Looking for a programming job within a research group is certainly the approach which is safest and most likely to succeed. There are plenty of positions for programmers in medium sized physical science projects. Unfortunately, at least in my experience (US academic astrophysics), a rather large number of jobs are already filled long before they are advertised publicly. Trying to get your first science job may take a long time without any insider contacts. But, getting your *second* science job will be much easier, assuming you do excellent work.
I've known and taught a few people who started a physics undergrad degree from scratch in their middle thirties. It's certainly possible, but it won't be entirely easy. If you decide to take that route, you'll be marked as an outsider from the start in a highly collaborative field. Expect to be referred to as "you know, the older guy" by your peers, and asked completely inappropriate questions by lecturers at the start of every semester. Meanwhile, you'll be working hundred hour weeks, as often as not on material that you find irrelevant and unengaging. And, you'll probably be facing seven to twelve years of school before you can apply for permanent positions with any hope of success. (The first for a MS and a high level tech job, the second for a PhD and a chance to do truly independent research. A BS and a lot of prior experience *can* land you an interesting job with some creative freedom, but its far from guranteed.)
Those who pull it off tend to be outgoing people with a great deal of personal charisma and an overwhelming passion for the field. It's not a decision to be taken lightly.
On the other hand, the field and the world in general would be better off if more people joined late in life. Anything that brings someone other than 20 year old white boys into the department is a good thing, if you ask me. But, the personal cost shouldn't be discounted.
If you can learn RPN, a used Palm T5 with an HP emulator will cost you less than a crappy TI (which are crap) plus will give you a huge advantage in college math/engineering courses later on.
I've got to disagree.
If you ask me, the single most important thing about a calculator is its user interface. In particular, button layout and the ease with which you can access commonly used functions, particularly with one hand while you read them off a blackboard/textbook/sheet of paper.
I still pull out my battered, ten year old TI-85 when I want to plug a bunch of physical constants into an equation, even when I'm sitting in front of a computer with orders of magnitude more processing power and infinite configurability.
As a mechanism for inputing data into a touch screen device, a calculator interface may be the best option. But as a replacement for a calculator with real mechanical buttons, a touchscreen running an emulator leaves out the only feature that makes a calculator actually useful in a world where everyone has a computer on thier desktop. It also leaves you with a device that's much easier to break, more likely to get stolen, and has dramatically reduced battery life.
I haven't looked at calculator in a decade or so, but the older TIs are built like bricks. You can drop them down a flight of stairs, and they'll run for months of weekly 40 hour problems sets on a set of AA batteries.
So there is correlation - build a healthy place and the people who live in there are healthy.
That's true. But, does building a healthy place create more healthy people, or does it just pull healthy people from elsewhere? (I note and appreciate that you were careful not to mention a causal relationship.)
Boulder also has some advantages: a wealthy population (median family income of 70K $US), internationally renown outdoor recreation within minutes of downtown, a large university population, and a reputation for progressive culture and politics that tends to pull such people from the surrounding states.
All those characteristics seem likely to give you both a healthy population and a city council interested in maintaining a walkable city. If you were to forcibly swap the populations with a boulder-sized Atlanta suburb, it's not at all clear what would happen. (Except that after a few weeks it would be possible to find edible food in Georgia, which might be nice for those poor souls who occasionally find themselves trapped there.)
That's not to say that making other cities walkable wouldn't in itself be a good thing, even if it didn't do anything to improve the health of most of their residents.
Why would anyone want to encourage people who don't want kids to have them? Maybe to validate their own decisions?
Yup. I've always been amazed by this response.
Although I've never felt the slightest urge to have children, I'm told by reliable people that it is a genuine sensation that many experience. Given that, I can sympathize with those who decide to have children even though I'd argue that they probably shouldn't, given their dedication to other things which aren't easily compatible with family life. (I do sometimes wonder how much better off we'd be if fewer great scientists and artists gave up on their careers in order to become mediocre parents and later resentful, bitter retirees.)
But the people who try to talk *others* into having kids astounds me. Even leaving off the global ethical issues and the welfare of the children, it seems insane to urge someone to commit to 20 years of *anything* about which they're not already overwhelmingly enthusiastic.
The linux audio users and linux audio developers lists are vibrant (perhaps overwhelming) and their archives and associated documents and HOWTOs contain more information than you could possibly want.
