They meant 'research' in the writer's sense - as in you read as much as you can to research a background to back up your fiction.
Research does not exclusively mean scientific or experimental research. E.g.: Plenty of scholarly research was done before the scientific method was even defined.
Though I wouldn't compare fiction writers' research to either, it is a perfectly valid use of the word.
What kind of work were those 10K req/sec on your own custom server doing? Was it a standard db-backed web app, or something more specialized and computationally intensive?
Not that I doubt the difference you saw - but I'm still skeptical of the 10:1 factor as applied to Facebook servers, which seem relatively standard webapp cycle (request -> datastore lookup -> html), *just from the programming language*.
Admittedly, I don't do PHP, so the language could be as bad and impossible to scale as you claim. But from their architecture description, I really doubt the request spends enough time (>= 90%?) in PHP computation to make *any* change there translate into a 10X improvement.
At least on my own experience, once you get into the '~1K req/sec' scenario on that type of webapp, the middle-tier code is rarely your main perf headache. You spend more of your time ensuring your data sources (sql+cache + any other services) keep up, and your middle-tier code spends much of its time waiting for whatever went out the socket to come back. There is always some perf improvement to make on the middle-tier, but if the request spends 90% of its time on the presentation layer... well, that's usually a perf bug, not by design.
There are good reasons to justify the cost of switching away from PHP - and they seem to be aware of them. But order-of-magnitude perf improvements would more likely result from architectural and datastore improvements - and sensibly enough, that seems where their efforts have focused.
To be honest, I always thought the real problem with the paywall was Salon's extremely high opinion of itself.
Their failure doesn't prove people will not pay for content - it just proved people would not pay for Salon.com content. That fact was not surprising to anyone but Salon itself.
Back in ye olde days, they had *much* better content than they have now - at least it was on my own daily website-checking browsing habits (incidentally, right before slashdot).
But their site never made sense as Paid Content: 99% opinion, ranging from reviews and pop culture commentary to self-indulgent political rants passing for analysis - all entertaining and very well written, but nothing I could live without. It's not like there was a shortage on the web of narcissistic blogs providing biased commentary on current events.
Their absurd advertising for their Premium subscription didn't help: asking you to subscribe 'save democracy and independent journalism', etc. It came off as arrogant and pretty much forced the reader to ask themselves whether it was really important journalism, something that made the world better, or a simple luxury they could live without.
Most people realized they didn't care that much about what Salon.com wanted to say - not more than subscribing to your local paper, or many better uses for 40 bucks a year.
By that explanation it sounds like an even worse idea. Why not use a foundation or any other donation schemes to fund open source projects?
This is a bad analogy made into an absurdity of a mission statement: what is the causal relation? how does this address the problem it claims to address? If you're offsetting your "bad code" to fund creating *new* open source code, who buys the offsets for any bad code from there?
The metaphor is already broken because "bad code" is not as objectively quantifiable as CO2 metric tons, but even assuming you can average that out ('25% of all code is bad')...
The point of buying Carbon Offsets is that the funds can be invested on cleaning up the carbon problem. This is like investing the "Carbon Offset" from a black shoes factory into expanding a red shoes factory. As much as you like red shoes, you're just increasing the total Carbon.
that's not surprising since iterating over a linq expression is a function evaluation - but how does that relate to dynamic typing?
afaik, both var and linq expressions are as statically typed as any ML function. just because it saves you some typing the obvious (or the inferable), doesn't mean things are not resolved at compile time.
His point is obvious enough to anyone who bothers to read the first two paragraphs: hiring *only* programmers who spent their free time coding is an absurd criteria - which may seem reasonable to kids right out of college, because they assume 'spare time' is and will be an abundant resource in their life.
You are arguing from a fallacious point. You assume that people who have never been held to a schedule in one area of their life are nonetheless ignorant of what a schedule is, and that they have never had to follow one. I sure hope you're not in charge of hiring anywhere.
>
I guess I must be feeding a troll, but I can't help but be a bit impressed: each individual sentence in this paragraph is a non-sequitur...
- I'm not arguing *Dziuba's point* (fallacious or not). "His point is obvious enough..." should be, well, obvious enough.
- Even for his argument, you haven't pointed out any actual fallacy on Dziuba's point.
There is no argument in neither his blog, nor on this thread, about schedules, task management, accountability or anything that could related to the supposed "assumption" you're talking about. It's an argument about work/life balance - life being, by definition, that which is outside of the schedule.
- *My* point was that if people are going to get that bothered by an argument, they should at least read it rather than rant against an strawman. So I don't know whether to be annoyed or validated by your reply.
This guy, though: He's like a professional, career-oriented brick mason, who sits around watching his 150-year-old red brick house crumble around him, while loudly proclaiming "I don't do masonry in my free time. So suck it, fellas!"
