The P's and R (Ruby) are ran by an interpreter, which essentially interprets the text into machine language.
That comment is misleading. The text is never translated to machine code. The interpreter reads the code and the interpreter does the work according to the rules of the language. Java takes an intermediate step and translates the code to bytecode, which can be more easily interpreted. However, so can python.
Also, your performance analysis is lacking. Code size is a big factor here. A perl program may, in a few short lines of code, accomplish something that would take a lot more java code. If you have a complex text processing program, perl may well be much faster. That's because the few short lines of perl code call a lot of highly optimized C text processing routines. The java code may do more of the work in Java code.
So, when evaluating the speed of a language, you should ask yourself: "Does parsing take a significant amount of time for my application?". Java may be fast, but if it doesn't call into the optimized C routines quickly enough, it's going to lose to players like perl, python, and ruby.
Go to LA. Go see the communities that live under the shadow of the gun.
Just out of curiosity, have you been to LA? I have, many times. I'll be driving through there this weekend, and I'll probably do a few things in the city. I'm not saying that I go through the most dangerous areas at night or anything. I just thought I'd mention it. Last time I was in LA my friend was telling me that he was shot in the ass once (he wasn't doing anything, just getting home at night or something).
LA has many problems. The police do not protect the poor communities, leaving them vulnerable to thugs. Most of these thugs posses weapons illegally.
There are between 200 and 250 million guns in the U.S. It would be impossible to reduce that number very much, which means that guns are available to criminals. Weapons in general aren't very hard to make, nor are they hard to smuggle. The only question remains whether guns are available to law-abiding citizens.
That's wonderful. You're willing to pay the price of civil liberty for a decrease in the murder rate. Murders overall are a very low percentage of deaths in the U.S. I value my civil liberties much more than that. And I really don't think there's any evidence that the rate of violent crime would drop by a factor of 95% if the government outlawed guns in the U.S.
I would just like to respond to the "libertarian ideals" comment. To me, libertarianism is about limited government. Not limited in size, but limited in function. A military is a necessary part of the government, and so is a police force. A responsible government would ensure that both of those government elements are strong enough to do their job (part of which is protecting private property). In Liberia and Rwanda, that is clearly not the case, and private property is not protected.
Another element of libertarianism is to break the government into smaller pieces, i.e. states. The states should have most of the government powers, and the federal government should mainly be used to ensure that rights are protected and the country is secured by the military.
You owning a gun is not going to make any criminal decide to stay clean instead.
That statement is suspect on the face of it. Guns owned by police clearly do deter criminal activity, so why not private citizens? But regardless, that was not my point. My point is that gun ownership is a civil right, not dependent on their ability to deter crime overall.
You wouldn't let anyone use statistics to take away our freedom of speech, I ask the same for freedom to own guns. There are bound to be situations in which gun control can lower violent crime, and others in which gun control raises violent crime. This is dependent on many factors, including the particular point in time in question.
When circumstances change and then all of a suddent you want your right to own guns back, it's too late.
Guns absolutely ARE a deterrent to oppressive government. You can see in wars from Vietnam to Iraq that cruise missles or battleships are not efficient weapons against small bands of people with guns.
Not only that, but the government should not have a monopoly on protecting me. I should be able to purchase locks, alarms, cameras, and guns; and I should be able to hire private detectives and body guards.
And as one final note, the Constitution clearly includes gun ownership in the Bill of Rights. I ask that you give the Second Amendment the same protection as the rest of the Constitution. We have enumerated rights, and it is important that we not allow the government to infringe on any of those rights, or else the government will easily be able to divide and conquer our rights. If you disagree with the Second Amendment, you can attempt to make another Amendment, and if that is really what the people want, it will pass.
You could make the argument that censorship reduces the incidence of offensive language, but that's a small benefit to receive in exchange for a civil right.
It's very difficult to really tell conclusively what long term effect gun control has on the incidence of violent crime in general. But regardless, I stand for the civil right for law-abiding citizens to produce, own, transport, and trade firearms. I believe that this right is an important deterrent for oppressive government.
Very interesting point. If you have a 16 way box, and intend to share it among many applications, each of those applications would need to be licensed for 16 processors. Is that what you're saying?