Personally, I've had very good experiences with:
ecasound (multitrack recording, processing, general all-around fiddling)
If some poor schmoe is able to do all your multi-year work in a few days then maybe you deserve it.
Perhaps, but a more relevant case is that some poor schmoe with the same knowledge and tools as you but with more time to devote to the analysis of a particular data set is able to do your work a little bit faster than you.
In the field where I work (cosmology with ground based CMB telescopes), a team of twenty people can easily work full time for years building an instrument. Tends of man-years are spent installing and calibrating it, figuring out where to point it, and writing software to capture and preprocess the data. Then someone has to manually set up and run the instrument while it's taking data.
At the end of the day, the "unreduced" data that comes streaming out of the back side of our instrument is the result of years of work. If we put it up live immediately, someone could easily take it and publish a map before we could, especially if they knew or could guess something about the form of the data, which wouldn't be too hard, given the proposals and conference talks floating around. What's more, our competitor would have the benefit of months of leisure during which to try out data analysis algorithms before the data appears, while we're all working 100 hour weeks just assembling and testing the thing.
Now, you can certainly bring a few dedicated analysis guys into the collaboration early on in an effort to get a full data analysis pipeline up and running as soon as possible. They're likely to have a slight advantage over competitors, given advanced knowledge of the instrument and form of the raw data. But, artificially broadening a collaboration solely in order to stand a fighting chance at competing with outsiders has its own drawbacks. Handing off all of the analysis to someone else isn't particularly more attractive just because the other guy agrees to put your name on the paper.
Under those circumstances, it seems entirely reasonable to me to demand that data be made public eventually, say when the first results paper based upon it is approved for publication, or after a fixed window of time. But, requiring immediate release provides little benefit could do a great deal of harm.
When it comes to a facilities instrument, like the Hubble, a more demanding release schedule may be appropriate. But, even when data formats and instrument parameters are fixed and researchers don't need to build instruments, it still takes a nontrivial amount of time and effort to choose an observing strategy, submit proposals, arrange for followups, etc. And, as in the parent poster's example, there are times when the rush to publish first could easily reward people who are willing to run with a barely-adequate measurement instead of those who do things right. (That happens anyway in some cases, but unrestricted immediate data release could only hurt.)
There's nothing difficult about adding 5/8 + 31/32, unless your school failed to teach you basic fractional arithematic. That's 51/32, I didn't even have count on my fingers, I could tell you that at a glance. And I'm really not very good at math. But come on, this is third grade stuff!
It may not be *difficult* in the absolute sense, but it's still a lot more difficult than *not* having to do it. There are a whole lot of us who have to do it often enough to be annoyed, but not quite often enough to memorize all possible combinations so that we can do it without thinking.
And when you ask, "is a 3/8 gap a loose fit to 7/62+7/32+1/16?" it becomes a genuine annoyance. I spend my days with physics grad students and expert machinists, and have only met one or two who wouldn't want check their answer with paper or calculator if they were working on something important.
By comparison, comparing.24 to.1+.06+.085 is utterly trivial.
Any modern human who isn't on crack would probably agree that choosing base-10 relationships is a good idea. We've got to pick something, and picking the same system for everything except where hardware limitations make it impractical is a good idea. But, it is interesting to ask how large a base unit ought to be.
It seems to me one ought to pick them so that human scale objects are usually of order 10 of them. That way you can generally refer to the size of something with within 10% with a single, small whole number, and have a very natural and intuitive sense of the unit.
In that sense, a foot is a pretty nice unit, because human scale is around 5-10 of them. A decimeter is also very nice, and would make a fine unit except that it is so rarely used that using it tends to hinder communication.
Likewise, a gram is not generally useful. A kilogram is great in every resepect except that having a base unit with a prefix is stupid.
But, personally, I'd rather spend my time pushing for calendar reform and a 10 "hour" day. SI is good enough.
First time I have heard the expression "American Units" - can any US readers clarify what you generally call this measuring system (inches, miles, pounds, etc) over there?
At least within US West coast physics departments, we usually refer to them as Imperial units and SI/MKS/CGS units in the classroom.
In the lab we call them English or metric. (As in the phrase, "Damn it, we forgot that the new pump flange doesn't use metric screws. Who wants to head back down the mountain to find an English hex driver set?")
