Even if it still effectively filters 50% of the remaining 75% of my email.. that's what, about 38%? SpamAssassin stops more than that before I even train its Bayesian brain. I'm also doubtful of its zero false positive claims.
The stated intent of this method (and am I the only person who has actually read the full 6-page proposal) is to reduce the computational overhead of filtering mail. The way this is intended to be used is to create whitelists and blacklists. Searching the header for a whitelist is many cases is less processor intensive that searching the whole message multiple times for spam phrases, or a fulltext baysian analysis.
And from the sounds of it, what makes it different from black(or white)lists? True, it's more sophisticated because it uses the whitelists of those on your whitelists, but why not just use a plain whitelist anyway?
The actual article that describes the algorithm explicitly states that it is a way to create whitelists based on existing information from your mailbox.
And how does this allow email from internet transactions or other non-social sources through? The article didn't seem to address that so clearly.
The algorithm has a fudge factor that assigns an ambiguous score for cases where the clustering coefficient can't be calculated.
The first is trivial and certain to succeed but has a Drawback to spammers: only send e-mail to single recpients. The drawback is this puts a much higher load on their servers since every message is sent individually.
True this method is strongest against dictionary spam and does not work against non-dictionary spam.
[i]The second method is to always include dummy addresses in the mailing list that the recpients probably have in their address books. For example, add the following names to the to-field: notifications@paypal.com and list-notication@ebay.com. Any recpieint that of the spam message that also has recieved e-mail from e-bay or pay-pal will trust the message.[/i]
Um, did you RTFA? (And perhaps most importantly, did anybody modding this article RTFA.)
The algorithm has nothing do do with addressbooks. Instead, it looks at friend of a friend networks as identified by mail headers.
For example, I work on a project with Bob, and Susan. A typical email message about the project will include my address, and their addresses in the header. The algorithm assumes that three first degree relationships exist: me-bob me-susan susan-bob
There are also three second-degree (friend of a friend relationships. me-susan-bob me-bob-susan susan- me-bob
The high ratio of second-degree/first-degree relationships gives susan and bob a higher score (3/3=1), and puts them on the whitelist.
With paypal.com, there is only one first-degree relationship: (paypal.comme) and no secondary relationships. The algorithm handles single relationship networks as a special case, and defines them as ambiguous.
With a typical dictionary attack, a spam comes with 50 email addresses in the header. However, because a dictionary attack relies on sequential or randomly generated usernames, the number of recipients who are part of my social network is low. So we have 50 first degree relationships, and lets say the spammer gets lucky and nails Susan and Bob as well. It still gets a low score. (2/50=.04)
One can do even better by planning ahead when harvesting e-mails. For example, if you harvest a set of e-mails from a pqarticular bulliten board you can make note of message cliques at the time of harvesting, and send messages in the same groupings. for good measure you also send the addresses of the buliten board admins as well.
This is a slightly better strategy. However, this only works if you use email from a member of the clique, and limit the recipient list to members of the clique.
But there is a serious problem with the strategy. The stated goal of the authors (did you RTFA?) is to increase the costs of spamming to the point where spamming is no longer economically profitable. Such a strategy would require research which is expensive.
Or create your own loss-leader messages. For example, send out some political action alert or anything that has some vlaue or use to most people, maybe a lottery drawing for a prize, or a discount subsciption to time magazine, so they will accpet the message. the sender does not have to be the same as your spammer address. Now you spam the crap out of them while including the trojan address in the to: field.
Once again RTFA. The algorithm has nothing to do with addressbooks. But you did raise one possible threat: spoofing. A spammer could not get integrated into my social network by offering a loss-leader (for the same reason that messages from ebay.com would not be whitelisted). A spammer could spoof a member of my social network. (For example, using Bob's address.) However, the problem here is economics. Bob would probably only be auto-whitelisted by 50 people. Thus spoofing Bob would only get you access to a small population, which defeats the entire economic rationale for spamming.
Yes, but what we know about life is it is in essense the struggle to organize against the natural order of the universe to decay into chaos.
I suppose if one wants to advance a mystical theory, that would make life possible everwhere. However, that is not a scientific theory.
