Slashdot Mirror

What version of the JDK are you using? Java 6 has improved speed considerably.

I'm on 1.6 and have been since beta releases of it, as it added some features I desperately needed. It improves the situation for some floating point algorithms, as I believe it uses processor intrinsic implementations of sin and cos, but not exp.

Here is are interesting benchmarks comparing floating point performance:

1) http://www.fourmilab.ch/fourmilog/archives/2005-08/000567.html

Java is 12% slower than C and Visual Basic.NET is 13% faster. This is based on the usage of an optical design ray tracing programme.

Again, this is using only raw floating point and trig. Sun's Java VM supports processor-based trig functions, but not other transcendental functions.

2) http://kano.net/javabench/

None of these are floating point benchmarks. I agree entirely that there is little reason to avoid using Java for integer number crunching.

Note: If you are raising numbers to an integer power you will almost certainly get better performance doing y = x*x*x than y = pow(x, 3.0) in any language

Agreed. My problem is very specifically calculating 1 / (1 + e^{-x}), which is somewhat slower in Java than in (e.g.) C, and can't be helped with any such transformation.

Here is are interesting benchmarks comparing floating point performance:

1) http://www.fourmilab.ch/fourmilog/archives/2005-08/000567.html

Java is 12% slower than C and Visual Basic.NET is 13% faster. This is based on the usage of an optical design ray tracing programme.

2) http://kano.net/javabench/

Note: If you are raising numbers to an integer power you will almost certainly get better performance doing y = x*x*x than y = pow(x, 3.0) in any language.

Why would LISP be interpreted? (that is your implication, is it not?).

http://www.ffconsultancy.com/languages/ray_tracer/index.html

Scheme (unadorned) is 30% slower that g++.

http://dan.corlan.net/bench.html

Choice quotes from the last page:

" Programming a 1 GHz Pentium [fastest, 1000$, 2002 PC] in Perl is like programming a 10 MHz something [overclocked, 50$, 1982 Z80-based ZX Spectrum] in FORTRAN!

This is no big surprise as we are looking at the classical trade-off between the power of a language (allowing the programmer to express something in a compact way) and the performace of an implementation (which classically was related to the language being close to the machine representation).

However, the huge exception is CommonLisp. Lisp is the most powerful language that is, representing the classical extreme choice for expressive power instead of efficient implementation. ...
For new projects in which one is free to choose the language, one should choose Lisp."

Here's one where Visual Basic .NET beats C:

http://www.fourmilab.ch/fourmilog/archives/2005-09/000588.html

Java beating C:

http://openmap.bbn.com/~kanderso/performance/java/index.html

Of course C is an extremely primitive language. It is possible to (on current hardware) write a C program to best most other systems. But, a proper program is much more difficult!

That would be great if the leading browser worked right, but it returns "Buy Now!" in IE (up to and including 7. Looks like it's finally fixed in 8).

Sounds like the do something like the free ent utility. It calculates a "randomness" of files. It can be quite useful to tell "data" from "encryption."

With current technology, speeding up and slowing down would each take at least months. If you managed to get up to 0.5 c, the trip (for the passengers) would take more than 34 years. If you got up to 0.9 c, the trip would take more than 9 years.

Very close to c, the trip would be nearly instantaneous (for the passengers).

(I assumed that gravity equivalent acceleration would be quite a feat, and used the table here:

http://www.fourmilab.ch/cship/timedial.html

)

One cubic foot of water is around 60 lbs.

I thought that number sounded a bit high as a gallon only weighs about 7 pounds, but sure enough, a cubic foot of water DOES weigh around 60 lbs. 62.42796 pounds to be exact. And a gallon is actually just over 8 pounds.

"If you are required..."

I don't think anyone in TFA is seriously suggesting that hand holding noobs be a requirement for publication and this is probably where the confusion sets in. I also understand that you may want to keep your own data close to your chest until you have extracted a paper out of it (ie: publish or perish).

"To be honest, if your institution does not foot the bill for subscription, try inter-library loans...[snip]...The problem with scientific publication is that you need to be terse. They're limited to 8-12 pages."

Einstein managed to get away with three elegant pages and zero refrences, chasing down the english translation in that link took a couple of minutes. I'm interested in quality not quantity, I would be delighted with the 8-12 pages at my finger-tips because like most educated laymen I do not have "too much time on my hands". The internet and afformentioned lack of time is the reason I have not set foot in a library for almost a decade and the last time I studied/taught at a tertiary institution was quite possibly before you were born...