Not that I would disagree with the rest of your post otherwise, but I'm not sure which 'this guy' you're talking about... Dziuba's blog doesn't fit the description above by any stretch. His point is obvious enough to anyone who bothers to read the first two paragraphs: hiring *only* programmers who spent their free time coding is an absurd criteria - which may seem reasonable to kids right out of college, because they assume 'spare time' is and will be an abundant resource in their life.
This seems to be a more typical case of the Slashdot summary having nothing to do with the linked article - and the Slashdot editor not bothering to even click on the link before posting.
Seconded: both on the terminology and using short stories.
Almost everywhere else in the world it's really all called "speculative fiction" for good reason - it is both more descriptive and more general than 'sci-fi'. You're also likely to find useful academic material and references under that label - which helps to put the material in context and connection with other literature. E.g.: Aldous Huxley, Jorge Luis Borges, Orwell all wrote 'sci-fi', but because they were 'serious writers' you'll probably not find those works under that heading.
I'd also strongly recommend focusing on short stories over novels (although not exclusively) - because they tend to be more effective in literature courses and workshops. Short stories just provide more opportunity to read different plot ideas, styles and approaches and discuss and compare them.
They *also* level the field for students with different interests: literature is subjective, and short stories mean more chances you'll find the type of fiction you love and understand better than anyone, and soon enough to motivate you. I remember many a highschool classmate who decided they didn't like reading fiction because they happened to hate the first 1-2 novels they had to read in class - and didn't like the classes because they were monopolized by the few who were *really* into those novels.
In contrast, in college workshops we used mostly short stories, which anyone could be expected to read within a day - so if someone found a story insufferably boring, they'd still finish it and participate in informed discussions about *why* they hated it and how they compared to other fiction... instead of giving up and dozing in class waiting for it all to end.
More to the point, an electronic copy is a sale to a different demographic...
The price of the printed books precludes individuals establishing significant libraries - no such restriction necessarily hinders the digital realm... and, hence, the potential market size should be considered when the price is set.
I think you have a bigger point there outside of the textbooks scenario.
When I grew up my family had a decent-size personal library, so I got into the habit of picking up books for reading with some frequency. Once I was living by myself I quickly realized that, after a few dozen, books have a larger ongoing cost besides their price of acquisition: in terms of space and preventing deterioration they're quite expensive.
They're heavy, require special furniture for storage, take up a lot of room and are vulnerable to heat, humidity, bugs, animals and all sorts of other environmental factors. And they're few specific editions are commodatized enough to be truly disposable (well, unless they're really bad), or rare enough to amortize the overhead.
Textbooks and reference material at least have the virtue of being a quantifiable investment - i.e.: you usually only *need* a few of those around if you have a library around, and they're expensive (and resaleable) enough, and reused enough, to be worth the space and care. So I don't see those changing that much. But for most people most of their *reading* material, both by quantity of books and frequency of use, does not meet that description.
Over time, I suspect physical books will inevitably be moving back to their original social place as cultural collectibles or luxury items. That doesn't mean people wouldn't buy them for aesthetics or personal reasons, or that the prices will be unreasonable - but like music records, mechanical watches and fountain pens, they're vulnerable to be displaced by the raw efficiency of the alternatives.
The 'problem' will not be that people buying paperbacks at 10-20 bucks today will demand a better e-book value. It's that most of the people willing to buy and read each e-book - at whatever price the market bears - wouldn't have considered buying each of those physical books in the first place. Once your demographic changes that much, your old price differentation stops working and you need to price for the new market (or your competitors will).
the problem is that the waste of time is not a function of what *you* have nothing meaningful to say, but whether *everyone in simultaneous communication with you* does.
in traditional communication tools, physical constraints and built-in inefficiencies act as noise filters. only one person can waste your time at once.
If the bus service near where you live is a waste of time that might be because you live in a place that can't be well-served by public transit.
Or lives in a place that is *underserved* by public transit - as is the case for many US cities.
There is an issue of critical mass with public transportation - gradual adoption doesn't make sense. Most people don't use it because they *can't* use it, because the routes are too few, inconvenient and unreliable to depend on them. But once you reach critical coverage on an area, and you don't have to wait >=1 hour for the bus anymore - things are *qualitatively* different and you have a chance to scale.
I understand your point that some places are too sparsely populated to make it cost effective. But the argument that you need a focal downtown and high density frankly doesn't make sense - many places in this planet don't match that description, and yet 'public' transportation is both omnipresent and effective far into the suburbs and small towns.
I put 'public' in quotes because often it is a mix of private and government-funded mass transit. When there is no public monopoly, it's often easier for small entrepeneurs to extend the official transit network into underserved areas at a smaller scale, for a small profit margin - since they don't have to deal with the politics (or the guarantees of service).
You may just be the only one who still bothers to read the article hoping for useful info.
New video card tech reviews are, almost always, all about vicarious genital measurements. Benchmarks for FPS, raw computational capacity, shader support, etc. all abound - as if it were 1996 and the high-end was still competing for mere adequacy.
It's not that 'the 1337' have taken over the tech evolution, it's that they're the only readers left for those publications as their focus became less relevant for the normal market.