I often take part in political protests, and have on occasion been arrested and held for days.
So, I put together a quick routine using perl and chron that dispatches email to my workplace, the local legal rep contact, and some friends. The later includes directions to a hidden key and asks them to feed my cat until they hear from me. I only enable the system when I'm expecting a significant risk of arrest. Once it's started, if I don't either log into the machine or send myself an email containing a specific string once every 24 hours, the alarm goes off.
Turns out it's never actually been used (except when testing.) I did get caught up in a surprise arrest not too long ago, but since my girlfriend was going to be at home and able to take care of any problems I didn't turn on the system.
But, if you ask me, trusting life-changing information to a php script is a really, really scary idea. Even my trivial "please feed my cat" letters included disclaimers explaining that they may have been falsely triggered.
Now, on the other hand, the possibility of spoofing dead man's letters from other people *does* sound promising.
Not to mention: What happens if there's a software or hardware goof? Even leaving aside malicious attacks and the possibility of bugs in your code, who'd want to trust life-changing information to the system clock on a single machine?
You could imagine hardening a system against some of the more obvious dangers. Using two severs in different countries which confer with each other before sending anything and which both contain part of the encryption key for your data would go a long way toward catching the obvious technical dangers. If you can keep them from knowing each other's address until they both trigger, all the better (eg, by having them both call out to a third server that has no record of their locations until contacted.)
But, there's still no way to distinguish between "I'm dead," "I'm in jail," and "I'm in a coma from which I'm expected to recover." Which could be rather an important distinction.
Yup. And nose around your local Ham swap meets and the back pages of a few ham radio periodicals while you're at it. (Might as well become a Ham too - never hurts to have yet another communication channel open in an emergency.)
Our radio club has run a 100% solar setup for the last five or six field days. Nothing excessive - just a couple hundred watts for around 48 hours. Last time our setup consisted of 4 2x6 foot panels, a commercial regulator / battery charger, three or four big marine lead-acid batteries, and a cheapo consumer inverter for running 110V equipment. I think the guys spent a few hundred bucks on the used solar panels and another hundred or two on the regulator.
That said, if you have to choose between a generator and a solar/battery optionfor a real short-term disaster, it seems to me you'd be crazy to go with solar. Solar's a great replacement for grid power or for stationary, long-term, remote installations; however, having to truck a solar kit *into* a disaster zone doesn't sound like fun.
Batteries are a pain in the ass to maintain, and they're a bigger pain in the ass to transport. (You can't just throw them in the basement for five years and expect them to work - you've got to keep them charged and test them periodically.) Solar panels are fragile and awkward to transport. And, since hurricanes were mentioned, portable solar panel installations become a lot more complicated if hey have to withstand strong winds.
I'd much rather hike in through damaged roads carrying a 400W generator and a gas can than an equivalent solar kit. What's more, the generator fuel will do double duty as backup fuel for a cook stove or a vehicle. And if things really get desperate, it's both more valuable and more readily obtainable if you need to borrow or barter with someone.
Uncharged batteries and solar panels may take a while to come on-line, but they're proven reliable and would provide a good long-term solution, (probably the most effective for the 50 watt load the poster mentions.)
Unless you've been very careful (and spent a great deal of money) when choosing your batteries, you'd be wise not to depend on batteries that have been stored discharged.
For ordinary wet lead acid or "gel cell" batteries, you ought to at the very least store them fully charged and give them a test and recharge every six months. Better is to store them hooked up to a smart charger at all times. (Preferably on the roof in a shed somewhere - charging batteries indoors can be dangerous.)
I suppose the only problem is that if you were to exceed the drain's capacity you'd end up with oil on the floor.
But, then again, the oil would only be a thin layer on top of all the water on the floor. Chances are if your server room looks like that, the oil will be the least of your problems.
Re:why are the only interesting materials only flu
on
5 Strangest Materials
·
· Score: 1
>Dry ice is not a fluid.
It's also a weirdly specific.
Am I the only person who thinks something like "cryogens" would be a more natural addition to a list whose other members are dilatants, auxetic materials, superfluids, and ferrofluids?
Sure, dry ice may be the cheapest and most readily available household cryogen, and that it sublimates is kind of cool, but every "neat" thing mentioned could be done with any cryogenic liquid. (Well, except the mosquito thing, which doesn't really have much to do with the properties of the dry ice at all.) What's more, there are plenty of neat things that you can do with colder materials that you can't do with dry ice.
