> KDE centralizes these features >so you don't have to be in a web >browser to use them.
>KDE: When you're in any application >press ALT+F2, type gg:foo, press ENTER >Firefox: Start or switch to Firefox, >CTRL+K, type foo, press ENTER -or- >Start or switch to Firefox, CTRL+L, >type "searchkeyword foo", press ENTER
But it takes a couple minutes of setup to be able to do the same thing with firefox (or any browser you prefer, such as my favorite, elinks).
The only question is why the KDE folks working on minicli didn't give their users the option to use a different browser in the first place. It would be a trivial change, and would make that feature a lot more useful. (I'd offer a possible answer, but it wouldn't be polite.)
Here's a quick example of how to get the same feature in firefox. Create an executable file somewhere in your path called "gg" containing this:
#---------------begin file gg #!/bin/bash
SEARCHURL="http://www.google.com/sear ch?hl=en&q=" SEARCHURL="${SEARCHURL}$1" shift f or i in $@; do SEARCHURL="${SEARCHURL}+$i" done
FIREFOX_PATH= "/usr/local/firefox"
if [ -z "`ps x | grep \"[0-9] ${FIREFOX_PATH}/firefox-bin\"`" ]; then # No MozillaFirebird running ${FIREFOX_PATH}/firefox $SEARCHURL else # MozillaFirebird running - open a new window ${FIREFOX_PATH}/firefox -a firefox -remote "openURL($SEARCHURL,new-tab)" fi
#---------------end file gg
Now type "gg some words" from the KDE minicli (or any terminal, the icewm bar, etc) and get the same effect.
E-books do have two real advantages. You can read them in the dark without a light source. You can, theoretically, bring your entire library with you rather than just one or two books. The second one is really the most compelling feature.
. . . and, you can become the most-watched guy on the subway. If you hang out in the places I do, standing around conspicuously holding a multi-hundred-dollar-device makes for some unpleasant social interactions.
I actually read a lot of ebooks. I work part of the day in a cleanroom, where I often have many 10 minute spells waiting for equipment with nothing to do. Since we can't bring in regular paper, a portable device is perfect. (In my case a first generation high-resolution palm pilot from a few years back that I got on the cheap.)
Also, as someone who sleeps significantly fewer hours than his mate, being able to read in bed with the lights off is a big plus. There's enough drm-free content out there from Project Guttenberg, Cory Doctorow, online SF mags, etc. to keep me occupied for a long time. I'm looking forward to trying out the Baen site mentioned here. (Sorry to see how terrible their web site looks. Not because I care -- its the books that matter, not their graphic design -- but because I want drm-free ebook distributors to succeed in the world, and their site screams "don't take me seriously.")
But, there's always going to be a place for paper books, if only because an individual paperback is almost worthless. It will be a long, long time before there's an electronic book reader that I'd feel good about reading on a bus in a developing country.
It constrains the possible geometries a bit, but has the great advantage of being super cheap and waterproof. I always imagined concentric poured concrete shells and sump-pumps, but if you can get an outer shell that's really impermeable, you're set.
I suppose cylindrical underground homes are less likely to appreciate in value than rectangular underground homes, but as we've already abandoned the home-as-investment model by moving underground in the first place, so who cares?
It seems to me neighborhoods would be a lot more interesting if people built the homes they most wanted to live in rather than the ones they think will net them the biggest profit on resale. We can't really be a society of people who think square boxes made of off-white painted dry wall are the best imaginable environment. Not sure how you convince an individual to make that choice, though. (For me, the choice is obvious. Which is why I'll almost certainly not become rich, although I have every intention of remaining happy...)
Perhaps the take away message from both posts is that power to control a system ought to reside with whoever is most invested in making sure it works, where "works" is defined by context. For a bank, not crashing the existing system may be the most important thing, and if the sysadmin is responsible for preventing that from happening, he should be given the power to do so.
For an academic, on the other hand, getting something to run at all, even if it locks up the machine and occasionally forces a hard reset is often good enough. Good enough, here, means that spending any more time on it that could be spent elsewhere would be wasteful.
In my experience (in physical sciences research labs), the only people genuinely invested in making sure software works are students and scientists who need it for a particular project. Full time computer people are great when we need to turn to someone for help, or for maintaining shared servers or backup systems, but when you're forced to turn to them for every minor change to a desktop it becomes impossible to get any work done. You spend all your time trying to convince them that something is worth doing, and then waiting for them to do it.
As far as I can tell, just about all the simulation, data analysis, and data acquisition code in the world is crap software that no self-respecting sysadmin would allow on a well run system. (The exception may be some of the stuff run by multi user computing facilities that hire real computer people, which I've never worked on.) But, it's how work gets done. Anything else would be prohibitively expensive.
We spend almost all of our computer time either writing our own buggy, amateurish software or trying to install and configure packages made by some other amateur coder who didn't pay nearly enough attention to portability and standards. It's the sort of annoying trial-and-error, one-off debugging that would just piss off a full time sysadmin, who is likely to either spend a *very* long time doing it or simply refuse and claim it impossible.
Just imagine what the average 8-5 computer admin type would say if you called them up on a friday evening and said, "we want to try reverting to the comedi libraries from cvs dated roughly six weeks ago just to see if maybe it fixes these random lockups that we've been experiencing. Let's also build the stuff from six months back too in case we want to try that. Can you have that done in time for Joe's run at midnight tonight?" They'd tell you to go to hell, with good reason. On the other hand, Joe may be perfectly happy to do the work himself on the off chance it will work. (In which case *not* blaming the computer admin guys when Joe trashes his system and has to reinstall everything is a very good idea.)
Asking a central administrator to debug code that's only ever going to run on a single dedicated machine in someone's lab for a week is a waste of everyone's time.
Whatever you choose to run through the conduit will be dated in a few years. But leaving yourself with as much easily accessible conduit as possible will always be useful. Use the largest diameter conduit your building codes will allow, and run it *everywhere*, and fill it with a bundle of pull-wires. Put in multiple panel boxes per wall, at floor, ceiling, and table rail height. Record their locations well, and then wall over them.
Then in the future when the residents decide to turn a living room into an office/nursery/VR-gymnasium, they can run whatever power or information standards they want through the conduit without destroying whole walls.
Anything beyond that is up to the fetish of the homeowner. My own personal dream home would be entirely underground, with the surface consisting only of light-pipe openings and ventilation embedded in garden space. (Where "garden" involves lots of native plants appropriate for the local climate and very little maintenance.) In exchange for some slight plumbing difficulties you get a gentle temperature year-round and dramatically increase the amount of outdoor-space available.
For infuriating, but far off-topic reasons, the place I work (an academic facility) has no choice but to purchase machines from Dell.
