There seem to be two different and orthogonal security concerns raised in the discussion. The one question actually asked is how to securely transmit the data to the consultant. For that, the correct answer is either GPG or SSH/SCP. Either of those insure that the channel itself cannot be intercepted.
The second question is whether the data is secure once the consultant has it. A corollary of this is whether it is *really* necessary to give the consultant data access. Perhaps the answer is "no", but I'm not privy to the internal issues that make "yes" at least a plausible answer. Once the data is in the consultants hands--no matter *how* it was transmitted--it's only as secure as the consultant keeps it. Whether it is an encrypted data channel like SSH, a hand-delivered medium, or whatever, if the consultant can decrypt it she could potentially expose it.
Bad choice! There have been fewer than a half-dozen documented cases of people surviving rabies for more than two weeks. On the other hand, tens of thousand number of people have survived AIDS for over 25 years (so far). Even taking base rates into account, your odds are orders of magnitude better with AIDS than rabies.
Researchers have also discovered that less than 1% of scientist produce more than 50% of the papers in top scientific journals. Clearly the system of peer review fails to address the hidden elitism of scientific work.
This very old, and very well known, fact that Wikipedia contriputions follow Zipf's Law (i.e. power-series distribution) gets repeated every six months as some conspiracy or cabal. What it amounts to is that a few Wikipedia users are very interested in contributing, and a lot more people are only slightly interested (or not at all). FWIW, I write that as someone in that top 1% of contributors (and can testify that my secret overlords have been seriously delinquent in paying the bribes).
It is a bit of an asinine question why Red Hat "tolerates" CentOS. Red Hat has no option here--nor should they. By distributing code or binaries that were created by people other than Red Hat, and licensed under GPL, Red Hat has explicitly agreed that CentOS (or anyone) has the right to do the same.
Red Hat is welcome to hold whatever opinion they want on whether they *like* CentOS to do what they do... but in the end, it's none of their damn business how someone else decides to distribute GPL'd code (within the license terms, of course... Red Hat is also a creator of a significant body of GPL code).
>>911 calls were the first thing I thought of, too. Any business owner who jams a call about somebody having a heart attack would be sued into oblivion, and deserve it.
I would hope personally that anyone (business owner or patron) who blocked 911 access illegally would suffer more than merely civil action. They would deserve prison time (i.e. criminal charges). Reckless endangerment of human life isn't something to treat lightly.
While I wouldn't be at all surprised for graduate departments in the USA to be majority non-USA citizens, it sounds rather off for any given department to be 90% Indian. There are certainly many very bright Indian computer science students who attend USA universities (as well as many in other departments, probably more in technical than humanities areas), but it's not *that* skewed towards India only. Many other nations also provide excellent candidates to USAian graduate programs as well; making most a healthy mixture of national origins.
As to the question itself: my reaction is *so what*?! It's great for the USA to have the best and brightest from other nations study here. Many of them will choose to stay in the USA longer term; those who don't will contribute to the intellectual development of the rest of the world, which is an equally excellent thing. Students who happen to be USA citizens are hardly being rejected on those grounds from US universities (though the lackluster quality of US education at lower levels than grad school certainly contributes to an underrepresentation of USA citizens at the top of intellectual achievement.
Moreover, take a look at Lynn Margulis' work about endosymbiosis. It now seems almost certain that eukayriotic cells (with multiple orgenelles) evolved from far simpler prokaryotic forms. Even those simpler cells no doubt arose somewhat incrementally from "pre-life" chemical reactions involving RNA and proteins outside functioning cells.
What preceded Darwinian evolution proper was indeed different than what you get via transmission of information encoding DNA segments, but it was certainly NOT some "miracle moment". Give a nice hydrocarbon soup a few million years, and some molecular regularities are quite likely to reoccur.
A number of posters have commented that Apple probably did this to insure their ability to relicense the CUPS code, or at least keep it under GPL2 rather than GPL3. That's probably true enough. But there's no particular harm in it either.
If Apple creates a branch of CUPS under some restrictive, proprietary license, so what? All the other developers in the world can take the last GPL(2) version, and enhance it however they like. They might not have access to Apple's enhancements, but there's no requirement for Apple to create them in the first place. For that matter, there's nothing that requires its primary developer to ever write another line of CUPS code either, so that's not something to count on either. Well, the contract with Apple might require such lines be written, but that's neither here nor there.
