'Dogma' is common in the sciences, but it implies something different than the formal definition you are thinking of. It is usually used to describe a highly simplified model of how a system works. It's just a useful way to think about something.
The most well known example is the central dogma of molecular biology. By the time you finish freshman molecular biology in college, you know that it is a gross simplification of how a cell works, but that it is a very good first approximation.
Chemistry is no different. The vast majority of chemical interactions involve the valence electrons. So how do you introduce the topic? You say 'all chemistry deals with valence electrons (cough, cough)'. If the students learn that, you're actually doing pretty well.
Once you get past the basics, you admit to the students that you might have fibbed, and that under unusually high energy conditions, the inner shell electrons actually can interact. Upper level chem courses have been teaching this for years - there are no surprises here.
All the article says is that a research group is predicting a previously unknown inner shell electron interaction under high energy conditions. While it is news, it is not shocking, and while it violates the 'dogma' that only valence electrons interact, it changes nothing about how the dogma will be taught.
Progress in science is made at the edges. What happens to this at high energies? How will these atoms behave at extremely low temperatures? The easy cases have been understood for years, if not centuries. This discovery doesn't change any of that. So this is cool, but not a fundamental break-though.
Now, if someone replicates this experimentally, and then figures out how to use it to make dilithium crystals to power their prototype warp drive, that'd be revolutionary.
This is not, however, the end of the story. The PTO's decision in Ex parte Wasynczuk provides one final twist so Kafkaesque as to strain credulity. ...
In sum, an innovative process is not patentable when operating on a single computer processor but is when operating on two processors, even though the Board recognizes that the process in the unpatentable claim "is essentially the method" set forth in the patentable claim.
I really hate law sometimes. Using two computers (instead of just one) means that the method is tied to a 'particular machine' instead of a general purpose computer.
So while this may kill of some current software patents, only a slight change in hardware is required to keep on patenting software. Why even bother? All this does is ensure that every software patent from this day forward will be tied to a 'particular machine'.
I harbor a secret hope that this will still render software patents ineffective, because to get a patent, you'll have to say that your process can only be run on a 'particular machine' comprised of two computers, while your competitors will just run a 'general purpose computer' and claim that they are not violating your patent. That's still probably wishful thinking, given my usual misunderstanding of how laws actually works.
Re:Never heard of Django before, now it's everwher
on
Practical Django Projects
·
· Score: 4, Informative
The Google App Engine already has the Django libraries available.
Have a little care, there. The GAE caused quite a stir in the Django community, because it is only a partial implementation of the database interface that Django normally uses. It is backed by BigTable, which is blazing fast, but not a full blown relational database. If that works for you, go for it - it looks like a sweet platform for certain kinds of projects.
As for your question about Java->Python, I'm a former C++ convert myself, but I can help a little here. For some 'compile time' checking, look at PyLint It may check too much for you, but you can turn off the stuff you don't want.
As for unit testing, PyUnit is a pretty straight port of JUnit, so that should look familiar. However, I actually find nose to be a little better. It has many of the same capabilities, but with less boilerplate needed, and it integrates well with any existing PyUnit or DocTest tests.
SpaceX has a rocket with slightly higher payload than the Ares I scheduled to fly in 2010.
Didn't know that. Thanks for pointing it out.
I still have a major problem with anything SpaceX at this point, though. It isn't flight proven. Both Atlas and Delta are riding streaks of ~90 successful launches in a row, spanning several vehicles. The Atlas V and has had 14 launches, with one minor upper stage glitch (this wouldn't have killed anyone). Delta IV has launched 8 times, one partial failure in an experimental heavy launch configuration (wouldn't have killed anyone).
SpaceX has launched twice - two total failures, at least one of which would have killed the crew. There's a good quote from Elon Musk here.
"I think they had something like 12 Atlas failures before the 13th one was success. To get this far on our second launch being an all-new rocket -- new main engine, new first stage, new second stage engine, new second stage, new fairing, new launch pad system, with so many new things -- to have gotten this far is great."
SpaceX has plans to man-rate their line of rockets. But it is a long way off, and they need a few successful launches before you could actually strap a human on.
Any way you look at it though, NASA is flat wasting years of time and $5Billion in cool hard cash developing the Ares I.
Yeah, including some real alternatives, that can actually get into the orbits NASA needs to get to, rather then just barely out of the atmosphere (where you can tell a tourist that they are 'in space'. Like this or this.
The US Government has already funded the development of not one, but two rockets with the kinds of capabilities they need. They are flight proven, expandable to handle all sorts of loads, and available right now, not whenever Ares will slip out to. Add a little redundancy in a couple of systems, and have them ready to launch American astronauts into space in two years.
SpaceX is cool, and is probably the direction that the future of American space exploration needs to go. But it is not ready, it is not proven and it doesn't come close to the kinds of payload capacity or reliability that we need now. Check back around the time when Ares is supposed to be done to see what SpaceX is up to. In the mean time, quit screwing around developing a rocket similar, but slightly different from, the two perfectly good commercially available ones that are already up and running.
Actually, some of the very first Ares prototypes were based on the Atlas V, and the Russian designed RD-180 engines (which slightly outperform the RS-68s, and make anything shuttle derived look like a class-D model rocket engine). My wife worked on some of the prototype ground software systems.
At the time, Lockheed (now United Launch Alliance) put in a serious proposal to man-rate the Atlas V. NASA scoffed, which strikes me as a little strange. The shuttle technology is based on a 30 year old design, is planned to top out at about 20 total missions, and has blown up two orbiters. Both Atlas and Delta have newer flight-proven technology, better launch track records then the shuttle, lower prices, etc. Delta has the capabilities to add additional stages, Atlas intends to shortly, and both Atlas and Delta can add additional liquid or solid boosters as needed to handle huge payloads.
There is still noise about this idea now - click here for more information. But it won't happen.
These rockets are not in development - they exist now, and have a similar or better performance profile to anything that Ares or Jupiter have been proposed to have. Yea, you'd have to add redundancy to some of the systems, but you wouldn't have to redesign.
Meanwhile, who does have the capabilities to put people into space right now? The Russians, on well-engineered, cheap-to-produce ballistic missile-style vehicles. And while NASA flounders around redesigning 30 year old shuttle derived technology, and watching their launch date slip out for years, if not decades, we will be dependent on the venerable Soyuz keep Americans in space.
NASA is technologically bankrupt. All the great engineers they had 30 years ago are in private industry. But they still have some of the best bureaucracy and politics!
There is just one problem with this - getting timely, reliable data.
I work with flu, and the epidemiologists I know would love to have a system that could "facilitate early outbreak detection, increase public awareness of outbreaks prior to formal recognition, and provide an integrated and contextualised view of global health information."
A few sites that should help do this for flu are coming online, but the biggest impediment is getting timely, reliable, geo-tagged data. The local physicians know that an outbreak is starting, but it takes a while for samples to filter up to the state and country levels, be reliably analyzed, and then be uploaded to a tracking website.
While better websites will help in this process, the bottleneck occurs much earlier in the process. Samples come trickling in from third-world countries months after they were collected (or sometimes after the flu season is over!) and they might have the name of a city associated with them. That's not much help when you have to pick next season's vaccine halfway through this season so that it'll be ready in time.
And don't think that asking the local physicians to use a website is going to help - not for flu or any other disease. Most of the time, you are likely to get a hand-written sheet, partially filled out, with five or six columns of basic information for each sample. They don't have time to do any more then that.
Sure, the pandemic threats get a bit more attention, but the seasonal stuff still kills hundreds of thousands of people every year. You want to help? Start with those third-world clinics. For most diseases, the CDC and WHO still have to get people out on location to do good surveillance, and a website just isn't going to change that.
Easier answer - pick books that are an easy read and a fun introduction into the genre that you are trying to get your kids interested in.
My dad read The Hobbit to my sister and I when I was in ~3rd grade, then started into some David Eddings. By half way through the first book, we were moving too slow, so I stole the book and tore through the rest of the series. Once I started, there was no looking back - I went on to read everything I could get my hands on. Sci-fi, fantasy, mysteries, classics - whatever. All I needed was a starting point.