Each and every time, we have to have someone look up all the major hardware components to make sure they are linux compatible. Then they ship us a machine with windows on it. We wipe it, then we install linux from scratch.
If they offered us a linux option, it would be great. We would probably put our own distro on it anyway, but not having to worry about hardware issues would be a huge benefit. If they force us to first wipe some other distro, then we're not any worse off than we are now. If they offer a little "no OS" checkbox and a few bucks off the price, all the better. (That is, assuming they don't do something really nasty like provide proprietary drivers that only work with eccentrically patched kernels... which wouldn't be entirely surprising, except in so far as it would require more effort than just choosing well supported hardware in the first place.)
I agree with previous posters: complaining about diversity among distributions is nonsense. Pick a popular one, and you'll satisfy those who demand a works-out-of-the-box solution, and you'll come a lot closer to satisfying the rest of us.
As a fan of a distro that's no longer among the top 5, I don't expect to find my favorite distro on a vendor's PC. (Besides, even if they launched a Dell Slackware line, I'm sure their partition scheme wouldn't be perfect for me anyway...)
I *do* expect them to offer a real linux option, with comparable features to their windows line and advertised in such a way that it stands a snowball's chance in hell of not failing miserably.
Waiting to offer a genuine linux option until linux converges on a single type of distribution is, like most of Dell's history of linux "support", either foolish or a disingenuous strategy meant to fail.
But, I've worked in three somewhat different academic research environments.
1 - One central admin for all the desktop machines in a massive department, no one else gets root on any machine.
2 - One central admin who is mostly an advisor, people are allowed to administer their own desktop machines if they want.
3 - Free-for-all, in which most groups have one or two principle computer gurus who handle multi user servers and almost everyone administers their own desktop machines.
#3 is far and away the best. In #2, no one that I knew of actually took them up on the remote administration option, essentially reducing it to #3. #1 was a nightmate for everyone. When the deparment computing committee tried to talk everyone into switching to something closer to #1, we all resisted fiercely and eventually they backed down.
In an environment where people are actually using their computers as research tools, rather than as expensive notepads with which to writeup the results of their research, it pays to place control at the lowest feasible level. Every time a user is forced to ask someone else to fiddle with software, it adds *days* to what should be simple tasks.
Sure, you create an occasional security risk when a bad user fails to install patches. But, there's no comparison between the number of man hours spent on dealing with those sort of incidents and the amount of wasted energy in trying forcing every minor change to go through a central administrator.
In a computer lab or a corporate environment, you might be able to make a case for central administration. For academics, it's just crazy. (And I suspect enforcing it will just drive everyone to switch to personal laptops instead, in addition to pissing them all off.)
setup all of your environment variables there (aliases and anything else you want), and have SSH pass your environment variables when connecting to a new box.
Hmmm. I've always thought of aliases and variables as very different sorts of beasts. Is there a way to manipulate aliases as though they were environment variables? (In one of the common shells?) Sounds like it could be a handy trick.
Although I hesitate to offer any advice that will increase productivity in a company so unpleasant that they won't let their employees create.profile files, here's something that may work.
First of all, if you can possibly find a way to create your aliases within the shell of the remote machine, do so. There are two reasons:
1 - Only the remote shell knows enough about the context of your input to interpret aliases reliably. Your ssh client will be flying blind, and trying to get something that's functional enough to be useful and not dangerous will be tricky. Unless you do something very restrictive, such as making a special unique "I'm invoking an alias" command and then expanding the alias on the screen immediately, you're in for trouble. (One might argue that isn't really a proper alias at all, but rather a keystroke macro.) Anything more advanced will crash and burn the first time you try to open a file that happens to have the same name as one of your aliases.
If all you ever want to do is enter sequential shell commands, you could possibly fudge it by finding a way to send whole lines rather than characters to the remote machine and then moving all the readline stuff into your client. The out-of-the-box example would be running ssh within a shell within emacs. But, you'll lose meaningful tab completion and other context sensitive goodies, and it will break any program that needs to interact with a real terminal. And, you'll have to install custom software on the client machines, which one might expect to be a bigger deal than installing a single.profile file.
2 - The people who wrote your shell already worked out all the bugs when implementing aliases decades ago. You're less likely to find nasty surprises if you use a tool that's been used millions of times than something cobbled together by yourself. Sounds like nasty surprises aren't something your bosses will like.
Given that, here's one option that will probably work in most environments.
Create an ascii text file containing all your aliases in a format that can be evaluated without linebreaks. For a bash example, a series of lines such as the following:
alias hihi='echo "this is an alias"';
alias lt='ls -lrt';
Then put it on the web somewhere.
When you log in to the remote machine, run something like this (again using bash as an example):
Assuming the remote machine is connected to the web and has lynx, you're done. Make sure you read the output carefully at least once before you evaluate it in case there are any funny surprises due to line folding or nested quotes.
Now you've reduced things to typing in a single string every time you ssh in. There's probably some way to make a windows ssh client execute an arbitrary command and within an interactive shell, which could save you even that annoyance. (But it's been a long time since I've used windows, so I wouldn't have a clue how to do it. With openssh on a unix machine, it would be easy. I'm sure a windows equivalent exists.)
. . . (the consoles, the displays, the living quarters, the cargo storage, the doors and so on) are all items that could exist in the real world.
I don't disagree with you, but I do wonder if in thirty years they'll look as dated as the TOS toggle switches and goosneck viewers. (Don't get me wrong - I love TOS - but it does require a rather sympathetic viewing to accept their gizmos as the very best of 23rd century technology.)
So far, the TNG environment seems to have held up pretty well. Granted, they have the advantage of being much more recent and the luxury of a huge budget and top notch film crew. But, now coming up on almost twenty years since the pilot, the beautiful flat panel displays and miniaturized gizmos still work well.
If they had restricted themselves to available corner-store tech, I doubt that would be true. cm-thick tablet computers with touchscreens may not have been a revolutionary idea in 1987, but with six years before the Apple Newton hit the market, it wasn't quite off-the-shelf material.
When our children watch the NX-01 screens on their wall-sized high definition displays, or when their children watch them on their neural implants, they may not be so convincing.
There's certainly value in learning to do things by hand the first time around, especially if one is in a course dedicated to that particular thing. Banning symbolic integrators in calculus class or computer algebra systems in an linear algebra class is a very good idea. (I suppose my definition of a "calculator" may be a little outdated.) And, a well designed problem really shouldn't require complicated math tricks anyway, unless it's a well designed problem in complicated math tricks class.
On the other hand, requiring that someone who's already been through calculus do trig substitutions by hand each and every time they meet an appropriate integral, or forcing them to solve big systems of algebraic equations by hand in a quantum class seems pointless to me. There may be some cases where you can find an appropriate form for an answer more quickly by hand, or gain some insight into what's actually going on. But the vast majority of the time I find I just end up doing exactly what the computer would have done - a bunch of mechanical operations with no real physical meaning - except that I do it far more slowly and am more likely to make mistakes. (Requiring someon to multiply long strings of physical constants together without a four function calculator is even crazier, although I've met a few old farts who insist it's a vital skill and invent absurd scenarios to justify that claim.)
