I'm no fan of Java-based curricula, for the same reason I'd be no fan of Fortran-based curricula. Computing isn't about one language. Each language and system shows you one hyperplane of a vast multidimensional space. The best programmers know lots of languages, and choose wisely among them — or even create new ones when appropriate.
In the production world, there are times where some C++ or Java code is appropriate... and there are times when what you want is a couple of lines of shellscript and some pipes... and there are times when the most sensible algorithm for something can't be neatly expressed in a language like C++ or Java, and really requires something like Common Lisp or Haskell. If you need to exploit multiple processors without getting bogged down in locking bullshit and race conditions, you're much better off using Go than Java.
(Just last night, at a meetup, I was talking with two bright young physicists who reported that their universities don't do a good enough job of teaching Fortran, which is the language they actually need to do their job. Scientific computing still relies heavily on Fortran, Matlab, and other languages well removed from what's trendy in the CS department — no matter if that CS department is in the Java, Haskell, or Python camp. But if you want to learn to write good Fortran, you basically need a mentor in the physics department with time to teach you.)
And there are times when the right thing to do is to create a new language, whether a domain-specific language or a new approach on general-purpose computing. There's a good reason Rob Pike came up with Sawzall, a logs-analysis DSL that compiles to arbitrarily parallel mapreduces; and then Go, a C-like systems language with a rocket engine of concurrency built in.
(And there's a good reason a lot of people adopting Go have been coming not from the C++/Java camps that the Go developers expected, but from Python and Ruby: because Go gives you the raw speed of a concurrent and native-compiled language, plus libraries designed by actual engineers, without a lot of the verbose bullshit of C++ or Java. Would I recommend Go as a first language? I'm not so sure about that....)
What would an optimal computing curriculum look like? I have no freakin' clue. It would have to cover particular basics — variable binding, iteration, recursion, sequencing, data structures, libraries and APIs, concurrency — no matter what the language. But it can't leave its students thinking that one language is Intuitive and the other ones are Just Gratuitously Weird... and that's too much of what I see from young programmers in industry today.
A few years back, I considered uninstalling Flash, but there was Homestar Runner. Now, I'd consider uninstalling it, but there's the animated segments of Homestuck.
If I uninstall Flash, won't I just miss out on the next awesome cartoon whose name matches the regex/^Homest.*/ ?
Sure, I know and like DNSBLs including Spamhaus's, but this is a distinct application from XBL. Specifically, removal needs to be rapid in order for it to be useful for rejecting customer Web traffic. That's an engineering requirement that email anti-spam systems don't have, since SMTP is designed to retry for days if necessary to get a message through. Moreover, hosts that send any legitimate email are very few compared to hosts that send Web requests; and even though email admins are frequently dense, unresponsive, or victim-blaming, they're still a level above typical users in knowing what the fuck is going on with their computer.
One approach would be to have each DDoS victim continually (e.g. every hour) assert which addresses were attacking it, and only list those addresses which are currently attacking. This way, as soon as a host stops attacking, it will drop off the list. This has weaknesses — for instance, an attacker can use your host all night while you're not using it, without you noticing — but it's still an improvement over what we have today. And it still depends on each subscribing site having a good enough backchannel to the listing service to stay open during the DDoS. Back in the day we'd do it with a dedicated modem line — the bandwidth requirements are really quite minimal — but nobody knows what that is any more.
Sites under DoS attack should publish (through a channel not congested by the attack) a list of the IP addresses attacking them, through some trustworthy third party. Then, other sites should subscribe to that list and refuse service to those addresses until they clean up and stop attacking.
For instance, consider your uncle who uses AOL. His computer is infected with botnet garbage and is participating in a DoS attack against (say) Slashdot. Slashdot sends a list of attacking IPs, including your uncle's, to Team Cymru (the third party). Cymru aggregates these and publishes a list, updated every three hours. AOL subscribes to that list. When your uncle goes to check his AOL email, he gets an error: "We regret to inform you, your computer has been hacked, and is being used by criminals to break the Internet. You can't get to your AOL email until you kick the criminals off by installing an antivirus program and running a full scan. Click here to install Kaspersky Antivirus for free. Thank you for helping keep criminals from breaking everyone's Internet. Sincerely, Tim Armstrong, CEO, AOL."
Then your uncle gets mad and calls up AOL and complains. They try walking him through using the antivirus program, but he just curses them out and says he'll go to Hotmail instead. He tries... but Hotmail also subscribes to the same list and tells him the same thing: "Your computer is infected with malware and is being used to attack other sites on the Internet. You cannot obtain a Hotmail account until your computer is clean. Click here to install Microsoft Antivirus." He gives up and calls AOL back, and they help him get his computer cleaned up. Within half an hour, it's off the botnet; and within three hours, it's off the list of attacking hosts, and your uncle can get his AOL email again.
I have IPv6 through my ISP, Sonic.net. Whenever I use BitTorrent, I see plenty of IPv6 hosts. The reason is pretty obvious to me: if you're passing IPv6 through your home router, you have an externally-reachable IPv6 address... but you may not have an externally-reachable IPv4 address thanks to your home router's NAT.
Presumably, this means that one incentive for home users getting IPv6 is to get a better-connected BitTorrent network. BitTorrent is pretty popular, but ISPs are never going to tell you "Get IPv6 so you can download movies... er, I mean, Ubuntu Live CDs!... faster."
The easiest place to reach with the mouse is the current position. The second easiest is the four corners of the screen. The third easiest are the four sides of the screen. The hardest place is a square in the middle of the screen. Ancient UI guidelines are still relevant today.
Yep. This a corollary of Fitts's Law and while it's often associated with the design of the Macintosh menu bar, the underlying research dates to 1954, thirty years before the Mac.
