He wants to see how you think, and you didn't seem to make that obvious, you were more interesting in answering questions with academic answers not immediately useful for the real world. His question for the phonebook asked "how long" it takes to look up the name, and just reciting O(1) isn't the full answer to this. You're right that logs of diffeerent bases are only related by a multiplicative factor, but if someone wants to know how many comparitive lookups you need what reason could you possibly want for expressing this in any base other than two? (I'm a physicist, not a comp-sci guy, so if there is an answer to that I'd be curious to know). To answer how long, you need to know how long each lookup takes and how many lookups you would need to perform (assuming he wanted an answer in time). You were like a politician, and answered a different question than the one he asked.
You also made it clearly obvious to the interviewer that you would be a very difficult guy to work with, Ie, if you're of average google hiring intelligence and experience, half of your coworkers at google would be less smart or skilled as you. And if someone needs help understanding big(O) notation for their project and asked you to help them, you might be a dick to them, as per your interview. They were probably looking for base 26, actually. It's the number of letters in our alphabet, so a trie (which is an m-tree where each path from the root to a leaf denotes a sequence) would have 26 children at each node. Although this may have been what they were looking for, Knuth proved that any m-tree can be expressed as a binary tree with only a constant factor difference, which puts it in the same bucket as differences between machine architectures.
Big O notation is one of the foundational concepts in computer science, right after the Church-Turing thesis and the proof of the undecidability of the halting problem. As a physicist, it would be similar to applying to a highly technical company (one that, say, makes scanning electron microscopes or maybe helium refrigeration equipment), and having to explain to one of your future coworkers that work should be measured in Joules or electron-volts instead of man-hours.
The Declaration of Independence used the "pursuit of happiness" phrase, whereas the Constitution used the "life, liberty, and property" phrase. However, Jefferson got his phrase from "life, liberty, and property", which came from Locke's philosophy of the role of the state. Many historians suggest that the rephrase was a deliberate attempt to avoid enshrining property as a founding principle of the nascent United States, in order to undermine arguments in favor of slavery that were often founded on the presumption of property rights.
This is part of the oath sworn by members of the U.S. Military: I, , do solemnly swear (or affirm) to uphold the Constitution of the United States against all enemies, foreign and domestic, and to obey the orders of the President and the officers appointed over me, (so help me god). The parts in the parentheses are optional.
If a big rocket falls on your house, it's not difficult to figure out who to sue; hence, regulation of this is not necessary.
Even if they do only care about profits, losing your rockets and buying someone a new house is never good for business. If their shortsightedness causes them to neglect these concerns, they deserve to fail.
Reports that include the phrase "more than all the elephants on the planet" are always welcome.
Interviewer: "Mr. Torvalds, what would it take to get you to pay for an SCO license?" Linus: More than all the elephants on the planet.
Earlier today, the RIAA announced that it would be suing filesharers. The damages? In dollars, more than all the elephants on the planet (if you sold them all at current market prices).
A new release of Mozilla Thunderbird is out. A huge party is planned. When the developers were asked how much ass the party would kick, the developers simply said, "more than all the elephants on the planet"
This is just search engine software. First of all, no one needs to make money from this, they can just use it for their own websites or intranets. Second, a company could only spider the domains of companies that pay for it, or only index the, say, PDF's of domains that pay for it. Subscription based services could contract with the company to write a specific spider that indexes the pages behind the subscription (or pay-per-view documents).
Or, Yahoo could use it. Yahoo's still a portal, and the popularity of their site brings in money. Hiring a full-time developer to work on the software is not out of the question if they think they can surpass google one day.
Instead of modding me down, post a reply telling me why forcing a copyright holder to allow free sharing of his work is good public policy. I want to believe, I just haven't heard a satisfactory argument yet.
Your wish is my command.
most of them boil down to a general denial of property rights or good reasons why the artist/label/retailer would benefit if they decided to allow copying. If I haven't heard a satisfactory defense (and I'm looking)
Far from denying property rights, it is easy to demonstrate that patents and copyrights undermine the property rights of several parties.
