One thing I noticed, which may have to do with the “file-system clutter” you referred to: When I read breadbox’s comment, I didn’t know where the excerpt he quoted came from, so I had my browser search for it, but couldn’t find it because the quote is in a knowl rather than the text. Fortunately I chose “Talmud” to search on rather than, say “exegesis,” so the search took me to the one paragraph with that keyword, and I guessed the quote was in one of the knowls there, so it wasn’t that hard to find. But it would be nice if there were something that facilitated this sort of search.

For years I’ve been daydreaming about a project called the Talmud of Science. We take a dozen or so deep and difficult papers—the kind that everybody knows but nobody reads—and we do the exegesis thing on them, with expert commentary, historical background, marginal glosses, added illustrations, and so on.

Charles Petzold did something almost exactly like what you’re describing in his book The Annotated Turing.

If you don’t need to support outdated browsers, you can get around the cross-domain restriction with HTML5’s postMessage function. That allows (relatively) safe transport of data between cooperating web pages (both sides have to be set up to allow the data transfer, and the receiver can’t see anything that the sender doesn’t explicitly send). It’s supported in recent versions of all the major browsers. There are good examples here: http://davidwalsh.name/window-postmessage and here: http://davidwalsh.name/window-iframe

As for *why* I prefer the knowlly solution: To my taste, it’s the better compromise between unobtrusiveness (when the note is closed) and frictionless accessibility (when it’s open). Anything that involves scrolling the window causes me to lose my place in the main text. Because a “page” on the web can have unlimited length and isn’t all within view at the same time, the Wiki notes seem more like endnotes than footnotes. You need to leave a bookmark to find your way back.

To my surprise, though, I find that the knowls do work, after a fashion, in my newsreader (NetNewsWire 3.3). They lose their styling, but the Javascript works, sliding the text open and closed at the appropriate spot.

One of the remarkable features of the Web—perhaps unique among media—is how much power and control the reader retains. The concept was there at the start, 20 years ago, with the idea that the author would mark up structure and semantics, and the reader would determine how it’s displayed. That’s a division of responsibility that was never possible with printed documents, and it goes against the grain in publishing formats like PDF. How long will it survive in the world of HTML/CSS/JavaScript? I don’t know, but I still take offense when an outfit like the Times has the effrontery to think they can fool around with my mouse buttons.

(And by the way, I’m a fan of the Times. Needless to say, there are lots of web sites that do far sleazier things. It’s because I respect the Times that I’m most bothered by their misbehavior.)

Joking aside I remain a bit skeptical of the current Standard Cosmology, and very concerned that cosmologists & astrophysicists have overstated the case for DE/DM/Inflation. If any of these mechanisms don’t pan out you can bet that opponents of reason—who use every appearance of revision to discount the validity of science—will have this as further fodder. So I would just like to see modern cosmology all presented a tiny bit more tentatively, as science often should be.

From my perspective the problem with the whole situation is that we haven’t really verified General Relativity on the galactic and intergalactic scales, which make DM/DE/Inflation seem like ad hoc corrections to keep Big Bang + GR consistent with observation. I’m not yet so convinced the assumption that most matter and energy in the universe is exotic and the behavior of spacetime curvature is uniform across scales is any more reasonable than assuming most of the matter and energy is the same while the behavior of spacetime changes with scale. (Though I suppose the last bit is essentially all dark energy says anyway.) When discussing it with experts I hear arguments about the accuracy of Big Bang nucleosynthesis predictions and the CMB fit—I’ll be much more convinced if one of the many searches for dark matter succeeds! (And how many DM searches must fail before it starts to look like aether?)

However, I’m no expert, so I concede that the opinions of those with a greater grasp of the relevant concepts should be valued far more than those (like me) who’s understanding is lacking.

On the other hand, occasionally in discussions with expert proponents of Standard Cosmology they’ll concede that the alternatives are not yet completely ruled out. It seems we’ve succeeded in explaining everything within arms length, and without a bigger ruler we’re stuck with a degree of mere speculation over which rules extend elsewhere and which ones will need revising. (Or maybe the small scale will save us—perhaps a mature theory of quantum gravity would make sense of it all!)

An illuminating experiment would be to get very young children (2 or 3 years old?), who have not been exposed to any cultural depictions of stars, show them the night sky, and ask them to draw what they saw. (Somehow I suspect we’ll see spikes. But we should try and see.)

http://iopscience.iop.org/0963-9659/7/1/002