"If you want to know the research, do your homework and study the subject carefully for a few years. Then you'll appreciate whatever data or paper the scientists are publishing.

Precisely why I chose to use the folk at realclimate as an example, following the science for 25+yrs does not make me a climatologist but it has given me a deep understanding of what they are banging on about.

And once in a while 'The Supreme Being' hits us with a real doozy! http://www.fourmilab.ch/documents/OhMyGodParticle/

Already done:

http://www.fourmilab.ch/hackdiet/www/hackdiet.html
http://www.physicsdiet.com/

the Mayan calendar merely resets at that date. similar to how computers were expected to reset at y2k, it was not that they expected the world to end they just did not include dates after that much like our calendar does not include specifically year numbers for after 9999

Actually it's even less catastrophic than that. The Mayan long count calendar is based on a hierarchical system of cycles, called kin (1 day), winal (20 days), tun (18 winal), katun (20 tun) and baktun (20 katun). Dates are indicated by giving the position in the relative cycle, so today, November 27, 2008, would usually be quoted as 12.19.15.15.15 in the Long Count calendar. You can check out the conversion formula e.g. in the source code for Fourmilab's calendar converter.

The five-position notation for the long count has a cycle length of 2,880,000 days, or approximately 7885 years, ranging from 11 August 3113 BC (0.0.0.0.0) to October 12, 4772 (19.19.19.17.19). If that isn't enough, there are higher-order cycles as well - a pictun of 20 baktun with a cycle length of some 150000 years, a kalabtun of 20 pictun with a cycle length of 3.15 million years and so on. These are conventionally omitted in notation, because dates from these cycles are rarely met in Mayan astrology (or elsewhere for that matter), but there is a mechanism for expressing them.

Thursday, December 20, 2012 is 12.19.19.17.19, and all that happens on December 21 is that all cycles reset and the baktun gets incremented by one, to 13.0.0.0.0. Some New Age freaks interpreted this as the end of the world because of a rather arcane interpretation of the significance of the 13th baktun cycle in a previous world, but even in 4772 the calendar won't "overflow", it will just shift to the next higher-order cycle.

Why is everyone so hung up on an energy-positive reaction? Antimatter is the ultimate in fuel for space-propulsion as it produces the highest theoretical amount of energy for the least possible mass. (i.e. 100% conversion - losses to nuetrinos that cannot be captured) This plays well into the rocket formula, giving antimatter drives a specific impulse unattainable with other rocket methodologies. In fact, the far-flung future may see c-ships traveling the stars based on matter-antimatter drives.

What I want verified is not if this process is energy efficient or not. I want to know if this process is several orders of magnitude more efficient than the current Fermilab and CERN processes.

Sure, you could use the antimatter to fuel some other reaction as you suggested, but then again we already have fission reactors which produce net positive energy.

Once again, antimatter catalyzation makes the fuel more efficient for its weight and thus plays well into the rocket formula.

If your random number generator is truly random, a single pass will scatter your dataset. The problem Monte Carlo simulations can run into is not the number of passes through the randomizer but relying on a crappy non-random generator.

If the tech community makes enough buzz about this, it's likely that we can put the pin back in this grenade.

This ironically reminds me of the title of a nice (and rather old) text, which sadly sounds almost like prophecy now.

That's the energy of a single atom. The overall energy of the entire beam is actually quite surprising. I've heard the beam has the same energy as a carrier at 5 knots. Obviously, this is not a dangerous amount of energy. The ohmygod particle had a much higher single-particle energy, for example.

Has anyone here read about the "Oh-My-God particle"? A proton detected in 1991 with an energy of 3.2±0.9×10^20 eV - that's 51 Joules, an energy you'd expect for a macroscopic object and 10 million times more than the maximum the LHC can produce (7 Tev).

The linked page has some of the relativistic properties calculated for that proton including that "After traveling one light year, the particle would be only 0.15 femtoseconds -- 46 nanometres -- behind a photon that left at the same time."

You can use BetterMail for a secure connection to Gmail, but Google still has all your messages and they're unencrypted when they go out from there. In this case store and forward is not your friend.

You could use a simple encryption tool like this one. It's a little less difficult than a system that requires a key exchange but it's also less secure. And there's still a decryption process. Copy, paste, type pass phrase, read.