Historically, recording and distribution have been natural partners. When you recorded something, something was modified to include that, and this was done on a large scale as the manufacture of vinyl albums and CD's. When the information age came about, it became possible to separate these tasks. Now, in telecommunications there is a natural synergy between previously heterogeneous media because networks can operate on information without any knowledge of the information, which allows for great economies of scale. These networks were already in place for the distribution of other types of media -- telephone conversations were their genesis and TCP packets were common. So when it became feasible for music to be distributed digitally in a format that had value, there was already a distribution network available. It's also more economical to distribute information digitally, although this is difficult to show.
The RIAA companies already have a distribution network in place, and it happens to be highly interoperable with the digital distribution system currently called piracy because the piratical network is simply lossy digital compression of the information distributed by the RIAA's chosen distribution method (plus or minus ID3). Thus the current value of the RIAA's current distribution network is diminished substantially by the introduction of the more efficient piracy. Its value is essentially the aggregate value customers place on the "officialness" of the content as well as the higher quality of CD's.
In order to crush the new distribution network, it's necessary to crush the new distribution network, which is owned by several parties. The network is owned by the telecoms, the ends are owned by ISPs, and the ISP connects to computers, which are owned by consumers.
The RIAA has ordered a trifecta of infringements on the rights of these three parties. First the DMCA mandates the externalization of the cost of discovery off of the RIAA onto the telecoms, which own the networks. Externalization means that the telecoms have to pay for the use of the telecom's property, which is antithetical to property rights, wherein the owners of a thing incur the costs and benefits of the thing. Second, the DMCA mandates that ISPs disconnect customers, which is antithetical to the right of a property owner to control what is done with the owner's property. Thirdly, the RIAA is suing customers of the ISPs, where if they lose, they must give their property to the copyright holder, which is antithetical to the right of a property owner to transfer or not transfer his property to another at his own discretion.
Redefining property rights to not include copyrighted material, one must discard all of the rights of a property owner to the physical strata on which these ideas lie. In particular, the RIAA and/or MPAA claim partial property rights to the following things owned by *you*: DVD's, DVD players, DVD burners, computers, CD's, CD players, CD burners, cable modem or modem, TV, VCR, MP3 player, game console, and PVR.
Some argue that there is a morality issue to distributing other people's ideas. I argue that this is simply a confusion of the actual, related issues of plagiarism, which is claiming work you did not do as your own, and putting someone else's name on something *you* did. In every filesharing system, this is strongly systematically discouraged, excep
A $2 million endowment can last forever. If wisely invested, the dividends on $2 million can pay $50 000 per year and still grow. Not enough to employ fulltime developers, but probably enough for bandwidth costs.
In particular, I'm thinking of China's widespread opium addiction leading up to (and following) the Opium Wars. Opium was, of course, illegal.
Heck, one of the common themes in the Tao Te Ching is that you shouldn't pass laws no one will follow so that the people don't lose respect for the laws or your government. Sounds kinda familiar, doesn't it?
You mean like Linus needed to have the resources AT&T did when it first developed unix? Or how KDE had to have the resources of windows?
Hell, Gnu/Linux has 5% market share, and most accounts put it on at least a techinical plateau with Windows, which has 90%. That would seem to make Linux 18 times more efficient than Windows.
By the same ratio, it would only require 5 full-time developers over two years, or 1 full-time developers and 18 part-time developers who average 2 hours a day. This is not unreasonable.
And this is a low estimate. A more accurate but controversial estimate would likely yield a much greater ratio.
Yeah, it'd be nice if Mozilla had some more convenient keys than this, but Enter should not be that key. Currently, find-as-you-type is set to check only the text in links by default, so if you use that default, to find a link, you can then press enter to load the page linked to by the link you found.
Personally, I'd rather have it use [ and ] or , and . or even just ; and shift-;
It is because software is supposed to bend to the will of the user, not the other way around. And that is why software is so feature-laden, so mandatorilly configurable.
If you wrote yourself a business app without configuration, you are dictating to your customers exactly how they will do the business they intend to use your software to do. That's great if your customer either does not do business in this area yet, or if they already do business the same way you expect them to.
This is exactly what changed during the industrial revolution. There may have been objections that automated seamstresses would need to manually tailor their clothing for the customer. This is obviously not true, as evidenced by the existence of T-shirts with standard sizes.
