Musings

This story broke a while back, but I never did get around to posting on it.

There’s always been a lot of talk about alternative fuels (Hydrogen-powered cars, Mr. President? Gimme a break!). But, aside from some pork-barrel (ethanol) subsidies for agri-business, not a whole lot of action.

Over in Britain, however, where taxes make fuel much more expensive than here, people have been taking matters into their own hands. Turns out that slightly-modified vegetable oil works as well as diesel, produces fewer emissions and — instead of that stinky diesel smell — has a pleasant, hunger-inducing aroma.

According to the Manchester Guardian, the authorities are none too pleased. But it’s all perfectly legal, provided you pay your taxes.

MovableType 2.6.x added a CreativeCommons Licence module (off by default, but all too easily turned on). This has touched off a bit of a brouhaha, as various prominent bloggers grapple with the implications of turning it on.

A very good legal analysis of these implications is helpful.

Personally, I thought that the CC licence I chose for this blog closely mirrors the prevailing ethic in the scientific community. I’m happy with that choice, and I know that I will surely have to live with it.

Previously, I asked whether there was any relation between the de Sitter solutions of compactified string theory with fluxes (and branes) and the explicit gauged supergravity solutions of Trigiante et al.

One good clue about how to go about understanding the connection (if there is one) is the recent paper of Andrianopoli et al. They show explicitly the connection between string compactifications with fluxes and effective gauged supergravity theories.

Perhaps that’s not totally surprising. After all both achieve a reduction in the number of moduli. But, in the former case, I’m used to thinking about this in terms of a flux-induced superpotential. The lifting of the flat directions in gauged supergravity also involves the Higgs mechanism. So the connection between the two approaches is less than obvious (at least to me).

Of course, Andrianopoli et al only discuss abelian gaugings, whereas Trigiante et al need a particular noncompact nonabelian gauging. And, on the string side, Kachru et al have a space-filling anti-D3 brane. So the connection is still a little obscure…

The really cool people have, not just an RSS feed for their blog entries, but also a feed of recently-posted comments. That way, if you’re following the discussion of some particular entry, you don’t have to check back to see if new comments have been posted.

Not wishing to seem uncool, I added a Comments Feed to this blog.

Of course, some people think they’re even cooler by having a separate RSS Comment Feed for each individual blog entry. That’s extremely tedious to deal with from a reader’s point of view. To make it bearable, NewsAggregators could give the option of automatically unsubscribing feeds that have not been updated in a while, and provide some sort of collapsible nested list of subscriptions (to deal with the sudden proliferation of subscriptions to similarly-named feeds).
With present technology, I’d have to have an order of magnitude or two more comment traffic on this blog to make individual comment feeds worth the extra trouble.

No, what really would be cool are Threaded Comments. But I think I’ll wait till Phil Ringalda does it to his blog.

Update: Dang!

At $14.95, it was hard to pass up the Keynote+iLife bundle from Apple. Which means that I’ve gotten to spend a bit of time playing around with Keynote.

After the initial amusement of watching bullet points bounce and twirl and slide across the screen as they appeared and disappeared from my slides, I got down to the serious business.

How to do equations? The obvious answer is EquationService. It’s best known as a providing a “send text, return PDF” service. For those Applications which accept such services (Mail, TextEdit), you can highlight a block of tex code, hit ⌘+\ and EquationService will run it through pdftex and return a PDF fragment in its place.

Keynote, alas, is not such an application. Fortunately, EquationServices can also function as a handy-dandy standalone equation editor.

You type TeX code in the upper pane, and then drag the resulting PDF from the lower pane into the target application. Keynote does accept drag 'n drop PDF.

By itself, this would be pretty useless. Every time you wanted to edit an equation, you’d tediously have to recreate it in EquationService. Fortunately, EquationService provides an untypeset facility. It embeds the original TeX code in the PDF file, and if you drag an EquationService-created PDF file back onto the lower pane, the TeX code appears in the upper pane for you to edit.

Or, at least, that’s how it works in Mail. If you try to drag the PDF from TextEdit, nothing happens. Curiously, you can drag the PDF from TextEdit into Mail and then from Mail into EquationService.

With Keynote, however…

• You can’t drag 'n drop from Keynote. If you try to drag the PDF out of the Keynote window, all you succeed in doing is moving it to some irretrievable location offscreen. (Where were Apple’s HIG people on this one?)
• You can copy the PDF file. But then EquationService won’t accept the paste.
• Mail will accept the paste. But if you drag it from there to EquationService, you get the message

  Error encountered reading source from PDF

So I’m stymied.