So life could exist anywhere in any imaginable form. One day we might be able to create "living" machines by our definition of life. Life is simply organized matter. And we haven't even explored all forms of matter in our little corner of our solar system let alone this universe.
I would argue that we have explored pretty much all the forms of matter that are abundant in the solar system. We know the properties of HCNPOS to a high degree of certainty, and we know that the properties of electron shells are tightly constrained by what apparently is fundamental laws and forces of the univese.
Thus it is quite possible to lay odds on how likely it is for life to develop in any location.
Probably because (as Asimov pointed out in a great collection of essays titled "The Tragedy of the Moon") what we know about chemistry suggests that life favors a sweet-spot of conditions. These conditions include an abundant diversity of chemicals, a reasonable temperature range, and a reasonable range of temperatures. At this sweet-spot the creation of complex molecules is probable. Outside of this sweet-spot increasingly improbable.
Re:Cemeteries are landfills
on
Space Burial
·
· Score: 4, Informative
This is true, a lot of the really old cemetaries in Europe reuse graves ever few generations. The United States has not been around quite long enough to require charnel houses (where bones are stacked to make way for new graves) except in New Orleans where the the previous occupant was pushed to the back of the crypt.
There is only so much depth you can get out of a monkey... "me is mad" and "me is looking at the pretty lady wondering how to fit my dick in there" is just about it. That has to be stretched for 100 minutes.
Interesting that you bring this up. LotR had a main character capable of only three expressions (eye wide, eye moving, eye glaring). Another Oscar contender has a horse as a central character, not a talking horse, but a horse that looks and talks like a horse.
Just like Sauron and Seabiscuit, King Kong is a plot device. The real drama happens with the people around Kong.
I think there is a lot of room for a good frankenstein movie. I don't mean the tall slow moving platfor show wearing grunt abomination. I mean an abomination like the one in the book. That has potential.
DeNiro wasn't bad if he could have gone opposite anyone else but Branagh. I find it interesting that the only things he's done lately that have not irritated the heck out of me are things like Harry Potter where he is supposed to irritate the heck out of me.
King Kong - OK, as modern weaponry gets to the point where one missle, launched from a boot 100 miles away could kill it, its a toughy. However, an expedition that goes horribly wrong has a lot of potential. Not with just the ape, but any other unusual creatures that may exist in Kongs jungle.
Well, there is that, but the other half of the story is that the forces of corporate greed know that these things are dangerous and bring it back anyway, unleashing a primal force of nature on the city.
This is going to be a big dissappointment for fans of LOTR expecting more of the same from Peter Jackson.
It seems to me that there is an inherent contradiction between the title of your post and the first sentence. Jackson made his career with edgy genre-busting films. In fact, LotR is in many ways his worst work in the last 10 years. That does not mean that it's not a fine pice of work, but compared to Dangerous Creatures and even The Frighteners the story is lacking. Really not any fault of the writers, it was just the naure of that wondrous hydra of a narrative that Tolkien created, that you really can't condense or summarize it successfully.
The nice thing about Kong (70 years old, not 50 years old), is that since it was written as a story to be told from start to finish in 100 minutes, it is the perfect size for a film adaptation. There is enough there to fill the attention of the audience for an afternoon, with enough wiggle room for Jackson to put his own stamp on it.
LotR fans are not necessarily Peter Jackson fans and Peter Jackson fans are not necessarily fans of LotR. I'd much rather see him do another Dangerous Creatures or Forgotten Silver (that managed to bamboozle fair number of New Zealaders into thinking that he really did discover that all of the major inovations we take for granted with film were invented in New Zealand and lost.) In fact, Jackson quipped that he promised Fran Walsh, his partner (professionally and personally) a low-budget, low-stress art film.
Don't forget what a flop the remake of Godzilla was.
The big problem with Godzilla is that the original Godzilla was a product of a specific time (post-WWII reconstruction) and place (Japan). Godzilla just does not translate well.
But on the other hand, we see a remake of The Wolfman about every 10 years, a remake of Dracula every 10 years. A remake of Hamlet ever generation. Why not do the Gilglamesh of monster movies, King Kong?