We were all there back then - building the machines, overclocking the cards, buying the extra cooling fans - because you needed to get those extra FPS to get an *adequate* gaming experience (and to admit to friends & family what you spend 300-400 bucks with a straight face). But for the last few years, any mid-range card can handle practically any new games and media tasks pretty well.
I still consider myself a PC gamer, but at this point I've probably changed my video card ~4 times without looking for a raw perf improvement. Today, my top 5 video card priorities are very different:
1) Is it relatively quiet? Is the background noise *not* comparable to a 747? 2) Is it small enough to fit on a case without having to design with a freaking airflow plan? Or does it come with a built-in ginormous fan, and take more space than the motherboard? 3) Is it *really* quiet? i.e.: Can I add the card without increasing the base noise level of a well ventilated PC? 4) Does the ratio of price vs qualitative difference not seem utterly ridiculous? 5) If it happens to be a bit faster than the current model - that'd be nice.
I've always been skeptical of the claims about curing diseases - which is what I really understood as reflexology - because I haven't seen any consistent evidence.
But the idea that applying localized pressure relieves pain seems sensible to me, and matches my own anecdotical experience (anecdotical by population size, but has been consistent). For that matter, so does a good movie or a deep conversation.
Focused sensation distracts from 'normal' sensation, and it is easier to focus on sharp, local stimuli over dull, distributed discomfort or pain. Never thought it had much to do with reflexology - although I could see how it could support its less ambitious claims.
"queue" in joke about OO responsiveness, Java dependencies, and the validity of the original statement.
I kid, I kid...
I haven't really been actively using OO for quite a few years, I don't know if that criticism is still as frequent - or valid - as it was back in ye olde days.
It's worth noting in this context that if you've ever seen the movie "Stand and Deliver", one way Jaime Escalante got those great results was by giving his students access to self-paced homeschooling math workbooks rather than using the standard school-provided curriculum.
Yes, but another crucial thing he did was to *push* those students beyond their own expectations to learn things they didn't think themselves capable of learning.
It's actually a great example story - and matches what I've seen from the very few great teachers I've met. Self-motivation is critical, but you need experience to be able to set to goals, and some empathy and art to keep students moving towards those goals even when their motivation wanes or they have the wrong perspective (from inexperience).
The quote that comes to mind is 'you only see the turn, you don't see the road ahead'.
A purely self-directed, self-motivated student is more likely to give up on the first real challenge - because they can't realize that, as much 'natural learner' as anyone can be, you're always going to find something that is *hard* for you to learn and understand. And that it is *normal*, that with some persistence they'll learn it, and that very often this is important for their future.
The problem is that the GP is also 100% correct - lots of college curricula pretend to cover team-work, but few courses (if any) can teach effective team dynamics. Actually, they unintentionally teach the opposite, as the parent described.
A good experience for team dynamics needs attributes that are omnipresent at work, but are difficult to have in project course-work:
- Project size must *require* a team - not be easily achievable by a single person. - Work needs diverse skills and complementary roles in equal(ish) measure - Results persistence after the project - in the real world, your initial 'grade' is almost irrelevant, long-term quality is what matters. - Social persistence after the project - in the real world, you'll see and work with this people every day long after a single task is done.
The typical "team project" crammed on half a semester course (or less) is the antithesis of this: too small and too short, heavily biased towards one-two skills (so everything else are 'slacker' tasks), and you can choose to never see your teammates again after completion.
Of course, if you can more easily complete everything by yourself by hacking it all together over a few weeks - you would be a fool to do otherwise. So most smart people end up doing exactly that, because what matters is the grade.
But in the workplace, most Real Work *requires* teamwork because it is simply bigger, more complex, and requires more complementary skills and expertise than an artificial CS assignment. And you cannot piss off, or even under-utilize, your peers without burning important bridges, because you'll typically work with the same people over *years*, not weeks. And you need to depend on peer feedback, and on people with complementary roles and skills you don't have (and do require a lot of work) - because a feature gap or quality issue can follow you for a long time.
Sadly, typical college projects seem to train the smartest students to be lone programmers - try to do everything themselves, assume theirs are the only 'real work' skills and they're the one indispensable worker in the operation... and if that doesn't scale, it must be the 'assignment' was broken and doomed anyway.
Fortunately most people learn some teamwork somewhere, but it doesn't seem to be through college.
Exception *might* be team sports... I've never been a big fan of sports, neither practice nor spectator, and before working on the Real World always thought the whole 'teaches teamwork' idea highly overrated. But it does seem to have characteristics missing from course projects, and (anecdotically) I've noticed people with that background tend to grok some of the team dynamics and social subtext more easily than me (which is admittedly not a high bar).
Trying to map the "Security through obscurity..." truism to this context ignores the original meaning, and gets silly pretty quickly.
Real-life security is simply the probability of not being the victim of a successful attack.
