In other words, as a system administrator, you're our natural enemy.:)
I joke, I joke. Sorta.
But seriously, you're the guys who write the stuff that we then have to make work. You think your job ends when the product is shipped... but it actually only begins for us then.
"I'm an IT Professional. Do I know anything about fixing windoze? HELL NO! Do people assume I know about PCs and wireless routers and (sheeze!) printers? "
So... you write code which will be accessed in an environment that you admit you don't understand, and you're *proud* of this?
As an IT person, whose job it is to take programs written by developers and then figure out how to shoehorn them into our system, I often find myself scratching my head and wondering "what were the developers thinking with this? Do they care how hard it is to install their software and make it work with a real-world environment? Have they ever actually *used* a real-world networked environment and realise how complicated it is? Or do they just live in a little developer bubble where they only ever use their own personal machines and don't have to wrestle with deployment and management policies?"
Your attitude makes me think that perhaps there's some truth to this feeling.
"The thing about end users is, they generally don't know what they want."
I don't think so. The definition of end users is that they know what they *want* - they just don't necessarily know how to *achieve* it given the tools at hand. They are neither dumb nor confused nor a lesser species. They simply have different goals and domain expertise than you. They don't understand the foundations of the computer system because it's not their job to - their job is to get stuff done, and the computer either assists or obstructs this. If it obstructs, it is an error which needs to be fixed.
Also, EVERYONE is an end user to some process, and the provider for another.
"Just because you don't know how to use non garbage collected memory does not mean that functions that don't use it should be phased out."
Actually, yes, it does.
The issue is not "some programmers don't *know* how to use unsafe potentially root-hole-exposing misfeatures", but "can we mathematically PROVE with 100% certainty that in EVERY case where unsafe APIs are used, there is NO possibility of error, in any way, by any person, at any time, in any possible configuration, for even the slightest tiny exposure."
Because If you CAN'T prove with 100% certainty that there is no possible error, then it's 100% certain that my system running your rockstar C++ kernel god code IS rootable, and WILL be pwned. That's where all my monthly WSUS and apt-get patches come from: people like you who think they're smart enough to use unsafe APIs correctly, but whoops, turns out they aren't.
Yes, that's why we have botnets. Because of you! Congratulations.
I don't want my system to be pwned by your bad code. The simplest 100% certain way of me not getting pwned is for me to deprecate the unsafe APIs that you used to write the code that gets me pwned.
"I need those functions to do my job since I write network based apps that require memory reuse and pointers to be able to process data as quickly as possible"
No, you do NOT need to be able to "process data as quickly as possible". A fast root hole is just a faster way to lose. You need to make your code 100% secure FIRST, then make it as fast as you can with what's left over.
Don't like managed code? Too bad. I don't like being a zombie. Deprecate those APIs now, kthx.
"As such these guys virtually never see a piece of data all the way thru the computer. Their customers are other pieces of software."
More and more, that's going to be the normal experience for all programming. The programmer who asks for, and thinks his software is so important that it requires, bare-metal machine access (or even that it needs to be "installed" on a particular machine) is going to get strange looks. Orchestrating components and data and playing nicely with others will be job one.
And that's the way it should be, if we're going to make any progress. How many industrial shops refuse to use any prefabricated components or deal with any suppliers and insist on smelting their own ore and growing their own hemp for rope? Yeah, we kinda grew past that point around the 18th century. But in software, it's still considered a badge of manliness to code all the way down to the bare metal and you're some kind of wimp if you don't.
"Somehow in the last 30 years we lost the ability to undertake large infrastructure, which you would think given the wealth, technology, etc... that it would be easier."
I think Vietnam was the tipping point. A huge military expenditure forced the curtailment of domestic infrastructure subsidies. Which led to industrial stagnation, and those of us who were kids in the 70s remember it being a bit grim: strikes, inflation, gas price rises, quality problems, and so on.
Then came Reagan like the white knight, and his "solution" for Morning In America was to deregulate, which let private infrastructure companies morph into Enron-like shell games. Finance became the new "sunrise industry", alongside microcomputing and networking - the focus was on production of information rather than investment in the old crumbling infrastructure. It was easier and cheaper to make profits by repackaging ownership and debt than doing the hard work. Image, not substance, was what the free market rewarded, so that's what we got.