There are certainly people who I respect who will proudly tell you they do all their analytic work by hand and then sometimes check it using computers. I usually do exactly the opposite. Perhaps I'm missing out on an occasional insight - but I'd argue that I also have more time to spend looking for insights elsewhere. (Whether or not I find them is quite another matter, of course.)
A good analogy is in the hiring of a mathematician, and basing their being hired on their speed of their computational skills sans calculator.
A beautiful analogy.
And, one that brings us back to the author of the main article, who says in a throw-away comment, "Math is a different matter. No student should be allowed to bring a calculator into a math class. Ever."
It seems that in math class, as well as in his analysis of subject at hand, he missed the point completely.
I've heard, "I've never been able to spell well" as an excuse for poor spelling for years, but what that really says is, "I don't consider the details of proper spelling important, since people will understand what I mean anyway. If someone can't be bothered to attend to the relatively easy-to-manage detail of correct spelling, why should I think they'll pay enough attention to other details that aren't as easy to manage?"
Translation: I cannot conceive of the existence of someone whose natural abilities and strengths differ from mine, and therefore they must be lazy bastards who just don't care enough to bother learning what I consider easy.
Just imagine the response if I were to say, "I've heard, `I've never been able to learn multivariable calculus' as an excuse for a poor understanding of basic physics for years, but what that really says is, `I don't consider the details of the world around me important, since other people will probably overlook any mistakes I make when describing it.' If someone can't be bothered to understand the most basic aspects of the physical world with which they interact on a daily basis, why should I think they'll pay enough attention to other details that aren't as fundamental and immediately applicable to their daily lives?"
People would call me a lunatic if I said something like that in public, and yet writing a textbook that derives all of multivariable calculus and its applications from scratch is a trivial task compared to, say, memorizing the correct spelling for the 5'000 most common English words.
Spelling is by no means an easy task for everyone, even for many of us who haven't been diagnosed with a "legitimate difficulty with spelling." It may be true that, if we chose to dedicate a significant portion of our lives to memorizing words, we could achieve the level of competency that those with a natural affinity for the subject display. But then we'd never get anything done in those areas for which we have a genuine talent.
Fortunately, there are now tools available which allow poor spellers to communicate effectively even with those too narrow minded to overlook poor spelling. Expecting someone to use correct spelling when publishing electronic text is perfectly reasonable. Anyone who allows spelling errors to appear in electronic documents *is* being lazy, since there are so many free and painless tools available to translate our text into correctly spelled words.
To require that an applicant be capable of somehow generating correctly spelled text when necessary is appropriate; however, to demand that an applicant spell well when writing in pen and ink, when they are being hired for a job that doesn't require doing so on a regular basis, is just silly. Judge us by the job for which we are being hired, not by how closely our skill set happens to match yours.
The language itself, and the ability for a person to appropriately and effectively communicate concepts and ideas, has nothing to do with whether it is written on bark, slate, or paper, written with chalk, pencil, or pen.
Just add the implicit "computerized text editor" to the end of your list, and it becomes clear that this is not only relevant to the current conversation, it is a direct response to the article. Our self proclaimed luddite's argument (in the most generous possible sense of the word) is that email and instant messaging are the reason people can not appropriately and effectively communicate.
Another approach may be to find a university with fab facilities and then hire a recent grad to run your process for you.
Around here (UC Berkeley microlab) it's not too uncommon for someone who has recently graduated to start their own company and pay to use the facility. It will still be expensive, but it could be less expensive than trying to get a pro fab house to do the work. You'll probably have to dish out either a decent paycheck or some intellectual buy-in on the project in order to convince a competent person to work for you, though.
On the other hand, I've seen very few people doing vanilla transistor/resistor/capacitor type circuits in academic labs. It's possible that for standard designs and technology from a couple decades ago, you can find a professional shop to do the work at a lower cost.
Writers *do* convey tone, usually at the expense of brevity. I would say that most emails are written with brevity in mind (right or wrong, I'm not arguing) and tone is lost. Plus, writing with tone may sound stilted, even assuming that most people are capable of writing well enough to convey tone if they actually took the time to do so:)
Those of us who've had the occasional misfortune to grade college papers might question that last assumption. I can't say I'd be particularly surprised to hear that, five times out of ten, people fail to successfully communicate tone even when they have the luxury of two weeks prep time and multiple drafts.
As an aside, I should point out that I'm not by any means going to decry the failure of education or popular ignorance. There are an awful lot of things in this world that are more important than being able to successfully express or interpret tone in written work. It is, on the whole, pretty far down on the list of useful skills which most of us do not possess; somewhere between not knowing how to hygienically handle uncooked meat and not being able to diagnose common automotive problems. After a few drinks, I might even go one step further and argue that the tyranny of an education system overwhelmingly focused upon the humanities has robbed most people of the opportunity to learn something genuinely useful or interesting while in school. (10 points to anyone who can tell me whether or not that last statement was sincere. 20 points to anyone able to touch off a three day flame war on the topic.)
Back to the topic at hand, one might ask whether the problem lies primarily with the writer or the reader. Are there writers who are universally understood? Are there readers who are able to correctly decipher tone far more often than average?
Conducting a similar study involving multiple pairings would tell us whether it is possible to be particularly good at either task. That, in turn, might give us some hope for improvement.
Intuitively, it seems that there almost certainly exceptional writers, even when we confine ourselves to informal, chatty text. There are plenty of dialogue driven short stories and plays in which authors convey tone with precision despite the absence of internal context. (When Mark Twain and Nabokov exchange email, I imagine there are few misunderstandings.) How one ought to go about convincing the world to learn from them is another matter.
As is the question of whether misunderstandings explain flame wars. An awful lot of those seem to occur because people have *correctly* interpreted their opponents' statements. In a world without consequence, some people discover that it's fun to be an asshole. Then again, an asshole who expresses himself with precision and wit is a lot more fun to read than the participants in most flame wars.
These are not "hourly employees". If they were, they would get paid overtime or even the government would be violating labor laws. These are exempt [about.com] employees. They are not paid by the hour, so they can't be workng dozens more hours than they are getting paid for.
In many cases, that's true. It's not obviously true when it comes to students and post-docs who often submit monthly time cards with explicit hours listed on them and work a whole lot more than gets listed. One could probably make a case out of it, except that it's voluntary and the last thing in the world we want is someone to tell us not to work extra hours. (There are also plenty of examples of hourly employees who really are being taken advantage of, but they tend not to be at research labs, and using skype at work obviously isn't their biggest concern.)