My wife, who grew up in a house where the TV was a family activity, never thought she liked reading, until I started reading some of those books to her in the evenings. It's become one of our favorite ways to spend time together. We've read Tolkien, Eddings, the Harry Potter series, and Narnia so far. We're doing A Wrinkle in Time right now, and I'm thinking we might try Dune next. In addition, she's now started to read quite a bit more on her own - some literary classics, some sci-fi, but all stuff she picked out herself. Most of it, I've never read. All she needed was a starting point.
To get someone started, pick something that has a good story and dive it. Show them that reading can be an extremely enjoyable activity. Provide a little direction toward a couple of books at the right reading level, then sit back, and smile and nod knowingly when they start telling you how great one of your favorite books is, or how they've discovered this great new series...
I have not read your articles. While I have a PhD in chemistry, it isn't in climatology. And I am here to tell you, in no uncertain terms, that everything you think you know about AGW is wrong. For *exactly* the same reasons that teaching ID as science is wrong.
Between 2003 and 2007, I worked in a room with a bunch of atmospheric chemists from the lab next door. They had read all of those articles, and hundreds more. They were all true believers - global warming is very real, and is caused by humans, period-end-of-discussion.
Except, in all of the seminars that I attended, they never said that outright. They would put up this graph showing a bunch of different sources of global warming and global cooling with error bars bigger than the data points. Which is a scientist's way of saying "this system is so complex that we can only guess at what is really going on - check back in a few decades if you want to know how this all turns out". They could not, in a room full of scientists, make a rational statement about AGW.
One day, toward the end of my PhD, our division seminar was about starting a university-wide effort on global warming. Many non-scientists attended. And at that *one* seminar, the true believers showed their colors. "AGW is real and it will kill us all!" they shouted.
The point I am making here is a very simple one. Blind faith is not science. Blind faith comes in many, many forms, and sometimes from the most unexpected of places.
The true believers in my story include some of the best atmospheric chemists America has - the CU Boulder/NIST/NOAA triangle forms the highest density of atmospheric researchers anywhere in the world, and includes many national academy professor types.
These people are caught in a strange, three-way hard place.
* On one hand, they are trained scientists - if you have not tried to disprove something like AGW every which way and failed every time, you can't claim it as truth. This is science, and it is a harsh mistress.
* There is the fact that it is almost impossible for these people to get funding unless they claim that their research will somehow lead to a better understanding of AGW. This is about money and the longevity of your career - will you sacrifice everything you have to be contrarian? Some of the greatest historical scientists did, becoming martyrs in the process.
* And then, their is the fact that they truly believe. This is the political, emotional, groupthink angle. In our current society, you are an outcast unless you believe the right things (like AGW). This is an irrational, emotional angle, and it runs deep in every one of us, hidden beneath a thin vernier of logic and rational thought.
And so, as a scientist in this position, you never say AGW in seminars with other scientists - but you do to non-scientists, the media, the funding agencies. You search for evidence to support AGW, because you believe, even as you search for evidence to disprove it, because you are trying to be a good scientist.
Most importantly, you judge papers you review only on their science - never their speculation. This means the materials/methods/experimental section - never the conclusions, where you are allowed to speculation about what this research *might* mean. As long as everyone does it that way, you can maintain the illusion that you are somehow separating what you *know* from what you *believe*, and thereby claim to remain objective.
Just because you read it in a peer-reviewed paper doesn't make it true. If the paper in question is regarding a politically/emotionally-charged topic, you can safely assume that the only facts in the paper are the hard numbers, with reported errors, that have been independently verified, and have stood up to repeated scrutiny over the period of several years. And you should probably take even those with a grain of salt.
Why am I ranting about AGW in a creationism vs evolution article? Because the problem that both fields have is the same one. True
But filing for frivolous patents is just the way you do business these days... It's just an indication that... the patent system is in serious need of reform.
The pathology runs deeper than you are implying. Mike Dillon , general counsel for Sun, goes so far as to compare patent portfolios to a cold war style nuclear arms race. Far from being 'frivolous', most companies today think that if they don't try to patent every little thing, they'll be sued into oblivion by the companies that do. It is this defensive mentality, not the patent trolls, that has lead to the current state of the patent system.
As someone whose first patent was published just this week, it is a little bittersweet. I'm happy that I was involved in a project that generated some truly innovative stuff, and that I've actually made a little money from the licensing of the rights to the patent. But the innovative parts of the project are not what got patented, the company that licensed the technology appears to be sitting on it as part of their defensive war chest, and I've had to battle their legal counsel to avoid their repeated attempts to get a patent on the software that I wrote for the project (which is useful, but not even remotely innovative).
I think as a patent holder, I see the needs for reform even more clearly now. Much like the cold war, I don't see the 'arms race' slowing down until company is holding so many patents that no one can make a move in any arena for fear of being destroyed by someone else. Mutually assured destruction is a technique for maintaining the status quo, and eventually, it will kill the ability to truly innovate in America. And when it does, another number of other nations will gladly step in and relieve us of our 'superpower' status.
The point is, our troop was nothing more than boys and their dads. We don't have some clergy like the church ruling our actions. In fact the scout leaders FORCED us to do EVERYTHING. We planned the trips, the meals, the transportation, the meetings, the lessons. They merely assisted and guided. What this means is that all the talk I hear now of homophobia and anti-atheist discrimination is a kind of surprise. It NEVER came up in my troop, I'd say a good majority of them weren't associated with any religion. Seconded. I am an eagle scout, and I never had an experience even remotely like what is being described in this thread. Our troop was active outdoors, produced a crap load of eagle scouts, and while we met in a church, religion was basically a non-issue.
If anything, it was one of the most *inclusive* environments that I have ever been associated with. We had all sorts of kids that would have been classified as trouble makers in any other environment, and they came out of that program with purpose and self-confidence. Not because our adults imposed any particular moral values on us, but rather because they provided a safe environment for a bunch of kids, then encouraged us to learn something. No one really cared what - as long as you were working to better yourself. As much of the responsibilities as possible were put on the scouts to serve as a chance to learn about handling responsibility.
Much like the parent poster, our troop was just a bunch of guys and their dads, along with a few active moms. The values of they troop reflected the values of the people in it. We didn't have any openly gay families or rabid atheists, but neither did we have any bible-thumping types.
I suspect that, like any situation where people are allowed to interact, most of the stories that have made the news did so because two extremists dug their heels in and started shouting a lot. I suspect that you'd find the overall attitude of the organization to run closer to "Why are you guys shouting at each other? Are we going to climb this mountain, or what?"
With that said, the summary makes it sound like the request for help from the open-source community comes from the highest levels of the organization. And it sounds a little out of tune with what they usually teach. Seems to me that there are plenty of people that are already active and involved that could probably help them.
For perspective, I was raised Christian, but without any serious family conviction. I've since become what my friends describe as a "non-practicing atheist". Scouts had approximately zero influence on how I view religion.
I think these charts reflect nothing. Only if you read the charts at face value. Instead of trying to interpret messy search volume data, ask what the charts, considered together instead of individually, say about the circumstances which generated the data.
While you can't draw any hard statistical conclusions due to heavy sampling bias (which TFA acknowledges), you can make some interesting observations.
Here's what I notice. The models worked pretty well for the Democratic primaries, but hardly predicted anything about the Republican primaries. This says something about the internet usage patterns of the voter demographics for the two parties. Democrats appear more likely to use the internet to get their information, while Republicans appear to prefer other media sources, probably traditional outlets like newspapers and TV.
Those are broad generalizations, but they are still interesting, and they do provide some insight into why Barack Obama's internet presence appears to be politically relevant, where Ron Paul's internet presence was just a fad.
The data is what it is. It says nothing more or less than what you see on the page when considered in isolation. Look past the presentation of the data, and you may just find real signal amidst all the noise.
I'm a recent PhD graduate from a good public university chemistry program. No formal programming was required for chemistry, physics, or in any of the engineering programs that I had contact with. Rather then try to make this sort of thing a formal requirement, the profs that needed it had just started requiring their own grad students to go take a C 101 course in the computer science department, then supplementing that with what ever language they liked to use in their own lab.
That said, I'd have like to see a requirement for some programming. More than anything, it would have helped to educate the profs as to what tools were actually available. During my time, I was asked to do various tasks in Matlab, Mathmatica, Excel, C, Fortran, Perl, and Python.
