Pu-238 is much too unstable to use in a thermonuclear device; too much of it will cause a premature partial detonation. However, it makes a dandy small scale energy source. Plutonium batteries were manufactured in the 1970s for devices such as pacemakers that needed a long term service-free power supply. Also, strontium and cesium based radiothermal generators were used in the USSR to power such things as remote lighthouses that would've been hideously inconvenient to supply fuel to in the winter.
As for U-235, I think one of the most inventive uses I've seen is powering a nuclear saltwater steam rocket engine for interplanetary use. Just watch where you point it, the exhaust is really nasty.
In the city of Ankh-Morpork the head of the Guild of Lawyers is in fact a zombie; he was executed but simply refuses to die until his firm's invoices for his defense at trial are paid by his descendants. His partners are vampires. Makes one wonder...
The bank was the fraud victim, you're collateral damage. Er, um, no pun intended...
After the fraud uses your personal information to take money from a third party creditor, said creditor unfairly trashes your reputation, since that's the easiest recourse they have. Actual damages inflicted by the creditor in what looks to me like a defamation case might well be difficult to demonstrate, but not impossible: that nasty little clause in your credit card agreements that makes everything go to 31.99% APR if anything derogatory appears on your credit report means the defamation is costing you actual cash.
I haven't looked at the reciprocal one, but as I see it the difference between MS-PL and other open source licenses is that MS-PL is written to maximize link-compatibility rather than copy-paste-compatibility.
It's often very difficult to migrate an application, piece by piece, to a new design. Look at Microsoft Office, for example.
Microsoft Office is a very atypical example though -- they're trying to maintain ~10 years of backward compatibility while moving the design forward. In any case, though, the usual cause I've found for difficulty changing the design of existing software is a combination of poor/insufficient/leaky abstraction and a crippling shortage of test resources.
The poor abstraction is often caused by people who can manage to write COBOL or FORTRAN style code in whatever language you put in front of them. 5000 line methods chock full of if, switch, and such. One shop I've worked at had one such routine -- in Java no less -- consisting of if (customerID == foo-id) { do-foo-special-work; } else if (customerID == bar-id) { do-bar-special-work; } else... on and on for pages.
The lack of test resources is more of an organizational problem. When the pointy-haired boss dictates that there will be one tester per 12 programmers, and that that tester will be in another city if not on the opposite side of the planet, and you will not be given time to build a test framework... well... sorry, and thank you for playing.
If you're right, this is an attempt to stop a certain kind of behavior by a large company. Isn't that what the legislature is for?
#include "disclaimer.h"
#include "not_lawyer.h"
I'm not so sure, but at least in contract law, courts seem to encourage the use of "self-help" measures when it looks like things are going sour, before getting the courts involved. If you squint a bit this could be considered a self-help measure of a different sort. Besides, we've all seen what a hash legislatures seem to make of things -- sometimes it's even a cryptographically strong hash: it's infeasible to figure out the purpose of the legislation by reading the bill.:-)
Agriculture is probably humanity's most powerful invention, and why hasn't it turned on us to kill us? Because we control it? Or is it simply because it helps us and that is what it's primary purpose is?
Unfortunately, it seems to have partially backfired in the last century both in dietary and land use terms.
No sooner would your spreadsheet application spontaneously become a 3D game engine than an intelligence designed for help spontaneously become a harmful entity.
"The greatest dangers to liberty lurk in insidious encroachment by men of zeal, well-meaning but without understanding." -- Louis D. Brandeis
Intelligence is not wisdom. Perhaps an exponentially higher intelligence might consistently find wisdom, but for now the two seem to be orthogonal.
Perhaps the best argument for the singularity is that we're apparently screwed without it, so better to jump into the unknown than stay on our apparently screwage-bound course.
A query on a row store has to query entire rows, which means you'll often end up hitting fields you don't give a damn about while looking for the specific fields you want to return.