Sadly, it hasn't been well learned on a lot of systems. Although Windows and Ubuntu both put a useful menu in a corner, few systems but the Mac make really effective use of the screen edge. Windows and many Linux desktops occupy much of one whole screen edge with a rarely used application switcher; but most users switch applications by pointing and clicking, or using keyboard shortcuts like Alt-Tab.
One big win that a lot of systems have benefited from, though, is contextual menus, which take advantage of the current position.
One problem is that as the size of a corporation increases, the influences on its behavior may become dominated by principal-agent problems and specific motivations of individuals within the corporation. It's easy for the members of a ten-person startup to keep "increasing shareholder value" in mind, but in a ten-thousand-person company, a middle manager or mid-ranking engineer may be much more interested in his or her next quarterly review or promotion.
Furthermore, the internal economy of a large corporation is a command economy, not a free market. In a free market, decision makers can count on prices to show them which goods are the most efficient choices, or which products may be the most lucrative. But within a large corporation, management is expected to know how best to apportion budgets, wages, investments, etc. â" all without the benefit of a pricing mechanism that accurately reflects (internal) needs. And as the corporation gets larger and more heavily capitalized, it becomes more and more different from the outside world, so external signals (such as the prevailing wages in the industry) become less relevant to internal decision-making.
I saw it about a week ago. Overall, my biggest impression was one of missed potential.
(Note, here I'm talking primarily about the story and the world-building, not about the cinematography.)
The overall structure was a weakness from the start. Sam Flynn turns out to be yet anotherPrince Harry character: the heir to the throne who goofs around and avoids his inherited position until he's handed a confrontation that forces him to prove himself, at which point he rises to the occasion as a True Prince. We've seen this before; it's the usual aristocratic nonsense: worth is not achieved, but inherited and then revealed.
Contrast the original: Kevin Flynn was an honest working hacker who was forced to go rogue when he was screwed over by a yuppie coworker. Kevin's triumph was to prove himself as a creator. He set out with the aim of showing that he and not Ed Dillinger was the author of Space Paranoids; and in the end, he accomplished that goal, but in a way that -- through his creative "User power" -- changed the Programs' world for the better.
Sam isn't a creator. He sets out with no particular goals of his own; he is handed all his goals by his inheritance. Kevin Flynn was a creative adult seeking justice; Sam Flynn is an irresponsible rich boy growing up. And that's a story that's been played out far too many times.
One of Legacy's few big world-building ideas is the emergence of the Isos: Programs evolved from the System itself, rather than being created in the image of a User. This could have been huge. But instead it is presented merely to give Sam's love interest a tragic backstory. The war is over; the Isos lost, here's the last surviving princess of a dead race. Give her a hug.
The political vision of the System in Tron is more complex. There are old powers in the System that defy the MCP's regime at personal risk to themselves: Dumont at the I/O Tower. The MCP's assimilation of the whole System into itself is not complete; it can be resisted. In Legacy, CLU's genocide of the Isos is over and done with... and nobody even bothers to say, "Sam, you dickhead, if you'd logged in yesterday, you could have stopped the fucking Holocaust."
Another new world-building idea is the possibility that a Program could use the laser terminal to escape into the real world: that the laser wasn't limited to objects that originated in the real world (oranges or Kevin Flynns), but could also play back a Program into human form. Thus Quorra's escape; thus CLU's threat to invade our world with armies of Programs.
Well, Tron's MCP didn't need armies to take over the world. The MCP could just hack the Pentagon. In Tron, the deep entanglement of the real world and the System is made clear: the MCP can threaten Dillinger not with armies materializing in ENCOM's laser bay, but with the legal and political forces native to our world.
Ironically enough, the 1982 vision has more in common with today's Internet-enabled reality than the 2010 version. As far as we know, the System in Legacy isn't even on the Net: it's a dusty minicomputer sitting in the basement of Flynn's Arcade with barely enough connectivity to reach Alan Bradley's pager.
Ultimately, CLU is much less of a real-world threat than the MCP. The MCP had taken over the System that ENCOM used to do its business, and was extending tentacles into banks, major governments, and who knows what else. CLU's domain is that one minicomputer; the big threat would be shut off if Alan or Sam had just unplugged the laser terminal.
Both of the above two problems point at a bigger problem with Legacy: it ultimately doesn't take Programs and the System seriously as an independent sort of intelligent existence rather than a mere imitation of our wor
Your digital camera may embed metadata into photographs with the camera's serial number or your location.
Record your location? Sure, if it's a smartphone with GPS. For standalone cameras, GPS is not exactly a common feature. There are about two models of pocket digital camera on the market that have GPS, and not very many SLRs with it either... go look. Those that have it make no secret of it; it's actually a big marketing point for people who want to record where they've been taking pictures.
As for smartphone models, I don't know about the Apple or Windows offerings, but Android's camera app exposes it as an option right on the main screen, next to the flash and focus settings... and I'm pretty sure it defaults to off. People turn this on because they actively want it.
Rather than scaring people about what their devices might be recording, it would be a lot more useful to tell people how to find out what tags are on their photos. For instance, the Linux command line program "exiftags" will tell you this kind of stuff: (Picked from a random image file I had lying around on my laptop.)
Camera-Specific Properties:
Equipment Make: OLYMPUS OPTICAL CO.,LTD Camera Model: C2500L Camera Software: Adobe Photoshop CS Macintosh Maximum Lens Aperture: f/2.6
Image-Specific Properties:
Image Orientation: Top, Left-Hand Horizontal Resolution: 173 dpi Vertical Resolution: 173 dpi Image Created: 2004:02:27 18:52:21 Exposure Time: 1/5 sec F-Number: f/6.9 Exposure Program: Manual ISO Speed Rating: 100 Exposure Bias: 0 EV Metering Mode: Center Weighted Average Flash: No Flash Focal Length: 20.70 mm Color Space Information: Uncalibrated Image Width: 736 Image Height: 767
If your language actually uses the character U+23C7 "DENTISTRY SYMBOL LIGHT DOWN AND HORIZONTAL WITH WAVE" as an operator, your editor will let you type it with a simple keyboard combination, like Compose-T-~. If you're using U.S. Windows and have to resort to Alt+numbers to type things, you're silly.
