Consider how many things could be coming at your car from the periphery that the human eye would not be able to detect. Computer systems can have more sensors with longer range. Computers can track more objects coming from more directions than the human eye can track simultaneously.
I can imagine a situation where, under normal situations, thousands of cars are able to drive within close proximity of each other, during optimal conditions, via autonomous means.
Ten, 20 years after automation becomes commonplace in automobiles, what happens when "something", like a goat in the road, throws a wrench into the gears? After a generation of automation, conditions which may have previously occurred would do so no longer. The smallest hickup/erratic change could result in a systemic hickup in one car, which would cause a cascade of problems in the software resulting in mass collisions.
Unless you put cars on tracks in an automated fashion (think of Asimov's conveyors, but with tracks, for cars), I don't see this being reasonably possible or safe. It's one of those "90% problems": it's easy to solve for x to 90% of a solution, but actually hitting 100% is worse than exponentially difficult, if not impossible.
The only people who think like this seem to be without any sort of corporeal responsibility. Their only perceived responsibilities are to themselves - their self-satisfaction.
They don't drive, because a car is a liability and a cost. Easier to mooch off of others.
They don't have families, because they're too immature and/or irresponsible to realize the benefit such things provide to society.
They don't own homes, because a mortgage (and the associated payments) demand stability and willpower to resist compulsive urges.
They're able to pay for small, single-person (or shared) apartments near their place of work because of the aforementioned lack of constraints. It's pretty easy to pay 1800/month for a loft apartment when it's just you living there and you haven't much more than a bottle of Jack Daniels and a pile of $300 shirts.
congratulations, we've got a new Unit
on
Largest Genome Ever
·
· Score: 2, Funny
Congratulations, we now have a new unit of measurement to join the myriad:
* Libraries of Congress * Landmasses of Texas * States of Massachusetts * California Economies * Lines of Code * Man-Hours * Kilobits per second
As someone who has written a couple books (er, attempting to write a couple books) I can certainly see the value - particularly to someone with limited time and who is not a 'professional' author.
A story is a wholly linear thing; it's not like the common structure of a Word document or progmatic functions, which can be broken up into non-sequential components. If he were to write a novel in that fashion ("this function references this, that function references that, so I'll write them in that order - but in reverse") he'd never get the book done. He'd sit down for that hour, write for 20 minutes, and then start rearranging shit "because it doesn't work".
An author will work his ass off to get something - anything - written, just to get it done. He's got to push through about 100k-120k words to result in 60-80k of 'finished book' words. You then take that, and begin mixing, editing, and rearranging the material, possibly rewriting some of it in the process to make it fit. (At least, that's one way that people write).
Many books are made like movies - on the cutting table. I can see vi, or ed, being useful for this kind of thing.
What, are you kidding me? Half the people I knew growing up lived on farms or ranches and I never saw anything but:
* Parents working from 6a-8pm (or more) every damn day (except for Sunday, in which cases they tried to do nothing but usually got drafted for something). * Moms who cooked 2-3 full meals, every day. * Certain parts of the year were full-time jobs - and by that, I mean you get maybe 2-3 hours of sleep between your days, and eat while you're working: branding season, haying season, harvest season. Other parts of the year are full of planning for the next, or waiting for things to grow - but there is always "one more thing" to do (much like, say, IT/sysadmin work). * Parents who wore tattered rags (comparably) and patched clothes so their children could have hot meals at school. * The children had 1-2 hours of chores every day, more on summers. * When the parents retired, the kids took over the farm. Many of these farms have/had multiple houses for multiple generations of family members (as well as hired workers). When they got married or wanted to live on their own, there was 'free' housing available.
This is a way of life, and not a particularly bad one. It is full of leisure time if you care to take it (during certain parts of the year), and nobody is your boss but yourself (and your wife and/or parents, of course). It has a lot of intrinsic value which is not present in most other career choices, allowing for cohesive families, land ownership, and freedom.
You speak as someone who doesn't know and who couldn't care - a modern 'educated' urban bigot.
Then why were they producing games like Oni and having other developers do the Mac ports?
The purchase of Bungie killed a lot of awesomeness, namely Oni 2 and multiplayer support for the original - a real shame, due to the unique awesomeness that Oni was (vs. Halo, which was more of the same - but with vehicles!).
You have to admit, it'll be a pairing of historical kinship: they both produce multiple closed variants of something commonly available for free which is known for wanton security flaws, bloat, and poor design.
Adobe's products will fit nicely into Microsoft's portfolio.
As one of those people who are always yelling about "green" energy which is no good, let me defend myself.