So build a covering index. What they seem to be driving at is aggregation performance, though, not just raw read performance; for that you need to build a materialized view, and not all RDBMS support those conveniently so you end up building a horrid mess of triggers or just rebuilding your proto-aggregate table out of a cron job. Column stores have an easier time of aggregation (just scan the column and fold) unless you are doing something that really is row-oriented, like SUM(A.x + B.y * B.n) over some huge wretched join with conditions have absolutely nothing to do with what you're aggregating.
Probably a better policy is to have a DBMS that supports a variety of table store formats. MySQL wins that point, and would probably run away with it if it supported a column store format. Second best is to have a variety of index store formats, like Postgres. Or you can always have just one trick but optimize it to the hilt.
Right.. sot he first chapter goes on an on how hot it is that once you assign a value to a variable, it can no longer be changed. You see, that's super hot, since you can assign Pi = 3.14 and then you can't change it.
It's called static single assignment form. It's arguable whether requiring the programmer to write in this formal style is better than letting the compiler convert the code though -- sort of like the holy war over manifest versus inferred types.
I believe concurrency will in the end be most implemented via RAM transactions, which proves a simple and effective model for handling race conditions, while retaining the look of the separate procedures as imperative sequential code as we know it today.
Software transactional memory is nifty, and there are implementations like SXM for.Net by Maurice Herlihy, but the bizarre contortions required to get between an object and its fields at runtime (in SXM's case, wrapping all fields in properties and using a runtime-generated subclass that overrides the setters and getters to monitor the transaction) feel more unnatural to me than SSA form.
Actually the metric I was thinking of is applicable to both firemen and help desk slaves: median and 95th percentile time to resolution of a call. Only difference is in exactly how the call is resolved. Different kinds of calls vary wildly in scale though: perhaps group them into buckets by issue type (single-family home, apartment building, brushfire, towering inferno, unauthorized possession of reading material, etc.) Median gives you a typical value, 95th percentile points out peaks and trouble spots while allowing you to throw out some outliers.
See the flamewar on the OSI mailing lists; all the above concerns and more have been aired.
No, seriously... large complex component-based software system + GPL = instant holy wars, because the line where one work ends and another begins is no longer clear.
"Dammit Jim, I'm an engineer, not an attorney!", but it seems to me that in practice, the GPL's process-boundary condition becomes little more than a performance issue because you have to use message passing over some kind of communications link instead of loading in-process. For additional flavor, release the "client side" of the message passing bits under the AFL or similar, and the "server side" obligingly under the GPL.
About 10 years ago, I used a Sun workstation for a while. It had an optical mouse which required a mouse pad (with a grid on it). Despite using that system for about 1.5 years, I never got used to having limits on where I could put the mouse. I now avoid mouse pads like the plague.
I actually prefer those to conventional ball mice. The most amazingly trouble-free mechanical contact mouse design I ever saw was the Hawley/Digital "flying saucer" mouse featured on VAXen and DECstations, which had two circular wheels set at shallow angles on its underside -- there was no mechanism to carry mouse gunk inside.
Ah yes, the Ken Thompson hack. I wonder if what's happening in software is analogous to what's become of the United States since the early days -- it's a citizen's responsibility to know every line of the Code (both meanings), but the Code has bloated to the point where that is a practical impossibility for anyone, even a professional in the field. I won't go into the matter of closed-source Code in the legal sphere; John Gilmore is probably the one to ask about that.
So split the difference. LGPL enforces a return of contributions to the library while allowing the code to be included in proprietary closed-source products.
There is a Nevada regulatory requirement (PDF, see section 1.080.2a) that a gaming device must refuse to execute non-blessed code. While this would conflict with anti-tivoization language, it isn't necessarily the author's choice, and it need not conflict with the spirit of Free software: if you want to put your modified code in a gaming device in front of real customers, just get your friendly local regulatory agency to sign off on your modifications. The regulators' mission is to keep everyone honest, not stifle development.
The best rationale for allowing "tivoization" I can think of is in environments like gaming where there is generally a regulatory requirement that the box (a) won't come up unless all signatures pass muster, and (b) enforces jurisdictional policy. A number of electronic gaming machines run Linux, and an anti-tivoization constraint would put manufacturers in a regulatory double-bind.