If you look at early 80s science fiction, like the cyberpunks, you see a lot of sunny optimism, even mixed in with terror, of how efficient private companies were going to be at building infrastructure. But that didn't happen except in computing, and I'm kind of surprised as to why even that occurred - I presume the Pentagon and Wall Street were the main drivers there.
Clinton slowed back a bit but kept mostly on the same privatisation track, and W accelerated it again. Now Obama's trying to reinvest in social infrastructure (healthcare) and gets called the worst of names for that. Far from Kennedy's space race era, half of the USA now sees the mere idea of national-level investment in anything but war as inherently evil. As an outsider, I don't understand why, but I can see the effects.
Space, for instance, was really all just about the ICBM buildout. Once the Minutemen were built, and the military got their spy and comms networks, and computers had shown that a manned space presence wasn't necessary to achieve the military objectives... there wasn't a whole lot left to do. Just more commsats.
Infrastructure is a hard problem to start with. When there's a political movement which actively believes even having a shared infrastructure to be a bad thing and that it's a moral duty to prevent those who don't have their own capital reserves from getting access to services... it gets a lot harder.
"Once again, the crowd that wants us to cut back our carbon emissions comes up with things we can't do rather than some suggestions."
Yes, because of course a group that's worried about the effect of human technologies on the environment would just love to trade one hazard for another. No, it must be that those greenies are hypocritical and just want to stop us having fun.
Who are they to tell us what is or isn't healthy for ecosystem. Pssht. What's this ecosystem think it is, anyway? Some kind of big wuss? Radiation's good for it.
"Today our understanding is much greater, and we have very advanced computer models to help the design process. Ever wondered why modern bridges and buildings are much more 'delicate' than older behemoths? Because we can compute the actual behavior of the structures to much higher precision and accuracy, so the needed safety margin is less."
Which sounds remarkably like the logic which led to the banking collapse. Increasing modelling sophistication leading to increased "returns on investment" by stripping away the "inefficient" safety margins.
Fortunately, the world is rapidly becoming a safer, more predictable place, in which black swan events never occur. All in all, this sounds like a sound way to run an industry and I heartily approve!
"Claiming that citation equals proof is an implicit appeal to authority and therefore fallacious."
That would be relevant if Wikipedia were claiming that citation equals proof. But they don't. Citation means *verifiability*, which is the exact opposite of appeal to authority. It means "don't just take my word for it, read the source material".
Do you *want* the Wikipedia to be a place where people can make random unjustified claims and say "trust me, I'm a subject matter expert even though I can't show you why"?
"His citation was to a fanciful coffee table reference book published before the system in question was declassified, and which was widely cited elsewhere on the web. My citation was to a professional academic analysis written a decade after the system was declassified, but which existed only in a few thousand hard copies."
This is a problem caused by intellectual property, basically. You say a resource exists which verifies your edit - but if you can't share that with the class, then it might as well not exist. It really does become your word against someone else's, and sorry, but just you *saying* that you've worked on the system in question really isn't enough.
The sensible answer would be for you to scan the pages in question from the book in question (or preferably the whole book), and put it up in a stable, permanent place on the Web semantically linked to the ISBN of the book, and then reference that from Wikipedia.
But of course we can't do that because it's nasty and illegal and copyright-violating.
Don't blame Wikipedia. Blame copyright. That's what's stopping you from being able to verify your claimed knowledge.
"it's contributors, being people with few or no credentials, often post POV articles and then use very biased sources such as blogs, etc."
Wait - so the GP is claiming that Wikipedia sucks because random nobodies CAN'T post articles about random POV cruft with no citations -- and you're *agreeing* by saying that Wikipedia sucks because random nobodies CAN post articles about random POV cruft with no citations?
You can't both be right. Seems to me like Wikipedia is walking a delicate balance between your two opinions and doing fine.
"was always going to primarily attract the type of person who is not interesting in providing knowledge for all, but only those for whom its articles are personal prestige projects, intended to impress only themselves and their imagined audience."
Sounds like it's doing a great job of emulating the hallowed halls of academia, then...:)
"And you have absolutely no idea what a huge PITA an FOI can be for a government agency. It is not a simple process, and it is not a short process, especially when there are pre-existing contractual obligations on that data. If you are a government agency contracting with a commercial entity, you cannot simply "turn over the data". It doesn't work that way."