But, even if you choose to argue that everyone involved is an exempt employee, then the claim that they're using skype on company time is no more valid.
You know, a job is a job. Stop whining, or find a better employer.
I'm not whining. I love what I do, and the institution in which I work, and the people around me. Wouldn't trade it for anything.
But, if for no good reason someone decided to block internet telephony, it would annoy me. It wouldn't be a big deal, but it also wouldn't serve any legitimate purpose. When people do pointless things that make my life more difficult (if only slightly), it bugs me.
I've pulled been unemployed, and I've been underemployed, and it sucks.
Hard to imagine, what with the charming personality you've displayed in this post.
There are worse problems to have than not being able to Skype at work.
Yup. But a tiny irritant is still an irritant.
I'm glad to have a job great enough that blocking skype at work actually would be a significant problem.
That assumes that the time you spend using their computer system isn't "your own time."
While that's true for an hourly wage type in an office building, it certainly isn't true for everyone, especially at a government lab.
At least around here, it isn't too uncommon for people to work many dozens of more hours than they're officially paid for in a week. In situations like that, allowing them to do something personal with the network that has negligable impact on anyone else is a no-brainer. Restricting it would just piss them off for no reason.
In my case, I'm in the lab during just about all hours that it's polite to call people in the local time zones. If I want to talk to anyone on the phone, I pretty much have to do it from work. That gives me two choices: use a cell phone and pay an outrageous fee, or use an online application like Skype. (Or use the land line and *really* abuse the lab resources.)
Fortunately, there's been no attempt to ban Skype or other telephony apps here.
At some of the security conscious national labs, there may be more restrictions. But applying them system wide to an institute as huge as Max Planck is just crazy.
If you ask me, any place that you're allowed to bring a cell phone, you should be allowed to install a telephone application. (Unless perhaps you're out in the field in some remote location with very limited network bandwidth.)
"Professor John Smith <jsmith@somewackydomain.com>"
Didn't realize the tags would get eaten even when posting in plain text. (Clearly this is some new definition of "plain old text" of which I was not previously aware...)
I can certainly see the advantage of using a personal domain for email. In particular, using a domain that isn't your isp is a must. I've known people trapped for years with a terrible ISP by the enormous amount of work required to change addresses.
But, it could also lead to serious trouble if your operational identity is closely tied to the where you work. If you're communicating with someone as a representative of your institution (or using your association with the institution to try to get something done that would be otherwise difficult), starting off with a homebrew email domain is risky.
For an academic, it strikes me as a particularly dangerous. Just imagine what your first thought would be if you received a cold letter from "Professor John Smith ". I'd guess that it won't be, "Oh, that must be that guy with a beard I chatted with at a conference last year." More likely is something along the lines of, "Is this spam? Some crank? Should I bother to open it to investigate?"
In a world where most email isn't worth reading and most people get too much of the stuff that is, it is a good idea to make your headers as obviously legitimate as possible. For an academic who probably has a fixed term of many years and can expect months of notice before an account is cancelled, changing addresses isn't really a huge problem.
Adding a personal address for friends and family can't hurt. But, if you're like me, the distinction between friends and colleagues is often imprecise. Even when it's not, juggling two different from-addresses and remembering who gets which is a pain.
Administering your own machine within your workplace may be a decent compromise, although you could lose your transitional buffer that way. Convincing your workplace to let you set up a.forward file and leave your account intact (if inaccessible to logins) for a few months is going to be a lot easier than convincing them to leave a personal machine running.
Because we are much more facile at using spoken language to be precise than we are at using mouse+keyboard to be precise, a "G+AUI" (graphical+audio user interface) should, in principle, be much more powerful than a GUI.
I'm not convinced that spoken language is more precise than any other form of interface. In fact, I'd suggest just the opposite.
When one wishes to communicate anything with precision, writing it down is likely to lead to far better results. For the really demanding material, diagrams, equations, and structured text generally accomplish the task much more easily than prose. Spoken language, on the other hand, is ideal for transmitting large amounts of imprecise information with little effort. (Well, that and poetry, which is by no means a trivial aspect of language, but one that seems largly unrelated to computer interfaces.)
If you forcibly deprive someone of sleep, they end up with physical brain damage and then die.
Got a reference for that? I'd be interested to hear about it. It might well be true, but it's certainly new to me. I'm not expert by any means. But, as an amateur sleep deprivation enthusiast I'm not totally ignorant of the subject.
I've never heard of humans dying, or suffering any long term ill effects associated with sleep deprivation. (Not counting what happens when you drive your car into a tree or hallucinate while in your boss's office.)
It's certainly true that when you deprive rats of sleep, they do become ill and die. But humans aren't rats. In particular, humans are able to rest while awake, we don't curl up with our tails every night to conserve body heat, and when we decide to engage in sleep deprivation it doesn't involve being repeatedly shocked or pushed around with a paddle for weeks on end without any explanation. (As I recall, everything in the above paragraph comes more or less directly from the excellent book by James Horne, Why We Sleep.)
You'd have to be going a lot faster than 6 times our current rotation to fall off.
The short proof - a low earth orbit satelite orbits in a little more than an hour, and it's only a tiny distance from the earth's surface compared to the earth's radius.
The long proof: the acceleration needed to keep us stuck to the earth is given by omega^2 r, where omega is our angular velocity. For omega = 2pi/day, that gives a=0.0034 m/s, or around 0.34% of the acceleration gravity at the earths surface.
A scales as omega^2, so if we increase the earth's rotation by a factor of 6, it's still only 12% of gravity. You'd just be able to jump a little higher and climb mountains a little less high (without oxygen tanks).
To fall off (well, to orbit at the earth's surface, if that counts as falling off), you need to get up to 17 times our current rotation speed. Of course the earth would change a bit as well, so it might not be as much fun as one might think.
Yup, the linux version of lynx does that too. (Or at least it did last time I used it.)
For some reason links doesn't. (Or at least didn't the last time I used it.) Elinks does, however.
I've also been amazed that, as far as I know, the only way to get the same functionality in mozilla/firefox is with a badly maintained third party plugin called moxez. Trying to write anything more than a sentence long with a built in browser editor drives me up a wall. It's hard to believe that among all the mozilla developers out there no one else thinks it am important feature.
On a side note, how does one post on/. with Lynx / Links (text mode) / eLinks?
Hmm. Perhaps I'm missing something. Why should it be any different from posting with graphical browser?
If you're not logged into an account, you have to invoke an external viewer to read the capcha. But, if you log in and allow cookies, everything works without a hitch.
Were there really people who were smart enough to be interested in slashdot, but too stupid to figure out the section concept?
Clearly the answer is yes.
The immediate followup question - have they any chance in hell at understanding the new configuration options?