That's pretty much the whole reason why 80(-ish) columns remains the correct limit. It's not some magic number per-se--if old terminals had had 75 or 90 columns instead, those wouldn't be obviously absurd values. But reading 120, or 200, or 500 character lines of code imposes a very high extra cognitive burden.
Yeah sure, for some particular programming languages and libraries, and using some specific IDEs, *part* of the harmful effect of excessively long lines can be masked. As a number of people have noticed, it's pretty much the same reason that decent publishers never make readers read more than about the same number of characters on one line of text. And it's probably a pretty good argument against the absurdly verbose naming conventions of languages like Java. Even in Java, however, breaking lines for readability is still a good idea: If one identifier is itself already sentence-like, that's a reasonable motive to let that identifier stand more-or-less alone on a line.
I've recently worked a place where I had dual 30" monitors at my workstation, as one extended desktop. In other words, plenty of space for almost unlimited width editors or terminals. After experimentation, I find that anything past about 90 characters width in a terminal (i.e. bash, or psql, or ipython, etc) causes far more distraction in trying to read a line than it saves in preventing lines from splitting. These terminal applications naturally handle wrapping in a sufficiently reasonable way. The solution to presenting more information is using multiple windows, or panes, or tabs to present 80-ish column information sources *next to* each other; it is NOT to make one information display excessively wide.
The point of why this is a bad move for OLPC isn't just about what's bad with the Cuba export ban specifically. That ban *is* indeed stupid, but this also subverts the international intention of the OLPC project to the narrow whims the US administration.
Perhaps some other country or countries will be declared official enemies next year. Especially if, say, MS and Intel can persuade a US administration that a mandate for Free Software in, say, Peru or Bolivia, is "contrary to US interests". Or even if such a ban is declared for completely unrelated reasons, the OLPC should not allow itself to be derailed by partisan or sensationalist whims of a USA administration.
Ummm.. "In the UK, if a million were sold there you'd have a 1/54 chance [or so] of knowing someone who owned a Zune. In Canada, it'd be about 1/32 or so."
It seems to me that some people actually have more than one acquaintance! Well, also the topology of acquaintance probably doesn't form a ring either, but that's a lesser quibble.
So if 1/54 people in UK had a Zune, presumably each of them would know *several* people: heck, they might even encounter as many as a dozen people in the course of their daily lives:-). Hence bringing the odds to almost 1/4.
In any case, the Microsoft-shill unquestionably lies about the numbers, so I'm sure it's not 1M sold. But let's still keep basic arithmetic a little more on track.
Hmm... the entirety of my career over the last eight years or so has been "writing term papers". I'm rather well known for it, even, albeit only in certain programming language areas.
When I write CS/programming articles for IBM, Intel, O'Reilly and the like, there are indeed a few differences in citation style and organization; but it's pretty darn close to the same skills that go into making a term paper. I argue points. I do research. I cite sources. I stick to a word limit. I try to make the language flow as smoothly and clearly as possible.
Too bad I wasn't at this unnammed (and fictional, obviously--I don't actually believe the anecdote)--school. The professor would owe me somewhere around $10k. (Even the consulting and employment I've done other than writing per se has mostly involved documentation, research, verbal composition, etc; though only as part of the work).
Does anyone remember when an April Fool's joke was at least supposed to be clever or humorous.
The idea that Moore might play the Scully character rather than Anderson doing so is a perfectly plausible one... that just happens to be false. Not false in any interesting or obvious way, just not something that happened to come about. It's about as clever making a false claim about exactly which actor (or roughly the right age and look) might be in the next James Bond film, or who might play the next villain in a Spiderman.
Or hell, it's about as interesting as pulling everyone's leg with the "outrageous" claim I ate camembert cheeses sandwich rather than gouda... wow! I fooled you.
The whole point of the VHS/Betamax battle was their use of incompatible media that prevented interoperability. It was a real hinderance to have a tape that you could play upstairs but not downstairs, or that you could play at your house but not your friend's house. I know some people think the wrong winner emerged (I know there were pros and cons; though it's been a while since I remembered them all), but even if so, having just one standard is way better.