Simply: I dislike most green energy because it isn't "green". Sure, you're not burning petrochemicals to get the fuel, but that does not make them "green", or even (in many cases, better).
Wind turbines, for instance, are environmentally destructive in use, have a relatively small life span, and use up copious amounts of natural resources in their construction (metal, cement). Their construction is disproportionately energy intensive - something which takes a very long time to recoup.
Solar is approaching practicality while at the same time decreasing in cost. It's improving, but the cost of these panels is largely demonstrative of their environmental impact (even after subsidy). Economy of scale is not the issue: production is running at 110% due to demand, yet prices remain high due to the materials and processing which must go into production.
Bio-diesel is another big one, historically (though I understand this has changed in the past couple years - I've not been paying much attention to it).
My particular pet peeve, at this point, is the EV - particularly the EVs which use small gas engines and lithium based batteries. Why in the world would anyone think these are a good idea? Aside from the limited supply of lithium and its relatively short lifespan (lead-acid gets 5 years, why would you expect more than that from lithium?), it is extremely costly to produce and recycle. This cost translates to environmental impact, and is largely the result of limited supply and the toxic nature of lithium. Then there's the fact that these EVs use gas engines instead of diesels, which makes absolutely no sense when considering the amperage potential of a diesel vs. a gas engine, the relative lack of energy density in gasoline when compared to diesel, and the additional refinement necessary over diesel for gasoline production. Not terribly green.
When you eliminate the gas engine, why do batteries at all? Just generate electricity directly from something like a diesel turbine engine and use that to directly power your wheel engines (or convert an older vehicle to use a variable speed electric motor). (For increased performance, you could add a battery 'capacitor' for starts and the like, of course.)
California's roadway system is another example. If you want to go 'green', improving your cities might be a better way to do so instead of improving your interstates or installing billions in 'public transit'. The cost should not be so dissimilar unless politicians get 'overly involved', and the result will be better cities.
I am against drastic, disruptive technological change, because there is not enough time to judge the practical long-term results of such changes. What will they do to the economy? What will they do to the environment? And so on.
We've had the technology for decades to create 'green' stuff, yet we're ignoring those improvements largely in favor for 'more technology'. Case in point: a 2010 Sienna claims 17 city, 23 highway - for a combined 19mpg. However, these #s are more than likely based off the in-car computer, which almost invariably seems to over-estimate the mileage). Meanwhile, a 20-year-old full-size van (truck chassis, more room, etc.) with a turbo charged inline 6 (Windsor) will get about 2-4mpg better than that with similar driving patterns and be able to haul a hell of a lot more, nevermind hitch a trailer full of tools behind it and not fall apart in the process. The technology in that 'old' van is 20 years old, and does a better job to boot in pretty much every comparison.
Personally, I prefer wide columns (to a point), because it breaks the paragraphs up more, allowing me to more easily identify where I am if I lose my place. It's much harder to lose your place in a paragraph (skipping over lines, for instance) when there are only 2-3 lines per paragraph. Skipping paragraphs might be a problem then, but I've yet to see it - they're spaced far enough apart that this does not tend to happen.
If the font is really small or poorly spaced (clustered together/poor line spacing), such as it is in some books, I could see it being a problem.
How often do you have monitors die? I've yet to have an LCD die, and only seen a scant handful die (3 or 4?) for our clients in the past decade (thousands of deployed), and only a few more laptop CCFLs/capacitors (not worth replacing, IMO - too damn difficult). Granted, I don't pay much attention to it, but compared to CRTs (which would have the tube go out or lose calibration), it's most certainly trivial.
Personally, I'm very pleased with my 1680x1050 screens. I've got two identical 22" Acers which sit on my desks - one at home, one at work.
They're perfect for multiple terminals - I can essentially fit 4 tiled 80x60 terminals on the screen and still have them all usable and readable from some distance away.
They're also good for media viewing and gaming.
Now, 900 horizonal pixels is bullshit. OK, it's great if you're going to be doing nothing but watch movies, I suppose (or maybe not - seems a little shallow to me). But that's not even enough for flipping the monitor and using it to read documents: it's crap. 800 pixels was barely wide enough 10 years ago for such tasks, and now it's almost impossible given all the 'software borders' we've got in most software.
I think we can chalk it up to the "stupid media consumer" culture. People want those wide screens, damn it - and in order to manufacture increasingly cheap monitors at larger sizes, they drop the pixel count, and widen the screen (lowering the DPI).