I've used Linux for about 6 years, and FreeBSD for about 5. However, I didn't heavily use FreeBSD until a year ago. What really amazes me is that I can try new operating systems that are all UNIX-like, and yet still each has aspects that really impress me.
I downloaded a couple sets of CDs from sun. Downloading a few GB over DSL, the registration time was definately not the major source of delay in the download process. It is more annoying than downloading FreeBSD, but didn't seem too difficult.
If their new line of processors is successful, that will go a long way. From what I understand, Solaris is gaining interest very quickly, and with it, Sun mindshare.
If they have a leading chip for common server loads, and a leading OS for common server tasks, the support contracts and hardware sales are bound to roll in.
It's sad that Sun is looking at so much financial trouble right now (assuming they actually are, and it might not actually be that bad). Solaris 10 is really quite amazing software, and their chip designs look very promising. I think it would take Linux/FreeBSD a long time to catch up to some of the things in Solaris 10 like ZFS and DTrace.
I'm really pretty new to Solaris, however as soon as I started using it I could tell that they did real research on the kernel. I was very impressed after using Linux/FreeBSD for a long time. I can't think of any other companies actually doing that kind of research still. IBM does a lot of Linux development, but I sometimes get the feeling it's more to just make Linux into an AIX replacement (not that I know much about AIX).
As usual, MySQL is intentionally being confusing about their licensing. You can't GPL a protocol, only the implementation of a protocol.
However, the protocol itself still every bit as proprietary a format as an MS Word document. It just happens that they license their implementation to you under the GPL. You can reverse engineer it, just like the MS word doc format, but MySQL AB can always make subtle changes to break it (just like MS can do to the word format).
It's proprietary, and you can reverse engineer it.
You could use the reverse engineering argument for MS Word documents as well. However, Microsoft can make subtle changes to the format designed to break 3rd party implementations, or even break their own old versions. MySQL can do the same, because since there is no standard document for the protocol, nobody can file a bug report regarding a 3rd party implementation.
Can you please provide a link to the open, published specification? I'd like to see it so that I know that anyone can implement that spec under whatever license they feel most comfortable with.
The MySQL wire protocol is proprietary. You would have to reverse engineer the product to create code under a different license. They did that on purpose.
That is someone who read the source files and produced a spec for a hypothetical re-implementation. The problem with that is that MySQL has made no promises to adhere to that spec, so it's quite possible that MySQL could manipulate the protocol to break 3rd party implementations without breaking their own. And nobody would have a standard document that they could use to file a bug report against. That's the definition of a proprietary protocol.
It is not a fact that there are too many people. Saying that the world would be better off one way or another is highly opinionated. Who judges whether the world is better or worse?
The way "environmentalists" get that much money is because there is usually some business interest behind it. The more regulations there are, and the more hoops people have to jump through to get work done, the worse for small business and the better for large business. Most people miss the last point, that large businesses and government go quite well together. And an environmental issue is an easy way for the large businesses to summon the powers of government to do their bidding.
Especially for filesystems that have relatively low usage, this method is megafast because it actually mirrors only the contents of the filesystem
Not only that, but due to zfs's design, if there is a power failure the raid does not need to be resync'd, nor does it need any kind of nvram. It really removes a lot of the headaches of raid.
There is a lot of work going on to make Solaris and Linux work well together. You can use zones to set up linux on a Solaris box, but I'm a little unclear on the details. I'll be checking that out soon though.
I think the porting of ZFS would be reasonable to do, but I'm sure it would take some work.
DTrace seems like it would require a LOT of work. All the work of DTrace was not the userspace application, but all of the hooks added into the OS at every level. So basically it would be repeating all the work on Linux rather than merely a port.
The main thing to take note of is that Sun is actually innovating. There's a lot of actual research being done on the OS kernel itself, which not too many companies are taking seriously. Linux is great, but I don't think it's going to make Solaris obsolete any time soon. There are also some major differences in overall design philosophy.
I suggest you actually look at Solaris. There are some amazing capabilities in there way ahead of Linux.
ZFS is a filesystem that can do raid5-like storage or mirrors. Filesystems can share a common storage pool. You can make snapshots instantly, and at any time you want you can roll back to that snapshot (transactions). Everything about it is very cool, check it out.