In cryptography, the "Security through obscurity is no security" applies to the specific case of making the security system *opaque* - hiding design and implementation of the lock has limited benefits against a determined attacker, shrinking rapidly to zero as more information and computation capacity is available. It assumes security is a binary concept instead of a probabilistic one because *in this specific context* it's a good approximation.
That's completely besides the point here, as in many real-life situations - this is not about the lock design, or the key, or leaving the door unlocked in the first place. It's not about whether the house is "Secure" by some academic definition against an hypothetical Oscar or Mallory determined to attack Mr.X's security. Here there's only Mr.X and the other millions of people with access to the Inter-webs who didn't know he existed in the first place.
The problem is about how Mr.X *becomes a target* in the first place - whether he's a cop or a typical private citizen. Compiling freely available information does more than caching computation - it *singles-out* Mr.X and makes him a visible target, indexed by search-able criteria, more easily accessible to people who wouldn't have bothered to look for Mr.X before.
That may not change the total security profile of the raw data - but it absolutely makes Mr.X *specifically* more exposed. Being a top-10 Google result for "cop off-duty personal address" is simply *not* the same thing as having your info available in principle, across several databases, as an extra record among thousands or millions of entries.
You can't just yell "Fire!" in a crowded theater to make an argument about fire safety codes, and then claim no responsibility because you didn't make the building, and 'someone else could do the same anyway'.
Technically, so can the stalking ex-girlfriend - or the paranoid employer who thinks the cashier is stealing, or the obsessive neighbor thinking anyone new to the neighborhood is a drug-dealer / serial-killer.
But it usually takes more than that argument to call it "investigative reporting" instead of harassment when they start monitoring 24/7.
You also gotta love the interest groups that have sprung up around the issue. MADD has morphed over the years from an organization with a laudable enough goal (reduce drunk driving deaths) into a neo-prohibitionist organization that is waging a war on all drinking. If they had their way, booze would be taxed at a higher rate than tobbaco and every car sold in the US would have an ignition interlock system.
Dude, I sort of agree with you, but that's hardly the definition of neo-prohibitionist behavior.
From a purely health/social perspective, alcohol *should* be taxed at a higher rate than tobacco. There are already laws against public smoking in most places with their own fines, so the taxed risks are for the smoker's health costs. Alcohol has both long-term health costs for the drinker, *and* the risks to others from coordination impairment.
The simple reason that doesn't happen is because there are more of us drinkers than there are smoking voters. That's not bad from my POV (I enjoy a drink, or ten, don't smoke) - but to expect the opposite is hardly irrational.
Not really - if anything, it removes the ability to take the code and redistribute for free on *Apple's platform* (without taking a loss).
But that's an Apple AppStore problem (or a 'feature') as a distribution channel for any software - not an issue with the GPL or what the authors are doing.
They seem to be doing the right thing with the code - they're essentially passing on the distribution costs, much like they'd have to do if they printed it on CDs for retail.
IFF Apple were to open their platform to other distribution channels, then automatically this would be a non-issue. That is unlikely to happen, but since their GLP-ed code is available, perhaps it could be used in the future to port it to other rich Touch interfaces that may not have such restrictions.
If someone at a company decides to use this tool, unbeknownst to the company and the other party is also using it, then the email becoming garbled and eventually deleted could become a problem should the company ever go to court.
Would that be different than if someone decided to use normal cryptography, and encrypt their emails to another party unbeknownst to the same company?
In both cases, the company would not be able to provide the cryptographic key to decipher the messages.
I'm getting the impression most people are looking at this as a solution to traditional crypto problems (corporate assets, legal docs, DRM), where it doesn't seem that relevant.
All of those can be considered valuable assets - in the sense that there is a concrete monetary value attached to them, in terms of business value, liability, or expected profits from sale. 'Vanishing' *valuable* data from the original source doesn't seem terribly useful, because if it is considered *valuable* it is likely to be copied/backed-up... and each recipient increases that likelihood.
IFF the information is very valuable, the usual security concerns about protecting against a resourceful untrusted party apply... and since the people with the key are trusted they can copy the data to a non-vanishing target anyway and break the loop. The more valuable, the more likley it will happen for non-malicious reasons, precisely because it will 'vanish'.
Where this seems useful is on large sets of *cheap* individual data more commonly shared online, for which individuals currently lack a good sense of value-as-privacy.
In general, we don't have yet a good instinct to preserve privacy online, at least partly because people are not used to communication networks being both *frictionless* and *persistent*.
If just a single assumption was broken, it wouldn't be a big deal... but the game has changed because *both* of these now occur transparently online.
We do not treat most personal data as information assets, nor do we really anticipate their aggregation, because for millenia *trivial* information would erode over time due to natural processes... people forget, business cards get lost, post-it notes become unreadable, etc. And in the context of that single communication, we could trust our estimate of how trivial that information was for both parties.
We're used to some erosion of information in real life - if in the middle of an in-person conversation we share freely a phone number or a mail address or some personal story with a friend, we don't expect that to be readable for millions of unknown people, or for it to be easily queryable years later, whenever anyone decides... for free.