It sure would be great if the Keynote engineers got together with Doug Rowland (the author of EquationService) and made the two work seamlessly together. In the meantime, I’ll stick to more primitive technologies.

Ars Technica has a great article on Moore’s Law, that oft-misquoted 1965 prediction about transistor budgets and manufacturing costs.

Moore’s Law is so perennially protean because its putative formulator never quite gave it a precise formulation. Rather, using prose, graphs, and a cartoon Moore wove together a collection of observations and insights in order to outline a cluster of trends that would change the way we live and work. In the main, Moore was right, and many of his specific predictions have come true over the years. The press, on the other hand, has met with mixed results in its attempts to sort out exactly what Moore said and, more importantly, what he meant. The present article represents my humble attempt to bring some order to the chaos of almost four decades of reporting and misreporting on an unbelievably complex industrial/social/psychological phenomenon.

Not too many other predictions about the future look so accurate nearly forty years on. So Moore’s 4 page paper deserves a careful look.

So I thought I’d say some more about the relation between Bobby Acharya’s paper on moduli-fixing in M-theory (which I’ve blogged about before) and the work of Kachru et al that I wrote about here.

Recall that the latter proceed in three steps

1. The flux-induced superpotential (in the Type-IIB orientifold description)

   (1)$$\int_M (F_3-\tau H_3)\wedge \Omega$$

   fixes the complex structure and the string coupling, leaving the Kahler modulus, $\rho$ (assume just one), as a flat direction.

2. They then guess at the structure of the the nonperturbative superpotential for $\rho$. With $\rho$ fixed, we end up with a supersymmetric solution in 4D anti-de Sitter space.
3. They introduce supersymmetry-breaking in the form of anti-D3 brane(s). This contribution to the potential for $\rho$ has its coefficient fine-tuned so as to raise the previous anti-de Sitter minimum to slightly above zero, producing a non-supersymmetric metastable solution with a small positive cosmological constant.

M-theory compactified on a manifold $X$ of $G_2$-holonomy also has a flux-induced superpotential

where $\phi$ is the $G_2$ structure. In addition, Bobby argues that if $X$ is fibered over a 3-manifold $Q$, with the generic fiber having an ALE singularity corresponding to the simply-laced gauge group $G$, there’s a further contribution to the superpotential that looks like a complex Chern-Simons term

where $\mathcal{A}=A+iB$. $A$ is the $G$ gauge connection on $Q$ and $B$ is a 1-form in the adjoint of $G$ (the twisted version of the 3 scalars in the 7D gauge multiplet).

The critical points of $W_2$ are flat (complexified) $G$-connections on $Q$ and on the space of critical points, we can write $W_2=c_1+i c_2$ for some constants $c_{1,2}$. The combination $W_1+W_2$ lifts all the flat directions, producing, as above, a supersymmetric solution in 4D anti-de Sitter space.

Bobby argues that the supergravity computation that led to this is reliable provided $c_2$ is large. Unfortunately, this excludes the familiar candidates for $Q$, like $S_3$ or $S^3/\mathbb{Z}_n$ (which have “known” heterotic duals). $Q$ must be a hyperbolic 3-manifold (yuck!).

Anyway, we have achieved points 1 and 2 above with no fudging whatsoever. This puts us in comparatively better shape to understand step 3. If we can introduce supersymmetry-breaking in the M-theory formulation, we might actually be able to say something reliable about the resulting de Sitter vacuum.

PHP 4.3.1 is out, fixing a CGI vulnerability.

MovableType has been bumped up to version 2.6.2, also for security reasons. (This, hot on the heels of the 2.6 and 2.6.1 releases.)

If you were expecting some interesting physics posts, now that my new software is up and running, you’ll have to wait a little longer.

Oh, yeah. And as I wrote this, my son awoke, happy as a clam, running a 103°F fever. Aaarrgh…

A really nice post recently by Aaron Bergman on the mess we’ve gotten ourselves into with North Korea. The one element, that Aaron fails to mention, contributing to the Administration’s frightfully bad decisions leading up to the current mess is their visceral “Whatever Clinton did, we shall do contrawise” (e.g., with respect to the Agreed Framework).

He also has a nice summary and some links about this week’s WMAP data.

