kukla wrote:... I have the feeling I'm heading for a divorce if I try to add another "obstacle" to my wife's browsing experience. We share this computer and, with some occasional grumbling, she's now putting up with NoScript. ...
You can probably imagine how many times we read that here: "I love NoScript, but my (spouse, parents, children, Grandma, Significant Other... ) can't handle it..."
I sympathize!
I didn't realize that NoScript wasn't completely effective in dealing with Cross Site requests.
Little Snitch for OS X seems to have caught this particular one from viacom.
I think the developers can explain the differences better than I can.

First, please review
this NS FAQ, then please review
the Request Policy FAQ, especially where he
compares RP to NS, though I believe he understates NS's protection against CSRF, probably by not being fully familiar with
ABE.
As I type this, script from informaction.com, where this site is hosted, is allowed, but requests to tinypic.com are blocked by RP, not because I don't trust timypic, but because I don't need it at this time. Principle of Least Privilege, and why use the bandwidth? Tinypic is not trying to run script here, and so doesn't show in NS menu.
I hope those answers will explain the difference. Note that RP's dev strongly urges users to use NS:
Justin Samuel wrote:NoScript is an amazing extension and is absolutely essential (like RequestPolicy) to using Firefox securely. It is best to use both RequestPolicy and NoScript.
while RP was first recommended to me by NS dev Giorgio Maone himself. I'm guessing that Giorgio doesn't return the favor on his page because, as said above, too many novices panic at first seeing NS, so asking them to install yet another "need-a-user-decision" add-on is liable to make them throw up their hands and walk away. Most who use RP are tech-savvy enough to consider NS a piece of cake, and *probably* already have NS.
Firefox Add-ons shows 30x as many NS users as RP users.
btw, I didn't realize that
Firefox Add-ons now shows "users" vs. "downloads", as I haven't added any new add-ons in a while. I think NS d/ls are probably about 90 million by now, and not sure where Moz. gets the stats on unique users, or how they know. If my large family or fraternity house has 20 users all using one computer, and all actively using NS... If I own two computers (which I do, bought three years apart) and d/l NS twice, it means I liked it on the first one enough to put it on the second one.
As for the YT videos, it was my understanding -- which may need correcting -- that malicious code can't easily make it through YT. Do you know of any specific exploits that have occurred via YT videos?
I don't actually visit YT that often, have never uploaded anything, and don't know the mechanics behind it. It was just an example - there are lots of other sites where a Flash vuln could be exploited, such as an ad on a legit page, though most often, through the old link-in-e-mail trick. "Paris Hilton video
here!" 
-- OK, that's probably not going to snag your wife, but the very same low-tech users who find NS intimidating are the ones most likely to fall for some kind of social engineering, no?
I don't think YouTube allows uploading arbitrary files directly. I've never tried posting something to YouTube, but my vague impression is that you send YouTube a video file, and then if YouTube can parse and process it, that file gets wrapped up in YouTube's Flash player wrapper for delivery. It would probably be difficult to get a Flash exploit attached to a video in such a manner that it would survive the whole process, although I can't prove it would be impossible.
If YT does any selectivity at all, how do all those copyrighted songs and records get through? Seriously, I've wondered. Every once in a while, a song I like gets taken down for copyright violation, but there are ten other versions of it - original CD or vinyl record, live performances, etc. Only when the copyright owner complains does something happen. So it didn't seem like there was much vetting. Given the number uploaded every day, seems impossible - and look at the amateur street vids from Libya, Syria, Egypt, etc.
I'm guessing one could insert a bit of malicious code into an otherwise-legit video, but I've never tried it and don't intend to.
Does anyone else have more info on whether this is possible, and what safeguards YT uses, if any?
I played with RequestPolicy a bit. Too bad there's no way to toggle it on/off. That way, I could use it on a per site basis when I think a site might be kind of dicey.
We close this post with good news! If your wife can do one right-click and one left-click, she can toggle it off for her session, and you get the machine back with the full protection in place, or can toggle it off the same way.
Wife: R-click RP red flag logo. L-click "Temporarily allow all requests". She'll browse as if it weren't there.
You either start a new session (close browser and restart), or just do what she did - R-click the now-yellow logo, and L-click the same "Temp' permission, which removes the checkmark.
Anything that can be done or undone in two clicks is a toggle in my book.
- Tom