>What sucked me in, perhaps not you, and >got me to watch was the start where they >show some physicist trying to do garage >science and capturing the feel of it so >perfectly.
As a physics grad student with a dumpster-diving habit, I've got to side with the parent poster. Sure, Primer got a lot of the details right; however, they never managed to use those details to construct anything remotely interesting.
Of all the questions one can ask about time travel, "what happens when two greedy, narrow minded little shits discover it and use it for trivial, boring pursuits?" isn't one that immediately jumps to my mind.
I'm usually happy when films are willing to take on unsympathetic characters and plots grounded in boring, disappointing realism. But it's pretty hard to stomach a film entirely about people one cannot like, engaged in activities with no large scale consequence or philosophical import, speaking to each other in realistically banal language for two hours.
It's become my standard example to demonstrate that intelligent, knowledgeable people can take a great premise and a competent film crew and still manage to produce a movie which is unwatchable. (It is by no means the worst film I've seen - but then I've spent rather more time at no-budget experimental film festivals than most.)
The disappointing thing is that it *could* have been great. Every discussion I've had with people about the movie has been far more interesting than the movie itself. Taking such a great framework and managing to construct something so completely unengaging must have required considerable effort.
True enough. But, the parent's second point is still valid. Software that encodes your audio (presumably into some highly compressed low bitrate format) and then sends the data out only during a call would escape detection.
While I don't know for a fact that modern cell phones could be made to do this with only software changes, it doesn't seem too unlikely. (Whether those software changes could be initiated remotely without any user interventions is another question. One would sure *hope* not - but I wouldn't put it past cell phone makers to install a remote back door.)
>So how is this implemented without >storing the actual password, rather >than the salted hash?
Good call. I'm no crypto expert, but it sounds like a sure indication that something at best trivially related to the real password is stored somewhere.
Now, if the routine that changes the password asks for the current password before asking for the new password, and if you aren't worried about people making back-to-back changes in order to get around the restriction, the old password needn't ever be written to disk.
If you ask me, the real question is why there was valuable data floating around in clear text on the network to sniff?
Sure, these guys had to jump through some hoops to get into the building, and they could have been caught if employees had been more wary.
But, there are hundreds of people who have legitimate access to the building who could easily do the same thing. Everyone from the real photocopier technician to random bank employees has the oppertunity to do the same thing daily, with far less risk of discovery.
I'm a bit skeptical about devices actively sensing water - my guess would be that the water shorts out things that aren't meant to be shorted and the device shuts down either because some built in protection circuits get tripped or because individual ICs see a big drop in their supply voltages and behave strangely. But, the end result is the same.
Removing all power sources from the device and opening it up as much as possible as soon as possible is a good idea.
If you really want to go nuts, a thorough rinse in isopropanol (followed optionally by methanol) followed by a blow dry with compressed air or a heat gun on low does a nice job of removing any residue and cleaning things up. If you're dealing with circuit boards or connectors with incompletely removed rosin flux, you will end up with nasty white crap all over. If that bugs you, a good scrub with a toothbrush in isopropanol will clean it up.
I've never tried it, but I suspect medicine-cabinet-grade 87% iso rubbing alcohol would probably work fine.
>Yeah, because that's such an easily solvable >problem. Damn Apple for skimping and not using super >future battery technology. I mean come on, this >criticism is slightly baseless.
Ah, but it *IS* an easy problem to solve. All it takes is somewhere between two and four screws and an extra part number with which one can order replacement batteries.
Making the bits which will fail first easy to replace isn't a particularly subtle design goal. The people who manufactured my wristwatch, pda, cell phone, camera, portable cd player, dat machine, ham radio gear, and every cordless power tool and piece of battery powered test gear I've ever touched seem to have had no trouble with it. Most of them even include a replacement part number right on the battery case. In the case of my wristwatch (all puns intended), they managed to do it without even needing those extra scews, and the damn thing is even waterproof.
There's no excuse for designing a battery powered device without making it easy to replace the batteries. Unless, of course, you count, "we thought it would be profitable to fuck over our customers" as an excuse.
>Capitalism is about making things more effective. >When will people stop complaining about small shops >closing up, people being laid off, et cetera? >That's the idea!