This tech seems valuable precisely for that scenario because it resets the result of an online communication to those expectations, so we don't have to choose between giving up the idea of privacy or treating everything we say on casual conversation as some kind of 'information asset'.
Slashdot Mirror
They meant 'research' in the writer's sense - as in you read as much as you can to research a background to back up your fiction.
Research does not exclusively mean scientific or experimental research. E.g.: Plenty of scholarly research was done before the scientific method was even defined.
Though I wouldn't compare fiction writers' research to either, it is a perfectly valid use of the word.
Alternatively, you could just use a fake book cover to read something that really applies to your role.
For example: http://despair.com/manageredition.html
What kind of work were those 10K req/sec on your own custom server doing? Was it a standard db-backed web app, or something more specialized and computationally intensive?
Not that I doubt the difference you saw - but I'm still skeptical of the 10:1 factor as applied to Facebook servers, which seem relatively standard webapp cycle (request -> datastore lookup -> html), *just from the programming language*.
Admittedly, I don't do PHP, so the language could be as bad and impossible to scale as you claim. But from their architecture description, I really doubt the request spends enough time (>= 90%?) in PHP computation to make *any* change there translate into a 10X improvement.
At least on my own experience, once you get into the '~1K req/sec' scenario on that type of webapp, the middle-tier code is rarely your main perf headache. You spend more of your time ensuring your data sources (sql+cache + any other services) keep up, and your middle-tier code spends much of its time waiting for whatever went out the socket to come back. There is always some perf improvement to make on the middle-tier, but if the request spends 90% of its time on the presentation layer... well, that's usually a perf bug, not by design.
There are good reasons to justify the cost of switching away from PHP - and they seem to be aware of them. But order-of-magnitude perf improvements would more likely result from architectural and datastore improvements - and sensibly enough, that seems where their efforts have focused.
To be honest, I always thought the real problem with the paywall was Salon's extremely high opinion of itself.
Their failure doesn't prove people will not pay for content - it just proved people would not pay for Salon.com content. That fact was not surprising to anyone but Salon itself.
Back in ye olde days, they had *much* better content than they have now - at least it was on my own daily website-checking browsing habits (incidentally, right before slashdot).
But their site never made sense as Paid Content: 99% opinion, ranging from reviews and pop culture commentary to self-indulgent political rants passing for analysis - all entertaining and very well written, but nothing I could live without. It's not like there was a shortage on the web of narcissistic blogs providing biased commentary on current events.
Their absurd advertising for their Premium subscription didn't help: asking you to subscribe 'save democracy and independent journalism', etc. It came off as arrogant and pretty much forced the reader to ask themselves whether it was really important journalism, something that made the world better, or a simple luxury they could live without.
Most people realized they didn't care that much about what Salon.com wanted to say - not more than subscribing to your local paper, or many better uses for 40 bucks a year.
By that explanation it sounds like an even worse idea. Why not use a foundation or any other donation schemes to fund open source projects?
This is a bad analogy made into an absurdity of a mission statement: what is the causal relation? how does this address the problem it claims to address? If you're offsetting your "bad code" to fund creating *new* open source code, who buys the offsets for any bad code from there?
The metaphor is already broken because "bad code" is not as objectively quantifiable as CO2 metric tons, but even assuming you can average that out ('25% of all code is bad')...
The point of buying Carbon Offsets is that the funds can be invested on cleaning up the carbon problem. This is like investing the "Carbon Offset" from a black shoes factory into expanding a red shoes factory. As much as you like red shoes, you're just increasing the total Carbon.
that's not surprising since iterating over a linq expression is a function evaluation - but how does that relate to dynamic typing?
afaik, both var and linq expressions are as statically typed as any ML function. just because it saves you some typing the obvious (or the inferable), doesn't mean things are not resolved at compile time.
Do you mean Lisp without packages or any other language support for real modularity?
There are many neat things about javascript. Scaling to large codebases is not one of those things.
His point is obvious enough to anyone who bothers to read the first two paragraphs: hiring *only* programmers who spent their free time coding is an absurd criteria - which may seem reasonable to kids right out of college, because they assume 'spare time' is and will be an abundant resource in their life.
You are arguing from a fallacious point. You assume that people who have never been held to a schedule in one area of their life are nonetheless ignorant of what a schedule is, and that they have never had to follow one. I sure hope you're not in charge of hiring anywhere.
>
I guess I must be feeding a troll, but I can't help but be a bit impressed: each individual sentence in this paragraph is a non-sequitur...
- I'm not arguing *Dziuba's point* (fallacious or not). "His point is obvious enough..." should be, well, obvious enough.
- Even for his argument, you haven't pointed out any actual fallacy on Dziuba's point.
There is no argument in neither his blog, nor on this thread, about schedules, task management, accountability or anything that could related to the supposed "assumption" you're talking about. It's an argument about work/life balance - life being, by definition, that which is outside of the schedule.
- *My* point was that if people are going to get that bothered by an argument, they should at least read it rather than rant against an strawman. So I don't know whether to be annoyed or validated by your reply.