Nothing like the same exclusionary principle applies to Plasma/LCD. I can perfectly well have a plasma screen upstairs and an LCD downstairs. There may be pros and cons to each type of screen, but there is no interoperability issue whatsoever. If I want to carry my existing DVD player from one TV to the other, either one equally well accepts the input signal; or likewise to hook my cable signal to either. It's true a few different signal types compete: NTSC, S-Video, DVI, etc; but the set of inputs a TV accepts have nothing to do with the display technology it uses.
It could very well prove that one technology becomes sufficiently better in almost all respects that it "wins". Personally, I still hope for OLEDs to get far enough along eventually to displace both of them. But that's just details of price, resolution, durability, brightness, etc. There's no exclusionary principle anywhere here.
Plus the fact that the anti-virus companies don't like the competition from Consumer Reports; after all, it's those companies that themselves create most of the "proof-of-concept" viruses to scare potential buyers (especially to create scares of vulnerability on OSX, Linux, BSD, etc... where no real vulnerability exists).
I have also received spam with a "hammy" initial portion for a number of years. That is, a text block having nothing to do with selling me drugs, making my penis larger, or suggesting I look at porn, occurs in (usually at the beginning) of a message. Mostly it seems to be semi-grammatical stuff with commonplace words, I'm not sure where it comes from exactly.
However, more recently, I have had the feeling that the pseudo-ham seems more targeted at me. That is, the words chosen seem to be ones that have something to do with my own, somewhat unusual, intersts. It is hard to be sure--it's not like any of these areas are unique to me. But most people, say, are not necessarily interested in both Python programming and postmodern philosophy. Usually this latest batch has a graphic attached with a "hot tip" on some stock. I sort of wonder if the spammers are taking the effort to extract words from one of the very public places my email address occurs, which would often have those same words on them.
Then again, it might just be the "horoscope effect": y'know, when you read a horoscope or the like, you can sort of imagine the prediction is actually relevant to you personally if you ignore half of it and read the other half loosely or metaphorically. Maybe I'm reading more personalization into the keywords than really exists.
I'd enjoy it if readers would like to take a look at my coverage of the event for IBM developerWorks. I think I interviewed some interesting people, and generally report on a number of sessions Slashdotters might like. Look at:
I strongly encourage you to download and read my book, free of charge (http://gnosis.cx/TPiP/). Readers have been able to do this since before the day the dead-tree version was first published. FWIW, I actually have earned past advance, so I *am* starting to get a little bit for each sale; most books don't do that though.
And music labels are a lot worse for artists than semi-academic book publishers. As a rule anyway, a small number of top selling artists get some real money, and better terms. But for the vast majority of music label contracts, artists don't even get a real advance the way book authors do. I got an advance, and I would get to keep it even if my book never sold a copy... it wasn't huge, but it was mine. In contrast, if I had recorded a musical work, my "advance" would actually be an IOU to the record label. If I didn't sell any copies, I'd owe the money back to the label (and they are quite happy to sue artists for it). And as with books, most musical recordings don't "earn past the advance".
So indeed, for 95%+ of recorded artists, the amount *they* get from a sale of a CD at Tower Records is not SMALL, it's NADA, ZERO, ZILCH. Record labels get a non-zero amount from these sales, but not artists.
So in other words, it's exactly the same as buying the record from Circuit City or Tower Records... from which the artists get the same amount of money: none.
It's moderately interesting that HP has managed to sell more things than IBM has. But selling a whole lot of low-margin low-end systems doesn't really make for a bigger company overall. IBM still seems to have a better focus (despite its huge size), as well as better margins. Of course, no one has the huge margins than a monopoly gets you; but IBM is one of those companies that actually earns its money relatively honestly.
Slashdot Mirror
There seem to be two different and orthogonal security concerns raised in the discussion. The one question actually asked is how to securely transmit the data to the consultant. For that, the correct answer is either GPG or SSH/SCP. Either of those insure that the channel itself cannot be intercepted.