True, at least in part.
Which is one of the reasons some of us think it's rather a bad idea.
I share your surprise that so many people complain about the necessary and inevitable result of a system while continuing to support it wholeheartedly at both the polls and the cash register. I just wish they'd stop the later rather than the former.
Slashdot Mirror
Looking for a programming job within a research group is certainly the approach which is safest and most likely to succeed. There are plenty of positions for programmers in medium sized physical science projects. Unfortunately, at least in my experience (US academic astrophysics), a rather large number of jobs are already filled long before they are advertised publicly. Trying to get your first science job may take a long time without any insider contacts. But, getting your *second* science job will be much easier, assuming you do excellent work.
I've known and taught a few people who started a physics undergrad degree from scratch in their middle thirties. It's certainly possible, but it won't be entirely easy. If you decide to take that route, you'll be marked as an outsider from the start in a highly collaborative field. Expect to be referred to as "you know, the older guy" by your peers, and asked completely inappropriate questions by lecturers at the start of every semester. Meanwhile, you'll be working hundred hour weeks, as often as not on material that you find irrelevant and unengaging. And, you'll probably be facing seven to twelve years of school before you can apply for permanent positions with any hope of success. (The first for a MS and a high level tech job, the second for a PhD and a chance to do truly independent research. A BS and a lot of prior experience *can* land you an interesting job with some creative freedom, but its far from guranteed.)
Those who pull it off tend to be outgoing people with a great deal of personal charisma and an overwhelming passion for the field. It's not a decision to be taken lightly.
On the other hand, the field and the world in general would be better off if more people joined late in life. Anything that brings someone other than 20 year old white boys into the department is a good thing, if you ask me. But, the personal cost shouldn't be discounted.
If you ask me, the single most important thing about a calculator is its user interface. In particular, button layout and the ease with which you can access commonly used functions, particularly with one hand while you read them off a blackboard/textbook/sheet of paper.
I still pull out my battered, ten year old TI-85 when I want to plug a bunch of physical constants into an equation, even when I'm sitting in front of a computer with orders of magnitude more processing power and infinite configurability.
As a mechanism for inputing data into a touch screen device, a calculator interface may be the best option. But as a replacement for a calculator with real mechanical buttons, a touchscreen running an emulator leaves out the only feature that makes a calculator actually useful in a world where everyone has a computer on thier desktop. It also leaves you with a device that's much easier to break, more likely to get stolen, and has dramatically reduced battery life.
I haven't looked at calculator in a decade or so, but the older TIs are built like bricks. You can drop them down a flight of stairs, and they'll run for months of weekly 40 hour problems sets on a set of AA batteries.
I quite like my pda, but a calculator it ain't.
Intersting. Somehow I had missed that point. Thanks.
That's true. But, does building a healthy place create more healthy people, or does it just pull healthy people from elsewhere? (I note and appreciate that you were careful not to mention a causal relationship.)
Boulder also has some advantages: a wealthy population (median family income of 70K $US), internationally renown outdoor recreation within minutes of downtown, a large university population, and a reputation for progressive culture and politics that tends to pull such people from the surrounding states.
All those characteristics seem likely to give you both a healthy population and a city council interested in maintaining a walkable city. If you were to forcibly swap the populations with a boulder-sized Atlanta suburb, it's not at all clear what would happen. (Except that after a few weeks it would be possible to find edible food in Georgia, which might be nice for those poor souls who occasionally find themselves trapped there.)
That's not to say that making other cities walkable wouldn't in itself be a good thing, even if it didn't do anything to improve the health of most of their residents.
Yup. I've always been amazed by this response.
Although I've never felt the slightest urge to have children, I'm told by reliable people that it is a genuine sensation that many experience. Given that, I can sympathize with those who decide to have children even though I'd argue that they probably shouldn't, given their dedication to other things which aren't easily compatible with family life. (I do sometimes wonder how much better off we'd be if fewer great scientists and artists gave up on their careers in order to become mediocre parents and later resentful, bitter retirees.)
But the people who try to talk *others* into having kids astounds me. Even leaving off the global ethical issues and the welfare of the children, it seems insane to urge someone to commit to 20 years of *anything* about which they're not already overwhelmingly enthusiastic.