The people using Matlab and Mathmatica were only doing so because they didn't know any other programming environments. While these are fantastic for pure math problems, they aren't really general purpose enough for most of what you really need to do.
The people using Fortran did so because 'it is *the* language of scientific computing, and it has been around forever, and will be around forever'. I have never been in so much pain in my life. The simplest of tasks, (file I/O, etc) took longer to write then doing the entire program (I/O, analysis, graphing) in a higher level language like Python. I tried to explain that most code that gets written in an academic lab was probably going to be used a couple of times to do some one-off analysis, and then discarded - it was only worth the time to write it in Fortran if the code was really going to be immortal, and then only if it needed to be fast. That fell on deaf ears, because Fortran was the *only* worth language, in many people's eyes.
The people that only knew Excel were by far the worst - 'when all you have is a hammer...'. If you are a VBA wizard, you might be able to do something good with it. But it can't do graphs more complicated than linear regressions, you are limited to 255 columns and 65k rows, and the IF statement available in the normal worksheet doesn't count as real flow control. Most of my problems in grad school were fixed by taking away peoples precious spreadsheets, wandering off and hacking some Python together, and giving them back the answer.
The exact language matters much less then what you can do with it. Once people realized that I could solve problems quickly and easily that they didn't even know how to start on, they started coming to me with all sorts of requests. I'm currently starting to make a career of it, because too many people in my field don't know any programming.
Here's what I would do. Find a real world data set, and do some basic analysis on it. A couple of simple transforms, and some graphing. Assign homework to all of the profs you are having this discussion with. The methods they use don't matter - only the results. Then get back together and compare notes.
It should be easy to pick a large enough dataset, a transformation that is easy to express in mathematically, but almost impossible to do without some basic numerical analysis library computationally, and a graph that simply cannot even be attempted in Excel. I doubt you'll have to contrive an example - I ran across those sorts of datasets all the time. Bonus points if it comes from one of your colleagues published work where some grad student nearly killed themselves trying to do it in Excel.
I prefer Python to Java for one very simple reason - it is the most pragmatic development language I've ever used (Perl included).
The dynamic typing, memory management, etc. mean that on average, developing a program in Python is ~10x faster than doing the same program in C. Note that what most C hackers are actually scared of is weak typing, and Python is a strongly typed language (Perl is what you are really scared of).
The object model is cleaner than any other language I've ever worked with, but you don't need the overhead of OO to write one-off scripts, like you do in Java.
The core language is clean, and there is a great emphasis on making programs easy to read, which means that maintenance for Python programs is *much* easier than any other language I've ever worked with.
The whitespace thing is a non-issue. You were going to indent your code anyway, weren't you? My indentation habits really didn't change much at all when I started using Python.
The standard library, while not covering the breadth that Java's does, vastly simplifies all of my common daily tasks. There are also bindings to a huge number of toolkits that were written for other languages (wxWindows, etc.).
Python plays well with other languages. Hardcore algorithms can be written in C and called from Python. Jython sits on top of Java, and offers all the ease of development of the Python language with the underlying breadth of Java's library. IronPython is the.Net version. Python makes a nearly perfect 'glue' language to manage any number of other programs written in any number of other languages with any number of I/O formats.
I could go on, but it comes down to this. Python doesn't do everything. If you must have pure speed, try C. If you are developing a huge code base, take advantage of the breadth of Java's library. If you need really good threading, check out Erlang. For pure text parsing, nothing has managed to touch Perl yet. Ada even has a place for those who demand the strictest typing. But, if you are doing prototyping, RAD, one-off scripts, mid-sized application development, or anything that could be done in Perl but needs to be maintained for more than 5 minutes, Python deserves a look.
Python does most things very well, and almost nothing badly. And unless you know you *need* something like C's speed or Java's massive library, you'll find that you can develop the application so much faster in Python that there is rarely a reason to go anywhere else. For most projects I've been involved with, development time matters much, much more than an extra couple seconds of execution time.
Python lets me do everything I want to do, faster, easier, and more readably. Is it perfect for every task? No, but it's pretty damn good for mine.
Seems to me that most of the people promoting DVCS have used them and have seen the light. Once you use a DVCS on a project you don't want to go back to the bad old way of doing things.
Indeed. The number one complaint I hear about them is that they aren't a centralized system.
As a one man programming show and a mercurial user, I set up a centralized repository for myself. That was the clean copy - it always worked. I'd have 1-4 other copies out at any given time for whatever I was working on. It offered extreme flexibility, and at the same time, all the usual advantages of having a centralized repository. I know with mercurial (and probably most of the rest of them) that there are ways to formalize that, if it is what you need.
I hear the same things from people who have worked on larger teams with these kinds of tools.
Distributed systems let you work the way you think, rather than making you think about the way the tool works.
Their argument is based off a strictly capitalist view. If you consider the notion that there is no way to claim your work as your own once it is under the GPL and generate a profit directly from it, in this world view, it's a waste. They see GPL as a trap where once entered, there is no escape....
That said, I do personally like to be able to release closed source versions of things, and allow others to do the same. The BSD and Eclipse licenses appeal to me more than the GPL. I consider myself a capitalist, and this doesn't ring true at all. The closed source way of doing things tends to restrict the flow of information (in this case, source code), creating an artificial scarcity on a product that has essentially zero marginal cost. Bill Gates considers this to be a good thing for one reason - it makes him piles of money. However, he leans heavily on the current snafu that is America's IP laws to continue to generate massive profits long after he has paid off his costs. Clearly there is an inefficiency in the market that would quickly be fixed if we were really discussing capitalism, and not greed.
I have no problem with closed source software - if the company that wrote it profits from it. I have huge problems with Microsoft abusing its monopolistic position in the market to run competitors out of business.
What it comes down to as an individual is this - if you license your code under BSD/MIT licenses, you are giving up your right to claim that code as your own. That is your choice, and I won't lambaste you for making it. You should know, however, that Microsoft, or a company like them, will pick it up and make a fuck-ton 'o cash on it by leveraging some other monopolistic position that you don't have.
I GPL as much of my code as I can, for one very simple reason. It makes it so that I can continue to use it long after it would have rotted away locked up by this company or that university. I keep the rights to use/profit from what I wrote, which is highly advantageous to me, considering I don't have the legal weight of a multi-national corp backing me up.
Also, from the capitalism standpoint, my code is out there, and any company can pick it up and use it, sell it, whatever. If they can make a profit from it, more power to them. But they cannot (legally) distribute it without releasing the code. And if it gets to be profitable enough, the nature of the GPL will allow someone else to step in, overcome the barriers to entry, and undercut the profit. Capitalism at its finest - if you get too greedy, someone will steal your customers.
Because it limits the ability of a corporation to throw up enormous barriers to entry, the GPL is probably the most capitalist of the open source licenses. *It is for precisely this reason that Billy G. doesn't like it.* He knows that in a perfectly capitalist society, he wouldn't be richer than God and married to an (aging) supermodel. Someone would have undercut him years ago, driven innovation in the OS industry, and left MS in the dust.
More to the point, he's afraid that the FOSS community is in the process of doing that as we speak. MS has a strangle-hold on the desktop, but they are woefully behind on mobile/ultra-portable platforms, and are flailing to catch up with Linux (eee, anyone?). Additionally, they are woefully behind on the web, and have been chasing Google for years now, and they are woefully behind in the MP3 market, as the iPod continues to dominate after ~6 years.
GPL'ed code adds value for anyone who wants it. It may or may not generate profit for anyone who wants it. But at least I can benefit from continued access to it, even if a big corp picks it up.
Bill Gates just wants to profit from work he didn't do himself, and didn't have to pay for. That's why he doesn't like the GPL, and I do.
The other responses to this comment always make the problem *way* harder than it has to be. The problem that most people have here is that 1/3, 1/2, and 2/3 (the relevant fractions in the problem) are all very close to each other, and even smart math types get confused about it because the answer isn't intuitive.
There is a *much easier* way to get the right answer - increase the number of doors to, say 100. There are 99 goats, and one car.
* You pick one door. Your chances of picking the right door: 1/100.
* Monty opens 98 doors, revealing 98 goats.
* Your chance of getting the car if you *don't switch*: 1/100. You picked that door before you had seen any goats. It was a random guess, and you can do no better than random chance in picking the right one.
