Waldrop on writing!
Okay, so it's really Waldrop on applying to Clarion, but much of what he says is equally applicable to writing in general.
He starts by pointing to the Turkey City Lexicon, which is required reading for all aspiring authors. It's been some years since I last read that, and there's a bunch of stuff there I hadn't remembered; makes me want to include a link to that page in many of our rejection letters. I've never been fond of most of the actual terminology in the Lexicon, but I do like the codification of a lot of the concepts; a good checklist of things not to do. I'm surprised, though, to see "Slipstream Story" among the insulting terms, defined as "Non-SF story which is so ontologically distorted or related in such a bizarrely non-realist fashion that it cannot pass muster as commercial mainstream fiction and therefore seeks shelter in the SF or fantasy genre."
(Another time I'll talk about the differences between 'commercial mainstream fiction' and 'literary fiction'; a distinction I glossed over in my slipstream/genre editorial, but I think it's an important one to make, particularly because sf people seem generally to regard anything sans sfnal elements as "mainstream.")
All of Howard's advice is good advice (though I'm not totally sure I understand the part about the dead deer), but the item that pleases me most is something he got from Damon Knight: "if you have a great opening line, your story starts on page 3. . . . You have spent the first two pages not on what happens in the story, but on justifying your opening line." I can't tell you how refreshing it is to hear someone advocating not starting a story with a punchy hook. (To mix metaphors.) I get particularly annoyed by punchy hooks that turn out not to be true: "Gregor Samsa awoke one morning to find himself transformed into a gigantic insect. Well, actually, he wasn't really an insect, he just kind of felt sort of insect-like, y'know? Because the world around him was so oppressive and stuff. So anyway, really he just woke up feeling bad."
I also particularly like Howard's quote from James Blish: "Most sf isn't about anything: the sf that is about something is usually about other sf."