The problem isn't that math isn't important. The problem is that the math being taught isn't important.
Yes. Exactly.
Fuck calculus. You don't need it unless you're going into one of a few specific fields. But there are whole swaths of math that most folks completely miss, that are directly applicable to everyday life:
Probability and statistics. No, not for understanding the census, nor for gambling -- rather, for understanding what's meant by words like "evidence". Bayesian probability can be taught to anyone who can understand percentages and division, and it can be straightforwardly applied to reasoning about the everyday world.
Proof and logic. The notion of logical proof has been around since Aristotle, but symbolic logic is much newer. Nonetheless, the notion of logical validity of an argument, of conclusions following from premises, is directly applicable to all sorts of real-world decision-making. Logic is also an obvious point to dovetail math into the humanities, via the analysis of written arguments.
Abstract algebra. Not the proofs, nor the deep abstractions, but rather the notions of properties such as commutativity, associativity, etc. and the idea that these can be applied to any sorts of operations, not just "mathematical" ones. Does it matter if you mix the eggs in before the butter? Do you need to do X separately to A, B, and C, or can you put A+B+C together and then do X all at once? The notion that some situations or problems have the same structure as others is itself pretty powerful. (And lends itself to comparison with the literary idea of analogy.)
... a small one. Here's what our policy to prevent piracy would have been:
Please don't pirate stuff too much. If we get notices saying that you're pirating stuff and asking you to quit, we'll call you in to the office and give them to you. If we get court orders telling us to give them your name, we'll probably have to do that, since we can't afford lawyers much.
If you really have to pirate stuff, please at least try to leech it off of your friends on the LAN rather than flooding our dinky little Internet uplink. Because if you do that, we'll probably end up blocking your IP address for a while so that email and our Debian updates can get in.
And while you're at it, here's the address of the porn server that some freshman set up. Get your porn over there, please don't mirror all of abbywinters.com over our connection.
Check out the huge regex at the bottom of the RFC 5322 compliant validator from CPAN:
Honestly, this sort of thing is an example of overusing regex when it's the only parsing tool they know. Regex becomes unwieldy when you put too much of it in one place -- but this is because regex is unwieldy, not because the problem of parsing email addresses is fundamentally hard. Parsing email addresses is a case for a modular parser such as Parsec (or any of its ports and imitators)... which will give you the added advantage of useful error messages on invalid input, instead of just a match failure.
Moreover, isn't it kind of silly to point at an example of someone already having written the code to do something as a way of saying that doing it is difficult? In code, once it's already been done once, correctly, it doesn't need to be done again. If you think CPAN's huge regex (or any other implementation) is correct, and you've tested it to your satisfaction, you don't need to reimplement it; just use it.
For folks who want to record the cops (or anyone else) and be sure that the footage will get to the world instead of being destroyed when they steal your camera phone: check out the Gandhicam project. This is an app for your Android phone that lets you take pictures or video and automatically send it to the net, either by HTTP upload or by email.
This doesn't stop them from filing criminal charges afterward, but that's why you donate to the ACLU and the EFF.
Re:Seven years for eight hours work
on
Novell Wins vs. SCO
·
· Score: 2, Interesting
Really, the only way that SCO was going to recover was with a court victory, and while the probability of that wasn't 0, it was as damn near to it as possible for practical applications.
There are people who believe things out of spite. Remember when the SCO case got started? There were plenty of folks -- chiefly in the "open-source haters" end of the trade press, but I met a few in industry, too -- who dearly wanted to see the "upstart" Linux smacked down hard.
It may be hard to believe it now that Linux is everywhere in the industry -- from the datacenter to the cell phone, from the Oracle database server to the displays on the backs of airplane seats -- but just a few years ago, plenty of people would have called you an "open-source zealot" if you said that it was worth using anywhere at all in business. And lots of traditional business people really wanted to see Linux dry up and blow away. Plenty of those people would have put hope, and a few bucks, behind the SCO suit.
Having studied eight foreign languages (French, Spanish, German, Latin, Ancient Greek, Russian, Japanese, and Finnish) in my life, and after talking this theory over with friends who have attained fluency in some really different languages (e.g. Spanish and Bahasa Melayu), I feel safe in stating this here in pretty strong terms:
The only way to learn a language is to use it.
The only sort of "classroom" language class that works worth a damn is an immersion class, in which during the class period you do not speak any language other than the one you're studying. Even classroom instructions ("Open your book to page 23") are in the language, once you've learned numerals.
The worst language classes I've taken have been ones in which the foreign language being studied is treated as a matter of abstract grammar and vocabulary to be memorized, not used... and in which the teacher spends most of their time telling anecdotes in English about their experiences in the culture in question. I took two years of Russian in high school and a year of it in college -- and forgot more Russian than I learned in that last year, since the teacher spent the class time telling stories (in English!) about run-ins with the KGB, instead of helping us practice speaking and reading Russian.
As regards Chinese: I've never studied Chinese, but I have studied Japanese including kanji, albeit only to the extent of a couple hundred kanji. The above applies fully to kanji, and I expect it applies to hanzi (Chinese characters) as well -- in order to learn them, you have to use them. Write them. Come up with silly sentences and write those. Don't just use flash cards and memorization; come up with things that you want to say in Chinese (even if just to be silly) and say those things with hanzi.