If there's something that's easy to implement and lets you exchange encrypted messages with other email clients that don't support your encryption scheme, then I don't know about it. Far as I know you have to make a decision to encrypt or not every time you send a message. When you're sending to a compatible client you can at least encrypt the body of the message, but as far as I'm aware, that's the state of the art.

The Hacker's Diet

Provides statistics, weight tracking, and a simple, yet rigorous workout plan. It works when one sticks to it, heh. I need to get back to sticking to it.

I personally follows this http://www.fourmilab.ch/hackdiet/ There's a useful training program in it

As I was only kidding about the above, try the Hacker's Diet: http://www.fourmilab.ch/hackdiet/ It was written by the creator of AutoDesk, He used to be overweight and created this diet for himself. Coupled with just a basic exercise regimen (take a daily walk and keep doing your situps and push ups) and you should be fine!

Biking to work is good. I did that for a while before my job moved too far away. (No, I will not sell my home in a buyer's market to move closer to work)

I can still bike partway, and take a bus the remainder. Our local transit authority has bike racks on all of the routes.

During the summer, I swim at least three times a day, once before breakfast, once when I get home from work, and once just before bed.

During the rest of the year, I do push-ups, sit-ups, stretches and jumping jacks, following the programme in The Hacker's Diet , with a minor modification: I skip the lifts, because I don't need them.

At work, I get up frequently, and choose to pay visits to people rather than phoning or emailing them. I use the stairs, and not the elevators.

Weather permitting, I will go for a walk at lunch time.

In other words, I do a little bit at every opportunity.

Just like the post said, awesome training part in the hackers diet:

http://www.fourmilab.ch/hackdiet/www/chapter1_2_5.html

Nice toy!

I am using textogif to produce images with formulas when I happen to need them.

"The Hacker's Diet" is another view, with commentary on how to manage what you eat as well:

http://www.fourmilab.ch/hackdiet/

John Walker's "The Hacker's Diet" includes a really good 15 minute progressive daily exercise routine that can be done anywhere, as well as an excellent technique for losing weight and keeping it off (clue - eat less!).

John Walker, one of the founders of Autodesk, has a pretty easy looking plan in his book, The Hacker's Diet . I haven't tried it myself, but I would if I didn't have wooden floors. :-)

Stronglifts 5x5 - http://stronglifts.com/stronglifts-5x5-beginner-strength-training-program/
The Hacker's Diet - http://www.fourmilab.ch/hackdiet/www/hackdiet.html

I used these two as a guide and have gone from 6'2" 165lbs squatting 100lbs to squatting 210lbs in less than three months. Feels awesome.

Find something you enjoy doing and enjoy it, and you'll do it. As long as you think of exercise as a chore you'll never be motivated to do it. It shouldn't be like flossing your teeth or taking out the trash.

Myself, I bicycle. I rode 4,000 miles in 2007 and loved every minute of it. I ride with the local club and go on group rides. I'm 42 and this year I am the fastest I have ever been and am finally able to ride fast enough to keep up (barely) with the "fast" speed groups.

If there's no form of sport or exercise you can find that you like, then the least you can do is approach weight loss in geeky fashion and check out http://www.fourmilab.ch/hackdiet/www/hackdiet.html "The Hacker's Diet."

Garbage in, Garbage out.

All that said, it really is best to find some sort of good aerobic exercise you enjoy, an hour a day, every other day at least. It helps manage weight, kills depression, clears your mind, increases self confidence. Yes, all that. ...and I -am- a geek.

and if you subtract the time spent exercising or recovering, it's not like you'll live noticeably longer anyhow.

As I understand it, you're incorrect. An hour spent exercising will, on average, make you live a few hours longer. Citation, though I don't know how accurate his claim is. It will also make you more healthy, so you'll enjoy the hours that you are alive more.

(You never really appreciate your health properly until you don't have it ...)

I'm not sure how geeks do it, but hackers do it by following the hacker diet of course:

Hacker's Diet

Complete with exercise routines!

Is exactly what you need. It details an exercise regimen that can be performed anywhere you are, without anyone else knowing if you don't want them to, can be performed in 15 minutes or less, and ramps up as you get more fit.

You can find an online copy of The Hacker's Diet here.

http://www.fourmilab.ch/hackdiet/

Has online and Palm tools to track you weight and completion of their exercise ladder, based on exercises that can be done in a small space with no equipment.

How come nobody has mentioned this one yet? I mean, come on! It has fitness tips as well as diet since you really need both to get the pounds off.

Hacker Diet

John Walker has written about this: The Hacker's Diet.

