Showing posts with label computers. Show all posts
Showing posts with label computers. Show all posts
Monday, May 17, 2010
Intelligence Games and Computer Worms
Fascinating and eerie description of the evolution of a sophisticated computer worm, and the hunt for its mysterious controller. Whoever this is has got millions of computers under his control...whenever he decides to activate the worm.
Wednesday, April 15, 2009
An amazing mechanical pocket calculator
The Curta pocket calculator, designed from within a German concentration camp, of all things. Absolutely amazing device, with about all the details you'd want, here. And yet, more here. And here's an online simulator.
Monday, November 10, 2008
In which tech writers forget there's a Rest Of The World
Remarkably narrow-minded article on "Five Useless Gadgets You Should Throw in the Trash" by a writer for Wired. (Of course, Wired) He thinks that printers, scanners, faxes, DVD or CD drives, and regular telephones (landlines, not cell phones) are "useless." He has a few caveats, but not enough to make up for the poor thinking of the rest of the article. This is the kind of thinking I'd liken to socialist central planning--"Why do you need your car? Take a bus everywhere! Why do you need your own house? Just rent an apartment--and there's no grass to mow!" He actually thinks that a printing "service" can replace your printer. I wonder how long their printouts take to get to you. But then, I'm sure you've never printed anything out that you couldn't simply wait a couple of days for, right?
I'm not going to call the author "stupid," as many have (or maybe they just meant the article--there's a decent case for that), but he's amazingly short-sighted.
The comment by "Sean" is priceless.
I'm not going to call the author "stupid," as many have (or maybe they just meant the article--there's a decent case for that), but he's amazingly short-sighted.
The comment by "Sean" is priceless.
Monday, August 11, 2008
Servers in [European] Georgia not responding
Or at least not the one I tried. Not surprising, I reckon, but that tells you how widespread the damage is. I wonder if this is physical damage to the hosting site, heavy traffic, or if it's Russian hackers doing a bit of cyber warfare.
Sunday, May 11, 2008
The best data recovery I've ever heard of
I see here that a computer hard drive from the space shuttle Columbia has been restored, and 99% of the experimental data on it has been recovered. Wow. This, despite the fact that the outside of the drive was a melted lump of metal and plastic, the seal had broken open, and part of the disk had been pitted with dust-sized debris. But because the computer was running DOS, which stores data in one place at a time, rather than scattering it across the disk, it chanced that the damage was not to the place the experimental data was stored, and 99% of the experiment's results were recovered.
The experiment probed the effect called "shear thinning," which is how substances like canned whipped cream come out like a liquid but then stiffen. They measured shear thinning in xenon near its critical point. NASA's write-up is here (with photos!), and this is a link to the article itself, coming out in Phys. Rev. E.
I think I remember some acquaintances or friends of friends at NASA who had another experiment on Columbia, but I don't know if any of their data were recovered.
The experiment probed the effect called "shear thinning," which is how substances like canned whipped cream come out like a liquid but then stiffen. They measured shear thinning in xenon near its critical point. NASA's write-up is here (with photos!), and this is a link to the article itself, coming out in Phys. Rev. E.
I think I remember some acquaintances or friends of friends at NASA who had another experiment on Columbia, but I don't know if any of their data were recovered.
Sunday, April 27, 2008
At last, at long last!
(Also via Instapundit) Holographic computer memory finally comes to market. Next month, in fact. Man, I've been reading about this for well on 20 years, and I can't believe they're finally selling them.
The drive is $18,000, at least for now, but each removable disk holds 300 GB. And it's photographic, which means it's got a much longer life than most digital media (they're saying 50 years for this). I wonder if this will be a solution for the huge memory requirements of the LSST observatory, which will be recording half a petabyte (1 petabyte=1000 terabytes) of data per month. For 10 years. Imagine adding 500 1-TB hard drives every month to some warehouse-sized room. That's the amount of computer storage they're going to need. Now, the LSST doesn't come on line until 2012, and we can expect disk space to get denser and denser by then, but it's an amazing amount of material they're going to be storing. I'll bet they're looking at these holographic drives as a possibility.