I make no claims at web navigation brilliance (although I suppose I should be proud to have immediately intuited what "2 more" meant), but I found the new interface damned confusing. Thought it was a display bug at first and started to write up a report before realizing that it was intentional.
There's nothing in the arrangement of little black boxes to indicate that you're reading a single row of options with two components each. It's more or less impossible to understand without reading the "learn more" page. Some sort of vertical elements to group each column sure would help. Adding a row label that says something like "main articles from this section" and "section only articles" would clarify things tremendously. Even the text on the "learn more" page is confusing if, like me, you don't instantly recall the fact that every story including those on the main page is categorized in a section.
But at least my long running (although not particularly passionate) complaint has been resolved. It used to be that the "n more" tags referred only to the vanilla main page, not to the actual main page each person saw. Thus, if you placed everything from particular sections on the main page, the "n more" tags always read high. Of course if you happen to remember which categories you dumped on the main page, you could filter it by hand. But, those of us with poor memories spent a lot of time clocking on categories and finding no new stories.
Now if they didn't hard limit the max stories option to 30, everything would be perfect.
Slashdot Mirror
> KDE centralizes these features
/bin/bash
r ch?hl=en&q="
f or i in $@;
= "/usr/local/firefox"
>so you don't have to be in a web
>browser to use them.
>KDE: When you're in any application
>press ALT+F2, type gg:foo, press ENTER
>Firefox: Start or switch to Firefox,
>CTRL+K, type foo, press ENTER -or-
>Start or switch to Firefox, CTRL+L,
>type "searchkeyword foo", press ENTER
But it takes a couple minutes of setup to be able to do the same thing with firefox (or any browser you prefer, such as my favorite, elinks).
The only question is why the KDE folks working on minicli didn't give their users the option to use a different browser in the first place. It would be a trivial change, and would make that feature a lot more useful. (I'd offer a possible answer, but it wouldn't be polite.)
Here's a quick example of how to get the same feature in firefox. Create an executable file somewhere in your path called "gg" containing this:
#---------------begin file gg
#!
SEARCHURL="http://www.google.com/sea
SEARCHURL="${SEARCHURL}$1"
shift
do
SEARCHURL="${SEARCHURL}+$i"
done
FIREFOX_PATH
if [ -z "`ps x | grep \"[0-9] ${FIREFOX_PATH}/firefox-bin\"`" ]; then
# No MozillaFirebird running
${FIREFOX_PATH}/firefox $SEARCHURL
else
# MozillaFirebird running - open a new window
${FIREFOX_PATH}/firefox -a firefox -remote "openURL($SEARCHURL,new-tab)"
fi
#---------------end file gg
Now type "gg some words" from the KDE minicli (or any terminal, the icewm bar, etc) and get the same effect.
. . . and, you can become the most-watched guy on the subway. If you hang out in the places I do, standing around conspicuously holding a multi-hundred-dollar-device makes for some unpleasant social interactions.
I actually read a lot of ebooks. I work part of the day in a cleanroom, where I often have many 10 minute spells waiting for equipment with nothing to do. Since we can't bring in regular paper, a portable device is perfect. (In my case a first generation high-resolution palm pilot from a few years back that I got on the cheap.)
Also, as someone who sleeps significantly fewer hours than his mate, being able to read in bed with the lights off is a big plus. There's enough drm-free content out there from Project Guttenberg, Cory Doctorow, online SF mags, etc. to keep me occupied for a long time. I'm looking forward to trying out the Baen site mentioned here. (Sorry to see how terrible their web site looks. Not because I care -- its the books that matter, not their graphic design -- but because I want drm-free ebook distributors to succeed in the world, and their site screams "don't take me seriously.")
But, there's always going to be a place for paper books, if only because an individual paperback is almost worthless. It will be a long, long time before there's an electronic book reader that I'd feel good about reading on a bus in a developing country.
That's a good idea.
It constrains the possible geometries a bit, but has the great advantage of being super cheap and waterproof. I always imagined concentric poured concrete shells and sump-pumps, but if you can get an outer shell that's really impermeable, you're set.
I suppose cylindrical underground homes are less likely to appreciate in value than rectangular underground homes, but as we've already abandoned the home-as-investment model by moving underground in the first place, so who cares?
It seems to me neighborhoods would be a lot more interesting if people built the homes they most wanted to live in rather than the ones they think will net them the biggest profit on resale. We can't really be a society of people who think square boxes made of off-white painted dry wall are the best imaginable environment. Not sure how you convince an individual to make that choice, though. (For me, the choice is obvious. Which is why I'll almost certainly not become rich, although I have every intention of remaining happy...)
Perhaps the take away message from both posts is that power to control a system ought to reside with whoever is most invested in making sure it works, where "works" is defined by context. For a bank, not crashing the existing system may be the most important thing, and if the sysadmin is responsible for preventing that from happening, he should be given the power to do so.
For an academic, on the other hand, getting something to run at all, even if it locks up the machine and occasionally forces a hard reset is often good enough. Good enough, here, means that spending any more time on it that could be spent elsewhere would be wasteful.
In my experience (in physical sciences research labs), the only people genuinely invested in making sure software works are students and scientists who need it for a particular project. Full time computer people are great when we need to turn to someone for help, or for maintaining shared servers or backup systems, but when you're forced to turn to them for every minor change to a desktop it becomes impossible to get any work done. You spend all your time trying to convince them that something is worth doing, and then waiting for them to do it.
As far as I can tell, just about all the simulation, data analysis, and data acquisition code in the world is crap software that no self-respecting sysadmin would allow on a well run system. (The exception may be some of the stuff run by multi user computing facilities that hire real computer people, which I've never worked on.) But, it's how work gets done. Anything else would be prohibitively expensive.
We spend almost all of our computer time either writing our own buggy, amateurish software or trying to install and configure packages made by some other amateur coder who didn't pay nearly enough attention to portability and standards. It's the sort of annoying trial-and-error, one-off debugging that would just piss off a full time sysadmin, who is likely to either spend a *very* long time doing it or simply refuse and claim it impossible.
Just imagine what the average 8-5 computer admin type would say if you called them up on a friday evening and said, "we want to try reverting to the comedi libraries from cvs dated roughly six weeks ago just to see if maybe it fixes these random lockups that we've been experiencing. Let's also build the stuff from six months back too in case we want to try that. Can you have that done in time for Joe's run at midnight tonight?" They'd tell you to go to hell, with good reason. On the other hand, Joe may be perfectly happy to do the work himself on the off chance it will work. (In which case *not* blaming the computer admin guys when Joe trashes his system and has to reinstall everything is a very good idea.)
Asking a central administrator to debug code that's only ever going to run on a single dedicated machine in someone's lab for a week is a waste of everyone's time.
Conduit is the best possible answer.
