I haven't been able to find a database of the shortest complete sentences in various languages.
I'm sure about Latin "i", Portuguese "é", English "go"/"do"/"be", pretty sure of Spanish "es" (also "va", "ve", "sé", "he"(?), "ha"(?), "hé"(?)), and fairly sure of German "geh" (also "sag"/"sei"/"hol"/"übt" and possibly some other three-letter imperatives).
If Hebrew "יש" is a complete sentence, it would probably win.
Maybe I should put up a web page and solicit contributions.
Update: It's not clear to me whether the -e on the familiar German imperative should be considered obligatory (my textbook didn't use it and I found a textbook claiming it's considered optional); without it we can probably form "üb" and maybe "sä" and "öl".
Update 2: if we consider "fi" a mere interjection, the Esperanto winner will be the imperative "amu" and about 15 other possibilities.
I wonder why some young people who hear that 'fish' is the same thing as 'fish' are totally unperturbed by this, while I was so disturbed to learn this fact that I never intentionally ate fish again (for more than a quarter-century so far). I learned this at almost exactly the same age as Iris did in the story linked to above and in almost precisely the same way.
"In Soviet Russia, all your base can has you!"
I've been out of touch with a lot of people since the new year.
Here's what I've been up to:
I had a great time hunting with Codex Bodley in the
2008 MIT Mystery Hunt.
This year I felt that I made a larger contribution to the solving
process than either of the two previous years that I've hunted with
Codex. Here are the puzzles that I particularly remember working on:
Hearing
Voices (MacOS synthesized voices sound files with movie quotes) --
I knew that having been a computer lab monitor in high school would
come in handy some day! We solved this puzzle in a fairly
straightforward fashion, and I did a lot of the work on it.
Character
Witness (cartoon characters) -- this puzzle had a PDF containing
two pages with what was, it turned out, meant to be the same image on
each page. Unfortunately, the puzzle creators managed to make the
resolution of the bitmaps different, so I spent a long time adding,
subtracting, and otherwise comparing the two versions, to no purpose.
Ecolocation
(Latin, Name of the Rose) -- I enjoyed the use of Latin
in this puzzle and fairly quickly figured out all the Latin questions
and answers (some of which were pretty funny). Unfortunately, we only
got about halfway through because we were unable to figure out what
to do with the answers. (They were supposed to give us another Latin
text cryptically instructing us to find the name Ehrich Weiss on the
Harvard Bridge, insert "der", change the color to red, and translate
into English, which would have yielded "Eric the Red". On the
other hand, we did backsolve this puzzle.)
Talk to Me
(foreign language) -- despite having people who together knew a dozen
or two natural languages listen to this over and over again, we
never made any progress on it at all, and even ended up concluding
that it was probably not based on any natural language. It seems that
the Navajo code talker code remains quite effective to this day!
Tragedy
(cellular automata) -- this was a multi-layered puzzle and I noticed
a few of the layers (but there kept on being more and more layers
within layers)...
Instant
Replay (pinball) -- I helped identify a few of the pinball
machines (at one point even downloading and reading a PDF of a
Spanish-language pinball machine maintenance manual to verify a guess),
and we managed to solve this one.
The Dungeon
(video game) -- we all realized that there was something funny about
this video game (apart from the fearsome "vicious algebraic topologist"),
but nobody ever managed to figure out that it was actually Tic-Tac-Toe (!)
in the guise of a text adventure!
Son of
the Realm of Unspeakable Chaos (constructed language) -- this
was my favorite puzzle of the Mystery Hunt. (Like Ecolocation, it
was composed by Kevin Wald.) This puzzle involved deciphering a
constructed language called Chaotic, of which only tiny samples
existed from previous years' hunts. This time around there was an
entire dialogue purportedly representing the end of a mystery novel;
after an hour or two of deciphering, we realized that it was actually
based on the game of Clue! Deciphering the entire story took me and
two teammates about six hours, and was incredibly satisfying as more
and more details and jokes steadily emerged. To commemorate this experience,
I've had
red shirts
made bearing the Chaotic text of the punchline of the story (an
explanation -- in Chaotic -- of the notion of
redshirts
in the Star Trek universe, which turns out to be essential to understanding
a clue given by the murder victim in the form of a metaphor). Despite
fully deciphering the story, we were focused on the internal structure
and assumed that the answer to the puzzle had something to do with the
solution to the mystery (for example, that the answer to the puzzle
should be the name of the murderer in the story). It turned out that,
in typical Mystery Hunt fashion, the story was merely a set-up to teach
us the Chaotic language and mention, in passing, figures that could be
loosely interpreted (with the help of a detail in the story) as maritime
signal flags which spelled out words in Chaotic that could be translated into
an English phrase. Although we didn't solve the puzzle (and neither,
apparently, did any other team!), I'm very grateful to Kevin Wald for the
wild ride.
(The idea of a mystery story where the solution to the mystery does not directly yield the answer to the puzzle also turned up in The
Deadly Hobby of Murder, which I had worked on earlier; that one had three layers of meaning which had to be combined to obtain the answer, but it could be solved with only a knowledge of English, heraldry terms, and computer fonts... so, naturally, we got it in short order!)