The drive is $18,000, at least for now, but each removable disk holds 300 GB. And it's photographic, which means it's got a much longer life than most digital media (they're saying 50 years for this). I wonder if this will be a solution for the huge memory requirements of the LSST observatory, which will be recording half a petabyte (1 petabyte=1000 terabytes) of data per month. For 10 years. Imagine adding 500 1-TB hard drives every month to some warehouse-sized room. That's the amount of computer storage they're going to need. Now, the LSST doesn't come on line until 2012, and we can expect disk space to get denser and denser by then, but it's an amazing amount of material they're going to be storing. I'll bet they're looking at these holographic drives as a possibility.
Sunday, February 24, 2008
Neat, fascinating weirdness in computer science
I mostly use computers for straightforward computational work (and for playing around on the internet), and I don't do a lot of fancy programming. For my astronomy work, I use IDL, IRAF, and SuperMongo (SM). IDL has a kinship with FORTRAN that makes it easy to pick up and simple to crunch your way through a heavy computational problem, and it's adapted for image processing, which we do a lot of in astrophysics. IRAF is its evil-mad-scientist cousin, which is also for image processing, but screwy to deal with and has a lot of black boxes. And SM is a language for making publication-quality plots, as well as calculating things from large tables.
When it comes to writing a quick program to calcuate what I need, I prefer IDL and its FORTRAN-like simplicity and directness. But I have a fascination with the varieties of programming languages out there, and especially the totally different concepts behind some of them. It's like breaking out of the Indo-European languages most of us reading this are familiar with (say, English, German, Spanish, Latin, Russian, and even Sanskrit and ancient Hittite), with their conjugations of verbs and declensions of nouns...and seeing how the Semitic languages work (Hebrew verbs have gender, and expressing the genitive case is done by putting two nouns next to each other). Or Vietnamese, which has neither plurals nor past tense, and differences in meaning come in part through the tone of voice.
In that vein, take a look at some of the "Esoteric Programming Languages" listed here in Wikipedia. If your familiar with the simpler BASIC or FORTRAN, see how different the concepts can be and still get the job done. One of my favorites is perhaps the extreme case of a single-instruction language. If chosen carefully, you can do any computation with clever arrangements of that one instruction!
I've taken too long in posting this and want to go on to something else, so I'll get back to it later--more weirdness in programming to come, as well as how this connects with frontiers of physics!
When it comes to writing a quick program to calcuate what I need, I prefer IDL and its FORTRAN-like simplicity and directness. But I have a fascination with the varieties of programming languages out there, and especially the totally different concepts behind some of them. It's like breaking out of the Indo-European languages most of us reading this are familiar with (say, English, German, Spanish, Latin, Russian, and even Sanskrit and ancient Hittite), with their conjugations of verbs and declensions of nouns...and seeing how the Semitic languages work (Hebrew verbs have gender, and expressing the genitive case is done by putting two nouns next to each other). Or Vietnamese, which has neither plurals nor past tense, and differences in meaning come in part through the tone of voice.
In that vein, take a look at some of the "Esoteric Programming Languages" listed here in Wikipedia. If your familiar with the simpler BASIC or FORTRAN, see how different the concepts can be and still get the job done. One of my favorites is perhaps the extreme case of a single-instruction language. If chosen carefully, you can do any computation with clever arrangements of that one instruction!
I've taken too long in posting this and want to go on to something else, so I'll get back to it later--more weirdness in programming to come, as well as how this connects with frontiers of physics!
Friday, March 23, 2007
Death of FORTRAN creator
John Derbyshire reports the death of John Backus, who came up with FORTRAN, the first computer language. I'm linking to his Corner post, rather than to the original obituary directly (he links to it), just so you can read his comments, too.
For myself, I like FORTRAN. I've also programmed in C, but most of my C and FORTRAN work was in particle physics simulations, for which I used straightforward mathematical equations. FORTRAN had most of them already built in, whereas in C, I'd have to go and define them, which was a pain. Plus, for FORTRAN, we could directly integrate the CERNLIB particle physics subroutines from CERN. Handy.
And I'm glad to see that FORTRAN still has an active life among the supercomputer guys, like my wife, who's using parallel FORTRAN every day.
For myself, I like FORTRAN. I've also programmed in C, but most of my C and FORTRAN work was in particle physics simulations, for which I used straightforward mathematical equations. FORTRAN had most of them already built in, whereas in C, I'd have to go and define them, which was a pain. Plus, for FORTRAN, we could directly integrate the CERNLIB particle physics subroutines from CERN. Handy.
And I'm glad to see that FORTRAN still has an active life among the supercomputer guys, like my wife, who's using parallel FORTRAN every day.
Subscribe to:
Posts (Atom)