The difference is that unless it's a kernel update Linux doesn't really need a reboot on update.
A C library update is pretty noticeable too; you might be able to keep the kernel up, but there's not a lot of point given that virtually every user process is entangled with the library being updated. OTOH, if you're having to update the C library on a regular basis, you've got pretty serious problems anyway...
Strongly typed languages like C++ could make the SQL injection problem go away by using SQL libraries that only accept strings of type "safe_string" (or whatever). The compiler will enforce the rest.
That'll only cover 99% of the cases. (Hey, it's a very useful subset!) The rest require evil string building because the SQL language itself doesn't allow for anything better; the awkward cases are where you want to substitute something other than a simple value, such as a whole WHERE clause (you could pass in a string to a stored procedure which reinterprets that as SQL, but that's hardly going to help with injection problems).
Unfortunately, there's no substitute for taking care and having someone else (preferably someone very suspicious and knowledgeable) review the code and the deployment configuration too.
How on earth can you think that malloc()/free()/strcat()/strdup() and raw pointers is good programming practice?
It's pretty horrible, but it doesn't have the deployment horrors that C++ does. (I've worked with a few programs that were C++ and building redistributable binaries of them was always a pain; when the developer switched to using C, the problems went away.)
The real problem of C++ (apart from its half-assed-ness in the OO department, from a Smalltalk perspective:-)) is that it tends to bind consumers of interfaces very closely to the implementations of those interfaces. Yes, this makes the object code fast, but the cost is that it also makes it much more brittle. Having to recompile the world (a slow business with C++, unlike most other languages) just because of the addition of a private variable to a class definition is not a winning strategy, and PIMPL is a band aid. (It also tempts programmers into writing inline methods to "simplify" access to the implementation fields and methods, all of which is heading rapidly back into the hell of tight binding.)
Although I will agree that language choice *usually* matters far less than algorithmic choices and occasionally people jump to a language change in a project to alleviate slowness only to end up not significantly better than they started because of glaring design problems that dwarf the language performance concerns.
Language choice can make a difference, but that's usually because higher level programming languages tend to have very well written string and buffer management libraries that avoid most of the problems that bedevil a lot of code in C and C++ (and Java too). It's not that the lower level languages can't be used efficiently, but rather that a substantial fraction of their practitioners simply don't use their tools well. The higher level languages, because they conceal more, can hide enough gory details that they can actually keep enough useful metadata around to be able to optimize algorithmically. Well, some of the time.
The next time I see someone using strcat() in an inner loop (or += over Strings in Java) I think I'll scream.
The days where we could just say "it's ok, buy a bigger server" or "next year we'll have enough computing power" are over. We now have to do more with what we've got, and that means more efficient programming.
The days of the hardware being trivially the source of all speedups required are long gone. Also long gone are the massive shared memory machines; they really didn't scale without the use of enormous amounts of money and that never really changed (the real cost of supercomputers back at around 2000 was in the funky backplane interconnect). Now, we have to learn to do more with message passing (that scales far better, in many ways) and we need to learn to only pass around the data that's necessary (because such communications cost).
The "downside" of this is that all those programmers who decided that parallel programming was all about throwing shared memory and threads at a problem before sprinkling in locks to stop things crashing, well those programmers' skills are pretty low value now. The tech they picked can't ever scale up or out. People used to network programming are in a better position; after all, that's clearly about true asynchronous coding...
It's a nice principle, but lots of recent precedents (East Texas anyone?) are weakening that defense in favor of the trolls.
The key is whether the two parties in the suit were talking about licensing in the first place, not whether it reached the attention of the court. The court is just there to resolve the dispute given that the parties don't agree.
Not at all, nor is it needed since this is not a trademark case this is a patent case. It is completely legal to wait until the infringer is making the most money to sue. It is also strategically intelligent.
While the principle of laches means that it is not a good idea (as a patent holder) to wait to contact the potential infringer until they've made a massive profit, if the holder and alleged infringer have been in contact (even if the alleged infringer has then just blown off the holder) then the principle won't hold. Legally, the defendant will know about the possibility of a claim against it and will therefore be able to take reasonable steps to minimize any potential harm.