I don't see a path back until the 'hottest craze consumer crap' surpasses current desktop LCD resolutions. They'll probably converge at one point, and desktop LCD prices will go up on account of technology consolidation ("it's the same as a TV, so we'll charge just as much"). It's somewhat disheartening, particularly when you consider that CRTs from 10 years ago could do 2-4x the resolution of a current LCD, at a higher DPI/at the same size.
Unless every country had a space elevator, we would quickly see denial attacks against said space elevators (or attempts to control them). "Free, unlimited energy" is a game changer.
A saner approach would probably be narrow band microwave, I'd think.
From what I've seen of Libertarians, this is pretty anti-libertarian in approach - but not for the reasons you state.
Libertarians want government, they just want "small government", particularly as the seat of power is far. Jurisdiction over an area is managed by the smallest form of government which covers said area. Federal government should be very, very small. State government would be (proportionately) somewhat larger. Larger still would be your city and municipality governments, with township councils also being larger and more significant in peoples' lives.
As it stands now, the opposite is true. Local governments are disregarded as ineffectual and useless; state government are viewed as handymen for the Federal government. Anything which impacts a person is typically seen as coming from the Federal government (unless you're living in a micro-country like California or NY, which views itself as the end-all, be-all, and treats you like they must provide every single service, disregarding the fact that you also have to pay out 30%+ of your income to the Federal government).
* Roadways? Local, county, or at the most state, managed. * Police? Why would funding for police come from the Federal level, when it is by definition a local issue? Do most places (by area) even need "policing"? Certainly not. * Healthcare? Under no uncertain terms, this shouldn't have government involvement at all. The "common good" clause is bullshit in this regard (as evidenced by the decreases in services AND increases in costs for those who need them when the government gets involved with regulation of said healthcare).
Government's role should not be one of authority, but of coordination. At the local level, this is easier to enact than at the Federal or State level, because you can actually hold government accountable and it is easier to be involved.
Hell, where there is no functional government, the community has to pick up the slack, anyway (community watch groups and cooperative fire departments, for instance). When the government is not saying "we are here to help", people realize they are on their own - and band together to help themselves (and by coincidence, others) in the process.
I have a friend who is a fire fighter in some of the most rural land in this country. It is, in essence, a "militia fire brigade": he and others have pickup trucks with water tanks, hoses, and the like (many of them on removable beds which can be quickly swapped out with, say, a welding bed). If (when - this is very dry country) there is a fire within the community (I believe it's a cooperative fire department in this regard, with dues), everyone gets a call and anyone able to respond, does. The responders get paid for their time on the fire, as well as a monthly stipend for maintaining the coop equipment. It's quite amazing to see the response to one of these fires: a half trucks with 300-500 gal or so (such as this: truck water tanks) each show up independently within a fairly short (30 minutes - many traveling at over 60mph down gravel roads to get there) period of time; each truck is blazoned in different colors, brands, and door badges ("JD's Welding and Fire Control", "Mike's Metal Works", "Jimbob Double-C Ranch").
If the burden of self reliance is there without the false impression of government help, then people will step up. There will always be fools who do not heed the call - but then they get billed and (if not paid) sued. Not putting out the fire, as this city did, is pretty much par for the course of government: "we can't help you because you didn't pay". Well, gee, thanks. An individual - a human - is more likely to rely upon his judgement and respond in a humane fashion: we'll take care of the problem, but your ass is mine afterwards if you fail your obligations at this point.
IPv6 isn't hard because everyone's waiting for someone else to take the plunge. It's because it's not necessary for most environments, it's difficult to comprehend,
The comprehension part: brains stop being able to group and separate things at about 6 independent objects:
Yet then you've got : instead of a much saner . for separation, which is more visually difficult to parse. There is a good reason why quad-dot notation became default for IPv6, and it probably has something to do with anything else being somewhat difficult to process/group.
Sure, you'll have shorter addresses:
0:0:0:0:0:0:101.45.75.219 or:::::::101.45.75.219
But is that really much easier?
Do I, personally, really need something like 5 billion personal addresses available to me?
Nevermind the frustrating complexity of subnetting with IPv6.
This is much more than just building it for public display. The idea is to demonstrate that it was, indeed, a fully functional device, and to give credit where credit is due.
Probably Debian would have been OK, but I was finding admin of most Linux distros a pain for exactly these reasons. I couldn't find a layer where I could do everything that I needed to do without worrying about one thing stepping on another.