The second question is whether the data is secure once the consultant has it. A corollary of this is whether it is *really* necessary to give the consultant data access. Perhaps the answer is "no", but I'm not privy to the internal issues that make "yes" at least a plausible answer. Once the data is in the consultants hands--no matter *how* it was transmitted--it's only as secure as the consultant keeps it. Whether it is an encrypted data channel like SSH, a hand-delivered medium, or whatever, if the consultant can decrypt it she could potentially expose it.
The question seems slightly misguided, and the answers suggesting sqlite3 seem very good. However, if you want a literal answer, I would think that Python's csv module does what the poster requested: http://docs.python.org/lib/csv-examples.html. Well, if you're inclined in a different way, you could use Text:CSV: http://www.cpan.org/modules/by-module/Text/Text-CSV-0.01.readme
>I'd rather catch rabies than AIDS
Bad choice! There have been fewer than a half-dozen documented cases of people surviving rabies for more than two weeks. On the other hand, tens of thousand number of people have survived AIDS for over 25 years (so far). Even taking base rates into account, your odds are orders of magnitude better with AIDS than rabies.
Researchers have also discovered that less than 1% of scientist produce more than 50% of the papers in top scientific journals. Clearly the system of peer review fails to address the hidden elitism of scientific work.
This very old, and very well known, fact that Wikipedia contriputions follow Zipf's Law (i.e. power-series distribution) gets repeated every six months as some conspiracy or cabal. What it amounts to is that a few Wikipedia users are very interested in contributing, and a lot more people are only slightly interested (or not at all). FWIW, I write that as someone in that top 1% of contributors (and can testify that my secret overlords have been seriously delinquent in paying the bribes).
News at 11.
Even easier is FireGPG (http://firegpg.tuxfamily.org/).
It is a bit of an asinine question why Red Hat "tolerates" CentOS. Red Hat has no option here--nor should they. By distributing code or binaries that were created by people other than Red Hat, and licensed under GPL, Red Hat has explicitly agreed that CentOS (or anyone) has the right to do the same.
Red Hat is welcome to hold whatever opinion they want on whether they *like* CentOS to do what they do... but in the end, it's none of their damn business how someone else decides to distribute GPL'd code (within the license terms, of course... Red Hat is also a creator of a significant body of GPL code).
>>911 calls were the first thing I thought of, too. Any business owner who jams a call about somebody having a heart attack would be sued into oblivion, and deserve it.
I would hope personally that anyone (business owner or patron) who blocked 911 access illegally would suffer more than merely civil action. They would deserve prison time (i.e. criminal charges). Reckless endangerment of human life isn't something to treat lightly.
While I wouldn't be at all surprised for graduate departments in the USA to be majority non-USA citizens, it sounds rather off for any given department to be 90% Indian. There are certainly many very bright Indian computer science students who attend USA universities (as well as many in other departments, probably more in technical than humanities areas), but it's not *that* skewed towards India only. Many other nations also provide excellent candidates to USAian graduate programs as well; making most a healthy mixture of national origins.
As to the question itself: my reaction is *so what*?! It's great for the USA to have the best and brightest from other nations study here. Many of them will choose to stay in the USA longer term; those who don't will contribute to the intellectual development of the rest of the world, which is an equally excellent thing. Students who happen to be USA citizens are hardly being rejected on those grounds from US universities (though the lackluster quality of US education at lower levels than grad school certainly contributes to an underrepresentation of USA citizens at the top of intellectual achievement.
Moreover, take a look at Lynn Margulis' work about endosymbiosis. It now seems almost certain that eukayriotic cells (with multiple orgenelles) evolved from far simpler prokaryotic forms. Even those simpler cells no doubt arose somewhat incrementally from "pre-life" chemical reactions involving RNA and proteins outside functioning cells.
What preceded Darwinian evolution proper was indeed different than what you get via transmission of information encoding DNA segments, but it was certainly NOT some "miracle moment". Give a nice hydrocarbon soup a few million years, and some molecular regularities are quite likely to reoccur.
A number of posters have commented that Apple probably did this to insure their ability to relicense the CUPS code, or at least keep it under GPL2 rather than GPL3. That's probably true enough. But there's no particular harm in it either.