The protection the GPL gives you is exactly that auditing power.
However, this is not a feature unique to the GPL. It exists within any free software license, and with any company that wishes to expose their source code to an audit under a non-disclosure agreement (such as the Microsoft "shared source" license.)
The GPL ensures that you will have source you can audit. If you get GPL software from a no-name vendor and they refuse to give you the source, the GPL empowers you to take it from them with subpoenas and stuff.
Certainly, it ensures that you will have source you can audit. Whether that source is the complete source of the binaries is the question raised by the Dev X article. One can assume that if no-name vendor is willing to court criminal charges to slip a trojan horse into the code, that they would not be concerned about "subpoenas and stuff." Which as we both know is closing the barn door after all the trojan horses are out.
Well, I think there are other trust models that the Open Source movement can invoke. I trust that the treatment options given to me by my doctors are safe because the studies documenting those treatments are published for peer review. I trust my doctor even more if the treatment has been on the market for a few years.
It seems interesting that nobody argues that X-ray radiography would be safer if the methods for producing an X-ray radiograph were trade secrets held by individual companies.
That's nto the point. The point is that the agency knows they are entitled to the source for the software, so they have absolutely no reason to run the binary without checking it against the source.
There are probably dozens of reasons why an agency might not want to go through the effort of compiling a duplicate copy of the binaries from source and checking them against a distributed copy.
I certainly agree that an open-source model is safer than a closed-source model (for the same reasons that we have peer review for medical procedures.) However, this is no excuse for bad arguments. The GPL does nothing to prevent the kind of attack mentioned in the Dev X article (a group of disgruntled open-source programmers slipping a trojan horse into a binary distribution.)
Ok, he needs a lesson in reading comprehension, or he needs to hire a lawyer to interpret the GPL for him. Because as we all know, and love, the GPL requires that the source used to make the binary you have just distributed be made available to the person you gave it to...
He ignores the GPL quite blatantly here, and that is the government's insurance that the binary they run will be as secure as they can make it.
Why would a black hat, about to commit a federal offense in planting a trojan horse, be concerned about Copyright Infringement or breach of contract?
Nope, but you also cannot trust Thugs R' Us Locksmiths.
Actually, most locksmiths are bonded and advertise their bonded status. This provides stronger incentives for honesty than for breaking into your house.
Perhaps the most important difference is that the GPL is designed around one specific type of creative work while the CC licenses are intended to be generic over multiple classes of creative works. For example, the thought of applying the GPL to dramatic productions and mixed-media artwork where the concept of "source code" is problematic makes my head spin. How do you distribute the source code of an improv jaz session?
The CC does include a copyleft license known as "Share and share alike".
The collaborative filter fuzzy matches your likes/dislikes up with people who have similar likes/dislikes, then recommends stuff that doesn't overlap from those other people on the assumption that their other likes/dislikes will be close to your own.
Which completely misses my point. Collaborative filtering depends on having a critical mass interested in that piece. This locks out the niche and the local.
In addition, I disagree that Slashdot and kuro5in moderation has resulted in "high quality". What I see is that ratings on both boards are inflated to an unreasonable degree, with a high ammount of bullshit getting modded up. Even in this day of collaborative everything, there is still a very real need for expert opinions.
The good record store is an invaluable service. The best music I've been able to get this year has happened by walking up to a record store owner and asking, "hey, what do you listen to that I would not be able to find in Bloomington, IN?" The old fashioned Disk Jockey at one time used to be a very good source for recommending the best music. However, finding good DJs on the dial unfettered by corporate playlists is getting pretty hard.
What will replace them. Heck, I don't know. But I really don't see the good quality record store dying any time soon. (Note that a division of Tower or Virgin does not count as high quality.) One of the best record stores in town specializes exclusively in classical music and has managed to survive the repeated death of classical as a musical form.
Why depend on a single employee for an answer to that question when you can query millions of people for a much better answer using collaborative filtering?
Except for the fact that the "millions of people" out there like music that I personaly hate.
I don't want to know what "millions of people" like. I want to know what is good, that I won't hear on the radio because it is too niche (for example solo albums by blues harp players), too local (part of my RIAA boycott includes patronizing locally produced music), or too edgy for mass appeal.