Here's the letter from NOAA to the Lunar X participants that outlines how this is pursuent to the Land Remote Sensing Policy Act of 1992 - says it may take up to 120 days to obtain the license - think about that before you take your first picure! ;-)

In the meantime, you can use existing satellite photos to image your house and here's a cool way to get a nifty Earth view.

Way lower, here, can be as much as a factor of ten million.

Here's a nerdy but popular account of an extreme high energy cosmic ray detected at the Fly's Eye II. And that's just what we've detected in a few decades of running small detectors. What the planet has intercepted in the last few billion years must be even more staggering.

The legal system is a lottery anywhere in the world.

Incidentally, John Walker pretty much foresaw this whole business in his 2003 document the digital imprimatur.

It makes interesting reading to say the least and if his future view of the headwind for 'p2p' is correct this is really only just beginning.

In most if not all cases, you will find applications that are as good as if not better than the PalmOS version.

Ah, but can I get the source? It's not essential, but it's a definite important feature for me.

TomTom - Exists for PalmOS and WM. In my opinion the WM version seems to work MUCH better and is far easier to set up.

I haven't tried this, but I was not aware that the GPS was usable by apps on Treo 650.

TCPMP media player - Exists for both platforms, has somewhat extended codec support compared to PalmOS on a Windows Mobile device

Yeah, I prefer TCPMP over RealPlayer, as TCPMP will play OGGs (which is what I rip to by default).

Web browser - The PalmOS web browser Just Plain Sucks in every way possible. Even Pocket IE is better and it isn't that hot (there are other options for WM)

Yes, which is why I've switched to Opera. I'm not happy that I can't get source to Opera, but I can't get source to any of the others and Opera seems to work better.

Java - The Java environment for PalmOS is utter and total crap and when I tried it on my Treo 650 not a single app I tried would work.

I managed to get ahold of the IMB java kit for PalmOS so I could run Opera. Seems to work pretty well, but then I don't use it for anything besides Opera; I'm not a big fan of Java myself.

GMail's Java app works great on my Tilt.

I run my own mail/web server, which seems to work fine with Opera and the mail client that comes with PalmOS. My only gripe there is that the PalmOS mail client doesn't support aliases.

Google Maps - Don't think there's a PalmOS client that can come anywhere close to Google's Windows Mobile native app

Don't know; haven't use the windows mobile version, but the PalmOS version of Google Maps seems to work fine for me.

Instant Messaging - Half of the PalmOS IM app vendors seem to have gone out of business or stopped supporting the app. I've found FAR more choices for WM.

IM isn't a priority for me (I hardly use it). I'm pretty sure there are open source IM clients for PalmOS however.

You don't exactly provide details of what apps you're using.

Here's a list:

• Little John - Console emulator for playing classic games.
• pFuel - Keeps track of gas mileage.
• FreeCoins - Accounting software.
• Keyring - Keeps track of passwords.
• pssh - SSH client.
• Plucker - EBook reader.
• LispMe - Scheme interpreter.
• Eat Watch - Weight tracking.

The thing is, PalmOS works very well for me, and is entirely compatible with Linux (or any other platform I choose to move to). If I were going to move to a new mobile platform, the last place I would look is to Microsoft (or Apple for that matter). Not only would I need to have the apps I use above (which I have source to, so I could port given a decent development environment that runs on my desktop of choice), but I would prefer to move to something more open, not less. Something more hackable, not something that gets in my way when I try to do something the big corps don't like.

I guess these IT guys have never heard of The Hacker's Diet. The body is a big machine, and how much is weighs is just an engineering problem.

A lot of people want to insist that it's full of vagaries and hocus-pocus, but it's really not. It's all about calorie intake vs. calories used.

I eat fast food all the time, and I've LOST 10 pounds in the last month and a half. I just count calories. A large diet coke and a double cheeseburger (no fries!) from McDonald's is only 440 calories, and reasonably filling. McChicken is 360, but a little less because I get it without mayo.. Chicken Nuggets are 420 for a ten piece.

It's probably not very nutritious to eat those things, but it's certainly not going to make you gain weight as long as you control your potions. I take a multi-vitamin daily, and do eat lots of fruits. Mostly because they're low calorie, yummy, and convenient.

I suggest the GP check out The Hacker's Diet. Your body is a machine, and losing weight is just an engineering problem.

For losing weight, the Hacker's Diet (google it) and exercise is working for me

Sir, I salute you. I found it here, I've never heard of this method but I find it fascinating.