The other half of the equation, of course, is to get someone who is fluent to respond to your crude childish attempts at speaking and writing. That's the point of a good language class: you get to make the sort of errors that a little kid makes, and they correct you. That method of language acquisition works for little kids, and it works for adults too if they're willing to be childish for a while.
The point of the article is not to accuse Microsoft of wrongdoing. It's to use a highly visible example -- a screen that will be presented to every Windows user in Europe! -- to teach a computer science lesson to programmers.
Unfortunately there are a few cases you can't do that. No way to use a prepared statement for an "IN" clause, for instance.
You could always use the technique the DBMS gives you for encapsulating complex query logic and passing parameters into it...
... stored procedures!
Cue screams of terror; cue fist-waving and incoherent irate noises. These will be heard from people who never cared to actually learn how their DBMS works, or who think of the DBMS as some foreign thing that their application throws data at once in a while.
(Your application does not use a DBMS. Rather, a DBMS is one of your application's components. If your app contains a DBMS but you don't know what it's doing, you are in the same camp as "programmers" who do not know how their language's object model works, or who are a little rusty on whether a loop with a false loop condition executes zero times or once. That is: you are not a programmer; you are a burger-flipper who sometimes farts out bad code.)
Stored procedures are not the chapter of your DBMS manual that was put in there to pad it out so it qualified for the better binding at the publisher. Nor were they put in there just to entertain your DBA. Stored procedures let you simplify the interface between your DBMS and the rest of your application, and this can radically improve your overall security.
After Google's CEO's comments about privacy is only wanted by wrongdoers
Except, of course, that he never said that. He was asked in an interview whether users should consider Google as a "trusted friend" -- and he said no. He said that if you're doing something that you don't want anyone to know about, doing it on Google is a bad idea... since Google is just as subject to U.S. law, including the USA PATRIOT Act, as any other company is.
He didn't say that only wrongdoers want privacy and that everyone should trust Google. He said that if you want perfect privacy, you can't get it from Google, because the law doesn't allow it. That's pretty much the opposite!
People spend a lot of effort blathering about which programming language is best to use for teaching. But the hard part of programming is not the programming language. It's the logical thinking skills; the abstract concepts like function and algorithm and data structure and type; the reasoned approach to breaking a problem down and seeing algorithms and patterns; the ability to learn new tools such as a utility or an API and put them together usefully.
These things transcend language. Yes, you will probably use different algorithms or data structures in Python on a Linux box than in C on a microcontroller, but you will use largely the same sort of thinking skills. You will approach writing code differently in Lisp than in Java, but in both you will be combining known parts in a new structure to accomplish a task.
And it is these abstract skills -- especially the skill of abstracting, of recognizing and using patterns -- which separate those who learn to program well from those who do not. (And this is different again from being a successful professional programmer, which entails a quite different set of skills.)
A monad is not a specific data structure (like an array) or a language feature (like a loop or conditional). It is not something that is put together by the compiler (like a compiled function) and it is not magic, either.
"Being a monad" is a mathematical property, like "being commutative" or "being symmetric". It has to do with values following a particular pattern, having operations defined on them that interact in a particular way. (These are called "monad laws", and they're the same sort of thing as the "identities" you get in high school algebra.)
Any sort of thing that follows the monad laws is a monad. In Haskell, we say that any type that follows the monad laws can be made a member of typeclassMonad. Casually, we say that "lists are a monad" and "IO actions are a monad".
What does it mean to be a monad? Roughly, it means that you can chain operations together. You can map a function over a list, or filter the list's elements to get a new list, or more generally "fold" a list by recursively applying an operation to collect a result. (Like the factorial function repeatedly applies multiplication.) Also you can join a list of lists together into a single big list.
It turns out that monads can represent the idea of "sequence" quite well. Lists are, after all, a sort of sequence. But so are IO actions -- all those "non-functional" things a program has to do, like read a file from disk, or write its contents to the terminal. IO actions form a sequence in time: if you want to write the file's contents to the terminal, you have to read them from the disk first.
The way that this works, under the hood, is to express the "later" actions as having the "earlier" actions' results as function arguments. Since a function's arguments have to be evaluated before the function can be applied, the disk gets read before the terminal-writing happens.
So monads are not a loophole that lets you cheat and do non-functional things inside a functional language. Rather, they are a model that lets you describe IO actions as just another sort of functional value. The only bit of "magic" necessary is that the main function of a program has the IO type.
most adults do not really believe the tenets of any major religion anyway.
Of course not. The question is, do most adults believe some of the tenets?
That depends profoundly on what you mean. Most religious people, most of the time, do not permit their religion to get in the way of their common sense or their common decency. When questioned on a point of religious belief, they will reliably respond according to their understanding of doctrine... but with equal or greater reliability, when presented with a practical challenge in life, they will respond based on an ordinary secular understanding of how the world works.
An example: Martyrdom. We know from the lives of the saints, and from the acts of certain modern-day counterparts, that there are those who put their faith in God ahead of their self-preservation, placing themselves in harm's way in the service of faith. We know, also, that only a tiny minority of religionists do this... even among those who claim to revere the saints and martyrs who do, and who when questioned on the matter state clearly their belief in heaven. In gist, people who claim belief in life-after-death act just as fervently to avoid death as those who do not claim that belief.
This is what people mean when they question whether religionists really do believe what they say they do. Most of the time, when we say "belief", we mean something that we are willing to rely upon: if you believe it is cold, you do not wear shorts; if you believe that Jane does not love you, you do not propose marriage to her. Your belief is detectable in your actions which rely upon it, which are explained by it.
Dennett presents the notion of "believing in belief" -- the idea that many people think that religious belief is a good thing, a thing to be desired, indeed one which you should pretend to, or "fake it until you make it". People who believe in belief can be expected to say they believe religious claims, but not to actually rely on them. This seems to be a cogent explanation of the way that most "religious" people actually deal with the real world.