And that's the problem. Customers frequently don't know how they do business, and forcing them to articulate their current processes leads to them facing this unpleasant truth. Sometimes they tell you the wrong thing. Sometimes they deliberately tell you the wrong thing.
Those companies that do should die. They will be unable to acquire business software that fulfills their needs and thus be unable to compete. But vendors and buyers these days both think they need shrinkwrap software to buy and sell. In other words, they think they need toothbrushes instead of the machine that spits out their plastic handles. The machine needed to be retooled to spit out handles. It needed to be hooked up to a conveyor, which needed to be hooked up to a machine to attach the bristles.
We don't need to hire machinists to lathe out toothbrushes (shrinkwrapware) one by one anymore, we need software that can be retooled (BSD/GPL?). We need software that can be hooked up to a conveyor (SOAP?), which we can hook up to other software (bonobo? XML? lisp sexps?).
It sounds like they're going to suggest that someone in IBM leaked SCO's source code to a Linux developer and that developer submitted it to the kernel. That would explain why they're not releasing the portions of code in question -- in case that developer wipes all the evidence of having had the leaked code.
The evidence wouldn't be on lkml because it would have been private emails. Neither would it be in the various historical linux versions.
I'm surprised I haven't read any comments about how we should create a better protocol than the current one. When email gets flooded with so much spam it becomes unusable, you want a successor protocol, but you don't want the same thing for Usenet?
AFAIK, Usenet lists are mostly unmoderated, and the ones that are have moderators verify every post before making it public. Perhaps we should include ways of moderation that are more like the web-based forums everywhere (EZboard, YaBB, Ikonboard, etc.). Or perhaps something more like Slashdot.
We have to at least attempt a technical solution; after all, we're geeks.
If you have 97.8 Trillion, and you need to distribute it 545 ways (435 representatives, 100 senators, 1 president, 9 justices), you get to bribe each one with $179 million.
You could buy outright fascism and have it tomorrow with that kind of money.
Before capitalism, it was thought that there was a constant amount of wealth in the world. Then modern economics came along and pointed to a scenario like this:
A has an item Q. A and B value item Q differently. For simplicity, let's say Q is worth $2 to B and $1 to A. Then, they set up a trade. In order for a trade to occur, both parties must see the trade as beneficial to them. Let's say B gives A $1.5 for Q.
Let's pretend for a moment that there is nothing else in the world. Then the sum of all values in the world was $1 (A's valuation of Q) + $1.5 (B's pocket change) = $2.5 After the trade, the total amount of wealth is $2 (B's valuation of Q) + $1.5 (A's new pocket change) = $3.5
The amount is greater because A and B both have more of the things they want (Q and money). A is better off because he has $1.5 instead of Q (which is worth $1 to him). B is better off because he has Q (which is worth $2 to him) instead of just $1.5 in cash.
Extrapolate this to hundreds of thousands of trades, and you can envision how economic growth theoretically makes everyone happier.
Thus, forcing people to give a number to how much they want things facilitates a redistribution of items that makes everyone get more of what they want. Whether this reduces suffering and increases joy is up to you, but it probably won't have the reverse effect.
Dunno. I get turned off to think that if I miss a tab somewhere I'll get a compiler error. A brace, sure. But whitespace??
I didn't like it either, at first. Then I got used to it. Then I really liked it.
In, let's say, C (I don't know Perl very well), an if statement can be followed by a single statement or a block in braces. To, lets say, add debugging information, one has to enclose the statement in braces. In Python, one needs only to add more statements.
Also, the structure of the code is automatically visible, and errors in indentation stick out like a sore thumb. It's much harder to tell with braces. (especially with the optional braces that C has)
If you use Vi, vi displays the tabs with the same number of spaces as Python equates with them (by default, anyway).
The delimiter symbols ({}, [], () ) are reserved for container data structures instead of syntax. {} denotes a dictionary, [] a list (variable length array), and () a tuple (fixed length array; also used for function calls). This makes the syntax for manipulating data structures cleaner and simpler.