You should've gone with Debian. Unlike Debian, FreeBSD will walk all over you when it comes to upgrading your ports. You have 'multiple layers' of software.
freebsd-upgrade is nice, but unfortunately it only does GENERIC. Guess what? GENERIC sucks, so unless you can 'get by' with the support it provides you're going to have to be doing kernel upgrades manually, too.
The best way to upgrade ports in FreeBSD is to not have to.
I don't think you really understand systems administration. 'Users,' or in this case admins, don't typically do stuff once. Furthermore, they need to know what he did and how to do it again (i.e. new server or whatever) or just remember what he did.
Are you kidding? In a small business environment (several hundred users) I rarely get to do the same thing twice. Every day it's on to something else - something different.
Yes, there is a subset of things which I do repetitively on a daily basis. But that is a very small subset. Usually, I'm trying to redo or redesign something done poorly, or fix a problem in some exotic application. Never a dull day!
One-off stuff isn't common and is a sign of poor administration (i.e. tracking changes and following processes).
If I have to make the same configuration multiple times, I am administering poorly.
I should not have to create the same user more than once for different platforms or applications: I create them in the directory, assign attributes, and I am done.
I should not have to edit a dozen different hosts to change the location of my DNS servers. I should not have to do repetitive shit: repetitive shit is why we have scripting and hundreds of management tools to reduce the amount of repetition.
I agree: the "sysadmins should be programmers" assumption is not only a flawed one, but the assumption leads to poor system processes.
Yes, you should be able to write basic system and backup scripts. But programming as an activity is an intrinsically time-consuming task. Sysadmin work, however, often requires "improvements per hour". They are fundamentally incompatible and such a shared role results in your 'programming admin' not doing either job well.
If you hire a programmer to do your sysadmin work (or one who does heavy programming) you are substantially increasing your management responsibilities with every program they write. Those tools have to be maintained or replaced. This costs money, and increases the cost of actually going forward from these 'elite admins'.
What's worse: most of these idiots don't document their code or even their tools in a document repository. They might use cvs or the like to manage them, but chances are they're "good" and just keep a revision or two at a time of a thousand+ line program. They'll write lots of 'glue' programs like this, resulting in inter-dependence amongst your hosts and systems - none of which will likely be documented well, and none will be anything near 'standard'.
On the flip side, a CLI allows a 'competent' administrator the possibility of making the environment overly complex.
If you've only got (say) 2-3 options for setting SNMP on Windows, it's hard to make your monitoring environment too complex. If the 'default' is already mostly set up for you (Active Directory) vs., say, Linux + Samba + LDAP + krb5, you're going to have many more ways to deviate from a 'standard' due to the CLI nature of things and open-ended nature of CLI configuration and management.
There is one way to add a service in Windows, whether it's via GUI or otherwise. In Linux or BSD, there are standard ways, but don't put it past someone to do it differently.
Also, a good CLI makes it really trivial for a bad administrator to make complex GUIs with no CLI command line. That's a lot of fun.
Basically, my point is: a bad administrator (who would probably be a competent Windows admin) could easily over-complicate things on a primarily-CLI operating system, whether due to not enough knowledge or two much with too little systems understanding.
Those systems that are GUI configured are certainly easier to rebuild or migrate. They're usually not as complex.
Let's keep the server tasks off the clients, thanks. It's enough having a handful of Windows clients on a network in a 'trusted' role; no need to add another twelve.
And if Windows and other operating systems were to implement bittorrent based updates, you'd have ISPs against it, too: it'd be nothing but a cost to them, because their upstream costs a lot more than downstream. Their entire business model is geared towards a one-way data pipe, and anything extra (upstream) is there out of necessity of acknowledged communications and low failure rates (TCP).
Nevermind the security issues to clients by instituting such a scheme for 'security updates'. BT would have to fundamentally change before it could be relied upon for such a mechanism.
If I see you downloading a patch/new update to a broken/vulnerable package, I know that for at least several seconds (and longer if the package is earlier in the alphabet) you've got a vulnerable box. So I launch my exploit against your IP and gain root on your machine before you're able to actually apply the new binary, and you're cooked.
Think that's far fetched? Imagine if (when) it's automated. Nevermind hitting every IP - you can just wait for hosts to announce their vulnerability.
Slashdot Mirror
The first set of instructions. They're so easy a computer can follow them.
Consider how many things could be coming at your car from the periphery that the human eye would not be able to detect. Computer systems can have more sensors with longer range. Computers can track more objects coming from more directions than the human eye can track simultaneously.
I can imagine a situation where, under normal situations, thousands of cars are able to drive within close proximity of each other, during optimal conditions, via autonomous means.