(P.S. Confidential to Kevin Wald: the modern and ancient Greek terms for "galaxy" are γαλαξίες and γαλαξίας, respectively, not γαλαξία; the latter is the name of "a kind of milk-frumenty". Trido di py Liddell-paidagogos py Scott-paidagogos py Jones-paidagogos fi.)
I've been working on Lest We Remember: Cold Boot Attacks on Encryption Keys (and now it's finally public!). I'm very proud of this project.
It's funny when different people do different parts of the translation (or translation revisions) for the same software, resulting in an actual inconsistency. I just got this amusing dialog in OpenOffice:
We could translate this as "The recovery of the documents was completed. Click 'Finish' [Finalizar] to view the documents" -- but there is no 'Finish' button, just (let's say) a 'Done' [Concluir] button. Very likely there once was a 'Finalizar' button at one point in the past, but some translator probably decided that 'Concluir' was a little more idiomatic... without realizing that there was other text referring ot the 'Finalizar' button.
Of course, this kind of inconsistency can happen any time people are changing software interfaces or documentation for any reason; it's not specific to translators updating translations.
Wikipedia says that "IM security providers created new products to be installed in corporate networks for the purpose of archiving, content-scanning, and security-scanning IM traffic". I found this jarring because I normally think of "IM security products" as those products like OTR that prevent archiving and monitoring of IM communications, rather than those products that perform such monitoring.
Lorenz had found that close is no cigar:
What now is near, in future may be far.
Then all may be omnipotent, yet none can see.
Timor mortis conturbat me.
This Lojban Reference Grammar example sounds pretty romantic:
mi ce do girzu
.i lu'o ri gunma
.i vu'i ri porsi
In English:
The set of you and me is a set.
The mass of you and me is a mass.
The sequence of you and me is a sequence.
(Lojban Reference Grammar, example 10.9)
Awwwww...
I saw a used clothing store called "Chic Saal".
I kind of want to visit Kiel so that I can board a train with a sign reading "Ziel: Kiel".
"I think we're not going to be in Rome after all."
"What's the matter? Forum non conveniens?"
(I'm posting this here so that search engines will be able to find the serial number.)
My bike, a Marin Belvedere hybrid with serial number C17609D008, was stolen from the South of Market Area in San Francisco during the last week of July, 2008. At the time it was stolen it had a dark metallic color and lots of stickers, but it may have been repainted; the silver Marin logo badge on the front was bashed up on one side. Since only the rear wheel was stolen with the bike, and the rear wheel was not original, the bike might be resold with mismatched wheels.
If you see this bike, please contact the San Francisco police department at 415-553-1392 and cite case #080824381.
So there's been a lot of controversy recently about how Firefox handles self-signed certificates. In general, it makes it rather difficult and scary to visit an SSL website with a self-signed certificate rather than a certificate-authority-signed certificate.
There seem to be much better ways of dealing with self-signed certificates, like the clever new Perspectives system, which would address the overwhelmingly most plausible man-in-the-middle attack models in the Western world.
Having just seen a presentation at eBay on phishing, I'm curious whether Firefox's default behavior is aimed at preventing man-in-the-middle attacks or whether the Firefox developers actually want to discourage the use of self-signed certificates in general, apart from the extent to which they expose users to man-in-the-middle risks. Does it bother the Firefox developers more that you might be getting spied upon (or getting data injected) by the operator of your local access network, or that Firefox might correctly assert that the self-signed SSL certificate is valid when you visit a botnet-hosted fraudulent phishing site over SSL?
There's actually a tension in practice between the use of SSL to identify sites (especially to make them appear legitimate or trustworthy) and the use of the SSL to protect against eavesdropping or spoofing by network operators. If you want to protect against eavesdropping, you would ideally find some way to make SSL use cheap and ubiquitous so that every site has a convenient path to enabling SSL or even making it mandatory. If you want to use SSL to indicate that a site is legitimate or trustworthy, you might want to force people to go through the path of conducting a commercial transaction with a CA before they can get the lock icon, etc., to display for users visiting their pages. You might even prefer that the lock icon show up relatively rarely so that users are reluctant to give personal information when it doesn't appear -- not because of the traditional eavesdropping concerns, but because of the more novel phishing concerns.
Many people may have noticed that I'm rather fond of self-referential t-shirts that are the sole result of something happening. So I couldn't resist putting together another one recently.
I wrote the text for this one; it says "amicus meus Romam petivit nec eo itinere donum ullum recepi nisi hanc camisiam Latino sermone inscriptam" -- roughly "my friend went to Rome and all I got was this Latin t-shirt", but a little more elegantly rendered. It was handsomely designed by Mako on my way home from the trip in question (using his excellent Inkscape skills), and you can now buy it in
men's or
women's sizes.
But in order to make the substance of the shirt come true on a wider scale, I'll also offer one as a present to the first four friends to request one. Just let me know your size and color. Operators are standing by!
(Note that because of grammatical gender problems, this shirt is only correct if the person as a result of whose trip one obtains it is male. I'm sure Mako would be happy to prepare an "amica mea Romam petivit" version on request!)
How come Peter's grep matches the string 'foo^bar' only with the pattern 'foo\^bar', but my grep matches the string 'foo^bar' with the pattern 'foo^bar'? (I think both are GNU grep on up-to-date systems, though I didn't compare the GNU grep versions.)
[Main]
Contact: Seth David Schoen