If Apple creates a branch of CUPS under some restrictive, proprietary license, so what? All the other developers in the world can take the last GPL(2) version, and enhance it however they like. They might not have access to Apple's enhancements, but there's no requirement for Apple to create them in the first place. For that matter, there's nothing that requires its primary developer to ever write another line of CUPS code either, so that's not something to count on either. Well, the contract with Apple might require such lines be written, but that's neither here nor there.
That's pretty much the whole reason why 80(-ish) columns remains the correct limit. It's not some magic number per-se--if old terminals had had 75 or 90 columns instead, those wouldn't be obviously absurd values. But reading 120, or 200, or 500 character lines of code imposes a very high extra cognitive burden.
Yeah sure, for some particular programming languages and libraries, and using some specific IDEs, *part* of the harmful effect of excessively long lines can be masked. As a number of people have noticed, it's pretty much the same reason that decent publishers never make readers read more than about the same number of characters on one line of text. And it's probably a pretty good argument against the absurdly verbose naming conventions of languages like Java. Even in Java, however, breaking lines for readability is still a good idea: If one identifier is itself already sentence-like, that's a reasonable motive to let that identifier stand more-or-less alone on a line.
I've recently worked a place where I had dual 30" monitors at my workstation, as one extended desktop. In other words, plenty of space for almost unlimited width editors or terminals. After experimentation, I find that anything past about 90 characters width in a terminal (i.e. bash, or psql, or ipython, etc) causes far more distraction in trying to read a line than it saves in preventing lines from splitting. These terminal applications naturally handle wrapping in a sufficiently reasonable way. The solution to presenting more information is using multiple windows, or panes, or tabs to present 80-ish column information sources *next to* each other; it is NOT to make one information display excessively wide.
The point of why this is a bad move for OLPC isn't just about what's bad with the Cuba export ban specifically. That ban *is* indeed stupid, but this also subverts the international intention of the OLPC project to the narrow whims the US administration.
Perhaps some other country or countries will be declared official enemies next year. Especially if, say, MS and Intel can persuade a US administration that a mandate for Free Software in, say, Peru or Bolivia, is "contrary to US interests". Or even if such a ban is declared for completely unrelated reasons, the OLPC should not allow itself to be derailed by partisan or sensationalist whims of a USA administration.
WTF? Why shouldn't you have the right to modify GPL code on your IBM mainframe?!
Sounds like big-time FUD to me.
Ummm.. "In the UK, if a million were sold there you'd have a 1/54 chance [or so] of knowing someone who owned a Zune. In Canada, it'd be about 1/32 or so."
:-). Hence bringing the odds to almost 1/4.
It seems to me that some people actually have more than one acquaintance! Well, also the topology of acquaintance probably doesn't form a ring either, but that's a lesser quibble.
So if 1/54 people in UK had a Zune, presumably each of them would know *several* people: heck, they might even encounter as many as a dozen people in the course of their daily lives
In any case, the Microsoft-shill unquestionably lies about the numbers, so I'm sure it's not 1M sold. But let's still keep basic arithmetic a little more on track.
Hmm... the entirety of my career over the last eight years or so has been "writing term papers". I'm rather well known for it, even, albeit only in certain programming language areas.
When I write CS/programming articles for IBM, Intel, O'Reilly and the like, there are indeed a few differences in citation style and organization; but it's pretty darn close to the same skills that go into making a term paper. I argue points. I do research. I cite sources. I stick to a word limit. I try to make the language flow as smoothly and clearly as possible.
Too bad I wasn't at this unnammed (and fictional, obviously--I don't actually believe the anecdote)--school. The professor would owe me somewhere around $10k. (Even the consulting and employment I've done other than writing per se has mostly involved documentation, research, verbal composition, etc; though only as part of the work).
Does anyone remember when an April Fool's joke was at least supposed to be clever or humorous.
The idea that Moore might play the Scully character rather than Anderson doing so is a perfectly plausible one... that just happens to be false. Not false in any interesting or obvious way, just not something that happened to come about. It's about as clever making a false claim about exactly which actor (or roughly the right age and look) might be in the next James Bond film, or who might play the next villain in a Spiderman.
Or hell, it's about as interesting as pulling everyone's leg with the "outrageous" claim I ate camembert cheeses sandwich rather than gouda... wow! I fooled you.
Boy is the Betamax analogy pointless!