However, if they can make a name for themselves in certain areas like The Exclusive Company has, then they will do just fine.
True, the record store has one main advantage to me that keeps me coming back: service. The best music I've managed to get recently has come from walking into a tiny little record store and saying "what do you recommend that I can't get in Bloomington, IN."
I don't like getting gadgetry as a gift. It is either too much like work, and if it is something I really really want, I'm going to be very picky about the brand, model and specifications.
My suggestion is don't spend huge quantities of money. A well-designed mix tape or CD is always a winner. A bag with a nice variety of hand picked coffee and other goodies for his taste is also good.
Don't get anything marketed as a "valentine's gift". The fun wears off after, oh, 30 minutes.
PDAs have their role, but they can be slow. Plus, I can't jot something down and tape it do a doorway or under a windshield wiper with an LCD screen.
One of the nice things about pen and paper is the wide variety of formats ranging from tiny post-its to table-sized sheets of butcher paper. Digital sketching is just plain painful.
1: The memo only applies to standardizing internal documentation for one department.
2: Courier is "time-honored" only in that it was the ubiquitous typeface for single-font devices like typewriters and ascii printers, as well as degrading nicely to dot matrix. Monospace is a pain to read in extended printed documents.
3: The article calls the new rules draconian, in spite of the fact that previously, Courier New 12 was mandated for all official documents!
True, but for a long time you were only able to use Times New Roman or similar proportional fonts if you had a professional typesetter and thousands of dollars in equipment. I suspect that Courier New was chosen back in the early days of word processing when you still had offices using a flavor of Courier on typewriters and the dreaded daisy wheel printer.
I'm probably the lone standout on this, but standardizing on a proportional serif typeface is overdue. I don't see this as a big deal in terms of money or cost. They probably just picked what a bunch of academic journals are recommending for submissions.
Slashdot Mirror
Even if it still effectively filters 50% of the remaining 75% of my email.. that's what, about 38%? SpamAssassin stops more than that before I even train its Bayesian brain. I'm also doubtful of its zero false positive claims.
The stated intent of this method (and am I the only person who has actually read the full 6-page proposal) is to reduce the computational overhead of filtering mail. The way this is intended to be used is to create whitelists and blacklists. Searching the header for a whitelist is many cases is less processor intensive that searching the whole message multiple times for spam phrases, or a fulltext baysian analysis.
In which case, it gets put into the chunk that is filtered through other means.
And from the sounds of it, what makes it different from black(or white)lists? True, it's more sophisticated because it uses the whitelists of those on your whitelists, but why not just use a plain whitelist anyway?
The actual article that describes the algorithm explicitly states that it is a way to create whitelists based on existing information from your mailbox.
And how does this allow email from internet transactions or other non-social sources through? The article didn't seem to address that so clearly.
The algorithm has a fudge factor that assigns an ambiguous score for cases where the clustering coefficient can't be calculated.
The first is trivial and certain to succeed but has a Drawback to spammers: only send e-mail to single recpients. The drawback is this puts a much higher load on their servers since every message is sent individually.
- me-bob
True this method is strongest against dictionary spam and does not work against non-dictionary spam.
[i]The second method is to always include dummy addresses in the mailing list that the recpients probably have in their address books. For example, add the following names to the to-field: notifications@paypal.com and list-notication@ebay.com.
Any recpieint that of the spam message that also has recieved e-mail from e-bay or pay-pal will trust the message.[/i]
Um, did you RTFA? (And perhaps most importantly, did anybody modding this article RTFA.)
The algorithm has nothing do do with addressbooks. Instead, it looks at friend of a friend networks as identified by mail headers.
For example, I work on a project with Bob, and Susan. A typical email message about the project will include my address, and their addresses in the header. The algorithm assumes that three first degree relationships exist:
me-bob
me-susan
susan-bob
There are also three second-degree (friend of a friend relationships.
me-susan-bob
me-bob-susan
susan
The high ratio of second-degree/first-degree relationships gives susan and bob a higher score (3/3=1), and puts them on the whitelist.