DTrace is also amazing. You can observe almost anything about a running program with negligable performance impact. It will break the information down for you statistically so you can tell that, for example, 1% of the time a given function call takes 1000 times longer than average.
It's also got containers and zones, and a service manager.
I have been using Linux and FreeBSD for a long time. I am just getting interested in Solaris, and I am very impressed.
You make some valid legal points. But I'm not sure how many companies really emerge from antitrust lawsuits with any real economic benefit. If I see AMD get some money in the next year or two, and then use it to make amazing products that blow away the market, you (and AMD) will be vindicated. Otherwise, not.
Like I said, sometimes a lawsuit is necessary. I'm not really qualified to start breaking down specific medical malpractice lawsuits. I'm sure there are valid and frivilous ones just like everything else. From what I understand now, malpractice insurance is a huge cost for many physicians. Some types of doctors pay $100k/yr for malpractice insurance just because of the type of doctor they are and the state they practice in.
I guess that's fine, but then you can't complain about a $250 bill to talk to the doctor for 10 minutes (that actually happened the last time I saw the doctor, by the way). To me, it seems more economically efficient to talk to 2-4 cheap doctors than one super-expensive one.
The result is that now most people get medical advice from family members and rarely see the doctor. What do you think? I'm not asking rhetorically, I want to know. You probably know more than I do. Perhaps it might be better to limit liability, thus reducing the cost of a doctor's visit, thus allowing patients more frequent visits (or phone consultations or even house calls), and second (or 3rd+) opinions.
Slashdot Mirror
The P's and R (Ruby) are ran by an interpreter, which essentially interprets the text into machine language.
That comment is misleading. The text is never translated to machine code. The interpreter reads the code and the interpreter does the work according to the rules of the language. Java takes an intermediate step and translates the code to bytecode, which can be more easily interpreted. However, so can python.
Also, your performance analysis is lacking. Code size is a big factor here. A perl program may, in a few short lines of code, accomplish something that would take a lot more java code. If you have a complex text processing program, perl may well be much faster. That's because the few short lines of perl code call a lot of highly optimized C text processing routines. The java code may do more of the work in Java code.
So, when evaluating the speed of a language, you should ask yourself: "Does parsing take a significant amount of time for my application?". Java may be fast, but if it doesn't call into the optimized C routines quickly enough, it's going to lose to players like perl, python, and ruby.
In other news, I learned two new words today :)
http://m-w.com/dictionary/putatively
http://m-w.com/dictionary/pejorative
Not that I've never heard those words before, but to really understand a word it's good to look it up on more than one occasion.
Go to LA. Go see the communities that live under the shadow of the gun.
Just out of curiosity, have you been to LA? I have, many times. I'll be driving through there this weekend, and I'll probably do a few things in the city. I'm not saying that I go through the most dangerous areas at night or anything. I just thought I'd mention it. Last time I was in LA my friend was telling me that he was shot in the ass once (he wasn't doing anything, just getting home at night or something).
LA has many problems. The police do not protect the poor communities, leaving them vulnerable to thugs. Most of these thugs posses weapons illegally.
There are between 200 and 250 million guns in the U.S. It would be impossible to reduce that number very much, which means that guns are available to criminals. Weapons in general aren't very hard to make, nor are they hard to smuggle. The only question remains whether guns are available to law-abiding citizens.
Our murder rate is 20 times lower than the US.
That's wonderful. You're willing to pay the price of civil liberty for a decrease in the murder rate. Murders overall are a very low percentage of deaths in the U.S. I value my civil liberties much more than that. And I really don't think there's any evidence that the rate of violent crime would drop by a factor of 95% if the government outlawed guns in the U.S.
I would just like to respond to the "libertarian ideals" comment. To me, libertarianism is about limited government. Not limited in size, but limited in function. A military is a necessary part of the government, and so is a police force. A responsible government would ensure that both of those government elements are strong enough to do their job (part of which is protecting private property). In Liberia and Rwanda, that is clearly not the case, and private property is not protected.
Another element of libertarianism is to break the government into smaller pieces, i.e. states. The states should have most of the government powers, and the federal government should mainly be used to ensure that rights are protected and the country is secured by the military.