The whole point of the VHS/Betamax battle was their use of incompatible media that prevented interoperability. It was a real hinderance to have a tape that you could play upstairs but not downstairs, or that you could play at your house but not your friend's house. I know some people think the wrong winner emerged (I know there were pros and cons; though it's been a while since I remembered them all), but even if so, having just one standard is way better.
Nothing like the same exclusionary principle applies to Plasma/LCD. I can perfectly well have a plasma screen upstairs and an LCD downstairs. There may be pros and cons to each type of screen, but there is no interoperability issue whatsoever. If I want to carry my existing DVD player from one TV to the other, either one equally well accepts the input signal; or likewise to hook my cable signal to either. It's true a few different signal types compete: NTSC, S-Video, DVI, etc; but the set of inputs a TV accepts have nothing to do with the display technology it uses.
It could very well prove that one technology becomes sufficiently better in almost all respects that it "wins". Personally, I still hope for OLEDs to get far enough along eventually to displace both of them. But that's just details of price, resolution, durability, brightness, etc. There's no exclusionary principle anywhere here.
Plus the fact that the anti-virus companies don't like the competition from Consumer Reports; after all, it's those companies that themselves create most of the "proof-of-concept" viruses to scare potential buyers (especially to create scares of vulnerability on OSX, Linux, BSD, etc... where no real vulnerability exists).
No g*d d*mn Dvorak! Seriously, can't we just follow a simple rule to make Slashdot a little bit better. Just don't feed the troll!
I have also received spam with a "hammy" initial portion for a number of years. That is, a text block having nothing to do with selling me drugs, making my penis larger, or suggesting I look at porn, occurs in (usually at the beginning) of a message. Mostly it seems to be semi-grammatical stuff with commonplace words, I'm not sure where it comes from exactly.
However, more recently, I have had the feeling that the pseudo-ham seems more targeted at me. That is, the words chosen seem to be ones that have something to do with my own, somewhat unusual, intersts. It is hard to be sure--it's not like any of these areas are unique to me. But most people, say, are not necessarily interested in both Python programming and postmodern philosophy. Usually this latest batch has a graphic attached with a "hot tip" on some stock. I sort of wonder if the spammers are taking the effort to extract words from one of the very public places my email address occurs, which would often have those same words on them.
Then again, it might just be the "horoscope effect": y'know, when you read a horoscope or the like, you can sort of imagine the prediction is actually relevant to you personally if you ignore half of it and read the other half loosely or metaphorically. Maybe I'm reading more personalization into the keywords than really exists.
I'd enjoy it if readers would like to take a look at my coverage of the event for IBM developerWorks. I think I interviewed some interesting people, and generally report on a number of sessions Slashdotters might like. Look at:
a vidmertz
http://www-03.ibm.com/developerworks/blogs/page/d
I strongly encourage you to download and read my book, free of charge (http://gnosis.cx/TPiP/). Readers have been able to do this since before the day the dead-tree version was first published. FWIW, I actually have earned past advance, so I *am* starting to get a little bit for each sale; most books don't do that though.
And music labels are a lot worse for artists than semi-academic book publishers. As a rule anyway, a small number of top selling artists get some real money, and better terms. But for the vast majority of music label contracts, artists don't even get a real advance the way book authors do. I got an advance, and I would get to keep it even if my book never sold a copy... it wasn't huge, but it was mine. In contrast, if I had recorded a musical work, my "advance" would actually be an IOU to the record label. If I didn't sell any copies, I'd owe the money back to the label (and they are quite happy to sue artists for it). And as with books, most musical recordings don't "earn past the advance".
So indeed, for 95%+ of recorded artists, the amount *they* get from a sale of a CD at Tower Records is not SMALL, it's NADA, ZERO, ZILCH. Record labels get a non-zero amount from these sales, but not artists.
So in other words, it's exactly the same as buying the record from Circuit City or Tower Records... from which the artists get the same amount of money: none.
It's moderately interesting that HP has managed to sell more things than IBM has. But selling a whole lot of low-margin low-end systems doesn't really make for a bigger company overall. IBM still seems to have a better focus (despite its huge size), as well as better margins. Of course, no one has the huge margins than a monopoly gets you; but IBM is one of those companies that actually earns its money relatively honestly.