Older blog entries for schoen (starting at number 31)

General

Skip to the end of my diary entry, if you'd like. Thanks.

I stayed home Friday to deal with some things, and, in the evening, my friend Josh arrived from Montana for a visit. He left again early Monday morning.

On Saturday, we went to Berkeley, and then came back to San Francisco to have dinner with a friend. We found some nice places to eat.

I am trying to think of exciting free-software-relevant events from the weekend. I did buy a copy of the other Kernighan and Pike. It looks very good. I also admired some other computing books in Cody's; they have a rather large Linux section, larger by quite a bit now than their Macintosh section. (I don't know whether the situation is quite the same in bookstores outside the Bay Area.)

Linux has a pretty significant mind share now, if the number of books in Cody's dedicated to it is any indication. But a whole lot of those are the cheap mass-market application-specific computing paperbacks that are only valid for a particular software version. (People are now writing some of the same kinds of books about Linux that they've been writing about Visual Basic for a long time.) I don't like to buy those books; I'd rather have books that talk about general principles on a high level of abstraction, and then maybe some technical manuals for current versions of particular software. "Learning Red Hat Linux 6.0" doesn't appeal to me much more than "Learning Visual Basic 4.0".

I had some chats with Nick and Stephane and Rob, at various points.

I'll be working more on my law cite thing soon, and I hope that it might eventually turn into a vaguely useful piece of free software.

General

Skip to the end of my diary entry, if you'd like. Thanks.

(bneely has prevailed upon me to change how I phrase that.)

Happy Cinco de Mayo.

Professor Abelson's declaration in the New York DVD case inspired me to enter my copy of SICP into my database. Unfortunately, Abelson and Sussman are still not at the top of that list. I'm afraid that even the great Abelson can't beat the sort order of Edwin Abbott Abbott, author of Flatland: A Romance of Many Dimensions. It's hard for me to imagine who could beat Abbott. Does anyone know of an author with the last name "Aaron" who might be of interest to me? :-)

mdorman has a long and interesting diary entry which also happens to recommend Phil Agre's Red Rock Eater News Service. I've been a faithful subscriber for about a year, and I've already recommended it on my web site. If you are the sort of person who ... well, I don't know exactly what sort of person should read RRE. Maybe someone should ask Phil Agre.

I'll give the URL again as a hyperlink:

http://dlis.gseis.ucla.edu/people/pagre/rre.html

I am staying home today (Friday) to deal with some personal problems.

I have a note about demographics which I'll save for another time.

On the train (Thursday)

A passenger in my car had a seizure as the train was passing through the Burlingame station. That was scary, not least for the passenger himself. The Burlingame fire department were the first responders; they were there in about five minutes after the CalTrain conductors were notified. They had several paramedics with defibrilators, stethoscopes, oxygen, and lots of equipment I didn't recognize. After them came the Burlingame police, and finally an ambulance.

This is only the second time I've seen an emergency response. The first time was when I was in high school, and a neighbor collapsed from an allergic reaction to something he'd been prescribed. In both cases, I was deeply impressed by the responses. The people who responded were fast, knowledgeable, and professional.

In both cases, too, I felt very helpless. Back in high school, I did at least get to call for help; this time, there was nothing I could do. (I came within ten seconds of calling 911 on my cell phone, but I doubt it would have helped; I would have reached a dispatcher in San Francisco, and I hadn't even been paying enough attention to know what station we were in.)

Emergency responses are also an occasion for a lot of reflection (at least afterward!).

More law cite fun

My law cite --> URL script is now in a form where you can try it out a bit. It is primitive. It is possibly already useful to some people.

I appreciate dalke's tip on SPARK -- I'll definitely be taking a look at that. My e-mail address can be formed in the common way from UserDir URLs:

s/http:\/\/www\.\(.*\)\/~\(.*\)\//\2@\1/

and is also mentioned on my contact information page (buried way down on my web page somewhere).

ajk had some things to say about my problem, too -- I'd like to find some books on parsing and stuff, since I understand many of the concepts but never got that far in my formal CS education.

General

Click here to skip to the end of my diary entry. Thanks. (Oh, bneely, what's your objection to my use of "click here"? Do you, like me, not usually use a mouse with your web browser? Do you dislike the fact that this is phrased as an imperative? Or is it just that the use of the word "here" in links to refer to the links themselves is bad WWW style?)

I haven't had time to read all the diaries here in the past couple of days. Please excuse me if I've missed something.

I now have a plane ticket back to Massachusetts. I'll be there between Thursday, May 11, and Monday, May 15. I need to write a letter soon to make sure of seeing someone while I'm back.

Previous diary entry

A mistake in my previous diary entry erratum and a couple of things which should have been linked: OpenPatent License. MPAA's Hyperlink FAQ, DVD FAQ. (Is linking to those a copyright infringement?) T. C. May's views. Opposition to genome patents (and much of genetic engineering).

I can't find any news of the upcoming Mass in B minor performance. Maybe I misread the flyer, maybe it's just not announced on-line. If anybody knows about this (as lots of Advogato readers live in the Bay Area), please let me know.

Program design help

I now have a program which parses a few kinds of legal citations. It currently does this by using the re module in Python. (That module implements Perl-style regular expressions, which are quite powerful and flexible.)

I'm thinking of abandoning the regular expressions and writing a BNF grammar instead. (I don't know anything which can use BNF-style grammars with Python, though. I could no doubt learn bison and write in C instead.)

Does anyone have any opinion?

One reason that grammars are appealing is that a citation can have a recursive structure, e.g.

Party v. Party, Volume Reporter Page at Page (Year), Volume Reporter Page at Page, Volume Reporter Page at Page, Volume Reporter Page at Page.

So I could probably write some bison-style rules:

case: parties ", " startreporterlist;
parties: plaintiff " v. " defendant;
startreporterlist: firstreporterwithpage | firstreporterwithpage ", " reporterlist;
reporterlist: reporterwithpage ", " reporterlist | reporterwithpage;
firstreporterwithpage: reportercitewithpage " (" year ") "; reporterwithpage: reportercite | reportercite " at " page;
reportercite: volume " " reportername " " page;

These rules are actually only a tiny subset of what I'd eventually need to do, because there are so many citation forms out there.

As I think I mentioned before, the main reason for writing this is to prove that legal citations are formal and specific enough to be read directly by a computer. This might also be useful to people doing legal research, especially if they don't have a formal background in law.

I hereby threaten to try to write a parser for English phrase structure. (I did that once using the declarative database query language that's implemented in Scheme in SICP. It was slow. The basic idea is to write a lot of rules like "If A is a noun-phrase and B is a verb-phrase, then A B is a sentence." Then you see what the system can deduce about what strings are sentences.) But I guess I can work on that after I'm done with legal citations.