Whatever you choose to run through the conduit will be dated in a few years. But leaving yourself with as much easily accessible conduit as possible will always be useful. Use the largest diameter conduit your building codes will allow, and run it *everywhere*, and fill it with a bundle of pull-wires. Put in multiple panel boxes per wall, at floor, ceiling, and table rail height. Record their locations well, and then wall over them.
Then in the future when the residents decide to turn a living room into an office/nursery/VR-gymnasium, they can run whatever power or information standards they want through the conduit without destroying whole walls.
Anything beyond that is up to the fetish of the homeowner. My own personal dream home would be entirely underground, with the surface consisting only of light-pipe openings and ventilation embedded in garden space. (Where "garden" involves lots of native plants appropriate for the local climate and very little maintenance.) In exchange for some slight plumbing difficulties you get a gentle temperature year-round and dramatically increase the amount of outdoor-space available.
A very good point.
For infuriating, but far off-topic reasons, the place I work (an academic facility) has no choice but to purchase machines from Dell.
Each and every time, we have to have someone look up all the major hardware components to make sure they are linux compatible. Then they ship us a machine with windows on it. We wipe it, then we install linux from scratch.
If they offered us a linux option, it would be great. We would probably put our own distro on it anyway, but not having to worry about hardware issues would be a huge benefit. If they force us to first wipe some other distro, then we're not any worse off than we are now. If they offer a little "no OS" checkbox and a few bucks off the price, all the better. (That is, assuming they don't do something really nasty like provide proprietary drivers that only work with eccentrically patched kernels... which wouldn't be entirely surprising, except in so far as it would require more effort than just choosing well supported hardware in the first place.)
I agree with previous posters: complaining about diversity among distributions is nonsense. Pick a popular one, and you'll satisfy those who demand a works-out-of-the-box solution, and you'll come a lot closer to satisfying the rest of us.
As a fan of a distro that's no longer among the top 5, I don't expect to find my favorite distro on a vendor's PC. (Besides, even if they launched a Dell Slackware line, I'm sure their partition scheme wouldn't be perfect for me anyway...)
I *do* expect them to offer a real linux option, with comparable features to their windows line and advertised in such a way that it stands a snowball's chance in hell of not failing miserably.
Waiting to offer a genuine linux option until linux converges on a single type of distribution is, like most of Dell's history of linux "support", either foolish or a disingenuous strategy meant to fail.
But, I've worked in three somewhat different academic research environments.
1 - One central admin for all the desktop machines in a massive department, no one else gets root on any machine.
2 - One central admin who is mostly an advisor, people are allowed to administer their own desktop machines if they want.
3 - Free-for-all, in which most groups have one or two principle computer gurus who handle multi user servers and almost everyone administers their own desktop machines.
#3 is far and away the best. In #2, no one that I knew of actually took them up on the remote administration option, essentially reducing it to #3. #1 was a nightmate for everyone. When the deparment computing committee tried to talk everyone into switching to something closer to #1, we all resisted fiercely and eventually they backed down.
In an environment where people are actually using their computers as research tools, rather than as expensive notepads with which to writeup the results of their research, it pays to place control at the lowest feasible level. Every time a user is forced to ask someone else to fiddle with software, it adds *days* to what should be simple tasks.
Sure, you create an occasional security risk when a bad user fails to install patches. But, there's no comparison between the number of man hours spent on dealing with those sort of incidents and the amount of wasted energy in trying forcing every minor change to go through a central administrator.
In a computer lab or a corporate environment, you might be able to make a case for central administration. For academics, it's just crazy. (And I suspect enforcing it will just drive everyone to switch to personal laptops instead, in addition to pissing them all off.)
Hmmm. I've always thought of aliases and variables as very different sorts of beasts. Is there a way to manipulate aliases as though they were environment variables? (In one of the common shells?) Sounds like it could be a handy trick.
Although I hesitate to offer any advice that will increase productivity in a company so unpleasant that they won't let their employees create .profile files, here's something that may work.
.profile file.
First of all, if you can possibly find a way to create your aliases within the shell of the remote machine, do so. There are two reasons:
1 - Only the remote shell knows enough about the context of your input to interpret aliases reliably. Your ssh client will be flying blind, and trying to get something that's functional enough to be useful and not dangerous will be tricky. Unless you do something very restrictive, such as making a special unique "I'm invoking an alias" command and then expanding the alias on the screen immediately, you're in for trouble. (One might argue that isn't really a proper alias at all, but rather a keystroke macro.) Anything more advanced will crash and burn the first time you try to open a file that happens to have the same name as one of your aliases.
If all you ever want to do is enter sequential shell commands, you could possibly fudge it by finding a way to send whole lines rather than characters to the remote machine and then moving all the readline stuff into your client. The out-of-the-box example would be running ssh within a shell within emacs. But, you'll lose meaningful tab completion and other context sensitive goodies, and it will break any program that needs to interact with a real terminal. And, you'll have to install custom software on the client machines, which one might expect to be a bigger deal than installing a single
2 - The people who wrote your shell already worked out all the bugs when implementing aliases decades ago. You're less likely to find nasty surprises if you use a tool that's been used millions of times than something cobbled together by yourself. Sounds like nasty surprises aren't something your bosses will like.
Given that, here's one option that will probably work in most environments.
Create an ascii text file containing all your aliases in a format that can be evaluated without linebreaks. For a bash example, a series of lines such as the following:
alias hihi='echo "this is an alias"';
alias lt='ls -lrt';
Then put it on the web somewhere.
When you log in to the remote machine, run something like this (again using bash as an example):
eval `lynx -dump "http://someserver/somefile.bash.txt"`
Assuming the remote machine is connected to the web and has lynx, you're done. Make sure you read the output carefully at least once before you evaluate it in case there are any funny surprises due to line folding or nested quotes.
Now you've reduced things to typing in a single string every time you ssh in. There's probably some way to make a windows ssh client execute an arbitrary command and within an interactive shell, which could save you even that annoyance. (But it's been a long time since I've used windows, so I wouldn't have a clue how to do it. With openssh on a unix machine, it would be easy. I'm sure a windows equivalent exists.)
Good luck.
I don't disagree with you, but I do wonder if in thirty years they'll look as dated as the TOS toggle switches and goosneck viewers. (Don't get me wrong - I love TOS - but it does require a rather sympathetic viewing to accept their gizmos as the very best of 23rd century technology.)
So far, the TNG environment seems to have held up pretty well. Granted, they have the advantage of being much more recent and the luxury of a huge budget and top notch film crew. But, now coming up on almost twenty years since the pilot, the beautiful flat panel displays and miniaturized gizmos still work well.
If they had restricted themselves to available corner-store tech, I doubt that would be true. cm-thick tablet computers with touchscreens may not have been a revolutionary idea in 1987, but with six years before the Apple Newton hit the market, it wasn't quite off-the-shelf material.