* Your chance of getting the car if you *switch*: 99/100. Monty had to know where the car was, so that he didn't accidentally reveal it when he showed you 98 goats. If you picked wrong the first time (99/100), Monty left the one door with a car behind it closed, and that other door is the one that you want.
---
Different thought experiment:
Monty shows you 100 doors, again with 99 goats and 1 car. *Before you pick* he opens 98 doors, showing you 98 goats. What are your chances of getting the car now? 1/2 - just a random guess.
---
As you can see, the problem is much easier to reason through when the fractions are 1/100, 1/2, 99/100 instead of 1/3, 1/2, and 2/3.
I'm not sure that this helps explain why monkeys eating M&Ms 50 years ago confused somebody, but hopefully it clears up the whole Monty Hall thing once and for all.
Even if the tally was exactly right, in general you cannot prove a program correct by using only black box testing. There are simply too many possible inputs to have time to test for all but the most trivial inputs. For all we know, there's a backdoor or unintentional security vulnerability that can be used to alter election outcomes. We need to be able to examine the machine's inner workings to have any hope of verifying those are not problems with the voting machine. You are not wrong. But, this article raises an interesting point - while consistency checks won't prove that there is a bug/vulnerability/backdoor, they will raise red flags in a significant number of cases that *something* isn't right.
The problem with electronic voting in general is that there are a number of places where it can go wrong. Let's assume you do get the source code, and prove that it is correct. Can you also prove that this exact version is what is installed on every voting machine, in every polling location? Checksums are nice and all, but the checksumming software could be tampered with. Can you be sure that no other software is also installed that could alter the core application at run-time? Can you be positive that the results cannot be altered after they are already entered? Hell, can you be positive that the compiler used hasn't slipped something into the executable? As it stands right now, I think the answer to these questions is, collectively, no. Somewhere, there is a piece that is going to be extremely difficult to verify in all cases. It doesn't have to be much - a 1% error in the results would have swung a couple of the last elections. Some allege that this has already happened. (Hey, you all, in the back - with the tinfoil hats. Raise your hands...)
In addition to requiring open source code, we should also have in place an extensive system of consistency checks to ensure that we catch most of the obvious ways to rig an election. If there are not the same number of ballots cast as there are people casting, that's bad. If a number of votes get invalidated (because of hanging/dimpled/pregnant chads, or what have you) that's bad. If people that can't vote (say, because they are dead), somehow manage to, that's bad. All of these things have been used as evidence of voting fraud in the past - don't throw them out just because you 'validated' some random piece of code.
One more thing. If we are going to use electronic voting (and it seems like we are), you also need to get a voter-verifiable print-out - like, you know, on paper. This way, you can be sure that if something is wrong, it'll be caught on a hand-recount, and your vote will still mean something. This is really just the ultimate consistency check, and I don't see how we are going to get around the fact that without it, there will always be a way to tamper with the results. Check out http://coloradovoter.net/ for more - or look for a group of concerned citizens closer to where you live.
As someone who would like my vote to count, I think we should ban all voting machine manufactures that don't agree to these sorts of checks. If they are trying to avoid this for any reason, I think they've got a hidden agenda. No more excuses about proprietary source code - if you want your machines to be used, you submit to a battery of external reviews and consistency checks. No exceptions.
Restricting mega-prizes to the young may eliminate groundbreaking work by mid-career and early-second-career scientists.
Not only that, but it sends the wrong message to our children: Once you hit 30 you aren't worth as much.
I don't think this point can be over emphasized enough in this discussion.
The author is aiming this prize at me. I went to college on a academic full ride, cranked through a PhD in chemistry in 4 years on a hot project that got national media attention, and am currently trying to figure out what my career is going to be. I'm 27, which is extremely young for a PhD.
I am the wrong person to aim this at. You want to throw money at someone, you need to be targeting my PhD adviser. She has connections that I can't dream of, a funding rate that is basically unheard of, deserves a big chunk of credit for my success, has published major work in two very different fields, and, most importantly, she's currently in the prime of her career - age 45. She has now left the university and started a company - it's the only way for her career to continue to move forward with the grant situation as bad as it is right at the moment.
It takes a very long time to establish yourself as a superstar in the world of science. Nobody does it by age 30. The best of the best, with all of the breaks going their way, might do it by 35 - with the caveat that they have to specialize to such an extent that they can't even consider going after a big prize like this unless it is perfectly suited to their field. And unless you are already a on the path to becoming a superstar, you won't get a sniff of big money like we are discussing here.
Better yet - don't throw that money at anyone at all. Inevitably, some of it would stick, but far more resources would be wasted competing for it.
I'm rocking the boat in a localized fashion right now. I'm making a name for myself by being the programming/database guy in a room full of biologists. I'm don't have to the smartest guy in the room - I have access to an entirely different set of tools than anyone else does, and I can communicate with the biologists in ways that a normal programmer would never be able to, allowing me to make a huge impact fresh off of my PhD.
If you really want see progress made, without the high risk/high reward gambles, look to make progress in the gaps between fields. Engineers collaborating with traditional academic scientists. PhDs in two major fields, instead of just one. Collaborative projects between industry and government, academia and industry. Corporate think tanks like we used to have - really good R&D in industry is hard to come by these days, but many of our best advancements in the last 50 years came from these sorts of institutions. Improved math/comp sci training for scientists and engineers (I don't care how much you had, more would probably have done you good). A major, national-involvement project to tackle, on par with putting a man on the moon - real renewable energy looks like a good candidate right now.
This is the future of America, and most of the rest of the first world. We have outsourced our blue-collar jobs, the white-collar jobs are slowly going international, and our high standard of living looks unattractive when someone in India will do your job for half the cost, even if they only do it half as well.
The way forward is to move faster, drive innovation, reward the people that are superstars (regardless of age) with incentive packages that make them want to work harder. America has had this sort of system in place a few times before in history, and we have attracted the best and the brightest, both domestic and foreign, to get involved and make huge strides in many fields. Progress is made on the margins - any attempt to maintain the status quo or fund a regression to the mean kills us slowly. Throwing big money at science keeps mediocre talent in, wasting resources, when they should throw in the towel and move on to
The oped peice refers to religious faith as "belief without evidence."... Rather they wrote what they considered to be personal evidence, with the hopes that readers of their words would likewise seek for their own personal evidence Sounds like the words you read were not the words the author wrote. That's not an attack on you, just my observation.
"Evidence" is one of those tricky words with too many definitions to be of much use in a discussion about religion and science. To a scientist, "evidence" means "a controlled, reproducible, empirically observable phenomenon". The "personal evidence" you mention means "I believe I saw something with my own eyes". The problem with "personal evidence" from a scientific perspective is that while it very well may have been real, you can never verify it independently. That opens up a whole can of worms where you potentially start accepting mirages, hallucinations and selective memories as "evidence" of what you want to believe anyway.
I can easily find "personal evidence" to support God, the FSM, the IPU, and the pterodactyl that lives under my bed. None of it matters from a scientific perspective if I can't show you a reproducible phenomenon.
That's what the author was getting at when he says religion is "belief without evidence". But therein lies the fundamental problem of having this sort of discussion at all. The reason that the "laws of nature" aren't discussed much by scientists is that we don't really have any grounds on which to discuss them. They are observations that we have failed to find exceptions to, but we don't have any overarching model to explain *why* they are.
It has been stated on Slashdot many times before. Science is the domain of *how* or *what*. Philosophy and religion deal in *why*. These domains are complimentary, and there isn't really any overlap. When scientists try to explain *why* or religion tries to explain *how*, someone is going to get their panties in a bunch.
As for the memory stick one, my dear old 512MB Sandisk USB memory stick has been through the wash twice and survived fine. I've heard other people say the same thing. Anyone else have this happen to them? Anyone have a bigger storage medium go through the wash? I've got an 8 GB USB stick that got washed once. Seems to work better now - even with the little cap thing on, pocket lint always seems to get in to the contacts. The washing machine took care of that, and I found it before it got dried.
Now if only I could do that with all of my other computer components - my main desktop accumulates dust like no other. Maybe I'll throw the mobo in the dishwasher...
Care to point me to any scientific evidence that Tamiflu, Relenza, or any other such drug in the pipeline will save a single person from a pandemic type flu virus? Sure.
Search Google Scholar with "TamiFlu H5N1".