Ten, 20 years after automation becomes commonplace in automobiles, what happens when "something", like a goat in the road, throws a wrench into the gears? After a generation of automation, conditions which may have previously occurred would do so no longer. The smallest hickup/erratic change could result in a systemic hickup in one car, which would cause a cascade of problems in the software resulting in mass collisions.
Unless you put cars on tracks in an automated fashion (think of Asimov's conveyors, but with tracks, for cars), I don't see this being reasonably possible or safe. It's one of those "90% problems": it's easy to solve for x to 90% of a solution, but actually hitting 100% is worse than exponentially difficult, if not impossible.
The only people who think like this seem to be without any sort of corporeal responsibility. Their only perceived responsibilities are to themselves - their self-satisfaction.
They don't drive, because a car is a liability and a cost. Easier to mooch off of others.
They don't have families, because they're too immature and/or irresponsible to realize the benefit such things provide to society.
They don't own homes, because a mortgage (and the associated payments) demand stability and willpower to resist compulsive urges.
They're able to pay for small, single-person (or shared) apartments near their place of work because of the aforementioned lack of constraints. It's pretty easy to pay 1800/month for a loft apartment when it's just you living there and you haven't much more than a bottle of Jack Daniels and a pile of $300 shirts.
Congratulations, we now have a new unit of measurement to join the myriad:
* Libraries of Congress
* Landmasses of Texas
* States of Massachusetts
* California Economies
* Lines of Code
* Man-Hours
* Kilobits per second
Welcome to the fold, Big Bens!
As someone who has written a couple books (er, attempting to write a couple books) I can certainly see the value - particularly to someone with limited time and who is not a 'professional' author.
A story is a wholly linear thing; it's not like the common structure of a Word document or progmatic functions, which can be broken up into non-sequential components. If he were to write a novel in that fashion ("this function references this, that function references that, so I'll write them in that order - but in reverse") he'd never get the book done. He'd sit down for that hour, write for 20 minutes, and then start rearranging shit "because it doesn't work".
An author will work his ass off to get something - anything - written, just to get it done. He's got to push through about 100k-120k words to result in 60-80k of 'finished book' words. You then take that, and begin mixing, editing, and rearranging the material, possibly rewriting some of it in the process to make it fit. (At least, that's one way that people write).
Many books are made like movies - on the cutting table. I can see vi, or ed, being useful for this kind of thing.
What, are you kidding me? Half the people I knew growing up lived on farms or ranches and I never saw anything but:
* Parents working from 6a-8pm (or more) every damn day (except for Sunday, in which cases they tried to do nothing but usually got drafted for something).
* Moms who cooked 2-3 full meals, every day.
* Certain parts of the year were full-time jobs - and by that, I mean you get maybe 2-3 hours of sleep between your days, and eat while you're working: branding season, haying season, harvest season. Other parts of the year are full of planning for the next, or waiting for things to grow - but there is always "one more thing" to do (much like, say, IT/sysadmin work).
* Parents who wore tattered rags (comparably) and patched clothes so their children could have hot meals at school.
* The children had 1-2 hours of chores every day, more on summers.
* When the parents retired, the kids took over the farm. Many of these farms have/had multiple houses for multiple generations of family members (as well as hired workers). When they got married or wanted to live on their own, there was 'free' housing available.
This is a way of life, and not a particularly bad one. It is full of leisure time if you care to take it (during certain parts of the year), and nobody is your boss but yourself (and your wife and/or parents, of course). It has a lot of intrinsic value which is not present in most other career choices, allowing for cohesive families, land ownership, and freedom.
You speak as someone who doesn't know and who couldn't care - a modern 'educated' urban bigot.
Bullshit.
Then why were they producing games like Oni and having other developers do the Mac ports?
The purchase of Bungie killed a lot of awesomeness, namely Oni 2 and multiplayer support for the original - a real shame, due to the unique awesomeness that Oni was (vs. Halo, which was more of the same - but with vehicles!).
You have to admit, it'll be a pairing of historical kinship: they both produce multiple closed variants of something commonly available for free which is known for wanton security flaws, bloat, and poor design.
Adobe's products will fit nicely into Microsoft's portfolio.
As one of those people who are always yelling about "green" energy which is no good, let me defend myself.
Simply: I dislike most green energy because it isn't "green". Sure, you're not burning petrochemicals to get the fuel, but that does not make them "green", or even (in many cases, better).
Wind turbines, for instance, are environmentally destructive in use, have a relatively small life span, and use up copious amounts of natural resources in their construction (metal, cement). Their construction is disproportionately energy intensive - something which takes a very long time to recoup.