When our children watch the NX-01 screens on their wall-sized high definition displays, or when their children watch them on their neural implants, they may not be so convincing.
You make a very good point.
There's certainly value in learning to do things by hand the first time around, especially if one is in a course dedicated to that particular thing. Banning symbolic integrators in calculus class or computer algebra systems in an linear algebra class is a very good idea. (I suppose my definition of a "calculator" may be a little outdated.) And, a well designed problem really shouldn't require complicated math tricks anyway, unless it's a well designed problem in complicated math tricks class.
On the other hand, requiring that someone who's already been through calculus do trig substitutions by hand each and every time they meet an appropriate integral, or forcing them to solve big systems of algebraic equations by hand in a quantum class seems pointless to me. There may be some cases where you can find an appropriate form for an answer more quickly by hand, or gain some insight into what's actually going on. But the vast majority of the time I find I just end up doing exactly what the computer would have done - a bunch of mechanical operations with no real physical meaning - except that I do it far more slowly and am more likely to make mistakes. (Requiring someon to multiply long strings of physical constants together without a four function calculator is even crazier, although I've met a few old farts who insist it's a vital skill and invent absurd scenarios to justify that claim.)
There are certainly people who I respect who will proudly tell you they do all their analytic work by hand and then sometimes check it using computers. I usually do exactly the opposite. Perhaps I'm missing out on an occasional insight - but I'd argue that I also have more time to spend looking for insights elsewhere. (Whether or not I find them is quite another matter, of course.)
A beautiful analogy.
And, one that brings us back to the author of the main article, who says in a throw-away comment, "Math is a different matter. No student should be allowed to bring a calculator into a math class. Ever."
It seems that in math class, as well as in his analysis of subject at hand, he missed the point completely.
Translation: I cannot conceive of the existence of someone whose natural abilities and strengths differ from mine, and therefore they must be lazy bastards who just don't care enough to bother learning what I consider easy.
Just imagine the response if I were to say, "I've heard, `I've never been able to learn multivariable calculus' as an excuse for a poor understanding of basic physics for years, but what that really says is, `I don't consider the details of the world around me important, since other people will probably overlook any mistakes I make when describing it.' If someone can't be bothered to understand the most basic aspects of the physical world with which they interact on a daily basis, why should I think they'll pay enough attention to other details that aren't as fundamental and immediately applicable to their daily lives?"
People would call me a lunatic if I said something like that in public, and yet writing a textbook that derives all of multivariable calculus and its applications from scratch is a trivial task compared to, say, memorizing the correct spelling for the 5'000 most common English words.
Spelling is by no means an easy task for everyone, even for many of us who haven't been diagnosed with a "legitimate difficulty with spelling." It may be true that, if we chose to dedicate a significant portion of our lives to memorizing words, we could achieve the level of competency that those with a natural affinity for the subject display. But then we'd never get anything done in those areas for which we have a genuine talent.
Fortunately, there are now tools available which allow poor spellers to communicate effectively even with those too narrow minded to overlook poor spelling. Expecting someone to use correct spelling when publishing electronic text is perfectly reasonable. Anyone who allows spelling errors to appear in electronic documents *is* being lazy, since there are so many free and painless tools available to translate our text into correctly spelled words.
To require that an applicant be capable of somehow generating correctly spelled text when necessary is appropriate; however, to demand that an applicant spell well when writing in pen and ink, when they are being hired for a job that doesn't require doing so on a regular basis, is just silly. Judge us by the job for which we are being hired, not by how closely our skill set happens to match yours.
Another approach may be to find a university with fab facilities and then hire a recent grad to run your process for you.
Around here (UC Berkeley microlab) it's not too uncommon for someone who has recently graduated to start their own company and pay to use the facility. It will still be expensive, but it could be less expensive than trying to get a pro fab house to do the work. You'll probably have to dish out either a decent paycheck or some intellectual buy-in on the project in order to convince a competent person to work for you, though.
On the other hand, I've seen very few people doing vanilla transistor/resistor/capacitor type circuits in academic labs. It's possible that for standard designs and technology from a couple decades ago, you can find a professional shop to do the work at a lower cost.
Those of us who've had the occasional misfortune to grade college papers might question that last assumption. I can't say I'd be particularly surprised to hear that, five times out of ten, people fail to successfully communicate tone even when they have the luxury of two weeks prep time and multiple drafts.
As an aside, I should point out that I'm not by any means going to decry the failure of education or popular ignorance. There are an awful lot of things in this world that are more important than being able to successfully express or interpret tone in written work. It is, on the whole, pretty far down on the list of useful skills which most of us do not possess; somewhere between not knowing how to hygienically handle uncooked meat and not being able to diagnose common automotive problems. After a few drinks, I might even go one step further and argue that the tyranny of an education system overwhelmingly focused upon the humanities has robbed most people of the opportunity to learn something genuinely useful or interesting while in school. (10 points to anyone who can tell me whether or not that last statement was sincere. 20 points to anyone able to touch off a three day flame war on the topic.)
Back to the topic at hand, one might ask whether the problem lies primarily with the writer or the reader. Are there writers who are universally understood? Are there readers who are able to correctly decipher tone far more often than average?
Conducting a similar study involving multiple pairings would tell us whether it is possible to be particularly good at either task. That, in turn, might give us some hope for improvement.
Intuitively, it seems that there almost certainly exceptional writers, even when we confine ourselves to informal, chatty text. There are plenty of dialogue driven short stories and plays in which authors convey tone with precision despite the absence of internal context. (When Mark Twain and Nabokov exchange email, I imagine there are few misunderstandings.) How one ought to go about convincing the world to learn from them is another matter.
As is the question of whether misunderstandings explain flame wars. An awful lot of those seem to occur because people have *correctly* interpreted their opponents' statements. In a world without consequence, some people discover that it's fun to be an asshole. Then again, an asshole who expresses himself with precision and wit is a lot more fun to read than the participants in most flame wars.
In many cases, that's true. It's not obviously true when it comes to students and post-docs who often submit monthly time cards with explicit hours listed on them and work a whole lot more than gets listed. One could probably make a case out of it, except that it's voluntary and the last thing in the world we want is someone to tell us not to work extra hours. (There are also plenty of examples of hourly employees who really are being taken advantage of, but they tend not to be at research labs, and using skype at work obviously isn't their biggest concern.)
But, even if you choose to argue that everyone involved is an exempt employee, then the claim that they're using skype on company time is no more valid.
I'm not whining. I love what I do, and the institution in which I work, and the people around me. Wouldn't trade it for anything.
But, if for no good reason someone decided to block internet telephony, it would annoy me. It wouldn't be a big deal, but it also wouldn't serve any legitimate purpose. When people do pointless things that make my life more difficult (if only slightly), it bugs me.