What's unfortunate is that it can take a long time and a lot of money to bring a patent case to court. That's a fair part of what's wrong with the US patent system. (The other big problem was the habit of awarding patents for software and business methods for little more than stapling "on the internet!!!" onto the end of the claims, but that's another problem entirely.)
Google's results were exactly as shitty as everyone else's, for many years.
You forget just how bad things were before Google. I remember the bad old days, and they were terrible. Finding things with AltaVista was really extremely hit-and-miss, and Yahoo! was only good if they happened to want to index what you were interested in (invariably not for me). Google indexed more, and gave much more precise results too. (Nowadays, all search engines are much better than they used to be; the general level of deployed technology has moved on.)
What made Google a force was the simple fact that they didn't plaster ten metric megaasstons of bullshit all over their search page.
That was a nice bonus, but the fact that they took you to where you wanted to go... that was absolutely golden.
They've not done a damned positive thing for the US in these past years, if anything....
Of course they haven't. The Tea Party fanatics (coupled with a minor swing to the Reps in the House) have caused Congress to be even more obstructionist than normal, and they have to be involved because doing anything requires paying for it. If you can't borrow more, raise taxes, cut outlay on entitlements or reduce the size of military spending, is it any wonder that the feds are paralyzed? Nothing else in the federal budget is of enough size to matter in the big picture. And when politics is this bunged up, is it any wonder that other groups try to surreptitiously usurp power for themselves? (Hence the rise of all the tinpot fascistic tendencies.)
Would the situation be the same if the strengths of the parties in various parts of government were reversed? Probably not, given that the Dems don't seem to be captured nearly as thoroughly by their wingnuts, but it's really hard to say (and isn't the current situation anyway, so such speculation isn't really leading anywhere).
Incorporation itself does not mean you obligate yourself to the pursuit of profit over all else. Having outside investors does.
Even then, not really. Management usually has a lot of freedom to pursue objectives, especially if these are ones that have been widely described in communications both to the investors and to the general public. Though they have a responsibility to produce an operating profit and service their debts, they don't have a duty to increase revenue every quarter.
Of course, Wall St would love that to change as it would make their lives (as investors and investment advisers) much easier. But that's really just a mark of how dysfunctional Wall St is.
Provided you had the budget in the first place. A lot of people had to save hard just to be able to afford a Spectrum (and Sinclair's genius was in recognizing that this was actually a substantial market that the other players weren't really focusing on). There is truly no comparison between having a Spectrum and not having a BBC because you've got to save 2.5 times as much to be able to afford one...
Better graphics modes too. Colour addressable per pixel in various resolutions, plus a teletext mode that only used 1KB. And commands in the BASIC interpreter to draw on that screen too.
Unfortunately, the awesome graphics modes used so much memory that you had no space left for your program. Picking the graphics mode was a matter of trading off between having memory for code and memory for output. I remember all this from when I wrote (what I now know to be) my first IDE. For the BBC. The trick was I loaded pieces of the program off floppy disk when needed. (I still hate those DFS floppies; only having a maximum of 31 files per disk was very limiting. Didn't have the ADFS available.)
And sound. BBC Micro had 3 channels of tones, plus a white noise channel, all ADSR programmable. Speccy had a single bit attached to a speaker.
But the Z80 was (with a bit of careful coding) fast enough to use PCM to drive that speaker. If your assembly chops were up to it, you could do truly impressive things. The BBC had some very interesting hardware, but really wasn't all that fast and the (lovely for the time) BASIC implementation greatly restricted what you could do in mixed BASIC/ASM code (unless you didn't mind being stuck in ASM).
I don't think it's a matter of WP being broken at it's core. They have some of the best core developers I've seen work on any open source project. However, it is easy to fall out of the best practices for running a WP site. Also consider that it is the most largely growing CMS out of them all.
What they appear to have is a more subtle problem: it's not designed to Fail Safe. Get something wrong? Fail to update? Any problem, and you end up with some ability to do local damage and set in motion a full exploit. If the system failed safe, it wouldn't allow you to do anything until you'd proved that you were legit; any bugs would just result in a reduction in what could be done.