The first link on the results page takes you to an article by Roche scientists, http://jac.oxfordjournals.org/cgi/reprint/55/suppl_1/i5.pdf.
They have a financial interest in TamiFlu, so don't just take their word for it - feel free to read the all 95 of the references. Flu antivirals are well characterized, and mutations that cause resistance are well understood. There have been plenty of animal studies, and multiple case studies in humans. For further reading about those case studies, try http://content.nejm.org/cgi/reprint/353/25/2667.pdf. That article has additional discussion about the possibility of mutations during the current recommended treatment course.
Even for non-pandemic strains, the evidence that vaccines and antivirals have had any impact of flu death rates is extremely thin. Antivirals are currently being used to decrease morbidity and mortality caused by influenza. There is good statistical evidence, confirmed by multiple independent studies, that these work as advertised. And plenty of discussion about when they fail.
No such evidence exists. Served.
one must be careful not to make claims that aren't (sufficiently) supported by the data. On the other hand, making wild claims can sometimes make it easier to get future funding... Good point. If the support for the claims is overstated, you really end up making yourself look like an idiot. However:
In what I would call "hard science", you don't bother with correlation coefficients. I would quibble with this point, especially when venturing into new or poorly understood corners of the scientific world. Long before anyone understands a topic well enough to put together a rational model, *many* papers will be published and *many* speculative models will be proposed and ripped apart. The common element in the hard sciences is that you can propose a quantitative model that can in fact be ripped apart by better data. Relativity is a classic example - Einstein proposed models long before they could be tested that later turned out to be quantitatively predictive.
If you overstate your claims here, you still look like an idiot. But, if you don't speculate, the field doesn't advance. Soft science (like the article) has the same problem without the possibility of resolution via experiment - you've got to speculate about causation without overstating your conclusions.
If the OP had said 'Hey, I read the article, and those claims are way overstated. Correlation != Causation." I wouldn't have had any problem. Hell, I didn't even read the article. What I took umbrage to was reflexive "Correlation != Causation" without R'ing TFA. That is getting really old, and always gets moderated up - it is becoming a form of karma whoring.
Slashdot Mirror
'Dogma' is common in the sciences, but it implies something different than the formal definition you are thinking of. It is usually used to describe a highly simplified model of how a system works. It's just a useful way to think about something.
The most well known example is the central dogma of molecular biology. By the time you finish freshman molecular biology in college, you know that it is a gross simplification of how a cell works, but that it is a very good first approximation.
Chemistry is no different. The vast majority of chemical interactions involve the valence electrons. So how do you introduce the topic? You say 'all chemistry deals with valence electrons (cough, cough)'. If the students learn that, you're actually doing pretty well.
Once you get past the basics, you admit to the students that you might have fibbed, and that under unusually high energy conditions, the inner shell electrons actually can interact. Upper level chem courses have been teaching this for years - there are no surprises here.
All the article says is that a research group is predicting a previously unknown inner shell electron interaction under high energy conditions. While it is news, it is not shocking, and while it violates the 'dogma' that only valence electrons interact, it changes nothing about how the dogma will be taught.
Progress in science is made at the edges. What happens to this at high energies? How will these atoms behave at extremely low temperatures? The easy cases have been understood for years, if not centuries. This discovery doesn't change any of that. So this is cool, but not a fundamental break-though.
Now, if someone replicates this experimentally, and then figures out how to use it to make dilithium crystals to power their prototype warp drive, that'd be revolutionary.
From TFA
This is not, however, the end of the story. The PTO's decision in Ex parte Wasynczuk provides one final twist so Kafkaesque as to strain credulity.
...
In sum, an innovative process is not patentable when operating on a single computer processor but is when operating on two processors, even though the Board recognizes that the process in the unpatentable claim "is essentially the method" set forth in the patentable claim.
I really hate law sometimes. Using two computers (instead of just one) means that the method is tied to a 'particular machine' instead of a general purpose computer.
So while this may kill of some current software patents, only a slight change in hardware is required to keep on patenting software. Why even bother? All this does is ensure that every software patent from this day forward will be tied to a 'particular machine'.
I harbor a secret hope that this will still render software patents ineffective, because to get a patent, you'll have to say that your process can only be run on a 'particular machine' comprised of two computers, while your competitors will just run a 'general purpose computer' and claim that they are not violating your patent. That's still probably wishful thinking, given my usual misunderstanding of how laws actually works.
The Google App Engine already has the Django libraries available.
Have a little care, there. The GAE caused quite a stir in the Django community, because it is only a partial implementation of the database interface that Django normally uses. It is backed by BigTable, which is blazing fast, but not a full blown relational database. If that works for you, go for it - it looks like a sweet platform for certain kinds of projects.
As for your question about Java->Python, I'm a former C++ convert myself, but I can help a little here. For some 'compile time' checking, look at PyLint It may check too much for you, but you can turn off the stuff you don't want.
As for unit testing, PyUnit is a pretty straight port of JUnit, so that should look familiar. However, I actually find nose to be a little better. It has many of the same capabilities, but with less boilerplate needed, and it integrates well with any existing PyUnit or DocTest tests.
SpaceX has a rocket with slightly higher payload than the Ares I scheduled to fly in 2010.
Didn't know that. Thanks for pointing it out.
I still have a major problem with anything SpaceX at this point, though. It isn't flight proven. Both Atlas and Delta are riding streaks of ~90 successful launches in a row, spanning several vehicles. The Atlas V and has had 14 launches, with one minor upper stage glitch (this wouldn't have killed anyone). Delta IV has launched 8 times, one partial failure in an experimental heavy launch configuration (wouldn't have killed anyone).
SpaceX has launched twice - two total failures, at least one of which would have killed the crew. There's a good quote from Elon Musk here.
"I think they had something like 12 Atlas failures before the 13th one was success. To get this far on our second launch being an all-new rocket -- new main engine, new first stage, new second stage engine, new second stage, new fairing, new launch pad system, with so many new things -- to have gotten this far is great."
SpaceX has plans to man-rate their line of rockets. But it is a long way off, and they need a few successful launches before you could actually strap a human on.
Any way you look at it though, NASA is flat wasting years of time and $5Billion in cool hard cash developing the Ares I.
Yep. Or more.
There are alternatives [spacex.com].
Yeah, including some real alternatives, that can actually get into the orbits NASA needs to get to, rather then just barely out of the atmosphere (where you can tell a tourist that they are 'in space'. Like this or this.
The US Government has already funded the development of not one, but two rockets with the kinds of capabilities they need. They are flight proven, expandable to handle all sorts of loads, and available right now, not whenever Ares will slip out to. Add a little redundancy in a couple of systems, and have them ready to launch American astronauts into space in two years.
SpaceX is cool, and is probably the direction that the future of American space exploration needs to go. But it is not ready, it is not proven and it doesn't come close to the kinds of payload capacity or reliability that we need now. Check back around the time when Ares is supposed to be done to see what SpaceX is up to. In the mean time, quit screwing around developing a rocket similar, but slightly different from, the two perfectly good commercially available ones that are already up and running.
Actually, some of the very first Ares prototypes were based on the Atlas V, and the Russian designed RD-180 engines (which slightly outperform the RS-68s, and make anything shuttle derived look like a class-D model rocket engine). My wife worked on some of the prototype ground software systems.
At the time, Lockheed (now United Launch Alliance) put in a serious proposal to man-rate the Atlas V. NASA scoffed, which strikes me as a little strange. The shuttle technology is based on a 30 year old design, is planned to top out at about 20 total missions, and has blown up two orbiters. Both Atlas and Delta have newer flight-proven technology, better launch track records then the shuttle, lower prices, etc. Delta has the capabilities to add additional stages, Atlas intends to shortly, and both Atlas and Delta can add additional liquid or solid boosters as needed to handle huge payloads.
There is still noise about this idea now - click here for more information. But it won't happen.
These rockets are not in development - they exist now, and have a similar or better performance profile to anything that Ares or Jupiter have been proposed to have. Yea, you'd have to add redundancy to some of the systems, but you wouldn't have to redesign.
Meanwhile, who does have the capabilities to put people into space right now? The Russians, on well-engineered, cheap-to-produce ballistic missile-style vehicles. And while NASA flounders around redesigning 30 year old shuttle derived technology, and watching their launch date slip out for years, if not decades, we will be dependent on the venerable Soyuz keep Americans in space.