A good starting place is to nose around the sites hosted at http://portal.linuxaudio.org/
The linux audio users and linux audio developers lists are vibrant (perhaps overwhelming) and their archives and associated documents and HOWTOs contain more information than you could possibly want.
Personally, I've had very good experiences with:
ecasound (multitrack recording, processing, general all-around fiddling)
ardour (recording and mixing)
rosegarder (midi sequencing and scoring)
JACK (patchbay and tool interfacing)
Perhaps, but a more relevant case is that some poor schmoe with the same knowledge and tools as you but with more time to devote to the analysis of a particular data set is able to do your work a little bit faster than you.
In the field where I work (cosmology with ground based CMB telescopes), a team of twenty people can easily work full time for years building an instrument. Tends of man-years are spent installing and calibrating it, figuring out where to point it, and writing software to capture and preprocess the data. Then someone has to manually set up and run the instrument while it's taking data.
At the end of the day, the "unreduced" data that comes streaming out of the back side of our instrument is the result of years of work. If we put it up live immediately, someone could easily take it and publish a map before we could, especially if they knew or could guess something about the form of the data, which wouldn't be too hard, given the proposals and conference talks floating around. What's more, our competitor would have the benefit of months of leisure during which to try out data analysis algorithms before the data appears, while we're all working 100 hour weeks just assembling and testing the thing.
Now, you can certainly bring a few dedicated analysis guys into the collaboration early on in an effort to get a full data analysis pipeline up and running as soon as possible. They're likely to have a slight advantage over competitors, given advanced knowledge of the instrument and form of the raw data. But, artificially broadening a collaboration solely in order to stand a fighting chance at competing with outsiders has its own drawbacks. Handing off all of the analysis to someone else isn't particularly more attractive just because the other guy agrees to put your name on the paper.
Under those circumstances, it seems entirely reasonable to me to demand that data be made public eventually, say when the first results paper based upon it is approved for publication, or after a fixed window of time. But, requiring immediate release provides little benefit could do a great deal of harm.
When it comes to a facilities instrument, like the Hubble, a more demanding release schedule may be appropriate. But, even when data formats and instrument parameters are fixed and researchers don't need to build instruments, it still takes a nontrivial amount of time and effort to choose an observing strategy, submit proposals, arrange for followups, etc. And, as in the parent poster's example, there are times when the rush to publish first could easily reward people who are willing to run with a barely-adequate measurement instead of those who do things right. (That happens anyway in some cases, but unrestricted immediate data release could only hurt.)
It may not be *difficult* in the absolute sense, but it's still a lot more difficult than *not* having to do it. There are a whole lot of us who have to do it often enough to be annoyed, but not quite often enough to memorize all possible combinations so that we can do it without thinking.
And when you ask, "is a 3/8 gap a loose fit to 7/62+7/32+1/16?" it becomes a genuine annoyance. I spend my days with physics grad students and expert machinists, and have only met one or two who wouldn't want check their answer with paper or calculator if they were working on something important.
By comparison, comparing
Any modern human who isn't on crack would probably agree that choosing base-10 relationships is a good idea. We've got to pick something, and picking the same system for everything except where hardware limitations make it impractical is a good idea. But, it is interesting to ask how large a base unit ought to be.
It seems to me one ought to pick them so that human scale objects are usually of order 10 of them. That way you can generally refer to the size of something with within 10% with a single, small whole number, and have a very natural and intuitive sense of the unit.
In that sense, a foot is a pretty nice unit, because human scale is around 5-10 of them. A decimeter is also very nice, and would make a fine unit except that it is so rarely used that using it tends to hinder communication.
Likewise, a gram is not generally useful. A kilogram is great in every resepect except that having a base unit with a prefix is stupid.
But, personally, I'd rather spend my time pushing for calendar reform and a 10 "hour" day. SI is good enough.
Taken an astronomy or E&M class lately?
I'm usually a big fan of standards, doing E&M in Heaviside-Lorentz units sure is nice.
At least within US West coast physics departments, we usually refer to them as Imperial units and SI/MKS/CGS units in the classroom.
In the lab we call them English or metric. (As in the phrase, "Damn it, we forgot that the new pump flange doesn't use metric screws. Who wants to head back down the mountain to find an English hex driver set?")
I often take part in political protests, and have on occasion been arrested and held for days.