That said, I've got no idea how far you could get with creating a full CMS on the principles I described. Yes, I use the principle in my own web programming but there the set of operations is far more restricted. (I also don't support run-time updating.)
In this case, it's reasonable for The New York Review of Books to be somewhat US-centric. After all, its primary audience is in the US.
However, there continues to be a strong case for pooling scientific funding (and projects and instruments and...) across many countries, especially when those projects are very large. You're not going to get all the best people in the world working in one country in any mature field (for all sorts of complex reasons) and you do want the best people talking to each other. Once they start talking, they will come up with ideas for areas to research; those are the seeds of proposals and projects. Given all that, pooled funding also makes sense. Well, provided the various funding agencies agree; that doesn't always go smoothly...
Given all the above, the disappointing thing is that tNYRoB didn't pick up on this matter. It's a reasonably well respected publication that at least tries to be not too parochial. Pity they failed this time (if only perhaps in the choice of Steven Weinberg).
Old construction in the UK has very strange hot water plumbing. Vented hot water heaters and gravity feed for the hot water, pressure feed for cold.
It works if you don't have mixer taps. It even wasn't completely crazy originally, because the bore of the water supply to the property was probably tiny and so the pressure difference wasn't all that much. To be fair, that was due to the practice of using lead pipes, which need to be much thicker than copper or polycarbonate pipes and so tended to have smaller bores. Thankfully, lead's been long outlawed as a pipe material but the time to remove it from everywhere it was used is long (especially from rarely used pipes that aren't carrying water used for drinking).
Are you trying to say Oracle is less evil than Google? Are you at all familiar with Oracle?
Very few companies are more evil than Oracle — especially when it comes to licensing charges — with the exceptions of those that actually kill and maim people (Union Carbide and the Bhopal disaster is one of the worst examples I know of there).
Yet it always amazes me just how many people fails to understand such simple premises when it comes to science. They seem to think that failures aren't science and that nothing is learned from them.
But if you then ask them about gravity, they'll almost always talk about Newton, yet if you point out that Newton got it wrong (mostly right, but still wrong), they will invariably tell you "that's different".
That's unhelpful. Newton's work on gravitation was based on a lot of previous work by other scientists over a long time; experimental work, observational work and theoretical work. Newton was the guy who first pulled it all together into a coherent framework, but he couldn't have done it without Copernicus, Brahe, Kepler, Galileo, probably many others. What replaced Newtonian dynamics? Relativity (which reduces to a damn good approximation to Newtonian for most everyday activity) and that was Einstein building on a lot of work of others again (particularly Maxwell, Lorentz, Michelson, Morley, and especially Riemann, and Eddington was important in confirming GR) some of whom were definitely experimentalists.
All of that really just goes to show that science has always been about building on the work of others. There's no problem at all for some people to be theoreticians (they tend to be the ones to come up with ideas to unify areas) but you still need plenty of experimenters and observationalists too to tell the theory guys where it is worthwhile doing something and to check that the results of the theory apply to reality. Scientific theory, like computer programming (or anything else highly abstract), can easily slip into GIGO.
Astronomers also have their own favorite zero time, but use only years (with a decimal point and lots more digits to whatever precision they need at the moment).
Astronomers count in days since 12:00 January 1, 4713 BC (a Monday). They most certainly use fractions of days as well, where it makes sense.
If someone is subjected to videos that provokes extreme feelings of disgust or anger, it's probably reasonable to assume that all sorts of hormones and chemicals get released into the blood. Some of these could be responsible for the increase in [penile] girth.
Sounds like you've got the theme for some prize-winning research right there! Admittedly, it'd probably be the Ig Nobel Prize but it'd still be something...
Slashdot Mirror
The difference is that unless it's a kernel update Linux doesn't really need a reboot on update.
A C library update is pretty noticeable too; you might be able to keep the kernel up, but there's not a lot of point given that virtually every user process is entangled with the library being updated. OTOH, if you're having to update the C library on a regular basis, you've got pretty serious problems anyway...
Strongly typed languages like C++ could make the SQL injection problem go away by using SQL libraries that only accept strings of type "safe_string" (or whatever). The compiler will enforce the rest.
