I did a bit of testing - surprisingly, changing the UserAgent on Firefox to mimic Chrome seems to replicate the bug. From there I did a lot of digging- and indeed, while a lot of spoilers in almost any topic I could find didn't seem messed up at all, the posts containing the spoilers mentioned in this topic did have the glitch (they look more like what StraightFlame has posted; so maybe the initial glitch Egomassive posted has been 'fixed' (or rather, just changed). That said, again, I'm not actually running chrome, so I don't know how accurate this really is (but it's odd that it replicates some oddities either way). It could be that the bug(s) I report aren't entirely genuine to the real Chrome; but there's obviously still something to this problem.
I did a full comparison on both pages (the Firefox version and the "Faked" Chrome version (which is using the 'latest' version, as reported)). There was only a single difference between them; a javascript call at the bottom of the page which is merely called "smf_codefix()" -- it doesn't really explain what it's supposed to fix, but it does seem to alter post height on a per-post basis, which possibly explains why the spoilers are drastically overflowing into everything else, yet the post size doesn't match like it should, making it unreadable. It might also explain why I'm not seeing it happen on many other spoilers personally, if it's conditional (maybe those posts just got really unlucky). The real point here is that it only appeared when acting like Chrome - there's nothing of the sort in the Firefox version the page, and it's the only difference at all. Considering the nature of that code, it seems like someone was trying to fix a quirk a long time ago and it never really manifested as a problem. It's probably called based on what browser 'needs' the fix; explaining how I can trick it into loading by using a different UserAgent that 'needs it' (Chrome, in this case, apparently). (So as a side note; that's actually a possible temp fix if you can do that on Chrome; change it to 'appear' as Firefox). I'm guessing this code wasn't just added recently, and since Firefox (maybe others?) are absent of the code entirely the other javascript scripts or the php files themselves must dynamically put that there), it should be relatively harmless to remove it from loading for a quick check. Finding where it gets inserted might be tricky; I doubt it's really part of the php or html code by default, though, since it would appear on Firefox too-- it's easier to add things in realtime to a page than remove stuff hardcoded to a file already. So my guess it that it's part of a script somewhere that only gets added based on reported browser -- if it stops adding that code, I wonder if it would fix the browsers that it tries to mess with.
Anyway, just an idea!
EDIT; this is the relevant code:
<script language="JavaScript" type="text/javascript"><!-- // --><![CDATA[
window.addEventListener("load", smf_codeFix, false);
function smf_codeFix()
{
var codeFix = document.getElementsByTagName ? document.getElementsByTagName("div") : document.all.tags("div");
for (var i = 0; i < codeFix.length; i++)
{
if ((codeFix[i].className == "code" || codeFix[i].className == "post" || codeFix[i].className == "signature") && codeFix[i].offsetHeight < 20)
codeFix[i].style.height = (codeFix[i].offsetHeight + 20) + "px";
}
}
// ]]></script>