Well, as you can see from my two previous posts, I figured out how to do text filtering in MovableType 2.6.

I now have a plugin which offers two conversions: itex to MathML and itex to MathML with parbreaks. The former is a straight itex2MML conversion. The latter uses the “TeX” (or blogger’s) convention that two consecutive newlines start a new paragraph. (The bloggers also like the convention that a single newline inserts a <br />, but that would almost certainly mess up any heavy-duty TeX equation input.)

I will blog with them for a while, and if I don’t uncover any major bugs, will release the plugin for public consumption.

Update: You can download the plugin or peruse the README file. And what do y’all think of the garish MathML logo?

This is a test of the new itex2MML+parbreaks filter. Here is an inline equation: $2\sin(x)\cos(x)=\sin(2x)$.

And here is a display equation

And here is some more prose.

The text used to produce the above was:

This is a test of the new <code>itex2MML+parbreaks</code> filter. Here is an inline equation: $2\sin(x)\cos(x)=\sin(2x)$.

And here is a display equation
\[\int_0^\infty e^{-x^2} dx = \frac{\sqrt{\pi}}{2}\]
And here is some more prose.

This is a test of the new itex2MML filter. Here is an inline equation: $2\sin(x)\cos(x)=\sin(2x)$.

And here is a display equation

And here is some more prose.

The text used to produce the above was:

<p>This is a test of the new <code>itex2MML</code> filter. Here is an inline equation: $2\sin(x)\cos(x)=\sin(2x)$.</p>
<p>And here is a display equation</p>
\[\int_0^\infty e^{-x^2} dx =\frac{\sqrt{\pi}}{2}\]
<p>And here is some more prose.</p>

A busy evening in front of the computer.

MacOSX 10.2.4 was released, via SoftwareUpdate. I had to recompile and reinstall Sendmail 8.12.7, but otherwise, it went smoothly.

MovableType 2.6 was released, with the much anticipated Text Filtering plugin architecture. Now all I need to do is find some documentation on creating such plugins, and I’ll be able to create an itex2MML plugin.

And Kung-Log 1.3 was released, with support for the nifty new features of MT 2.6.

Woohoo!

“We don’t need no steenkin’ floppies!”

According to this Austin American Statesman article [link broken], Dell is planning to discontinue the floppy drive on their computers.

How … umh … innovative!
What will those wild-eyed revolutionaries think of next?

Thanks to Adam Kalsey’s SimpleComments plugin, “Comment” and “Trackback” listings have been merged into “Followups”.

Why click twice, when you can click once?

I’m very confused about the fate of metastable de Sitter universes like the one discussed previously.

They decay by nucleating bubbles of true vacuum (with an incredibly long lifetime). But, surely, such bubbles don’t coalesce, as they are surrounded by regions of de Sitter space, which expand exponentially.

You might have thought that since the true vacuum is (10 dimensional) flat space, we don’t have to beat our heads against the wall trying to understand quantum gravity in (eternal) de Sitter. But if the bubbles don’t coalesce, it doesn’t seem that we’re out of the woods yet!

Two wonderful CD’s have recently joined my library.

Ry Cooder & Manuel Galbán’s Mambo Sinuendo is beyond hip.

And Fanfare Ciocărlia’s Radio Paşcani kicks a** as only the premier 11-piece Gypsy brass band from Romania can.

I could make some comment about the crap proffered by the major record labels. But that would be superfluous and would, moreover, spoil the mood …

So Dave Hyatt wants to hear where we’d like Safari to go.

I’d like to see MathML support come to Safari (WebCore, really). I realize that ain’t exactly mass-market, but then neither is the demographic of Safari’s early-adopters. And MathML in WebCore would open up a bunch of nifty possibilities for other applications (Keynote, spreadsheets, …) which would like to render mathematical formulas.

Anyway, how does one get from here to there?

1. An XML parser (supposedly working in WebCore now).
2. Really good CSS-handling (from Dave’s posts, it sounds like WebCore has pretty much arrived).
3. A converter to translate the MathML box model to CSS frames.

That’s the approach the Mozilla MathML Project took for their implementation:

A mathematical expression can be represented as an aggregate set of boxes. These are the bounding boxes that would enclose mathematical entities (literal symbol, operator, delimiter, etc). With rules governing the positioning of these entities (subscript, superscript, fraction, etc), it is possible to construct the box-model in a recursive manner by traversing the parsing tree of the expression.