Patents

It's hard to co-ordinate any kind of anti-patent movement, in much the same way as it's hard to co-ordinate, say, an organized free software movement.

Some people think that {patents, proprietary software programs} are {morally wrong, improperly protected by legislation, impractical, inefficient, counterproductive} in {all cases, many general cases, certain situations}.

Sentiment against software, business method, and organism/variety/genome patents runs pretty high in certain circles. That's not necessarily in the Outside World, though. (Intellectual property in pharmaceutical products is pretty controversial too. You can imagine a world in which a cure for AIDS or cancer was known and published but nobody in most countries had the right to prepare and distribute it. Of course, economic considerations make this enormously unlikely.)

It's easier for people with the same overall perspective to agree on a strategy. Consider the difference between the GPL, releasing software to the public domain (or under an MIT-style license), and signing a (large, public) covenant among many programmers never to sue each other for copyright infringement. Depending on what people think the problems or opportunities are, different strategies will seem more appealing.

So, people who are against patents in general could sign a document saying that none of them will ever bring a patent infringement suit against anyone who has signed that document. But people who would sign such a thing are pretty unlikely to get ahold of patents in the first place, and definitely account for only a tiny percentage of patent holders.

I need to get ahold of Don and finish our discussion of the economics of patents. Well, it's only four weeks to the next Stammtisch...

Tuesday

I stayed in San Francisco on Tuesday and worked in Linuxcare's office there. I had lunch with Nick and Rob at a place called the Primo Patio Cafe (which has been mentioned on Advogato before). It was lots of fun.

Berkeley

I went to the last CalLUG meeting of the semester. It was unusually well attended. There were lots of novices there; did we make them feel welcome? I've heard two directly opposite opinions this semester -- that CalLUG doesn't do enough to make new and prospective users feel welcome, and that it does too much outreach and wastes time that way.

I feel that outreach to new and prospective users is incredibly important. On the other hand, I also expect them to be serious and dedicated if they do decide to become involved with Linux -- I guess my ideal is something like the martial arts instructors in the movies, who have no patience for students who lack commitment, but show a lot of concern for, and give a lot of attention to, those who seem to make the effort.

I got involved in a discussion of the "is free software changing the world, or is the world changing free software?" argument. (It's actually a normative question: "should free software try to change the world, or try to accomodate to people's existing expectations?".) I described the two extremes of opinion on this question and talked about GUIs versus command lines. One prospective Linux user at the meeting, when asked what attracted her to Linux, pronounced the lack of a required GUI the largest attraction. What, someone asked, did she have against graphics?

"I think they're decadent," she said.

Rounds of laughter.

No less an authority, I pointed out after she'd left, than Neal Stephenson suggested that many hacker-types had an instinct that graphical interface is decadent. I would much rather the LUG hadn't laughed at her; polite disagreement is fair.

I had dinner with a friend from CalLUG, and then quite a remarkable conversation. Thanks.

Wednesday

I went to SVLUG, where I heard Rasmus Lerdorf talking about PHP, and where there was much discussion about the recent events at Linuxcare. Thanks to Jim and Dave for rides.

A completely awesome list of people filing declarations in support of the defendant's motion against the anti-linking injunction -- there's a confusing sentence for you. These people made arguments that 2600 should be allowed to link to DeCSS. Take a look at this cast of characters! Professors, lawyers, computer scientists, security experts, Linux developers... (The actual motion is up at Cryptome by now.)

Haloch yelech u'vachoh

It has been said that having a girlfriend means allowing someone to call iopl(2) on you.

General

Click here to skip to the end of my diary entry. Thanks.

Weekend

On Saturday, as my previous entry discusses, I worked for several hours on finding a polynomial which would generate a particular sequence. I still haven't remembered where I learned the technique to do this, but it's pretty easy in principle.

I guess I could repeat the process in five or ten minutes now using the programs I wrote and downloaded.

I eventually made it to Berkeley. I had a good weekend there; I saw several friends, several of whom showed me great hospitality. We ended up eating a lot of Indian food through the weekend.

I also ran into my boss from the Lab. We discussed of the ideal efficiency of parallel algorithms (and the significance of context switch time, or conditional branch time, versus network latency). I explained an argument due to my ex-girlfriend (who doesn't know what "context switch time" or "network latency" are, but who didn't let that prevent her from making a very subtle argument about them), and we delved into it a bit further than the people with us might have appreciated. (My friend who watched our discussion said it had reminded her of a graduation party where a comedian told jokes in Hindi -- she could at least tell that the jokes were funny...)

On Sunday evening, I heard David Friedman speak about encryption and freedom on the Internet. Most of his talk was a summary of standard cryptoanarchist theory (along the lines of the writing of T. C. May, who got some credit). Unfortunately, in Berkeley, most people who would come to a talk about encryption and freedom on the Internet are already pretty familiar with cypherpunk ideas. Friedman managed to avoid claiming that encryption and remailers would dramatically erode governments' control over the physical world. Those claims are really hard to support, compared to the much solider claim that information will be very difficult to regulate if the use of encryption and networks continues to expand.

Professor Friedman autographed my copy of The Machinery of Freedom. Cool. Unfortunately, Vernor Vinge's True Names ... and Other Dangers is still out of print; many people in the audience had been looking for it.

A performance of Bach's Mass in B Minor is coming up next weekend or so. I hope a few of my friends will want to go with me.

Copyleft for organisms

graydon: I've talked to people quite a bit about genome patents and organism patents. Many of my friends are very unhappy that, say, plant varieties (or features of plants, or plant genes) are patentable under various legal systems. For some people, this is an issue of being opposed to genetic engineering in general.

I'm not opposed to genetic engineering, but organism patents make me unhappy for other reasons. So I started wondering about whether plants could be copylefted -- or, in general, could one make up an effective copyleft-style license for patent rights?

I've asked several people about this, including one or two intellectual property lawyers, and Richard M. Stallman. Stallman's opinion was unequivocal: copyleft licensing cannot work for patents, and political activism toward patent reform is necessary instead. Apparently, according to Stallman, the law of patents is so far removed from the law of copyrights that the appropriate legal strategies simply can't translate meaningfully from one realm to another.

When Stallman told me this, I was disappointed, because I wished that, if intellectual property rights are going to be enforced in organisms and their genomes, it would be nice to have something analogous to copyleft to try to keep some useful but novel plant varieties and such freely available to the public.

