So, who out there has one of $SUBJ in their config? I wanna know how you're invoking them, because I wanna make some changes.
There are some oddities in how the params are parsed. They both list "vgrav hgrav" in the arguments to them, and that's the Proper(tm) way to do it. However, due to historical quirks in how the config is parsed, a much looser constraint is applied to them. The parser is perfectly happy to accept "vgrav vgrav" and "hgrav hgrav" for instance. I don't really care about breaking that, because it's already broken (i.e., leads to a segfault as soon as you try using it). So, there are no valid configs that would break if that were fixed. The reversed case "hgrav vgrav" is more interesting, though. Because of scary details of how it's used[0], it actually mostly works right-ish if you give them in the wrong order. There are some intended behavioral bits that won't trigger right though. So far, I'm not worried about breaking that either. But there is one functional difference; if you specify the hgrav first, it takes precedence when laying things out. That is, "vgrav hgrav" fills in the columns first, while "hgrav vgrav" fills in the rows. That means that "fixing" the order _will_ cause a change in presumtively desired behavior. So, my question: is anybody doing that? It only _sorta_ works right now, and I'm about to make the parsing more careful and make it start failing completely. If nobody's relying on the odd-case behavior, that's a nice neat end of it. If somebody is, I'll have to add something else to handle that (maybe another keyword; for $REASONS it's not as easy as you'd think doing it inline). [0] Seriously, you don't want to know. You'll never sleep again. -- Matthew Fuller (MF4839) | [email protected] Systems/Network Administrator | http://www.over-yonder.net/~fullermd/ On the Internet, nobody can hear you scream.
