').

I'm connecting to the web using dhcp.

[Hu]

I suppose it is out of the question to use https, to prevent the broken device from mangling your stream? For that matter, can the broken device be fixed or taken offline, thereby benefiting everyone in the area?

[PhiJ]

No. The website doesn't seem to accept https (I'd be surprised if it did) and the network is not under my control. I filed a bug report and got a 'why can't you use IE, it works in IE' back.

Perhaps got a bit too irate at them for that, and haven't heard back since. I suppose it's possible that they're quietly fixing it.

[Dr.Willy]

Well, if you want something that is further back than Greasemonkey and thelike you should have a look at those filtering proxies like proxomitron and bfilter.

[Mad Merlin]