Mortal Engines, The Fall of the Kings, The Visual Display ..
Besides the three books I mention below, I did go ahead and read Dragonfrigate Wizard Halcyon Blithe. Like you probably expect, it was somewhat better than the last book -- the plot not quite so conventional, the exposition slightly less massive, the songs a bit less lame. The flaws are basically the same as last time, though, so it doesn't seem worth doing a full review about, though I do wish to lodge two complaints. 1) Lord Firstname and Lord Lastname are not interchangeable names for the same person. 2) So far there has been no interaction with a woman in a romantic way where it hasn't turned out to be a comic misunderstanding or the woman being secretly evil).

Now, onwards:


Mortal Engines (Philip Reeve): As an author, when you write a first sentence like "It was a dark, blustery afternoon in spring, and the city of London was chasing a small mining town across the dried-out bed of the old North Sea." then you pretty much dust your hands off and say "ok, that's the hard part of the book done". The zeppelins are just a bonus.

Structurally the book is a little weird: the primary protagonists get tossed out of London early on in the book, and spend most of the book exploring the land and seeing all the weird and interesting things there are to see outside. The problem is that the actual important plot developments are all in the city, so the author's introduced a secondary protagonist who stays in London and watches them go along, to let the reader follow things until the main protagonists intersect with the main plot again. Unfortunately, the secondary protagonist really is secondary -- this is the kind of sf book where character development takes a back seat to setting, and as a secondary she just doesn't have enough screen time to develop until almost the end of the book, at which time her character is written out of the narrative. What this means is the whole plotline with her feels kind of useless (there's even a bit where she says "oh, this is what fate put me here for!" and it turns out to be just to give an assist to a "real" protagonist -- gee, thanks fate).

However, the real point of the book is the setting, and the setting is awesome. Being me, I naturally thought how it would make a good RPG setting* -- you have the cities of various sizes moving around the landscape fighting and hiding and looking for food, and then you have zeppelins flying spy missions, an airship city in the clouds, killer robots and other old tech, complexly stratified social classes, and an anti-tractionist league working against the moving cities. But even just as a book it is pretty good.

I'm given to understand there are another three books in this series. I don't feel any particular need to read them, but I guess I may pick them up if I see them in the future.

*It has been pointed out to me that there actually is an RPG based on this, but it's one of those shared-world freeform-writing RPGs, not the thing of complex rules and mechanics and sourcebooks I was imagining.


The Fall of the Kings (Ellen Kushner and Delia Sherman): So Samuel reported all the words of the LORD to the people who were asking him for a king. He said, "These will be the ways of the king who will reign over you: he will take your sons and appoint them to his chariots and to be his horsemen, and to run before his chariots; ... He will take one-tenth of your flocks, and you shall be his slaves. And in that day you will cry out because of your king, whom you have chosen for yourselves; but the LORD will not answer you in that day."

This book does indeed provide the ending in blood and fire that I found lacking in the previous and following books. But is it greedy for me to demand yet more blood and more fire? What this book provides instead is love and madness and magic, and while these are all fine things, I'm not sure they are a totally satisfactory substitute. Depending on your relationship history, you may say that blood and fire and love and madness are all the same thing, and there's a certain truth in that, but the difference is this: the latter two end. You wake up one morning, weak and trembling after the fever has broken, and begin to recover your health and return to how you once were. Blood and fire, on the other hand, leave permanent scars, and everyone who touches or is touched by them is marked.

Which is a fancy-pants way of saying I thought this book had too many characters who did too many things that changed them too little. A man betrays his friends and pays no price for it except guilt (and that too fades); a woman stages an event that has the whole city talking and then leaves again, totally untouched and untroubled by it; two people are engaged to be married on a whim and just as casually disengaged (and we know all along the marriage, even if it lasts, will be totally unimportant). Perhaps the heart of the problem is that at the end of the book, the final question is not "What repercussions will his actions have?" but "Will anyone remember his actions at all?"

While I'm griping about things, hardly any of the romantic relationships felt plausible to me. This wasn't really an issue with the central relationship of the book, which had an explanation for being fey and weird, but the other relationships are all extremely shallow. I know Kushner can do this because I am totally willing to buy Alec and Richard's relationship in the first and third books, but here the only time you see male relationships* in this book, the characters involved are busy kissing hotly or clenching each other deeply, and there's no real chance for them to, like, relate.

*I'm talking about male-male relationships here -- the male-female relationships are friendships passed off as romances, and the one lesbian with a sizable role is another irritating deus ex machina who's too busy being awesome to have time to interact normally with people.

So why am I still remembering bits of this book vividly? I guess because the portrait of love and madness the authors give is really fucking compelling. There is an extended period of the book where some of the characters are wondering if magic is really real or not (spoiler: it is) and while this is going on, the stuff they are driven to is uncanny -- they race along the edge between what you might do if compelled by a strangled heart and what you might do if ensorcelled and wrapped in madness. This scene where one man comes back to his ex-partner, forced to make supplication like a beast and then is driven out to hunt again ... holy crap. Stunning.

Anyway, yeah, this book is an unusual mix. Also unusually, I think the mixed-up order that I read the three books of the series in (one, three, two) is probably the right one; that's the right plot-temporal order, it's the right thematic order (as the books move away from swordsmanship and noble-to-noble politics, and towards societal trends and history), and the protagonist of the third book appears in this one, but in a form that seems like it would only make the second depressing to read later (not because she's badly off, but because it decisively closes off a number of possibilities that seem fresh and open in the third).


The Visual Display of Quantitative Information (Edward Tufte): This was described to me by somebody as graph porn, which is catchy, but almost exactly wrong. The point of porn is to show only the good bits, of course, but also to show only the good bits that you're already familiar with -- you're not supposed to learn or realize anything*, just drool over the concept of the thing. Stimulation without education.

*Except possibly what to look for in a cable-repair guy.

This book, by contrast, is more like A Pattern Language or A Theory of Fun -- although there are plenty of explicit pictures of bits for people who like bits, the intent is to use them as props, to encourage the reader to analyze and dissect them. There are also plenty of pictures of bad graphs and designs, and the book dissects them as well.

Every aspect of graph design is covered: proportions, shading, graph type, data density, what to label and why, illustrations, deceptive statistics (including a helpful way to calculate the Lie Factor for any given graph) and more. If you have any interest in the subject this is an awesome book.

Also, despite the discussion at the beginning, this is one of those books I feel the irresistible urge to own rather than get from the library, just in case I have to graph something at a moment's notice. But it's not graph porn. It's graph erotica.

P.S. My favorite graph is on page 141 of the second edition. Triple-functioning data measure, clean lines, data-ink maximization = hottt! (ctate points out there are many other awesome graphs available on the same site).


Up next is yet another book in the Company series. lpsmith and I are really going to have to do co-reviews of this at some point.