I know of at least one person who is actually trying to draft a copyleft-style license for patent rights. (This despite the comments of Stallman and others that this strategy isn't effective.) I wish him luck.

One obvious practical difference between copyrights and patents (in the U.S.) is that copyrights exist immediately and automatically, with no filing fee. Patents require an application process, including a filing fee, and may or may not be granted. That definitely changes the economics of patent licensing.

May Day

I hope everybody had a happy May Day.

On the train

I tried to sleep on the way down, being very tired. On the way back, I started work on a "legal citation to URL convertor" (which you can try out here). One of the regulars on dvd-discuss proposed that a demonstration of the analogy between WWW hypertext links and legal citations might help the argument against the MPAA's anti-linking motion. (They asked the Southern District of New York to ban 2600 from linking to sites that mirror DeCSS.)

So one claim is that hypertext links are much like legal citations, because they are a precise, formal way to describe how to locate a particular resource. But they don't actually provide the resource: it's your job, if you want to look up a citation, or follow a link, to do the looking-up work. The actual resource will be provided by someone else; the citation just tells you where you ought to look and what you ought to look for.

Since a variety of legal authorities are now freely available on-line, we can dramatize this analogy by making a program which turns legal citations into URLs. I've now done this, but only for four authorities. (There are hundreds of legal authorities, and I don't know how to cite most of them. But I can learn!)

This might end up being a free software project -- I could try to expand this code into a library for manipulating legal citations (parsing them, generating them, and finding or explaining how to find what they refer to). That would be a fun way to learn the ins and outs of the whole thing.

No haircut

I never managed to get a haircut, because the places I wanted to go were closed on Sunday. Oops!

General

Click here to skip to the end of my diary entry. Thanks.

Jim told me about Japanese particles. (Think linguistics, not physics.)

I had some interesting e-mail exchanges with Rachel, although I wasn't as clear and well-organized as I'd like to have been. Her diary mentions the discussion between Eric and Larry Lessig (which I've been following -- it's a fantastic discussion, and I should say more about it sometime).

I read "What Russia Teaches Us Now" a while ago (probably shortly after it first came out); I this I followed the link from Huben's Critiques of Libertarianism site. I found it very depressing. (I remember reading something very similar in print in something my father sent me, probably in The Atlantic or something.)

Since I'm still planning to go hear David Friedman on Sunday, maybe I should take the opportunity to ask him about Russia. The problem is that Friedman isn't coming to talk about Machinery of Freedom sorts of things -- he's going to speak about economic analysis of law.

graydon's comments about LSD and Buddhism again remind me of the story of the Sirens in the Odyssey.

Nick announced that our roommate Karl is planning to move out, so there followed much discussion about whether we should try to get another roommate or not. (It's mostly a classic space vs. rent thing.)

The DVD cases have some interesting things coming up, most notably a "right to link" issue. The defense in the New York case is expected to respond very soon to a motion for an injunction which would forbid 2600 from putting links on its web site to of DeCSS (among other things).

This could be an important precedent, whether that motion is granted or denied.

BBC

Thanks to the good people of the Linuxcare web team, you can now download a BBC source tree directly from Linuxcare's web site. (Previously people were supposed to e-mail us and such.) I hope this will be helpful to sethcohn and the Lubbock Project.

Trading CS books for law books

I was going to trade a copy of K&R for a Blue Book, seeing as I had a spare K&R (sort of) and I know a lawyer who has a spare Blue Book. But she had already bought a copy of K&R. So, I get the Blue Book, and she gets an IOU.

Learning legal citation is reasonably useful, so that lawyers won't laugh at you when you write about the law. I've gotten as far as knowing how to cite Supreme Court decisions, USC, and CFR correctly in principle. There are still a lot of things I have absolutely no clue how to cite.

On the train

I listened to some music and thought about some things. I started a letter to Wolfgang. I finished my expense report for my trip to Southern California last month. No math recently; I've been trying to catch up on sleep and other things.

Book collecting

I reached the 500th book in my on-line catalog (which was Fraleigh's First Course in Abstract Algebra). And, speaking of algebra,

Let's do algebra instead of getting a haircut!

I wanted to go to Berkeley early on Saturday. Instead, I spent about three or four hours trying to simplify a polynomial. But I learned a lot from the process.

First, I got introduced to a library called SAML; then, I tried to build the Python stuff locally, and got stuck because I didn't have the Python API header files. While I was waiting for the Python source code to download over a 28.8 link, I ended up learning a little of the Samuel language, which exposes the string parsing functions of SAML, which was all I wanted to use in the first place.

I was doing this because a friend took her teacher certification exam this morning, and sent me a "complete this sequence" question:

1, 3, 7, 13, ___, 31, 43, 57

My friend guessed 17.

If you happen to want to try this problem yourself, don't read anything below this paragraph.

(Well, in case you didn't believe me, I'm giving you one more chance to stop reading now before you see my answer, which is, incidentally, not 15, in case your eyes are magically attracted to numbers in the paragraphs below where you are supposed to be reading...)

^L

OK, when I saw the problem, I did successive differences and immediately came up with the answer 21. I gave my friend a few justifications for why the answer should be 21 (the successive differences are the even numbers 2, 4, 6, 8...), and then I said "Now I have to come up with a justification for why 17 is correct, so that you can claim to be right anyway. :-)" -- so, three and a half hours later, having located, downloaded, built, and experimented with SAML and a bunch of Python programs that generate other Python programs or Samuel programs, I came up with

f(n) = (1/36)n^7 - (31/36)n^6 + (391/36)n^5 - (2581/36)n^4 + (2386/9)n^3 - (4882/9)n^2 + 563n - 223

This polynomial has the nice property of generating the sequence 1, 3, 7, 13, 17, 31, 43, 57 when applied to the integers, thus justifying my friend's answer. :-)

I forgot where I learned the trick I used for getting an expression for that. But the expression itself was

((-(1/5040)*(n-2)*(n-3)*(n-4)*(n-5)*(n-6)*(n-7)*(n-8))+((3/720)*(n-1)*(n-3)*(n-4)*(n-5)*(n-6)*(n-7)*(n-8))+(-(7/240)*(n-1)*(n-2)*(n-4)*(n-5)*(n-6)*(n-7)*(n-8))+((13/144)*(n-1)*(n-2)*(n-3)*(n-5)*(n-6)*(n-7)*(n-8))+(-(17/144)*(n-1)*(n-2)*(n-3)*(n-4)*(n-6)*(n-7)*(n-8))+((31/240)*(n-1)*(n-2)*(n-3)*(n-4)*(n-5)*(n-7)*(n-8))+(-(43/720)*(n-1)*(n-2)*(n-3)*(n-4)*(n-5)*(n-6)*(n-8))+((57/5040)*(n-1)*(n-2)*(n-3)*(n-4)*(n-5)*(n-6)*(n-7)))

which took SAML to simplify, as I wouldn't have wanted to do it by hand. There are some very pretty reasons why the signs alternate.