I think they would have to define a man as someone that is XY and a woman as someone that is XX, anything else doesn't qualify or gets removed based on medical grounds.
Go read about androgen insensitivity syndrome. A person with complete androgen insensitivity has the external appearance (including external genitals) of a normal female, and usually has been raised with a female identity, and so identifies as a girl or woman. But she has no menstrual period because she has no ovaries or uterus -- she has an XY genotype and undescended internal testes. Androgens (male hormones) create the male external appearance; in the absence of androgens, a fetus develops a female appearance -- and parents and society assign a female identity.
It seems to me that a reasonable approach to dealing fairly with intersexed people in sports would be to replace sex separation with weight-class separation. Human sex and gender turn out to be complex and fraught with both social and technical problems, and it is unreasonable to deny an athlete the chance to compete just because their genetic curiosity happens to be on the 23rd chromosome.
Slashdot Mirror
I'm no fan of Java-based curricula, for the same reason I'd be no fan of Fortran-based curricula. Computing isn't about one language. Each language and system shows you one hyperplane of a vast multidimensional space. The best programmers know lots of languages, and choose wisely among them — or even create new ones when appropriate.
In the production world, there are times where some C++ or Java code is appropriate ... and there are times when what you want is a couple of lines of shellscript and some pipes ... and there are times when the most sensible algorithm for something can't be neatly expressed in a language like C++ or Java, and really requires something like Common Lisp or Haskell. If you need to exploit multiple processors without getting bogged down in locking bullshit and race conditions, you're much better off using Go than Java.
(Just last night, at a meetup, I was talking with two bright young physicists who reported that their universities don't do a good enough job of teaching Fortran, which is the language they actually need to do their job. Scientific computing still relies heavily on Fortran, Matlab, and other languages well removed from what's trendy in the CS department — no matter if that CS department is in the Java, Haskell, or Python camp. But if you want to learn to write good Fortran, you basically need a mentor in the physics department with time to teach you.)
And there are times when the right thing to do is to create a new language, whether a domain-specific language or a new approach on general-purpose computing. There's a good reason Rob Pike came up with Sawzall, a logs-analysis DSL that compiles to arbitrarily parallel mapreduces; and then Go, a C-like systems language with a rocket engine of concurrency built in.
(And there's a good reason a lot of people adopting Go have been coming not from the C++/Java camps that the Go developers expected, but from Python and Ruby: because Go gives you the raw speed of a concurrent and native-compiled language, plus libraries designed by actual engineers, without a lot of the verbose bullshit of C++ or Java. Would I recommend Go as a first language? I'm not so sure about that ....)
What would an optimal computing curriculum look like? I have no freakin' clue. It would have to cover particular basics — variable binding, iteration, recursion, sequencing, data structures, libraries and APIs, concurrency — no matter what the language. But it can't leave its students thinking that one language is Intuitive and the other ones are Just Gratuitously Weird ... and that's too much of what I see from young programmers in industry today.
You never use a search engine while writing? They're awfully handy for fact-checking, looking up sources, and so on.
But I suppose those sorts of activities are not required these days ....
A few years back, I considered uninstalling Flash, but there was Homestar Runner. Now, I'd consider uninstalling it, but there's the animated segments of Homestuck.
If I uninstall Flash, won't I just miss out on the next awesome cartoon whose name matches the regex /^Homest.*/ ?
Sure, I know and like DNSBLs including Spamhaus's, but this is a distinct application from XBL. Specifically, removal needs to be rapid in order for it to be useful for rejecting customer Web traffic. That's an engineering requirement that email anti-spam systems don't have, since SMTP is designed to retry for days if necessary to get a message through. Moreover, hosts that send any legitimate email are very few compared to hosts that send Web requests; and even though email admins are frequently dense, unresponsive, or victim-blaming, they're still a level above typical users in knowing what the fuck is going on with their computer.
One approach would be to have each DDoS victim continually (e.g. every hour) assert which addresses were attacking it, and only list those addresses which are currently attacking. This way, as soon as a host stops attacking, it will drop off the list. This has weaknesses — for instance, an attacker can use your host all night while you're not using it, without you noticing — but it's still an improvement over what we have today. And it still depends on each subscribing site having a good enough backchannel to the listing service to stay open during the DDoS. Back in the day we'd do it with a dedicated modem line — the bandwidth requirements are really quite minimal — but nobody knows what that is any more.
Sites under DoS attack should publish (through a channel not congested by the attack) a list of the IP addresses attacking them, through some trustworthy third party. Then, other sites should subscribe to that list and refuse service to those addresses until they clean up and stop attacking.
For instance, consider your uncle who uses AOL. His computer is infected with botnet garbage and is participating in a DoS attack against (say) Slashdot. Slashdot sends a list of attacking IPs, including your uncle's, to Team Cymru (the third party). Cymru aggregates these and publishes a list, updated every three hours. AOL subscribes to that list. When your uncle goes to check his AOL email, he gets an error: "We regret to inform you, your computer has been hacked, and is being used by criminals to break the Internet. You can't get to your AOL email until you kick the criminals off by installing an antivirus program and running a full scan. Click here to install Kaspersky Antivirus for free. Thank you for helping keep criminals from breaking everyone's Internet. Sincerely, Tim Armstrong, CEO, AOL."
Then your uncle gets mad and calls up AOL and complains. They try walking him through using the antivirus program, but he just curses them out and says he'll go to Hotmail instead. He tries ... but Hotmail also subscribes to the same list and tells him the same thing: "Your computer is infected with malware and is being used to attack other sites on the Internet. You cannot obtain a Hotmail account until your computer is clean. Click here to install Microsoft Antivirus." He gives up and calls AOL back, and they help him get his computer cleaned up. Within half an hour, it's off the botnet; and within three hours, it's off the list of attacking hosts, and your uncle can get his AOL email again.