Then that's a huge problem, right there, with the practice of science. Data for public purposes needs to be made at least as open as Wikimedia Commons.
It's simply not acceptable to base global policy decisions on datasets with intellectual property restrictions. Nor is it acceptable to say 'we've got too much data' unless you're the LHC. Storage space is cheap now. We have terabyte removable drives for $100.
Sorry. That argument might have cut it in the 1950s. It doesn't anymore. Free the data. Open the science. Open the governance. Or don't expect the world to buy into policy decisions made on that data.
"Like Cardinal Richelieu said: “If you give me six lines written by the hand of the most honest of men, I will find something in them which will hang him”"
If those six lines were code, Cardinal Richelieu was apparently a DEFCON presenter.:)
And if you look at, say, the Linux kernel mailing lists you can see exactly this sort of honest, robust discussion. Strangely nobody has been arrested yet because of their LKML postings.
This is why transparency and openness in process right from the beginning is a Good Thing - and if we're talking about science affecting global policy, heck yeah it needed to have been open right from the start.
Fudging and pretending and then presenting an airbrushed data 'product' at the end - and then basing extremely controversial legislation on top of that product - just doesn't cut it. Any more than having voting machines based on secret code.
Open the science, all the way. This is the Facebook generation. We're getting used to being honest as a society. It's time publically-funded science caught up with the Linux kernel. (I can't believe I'm even reading the words "intellectual property" next to "climate science").
"Who are these "other people"? If they are people with an adequate background and understanding of the data and the science behind it, then you would have a point. If you mean the general public or someone with an agenda, then no."
That makes sense... in exactly the same way as that the Linux kernel, being a very specialised and technical piece of engineering only correctly understandable by experts, is safely held under lock and key to make sure the general public and people with agendas can't see it, because it might melt their little brains.
In other words, no sense at all.
I'm generally of the opinion that anthropomorphic global warming is real, and that the only people disputing (though I've been a bit bemused as to why there's a sudden trendy focus on CO2 rather than the more obvious environmental threats like deforestation and fish stock depletion) - but this level of obfuscation is really shaking my faith in the scientific community.
The scientific method is based on open and honest sharing of data. Most especially so when the data is heavily massaged with computer simulations, and describes a very complex system that we don't come close to understanding in detail.
It's not the emails that bug me - science is a robust disagreement and scientists air their views in less than diplomatic ways, okay - but it's the the fact that 1) the data wasn't available right from the outset, and b) CRU's immediate reaction to this leak as if they were a corporation whose trade secrets had been stolen.
This is public science, not military or industry, so there's no obligation for secrecy - but more to the point, it's science whose interpretation and outcome affects the whole planet, so there's actually a positive requirement for transparency. For these reasons, ALL the steps in the data manipulation pipeline need to be utterly transparent so that the whole process is completely above reproach.
George Monbiot doesn't like this and neither do I. Planetary-level science needs to quit this weird shroud of secrecy. Let the data be seen in all its warts. Open source the sucker.
If you read Popular Science and Popular Mechanics magazines from the 1950s (online now at Google Books btw - wonderful nostalgia trip!) you can see how gung-ho some sectors of the country were about Science. (And military - it's pretty scary now reading all about fairly hairy atomic and medical tests which would never pass ethics committees today - one hopes).
This is why I love UFO: Enemy Unknown. Half the game is fighting the alien menace - the other, harder, part is fighting those shortsighted fools from the funding committee who don't understand why you NEED a third containment facility in your Antarctic base.
Slashdot Mirror
"I'm not an "IT guy". I'm a software developer."
In other words, as a system administrator, you're our natural enemy. :)
I joke, I joke. Sorta.
But seriously, you're the guys who write the stuff that we then have to make work. You think your job ends when the product is shipped... but it actually only begins for us then.
Then comes the running, and the screaming.
"I'm an IT Professional. Do I know anything about fixing windoze? HELL NO! Do people assume I know about PCs and wireless routers and (sheeze!) printers? "
So... you write code which will be accessed in an environment that you admit you don't understand, and you're *proud* of this?
As an IT person, whose job it is to take programs written by developers and then figure out how to shoehorn them into our system, I often find myself scratching my head and wondering "what were the developers thinking with this? Do they care how hard it is to install their software and make it work with a real-world environment? Have they ever actually *used* a real-world networked environment and realise how complicated it is? Or do they just live in a little developer bubble where they only ever use their own personal machines and don't have to wrestle with deployment and management policies?"