With the object-oriented paradigm, each box can be viewed as an object that has its own specific properties and shares a common set of properties with other objects. With the CSS paradigm, each box can be viewed as a CSS frame that possibly embeds other CSS frames. Hence there is a direct correspondence between the two paradigms.

MathML offers two formats for representing an equation: presentational tags and semantic/content tags. Given an equation in either format, the MathML project will ultimately aim at constructing a lump of CSS frames that can then be passed onto Gecko for layout and display.

[Currently, they support only MathML Presentational Tags.]

Substitute “KHTML” for “Gecko” above, and one has the semblance of a plan.

(Yup, another MathML-enabled post. The usual caveats apply: you need the fonts and you need Mozilla to see the equations as they were intended.)

I gave a talk today in our Brown bag seminar about doublet-triplet splitting (and related matters) in String/M-Theory. It made me feel old because most of the audience were far too young (having been in elementary school at the time) to be familiar with the heterotic version of the story. But it also made me realize what had been (to me, at least) a nagging lack in the M-theory version of the story, as promulgated to date.

One of the great successes of the heterotic story was that one obtains the “successful” predictions of SU(5) SUSY GUTS (e. g. for ${sin}^2\left({\theta }_W\right)$), without some of the less-successful features (the doublet-triplet splitting problem, SU(5) relations among superpotential couplings, … ).

The origin of this is that, in the heterotic string, unlike in conventional GUTS, the generations do not form SU(5) multiplets. The vertex operators for quarks and leptons in a single “generation” are constructed using distinct chiral primaries of the internal SCFT. If there’s just a single $5+\bar{5}$ of Higgses, the corresponding chiral primaries must be invariant under the ${\mathbb{Z}}_n$ orbifolding. Hence they don’t give rise to massless colour triplets^*.

This feature of the heterotic story has, so far, been invisible in the M-theory approach. I strongly believe (at least in some cases) that it is still there, but it is a bit harder to tease out of the M-theory formulation.

^* If people are curious to see the details, I could put up some TeXed notes on the subject. I can’t be bothered to do it in itex.

Wow! That was weird. My web site was unavailable for about an hour this afternoon, till I noticed something amiss and restarted it.

From the logs, it looks like the trouble started at 16:49. The error_log shows

[Tue Feb 04 16:49:02 2003] [notice] child pid 12411 exit signal Bus error (10)
[Tue Feb 04 16:49:03 2003] [notice] child pid 12665 exit signal Bus error (10)

followed by a slew of such errors, up to 17:51, when I tried an apachectl restart. That didn’t work, and the error log recorded

[Tue Feb 04 17:51:25 2003] [notice] SIGHUP received. Attempting to restart
[Tue Feb 04 17:51:27 2003] [notice] seg fault or similar nasty error detected in the parent process

[Tue Feb 04 17:52:16 2003] [notice] Digest: generating secret for digest authentication ...
[Tue Feb 04 17:52:16 2003] [notice] Digest: done
[Tue Feb 04 17:52:17 2003] [warn] pid file /usr/local/apache2/logs/httpd.pid overwritten -- Unclean shutdown of previous Apache run?
[Tue Feb 04 17:52:17 2003] [notice] Apache/2.0.44 (Unix) DAV/2 PHP/4.3.0 configured -- resuming normal operations

In all these years, I’ve never had Apache die on me like that. Anyone know why it would suddenly up and die one sunny Tuesday afternoon?

For centuries, the Blood Libel — the accusation that Jews consumed Gentile children (or their blood) for ritual purposes — was a sure-fire way to whip the population into an antisemitic frenzy.

Today, the Arab Press is unabashed in evoking the Blood Libel, but in contemporary Europe, that would be deemed “unsophisticated.”

Oh no! We didn’t mean that. We were … umh … quoting Goya!

says the Independent.

(Al Quds, May 17, 2001)

(The Independent, Jan 27, 2003)

There’s a lively discussion over at Alas, a Blog — political cartoonist Barry Deutsch’s blog.

His blog is achingly beautifully designed. Seeing it makes me want to tear up and redo the design of mine.

Must … resist … !

Brad DeLong takes a break from a long series of excellent posts on the Bush economic “plan” to take Andrew Sullivan to task for attempting to average the numbers 4,6,4,5 and 8 and obtaining the answer “3—4.”

Hey Brad! Don’t be too hard on the guy!

Sully may not be the sharpest knife in the drawer, but he certainly is … umh … buff.