In any case, the business-product restriction has nothing to do with how the acquiring business is organized -- it's a question of the intended use of the product. Vanishingly few people buy electronic gaming machines for purposes other than to operate them as a business.
Trade and commerce, if they were not made of India rubber, would never manage to bounce over the obstacles which legislators are continually putting in their way; and, if one were to judge these men wholly by the effects of their actions, and not partly by their intentions, they would deserve to be classed and punished with those mischievous persons who put obstructions on the railroads. -- Henry David Thoreau
It seems that the prime function of legislative bodies is to generate well-intentioned but ill-conceived responses to the latest moral panic. Undoubtedly some fraction of the developer community will want to engage in 'hacktivism', but for most life and work must still go on. Competent developers know how to work around legislative obstruction and still do business: a jurisdiction configuration that specifies whether each of the system's jurisdiction-sensitive features are permitted, limited, and/or forbidden.
Gambling games are configured to permit or forbid features like electronic funds transfer and define the threshold at which wins are recorded for tax purposes.
Commercial cryptography toolkits are configured to specify which cipher suites are available, what maximum key strengths are used, and how the inclusion of law enforcement access features modifies the limits.
Even the GPLv2 itself permits geographic restriction of a license to allow compliance with jurisdictional issues.
You can't install software on a user, so anything a program needs to be successfully run after being installed needs to be HKLM.
Actually, HKCR is a composite view of HKCU\SOFTWARE\Classes overriding HKLM\SOFTWARE\Classes. User-specific registrations dominate machine-wide registrations. In theory, this is also what the install for "just me" option in many installers is for.
Unfortunately, all too often, software installed under a user with XP admin rights is still installed in %ProgramFiles%, often having unwanted side effects on HKLM or %AllUsersProfile%, and the Uninstall entry still only appears under the offending user's profile. A pox on badly written installers!
Ratings do NOT create censorship -- they just inform the consumer. It is sort of like complaining that food labels "censor" high-calorie fatty foods and create an underground market for twinkies.
Now, if certain retailers decide not to carry any "mature" ratings, that is up to the retailer. But there will always be places on the internet that you can order this stuff from. It is not like www.gogamer.com qualifies as an "underground."
Only problem here is that the video game business has gone all (RI|MP)AA in its distribution scheme - you need to go platinum and you need to sell through the "big box" stores or you're not even going to recover the cost of the (RI|MP)AA-grade media assets. If the "big box" stores won't sell your game you'll be lucky to sell 10k units. How many X rated movies have been blockbusters on the scale of, say, the recent Lord of the Rings movies?
Guy L. Steele sketched this amusing commentary on problems in '70s fluidic computing, one episode of the Crunchly saga now entwined with the Jargon File.
All they have proven is that turning the gene back on can alleviate the disease. This is no closer to a cure than any other single gene disease that could be fixed by putting a good copy in.
So Rett syndrome is a candidate for gene therapy, if the issues with gene therapy can be hammered out...
The title and summary are the biggest load of non-biologists trying to write about biology I've read in months.
If you haven't noticed by now, this whole site is the second biggest load of non-xists trying to write about x for just about any value of x. (Grins, ducks hail of hot grits from site regulars)
Slashdot Mirror
It's the same Iolo, proprietor of New World Arbalest.
As for U-235, I think one of the most inventive uses I've seen is powering a nuclear saltwater steam rocket engine for interplanetary use. Just watch where you point it, the exhaust is really nasty.
In the city of Ankh-Morpork the head of the Guild of Lawyers is in fact a zombie; he was executed but simply refuses to die until his firm's invoices for his defense at trial are paid by his descendants. His partners are vampires. Makes one wonder...
The bank was the fraud victim, you're collateral damage. Er, um, no pun intended...
