The syntax of values is specified separately for each property, but in any case, values are built from identifiers, strings, numbers, lengths, percentages, URIs, colors, angles, times, and frequencies.

A user agent must ignore a declaration with an invalid property name or an invalid value. Every CSS2 property has its own syntactic and semantic restrictions on the values it accepts.

In other words, Firefox 1.5 is doing things it shouldn’t do, and is not following the W3C specifications.

Xylescope is indeed excellent — I purchased my copy just yesterday, in fact, after using the trial first. Very, very blissful tool!

Anyone found a more efficient way around this?

Try running a webserver on your own localhost?

Xylescope functions as a Webkit-powered browser (as far as I can tell, anyway) and Webkit isn’t a PHP parser (logically). Opening PHP files would thus not make sense unless they’re pre-processed by a webserver, which you can run on your own computer with minimal effort.

And hey, Mike started off this post with sIFR 3 and FACE, I think it’s fine to say that here ;-)

Besides, Steph totally deserved that one. I’m sick of the juries giving out the money to the person they A) Like the best, B) Hate the least or C) Whoever got there on the other person’s back. It’s about the game, people!

In other words, Firefox 1.5 is doing things it shouldn’t do, and is not following the W3C specifications.

That’s not true, Firefox is not applying any of the invalid rules it encounteres and merely reports the fact that it’s not applying them.

I find the behaviour very usefull when developing since it notifies you of small stupid mistakes you might make in your CSS.

While you both have a point, my issue with it is that defaulting to showing errors about this isn’t what I consider to be ignoring. If people want to see errors in their CSS, let them check a checkbox or hit a button for it in the console, first. I can see how it’s useful, but forcing it upon you as the developer is taking it too far, in my opinion.

Also, in my experiences with it so far, it’s not useful but just plain annoying. I know where I’m using “invalid” CSS (valid, just not recognized, thus it should be ignored) — but it’s a hell of a job for me to bugfix Javascript now, what with all the “errors” that the JS Console is giving me.

And like Mike, I just don’t like being forced by a browser to use a more verbose hack just to clear the console. It’s not very (user)friendly, to me.

And the Firefox console is one of the most usefull around. Try Safari or Opera any other day, in terms of feedback for development.

Admittedly, Firefox’s console is more useful than Safari’s or Opera’s. For that reason, Firefox is my debugging-browser. I really don’t like it for my own browsing, if anything because it doesn’t support text-shadow or hooks into the nice Mac OS X behaviours like Safari. :-)

Aside from some unexplainable bugs related to its new handling of the overflow property

Care to elaborate on those ‘bugs’ ? As far as I can tell, Gecko is handling the overflow property fairly well (per CSS2.1 and CSS3 module). I remember your site has been messed up for the better part of the year in Gecko nightly builds, due to your incorrect handling of that property (yes that sidebar with overflow hidden and a huge left margin set, jumping away).

Jemaleddin: Oh yes, Danni definitely needs to gain some weight, but she looked pretty damn good on that Survivor reunion show. As opposed to Stephanie who looked, ummmm, well, let’s just say she looked better without the makeup, the home perm, and the overly dramatic shaved eyebrows. I wouldn’t peg Danni for the full-on Playmate Of The Month spread, but just one of those trademark “hey, this person’s a celebrity, here she is with very little on” deals.

Phillipe: Ok, regarding the Firefox bugs, nobody ever did get to the bottom of the overflow/sidebar issue, but seeing as the code was valid, the rules were followed, and it worked in every other browser including all previous versions of Firefox, I’m not willing to put the blame on my code at this point. This site used a very standard “float one column, use a margin on the other” layout and something about the new handling of overflow caused that not to work anymore. The only reason it was not noticeable on many other sites is that most people probably don’t have the overflow property in their layouts. Soooo… after getting enough complaints from people using Deer Park and other FF beta, I just modified my layout so that everything is floated. That seemed to fix the symptoms, but clearly not the disease. So then FF 1.5 comes out and someone on OS X notices that every single page on my site with over about 200 comments shows up as completely blank. Not only that, but if you try to drag the scrollbar down, you actually freeze up the entire browser. I’ve made a reduced test-case here. The “culprit” here is apparently that if you apply an overflow-x css property to any really long div (like one with 200 comments in it), all hell breaks loose. The only solution is to put an underscore before the css property (or use * HTML) to hide it from Firefox, but hey that’s another error for the console. :)