With paypal.com, there is only one first-degree relationship: (paypal.comme) and no secondary relationships. The algorithm handles single relationship networks as a special case, and defines them as ambiguous.
With a typical dictionary attack, a spam comes with 50 email addresses in the header. However, because a dictionary attack relies on sequential or randomly generated usernames, the number of recipients who are part of my social network is low. So we have 50 first degree relationships, and lets say the spammer gets lucky and nails Susan and Bob as well. It still gets a low score. (2/50=.04)
One can do even better by planning ahead when harvesting e-mails. For example, if you harvest a set of e-mails from a pqarticular bulliten board you can make note of message cliques at the time of harvesting, and send messages in the same groupings. for good measure you also send the addresses of the buliten board admins as well.
This is a slightly better strategy. However, this only works if you use email from a member of the clique, and limit the recipient list to members of the clique.
But there is a serious problem with the strategy. The stated goal of the authors (did you RTFA?) is to increase the costs of spamming to the point where spamming is no longer economically profitable. Such a strategy would require research which is expensive.
Or create your own loss-leader messages. For example, send out some political action alert or anything that has some vlaue or use to most people, maybe a lottery drawing for a prize, or a discount subsciption to time magazine, so they will accpet the message. the sender does not have to be the same as your spammer address. Now you spam the crap out of them while including the trojan address in the to: field.
Once again RTFA. The algorithm has nothing to do with addressbooks. But you did raise one possible threat: spoofing. A spammer could not get integrated into my social network by offering a loss-leader (for the same reason that messages from ebay.com would not be whitelisted). A spammer could spoof a member of my social network. (For example, using Bob's address.) However, the problem here is economics. Bob would probably only be auto-whitelisted by 50 people. Thus spoofing Bob would only get you access to a small population, which defeats the entire economic rationale for spamming.
Yes, but what we know about life is it is in essense the struggle to organize against the natural order of the universe to decay into chaos.
I suppose if one wants to advance a mystical theory, that would make life possible everwhere. However, that is not a scientific theory.
So life could exist anywhere in any imaginable form. One day we might be able to create "living" machines by our definition of life. Life is simply organized matter. And we haven't even explored all forms of matter in our little corner of our solar system let alone this universe.
I would argue that we have explored pretty much all the forms of matter that are abundant in the solar system. We know the properties of HCNPOS to a high degree of certainty, and we know that the properties of electron shells are tightly constrained by what apparently is fundamental laws and forces of the univese.
Thus it is quite possible to lay odds on how likely it is for life to develop in any location.
Probably because (as Asimov pointed out in a great collection of essays titled "The Tragedy of the Moon") what we know about chemistry suggests that life favors a sweet-spot of conditions. These conditions include an abundant diversity of chemicals, a reasonable temperature range, and a reasonable range of temperatures. At this sweet-spot the creation of complex molecules is probable. Outside of this sweet-spot increasingly improbable.
This is true, a lot of the really old cemetaries in Europe reuse graves ever few generations. The United States has not been around quite long enough to require charnel houses (where bones are stacked to make way for new graves) except in New Orleans where the the previous occupant was pushed to the back of the crypt.
There is only so much depth you can get out of a monkey ... "me is mad" and "me is looking at the pretty lady wondering how to fit my dick in there" is just about it. That has to be stretched for 100 minutes.
Interesting that you bring this up. LotR had a main character capable of only three expressions (eye wide, eye moving, eye glaring). Another Oscar contender has a horse as a central character, not a talking horse, but a horse that looks and talks like a horse.
Just like Sauron and Seabiscuit, King Kong is a plot device. The real drama happens with the people around Kong.
I think there is a lot of room for a good frankenstein movie. I don't mean the tall slow moving platfor show wearing grunt abomination.
I mean an abomination like the one in the book. That has potential.
DeNiro wasn't bad if he could have gone opposite anyone else but Branagh. I find it interesting that the only things he's done lately that have not irritated the heck out of me are things like Harry Potter where he is supposed to irritate the heck out of me.