Maybe I will publish a Python program which uses SAML to print a polynomial which generates an arbitrary integer sequence. The algebraic manipulation is the only missing piece, and it is kind of fun to be able to do this.

The only practical application I know, though, is proving the validity of completely absurd solutions to integer sequence completion puzzles.

The only problem is that this whole exercise made me something like five hours late getting to Berkeley, which ruined my plan of getting a haircut over there. However, people who implement free symbolic math libraries and languages rule!

General

Click here to skip to the end of my diary entry. Thanks.

My beard is coming back, gradually. I believe that there is probably great psychological significance in this.

When I wrote (in my discussion of evangelism recently) about the changing demographics of people trying Linux, I said that many people "don't becoming Linux users"; that should have been "don't become Linux users".

pvg

When I first started looking at the Pascha/pascho thing, I didn't know that "Pascha" was related to "Pesach". At least, I don't think I knew that. But it turned out that I was not the first person to think there was a connection.

The Greek term for Easter, pascha, has nothing in common with the verb paschein, "to suffer," although by the later symbolic writers it was connected with it; it is the Aramaic form of the Hebrew word pesach (transitus, passover).

lilo

I wonder where Rob went.

Telsa

Metamagical Themas is much "easier" than Godel, Escher, Bach, because you can pick and choose which short essays you're interested in; they aren't all so tightly interwoven. The articles in MT don't really depend on one another, so you won't necessarily get confused by reading them out of order or by picking and choosing among them. I thought GEB was a great challenge, and it has been my favorite book, but it's definitely kind of scary.

Metamagical Themas is a collection of short articles on different subjects -- nuclear war, sexism, linguistics, LISP, the Rubik's Cube, Chopin, game theory and evolutionary biology, calligraphy, artificial intelligence, numbers, parapsychology... it's pretty cool.

rachel

I saw your article about Linuxcare. So did a lot of other people at Linuxcare. We must be interesting to the outside world or something... sort of like Red Hat.

Metagato

Maybe somebody should just register metagato.org and...

On the train -- Monday

On the way down:

I agree with the recent comment about high-level languages and library functions, assuming I remember it correctly. :-)

It's amazing to think about the difference between people whose computers are likely to be CPU-bound and people whose computers are not. This is the underlying point of the claim that most people don't need computers as fast as those presently on the market.

Clearly, if you are using your computer to try to do computations, then you need all the CPU time you can get. I've been thinking about calculations of pi again, after looking through Pi: A Source Book. If you're not trying to calculate a constant, or perform a simulation, or generally run something CPU-intensive, then you don't need such a fancy computer. If you strace most of the processes running on a desktop system, you will see that most of them are blocked on I/O most of the time (unless they are doing computations, or are poorly written). I mean, let someone log in as you remotely, and "strace -p" various things you are running as you run them. (I've attached an strace to the editor I'm using to write this entry. By switching back and forth, I can see that vim spends almost all of its time in select(2), waiting for Linux to tell it that I've typed something. The computer is, of course, millions of times faster than I am, even though I am still the fastest typist I know.)

Load averages tell the same basic story.

Lewis Carroll, I've discovered, shared my objection to the premises of certain arithmetical proportion problems. (As I put it, "If nine women can have nine babies in nine months, how many women will it take to have one baby in one month?" I think this analogy is originally due to F. P. Brooks or one of his colleagues.)

Once upon a time, I was actually asked such a question on a test, and I objected. I think the problem was given in terms of pizzas.

Suppose C chefs are making pizza. It is observed that they can make P pizzas in time T. The naive model says that everything is proportional; there is a rate R=kC=P/T such that any number of chefs can make pizzas in a quantity proportional to the number of chefs and the time spent on the exercise. (Actually, the really silly assumption is that there is a constant k such that k=R/C: that is, that each chef adds a constant number of pizzas per hour to the overall rate of pizza production!)

This assumption is ridiculous in several ways:

1. The pizza-making process consists of several steps which must be performed in a particular order (for each pizza). These steps have different characteristics.
2. There might be an absolutely fixed-time component required for one step (e.g. every pizza must be baked for 11 minutes, so the pizza-baking step can never be reduced below 11 minutes, regardless of what resources are applied to it).
3. There might be a limit on parallelization of one step (e.g. the oven used to bake the pizzas might only hold 8 pizzas at once, so no more than 8 pizzas at a time can ever be in the baking stage).
4. There might be diminishing returns to some resource (e.g. the "too many cooks" problem: if there are more than 12 chefs in the kitchen at once, they start to bump into each other and slow each other down).
5. There might be imperfect divisibility of some resource (e.g. a chef might be distracted by having to spend only one second at a time working on each of 20,000 pizzas).
6. There might be some step which cannot be scaled up or down at all (e.g. maybe some aspect of pizza-making requires exactly C chefs, one for each of C distinct and necessary roles, and is not aided by the addition of more chefs, but would be impossible with fewer chefs).
7. Some steps might be subject to a strict serialization constraint (e.g. it is not possible to begin preparing meat pizzas at all until all of the kosher pizzas that the kitchen will ever produce have been finished; chefs who specialize in some aspect of meat pizza production must sit around and watch until the kosher pizzas are all done and out of the kitchen).

Infinite divisibility and arbitrary scaling of processes are problematic assumptions. If a problem says that we should assume that every aspect of a process can be scaled arbitrarily and linearly, then this problem has no connection to anything that people can do in the real world. (It might be close to things computers can do, though.)

If you really made a large number of pizzas, some steps would be labor-bound, some steps would be time-bound, some steps would be space-bound, and so on.

On the way back: I took a nap.

Here's what Carroll said (I subsequently found it again in John Fisher's The Magic of Lewis Carroll, pp. 140-142).

If 6 cats kill 6 rats in 6 minutes, how many will be needed to kill 100 rats in 50 minutes?

This is a good example of a phenomenon that often occurs in working problems in double proportion; the answer looks all right at first, but when we come to test it, we find that, owing to peculiar circumstances in the case, the solution is either impossible or else indefinite, and needing further data. in 6 minutes, how many will be needed to kill 100 rats in 50 minutes?

This is a good example of a phenomenon that often occurs in working problems in double proportion; the answer looks all right at first, but when we come to test it, we find that, owing to peculiar circumstances in the case, the solution is either impossible or else indefinite, and needing further data. [...]