Solar is approaching practicality while at the same time decreasing in cost. It's improving, but the cost of these panels is largely demonstrative of their environmental impact (even after subsidy). Economy of scale is not the issue: production is running at 110% due to demand, yet prices remain high due to the materials and processing which must go into production.
Bio-diesel is another big one, historically (though I understand this has changed in the past couple years - I've not been paying much attention to it).
My particular pet peeve, at this point, is the EV - particularly the EVs which use small gas engines and lithium based batteries. Why in the world would anyone think these are a good idea? Aside from the limited supply of lithium and its relatively short lifespan (lead-acid gets 5 years, why would you expect more than that from lithium?), it is extremely costly to produce and recycle. This cost translates to environmental impact, and is largely the result of limited supply and the toxic nature of lithium. Then there's the fact that these EVs use gas engines instead of diesels, which makes absolutely no sense when considering the amperage potential of a diesel vs. a gas engine, the relative lack of energy density in gasoline when compared to diesel, and the additional refinement necessary over diesel for gasoline production. Not terribly green.
When you eliminate the gas engine, why do batteries at all? Just generate electricity directly from something like a diesel turbine engine and use that to directly power your wheel engines (or convert an older vehicle to use a variable speed electric motor). (For increased performance, you could add a battery 'capacitor' for starts and the like, of course.)
California's roadway system is another example. If you want to go 'green', improving your cities might be a better way to do so instead of improving your interstates or installing billions in 'public transit'. The cost should not be so dissimilar unless politicians get 'overly involved', and the result will be better cities.
I am against drastic, disruptive technological change, because there is not enough time to judge the practical long-term results of such changes. What will they do to the economy? What will they do to the environment? And so on.
We've had the technology for decades to create 'green' stuff, yet we're ignoring those improvements largely in favor for 'more technology'. Case in point: a 2010 Sienna claims 17 city, 23 highway - for a combined 19mpg. However, these #s are more than likely based off the in-car computer, which almost invariably seems to over-estimate the mileage). Meanwhile, a 20-year-old full-size van (truck chassis, more room, etc.) with a turbo charged inline 6 (Windsor) will get about 2-4mpg better than that with similar driving patterns and be able to haul a hell of a lot more, nevermind hitch a trailer full of tools behind it and not fall apart in the process. The technology in that 'old' van is 20 years old, and does a better job to boot in pretty much every comparison.
Personally, I prefer wide columns (to a point), because it breaks the paragraphs up more, allowing me to more easily identify where I am if I lose my place. It's much harder to lose your place in a paragraph (skipping over lines, for instance) when there are only 2-3 lines per paragraph. Skipping paragraphs might be a problem then, but I've yet to see it - they're spaced far enough apart that this does not tend to happen.
If the font is really small or poorly spaced (clustered together/poor line spacing), such as it is in some books, I could see it being a problem.
How often do you have monitors die? I've yet to have an LCD die, and only seen a scant handful die (3 or 4?) for our clients in the past decade (thousands of deployed), and only a few more laptop CCFLs/capacitors (not worth replacing, IMO - too damn difficult). Granted, I don't pay much attention to it, but compared to CRTs (which would have the tube go out or lose calibration), it's most certainly trivial.
Personally, I'm very pleased with my 1680x1050 screens. I've got two identical 22" Acers which sit on my desks - one at home, one at work.
They're perfect for multiple terminals - I can essentially fit 4 tiled 80x60 terminals on the screen and still have them all usable and readable from some distance away.
They're also good for media viewing and gaming.
Now, 900 horizonal pixels is bullshit. OK, it's great if you're going to be doing nothing but watch movies, I suppose (or maybe not - seems a little shallow to me). But that's not even enough for flipping the monitor and using it to read documents: it's crap. 800 pixels was barely wide enough 10 years ago for such tasks, and now it's almost impossible given all the 'software borders' we've got in most software.
I think we can chalk it up to the "stupid media consumer" culture. People want those wide screens, damn it - and in order to manufacture increasingly cheap monitors at larger sizes, they drop the pixel count, and widen the screen (lowering the DPI).
I don't see a path back until the 'hottest craze consumer crap' surpasses current desktop LCD resolutions. They'll probably converge at one point, and desktop LCD prices will go up on account of technology consolidation ("it's the same as a TV, so we'll charge just as much"). It's somewhat disheartening, particularly when you consider that CRTs from 10 years ago could do 2-4x the resolution of a current LCD, at a higher DPI/at the same size.