Well, I suppose Daikatana will be coming out for iOS but not for Android, then?
I have IPv6 through my ISP, Sonic.net. Whenever I use BitTorrent, I see plenty of IPv6 hosts. The reason is pretty obvious to me: if you're passing IPv6 through your home router, you have an externally-reachable IPv6 address ... but you may not have an externally-reachable IPv4 address thanks to your home router's NAT.
Presumably, this means that one incentive for home users getting IPv6 is to get a better-connected BitTorrent network. BitTorrent is pretty popular, but ISPs are never going to tell you "Get IPv6 so you can download movies ... er, I mean, Ubuntu Live CDs! ... faster."
Yep. This a corollary of Fitts's Law and while it's often associated with the design of the Macintosh menu bar, the underlying research dates to 1954, thirty years before the Mac.
Sadly, it hasn't been well learned on a lot of systems. Although Windows and Ubuntu both put a useful menu in a corner, few systems but the Mac make really effective use of the screen edge. Windows and many Linux desktops occupy much of one whole screen edge with a rarely used application switcher; but most users switch applications by pointing and clicking, or using keyboard shortcuts like Alt-Tab.
One big win that a lot of systems have benefited from, though, is contextual menus, which take advantage of the current position.
Furthermore, the internal economy of a large corporation is a command economy, not a free market. In a free market, decision makers can count on prices to show them which goods are the most efficient choices, or which products may be the most lucrative. But within a large corporation, management is expected to know how best to apportion budgets, wages, investments, etc. â" all without the benefit of a pricing mechanism that accurately reflects (internal) needs. And as the corporation gets larger and more heavily capitalized, it becomes more and more different from the outside world, so external signals (such as the prevailing wages in the industry) become less relevant to internal decision-making.
I saw it about a week ago. Overall, my biggest impression was one of missed potential.
(Note, here I'm talking primarily about the story and the world-building, not about the cinematography.)
The overall structure was a weakness from the start. Sam Flynn turns out to be yet another Prince Harry character: the heir to the throne who goofs around and avoids his inherited position until he's handed a confrontation that forces him to prove himself, at which point he rises to the occasion as a True Prince. We've seen this before; it's the usual aristocratic nonsense: worth is not achieved, but inherited and then revealed.
Contrast the original: Kevin Flynn was an honest working hacker who was forced to go rogue when he was screwed over by a yuppie coworker. Kevin's triumph was to prove himself as a creator. He set out with the aim of showing that he and not Ed Dillinger was the author of Space Paranoids; and in the end, he accomplished that goal, but in a way that -- through his creative "User power" -- changed the Programs' world for the better.
Sam isn't a creator. He sets out with no particular goals of his own; he is handed all his goals by his inheritance. Kevin Flynn was a creative adult seeking justice; Sam Flynn is an irresponsible rich boy growing up. And that's a story that's been played out far too many times.
One of Legacy's few big world-building ideas is the emergence of the Isos: Programs evolved from the System itself, rather than being created in the image of a User. This could have been huge. But instead it is presented merely to give Sam's love interest a tragic backstory. The war is over; the Isos lost, here's the last surviving princess of a dead race. Give her a hug. ... and nobody even bothers to say, "Sam, you dickhead, if you'd logged in yesterday, you could have stopped the fucking Holocaust."
The political vision of the System in Tron is more complex. There are old powers in the System that defy the MCP's regime at personal risk to themselves: Dumont at the I/O Tower. The MCP's assimilation of the whole System into itself is not complete; it can be resisted. In Legacy, CLU's genocide of the Isos is over and done with
Another new world-building idea is the possibility that a Program could use the laser terminal to escape into the real world: that the laser wasn't limited to objects that originated in the real world (oranges or Kevin Flynns), but could also play back a Program into human form. Thus Quorra's escape; thus CLU's threat to invade our world with armies of Programs.
Well, Tron's MCP didn't need armies to take over the world. The MCP could just hack the Pentagon. In Tron, the deep entanglement of the real world and the System is made clear: the MCP can threaten Dillinger not with armies materializing in ENCOM's laser bay, but with the legal and political forces native to our world.
Ironically enough, the 1982 vision has more in common with today's Internet-enabled reality than the 2010 version. As far as we know, the System in Legacy isn't even on the Net: it's a dusty minicomputer sitting in the basement of Flynn's Arcade with barely enough connectivity to reach Alan Bradley's pager.
Ultimately, CLU is much less of a real-world threat than the MCP. The MCP had taken over the System that ENCOM used to do its business, and was extending tentacles into banks, major governments, and who knows what else. CLU's domain is that one minicomputer; the big threat would be shut off if Alan or Sam had just unplugged the laser terminal.
Both of the above two problems point at a bigger problem with Legacy: it ultimately doesn't take Programs and the System seriously as an independent sort of intelligent existence rather than a mere imitation of our wor
Record your location? Sure, if it's a smartphone with GPS. For standalone cameras, GPS is not exactly a common feature. There are about two models of pocket digital camera on the market that have GPS, and not very many SLRs with it either ... go look. Those that have it make no secret of it; it's actually a big marketing point for people who want to record where they've been taking pictures.
As for smartphone models, I don't know about the Apple or Windows offerings, but Android's camera app exposes it as an option right on the main screen, next to the flash and focus settings ... and I'm pretty sure it defaults to off. People turn this on because they actively want it.
Rather than scaring people about what their devices might be recording, it would be a lot more useful to tell people how to find out what tags are on their photos. For instance, the Linux command line program "exiftags" will tell you this kind of stuff: (Picked from a random image file I had lying around on my laptop.)