[W]hen we come to trace the history of this sanguinary scene through all its horrid details, we find that at the end of 48 minutes 96 rats are ead, and that there remain 4 live rates and 2 minutes to kill them in: the question is, can this be done?

Now there are at least four different ways in which the original feat, of 6 cats killing 6 rats in 6 minutes, may be achieved. For the sake of cleaness let us tabulate them:

A. All 6 cats are needed to kill a rat: and this they do in one minute, the other rats standing meekly by, waiting for their turn.

B. 3 cats are needed to kill a rat, and they do it in 2 minutes.

C. 2 cats are needed, and do it in 3 minutes.

D. Each cat kills a rat all by itself, and take[s] 6 minutes to do it. [...]

This, then, is an instance of a solution made "indefinite" by the circumstances of the case. If an instance of the "impossible" be desired, take the following: "If a cat can kill a rat in a minute, how many would be needed to kill it in the thousandth part of a second?" The mathematical answer, of course, is "60,000," and no doubt less than this would not suffice; but would 60,000 suffice? I doubt it very much. I fancy that at least 50,000 of the cats would never even see the rat, or have any idea of what was going on.

Or take this: "If a cat can kill a rat in a minute, how long would it be killing 60,000 rats?" Ah, how long, indeed! My private opinion is that the rats would kill the cat.

Python

I wrote my first Python one-liner Monday. It was fun. Thereupon, I also wrote the fragment below, which uses tuple, map, and divmod on the same line. Python's choices of built-in functions are pretty neat.

This is a way (I bet Python already provides a way) to print out integers formatted (US English locale-style) with commas every three places:

def comma(n):
        if n < 0:
                return "-" + comma(-n)
        if n < 1000:
                return str(n)
        return "%s,%s" % tuple(map(comma,divmod(n,1000)))

I think Python is something like a cross between BASIC and Scheme. Or maybe it's just my use of Python that's like a cross between BASIC and Scheme. :-)

Why doesn't the "%" string formatting operator allow you to use any sequence type, instead of just tuples? If it could work with a string and a list, I could say

return "%s,%s" % map(comma,divmod(n,1000))

without creating any particularly huge amount of ambiguity.

</BLOCKQUOTE>

On the train -- Tuesday and Wednesday

I took some naps and also ported my book collection web page generation script from awk to Python. The awk version is much faster than the Python version. But the Python version helped me find some typos in the database, mostly because there are very few things that will actually ever cause an error in awk. (Array overruns? Non-existent dictionary keys? No problem!)

Conference schedules

Why are the O'Reilly Open Source Conference and the Ottawa Linux Symposium happening at the same time?

Wisdom teeth

Now I am getting very worried about my wisdom teeth; as they come in, they hurt slightly in various ways. Teeth do hurt when they come in, which is no problem; my concern is whether the pain will persist and be a sign that I need to see a dentist about this.

Wednesday

After work, I went to the Linuxstammtisch at Tied House in Mountain View. I'd never been to a Linuxstammtisch before, even though they've been held for over a year. I was impressed with Mountain View. I used to be under the impression that Mountain View was some sort of huge suburban wasteland, like some of the parts of Southern California I've seen. It's not true!

I had this impression because I had never been to downtown Mountain View; all I had ever seen was computer companies, and heard other people's descriptions. So, downtown Mountain View is actually chock full of restaurants, and (wow) bookstores. These, for me, are the two largest direct benefits of cities. What am I to make of this? I would not necessarily have minded living in Mountain View -- although it would make getting to Berkeley quite difficult.

The restaurants around Castro Street are quite appealing, even if some of them look like the were taken out of an upscale mall. There is a nice variety of Asian food, and a lot of other things. I look forward to coming back and trying some of them -- and going book-hunting.

The Linuxstammtische are held at Tied House, a brew pub and restaurant in Mountain View. It was quite nice, and the food was very good, much better than anything I have ever found at a bar before. (Not that I spend a lot of time in bars.) I finally learned why it's called "Tied House": it's based on an old name for a tavern which served a particular brewery's beer (in this case, its own!), as opposed to a "free house", which served the beer brewed by the breweries of its choice. I wonder if the Department of Justice will investigate. :-)

I saw lots of people I knew at the Linuxstammtisch, including Deirdre, Don, Heather, and Jim. (Also Rick, who's not on Advogato.) South Bay Linux events seem to get a very large VA contingent. The conversations were pretty geeky, and pretty loud (with the background noise of Tied House to compete against). There were some forseeable conversations about intellectual property law, and Don told me about some interesting stuff he was thinking about.

Don is fun to talk to. (Sorry I didn't get to talk to other people so much -- it was pretty noisy and hard to hear.)

Latin morphology and inflection code

Perseus clearly has some nice Latin and Greek morphological analysis software. You can put in something like "anesti", and it immediately tells you something like "oh, that's prefix ana- + third person singular indicative of eimi", which is cool, which is hard for people to learn to do in real life, and which is somewhat more useful than a dictionary, especially for compounds and irregular verbs (or compounds of irregular verbs).

In eighth grade, I wrote some Latin verb conjugating software in BASICA (i.e. GWBASIC). It was pretty nice, with infinitives and most finite forms produced from principal parts of a first-conjugation verb. (I had trouble with some of the rules for what happens to stem vowels in other conjugations. It can be tricky, although it's possible to describe somehow.) Among the many problems with this (completely aside from the awful language I was using) was the fact that it only went one way, from a description of a form to the form.

Nowadays, I might try to write a Latin inflections library, or a Latin parsing library (presumably in Python). It would be fun. But I'm wondering whether Perseus or someone else has released code that's useful for manipulating Latin words. I don't necessarily want to reinvent the wheel, but I don't know how Perseus feels about free software or where their code comes from. Is there something I should look at?

General

This is longer than my usual diary entries, even, because it relates to the whole weekend. If you don't like this, click here to skip to the end of this entry (via a "#" link). Thanks.

I stayed in San Francisco on Friday and went to the Linuxcare company lunch and meeting. (Company lunch: pasta and pizza. This subsequently made me think about the lack of Passover-friendly food; see below.)

In the afternoon, Duncan and I worked on the Bootable Business Card. I hadn't had time to work on it at all recently, because I was out doing other work for Linuxcare.

I heard from my friend Anirvan, who's read my Advogato diary, and sent me a message about it. He had a number of very interesting comments on the phenomenon of web logs and web diaries.

Welcome to dmarti. ("Don Marti rules.")

Berkeley

In the evening, I went over to Berkeley. After having dinner with one friend, I went to Douglas Hofstadter's lecture, where I saw nine people I knew. (Hmmmm, you don't suppose I tend to meet people who would know who Douglas Hofstadter is, do you?)