Your attitude makes me think that perhaps there's some truth to this feeling.
"The thing about end users is, they generally don't know what they want."
I don't think so. The definition of end users is that they know what they *want* - they just don't necessarily know how to *achieve* it given the tools at hand. They are neither dumb nor confused nor a lesser species. They simply have different goals and domain expertise than you. They don't understand the foundations of the computer system because it's not their job to - their job is to get stuff done, and the computer either assists or obstructs this. If it obstructs, it is an error which needs to be fixed.
Also, EVERYONE is an end user to some process, and the provider for another.
"You're basically saying A > A since this "anyone" includes "you" too. ;)"
So.. he hates Microsoft so much he does an infinite recursion of Cantor diagonalisation every time he thinks of them?
That's a lot of hate!
"Just because you don't know how to use non garbage collected memory does not mean that functions that don't use it should be phased out."
Actually, yes, it does.
The issue is not "some programmers don't *know* how to use unsafe potentially root-hole-exposing misfeatures", but "can we mathematically PROVE with 100% certainty that in EVERY case where unsafe APIs are used, there is NO possibility of error, in any way, by any person, at any time, in any possible configuration, for even the slightest tiny exposure."
Because If you CAN'T prove with 100% certainty that there is no possible error, then it's 100% certain that my system running your rockstar C++ kernel god code IS rootable, and WILL be pwned. That's where all my monthly WSUS and apt-get patches come from: people like you who think they're smart enough to use unsafe APIs correctly, but whoops, turns out they aren't.
Yes, that's why we have botnets. Because of you! Congratulations.
I don't want my system to be pwned by your bad code. The simplest 100% certain way of me not getting pwned is for me to deprecate the unsafe APIs that you used to write the code that gets me pwned.
"I need those functions to do my job since I write network based apps that require memory reuse and pointers to be able to process data as quickly as possible"
No, you do NOT need to be able to "process data as quickly as possible". A fast root hole is just a faster way to lose. You need to make your code 100% secure FIRST, then make it as fast as you can with what's left over.
Don't like managed code? Too bad. I don't like being a zombie. Deprecate those APIs now, kthx.
"As such these guys virtually never see a piece of data all the way thru the computer. Their customers are other pieces of software."
More and more, that's going to be the normal experience for all programming. The programmer who asks for, and thinks his software is so important that it requires, bare-metal machine access (or even that it needs to be "installed" on a particular machine) is going to get strange looks. Orchestrating components and data and playing nicely with others will be job one.
And that's the way it should be, if we're going to make any progress. How many industrial shops refuse to use any prefabricated components or deal with any suppliers and insist on smelting their own ore and growing their own hemp for rope? Yeah, we kinda grew past that point around the 18th century. But in software, it's still considered a badge of manliness to code all the way down to the bare metal and you're some kind of wimp if you don't.
It's time software grew up.
"Somehow in the last 30 years we lost the ability to undertake large infrastructure, which you would think given the wealth, technology, etc... that it would be easier."
I think Vietnam was the tipping point. A huge military expenditure forced the curtailment of domestic infrastructure subsidies. Which led to industrial stagnation, and those of us who were kids in the 70s remember it being a bit grim: strikes, inflation, gas price rises, quality problems, and so on.
Then came Reagan like the white knight, and his "solution" for Morning In America was to deregulate, which let private infrastructure companies morph into Enron-like shell games. Finance became the new "sunrise industry", alongside microcomputing and networking - the focus was on production of information rather than investment in the old crumbling infrastructure. It was easier and cheaper to make profits by repackaging ownership and debt than doing the hard work. Image, not substance, was what the free market rewarded, so that's what we got.
If you look at early 80s science fiction, like the cyberpunks, you see a lot of sunny optimism, even mixed in with terror, of how efficient private companies were going to be at building infrastructure. But that didn't happen except in computing, and I'm kind of surprised as to why even that occurred - I presume the Pentagon and Wall Street were the main drivers there.
Clinton slowed back a bit but kept mostly on the same privatisation track, and W accelerated it again. Now Obama's trying to reinvest in social infrastructure (healthcare) and gets called the worst of names for that. Far from Kennedy's space race era, half of the USA now sees the mere idea of national-level investment in anything but war as inherently evil. As an outsider, I don't understand why, but I can see the effects.