After the fraud uses your personal information to take money from a third party creditor, said creditor unfairly trashes your reputation, since that's the easiest recourse they have. Actual damages inflicted by the creditor in what looks to me like a defamation case might well be difficult to demonstrate, but not impossible: that nasty little clause in your credit card agreements that makes everything go to 31.99% APR if anything derogatory appears on your credit report means the defamation is costing you actual cash.
I haven't looked at the reciprocal one, but as I see it the difference between MS-PL and other open source licenses is that MS-PL is written to maximize link-compatibility rather than copy-paste-compatibility.
Microsoft Office is a very atypical example though -- they're trying to maintain ~10 years of backward compatibility while moving the design forward. In any case, though, the usual cause I've found for difficulty changing the design of existing software is a combination of poor/insufficient/leaky abstraction and a crippling shortage of test resources.
The poor abstraction is often caused by people who can manage to write COBOL or FORTRAN style code in whatever language you put in front of them. 5000 line methods chock full of if, switch, and such. One shop I've worked at had one such routine -- in Java no less -- consisting of if (customerID == foo-id) { do-foo-special-work; } else if (customerID == bar-id) { do-bar-special-work; } else ... on and on for pages.
The lack of test resources is more of an organizational problem. When the pointy-haired boss dictates that there will be one tester per 12 programmers, and that that tester will be in another city if not on the opposite side of the planet, and you will not be given time to build a test framework... well... sorry, and thank you for playing.
#include "not_lawyer.h"
I'm not so sure, but at least in contract law, courts seem to encourage the use of "self-help" measures when it looks like things are going sour, before getting the courts involved. If you squint a bit this could be considered a self-help measure of a different sort. Besides, we've all seen what a hash legislatures seem to make of things -- sometimes it's even a cryptographically strong hash: it's infeasible to figure out the purpose of the legislation by reading the bill. :-)
Unfortunately, it seems to have partially backfired in the last century both in dietary and land use terms.
"The greatest dangers to liberty lurk in insidious encroachment by men of zeal, well-meaning but without understanding." -- Louis D. Brandeis
Intelligence is not wisdom. Perhaps an exponentially higher intelligence might consistently find wisdom, but for now the two seem to be orthogonal.
Perhaps the best argument for the singularity is that we're apparently screwed without it, so better to jump into the unknown than stay on our apparently screwage-bound course.
So build a covering index. What they seem to be driving at is aggregation performance, though, not just raw read performance; for that you need to build a materialized view, and not all RDBMS support those conveniently so you end up building a horrid mess of triggers or just rebuilding your proto-aggregate table out of a cron job. Column stores have an easier time of aggregation (just scan the column and fold) unless you are doing something that really is row-oriented, like SUM(A.x + B.y * B.n) over some huge wretched join with conditions have absolutely nothing to do with what you're aggregating.
Probably a better policy is to have a DBMS that supports a variety of table store formats. MySQL wins that point, and would probably run away with it if it supported a column store format. Second best is to have a variety of index store formats, like Postgres. Or you can always have just one trick but optimize it to the hilt.
It's called static single assignment form. It's arguable whether requiring the programmer to write in this formal style is better than letting the compiler convert the code though -- sort of like the holy war over manifest versus inferred types.
Software transactional memory is nifty, and there are implementations like SXM for .Net by Maurice Herlihy, but the bizarre contortions required to get between an object and its fields at runtime (in SXM's case, wrapping all fields in properties and using a runtime-generated subclass that overrides the setters and getters to monitor the transaction) feel more unnatural to me than SSA form.
Actually the metric I was thinking of is applicable to both firemen and help desk slaves: median and 95th percentile time to resolution of a call. Only difference is in exactly how the call is resolved. Different kinds of calls vary wildly in scale though: perhaps group them into buckets by issue type (single-family home, apartment building, brushfire, towering inferno, unauthorized possession of reading material, etc.) Median gives you a typical value, 95th percentile points out peaks and trouble spots while allowing you to throw out some outliers.
See the flamewar on the OSI mailing lists; all the above concerns and more have been aired.
No, seriously... large complex component-based software system + GPL = instant holy wars, because the line where one work ends and another begins is no longer clear.