Not that I would disagree with the rest of your post otherwise, but I'm not sure which 'this guy' you're talking about... Dziuba's blog doesn't fit the description above by any stretch. His point is obvious enough to anyone who bothers to read the first two paragraphs: hiring *only* programmers who spent their free time coding is an absurd criteria - which may seem reasonable to kids right out of college, because they assume 'spare time' is and will be an abundant resource in their life.
This seems to be a more typical case of the Slashdot summary having nothing to do with the linked article - and the Slashdot editor not bothering to even click on the link before posting.
Seconded: both on the terminology and using short stories.
Almost everywhere else in the world it's really all called "speculative fiction" for good reason - it is both more descriptive and more general than 'sci-fi'. You're also likely to find useful academic material and references under that label - which helps to put the material in context and connection with other literature. E.g.: Aldous Huxley, Jorge Luis Borges, Orwell all wrote 'sci-fi', but because they were 'serious writers' you'll probably not find those works under that heading.
I'd also strongly recommend focusing on short stories over novels (although not exclusively) - because they tend to be more effective in literature courses and workshops. Short stories just provide more opportunity to read different plot ideas, styles and approaches and discuss and compare them.
They *also* level the field for students with different interests: literature is subjective, and short stories mean more chances you'll find the type of fiction you love and understand better than anyone, and soon enough to motivate you. I remember many a highschool classmate who decided they didn't like reading fiction because they happened to hate the first 1-2 novels they had to read in class - and didn't like the classes because they were monopolized by the few who were *really* into those novels.
In contrast, in college workshops we used mostly short stories, which anyone could be expected to read within a day - so if someone found a story insufferably boring, they'd still finish it and participate in informed discussions about *why* they hated it and how they compared to other fiction... instead of giving up and dozing in class waiting for it all to end.
I think you have a bigger point there outside of the textbooks scenario.
When I grew up my family had a decent-size personal library, so I got into the habit of picking up books for reading with some frequency. Once I was living by myself I quickly realized that, after a few dozen, books have a larger ongoing cost besides their price of acquisition: in terms of space and preventing deterioration they're quite expensive.
They're heavy, require special furniture for storage, take up a lot of room and are vulnerable to heat, humidity, bugs, animals and all sorts of other environmental factors. And they're few specific editions are commodatized enough to be truly disposable (well, unless they're really bad), or rare enough to amortize the overhead.
Textbooks and reference material at least have the virtue of being a quantifiable investment - i.e.: you usually only *need* a few of those around if you have a library around, and they're expensive (and resaleable) enough, and reused enough, to be worth the space and care. So I don't see those changing that much. But for most people most of their *reading* material, both by quantity of books and frequency of use, does not meet that description.
Over time, I suspect physical books will inevitably be moving back to their original social place as cultural collectibles or luxury items. That doesn't mean people wouldn't buy them for aesthetics or personal reasons, or that the prices will be unreasonable - but like music records, mechanical watches and fountain pens, they're vulnerable to be displaced by the raw efficiency of the alternatives.
The 'problem' will not be that people buying paperbacks at 10-20 bucks today will demand a better e-book value. It's that most of the people willing to buy and read each e-book - at whatever price the market bears - wouldn't have considered buying each of those physical books in the first place. Once your demographic changes that much, your old price differentation stops working and you need to price for the new market (or your competitors will).
the problem is that the waste of time is not a function of what *you* have nothing meaningful to say, but whether *everyone in simultaneous communication with you* does.
in traditional communication tools, physical constraints and built-in inefficiencies act as noise filters. only one person can waste your time at once.
Or lives in a place that is *underserved* by public transit - as is the case for many US cities.
There is an issue of critical mass with public transportation - gradual adoption doesn't make sense. Most people don't use it because they *can't* use it, because the routes are too few, inconvenient and unreliable to depend on them. But once you reach critical coverage on an area, and you don't have to wait >=1 hour for the bus anymore - things are *qualitatively* different and you have a chance to scale.
I understand your point that some places are too sparsely populated to make it cost effective. But the argument that you need a focal downtown and high density frankly doesn't make sense - many places in this planet don't match that description, and yet 'public' transportation is both omnipresent and effective far into the suburbs and small towns.
I put 'public' in quotes because often it is a mix of private and government-funded mass transit. When there is no public monopoly, it's often easier for small entrepeneurs to extend the official transit network into underserved areas at a smaller scale, for a small profit margin - since they don't have to deal with the politics (or the guarantees of service).
You may just be the only one who still bothers to read the article hoping for useful info.
New video card tech reviews are, almost always, all about vicarious genital measurements. Benchmarks for FPS, raw computational capacity, shader support, etc. all abound - as if it were 1996 and the high-end was still competing for mere adequacy.
It's not that 'the 1337' have taken over the tech evolution, it's that they're the only readers left for those publications as their focus became less relevant for the normal market.