Space, for instance, was really all just about the ICBM buildout. Once the Minutemen were built, and the military got their spy and comms networks, and computers had shown that a manned space presence wasn't necessary to achieve the military objectives... there wasn't a whole lot left to do. Just more commsats.
Infrastructure is a hard problem to start with. When there's a political movement which actively believes even having a shared infrastructure to be a bad thing and that it's a moral duty to prevent those who don't have their own capital reserves from getting access to services... it gets a lot harder.
"Once again, the crowd that wants us to cut back our carbon emissions comes up with things we can't do rather than some suggestions."
Yes, because of course a group that's worried about the effect of human technologies on the environment would just love to trade one hazard for another. No, it must be that those greenies are hypocritical and just want to stop us having fun.
Who are they to tell us what is or isn't healthy for ecosystem. Pssht. What's this ecosystem think it is, anyway? Some kind of big wuss? Radiation's good for it.
"Today our understanding is much greater, and we have very advanced computer models to help the design process. Ever wondered why modern bridges and buildings are much more 'delicate' than older behemoths? Because we can compute the actual behavior of the structures to much higher precision and accuracy, so the needed safety margin is less."
Which sounds remarkably like the logic which led to the banking collapse. Increasing modelling sophistication leading to increased "returns on investment" by stripping away the "inefficient" safety margins.
Fortunately, the world is rapidly becoming a safer, more predictable place, in which black swan events never occur. All in all, this sounds like a sound way to run an industry and I heartily approve!
So we'll now have automatic 'Like', 'Dislike' and 'Eeeeeeagh my visual cortex where is the brain soap' responses?
"Claiming that citation equals proof is an implicit appeal to authority and therefore fallacious."
That would be relevant if Wikipedia were claiming that citation equals proof. But they don't. Citation means *verifiability*, which is the exact opposite of appeal to authority. It means "don't just take my word for it, read the source material".
Do you *want* the Wikipedia to be a place where people can make random unjustified claims and say "trust me, I'm a subject matter expert even though I can't show you why"?
"His citation was to a fanciful coffee table reference book published before the system in question was declassified, and which was widely cited elsewhere on the web. My citation was to a professional academic analysis written a decade after the system was declassified, but which existed only in a few thousand hard copies."
This is a problem caused by intellectual property, basically. You say a resource exists which verifies your edit - but if you can't share that with the class, then it might as well not exist. It really does become your word against someone else's, and sorry, but just you *saying* that you've worked on the system in question really isn't enough.
The sensible answer would be for you to scan the pages in question from the book in question (or preferably the whole book), and put it up in a stable, permanent place on the Web semantically linked to the ISBN of the book, and then reference that from Wikipedia.
But of course we can't do that because it's nasty and illegal and copyright-violating.
Don't blame Wikipedia. Blame copyright. That's what's stopping you from being able to verify your claimed knowledge.
"it's contributors, being people with few or no credentials, often post POV articles and then use very biased sources such as blogs, etc."
Wait - so the GP is claiming that Wikipedia sucks because random nobodies CAN'T post articles about random POV cruft with no citations -- and you're *agreeing* by saying that Wikipedia sucks because random nobodies CAN post articles about random POV cruft with no citations?
You can't both be right. Seems to me like Wikipedia is walking a delicate balance between your two opinions and doing fine.
"was always going to primarily attract the type of person who is not interesting in providing knowledge for all, but only those for whom its articles are personal prestige projects, intended to impress only themselves and their imagined audience."
Sounds like it's doing a great job of emulating the hallowed halls of academia, then... :)
"How in the world does one lauch a Word?"
Well, you poit your moue curor at the ion and cick the buton...
""Are 1 in 4 children really sexually abused by the Internet?"
Err NO"
Someone hasn't watched Demon Seed...
"And you have absolutely no idea what a huge PITA an FOI can be for a government agency. It is not a simple process, and it is not a short process, especially when there are pre-existing contractual obligations on that data. If you are a government agency contracting with a commercial entity, you cannot simply "turn over the data". It doesn't work that way."
Then that's a huge problem, right there, with the practice of science. Data for public purposes needs to be made at least as open as Wikimedia Commons.