I enjoyed Hofstadter's talk, and found it interesting. He discussed the various uses of "guy" (basically "a person" versus "a male person"), and some problems caused by these uses.

On the other hand, this talk was really not as exciting as other things Hofstadter has done. I heard him give a talk on translating poetry, a year or so ago, and it was more exciting. On the subject of sexism in language, Hofstadter's essay "A Person Paper on Purity in Language" (in Metamagical Themas) was much better, and I recommend it highly. There, he was completely dedicated to the point that sexism is language is an important and challenging issue, and he proved that point admirably.

One friend of mine was annoyed that Hofstadter had mixed descriptive linguistics (here are some ways that different people use "guy" in different situations) with prescriptive social activism (the implication that you should not use "guy" to refer to women, because of the negative consequences of doing so). Academic linguists usually do not mix prescriptive and descriptive treatments of language (and usually shy away from the prescriptive entirely, at least insisting that it does not belong to linguistics). My friend believed that Hofstadter crossed back and forth too carelessly.

I certainly respect Hofstadter a lot for his activism, as well as for his incredible academic work. (You can see some activist essays in Metamagical Themas, where there are several chapters on the challenges of preventing nuclear war between the United States and the Soviet Union, in addition to the discussions of sexism in language. Hofstadter was courageous to decide to present these issues to his geek-and-linguist audience, and he used his talents well when he did so.) But I can appreciate the criticism that activist commentary and academic analysis don't always mix well. I think that, in this case, Hofstadter's talk would have been better if it had been more firmly on one side or the other. Very possibly, fewer people would have come to hear it.

Of course, it's very possible to be led to moral conclusions through academic work -- but people who present them together as such are going to be on pretty thin ice a lot of the time.

I'd like to see Hofstadter come back and talk about translation or concepts or fluidity in lexical categories or metaphors (maybe he could debate George Lakoff?), and then give a separate talk about sexism. I'd attend both.

Philosophiae, Quinte, semper vaco

After the talk, I was invited to a friend's house for what was originally billed as a discussion of what we'd just heard. The basic idea was that a number of people interested in philosophy or linguistics or cognitive science should come for a party at which we would all chat about Hofstadter's ideas.

This was a good idea, although it didn't necessarily come off exactly as planned: the amount of Hofstadter discussion was fairly small compared to other stuff. We had some fairly wide-ranging conversations.

I really had a lot of fun, and met some very interesting people. Maybe I should try to throw parties and get people talking to each other. That is a new idea for me; would it count toward my collaboration rating? :-)

Agreeing with Paul

The "I Agree With Paul" campaign was a big deal and a big topic of discussion all day, as well as at the party I went to. The basic idea was that some evangelical Protestant student groups got someone (a Berkeley student named Paul Lai) to write a personal statement of faith, whereupon almost all of their membership expressed their agreement with Paul's statement, and spent Holy Week ("Paul Week"?) wearing orange "I agree with Paul" shirts. (I'd like to link to Paul's statement, but I don't believe that it's online anywhere; it was printed in an ad in the Daily Cal, and on a few flyers.)

Amazingly, the evangelists did not register iagreewithpaul.com, iagreewithpaul.org, agreewithpaul.com, or agreewithpaul.org -- which I really would somehow have expected of a Bay Area movement run by twenty-year-olds. Not even paulweek.org. Hmmmm.

A couple of non-evangelical Christian friends told me that they were unhappy with the campaign, either because they felt that it was either hurting the solemnity or seriousness of religion, or because they did not agree with Christian evangelism in general.

I thought that the campaign was very effective because it got so many people to think about it and talk about it (and to do things like spend time writing Advogato diary entries about it). In fact, other campus groups started using the "I agree with..." formula in their own advertising. (I saw "Do you agree with Hofstadter?" flyers!) Some anti-Christian or anti-evangelical students ended up wearing their own orange t-shirts with other slogans (mostly "I agree with Nietzsche", which is actually a more complicated proposition than some people may have realized).

From the point of view of "exposing people who have never really heard about it to the gospel", even getting people to wear shirts that say that they disagree with Paul (of Tarsus, no doubt, as well as Paul Lai) is a great success. Many evangelists have said that the biggest problem is when people ignore their message -- not when people disagree with it. Every person who sees an anti-Paul shirt will thereby at least think about what Paul had to say, right?

I mean, I remember that lots of Linux activists were really happy when Microsoft started criticizing Linux, and running anti-Linux material on its web site. There are many reasons for this, including a peculiar sense of validation. Previously, Microsoft had scrupulously and deliberately ignored Linux, in a way that made it look completely out of touch with the technical community. By expressing its view that Linux was worthless, Microsoft proved that it thought that Linux was at least a worthy competitor, worth spending time and money to discredit.

And, in fact, by criticizing Linux, Microsoft made (in some people's view, certainly not everyone's) the PR battle turn from a fight to prove that Linux exists to a fight to prove that Linux is better. (Remember when the big problem for the spread of Linux was that nobody outside of hard-core geek circles had heard of the thing? Now, for the first time, we have large numbers of people trying Linux who subsequently actually don't like it and don't becoming Linux users! Previously, self-selection and other factors made this sort of thing fairly uncommon.)

So, even spending the time to argue against something helps keep it on people's minds, and helps send the message that you believe, and can understand, that someone might well at least find it worth considering. It lends credibility, or, if not credibility, at least an appearance of significance. Every single person who wears a shirt for or against the "I agree with Paul" campaign helps send the message to the public that he or she thinks that the question of who Jesus was is actually a question worth spending time and effort on. (This diary entry is itself also a victory of exactly the same sort I'm describing.)

So this is the strange and indirect triumph of the campaign, completely outside of how many people are converted directly by what they hear on Sproul Plaza. But the tactics are certainly making some observers very uneasy.

I guess my biggest problem with the campaign is that, for some people, it seems to reduce religion (and fundamental questions of theism and theology) to the level of a t-shirt slogan. The idea of uniting around a formal, common statement of belief is pretty old, and pretty common in Christianity. But these statements are traditionally longer, more sophisticated, and then also more elaborately justified, discussed, and understood.

The Nicene Creed is the statement of faith with which I'm most familiar. It (like the MIT X license) is certainly very concise. But that creed is the result of very sophisticated thought and discussion, and has a very detailed justification. In some sense, it was a major intellectual accomplishment. (I realize that some people, including some people who believe the Nicene Creed, don't think that that is an obvious virtue of a religious doctrine.)