NASA is technologically bankrupt. All the great engineers they had 30 years ago are in private industry. But they still have some of the best bureaucracy and politics!
does that make me a housewife?
I think I need to go solder something...
There is just one problem with this - getting timely, reliable data.
I work with flu, and the epidemiologists I know would love to have a system that could "facilitate early outbreak detection, increase public awareness of outbreaks prior to formal recognition, and provide an integrated and contextualised view of global health information."
A few sites that should help do this for flu are coming online, but the biggest impediment is getting timely, reliable, geo-tagged data. The local physicians know that an outbreak is starting, but it takes a while for samples to filter up to the state and country levels, be reliably analyzed, and then be uploaded to a tracking website.
While better websites will help in this process, the bottleneck occurs much earlier in the process. Samples come trickling in from third-world countries months after they were collected (or sometimes after the flu season is over!) and they might have the name of a city associated with them. That's not much help when you have to pick next season's vaccine halfway through this season so that it'll be ready in time.
And don't think that asking the local physicians to use a website is going to help - not for flu or any other disease. Most of the time, you are likely to get a hand-written sheet, partially filled out, with five or six columns of basic information for each sample. They don't have time to do any more then that.
Sure, the pandemic threats get a bit more attention, but the seasonal stuff still kills hundreds of thousands of people every year. You want to help? Start with those third-world clinics. For most diseases, the CDC and WHO still have to get people out on location to do good surveillance, and a website just isn't going to change that.
Easier answer - pick books that are an easy read and a fun introduction into the genre that you are trying to get your kids interested in.
My dad read The Hobbit to my sister and I when I was in ~3rd grade, then started into some David Eddings. By half way through the first book, we were moving too slow, so I stole the book and tore through the rest of the series. Once I started, there was no looking back - I went on to read everything I could get my hands on. Sci-fi, fantasy, mysteries, classics - whatever. All I needed was a starting point.
My wife, who grew up in a house where the TV was a family activity, never thought she liked reading, until I started reading some of those books to her in the evenings. It's become one of our favorite ways to spend time together. We've read Tolkien, Eddings, the Harry Potter series, and Narnia so far. We're doing A Wrinkle in Time right now, and I'm thinking we might try Dune next. In addition, she's now started to read quite a bit more on her own - some literary classics, some sci-fi, but all stuff she picked out herself. Most of it, I've never read. All she needed was a starting point.
To get someone started, pick something that has a good story and dive it. Show them that reading can be an extremely enjoyable activity. Provide a little direction toward a couple of books at the right reading level, then sit back, and smile and nod knowingly when they start telling you how great one of your favorite books is, or how they've discovered this great new series...
I have not read your articles. While I have a PhD in chemistry, it isn't in climatology. And I am here to tell you, in no uncertain terms, that everything you think you know about AGW is wrong. For *exactly* the same reasons that teaching ID as science is wrong.
Between 2003 and 2007, I worked in a room with a bunch of atmospheric chemists from the lab next door. They had read all of those articles, and hundreds more. They were all true believers - global warming is very real, and is caused by humans, period-end-of-discussion.
Except, in all of the seminars that I attended, they never said that outright. They would put up this graph showing a bunch of different sources of global warming and global cooling with error bars bigger than the data points. Which is a scientist's way of saying "this system is so complex that we can only guess at what is really going on - check back in a few decades if you want to know how this all turns out". They could not, in a room full of scientists, make a rational statement about AGW.
One day, toward the end of my PhD, our division seminar was about starting a university-wide effort on global warming. Many non-scientists attended. And at that *one* seminar, the true believers showed their colors. "AGW is real and it will kill us all!" they shouted.
The point I am making here is a very simple one. Blind faith is not science. Blind faith comes in many, many forms, and sometimes from the most unexpected of places.
The true believers in my story include some of the best atmospheric chemists America has - the CU Boulder/NIST/NOAA triangle forms the highest density of atmospheric researchers anywhere in the world, and includes many national academy professor types.
These people are caught in a strange, three-way hard place.
* On one hand, they are trained scientists - if you have not tried to disprove something like AGW every which way and failed every time, you can't claim it as truth. This is science, and it is a harsh mistress.
* There is the fact that it is almost impossible for these people to get funding unless they claim that their research will somehow lead to a better understanding of AGW. This is about money and the longevity of your career - will you sacrifice everything you have to be contrarian? Some of the greatest historical scientists did, becoming martyrs in the process.
* And then, their is the fact that they truly believe. This is the political, emotional, groupthink angle. In our current society, you are an outcast unless you believe the right things (like AGW). This is an irrational, emotional angle, and it runs deep in every one of us, hidden beneath a thin vernier of logic and rational thought.
And so, as a scientist in this position, you never say AGW in seminars with other scientists - but you do to non-scientists, the media, the funding agencies. You search for evidence to support AGW, because you believe, even as you search for evidence to disprove it, because you are trying to be a good scientist.
Most importantly, you judge papers you review only on their science - never their speculation. This means the materials/methods/experimental section - never the conclusions, where you are allowed to speculation about what this research *might* mean. As long as everyone does it that way, you can maintain the illusion that you are somehow separating what you *know* from what you *believe*, and thereby claim to remain objective.
Just because you read it in a peer-reviewed paper doesn't make it true. If the paper in question is regarding a politically/emotionally-charged topic, you can safely assume that the only facts in the paper are the hard numbers, with reported errors, that have been independently verified, and have stood up to repeated scrutiny over the period of several years. And you should probably take even those with a grain of salt.
Why am I ranting about AGW in a creationism vs evolution article? Because the problem that both fields have is the same one. True
The pathology runs deeper than you are implying. Mike Dillon , general counsel for Sun, goes so far as to compare patent portfolios to a cold war style nuclear arms race. Far from being 'frivolous', most companies today think that if they don't try to patent every little thing, they'll be sued into oblivion by the companies that do. It is this defensive mentality, not the patent trolls, that has lead to the current state of the patent system.
As someone whose first patent was published just this week, it is a little bittersweet. I'm happy that I was involved in a project that generated some truly innovative stuff, and that I've actually made a little money from the licensing of the rights to the patent. But the innovative parts of the project are not what got patented, the company that licensed the technology appears to be sitting on it as part of their defensive war chest, and I've had to battle their legal counsel to avoid their repeated attempts to get a patent on the software that I wrote for the project (which is useful, but not even remotely innovative).
I think as a patent holder, I see the needs for reform even more clearly now. Much like the cold war, I don't see the 'arms race' slowing down until company is holding so many patents that no one can make a move in any arena for fear of being destroyed by someone else. Mutually assured destruction is a technique for maintaining the status quo, and eventually, it will kill the ability to truly innovate in America. And when it does, another number of other nations will gladly step in and relieve us of our 'superpower' status.
If anything, it was one of the most *inclusive* environments that I have ever been associated with. We had all sorts of kids that would have been classified as trouble makers in any other environment, and they came out of that program with purpose and self-confidence. Not because our adults imposed any particular moral values on us, but rather because they provided a safe environment for a bunch of kids, then encouraged us to learn something. No one really cared what - as long as you were working to better yourself. As much of the responsibilities as possible were put on the scouts to serve as a chance to learn about handling responsibility.
Much like the parent poster, our troop was just a bunch of guys and their dads, along with a few active moms. The values of they troop reflected the values of the people in it. We didn't have any openly gay families or rabid atheists, but neither did we have any bible-thumping types.
I suspect that, like any situation where people are allowed to interact, most of the stories that have made the news did so because two extremists dug their heels in and started shouting a lot. I suspect that you'd find the overall attitude of the organization to run closer to "Why are you guys shouting at each other? Are we going to climb this mountain, or what?"
With that said, the summary makes it sound like the request for help from the open-source community comes from the highest levels of the organization. And it sounds a little out of tune with what they usually teach. Seems to me that there are plenty of people that are already active and involved that could probably help them.
For perspective, I was raised Christian, but without any serious family conviction. I've since become what my friends describe as a "non-practicing atheist". Scouts had approximately zero influence on how I view religion.
While you can't draw any hard statistical conclusions due to heavy sampling bias (which TFA acknowledges), you can make some interesting observations.