It's simply not acceptable to base global policy decisions on datasets with intellectual property restrictions. Nor is it acceptable to say 'we've got too much data' unless you're the LHC. Storage space is cheap now. We have terabyte removable drives for $100.
Sorry. That argument might have cut it in the 1950s. It doesn't anymore. Free the data. Open the science. Open the governance. Or don't expect the world to buy into policy decisions made on that data.
Sure, but you need to be open about exactly what manipulations you made in order to validate your data-cleaning methodology, surely.
Obscuring that process is what's at issue here.
"Like Cardinal Richelieu said:
“If you give me six lines written by the hand of the most honest of men, I will find something in them which will hang him”"
If those six lines were code, Cardinal Richelieu was apparently a DEFCON presenter. :)
And if you look at, say, the Linux kernel mailing lists you can see exactly this sort of honest, robust discussion. Strangely nobody has been arrested yet because of their LKML postings.
This is why transparency and openness in process right from the beginning is a Good Thing - and if we're talking about science affecting global policy, heck yeah it needed to have been open right from the start.
Fudging and pretending and then presenting an airbrushed data 'product' at the end - and then basing extremely controversial legislation on top of that product - just doesn't cut it. Any more than having voting machines based on secret code.
Open the science, all the way. This is the Facebook generation. We're getting used to being honest as a society. It's time publically-funded science caught up with the Linux kernel. (I can't believe I'm even reading the words "intellectual property" next to "climate science").
"Who are these "other people"? If they are people with an adequate background and understanding of the data and the science behind it, then you would have a point. If you mean the general public or someone with an agenda, then no."
That makes sense... in exactly the same way as that the Linux kernel, being a very specialised and technical piece of engineering only correctly understandable by experts, is safely held under lock and key to make sure the general public and people with agendas can't see it, because it might melt their little brains.
In other words, no sense at all.
I'm generally of the opinion that anthropomorphic global warming is real, and that the only people disputing (though I've been a bit bemused as to why there's a sudden trendy focus on CO2 rather than the more obvious environmental threats like deforestation and fish stock depletion) - but this level of obfuscation is really shaking my faith in the scientific community.
The scientific method is based on open and honest sharing of data. Most especially so when the data is heavily massaged with computer simulations, and describes a very complex system that we don't come close to understanding in detail.
It's not the emails that bug me - science is a robust disagreement and scientists air their views in less than diplomatic ways, okay - but it's the the fact that 1) the data wasn't available right from the outset, and b) CRU's immediate reaction to this leak as if they were a corporation whose trade secrets had been stolen.
This is public science, not military or industry, so there's no obligation for secrecy - but more to the point, it's science whose interpretation and outcome affects the whole planet, so there's actually a positive requirement for transparency. For these reasons, ALL the steps in the data manipulation pipeline need to be utterly transparent so that the whole process is completely above reproach.
George Monbiot doesn't like this and neither do I. Planetary-level science needs to quit this weird shroud of secrecy. Let the data be seen in all its warts. Open source the sucker.
"we've decided to fast track certain predominantly Godless groups to eternal damnation. You're now stuck at work."
But with Internet access... so how is that Hell again?
"Which makes large, money-centric programs like this all about the people to whom the money goes: the unions."
Don't you mean "the workers"?
Last I checked union dues weren't 100%.
(Disclaimer: I work in tertiary education in New Zealand, as IT support, and I'm a proud union member. Education pays my wages.)
Followed rapidly by "I wonder how fast I can run 200 yards" and "Do tigers climb trees?"
If you read Popular Science and Popular Mechanics magazines from the 1950s (online now at Google Books btw - wonderful nostalgia trip!) you can see how gung-ho some sectors of the country were about Science. (And military - it's pretty scary now reading all about fairly hairy atomic and medical tests which would never pass ethics committees today - one hopes).
http://books.google.com/books?id=iSEDAAAAMBAJ&dq=popular+science+1950&rview=1&source=gbs_navlinks_s
http://books.google.com/books?id=K9kDAAAAMBAJ&dq=popular+mechanics+1950&rview=1&source=gbs_navlinks_s
There was a big war-era science boom.
Disconnect your power and stop delivering FedEx.
This is why I love UFO: Enemy Unknown. Half the game is fighting the alien menace - the other, harder, part is fighting those shortsighted fools from the funding committee who don't understand why you NEED a third containment facility in your Antarctic base.