Hard to imagine, what with the charming personality you've displayed in this post.
Yup. But a tiny irritant is still an irritant.
I'm glad to have a job great enough that blocking skype at work actually would be a significant problem.
That assumes that the time you spend using their computer system isn't "your own time."
While that's true for an hourly wage type in an office building, it certainly isn't true for everyone, especially at a government lab.
At least around here, it isn't too uncommon for people to work many dozens of more hours than they're officially paid for in a week. In situations like that, allowing them to do something personal with the network that has negligable impact on anyone else is a no-brainer. Restricting it would just piss them off for no reason.
In my case, I'm in the lab during just about all hours that it's polite to call people in the local time zones. If I want to talk to anyone on the phone, I pretty much have to do it from work. That gives me two choices: use a cell phone and pay an outrageous fee, or use an online application like Skype. (Or use the land line and *really* abuse the lab resources.)
Fortunately, there's been no attempt to ban Skype or other telephony apps here.
At some of the security conscious national labs, there may be more restrictions. But applying them system wide to an institute as huge as Max Planck is just crazy.
If you ask me, any place that you're allowed to bring a cell phone, you should be allowed to install a telephone application. (Unless perhaps you're out in the field in some remote location with very limited network bandwidth.)
Doh! That should have said,
"Professor John Smith <jsmith@somewackydomain.com>"
Didn't realize the tags would get eaten even when posting in plain text. (Clearly this is some new definition of "plain old text" of which I was not previously aware...)
I can certainly see the advantage of using a personal domain for email. In particular, using a domain that isn't your isp is a must. I've known people trapped for years with a terrible ISP by the enormous amount of work required to change addresses.
.forward file and leave your account intact (if inaccessible to logins) for a few months is going to be a lot easier than convincing them to leave a personal machine running.
But, it could also lead to serious trouble if your operational identity is closely tied to the where you work. If you're communicating with someone as a representative of your institution (or using your association with the institution to try to get something done that would be otherwise difficult), starting off with a homebrew email domain is risky.
For an academic, it strikes me as a particularly dangerous. Just imagine what your first thought would be if you received a cold letter from "Professor John Smith ". I'd guess that it won't be, "Oh, that must be that guy with a beard I chatted with at a conference last year." More likely is something along the lines of, "Is this spam? Some crank? Should I bother to open it to investigate?"
In a world where most email isn't worth reading and most people get too much of the stuff that is, it is a good idea to make your headers as obviously legitimate as possible. For an academic who probably has a fixed term of many years and can expect months of notice before an account is cancelled, changing addresses isn't really a huge problem.
Adding a personal address for friends and family can't hurt. But, if you're like me, the distinction between friends and colleagues is often imprecise. Even when it's not, juggling two different from-addresses and remembering who gets which is a pain.
Administering your own machine within your workplace may be a decent compromise, although you could lose your transitional buffer that way. Convincing your workplace to let you set up a
I'm not convinced that spoken language is more precise than any other form of interface. In fact, I'd suggest just the opposite.
When one wishes to communicate anything with precision, writing it down is likely to lead to far better results. For the really demanding material, diagrams, equations, and structured text generally accomplish the task much more easily than prose. Spoken language, on the other hand, is ideal for transmitting large amounts of imprecise information with little effort. (Well, that and poetry, which is by no means a trivial aspect of language, but one that seems largly unrelated to computer interfaces.)
Got a reference for that? I'd be interested to hear about it. It might well be true, but it's certainly new to me. I'm not expert by any means. But, as an amateur sleep deprivation enthusiast I'm not totally ignorant of the subject.
I've never heard of humans dying, or suffering any long term ill effects associated with sleep deprivation. (Not counting what happens when you drive your car into a tree or hallucinate while in your boss's office.)
It's certainly true that when you deprive rats of sleep, they do become ill and die. But humans aren't rats. In particular, humans are able to rest while awake, we don't curl up with our tails every night to conserve body heat, and when we decide to engage in sleep deprivation it doesn't involve being repeatedly shocked or pushed around with a paddle for weeks on end without any explanation. (As I recall, everything in the above paragraph comes more or less directly from the excellent book by James Horne, Why We Sleep.)
Nope.
You'd have to be going a lot faster than 6 times our current rotation to fall off.
The short proof - a low earth orbit satelite orbits in a little more than an hour, and it's only a tiny distance from the earth's surface compared to the earth's radius.
The long proof: the acceleration needed to keep us stuck to the earth is given by omega^2 r, where omega is our angular velocity. For omega = 2pi/day, that gives a=0.0034 m/s, or around 0.34% of the acceleration gravity at the earths surface.
A scales as omega^2, so if we increase the earth's rotation by a factor of 6, it's still only 12% of gravity. You'd just be able to jump a little higher and climb mountains a little less high (without oxygen tanks).
To fall off (well, to orbit at the earth's surface, if that counts as falling off), you need to get up to 17 times our current rotation speed. Of course the earth would change a bit as well, so it might not be as much fun as one might think.
For some reason links doesn't. (Or at least didn't the last time I used it.) Elinks does, however.
I've also been amazed that, as far as I know, the only way to get the same functionality in mozilla/firefox is with a badly maintained third party plugin called moxez. Trying to write anything more than a sentence long with a built in browser editor drives me up a wall. It's hard to believe that among all the mozilla developers out there no one else thinks it am important feature.
Hmm. Perhaps I'm missing something. Why should it be any different from posting with graphical browser?
If you're not logged into an account, you have to invoke an external viewer to read the capcha. But, if you log in and allow cookies, everything works without a hitch.
Clearly the answer is yes.
The immediate followup question - have they any chance in hell at understanding the new configuration options?
I make no claims at web navigation brilliance (although I suppose I should be proud to have immediately intuited what "2 more" meant), but I found the new interface damned confusing. Thought it was a display bug at first and started to write up a report before realizing that it was intentional.
There's nothing in the arrangement of little black boxes to indicate that you're reading a single row of options with two components each. It's more or less impossible to understand without reading the "learn more" page.
Some sort of vertical elements to group each column sure would help. Adding a row label that says something like "main articles from this section" and "section only articles" would clarify things tremendously. Even the text on the "learn more" page is confusing if, like me, you don't instantly recall the fact that every story including those on the main page is categorized in a section.
But at least my long running (although not particularly passionate) complaint has been resolved. It used to be that the "n more" tags referred only to the vanilla main page, not to the actual main page each person saw. Thus, if you placed everything from particular sections on the main page, the "n more" tags always read high. Of course if you happen to remember which categories you dumped on the main page, you could filter it by hand. But, those of us with poor memories spent a lot of time clocking on categories and finding no new stories.
Now if they didn't hard limit the max stories option to 30, everything would be perfect.