Here's what I notice. The models worked pretty well for the Democratic primaries, but hardly predicted anything about the Republican primaries. This says something about the internet usage patterns of the voter demographics for the two parties. Democrats appear more likely to use the internet to get their information, while Republicans appear to prefer other media sources, probably traditional outlets like newspapers and TV.
Those are broad generalizations, but they are still interesting, and they do provide some insight into why Barack Obama's internet presence appears to be politically relevant, where Ron Paul's internet presence was just a fad.
The data is what it is. It says nothing more or less than what you see on the page when considered in isolation. Look past the presentation of the data, and you may just find real signal amidst all the noise.
I'm a recent PhD graduate from a good public university chemistry program. No formal programming was required for chemistry, physics, or in any of the engineering programs that I had contact with. Rather then try to make this sort of thing a formal requirement, the profs that needed it had just started requiring their own grad students to go take a C 101 course in the computer science department, then supplementing that with what ever language they liked to use in their own lab.
That said, I'd have like to see a requirement for some programming. More than anything, it would have helped to educate the profs as to what tools were actually available. During my time, I was asked to do various tasks in Matlab, Mathmatica, Excel, C, Fortran, Perl, and Python.
The people using Matlab and Mathmatica were only doing so because they didn't know any other programming environments. While these are fantastic for pure math problems, they aren't really general purpose enough for most of what you really need to do.
The people using Fortran did so because 'it is *the* language of scientific computing, and it has been around forever, and will be around forever'. I have never been in so much pain in my life. The simplest of tasks, (file I/O, etc) took longer to write then doing the entire program (I/O, analysis, graphing) in a higher level language like Python. I tried to explain that most code that gets written in an academic lab was probably going to be used a couple of times to do some one-off analysis, and then discarded - it was only worth the time to write it in Fortran if the code was really going to be immortal, and then only if it needed to be fast. That fell on deaf ears, because Fortran was the *only* worth language, in many people's eyes.
The people that only knew Excel were by far the worst - 'when all you have is a hammer...'. If you are a VBA wizard, you might be able to do something good with it. But it can't do graphs more complicated than linear regressions, you are limited to 255 columns and 65k rows, and the IF statement available in the normal worksheet doesn't count as real flow control. Most of my problems in grad school were fixed by taking away peoples precious spreadsheets, wandering off and hacking some Python together, and giving them back the answer.
The exact language matters much less then what you can do with it. Once people realized that I could solve problems quickly and easily that they didn't even know how to start on, they started coming to me with all sorts of requests. I'm currently starting to make a career of it, because too many people in my field don't know any programming.
Here's what I would do. Find a real world data set, and do some basic analysis on it. A couple of simple transforms, and some graphing. Assign homework to all of the profs you are having this discussion with. The methods they use don't matter - only the results. Then get back together and compare notes.
It should be easy to pick a large enough dataset, a transformation that is easy to express in mathematically, but almost impossible to do without some basic numerical analysis library computationally, and a graph that simply cannot even be attempted in Excel. I doubt you'll have to contrive an example - I ran across those sorts of datasets all the time. Bonus points if it comes from one of your colleagues published work where some grad student nearly killed themselves trying to do it in Excel.
I prefer Python to Java for one very simple reason - it is the most pragmatic development language I've ever used (Perl included).
.Net version. Python makes a nearly perfect 'glue' language to manage any number of other programs written in any number of other languages with any number of I/O formats.
The dynamic typing, memory management, etc. mean that on average, developing a program in Python is ~10x faster than doing the same program in C. Note that what most C hackers are actually scared of is weak typing, and Python is a strongly typed language (Perl is what you are really scared of).
The object model is cleaner than any other language I've ever worked with, but you don't need the overhead of OO to write one-off scripts, like you do in Java.
The core language is clean, and there is a great emphasis on making programs easy to read, which means that maintenance for Python programs is *much* easier than any other language I've ever worked with.
The whitespace thing is a non-issue. You were going to indent your code anyway, weren't you? My indentation habits really didn't change much at all when I started using Python.
The standard library, while not covering the breadth that Java's does, vastly simplifies all of my common daily tasks. There are also bindings to a huge number of toolkits that were written for other languages (wxWindows, etc.).
Python plays well with other languages. Hardcore algorithms can be written in C and called from Python. Jython sits on top of Java, and offers all the ease of development of the Python language with the underlying breadth of Java's library. IronPython is the
I could go on, but it comes down to this. Python doesn't do everything. If you must have pure speed, try C. If you are developing a huge code base, take advantage of the breadth of Java's library. If you need really good threading, check out Erlang. For pure text parsing, nothing has managed to touch Perl yet. Ada even has a place for those who demand the strictest typing. But, if you are doing prototyping, RAD, one-off scripts, mid-sized application development, or anything that could be done in Perl but needs to be maintained for more than 5 minutes, Python deserves a look.
Python does most things very well, and almost nothing badly. And unless you know you *need* something like C's speed or Java's massive library, you'll find that you can develop the application so much faster in Python that there is rarely a reason to go anywhere else. For most projects I've been involved with, development time matters much, much more than an extra couple seconds of execution time.
Python lets me do everything I want to do, faster, easier, and more readably. Is it perfect for every task? No, but it's pretty damn good for mine.
Indeed. The number one complaint I hear about them is that they aren't a centralized system.
As a one man programming show and a mercurial user, I set up a centralized repository for myself. That was the clean copy - it always worked. I'd have 1-4 other copies out at any given time for whatever I was working on. It offered extreme flexibility, and at the same time, all the usual advantages of having a centralized repository. I know with mercurial (and probably most of the rest of them) that there are ways to formalize that, if it is what you need.
I hear the same things from people who have worked on larger teams with these kinds of tools.
Distributed systems let you work the way you think, rather than making you think about the way the tool works.
That said, I do personally like to be able to release closed source versions of things, and allow others to do the same. The BSD and Eclipse licenses appeal to me more than the GPL. I consider myself a capitalist, and this doesn't ring true at all. The closed source way of doing things tends to restrict the flow of information (in this case, source code), creating an artificial scarcity on a product that has essentially zero marginal cost. Bill Gates considers this to be a good thing for one reason - it makes him piles of money. However, he leans heavily on the current snafu that is America's IP laws to continue to generate massive profits long after he has paid off his costs. Clearly there is an inefficiency in the market that would quickly be fixed if we were really discussing capitalism, and not greed.
I have no problem with closed source software - if the company that wrote it profits from it. I have huge problems with Microsoft abusing its monopolistic position in the market to run competitors out of business.
What it comes down to as an individual is this - if you license your code under BSD/MIT licenses, you are giving up your right to claim that code as your own. That is your choice, and I won't lambaste you for making it. You should know, however, that Microsoft, or a company like them, will pick it up and make a fuck-ton 'o cash on it by leveraging some other monopolistic position that you don't have.
I GPL as much of my code as I can, for one very simple reason. It makes it so that I can continue to use it long after it would have rotted away locked up by this company or that university. I keep the rights to use/profit from what I wrote, which is highly advantageous to me, considering I don't have the legal weight of a multi-national corp backing me up.
Also, from the capitalism standpoint, my code is out there, and any company can pick it up and use it, sell it, whatever. If they can make a profit from it, more power to them. But they cannot (legally) distribute it without releasing the code. And if it gets to be profitable enough, the nature of the GPL will allow someone else to step in, overcome the barriers to entry, and undercut the profit. Capitalism at its finest - if you get too greedy, someone will steal your customers.
Because it limits the ability of a corporation to throw up enormous barriers to entry, the GPL is probably the most capitalist of the open source licenses. *It is for precisely this reason that Billy G. doesn't like it.* He knows that in a perfectly capitalist society, he wouldn't be richer than God and married to an (aging) supermodel. Someone would have undercut him years ago, driven innovation in the OS industry, and left MS in the dust.
More to the point, he's afraid that the FOSS community is in the process of doing that as we speak. MS has a strangle-hold on the desktop, but they are woefully behind on mobile/ultra-portable platforms, and are flailing to catch up with Linux (eee, anyone?). Additionally, they are woefully behind on the web, and have been chasing Google for years now, and they are woefully behind in the MP3 market, as the iPod continues to dominate after ~6 years.
GPL'ed code adds value for anyone who wants it. It may or may not generate profit for anyone who wants it. But at least I can benefit from continued access to it, even if a big corp picks it up.