That'll only cover 99% of the cases. (Hey, it's a very useful subset!) The rest require evil string building because the SQL language itself doesn't allow for anything better; the awkward cases are where you want to substitute something other than a simple value, such as a whole WHERE clause (you could pass in a string to a stored procedure which reinterprets that as SQL, but that's hardly going to help with injection problems).
Unfortunately, there's no substitute for taking care and having someone else (preferably someone very suspicious and knowledgeable) review the code and the deployment configuration too.
How on earth can you think that malloc()/free()/strcat()/strdup() and raw pointers is good programming practice?
It's pretty horrible, but it doesn't have the deployment horrors that C++ does. (I've worked with a few programs that were C++ and building redistributable binaries of them was always a pain; when the developer switched to using C, the problems went away.)
The real problem of C++ (apart from its half-assed-ness in the OO department, from a Smalltalk perspective :-)) is that it tends to bind consumers of interfaces very closely to the implementations of those interfaces. Yes, this makes the object code fast, but the cost is that it also makes it much more brittle. Having to recompile the world (a slow business with C++, unlike most other languages) just because of the addition of a private variable to a class definition is not a winning strategy, and PIMPL is a band aid. (It also tempts programmers into writing inline methods to "simplify" access to the implementation fields and methods, all of which is heading rapidly back into the hell of tight binding.)
Although I will agree that language choice *usually* matters far less than algorithmic choices and occasionally people jump to a language change in a project to alleviate slowness only to end up not significantly better than they started because of glaring design problems that dwarf the language performance concerns.
Language choice can make a difference, but that's usually because higher level programming languages tend to have very well written string and buffer management libraries that avoid most of the problems that bedevil a lot of code in C and C++ (and Java too). It's not that the lower level languages can't be used efficiently, but rather that a substantial fraction of their practitioners simply don't use their tools well. The higher level languages, because they conceal more, can hide enough gory details that they can actually keep enough useful metadata around to be able to optimize algorithmically. Well, some of the time.
The next time I see someone using strcat() in an inner loop (or += over Strings in Java) I think I'll scream.
The days where we could just say "it's ok, buy a bigger server" or "next year we'll have enough computing power" are over. We now have to do more with what we've got, and that means more efficient programming.
The days of the hardware being trivially the source of all speedups required are long gone. Also long gone are the massive shared memory machines; they really didn't scale without the use of enormous amounts of money and that never really changed (the real cost of supercomputers back at around 2000 was in the funky backplane interconnect). Now, we have to learn to do more with message passing (that scales far better, in many ways) and we need to learn to only pass around the data that's necessary (because such communications cost).
The "downside" of this is that all those programmers who decided that parallel programming was all about throwing shared memory and threads at a problem before sprinkling in locks to stop things crashing, well those programmers' skills are pretty low value now. The tech they picked can't ever scale up or out. People used to network programming are in a better position; after all, that's clearly about true asynchronous coding...
It's a nice principle, but lots of recent precedents (East Texas anyone?) are weakening that defense in favor of the trolls.
The key is whether the two parties in the suit were talking about licensing in the first place, not whether it reached the attention of the court. The court is just there to resolve the dispute given that the parties don't agree.
Not at all, nor is it needed since this is not a trademark case this is a patent case. It is completely legal to wait until the infringer is making the most money to sue. It is also strategically intelligent.
While the principle of laches means that it is not a good idea (as a patent holder) to wait to contact the potential infringer until they've made a massive profit, if the holder and alleged infringer have been in contact (even if the alleged infringer has then just blown off the holder) then the principle won't hold. Legally, the defendant will know about the possibility of a claim against it and will therefore be able to take reasonable steps to minimize any potential harm.
What's unfortunate is that it can take a long time and a lot of money to bring a patent case to court. That's a fair part of what's wrong with the US patent system. (The other big problem was the habit of awarding patents for software and business methods for little more than stapling "on the internet!!!" onto the end of the claims, but that's another problem entirely.)
Google's results were exactly as shitty as everyone else's, for many years.