We were all there back then - building the machines, overclocking the cards, buying the extra cooling fans - because you needed to get those extra FPS to get an *adequate* gaming experience (and to admit to friends & family what you spend 300-400 bucks with a straight face). But for the last few years, any mid-range card can handle practically any new games and media tasks pretty well.
I still consider myself a PC gamer, but at this point I've probably changed my video card ~4 times without looking for a raw perf improvement. Today, my top 5 video card priorities are very different:
1) Is it relatively quiet? Is the background noise *not* comparable to a 747?
2) Is it small enough to fit on a case without having to design with a freaking airflow plan? Or does it come with a built-in ginormous fan, and take more space than the motherboard?
3) Is it *really* quiet? i.e.: Can I add the card without increasing the base noise level of a well ventilated PC?
4) Does the ratio of price vs qualitative difference not seem utterly ridiculous?
5) If it happens to be a bit faster than the current model - that'd be nice.
Is the 'pain' part really reflexology?
I've always been skeptical of the claims about curing diseases - which is what I really understood as reflexology - because I haven't seen any consistent evidence.
But the idea that applying localized pressure relieves pain seems sensible to me, and matches my own anecdotical experience (anecdotical by population size, but has been consistent). For that matter, so does a good movie or a deep conversation.
Focused sensation distracts from 'normal' sensation, and it is easier to focus on sharp, local stimuli over dull, distributed discomfort or pain. Never thought it had much to do with reflexology - although I could see how it could support its less ambitious claims.
"queue" in joke about OO responsiveness, Java dependencies, and the validity of the original statement.
I kid, I kid...
I haven't really been actively using OO for quite a few years, I don't know if that criticism is still as frequent - or valid - as it was back in ye olde days.
It's worth noting in this context that if you've ever seen the movie "Stand and Deliver", one way Jaime Escalante got those great results was by giving his students access to self-paced homeschooling math workbooks rather than using the standard school-provided curriculum.
Yes, but another crucial thing he did was to *push* those students beyond their own expectations to learn things they didn't think themselves capable of learning.
It's actually a great example story - and matches what I've seen from the very few great teachers I've met. Self-motivation is critical, but you need experience to be able to set to goals, and some empathy and art to keep students moving towards those goals even when their motivation wanes or they have the wrong perspective (from inexperience).
The quote that comes to mind is 'you only see the turn, you don't see the road ahead'.
A purely self-directed, self-motivated student is more likely to give up on the first real challenge - because they can't realize that, as much 'natural learner' as anyone can be, you're always going to find something that is *hard* for you to learn and understand. And that it is *normal*, that with some persistence they'll learn it, and that very often this is important for their future.
Absolutely true for real working environments.
The problem is that the GP is also 100% correct - lots of college curricula pretend to cover team-work, but few courses (if any) can teach effective team dynamics. Actually, they unintentionally teach the opposite, as the parent described.
A good experience for team dynamics needs attributes that are omnipresent at work, but are difficult to have in project course-work:
- Project size must *require* a team - not be easily achievable by a single person.
- Work needs diverse skills and complementary roles in equal(ish) measure
- Results persistence after the project - in the real world, your initial 'grade' is almost irrelevant, long-term quality is what matters.
- Social persistence after the project - in the real world, you'll see and work with this people every day long after a single task is done.
The typical "team project" crammed on half a semester course (or less) is the antithesis of this: too small and too short, heavily biased towards one-two skills (so everything else are 'slacker' tasks), and you can choose to never see your teammates again after completion.
Of course, if you can more easily complete everything by yourself by hacking it all together over a few weeks - you would be a fool to do otherwise. So most smart people end up doing exactly that, because what matters is the grade.
But in the workplace, most Real Work *requires* teamwork because it is simply bigger, more complex, and requires more complementary skills and expertise than an artificial CS assignment. And you cannot piss off, or even under-utilize, your peers without burning important bridges, because you'll typically work with the same people over *years*, not weeks. And you need to depend on peer feedback, and on people with complementary roles and skills you don't have (and do require a lot of work) - because a feature gap or quality issue can follow you for a long time.
Sadly, typical college projects seem to train the smartest students to be lone programmers - try to do everything themselves, assume theirs are the only 'real work' skills and they're the one indispensable worker in the operation... and if that doesn't scale, it must be the 'assignment' was broken and doomed anyway.
Fortunately most people learn some teamwork somewhere, but it doesn't seem to be through college.
Exception *might* be team sports... I've never been a big fan of sports, neither practice nor spectator, and before working on the Real World always thought the whole 'teaches teamwork' idea highly overrated. But it does seem to have characteristics missing from course projects, and (anecdotically) I've noticed people with that background tend to grok some of the team dynamics and social subtext more easily than me (which is admittedly not a high bar).
Just because someone gets a check from a company, doesn't mean they're not lonely nerds hiding in their mother's basement :-)
Trying to map the "Security through obscurity..." truism to this context ignores the original meaning, and gets silly pretty quickly.
Real-life security is simply the probability of not being the victim of a successful attack.