Bill Gates just wants to profit from work he didn't do himself, and didn't have to pay for. That's why he doesn't like the GPL, and I do.
The other responses to this comment always make the problem *way* harder than it has to be. The problem that most people have here is that 1/3, 1/2, and 2/3 (the relevant fractions in the problem) are all very close to each other, and even smart math types get confused about it because the answer isn't intuitive.
There is a *much easier* way to get the right answer - increase the number of doors to, say 100. There are 99 goats, and one car.
* You pick one door. Your chances of picking the right door: 1/100.
* Monty opens 98 doors, revealing 98 goats.
* Your chance of getting the car if you *don't switch*: 1/100. You picked that door before you had seen any goats. It was a random guess, and you can do no better than random chance in picking the right one.
* Your chance of getting the car if you *switch*: 99/100. Monty had to know where the car was, so that he didn't accidentally reveal it when he showed you 98 goats. If you picked wrong the first time (99/100), Monty left the one door with a car behind it closed, and that other door is the one that you want.
---
Different thought experiment:
Monty shows you 100 doors, again with 99 goats and 1 car. *Before you pick* he opens 98 doors, showing you 98 goats. What are your chances of getting the car now? 1/2 - just a random guess.
---
As you can see, the problem is much easier to reason through when the fractions are 1/100, 1/2, 99/100 instead of 1/3, 1/2, and 2/3.
I'm not sure that this helps explain why monkeys eating M&Ms 50 years ago confused somebody, but hopefully it clears up the whole Monty Hall thing once and for all.
The problem with electronic voting in general is that there are a number of places where it can go wrong. Let's assume you do get the source code, and prove that it is correct. Can you also prove that this exact version is what is installed on every voting machine, in every polling location? Checksums are nice and all, but the checksumming software could be tampered with. Can you be sure that no other software is also installed that could alter the core application at run-time? Can you be positive that the results cannot be altered after they are already entered? Hell, can you be positive that the compiler used hasn't slipped something into the executable? As it stands right now, I think the answer to these questions is, collectively, no. Somewhere, there is a piece that is going to be extremely difficult to verify in all cases. It doesn't have to be much - a 1% error in the results would have swung a couple of the last elections. Some allege that this has already happened. (Hey, you all, in the back - with the tinfoil hats. Raise your hands...)
In addition to requiring open source code, we should also have in place an extensive system of consistency checks to ensure that we catch most of the obvious ways to rig an election. If there are not the same number of ballots cast as there are people casting, that's bad. If a number of votes get invalidated (because of hanging/dimpled/pregnant chads, or what have you) that's bad. If people that can't vote (say, because they are dead), somehow manage to, that's bad. All of these things have been used as evidence of voting fraud in the past - don't throw them out just because you 'validated' some random piece of code.
One more thing. If we are going to use electronic voting (and it seems like we are), you also need to get a voter-verifiable print-out - like, you know, on paper. This way, you can be sure that if something is wrong, it'll be caught on a hand-recount, and your vote will still mean something. This is really just the ultimate consistency check, and I don't see how we are going to get around the fact that without it, there will always be a way to tamper with the results. Check out http://coloradovoter.net/ for more - or look for a group of concerned citizens closer to where you live.
As someone who would like my vote to count, I think we should ban all voting machine manufactures that don't agree to these sorts of checks. If they are trying to avoid this for any reason, I think they've got a hidden agenda. No more excuses about proprietary source code - if you want your machines to be used, you submit to a battery of external reviews and consistency checks. No exceptions.
Restricting mega-prizes to the young may eliminate groundbreaking work by mid-career and early-second-career scientists.
Not only that, but it sends the wrong message to our children: Once you hit 30 you aren't worth as much.
I don't think this point can be over emphasized enough in this discussion.
The author is aiming this prize at me. I went to college on a academic full ride, cranked through a PhD in chemistry in 4 years on a hot project that got national media attention, and am currently trying to figure out what my career is going to be. I'm 27, which is extremely young for a PhD.
I am the wrong person to aim this at. You want to throw money at someone, you need to be targeting my PhD adviser. She has connections that I can't dream of, a funding rate that is basically unheard of, deserves a big chunk of credit for my success, has published major work in two very different fields, and, most importantly, she's currently in the prime of her career - age 45. She has now left the university and started a company - it's the only way for her career to continue to move forward with the grant situation as bad as it is right at the moment.
It takes a very long time to establish yourself as a superstar in the world of science. Nobody does it by age 30. The best of the best, with all of the breaks going their way, might do it by 35 - with the caveat that they have to specialize to such an extent that they can't even consider going after a big prize like this unless it is perfectly suited to their field. And unless you are already a on the path to becoming a superstar, you won't get a sniff of big money like we are discussing here.
Better yet - don't throw that money at anyone at all. Inevitably, some of it would stick, but far more resources would be wasted competing for it.
I'm rocking the boat in a localized fashion right now. I'm making a name for myself by being the programming/database guy in a room full of biologists. I'm don't have to the smartest guy in the room - I have access to an entirely different set of tools than anyone else does, and I can communicate with the biologists in ways that a normal programmer would never be able to, allowing me to make a huge impact fresh off of my PhD.
If you really want see progress made, without the high risk/high reward gambles, look to make progress in the gaps between fields. Engineers collaborating with traditional academic scientists. PhDs in two major fields, instead of just one. Collaborative projects between industry and government, academia and industry. Corporate think tanks like we used to have - really good R&D in industry is hard to come by these days, but many of our best advancements in the last 50 years came from these sorts of institutions. Improved math/comp sci training for scientists and engineers (I don't care how much you had, more would probably have done you good). A major, national-involvement project to tackle, on par with putting a man on the moon - real renewable energy looks like a good candidate right now.
This is the future of America, and most of the rest of the first world. We have outsourced our blue-collar jobs, the white-collar jobs are slowly going international, and our high standard of living looks unattractive when someone in India will do your job for half the cost, even if they only do it half as well.
The way forward is to move faster, drive innovation, reward the people that are superstars (regardless of age) with incentive packages that make them want to work harder. America has had this sort of system in place a few times before in history, and we have attracted the best and the brightest, both domestic and foreign, to get involved and make huge strides in many fields. Progress is made on the margins - any attempt to maintain the status quo or fund a regression to the mean kills us slowly. Throwing big money at science keeps mediocre talent in, wasting resources, when they should throw in the towel and move on to
"Evidence" is one of those tricky words with too many definitions to be of much use in a discussion about religion and science. To a scientist, "evidence" means "a controlled, reproducible, empirically observable phenomenon". The "personal evidence" you mention means "I believe I saw something with my own eyes". The problem with "personal evidence" from a scientific perspective is that while it very well may have been real, you can never verify it independently. That opens up a whole can of worms where you potentially start accepting mirages, hallucinations and selective memories as "evidence" of what you want to believe anyway.
I can easily find "personal evidence" to support God, the FSM, the IPU, and the pterodactyl that lives under my bed. None of it matters from a scientific perspective if I can't show you a reproducible phenomenon.
That's what the author was getting at when he says religion is "belief without evidence". But therein lies the fundamental problem of having this sort of discussion at all. The reason that the "laws of nature" aren't discussed much by scientists is that we don't really have any grounds on which to discuss them. They are observations that we have failed to find exceptions to, but we don't have any overarching model to explain *why* they are.
It has been stated on Slashdot many times before. Science is the domain of *how* or *what*. Philosophy and religion deal in *why*. These domains are complimentary, and there isn't really any overlap. When scientists try to explain *why* or religion tries to explain *how*, someone is going to get their panties in a bunch.
Now if only I could do that with all of my other computer components - my main desktop accumulates dust like no other. Maybe I'll throw the mobo in the dishwasher...
If you overstate your claims here, you still look like an idiot. But, if you don't speculate, the field doesn't advance. Soft science (like the article) has the same problem without the possibility of resolution via experiment - you've got to speculate about causation without overstating your conclusions.
If the OP had said 'Hey, I read the article, and those claims are way overstated. Correlation != Causation." I wouldn't have had any problem. Hell, I didn't even read the article. What I took umbrage to was reflexive "Correlation != Causation" without R'ing TFA. That is getting really old, and always gets moderated up - it is becoming a form of karma whoring.