So, I put together a quick routine using perl and chron that dispatches email to my workplace, the local legal rep contact, and some friends. The later includes directions to a hidden key and asks them to feed my cat until they hear from me. I only enable the system when I'm expecting a significant risk of arrest. Once it's started, if I don't either log into the machine or send myself an email containing a specific string once every 24 hours, the alarm goes off.
Turns out it's never actually been used (except when testing.) I did get caught up in a surprise arrest not too long ago, but since my girlfriend was going to be at home and able to take care of any problems I didn't turn on the system.
But, if you ask me, trusting life-changing information to a php script is a really, really scary idea. Even my trivial "please feed my cat" letters included disclaimers explaining that they may have been falsely triggered.
Now, on the other hand, the possibility of spoofing dead man's letters from other people *does* sound promising.
Not to mention: What happens if there's a software or hardware goof? Even leaving aside malicious attacks and the possibility of bugs in your code, who'd want to trust life-changing information to the system clock on a single machine?
You could imagine hardening a system against some of the more obvious dangers. Using two severs in different countries which confer with each other before sending anything and which both contain part of the encryption key for your data would go a long way toward catching the obvious technical dangers. If you can keep them from knowing each other's address until they both trigger, all the better (eg, by having them both call out to a third server that has no record of their locations until contacted.)
But, there's still no way to distinguish between "I'm dead," "I'm in jail," and "I'm in a coma from which I'm expected to recover." Which could be rather an important distinction.
Yup. And nose around your local Ham swap meets and the back pages of a few ham radio periodicals while you're at it. (Might as well become a Ham too - never hurts to have yet another communication channel open in an emergency.)
Our radio club has run a 100% solar setup for the last five or six field days. Nothing excessive - just a couple hundred watts for around 48 hours. Last time our setup consisted of 4 2x6 foot panels, a commercial regulator / battery charger, three or four big marine lead-acid batteries, and a cheapo consumer inverter for running 110V equipment. I think the guys spent a few hundred bucks on the used solar panels and another hundred or two on the regulator.
That said, if you have to choose between a generator and a solar/battery optionfor a real short-term disaster, it seems to me you'd be crazy to go with solar. Solar's a great replacement for grid power or for stationary, long-term, remote installations; however, having to truck a solar kit *into* a disaster zone doesn't sound like fun.
Batteries are a pain in the ass to maintain, and they're a bigger pain in the ass to transport. (You can't just throw them in the basement for five years and expect them to work - you've got to keep them charged and test them periodically.) Solar panels are fragile and awkward to transport. And, since hurricanes were mentioned, portable solar panel installations become a lot more complicated if hey have to withstand strong winds.
I'd much rather hike in through damaged roads carrying a 400W generator and a gas can than an equivalent solar kit. What's more, the generator fuel will do double duty as backup fuel for a cook stove or a vehicle. And if things really get desperate, it's both more valuable and more readily obtainable if you need to borrow or barter with someone.
Unless you've been very careful (and spent a great deal of money) when choosing your batteries, you'd be wise not to depend on batteries that have been stored discharged.
For ordinary wet lead acid or "gel cell" batteries, you ought to at the very least store them fully charged and give them a test and recharge every six months. Better is to store them hooked up to a smart charger at all times. (Preferably on the roof in a shed somewhere - charging batteries indoors can be dangerous.)
I suppose the only problem is that if you were to exceed the drain's capacity you'd end up with oil on the floor.
But, then again, the oil would only be a thin layer on top of all the water on the floor. Chances are if your server room looks like that, the oil will be the least of your problems.
>Dry ice is not a fluid.
It's also a weirdly specific.
Am I the only person who thinks something like "cryogens" would be a more natural addition to a list whose other members are dilatants, auxetic materials, superfluids, and ferrofluids?
Sure, dry ice may be the cheapest and most readily available household cryogen, and that it sublimates is kind of cool, but every "neat" thing mentioned could be done with any cryogenic liquid. (Well, except the mosquito thing, which doesn't really have much to do with the properties of the dry ice at all.) What's more, there are plenty of neat things that you can do with colder materials that you can't do with dry ice.
>What sucked me in, perhaps not you, and
>got me to watch was the start where they
>show some physicist trying to do garage
>science and capturing the feel of it so
>perfectly.
