It's generaly easy to find HTML entity names; do a google search for HTML entity OE or suchlike. (You don't even have to clickthrough for this one; Œ is Œ, as you can see reading the snippet of the first hit.) (BTW, for input you should support UTF-8 chars 338 and 339 (AKA /\x{152}/i), as well as /oe/i. The info on codepoints is in that same table, if you remember that the &#nnn; syntax uses decimal nnn.

Oh, and PM/Everything can pass-through any tag it's told to; it's being told not to passthrough <img> tags. I rather wonder why. I suppose there is a worry people will put <img> tags pointing to porn or other things undeseriable here. (On homenodes, it will allow them IFF the users's level supports homenode images, IIRC.)

Update: I almost forgot: when do you need an entity for curlies? They aren't metachars in HTML nor Everything.

Thank you! Very good info. No need for curly entities anymore since now I've memorized that they don't need 'em.

I will certainly study this table and thak you for the info on UTF-8.

I'm building a database for a wine merchant and I'm worrying about how to search/input accented characters and the OE. There are problems with Japanese keyboards and fonts for this, so I am thinking about converting OE to the entity on display, but to ignore accents when searching. Of course I'm probably going to write Word/PDF so that's a different story maybe.

About the <img> tag, yes the danger of porn uploads is obvious but seriously we could make very good use of diagrams, and we often talk about diagramming, so I would think that if l33t porn jokes is the danger then it should be connected to level privileges. Perhaps there is a security or diskspace/bandwidth problem with uploads, but I'm going to think about putting images on my server to go along with posts, if necessary to the discussion. Can't think of one now, though it might have been useful recently talking about Quantum:: in Ovid's thread.