The t-shirt slogan "I agree with Paul" is just not comparable to a religious tradition like that. It's just not as serious; it doesn't rely on or appeal to the same tradition of theological discussion. I guess that what I'm getting at is that religious t-shirt slogans seem sort of anti-intellectual to me, in comparison with other evangelical tactics like writing books. Last week, I was talking to someone about St. Anselm's ontological argument for the existence and perfection of God, and (shortly afterward) reading some of Phillip Johnson's work, like Reason in the Balance, and criticisms of it, like Tower of Babel. It seems to me that this is a much better (more respectful, more committed, more fruitful?) approach to the question of God's existence (and the truth of particular religions, claimed divine revelations, etc.) than the apparent alternative:

Oh yeah? Well, I have an orange shirt!

You do not!

Do too!

Do not!

Do too!

The last word probably belongs to the student who supposedly showed up somewhere in an "I agree with Kierkegaard" shirt.

Local law

I've been thinking further about kelly's comments about the "local law" provisions in licenses.

I'm familiar by now with the difference between a conditional license and a contract (although there are lots of efforts to prove that the GPL should be read as creating a contractual relationship, in order to make it irrevocable). I would like to know more about the consequences of different kinds of language in non-contractual licenses, and, for example, what happens if you put a severability clause or a non-severability clause in a non-contract license.

I'm still looking for ways to make people feel comfortable not requiring compliance with local law. :-) (This because I still think that doing so would violate the OSD, DSFG, etc., although I'm not sure I could convince people of this claim.)

There are also some subtle questions about the interaction between civil and criminal law (e.g. depraved indifference, contracts used to aid criminal purposes, other weird situations). The problem is that I originally only really wanted to learn about intellectual property stuff, but it seems impossible to have a good understanding of only one portion of a legal system -- like having a good understand of only one portion of science, maybe?

I'm probably going to need to think about the local law stuff, although maybe we can ask the lawyers who proposed this new license what they were trying to accomplish with that provision.

Food

Passover always makes me think a lot about what people eat and what they don't eat, and what they believe about what they eat, and what they believe about what other people eat, and what the social, political, and interpersonal consequences of these things are.

Some of these things are quite political, so to speak.

I mean, there was a big flamewar at my high school once upon a time on our e-mail and conferencing system about the school's food policies. Sorry, make that two flamewars. One was about whether animal rights supporters were right to wish that the school served only vegetarian food (the flamewar didn't reach the question of whether or not it was right for the school to serve meat, but only whether it was right for someone to wish that the school didn't serve meat). The other was about whether the school or students should take what might be called nonintuitive steps to satisfy the religious beliefs of Muslims at the school.

So, most people would probably say that the best approach is to speak of courtesy rather than moral obligation.

I could outline quickly a few possible reasons for not eating something:

• "I have a moral or religious objection to people in general eating X"
  • moral vegetarianism, religious vegetarianism
  • politically-related boycotts (labor, trade, environmental issues)
  • GMO activism

• "I have a moral or religious reason not to eat X myself at this time"
  • any of the above
  • religious dietary laws
  • beliefs about processed foods, organic food, etc.

• "I have a 'mandatory' health reason not to eat X myself"
  • I know someone who is allergic to gluten and someone who is allergic to garlic

• "I have an 'elective' health reason (or other explicit elective reason) not to eat X myself"
  • being "on a diet"
  • being vegetarian for long-term health reasons
  • lots of other things

• "I don't like X"
  • e.g., I don't like tomatoes

• "I don't feel like eating X"

What should you do if someone tells you any of the above things? How does what you should do depend on other aspects of the situation you're in?

There's a lot more I could say about this, but the issues go off in all different directions, and I don't know which thread to follow. I just know that satisfying people's beliefs and rules around food is quite difficult and provides examples of all kinds of other interesting controversies.

I've heard some people get very unhappy because a large entity (firm, school, conference...) did not properly account for their dietary rules. (Passover is a common occasion for this; perhaps someone at Linuxcare could have been upset that absolutely everything served at lunch was chametz.) On the other hand, I can probably find some examples where some of the same people would not be upset that other dietary rules hadn't been accomodated. So I guess that people tend to divide dietary rules into different categories, and have different intuitions about the relative importance of making accomodations for different rules.

Saturday

On Saturday, I went to Stacey's and got some fantastic stuff. I got a copy of Who is Fourier?, Recreations in the Theory of Numbers (recommended by someone here on Advogato), several other things, and a birthday present for a friend.

I went to dinner at Steps of Rome, where I met my friend Ben. He came by and we talked for quite a while.

Sunday

stephane mentioned Ukrainian Easter eggs, which prompted me to look up more information about Galicia. (My family on my mother's side were supposedly mostly Galician Jews, a century or two ago. According to what I found on the web, this means they were likely Austro-Hungarian subjects for most of the time they lived there. I don't know whether they lived in Western or Eastern Galicia.)

Galicia is one of these parts of Europe with a long and complex history. Well, I guess all of Europe is like that.

I also talked to Nick about the calculation of Easter. It is more complicated than I thought -- he tells me that Old Calendarist and New Calendarist Orthodox Christians celebrate Pascha on the same day (which may or may not be the same day the Western churches do, in a given year), even though they celebrate other holidays on different days. (So there are actually three different ways in wide use of calculating Christian holidays -- Western-style Gregorian, Eastern-style Gregorian, and Eastern-style Julian.)

Since Nick and I weren't celebrating Easter today, we went out for Indian food. We ran into Joey on Muni. Then we ended up at the new branch of Black Oak Books in the City, which was fun. (I know some of the owners of that store, and they happened to drop by while we were there.)

kjk is looking for Vernor Vinge's "True Names"; I've been looking for that, too, for two years, and I haven't found it, despite checking in a few dozen bookstores, including science fiction bookstores. The book is True Names ... and Other Dangers, and, well, they told me that it was going to be reprinted Real Soon Now. It's very definitely out of print. (By the way, it's "Vernor" rather than "Vernon".)

ObFreeSoftware: I first heard of Vernor Vinge through the Jargon File bibliography, and then subsequently through the work of the American libertarian anarchist economist David D. Friedman (they cite each other in connection with Vinge's story "The Ungoverned" and Friedman's book The Machinery of Freedom; Friedman's speaking on Sunday in Berkeley, and anyone who's interested can e-mail me for details). But it turns out that Vinge also has a connection to the free software world; he is an occasional advisor to the FSF, and he's been a judge on a panel for the first FSF free software award.

I wrote to the FSF to ask what Vinge's connection was; they wrote back and said, more or less, that he had been a long-time supporter of the concept of free software.

I get it!

General

I heard from several friends, which was nice.