As a physics grad student with a dumpster-diving habit, I've got to side with the parent poster. Sure, Primer got a lot of the details right; however, they never managed to use those details to construct anything remotely interesting.
Of all the questions one can ask about time travel, "what happens when two greedy, narrow minded little shits discover it and use it for trivial, boring pursuits?" isn't one that immediately jumps to my mind.
I'm usually happy when films are willing to take on unsympathetic characters and plots grounded in boring, disappointing realism. But it's pretty hard to stomach a film entirely about people one cannot like, engaged in activities with no large scale consequence or philosophical import, speaking to each other in realistically banal language for two hours.
It's become my standard example to demonstrate that intelligent, knowledgeable people can take a great premise and a competent film crew and still manage to produce a movie which is unwatchable. (It is by no means the worst film I've seen - but then I've spent rather more time at no-budget experimental film festivals than most.)
The disappointing thing is that it *could* have been great. Every discussion I've had with people about the movie has been far more interesting than the movie itself. Taking such a great framework and managing to construct something so completely unengaging must have required considerable effort.
True enough. But, the parent's second point is still valid. Software that encodes your audio (presumably into some highly compressed low bitrate format) and then sends the data out only during a call would escape detection.
While I don't know for a fact that modern cell phones could be made to do this with only software changes, it doesn't seem too unlikely. (Whether those software changes could be initiated remotely without any user interventions is another question. One would sure *hope* not - but I wouldn't put it past cell phone makers to install a remote back door.)
Or, it's a very *very* long passphrase.
>So how is this implemented without
>storing the actual password, rather
>than the salted hash?
Good call. I'm no crypto expert, but it sounds like a sure indication that something at best trivially related to the real password is stored somewhere.
Now, if the routine that changes the password asks for the current password before asking for the new password, and if you aren't worried about people making back-to-back changes in order to get around the restriction, the old password needn't ever be written to disk.
If you ask me, the real question is why there was valuable data floating around in clear text on the network to sniff?
Sure, these guys had to jump through some hoops to get into the building, and they could have been caught if employees had been more wary.
But, there are hundreds of people who have legitimate access to the building who could easily do the same thing. Everyone from the real photocopier technician to random bank employees has the oppertunity to do the same thing daily, with far less risk of discovery.
Yup.
I'm a bit skeptical about devices actively sensing water - my guess would be that the water shorts out things that aren't meant to be shorted and the device shuts down either because some built in protection circuits get tripped or because individual ICs see a big drop in their supply voltages and behave strangely. But, the end result is the same.
Removing all power sources from the device and opening it up as much as possible as soon as possible is a good idea.
If you really want to go nuts, a thorough rinse in isopropanol (followed optionally by methanol) followed by a blow dry with compressed air or a heat gun on low does a nice job of removing any residue and cleaning things up. If you're dealing with circuit boards or connectors with incompletely removed rosin flux, you will end up with nasty white crap all over. If that bugs you, a good scrub with a toothbrush in isopropanol will clean it up.
I've never tried it, but I suspect
medicine-cabinet-grade 87% iso rubbing alcohol would probably work fine.
>Yeah, because that's such an easily solvable
>problem. Damn Apple for skimping and not using super
>future battery technology. I mean come on, this
>criticism is slightly baseless.
Ah, but it *IS* an easy problem to solve. All it takes is somewhere between two and four screws and an extra part number with which one can order replacement batteries.
Making the bits which will fail first easy to replace isn't a particularly subtle design goal. The people who manufactured my wristwatch, pda, cell phone, camera, portable cd player, dat machine, ham radio gear, and every cordless power tool and piece of battery powered test gear I've ever touched seem to have had no trouble with it. Most of them even include a replacement part number right on the battery case. In the case of my wristwatch (all puns intended), they managed to do it without even needing those extra scews, and the damn thing is even waterproof.
There's no excuse for designing a battery powered device without making it easy to replace the batteries. Unless, of course, you count, "we thought it would be profitable to fuck over our customers" as an excuse.
>Capitalism is about making things more effective.
>When will people stop complaining about small shops
>closing up, people being laid off, et cetera?
>That's the idea!
True, at least in part.
Which is one of the reasons some of us think it's rather a bad idea.
I share your surprise that so many people complain about the necessary and inevitable result of a system while continuing to support it wholeheartedly at both the polls and the cash register. I just wish they'd stop the later rather than the former.