You forget just how bad things were before Google. I remember the bad old days, and they were terrible. Finding things with AltaVista was really extremely hit-and-miss, and Yahoo! was only good if they happened to want to index what you were interested in (invariably not for me). Google indexed more, and gave much more precise results too. (Nowadays, all search engines are much better than they used to be; the general level of deployed technology has moved on.)
What made Google a force was the simple fact that they didn't plaster ten metric megaasstons of bullshit all over their search page.
That was a nice bonus, but the fact that they took you to where you wanted to go... that was absolutely golden.
They've not done a damned positive thing for the US in these past years, if anything....
Of course they haven't. The Tea Party fanatics (coupled with a minor swing to the Reps in the House) have caused Congress to be even more obstructionist than normal, and they have to be involved because doing anything requires paying for it. If you can't borrow more, raise taxes, cut outlay on entitlements or reduce the size of military spending, is it any wonder that the feds are paralyzed? Nothing else in the federal budget is of enough size to matter in the big picture. And when politics is this bunged up, is it any wonder that other groups try to surreptitiously usurp power for themselves? (Hence the rise of all the tinpot fascistic tendencies.)
Would the situation be the same if the strengths of the parties in various parts of government were reversed? Probably not, given that the Dems don't seem to be captured nearly as thoroughly by their wingnuts, but it's really hard to say (and isn't the current situation anyway, so such speculation isn't really leading anywhere).
Steve (the Quarterback) ends up with [...] a paternity suite.
Does that come with some nice cushions and a coffee table? Or is it just some sofas large enough to sleep on when the going gets too rough?
Incorporation itself does not mean you obligate yourself to the pursuit of profit over all else. Having outside investors does.
Even then, not really. Management usually has a lot of freedom to pursue objectives, especially if these are ones that have been widely described in communications both to the investors and to the general public. Though they have a responsibility to produce an operating profit and service their debts, they don't have a duty to increase revenue every quarter.
Of course, Wall St would love that to change as it would make their lives (as investors and investment advisers) much easier. But that's really just a mark of how dysfunctional Wall St is.
Nearly three times as much...
But there was really no comparison.
Provided you had the budget in the first place. A lot of people had to save hard just to be able to afford a Spectrum (and Sinclair's genius was in recognizing that this was actually a substantial market that the other players weren't really focusing on). There is truly no comparison between having a Spectrum and not having a BBC because you've got to save 2.5 times as much to be able to afford one...
Better graphics modes too. Colour addressable per pixel in various resolutions, plus a teletext mode that only used 1KB. And commands in the BASIC interpreter to draw on that screen too.
Unfortunately, the awesome graphics modes used so much memory that you had no space left for your program. Picking the graphics mode was a matter of trading off between having memory for code and memory for output. I remember all this from when I wrote (what I now know to be) my first IDE. For the BBC. The trick was I loaded pieces of the program off floppy disk when needed. (I still hate those DFS floppies; only having a maximum of 31 files per disk was very limiting. Didn't have the ADFS available.)
And sound. BBC Micro had 3 channels of tones, plus a white noise channel, all ADSR programmable. Speccy had a single bit attached to a speaker.
But the Z80 was (with a bit of careful coding) fast enough to use PCM to drive that speaker. If your assembly chops were up to it, you could do truly impressive things. The BBC had some very interesting hardware, but really wasn't all that fast and the (lovely for the time) BASIC implementation greatly restricted what you could do in mixed BASIC/ASM code (unless you didn't mind being stuck in ASM).
I don't think it's a matter of WP being broken at it's core. They have some of the best core developers I've seen work on any open source project. However, it is easy to fall out of the best practices for running a WP site. Also consider that it is the most largely growing CMS out of them all.
What they appear to have is a more subtle problem: it's not designed to Fail Safe. Get something wrong? Fail to update? Any problem, and you end up with some ability to do local damage and set in motion a full exploit. If the system failed safe, it wouldn't allow you to do anything until you'd proved that you were legit; any bugs would just result in a reduction in what could be done.
That said, I've got no idea how far you could get with creating a full CMS on the principles I described. Yes, I use the principle in my own web programming but there the set of operations is far more restricted. (I also don't support run-time updating.)