King Kong - OK, as modern weaponry gets to the point where one missle, launched from a boot 100 miles away could kill it, its a toughy. However, an expedition that goes horribly wrong has a lot of potential. Not with just the ape, but any other unusual creatures that may exist in Kongs jungle.
Well, there is that, but the other half of the story is that the forces of corporate greed know that these things are dangerous and bring it back anyway, unleashing a primal force of nature on the city.
This is going to be a big dissappointment for fans of LOTR expecting more of the same from Peter Jackson.
It seems to me that there is an inherent contradiction between the title of your post and the first sentence. Jackson made his career with edgy genre-busting films. In fact, LotR is in many ways his worst work in the last 10 years. That does not mean that it's not a fine pice of work, but compared to Dangerous Creatures and even The Frighteners the story is lacking. Really not any fault of the writers, it was just the naure of that wondrous hydra of a narrative that Tolkien created, that you really can't condense or summarize it successfully.
The nice thing about Kong (70 years old, not 50 years old), is that since it was written as a story to be told from start to finish in 100 minutes, it is the perfect size for a film adaptation. There is enough there to fill the attention of the audience for an afternoon, with enough wiggle room for Jackson to put his own stamp on it.
LotR fans are not necessarily Peter Jackson fans and Peter Jackson fans are not necessarily fans of LotR. I'd much rather see him do another Dangerous Creatures or Forgotten Silver (that managed to bamboozle fair number of New Zealaders into thinking that he really did discover that all of the major inovations we take for granted with film were invented in New Zealand and lost.) In fact, Jackson quipped that he promised Fran Walsh, his partner (professionally and personally) a low-budget, low-stress art film.
Don't forget what a flop the remake of Godzilla was.
The big problem with Godzilla is that the original Godzilla was a product of a specific time (post-WWII reconstruction) and place (Japan). Godzilla just does not translate well.
But on the other hand, we see a remake of The Wolfman about every 10 years, a remake of Dracula every 10 years. A remake of Hamlet ever generation. Why not do the Gilglamesh of monster movies, King Kong?
The protection the GPL gives you is exactly that auditing power.
However, this is not a feature unique to the GPL. It exists within any free software license, and with any company that wishes to expose their source code to an audit under a non-disclosure agreement (such as the Microsoft "shared source" license.)
The GPL ensures that you will have source you can audit. If you get GPL software from a no-name vendor and they refuse to give you the source, the GPL empowers you to take it from them with subpoenas and stuff.
Certainly, it ensures that you will have source you can audit. Whether that source is the complete source of the binaries is the question raised by the Dev X article. One can assume that if no-name vendor is willing to court criminal charges to slip a trojan horse into the code, that they would not be concerned about "subpoenas and stuff." Which as we both know is closing the barn door after all the trojan horses are out.
Well, I think there are other trust models that the Open Source movement can invoke. I trust that the treatment options given to me by my doctors are safe because the studies documenting those treatments are published for peer review. I trust my doctor even more if the treatment has been on the market for a few years.
It seems interesting that nobody argues that X-ray radiography would be safer if the methods for producing an X-ray radiograph were trade secrets held by individual companies.
That's nto the point. The point is that the agency knows they are entitled to the source for the software, so they have absolutely no reason to run the binary without checking it against the source.
There are probably dozens of reasons why an agency might not want to go through the effort of compiling a duplicate copy of the binaries from source and checking them against a distributed copy.
I certainly agree that an open-source model is safer than a closed-source model (for the same reasons that we have peer review for medical procedures.) However, this is no excuse for bad arguments. The GPL does nothing to prevent the kind of attack mentioned in the Dev X article (a group of disgruntled open-source programmers slipping a trojan horse into a binary distribution.)
Ok, he needs a lesson in reading comprehension, or he needs to hire a lawyer to interpret the GPL for him. Because as we all know, and love, the GPL requires that the source used to make the binary you have just distributed be made available to the person you gave it to...
He ignores the GPL quite blatantly here, and that is the government's insurance that the binary they run will be as secure as they can make it.
Why would a black hat, about to commit a federal offense in planting a trojan horse, be concerned about Copyright Infringement or breach of contract?
Nope, but you also cannot trust Thugs R' Us Locksmiths.