On the train

Going down: I did a tiny bit of consulting on the phone, felt mostly tired and unproductive, and should have taken a nap.

I planned my mini-vacation to attend the Smith College graduation in Northampton on May 14.

Coming back: I did take a nap. I was exhausted, and the conductor had to wake me up when we got to San Francisco. "Last stop!"

Licenses and local law

I talked to Jim about "indifference" and "depraved indifference", and whether you can have legal liability under various theories from putting code in the public domain.

kelly says that licenses might require compliance with local law so that courts won't re-interpret the licenses unfavorably in case of a conflict with local law.

Wouldn't a severability clause or a non-severability clause work for this purpose? The GPL is clear enough on this:

4. You may not copy, modify, sublicense, or distribute the Program except as expressly provided under this License. [...]

7. If, as a consequence of a court judgment or allegation of patent infringement or for any other reason (not limited to patent issues), conditions are imposed on you (whether by court order, agreement or otherwise) that contradict the conditions of this License, they do not excuse you from the conditions of this License. If you cannot distribute so as to satisfy simultaneously your obligations under this License and any other pertinent obligations, then as a consequence you may not distribute the Program at all. For example, if a patent license would not permit royalty-free redistribution of the Program by all those who receive copies directly or indirectly through you, then the only way you could satisfy both it and this License would be to refrain entirely from distribution of the Program.

Other licenses have taken a very different approach and explicitly added severability clauses.

Nobody seems to think that either approach makes a license non-free. (It's the legal system's own fault, not the licensor's, if some jurisdiction isn't willing to respect the licensor's intentions!)

Is there something that a clause requiring compliance with local law gets you that a severability clause doesn't?

Speaking of preaching for Linux...

I got to see Don, who has my vote for Most Likely to Found the Real-Life Church of Emacs. All of you go read Linuxmanship now, if you haven't.

I'll put down the Linuxstammtische on my calendar now.

We're all mad here

$ zless glibc_2.1.3-8.diff.gz
   Per me si va ne la città dolente;
   per me si va ne l'etterno dolore;
   per me si va tra la perduta gente.

    Giustizia mosse il mio alto Fattore;
   fecemi la divina Potestate,
   la somma Sapienza e 'l primo Amore.

    Dinanzi a me non fuor cose create,
   se non etterne, e io etterno duro:
   Lasciate ogni speranza, voi ch' entrate.

Shared libraries

I've been learning a lot about shared libraries this week, thanks mostly to dhd and to some experimentation. (I've also heard from other folks like Nick and David on this subject.) I'm trying to learn how shared libraries are organized and how to manipulate them in various ways.

I'm sure that some little birds could tell me more about this subject if I could just find the right opportunity. This reminds me that I really want to go to OLS.

My basic spring and summer travel schedule, if I can make it work out, would include Northampton (Smith graduation, elsewhere noted), Ottawa (OLS), and Las Vegas (DefCon). I might also try to sneak off to Los Angeles for another CPTWG, if I get in the mood.

If I ever finish writing "Third Parties and the End-to-end Model", I could submit it to a conference.

I really, really wanted to go to Portland, OR, and Seattle, WA, ever since September. But things really haven't worked out for that trip. It's actually very sad. I've now heard that two friends of mine are planning to drive up there after they graduate from college this spring -- but I don't know that I'll be able to take time away from work to join them.

Dale Carnegie

Re kelly's diary: I think the book was "How to Win Friends and Influence People".

pie

We got my colocated machine (pie) back up fairly quickly, thanks in part to tomsrtbt. (It's a very impressive piece of work. Thanks, Tom.) I couldn't use the BBC because pie doesn't have a CD-ROM drive.

General

I continue to enjoy reading other people's diary entries, as well as their comments on or responses to my diary, even if I don't mention you by name. Thanks.

On the train

Leaving: slept. Coming back: should have slept, but instead wrote a poem. (I hadn't written a poem since December, and in fact this poem and that poem could be called counterparts.)

Work

I worked on cool stuff.

I regret not being able to talk more about Linuxcare. Everybody (who cares to do so) is talking about us, there are leaks and alleged leaks, there are controversies, there are rumors, there are slashdot flamewars, but there are precious few public opinions and comments of actual Linuxcare employees.

Hofstadter's talk

I got my sixth and seventh invitations to Hofstadter's talk. Thanks, Sumana. I hope to see you (and almost all of the people who invited me) there.

I'm amazed at how many of my friends know I'm a Hofstadter fan, and how many of them are actually Hofstadter fans themselves. I remember when I was the only person in my entire school who had heard of Douglas Hofstadter.

A sanitary and tidy way to eat buttered popcorn

Jim reports that, in a previous job, he knew someone who would eat popcorn with chopsticks, to avoid getting butter on his keyboard. That somehow seems to make a lot of sense to me.

Holidays; etymology

Holy Week -- evangelical Protestants at UC Berkeley are doing a sort of quiet but very visible outreach program to try to get people to talk to them about religion. I wonder whether other groups could use some of the same techniques. (I did preach for Linux on Sproul Plaza once, complete with readings from the Linux Bible, but I didn't have the same kind of organization that these Holy Week evangelists are practicing.)

Pesach -- there some disappointment within my family that I am not coming to celebrate Passover with them. I assume that, mutatis mutandis, my girlfriend probably would have taken me to a seder. As it happens, I'm not doing anything to observe the holiday.

I could probably get into the etymologies again this year, but lots of my friends have already heard the result. A year ago, I was still living in Berkeley, and I went and bought the Oxford Greek Dictionary (LSJ -- Liddell, Scott, and Jones, now also McKenzie) to resolve the question of whether "Pascha" (Orthodox Christian name for the Easter/Passover holiday) has to do with "pascho, paschein" (to suffer).

It seemed obvious to me that it should -- I heard about "paschal lambs" and so on, and it would make lots of intuitive sense to imagine a connection here. (In fact, Christian tradition definitely makes such a connection. See 1 Corinthians 5:7-8.)

To my astonishment, it turns out that "pascho" (from which we get all kinds of stuff -- "passion", "pathos", "pathetic", "sympathy", "pathology", "psychopathology", "passive", and so on) is thought to have absolutely no connection with "Pascha" (which is simply the Greek for the Hebrew proper noun "Pesach", which has to do with sparing, overlooking, omitting, but not suffering, experience, etc.). This is particularly funny in that the "Passion" and "Pascha" of Christian tradition commemorate the same event (sort of), and their Greek roots differ by only a single vowel, yet are not cognate.

Nick's father, a librarian and polyglot, reached the same conclusion, but I don't think he even had to look it up. :-)