In the last ten years I've put on about 50 pounds. One of the reasons why, I believe, is that when you're so engrossed in your computer work, you just don't take the time to eat right and/or exercise. And why should you? The computer stuff is so awesome, you never think about all that other stuff.

To make matters worse, most "health nuts" (and/or diet freaks) are so irritatingly moronic, you just get turned off on the whole thing entirely. Refreshing and inspiring to see a book/site written by a true engineer.

Thanks again.

Did you factor your drinking into your food intake? I was packing on the pounds a few months ago before I did a total accounting of everything I ate and drank and started tracking my weight daily. A few minor tweaks to my diet (the first couple weeks were rough I'll admit) and 4 months later I've lost over 30 lbs. I was only eating about 1800 calories a day but my drinking habits averaged over a week easily put me over 2300 calories. A couple cocktails or non-light beers add up hella fast, and even most "light" beers are packed with calories. This is what got me started and I'd say it's successful (and fairly illuminating) if you can stick to it.

Has someone actually checked the random output for cryptographic soundness?

The proper source of cryptographic-quality random bits in Linux is "/dev/random". Reading uninitialized memory is not a good source of entropy - it might be initialized to some constant value, especially if you're compiling with a debug allocator or running under a virtual machine monitor.

OpenSSL makes substantial efforts to get a good random starter. Unless someone did something so stupid that OpenSSL didn't use /dev/random, it should still work. OpenSSL is supposed to have a check for bad random seeds. Was that bypassed, or doesn't it work, or what?

Obtaining a good source of randomness is hard. Computers are rather deterministic. Historically, there have been major failures in this area. See "Venona" where the USSR was generating "one-time pads" by having people type random digits on typewriters. Arlington Hall, NSA's predecessor, cracked that. Humans aren't random enough. True random number generation requires special hardware, like a noise diode or a radiation source.

"Any one who considers arithmetical methods of producing random digits is, of course, in a state of sin." -- John von Neumann

Yes, he is still alive. John Walker's homepage is www.fourmilab.ch.

'What's the oldest piece of code you can get running? Either on emulation or on original hardware. Be creative, winner gets... well, kudos. But that's gotta count for something on Slashdot right? :^P'

How about this?:

http://www.fourmilab.ch/babbage

http://www.fourmilab.ch/babbage/applet.html

'What's the oldest piece of code you can get running? Either on emulation or on original hardware. Be creative, winner gets... well, kudos. But that's gotta count for something on Slashdot right? :^P'

How about this?:

http://www.fourmilab.ch/babbage

http://www.fourmilab.ch/babbage/applet.html

FANG, from 1972, is probably one of the oldest applications you can still download and run. It's a copying utility for UNIVAC mainframes. UNIVAC Exec 8 was way ahead of its time, with full support for threads, multiprocessors, and concurrent I/O from the late 1960s. FANG was one of the first applications to use that concurrency effectively. You could put in a series of commands to operate on multiple files, and it would do them as concurrently as possible, keeping track of any dependencies in the file copies.

Or fractal vegetables

Take a few steps back and consider your perspective. Try reading about engineers vs. managers: http://www.fourmilab.ch/hackdiet/e4/eatwatch.html (scroll halfway down)

Many computer guys tend to be alarmist and see the world in black and white. Many security firms rate problems only based on potential damage without consideration for existing mitigations elsewhere in the system or the reality of targeting from attackers. Consider your company's situation carefully.

If, after much deliberation, you are certain legitimate problems exist that must be fixed (versus managed) then talk to the managers in their language: build a business case. You work for a company, the company's job is to make money. Security costs money. You must clearly articulate how the security improvements will make money or stop the company from losing money. It's all engineering, in the end. It's just engineering with words and numbers.

Cheers.
- jj

Here are some links :

http://www.fourmilab.ch/babbage/

The obligatory 99-bottles-of-beer-on-the-wall in punched card Analytical Engine assembly language :

http://99-bottles-of-beer.net/language-babbage's-analytical-machine-79.html

Hmmm, I dare say that's shorter than the C# version, if you remove the comments. Oh and it will run Linux, if you have enough coal and are willing to wait a few years for X to load. ;) (it does have a graphical output device) As for a beowulf cluster, that might help performance, although your interconnect mechanism would probably be pneumatic ie. tubes (that's what the Internet is made of anyway right?) and the cluster size would require a few tens of millions of units. ;)

jdb2