In this case, it's reasonable for The New York Review of Books to be somewhat US-centric. After all, its primary audience is in the US.
However, there continues to be a strong case for pooling scientific funding (and projects and instruments and ...) across many countries, especially when those projects are very large. You're not going to get all the best people in the world working in one country in any mature field (for all sorts of complex reasons) and you do want the best people talking to each other. Once they start talking, they will come up with ideas for areas to research; those are the seeds of proposals and projects. Given all that, pooled funding also makes sense. Well, provided the various funding agencies agree; that doesn't always go smoothly...
Given all the above, the disappointing thing is that tNYRoB didn't pick up on this matter. It's a reasonably well respected publication that at least tries to be not too parochial. Pity they failed this time (if only perhaps in the choice of Steven Weinberg).
Old construction in the UK has very strange hot water plumbing. Vented hot water heaters and gravity feed for the hot water, pressure feed for cold.
It works if you don't have mixer taps. It even wasn't completely crazy originally, because the bore of the water supply to the property was probably tiny and so the pressure difference wasn't all that much. To be fair, that was due to the practice of using lead pipes, which need to be much thicker than copper or polycarbonate pipes and so tended to have smaller bores. Thankfully, lead's been long outlawed as a pipe material but the time to remove it from everywhere it was used is long (especially from rarely used pipes that aren't carrying water used for drinking).
Are you trying to say Oracle is less evil than Google? Are you at all familiar with Oracle?
Very few companies are more evil than Oracle — especially when it comes to licensing charges — with the exceptions of those that actually kill and maim people (Union Carbide and the Bhopal disaster is one of the worst examples I know of there).
They are public enemies and should be treated as such and stopped before this madness spreads to other industries.
Too late. The music and film industries have been selling us the same old rubbish over and over for decades.
Yes, and soon enough six million dollars will be mere pocket change!
Either your name is Mitt Romney or you're talking about "Quantitative Easing".
Yet it always amazes me just how many people fails to understand such simple premises when it comes to science. They seem to think that failures aren't science and that nothing is learned from them.
But if you then ask them about gravity, they'll almost always talk about Newton, yet if you point out that Newton got it wrong (mostly right, but still wrong), they will invariably tell you "that's different".
That's unhelpful. Newton's work on gravitation was based on a lot of previous work by other scientists over a long time; experimental work, observational work and theoretical work. Newton was the guy who first pulled it all together into a coherent framework, but he couldn't have done it without Copernicus, Brahe, Kepler, Galileo, probably many others. What replaced Newtonian dynamics? Relativity (which reduces to a damn good approximation to Newtonian for most everyday activity) and that was Einstein building on a lot of work of others again (particularly Maxwell, Lorentz, Michelson, Morley, and especially Riemann, and Eddington was important in confirming GR) some of whom were definitely experimentalists.
All of that really just goes to show that science has always been about building on the work of others. There's no problem at all for some people to be theoreticians (they tend to be the ones to come up with ideas to unify areas) but you still need plenty of experimenters and observationalists too to tell the theory guys where it is worthwhile doing something and to check that the results of the theory apply to reality. Scientific theory, like computer programming (or anything else highly abstract), can easily slip into GIGO.
And scientifically, it went around 750 miles in 3 minutes. In an atmosphere. That's a pretty damn awesome piece of engineering.
I hope the FIFA gets in all kinds of trouble over this.
They're too busy swimming in their money bins to care what you (or I) think, peasant! Now buy that merchandise!
Astronomers also have their own favorite zero time, but use only years (with a decimal point and lots more digits to whatever precision they need at the moment).
Astronomers count in days since 12:00 January 1, 4713 BC (a Monday). They most certainly use fractions of days as well, where it makes sense.
If someone is subjected to videos that provokes extreme feelings of disgust or anger, it's probably reasonable to assume that all sorts of hormones and chemicals get released into the blood. Some of these could be responsible for the increase in [penile] girth.
Sounds like you've got the theme for some prize-winning research right there! Admittedly, it'd probably be the Ig Nobel Prize but it'd still be something...
That's when I park the new car and go back to driving my '79 4x4.
Is that the one with the "impale driver's heart on crash" steering column option?