"Dammit Jim, I'm an engineer, not an attorney!", but it seems to me that in practice, the GPL's process-boundary condition becomes little more than a performance issue because you have to use message passing over some kind of communications link instead of loading in-process. For additional flavor, release the "client side" of the message passing bits under the AFL or similar, and the "server side" obligingly under the GPL.
I actually prefer those to conventional ball mice. The most amazingly trouble-free mechanical contact mouse design I ever saw was the Hawley/Digital "flying saucer" mouse featured on VAXen and DECstations, which had two circular wheels set at shallow angles on its underside -- there was no mechanism to carry mouse gunk inside.
Ah yes, the Ken Thompson hack. I wonder if what's happening in software is analogous to what's become of the United States since the early days -- it's a citizen's responsibility to know every line of the Code (both meanings), but the Code has bloated to the point where that is a practical impossibility for anyone, even a professional in the field. I won't go into the matter of closed-source Code in the legal sphere; John Gilmore is probably the one to ask about that.
So split the difference. LGPL enforces a return of contributions to the library while allowing the code to be included in proprietary closed-source products.
There is a Nevada regulatory requirement (PDF, see section 1.080.2a) that a gaming device must refuse to execute non-blessed code. While this would conflict with anti-tivoization language, it isn't necessarily the author's choice, and it need not conflict with the spirit of Free software: if you want to put your modified code in a gaming device in front of real customers, just get your friendly local regulatory agency to sign off on your modifications. The regulators' mission is to keep everyone honest, not stifle development.
The best rationale for allowing "tivoization" I can think of is in environments like gaming where there is generally a regulatory requirement that the box (a) won't come up unless all signatures pass muster, and (b) enforces jurisdictional policy. A number of electronic gaming machines run Linux, and an anti-tivoization constraint would put manufacturers in a regulatory double-bind.
In any case, the business-product restriction has nothing to do with how the acquiring business is organized -- it's a question of the intended use of the product. Vanishingly few people buy electronic gaming machines for purposes other than to operate them as a business.
It seems that the prime function of legislative bodies is to generate well-intentioned but ill-conceived responses to the latest moral panic. Undoubtedly some fraction of the developer community will want to engage in 'hacktivism', but for most life and work must still go on. Competent developers know how to work around legislative obstruction and still do business: a jurisdiction configuration that specifies whether each of the system's jurisdiction-sensitive features are permitted, limited, and/or forbidden.
Actually, HKCR is a composite view of HKCU\SOFTWARE\Classes overriding HKLM\SOFTWARE\Classes. User-specific registrations dominate machine-wide registrations. In theory, this is also what the install for "just me" option in many installers is for.
Unfortunately, all too often, software installed under a user with XP admin rights is still installed in %ProgramFiles%, often having unwanted side effects on HKLM or %AllUsersProfile%, and the Uninstall entry still only appears under the offending user's profile. A pox on badly written installers!
Have you ever heard the term "copycenter"? As in, "I don't care, just take it down to the copy center and make as many copies as you want."
Only problem here is that the video game business has gone all (RI|MP)AA in its distribution scheme - you need to go platinum and you need to sell through the "big box" stores or you're not even going to recover the cost of the (RI|MP)AA-grade media assets. If the "big box" stores won't sell your game you'll be lucky to sell 10k units. How many X rated movies have been blockbusters on the scale of, say, the recent Lord of the Rings movies?
Quoted for truth! Well, actually I think Windows 2000 was a similar "sweet spot".
Guy L. Steele sketched this amusing commentary on problems in '70s fluidic computing, one episode of the Crunchly saga now entwined with the Jargon File.
Around here the preferred phrase seems to be 'you insensitive clod', but I think we get your meaning.
So Rett syndrome is a candidate for gene therapy, if the issues with gene therapy can be hammered out...
If you haven't noticed by now, this whole site is the second biggest load of non-xists trying to write about x for just about any value of x. (Grins, ducks hail of hot grits from site regulars)