Unless every country had a space elevator, we would quickly see denial attacks against said space elevators (or attempts to control them). "Free, unlimited energy" is a game changer.
A saner approach would probably be narrow band microwave, I'd think.
From what I've seen of Libertarians, this is pretty anti-libertarian in approach - but not for the reasons you state.
Libertarians want government, they just want "small government", particularly as the seat of power is far. Jurisdiction over an area is managed by the smallest form of government which covers said area. Federal government should be very, very small. State government would be (proportionately) somewhat larger. Larger still would be your city and municipality governments, with township councils also being larger and more significant in peoples' lives.
As it stands now, the opposite is true. Local governments are disregarded as ineffectual and useless; state government are viewed as handymen for the Federal government. Anything which impacts a person is typically seen as coming from the Federal government (unless you're living in a micro-country like California or NY, which views itself as the end-all, be-all, and treats you like they must provide every single service, disregarding the fact that you also have to pay out 30%+ of your income to the Federal government).
* Roadways? Local, county, or at the most state, managed.
* Police? Why would funding for police come from the Federal level, when it is by definition a local issue? Do most places (by area) even need "policing"? Certainly not.
* Healthcare? Under no uncertain terms, this shouldn't have government involvement at all. The "common good" clause is bullshit in this regard (as evidenced by the decreases in services AND increases in costs for those who need them when the government gets involved with regulation of said healthcare).
Government's role should not be one of authority, but of coordination. At the local level, this is easier to enact than at the Federal or State level, because you can actually hold government accountable and it is easier to be involved.
Hell, where there is no functional government, the community has to pick up the slack, anyway (community watch groups and cooperative fire departments, for instance). When the government is not saying "we are here to help", people realize they are on their own - and band together to help themselves (and by coincidence, others) in the process.
I have a friend who is a fire fighter in some of the most rural land in this country. It is, in essence, a "militia fire brigade": he and others have pickup trucks with water tanks, hoses, and the like (many of them on removable beds which can be quickly swapped out with, say, a welding bed). If (when - this is very dry country) there is a fire within the community (I believe it's a cooperative fire department in this regard, with dues), everyone gets a call and anyone able to respond, does. The responders get paid for their time on the fire, as well as a monthly stipend for maintaining the coop equipment. It's quite amazing to see the response to one of these fires: a half trucks with 300-500 gal or so (such as this: truck water tanks) each show up independently within a fairly short (30 minutes - many traveling at over 60mph down gravel roads to get there) period of time; each truck is blazoned in different colors, brands, and door badges ("JD's Welding and Fire Control", "Mike's Metal Works", "Jimbob Double-C Ranch").
If the burden of self reliance is there without the false impression of government help, then people will step up. There will always be fools who do not heed the call - but then they get billed and (if not paid) sued. Not putting out the fire, as this city did, is pretty much par for the course of government: "we can't help you because you didn't pay". Well, gee, thanks. An individual - a human - is more likely to rely upon his judgement and respond in a humane fashion: we'll take care of the problem, but your ass is mine afterwards if you fail your obligations at this point.
IPv6 isn't hard because everyone's waiting for someone else to take the plunge. It's because it's not necessary for most environments, it's difficult to comprehend,
The comprehension part: brains stop being able to group and separate things at about 6 independent objects:
623
5452
23412
848328
The 7th becomes difficult:
2399495
What you have with IPv6 is something like this:
1111111111111111.1111111111111111.111111111111111.1111111111111111.1111111111111111.1111111111111111.1111111111111111.1111111111111111
Now, quick! Which of these segments is different than the address above?
1111111111111111.1111111111111111.1111111111111111.1111111111111111.1111111111111111.1111111111111111.1111111111111111.1111111111111111
Sure, you could write it in hex:
2001:0db8:3c4d:0015:0000:0000:abcd:ef12
Yet then you've got : instead of a much saner . for separation, which is more visually difficult to parse. There is a good reason why quad-dot notation became default for IPv6, and it probably has something to do with anything else being somewhat difficult to process/group.
Sure, you'll have shorter addresses:
0:0:0:0:0:0:101.45.75.219 :::::::101.45.75.219
or
But is that really much easier?
Do I, personally, really need something like 5 billion personal addresses available to me?
Nevermind the frustrating complexity of subnetting with IPv6.
Realistically, you should have your original install medium available: no Internet connection necessary.
Or were you suggesting stupid like "an infected machine can be disinfected with certainty"?
How did this idiot get modded insightful? Just because he posted early?
More like nobody else gives a damn about this kinda thing.