In cryptography, the "Security through obscurity is no security" applies to the specific case of making the security system *opaque* - hiding design and implementation of the lock has limited benefits against a determined attacker, shrinking rapidly to zero as more information and computation capacity is available. It assumes security is a binary concept instead of a probabilistic one because *in this specific context* it's a good approximation.
That's completely besides the point here, as in many real-life situations - this is not about the lock design, or the key, or leaving the door unlocked in the first place. It's not about whether the house is "Secure" by some academic definition against an hypothetical Oscar or Mallory determined to attack Mr.X's security. Here there's only Mr.X and the other millions of people with access to the Inter-webs who didn't know he existed in the first place.
The problem is about how Mr.X *becomes a target* in the first place - whether he's a cop or a typical private citizen. Compiling freely available information does more than caching computation - it *singles-out* Mr.X and makes him a visible target, indexed by search-able criteria, more easily accessible to people who wouldn't have bothered to look for Mr.X before.
That may not change the total security profile of the raw data - but it absolutely makes Mr.X *specifically* more exposed. Being a top-10 Google result for "cop off-duty personal address" is simply *not* the same thing as having your info available in principle, across several databases, as an extra record among thousands or millions of entries.
You can't just yell "Fire!" in a crowded theater to make an argument about fire safety codes, and then claim no responsibility because you didn't make the building, and 'someone else could do the same anyway'.
Technically, so can the stalking ex-girlfriend - or the paranoid employer who thinks the cashier is stealing, or the obsessive neighbor thinking anyone new to the neighborhood is a drug-dealer / serial-killer.
But it usually takes more than that argument to call it "investigative reporting" instead of harassment when they start monitoring 24/7.
Dude, I sort of agree with you, but that's hardly the definition of neo-prohibitionist behavior.
From a purely health/social perspective, alcohol *should* be taxed at a higher rate than tobacco. There are already laws against public smoking in most places with their own fines, so the taxed risks are for the smoker's health costs. Alcohol has both long-term health costs for the drinker, *and* the risks to others from coordination impairment.
The simple reason that doesn't happen is because there are more of us drinkers than there are smoking voters. That's not bad from my POV (I enjoy a drink, or ten, don't smoke) - but to expect the opposite is hardly irrational.
Not really - if anything, it removes the ability to take the code and redistribute for free on *Apple's platform* (without taking a loss).
But that's an Apple AppStore problem (or a 'feature') as a distribution channel for any software - not an issue with the GPL or what the authors are doing.
They seem to be doing the right thing with the code - they're essentially passing on the distribution costs, much like they'd have to do if they printed it on CDs for retail.
IFF Apple were to open their platform to other distribution channels, then automatically this would be a non-issue. That is unlikely to happen, but since their GLP-ed code is available, perhaps it could be used in the future to port it to other rich Touch interfaces that may not have such restrictions.
Oh crap!
Now that the first AC is not working on Linux tty code the chances of him leaving his basement and meeting the FreeBSD AC are much higher.
What happens if they meet? What would happen to the space-time continuum?
Would that be different than if someone decided to use normal cryptography, and encrypt their emails to another party unbeknownst to the same company?
In both cases, the company would not be able to provide the cryptographic key to decipher the messages.
I'm getting the impression most people are looking at this as a solution to traditional crypto problems (corporate assets, legal docs, DRM), where it doesn't seem that relevant.
All of those can be considered valuable assets - in the sense that there is a concrete monetary value attached to them, in terms of business value, liability, or expected profits from sale. 'Vanishing' *valuable* data from the original source doesn't seem terribly useful, because if it is considered *valuable* it is likely to be copied/backed-up... and each recipient increases that likelihood.
IFF the information is very valuable, the usual security concerns about protecting against a resourceful untrusted party apply... and since the people with the key are trusted they can copy the data to a non-vanishing target anyway and break the loop. The more valuable, the more likley it will happen for non-malicious reasons, precisely because it will 'vanish'.
Where this seems useful is on large sets of *cheap* individual data more commonly shared online, for which individuals currently lack a good sense of value-as-privacy.
In general, we don't have yet a good instinct to preserve privacy online, at least partly because people are not used to communication networks being both *frictionless* and *persistent*.
If just a single assumption was broken, it wouldn't be a big deal... but the game has changed because *both* of these now occur transparently online.
We do not treat most personal data as information assets, nor do we really anticipate their aggregation, because for millenia *trivial* information would erode over time due to natural processes... people forget, business cards get lost, post-it notes become unreadable, etc. And in the context of that single communication, we could trust our estimate of how trivial that information was for both parties.
We're used to some erosion of information in real life - if in the middle of an in-person conversation we share freely a phone number or a mail address or some personal story with a friend, we don't expect that to be readable for millions of unknown people, or for it to be easily queryable years later, whenever anyone decides... for free.
This tech seems valuable precisely for that scenario because it resets the result of an online communication to those expectations, so we don't have to choose between giving up the idea of privacy or treating everything we say on casual conversation as some kind of 'information asset'.