I'm not really a newbie (4 or 5 years coding C) at C/C++, but it's still more readable. Instead of adopting coding styles that make it easy to make sure that all our braces match (while throwing away vertical screen space), we should adopt languages that make it not matter.
Slashdot Mirror
You also made it clearly obvious to the interviewer that you would be a very difficult guy to work with, Ie, if you're of average google hiring intelligence and experience, half of your coworkers at google would be less smart or skilled as you. And if someone needs help understanding big(O) notation for their project and asked you to help them, you might be a dick to them, as per your interview. They were probably looking for base 26, actually. It's the number of letters in our alphabet, so a trie (which is an m-tree where each path from the root to a leaf denotes a sequence) would have 26 children at each node. Although this may have been what they were looking for, Knuth proved that any m-tree can be expressed as a binary tree with only a constant factor difference, which puts it in the same bucket as differences between machine architectures.
Big O notation is one of the foundational concepts in computer science, right after the Church-Turing thesis and the proof of the undecidability of the halting problem. As a physicist, it would be similar to applying to a highly technical company (one that, say, makes scanning electron microscopes or maybe helium refrigeration equipment), and having to explain to one of your future coworkers that work should be measured in Joules or electron-volts instead of man-hours.
In Firefox, there's an entry in the help menu called "Report Broken Website". Also, the Mozilla Foundation has a site evangelism page here: http://www.mozilla.org/projects/tech-evangelism/si te/procedures.html. Broken websites are recorded and tracked in Bugzilla.
The Declaration of Independence used the "pursuit of happiness" phrase, whereas the Constitution used the "life, liberty, and property" phrase. However, Jefferson got his phrase from "life, liberty, and property", which came from Locke's philosophy of the role of the state. Many historians suggest that the rephrase was a deliberate attempt to avoid enshrining property as a founding principle of the nascent United States, in order to undermine arguments in favor of slavery that were often founded on the presumption of property rights.
This is part of the oath sworn by members of the U.S. Military: I, , do solemnly swear (or affirm) to uphold the Constitution of the United States against all enemies, foreign and domestic, and to obey the orders of the President and the officers appointed over me, (so help me god). The parts in the parentheses are optional.
If a big rocket falls on your house, it's not difficult to figure out who to sue; hence, regulation of this is not necessary.
Even if they do only care about profits, losing your rockets and buying someone a new house is never good for business. If their shortsightedness causes them to neglect these concerns, they deserve to fail.
Where can I find your collection? I'll probably end up mirroring it, and listening to some of it.
Interviewer: "Mr. Torvalds, what would it take to get you to pay for an SCO license?"
Linus: More than all the elephants on the planet.
Earlier today, the RIAA announced that it would be suing filesharers. The damages? In dollars, more than all the elephants on the planet (if you sold them all at current market prices).
A new release of Mozilla Thunderbird is out. A huge party is planned. When the developers were asked how much ass the party would kick, the developers simply said, "more than all the elephants on the planet"
This is just search engine software. First of all, no one needs to make money from this, they can just use it for their own websites or intranets. Second, a company could only spider the domains of companies that pay for it, or only index the, say, PDF's of domains that pay for it. Subscription based services could contract with the company to write a specific spider that indexes the pages behind the subscription (or pay-per-view documents).
Or, Yahoo could use it. Yahoo's still a portal, and the popularity of their site brings in money. Hiring a full-time developer to work on the software is not out of the question if they think they can surpass google one day.
Instead of modding me down, post a reply telling me why forcing a copyright holder to allow free sharing of his work is good public policy. I want to believe, I just haven't heard a satisfactory argument yet.
Your wish is my command.
most of them boil down to a general denial of property rights or good reasons why the artist/label/retailer would benefit if they decided to allow copying. If I haven't heard a satisfactory defense (and I'm looking)
Far from denying property rights, it is easy to demonstrate that patents and copyrights undermine the property rights of several parties.
Historically, recording and distribution have been natural partners. When you recorded something, something was modified to include that, and this was done on a large scale as the manufacture of vinyl albums and CD's. When the information age came about, it became possible to separate these tasks. Now, in telecommunications there is a natural synergy between previously heterogeneous media because networks can operate on information without any knowledge of the information, which allows for great economies of scale. These networks were already in place for the distribution of other types of media -- telephone conversations were their genesis and TCP packets were common. So when it became feasible for music to be distributed digitally in a format that had value, there was already a distribution network available. It's also more economical to distribute information digitally, although this is difficult to show.