If your language actually uses the character U+23C7 "DENTISTRY SYMBOL LIGHT DOWN AND HORIZONTAL WITH WAVE" as an operator, your editor will let you type it with a simple keyboard combination, like Compose-T-~. If you're using U.S. Windows and have to resort to Alt+numbers to type things, you're silly.
Yes. Exactly.
Fuck calculus. You don't need it unless you're going into one of a few specific fields. But there are whole swaths of math that most folks completely miss, that are directly applicable to everyday life:
Probability and statistics. No, not for understanding the census, nor for gambling -- rather, for understanding what's meant by words like "evidence". Bayesian probability can be taught to anyone who can understand percentages and division, and it can be straightforwardly applied to reasoning about the everyday world.
Proof and logic. The notion of logical proof has been around since Aristotle, but symbolic logic is much newer. Nonetheless, the notion of logical validity of an argument, of conclusions following from premises, is directly applicable to all sorts of real-world decision-making. Logic is also an obvious point to dovetail math into the humanities, via the analysis of written arguments.
Abstract algebra. Not the proofs, nor the deep abstractions, but rather the notions of properties such as commutativity, associativity, etc. and the idea that these can be applied to any sorts of operations, not just "mathematical" ones. Does it matter if you mix the eggs in before the butter? Do you need to do X separately to A, B, and C, or can you put A+B+C together and then do X all at once? The notion that some situations or problems have the same structure as others is itself pretty powerful. (And lends itself to comparison with the literary idea of analogy.)
... a small one. Here's what our policy to prevent piracy would have been:
Please don't pirate stuff too much. If we get notices saying that you're pirating stuff and asking you to quit, we'll call you in to the office and give them to you. If we get court orders telling us to give them your name, we'll probably have to do that, since we can't afford lawyers much.
If you really have to pirate stuff, please at least try to leech it off of your friends on the LAN rather than flooding our dinky little Internet uplink. Because if you do that, we'll probably end up blocking your IP address for a while so that email and our Debian updates can get in.
And while you're at it, here's the address of the porn server that some freshman set up. Get your porn over there, please don't mirror all of abbywinters.com over our connection.
Honestly, this sort of thing is an example of overusing regex when it's the only parsing tool they know. Regex becomes unwieldy when you put too much of it in one place -- but this is because regex is unwieldy, not because the problem of parsing email addresses is fundamentally hard. Parsing email addresses is a case for a modular parser such as Parsec (or any of its ports and imitators) ... which will give you the added advantage of useful error messages on invalid input, instead of just a match failure.
Moreover, isn't it kind of silly to point at an example of someone already having written the code to do something as a way of saying that doing it is difficult? In code, once it's already been done once, correctly, it doesn't need to be done again. If you think CPAN's huge regex (or any other implementation) is correct, and you've tested it to your satisfaction, you don't need to reimplement it; just use it.
For folks who want to record the cops (or anyone else) and be sure that the footage will get to the world instead of being destroyed when they steal your camera phone: check out the Gandhicam project. This is an app for your Android phone that lets you take pictures or video and automatically send it to the net, either by HTTP upload or by email.
This doesn't stop them from filing criminal charges afterward, but that's why you donate to the ACLU and the EFF.
There are people who believe things out of spite. Remember when the SCO case got started? There were plenty of folks -- chiefly in the "open-source haters" end of the trade press, but I met a few in industry, too -- who dearly wanted to see the "upstart" Linux smacked down hard.
It may be hard to believe it now that Linux is everywhere in the industry -- from the datacenter to the cell phone, from the Oracle database server to the displays on the backs of airplane seats -- but just a few years ago, plenty of people would have called you an "open-source zealot" if you said that it was worth using anywhere at all in business. And lots of traditional business people really wanted to see Linux dry up and blow away. Plenty of those people would have put hope, and a few bucks, behind the SCO suit.
Having studied eight foreign languages (French, Spanish, German, Latin, Ancient Greek, Russian, Japanese, and Finnish) in my life, and after talking this theory over with friends who have attained fluency in some really different languages (e.g. Spanish and Bahasa Melayu), I feel safe in stating this here in pretty strong terms:
The only way to learn a language is to use it.
The only sort of "classroom" language class that works worth a damn is an immersion class, in which during the class period you do not speak any language other than the one you're studying. Even classroom instructions ("Open your book to page 23") are in the language, once you've learned numerals.
The worst language classes I've taken have been ones in which the foreign language being studied is treated as a matter of abstract grammar and vocabulary to be memorized, not used ... and in which the teacher spends most of their time telling anecdotes in English about their experiences in the culture in question. I took two years of Russian in high school and a year of it in college -- and forgot more Russian than I learned in that last year, since the teacher spent the class time telling stories (in English!) about run-ins with the KGB, instead of helping us practice speaking and reading Russian.
As regards Chinese: I've never studied Chinese, but I have studied Japanese including kanji, albeit only to the extent of a couple hundred kanji. The above applies fully to kanji, and I expect it applies to hanzi (Chinese characters) as well -- in order to learn them, you have to use them. Write them. Come up with silly sentences and write those. Don't just use flash cards and memorization; come up with things that you want to say in Chinese (even if just to be silly) and say those things with hanzi.
The other half of the equation, of course, is to get someone who is fluent to respond to your crude childish attempts at speaking and writing. That's the point of a good language class: you get to make the sort of errors that a little kid makes, and they correct you. That method of language acquisition works for little kids, and it works for adults too if they're willing to be childish for a while.
The point of the article is not to accuse Microsoft of wrongdoing. It's to use a highly visible example -- a screen that will be presented to every Windows user in Europe! -- to teach a computer science lesson to programmers.
You could always use the technique the DBMS gives you for encapsulating complex query logic and passing parameters into it ...
... stored procedures!