You owning a gun is not going to make any criminal decide to stay clean instead.
That statement is suspect on the face of it. Guns owned by police clearly do deter criminal activity, so why not private citizens? But regardless, that was not my point. My point is that gun ownership is a civil right, not dependent on their ability to deter crime overall.
You wouldn't let anyone use statistics to take away our freedom of speech, I ask the same for freedom to own guns. There are bound to be situations in which gun control can lower violent crime, and others in which gun control raises violent crime. This is dependent on many factors, including the particular point in time in question.
When circumstances change and then all of a suddent you want your right to own guns back, it's too late.
Guns absolutely ARE a deterrent to oppressive government. You can see in wars from Vietnam to Iraq that cruise missles or battleships are not efficient weapons against small bands of people with guns.
Not only that, but the government should not have a monopoly on protecting me. I should be able to purchase locks, alarms, cameras, and guns; and I should be able to hire private detectives and body guards.
And as one final note, the Constitution clearly includes gun ownership in the Bill of Rights. I ask that you give the Second Amendment the same protection as the rest of the Constitution. We have enumerated rights, and it is important that we not allow the government to infringe on any of those rights, or else the government will easily be able to divide and conquer our rights. If you disagree with the Second Amendment, you can attempt to make another Amendment, and if that is really what the people want, it will pass.
Gun ownership is a civil right.
You could make the argument that censorship reduces the incidence of offensive language, but that's a small benefit to receive in exchange for a civil right.
It's very difficult to really tell conclusively what long term effect gun control has on the incidence of violent crime in general. But regardless, I stand for the civil right for law-abiding citizens to produce, own, transport, and trade firearms. I believe that this right is an important deterrent for oppressive government.
Very interesting point. If you have a 16 way box, and intend to share it among many applications, each of those applications would need to be licensed for 16 processors. Is that what you're saying?
I hope so. I love FreeBSD!
I've used Linux for about 6 years, and FreeBSD for about 5. However, I didn't heavily use FreeBSD until a year ago. What really amazes me is that I can try new operating systems that are all UNIX-like, and yet still each has aspects that really impress me.
Does it have to do with export regulations?
I downloaded a couple sets of CDs from sun. Downloading a few GB over DSL, the registration time was definately not the major source of delay in the download process. It is more annoying than downloading FreeBSD, but didn't seem too difficult.
What does Sun need to do to succeed?
If their new line of processors is successful, that will go a long way. From what I understand, Solaris is gaining interest very quickly, and with it, Sun mindshare.
If they have a leading chip for common server loads, and a leading OS for common server tasks, the support contracts and hardware sales are bound to roll in.
It's sad that Sun is looking at so much financial trouble right now (assuming they actually are, and it might not actually be that bad). Solaris 10 is really quite amazing software, and their chip designs look very promising. I think it would take Linux/FreeBSD a long time to catch up to some of the things in Solaris 10 like ZFS and DTrace.
I'm really pretty new to Solaris, however as soon as I started using it I could tell that they did real research on the kernel. I was very impressed after using Linux/FreeBSD for a long time. I can't think of any other companies actually doing that kind of research still. IBM does a lot of Linux development, but I sometimes get the feeling it's more to just make Linux into an AIX replacement (not that I know much about AIX).
As usual, MySQL is intentionally being confusing about their licensing. You can't GPL a protocol, only the implementation of a protocol.
However, the protocol itself still every bit as proprietary a format as an MS Word document. It just happens that they license their implementation to you under the GPL. You can reverse engineer it, just like the MS word doc format, but MySQL AB can always make subtle changes to break it (just like MS can do to the word format).
It's proprietary, and you can reverse engineer it.
You could use the reverse engineering argument for MS Word documents as well. However, Microsoft can make subtle changes to the format designed to break 3rd party implementations, or even break their own old versions. MySQL can do the same, because since there is no standard document for the protocol, nobody can file a bug report regarding a 3rd party implementation.
Can you please provide a link to the open, published specification? I'd like to see it so that I know that anyone can implement that spec under whatever license they feel most comfortable with.