The RIAA companies already have a distribution network in place, and it happens to be highly interoperable with the digital distribution system currently called piracy because the piratical network is simply lossy digital compression of the information distributed by the RIAA's chosen distribution method (plus or minus ID3). Thus the current value of the RIAA's current distribution network is diminished substantially by the introduction of the more efficient piracy. Its value is essentially the aggregate value customers place on the "officialness" of the content as well as the higher quality of CD's.
In order to crush the new distribution network, it's necessary to crush the new distribution network, which is owned by several parties. The network is owned by the telecoms, the ends are owned by ISPs, and the ISP connects to computers, which are owned by consumers.
The RIAA has ordered a trifecta of infringements on the rights of these three parties. First the DMCA mandates the externalization of the cost of discovery off of the RIAA onto the telecoms, which own the networks. Externalization means that the telecoms have to pay for the use of the telecom's property, which is antithetical to property rights, wherein the owners of a thing incur the costs and benefits of the thing. Second, the DMCA mandates that ISPs disconnect customers, which is antithetical to the right of a property owner to control what is done with the owner's property. Thirdly, the RIAA is suing customers of the ISPs, where if they lose, they must give their property to the copyright holder, which is antithetical to the right of a property owner to transfer or not transfer his property to another at his own discretion.
Redefining property rights to not include copyrighted material, one must discard all of the rights of a property owner to the physical strata on which these ideas lie. In particular, the RIAA and/or MPAA claim partial property rights to the following things owned by *you*: DVD's, DVD players, DVD burners, computers, CD's, CD players, CD burners, cable modem or modem, TV, VCR, MP3 player, game console, and PVR.
Some argue that there is a morality issue to distributing other people's ideas. I argue that this is simply a confusion of the actual, related issues of plagiarism, which is claiming work you did not do as your own, and putting someone else's name on something *you* did. In every filesharing system, this is strongly systematically discouraged, excep
I see you didn't get the memo:. shtml
http://www.reason.com/0303/fe.jm.disarming
Short version:
No, you're wrong.
China had a huge drug problem before the communist regime came along. The solution? Execute the opium-heads.
I'd guess that that's why China has a lesser rate of incarceration -- a lot higher rate of execution.
A $2 million endowment can last forever. If wisely invested, the dividends on $2 million can pay $50 000 per year and still grow. Not enough to employ fulltime developers, but probably enough for bandwidth costs.
This is definately not unique to America.
In particular, I'm thinking of China's widespread opium addiction leading up to (and following) the Opium Wars. Opium was, of course, illegal.
Heck, one of the common themes in the Tao Te Ching is that you shouldn't pass laws no one will follow so that the people don't lose respect for the laws or your government. Sounds kinda familiar, doesn't it?
You mean like Linus needed to have the resources AT&T did when it first developed unix? Or how KDE had to have the resources of windows?
Hell, Gnu/Linux has 5% market share, and most accounts put it on at least a techinical plateau with Windows, which has 90%. That would seem to make Linux 18 times more efficient than Windows.
By the same ratio, it would only require 5 full-time developers over two years, or 1 full-time developers and 18 part-time developers who average 2 hours a day. This is not unreasonable.
And this is a low estimate. A more accurate but controversial estimate would likely yield a much greater ratio.
Yeah, it'd be nice if Mozilla had some more convenient keys than this, but Enter should not be that key. Currently, find-as-you-type is set to check only the text in links by default, so if you use that default, to find a link, you can then press enter to load the page linked to by the link you found.
Personally, I'd rather have it use [ and ] or , and . or even just ; and shift-;
[quote]even smells like Unix[/quote]Then let's hope the original Unix developers didn't use deodorant.
This is exactly what changed during the industrial revolution. There may have been objections that automated seamstresses would need to manually tailor their clothing for the customer. This is obviously not true, as evidenced by the existence of T-shirts with standard sizes.