Actually, most locksmiths are bonded and advertise their bonded status. This provides stronger incentives for honesty than for breaking into your house.
I think that this exchange (and many others just like it) highlights that freedom of choice is important.
Perhaps the most important difference is that the GPL is designed around one specific type of creative work while the CC licenses are intended to be generic over multiple classes of creative works. For example, the thought of applying the GPL to dramatic productions and mixed-media artwork where the concept of "source code" is problematic makes my head spin. How do you distribute the source code of an improv jaz session?
The CC does include a copyleft license known as "Share and share alike".
The collaborative filter fuzzy matches your likes/dislikes up with people who have similar likes/dislikes, then recommends stuff that doesn't overlap from those other people on the assumption that their other likes/dislikes will be close to your own.
Which completely misses my point. Collaborative filtering depends on having a critical mass interested in that piece. This locks out the niche and the local.
In addition, I disagree that Slashdot and kuro5in moderation has resulted in "high quality". What I see is that ratings on both boards are inflated to an unreasonable degree, with a high ammount of bullshit getting modded up. Even in this day of collaborative everything, there is still a very real need for expert opinions.
The good record store is an invaluable service. The best music I've been able to get this year has happened by walking up to a record store owner and asking, "hey, what do you listen to that I would not be able to find in Bloomington, IN?" The old fashioned Disk Jockey at one time used to be a very good source for recommending the best music. However, finding good DJs on the dial unfettered by corporate playlists is getting pretty hard.
What will replace them. Heck, I don't know. But I really don't see the good quality record store dying any time soon. (Note that a division of Tower or Virgin does not count as high quality.) One of the best record stores in town specializes exclusively in classical music and has managed to survive the repeated death of classical as a musical form.
Why depend on a single employee for an answer to that question when you can query millions of people for a much better answer using collaborative filtering?
Except for the fact that the "millions of people" out there like music that I personaly hate.
I don't want to know what "millions of people" like. I want to know what is good, that I won't hear on the radio because it is too niche (for example solo albums by blues harp players), too local (part of my RIAA boycott includes patronizing locally produced music), or too edgy for mass appeal.
However, if they can make a name for themselves in certain areas like The Exclusive Company has, then they will do just fine.
True, the record store has one main advantage to me that keeps me coming back: service. The best music I've managed to get recently has come from walking into a tiny little record store and saying "what do you recommend that I can't get in Bloomington, IN."
I don't like getting gadgetry as a gift. It is either too much like work, and if it is something I really really want, I'm going to be very picky about the brand, model and specifications.
My suggestion is don't spend huge quantities of money. A well-designed mix tape or CD is always a winner. A bag with a nice variety of hand picked coffee and other goodies for his taste is also good.
Don't get anything marketed as a "valentine's gift". The fun wears off after, oh, 30 minutes.
PDAs have their role, but they can be slow. Plus, I can't jot something down and tape it do a doorway or under a windshield wiper with an LCD screen.
One of the nice things about pen and paper is the wide variety of formats ranging from tiny post-its to table-sized sheets of butcher paper. Digital sketching is just plain painful.
Hasn't died yet, but it's coming. Film will be relegated to the fine arts only, next to oil paints and lithographs.
I wouldn't say that oil painting and lithography is dead either.
"US bans time-honored typeface"
What the heck?
1: The memo only applies to standardizing internal documentation for one department.
2: Courier is "time-honored" only in that it was the ubiquitous typeface for single-font devices like typewriters and ascii printers, as well as degrading nicely to dot matrix. Monospace is a pain to read in extended printed documents.
3: The article calls the new rules draconian, in spite of the fact that previously, Courier New 12 was mandated for all official documents!
True, but for a long time you were only able to use Times New Roman or similar proportional fonts if you had a professional typesetter and thousands of dollars in equipment. I suspect that Courier New was chosen back in the early days of word processing when you still had offices using a flavor of Courier on typewriters and the dreaded daisy wheel printer.
I'm probably the lone standout on this, but standardizing on a proportional serif typeface is overdue. I don't see this as a big deal in terms of money or cost. They probably just picked what a bunch of academic journals are recommending for submissions.