Cue screams of terror; cue fist-waving and incoherent irate noises. These will be heard from people who never cared to actually learn how their DBMS works, or who think of the DBMS as some foreign thing that their application throws data at once in a while.
(Your application does not use a DBMS. Rather, a DBMS is one of your application's components. If your app contains a DBMS but you don't know what it's doing, you are in the same camp as "programmers" who do not know how their language's object model works, or who are a little rusty on whether a loop with a false loop condition executes zero times or once. That is: you are not a programmer; you are a burger-flipper who sometimes farts out bad code.)
Stored procedures are not the chapter of your DBMS manual that was put in there to pad it out so it qualified for the better binding at the publisher. Nor were they put in there just to entertain your DBA. Stored procedures let you simplify the interface between your DBMS and the rest of your application, and this can radically improve your overall security.
Except, of course, that he never said that. He was asked in an interview whether users should consider Google as a "trusted friend" -- and he said no. He said that if you're doing something that you don't want anyone to know about, doing it on Google is a bad idea ... since Google is just as subject to U.S. law, including the USA PATRIOT Act, as any other company is.
He didn't say that only wrongdoers want privacy and that everyone should trust Google. He said that if you want perfect privacy, you can't get it from Google, because the law doesn't allow it. That's pretty much the opposite!
People spend a lot of effort blathering about which programming language is best to use for teaching. But the hard part of programming is not the programming language. It's the logical thinking skills; the abstract concepts like function and algorithm and data structure and type; the reasoned approach to breaking a problem down and seeing algorithms and patterns; the ability to learn new tools such as a utility or an API and put them together usefully.
These things transcend language. Yes, you will probably use different algorithms or data structures in Python on a Linux box than in C on a microcontroller, but you will use largely the same sort of thinking skills. You will approach writing code differently in Lisp than in Java, but in both you will be combining known parts in a new structure to accomplish a task.
And it is these abstract skills -- especially the skill of abstracting, of recognizing and using patterns -- which separate those who learn to program well from those who do not. (And this is different again from being a successful professional programmer, which entails a quite different set of skills.)
A monad is not a specific data structure (like an array) or a language feature (like a loop or conditional). It is not something that is put together by the compiler (like a compiled function) and it is not magic, either.
"Being a monad" is a mathematical property, like "being commutative" or "being symmetric". It has to do with values following a particular pattern, having operations defined on them that interact in a particular way. (These are called "monad laws", and they're the same sort of thing as the "identities" you get in high school algebra.)
Any sort of thing that follows the monad laws is a monad. In Haskell, we say that any type that follows the monad laws can be made a member of typeclass Monad. Casually, we say that "lists are a monad" and "IO actions are a monad".
What does it mean to be a monad? Roughly, it means that you can chain operations together. You can map a function over a list, or filter the list's elements to get a new list, or more generally "fold" a list by recursively applying an operation to collect a result. (Like the factorial function repeatedly applies multiplication.) Also you can join a list of lists together into a single big list.
It turns out that monads can represent the idea of "sequence" quite well. Lists are, after all, a sort of sequence. But so are IO actions -- all those "non-functional" things a program has to do, like read a file from disk, or write its contents to the terminal. IO actions form a sequence in time: if you want to write the file's contents to the terminal, you have to read them from the disk first.
The way that this works, under the hood, is to express the "later" actions as having the "earlier" actions' results as function arguments. Since a function's arguments have to be evaluated before the function can be applied, the disk gets read before the terminal-writing happens.
So monads are not a loophole that lets you cheat and do non-functional things inside a functional language. Rather, they are a model that lets you describe IO actions as just another sort of functional value. The only bit of "magic" necessary is that the main function of a program has the IO type.
That depends profoundly on what you mean. Most religious people, most of the time, do not permit their religion to get in the way of their common sense or their common decency. When questioned on a point of religious belief, they will reliably respond according to their understanding of doctrine ... but with equal or greater reliability, when presented with a practical challenge in life, they will respond based on an ordinary secular understanding of how the world works.
An example: Martyrdom. We know from the lives of the saints, and from the acts of certain modern-day counterparts, that there are those who put their faith in God ahead of their self-preservation, placing themselves in harm's way in the service of faith. We know, also, that only a tiny minority of religionists do this ... even among those who claim to revere the saints and martyrs who do, and who when questioned on the matter state clearly their belief in heaven. In gist, people who claim belief in life-after-death act just as fervently to avoid death as those who do not claim that belief.
This is what people mean when they question whether religionists really do believe what they say they do. Most of the time, when we say "belief", we mean something that we are willing to rely upon: if you believe it is cold, you do not wear shorts; if you believe that Jane does not love you, you do not propose marriage to her. Your belief is detectable in your actions which rely upon it, which are explained by it.
Dennett presents the notion of "believing in belief" -- the idea that many people think that religious belief is a good thing, a thing to be desired, indeed one which you should pretend to, or "fake it until you make it". People who believe in belief can be expected to say they believe religious claims, but not to actually rely on them. This seems to be a cogent explanation of the way that most "religious" people actually deal with the real world.
Go read about androgen insensitivity syndrome. A person with complete androgen insensitivity has the external appearance (including external genitals) of a normal female, and usually has been raised with a female identity, and so identifies as a girl or woman. But she has no menstrual period because she has no ovaries or uterus -- she has an XY genotype and undescended internal testes. Androgens (male hormones) create the male external appearance; in the absence of androgens, a fetus develops a female appearance -- and parents and society assign a female identity.
It seems to me that a reasonable approach to dealing fairly with intersexed people in sports would be to replace sex separation with weight-class separation. Human sex and gender turn out to be complex and fraught with both social and technical problems, and it is unreasonable to deny an athlete the chance to compete just because their genetic curiosity happens to be on the 23rd chromosome.