Those companies that do should die. They will be unable to acquire business software that fulfills their needs and thus be unable to compete. But vendors and buyers these days both think they need shrinkwrap software to buy and sell. In other words, they think they need toothbrushes instead of the machine that spits out their plastic handles. The machine needed to be retooled to spit out handles. It needed to be hooked up to a conveyor, which needed to be hooked up to a machine to attach the bristles.
We don't need to hire machinists to lathe out toothbrushes (shrinkwrapware) one by one anymore, we need software that can be retooled (BSD/GPL?). We need software that can be hooked up to a conveyor (SOAP?), which we can hook up to other software (bonobo? XML? lisp sexps?).
What's to stop someone from walking by with a laptop, and gaining access to all your data?
The off button.
It sounds like they're going to suggest that someone in IBM leaked SCO's source code to a Linux developer and that developer submitted it to the kernel. That would explain why they're not releasing the portions of code in question -- in case that developer wipes all the evidence of having had the leaked code.
The evidence wouldn't be on lkml because it would have been private emails. Neither would it be in the various historical linux versions.
I'm surprised I haven't read any comments about how we should create a better protocol than the current one. When email gets flooded with so much spam it becomes unusable, you want a successor protocol, but you don't want the same thing for Usenet?
AFAIK, Usenet lists are mostly unmoderated, and the ones that are have moderators verify every post before making it public. Perhaps we should include ways of moderation that are more like the web-based forums everywhere (EZboard, YaBB, Ikonboard, etc.). Or perhaps something more like Slashdot.
We have to at least attempt a technical solution; after all, we're geeks.
Damn. You must have sucked. Those things took at least two hours per try for me.
Let's do the math:
If you have 97.8 Trillion, and you need to distribute it 545 ways (435 representatives, 100 senators, 1 president, 9 justices), you get to bribe each one with $179 million.
You could buy outright fascism and have it tomorrow with that kind of money.
Before capitalism, it was thought that there was a constant amount of wealth in the world. Then modern economics came along and pointed to a scenario like this:
A has an item Q. A and B value item Q differently. For simplicity, let's say Q is worth $2 to B and $1 to A. Then, they set up a trade. In order for a trade to occur, both parties must see the trade as beneficial to them. Let's say B gives A $1.5 for Q.
Let's pretend for a moment that there is nothing else in the world. Then the sum of all values in the world was $1 (A's valuation of Q) + $1.5 (B's pocket change) = $2.5
After the trade, the total amount of wealth is $2 (B's valuation of Q) + $1.5 (A's new pocket change) = $3.5
The amount is greater because A and B both have more of the things they want (Q and money). A is better off because he has $1.5 instead of Q (which is worth $1 to him). B is better off because he has Q (which is worth $2 to him) instead of just $1.5 in cash.
Extrapolate this to hundreds of thousands of trades, and you can envision how economic growth theoretically makes everyone happier.
Thus, forcing people to give a number to how much they want things facilitates a redistribution of items that makes everyone get more of what they want. Whether this reduces suffering and increases joy is up to you, but it probably won't have the reverse effect.
Dunno. I get turned off to think that if I miss a tab somewhere I'll get a compiler error. A brace, sure. But whitespace??
I didn't like it either, at first. Then I got used to it. Then I really liked it.
In, let's say, C (I don't know Perl very well), an if statement can be followed by a single statement or a block in braces. To, lets say, add debugging information, one has to enclose the statement in braces. In Python, one needs only to add more statements.
Also, the structure of the code is automatically visible, and errors in indentation stick out like a sore thumb. It's much harder to tell with braces. (especially with the optional braces that C has)
If you use Vi, vi displays the tabs with the same number of spaces as Python equates with them (by default, anyway).
The delimiter symbols ({}, [], () ) are reserved for container data structures instead of syntax. {} denotes a dictionary, [] a list (variable length array), and () a tuple (fixed length array; also used for function calls). This makes the syntax for manipulating data structures cleaner and simpler.
I'm not really a newbie (4 or 5 years coding C) at C/C++, but it's still more readable. Instead of adopting coding styles that make it easy to make sure that all our braces match (while throwing away vertical screen space), we should adopt languages that make it not matter.
In this case, the obligatory Simpson's references really *are* obligatory.