/ 12572/OpenDocument-v1.0-os.pdfl .html
c ol.html
An example of a truly open standard is:
http://www.oasis-open.org/committees/download.php
Or also, the PostgreSQL wire protocol:
http://www.postgresql.org/docs/8.1/static/protoco
The MySQL wire protocol is proprietary. You would have to reverse engineer the product to create code under a different license. They did that on purpose.
Note, there are documents in existence, such as:
http://www.redferni.uklinux.net/mysql/MySQL-Proto
That is someone who read the source files and produced a spec for a hypothetical re-implementation. The problem with that is that MySQL has made no promises to adhere to that spec, so it's quite possible that MySQL could manipulate the protocol to break 3rd party implementations without breaking their own. And nobody would have a standard document that they could use to file a bug report against. That's the definition of a proprietary protocol.
It is not a fact that there are too many people. Saying that the world would be better off one way or another is highly opinionated. Who judges whether the world is better or worse?
The way "environmentalists" get that much money is because there is usually some business interest behind it. The more regulations there are, and the more hoops people have to jump through to get work done, the worse for small business and the better for large business. Most people miss the last point, that large businesses and government go quite well together. And an environmental issue is an easy way for the large businesses to summon the powers of government to do their bidding.
Working code is a lot more than "lots of promises"
Yes, their home-made opteron CPUs are a great example of how Sun does hardware....
Sun has some interesting technology in their UltraSPARC T1 processor. 32 threads of simultaneous execution and only 72 watts.
Especially for filesystems that have relatively low usage, this method is megafast because it actually mirrors only the contents of the filesystem
Not only that, but due to zfs's design, if there is a power failure the raid does not need to be resync'd, nor does it need any kind of nvram. It really removes a lot of the headaches of raid.
There is a lot of work going on to make Solaris and Linux work well together. You can use zones to set up linux on a Solaris box, but I'm a little unclear on the details. I'll be checking that out soon though.
I think the porting of ZFS would be reasonable to do, but I'm sure it would take some work.
DTrace seems like it would require a LOT of work. All the work of DTrace was not the userspace application, but all of the hooks added into the OS at every level. So basically it would be repeating all the work on Linux rather than merely a port.
The main thing to take note of is that Sun is actually innovating. There's a lot of actual research being done on the OS kernel itself, which not too many companies are taking seriously. Linux is great, but I don't think it's going to make Solaris obsolete any time soon. There are also some major differences in overall design philosophy.
I suggest you actually look at Solaris. There are some amazing capabilities in there way ahead of Linux.
ZFS is a filesystem that can do raid5-like storage or mirrors. Filesystems can share a common storage pool. You can make snapshots instantly, and at any time you want you can roll back to that snapshot (transactions). Everything about it is very cool, check it out.
DTrace is also amazing. You can observe almost anything about a running program with negligable performance impact. It will break the information down for you statistically so you can tell that, for example, 1% of the time a given function call takes 1000 times longer than average.
It's also got containers and zones, and a service manager.
I have been using Linux and FreeBSD for a long time. I am just getting interested in Solaris, and I am very impressed.
Don't worry about getting your money back. It will be a FreeVee.
("Running Man" reference)
You make some valid legal points. But I'm not sure how many companies really emerge from antitrust lawsuits with any real economic benefit. If I see AMD get some money in the next year or two, and then use it to make amazing products that blow away the market, you (and AMD) will be vindicated. Otherwise, not.
Like I said, sometimes a lawsuit is necessary. I'm not really qualified to start breaking down specific medical malpractice lawsuits. I'm sure there are valid and frivilous ones just like everything else. From what I understand now, malpractice insurance is a huge cost for many physicians. Some types of doctors pay $100k/yr for malpractice insurance just because of the type of doctor they are and the state they practice in.
I guess that's fine, but then you can't complain about a $250 bill to talk to the doctor for 10 minutes (that actually happened the last time I saw the doctor, by the way). To me, it seems more economically efficient to talk to 2-4 cheap doctors than one super-expensive one.
The result is that now most people get medical advice from family members and rarely see the doctor. What do you think? I'm not asking rhetorically, I want to know. You probably know more than I do. Perhaps it might be better to limit liability, thus reducing the cost of a doctor's visit, thus allowing patients more frequent visits (or phone consultations or even house calls), and second (or 3rd+) opinions.