This is much more than just building it for public display. The idea is to demonstrate that it was, indeed, a fully functional device, and to give credit where credit is due.
Probably Debian would have been OK, but I was finding admin of most Linux distros a pain for exactly these reasons. I couldn't find a layer where I could do everything that I needed to do without worrying about one thing stepping on another.
You should've gone with Debian. Unlike Debian, FreeBSD will walk all over you when it comes to upgrading your ports. You have 'multiple layers' of software.
freebsd-upgrade is nice, but unfortunately it only does GENERIC. Guess what? GENERIC sucks, so unless you can 'get by' with the support it provides you're going to have to be doing kernel upgrades manually, too.
The best way to upgrade ports in FreeBSD is to not have to.
I don't think you really understand systems administration. 'Users,' or in this case admins, don't typically do stuff once. Furthermore, they need to know what he did and how to do it again (i.e. new server or whatever) or just remember what he did.
Are you kidding? In a small business environment (several hundred users) I rarely get to do the same thing twice. Every day it's on to something else - something different.
Yes, there is a subset of things which I do repetitively on a daily basis. But that is a very small subset. Usually, I'm trying to redo or redesign something done poorly, or fix a problem in some exotic application. Never a dull day!
One-off stuff isn't common and is a sign of poor administration (i.e. tracking changes and following processes).
If I have to make the same configuration multiple times, I am administering poorly.
I should not have to create the same user more than once for different platforms or applications: I create them in the directory, assign attributes, and I am done.
I should not have to edit a dozen different hosts to change the location of my DNS servers. I should not have to do repetitive shit: repetitive shit is why we have scripting and hundreds of management tools to reduce the amount of repetition.
I agree: the "sysadmins should be programmers" assumption is not only a flawed one, but the assumption leads to poor system processes.
Yes, you should be able to write basic system and backup scripts. But programming as an activity is an intrinsically time-consuming task. Sysadmin work, however, often requires "improvements per hour". They are fundamentally incompatible and such a shared role results in your 'programming admin' not doing either job well.
If you hire a programmer to do your sysadmin work (or one who does heavy programming) you are substantially increasing your management responsibilities with every program they write. Those tools have to be maintained or replaced. This costs money, and increases the cost of actually going forward from these 'elite admins'.
What's worse: most of these idiots don't document their code or even their tools in a document repository. They might use cvs or the like to manage them, but chances are they're "good" and just keep a revision or two at a time of a thousand+ line program. They'll write lots of 'glue' programs like this, resulting in inter-dependence amongst your hosts and systems - none of which will likely be documented well, and none will be anything near 'standard'.
On the flip side, a CLI allows a 'competent' administrator the possibility of making the environment overly complex.
If you've only got (say) 2-3 options for setting SNMP on Windows, it's hard to make your monitoring environment too complex. If the 'default' is already mostly set up for you (Active Directory) vs., say, Linux + Samba + LDAP + krb5, you're going to have many more ways to deviate from a 'standard' due to the CLI nature of things and open-ended nature of CLI configuration and management.
There is one way to add a service in Windows, whether it's via GUI or otherwise. In Linux or BSD, there are standard ways, but don't put it past someone to do it differently.
Also, a good CLI makes it really trivial for a bad administrator to make complex GUIs with no CLI command line. That's a lot of fun.
Basically, my point is: a bad administrator (who would probably be a competent Windows admin) could easily over-complicate things on a primarily-CLI operating system, whether due to not enough knowledge or two much with too little systems understanding.
Those systems that are GUI configured are certainly easier to rebuild or migrate. They're usually not as complex.
They already have this, it's called WSUS.
Let's keep the server tasks off the clients, thanks. It's enough having a handful of Windows clients on a network in a 'trusted' role; no need to add another twelve.
And if Windows and other operating systems were to implement bittorrent based updates, you'd have ISPs against it, too: it'd be nothing but a cost to them, because their upstream costs a lot more than downstream. Their entire business model is geared towards a one-way data pipe, and anything extra (upstream) is there out of necessity of acknowledged communications and low failure rates (TCP).
Nevermind the security issues to clients by instituting such a scheme for 'security updates'. BT would have to fundamentally change before it could be relied upon for such a mechanism.
If I see you downloading a patch/new update to a broken/vulnerable package, I know that for at least several seconds (and longer if the package is earlier in the alphabet) you've got a vulnerable box. So I launch my exploit against your IP and gain root on your machine before you're able to actually apply the new binary, and you're cooked.
Think that's far fetched? Imagine if (when) it's automated. Nevermind hitting every IP - you can just wait for hosts to announce their vulnerability.