I'm sure you'd know better than my anecdotal experience.
In my world Novell Netware ran this space until post NT 4 and NT 4 wasn't released until 1996. By the time we were seeing and recommending Linux/Samba combinations as replacements Linux had already established itself running web boxes. But then I wasn't working in the enterprise back then. Most of the companies we did work for had under 50 seats.
Very interesting stuff. I didn't know about Samba's history on commercial UNIX. But that's all ancient history anyway.
Over the years I have to say I've found Samba to be my tool of choice in many roles and it is an absolutely indispensable part of my toolkit. I haven't had a chance to put Samba 4 through the paces yet but some of the new capabilities suggest that I might be able to eliminate windows entirely from the piece of network I'm working in now. So thank you Jeremy for your role in all of that!
Jeremy I have no doubt you know when Samba was started but started and widely adopted are two different things.
I certainly saw no sign of serious samba deployment until after serious Linux deployment and in most organization that was after the Linux boxes were first deployed (at least deployments that stuck) in roles that were previously owned by UNIX such as web/ftp/dns and for a while email but exchange ended up winning email. It wasn't until after a company already had Linux working and proven as a reliable source of a cheap dust collector that just works before they would risk more directly desktop interfacing tasks to it like file/print and auth.
Yup outlook/exchange. I've never understood what is so great a technical challenge in developing open solutions that replace these well but when something pops up it seems to focus on only one piece of this or it provides calendar and email but as separate pieces within the suite and not tying and integrating all the pieces together the way exchange does.
I'm sorry but you simply don't speak for what is the case in corporate america. The only place you see windows or samba needing to replace it is on desktops, file and print servers and often authentication ties (so as to support single sign-on) and exchange. In most of those cases it is only used because it easy and popular. And thanks to Samba 4 you could drop windows from all of that but exchange and the desktops. At the end of the day the main reason you use windows/samba for print/auth is that you already have it for your desktop+outlook combo.
This all represents a very small part of the corporate server world. You aren't going to find windows boxes serving webpages or anywhere in the data center. Most applications (especially the internal apps that drive a workplace) are going to be running on Linux and served up through a web interface.
Perhaps it is just me. When you say 'corporate america' my mind goes to at least Fortune 1000. I suppose technically you could include all the small businesses that don't really have an infrastructure beyond desktop/file/print/exchange but are legally structured as a s or c corp.
The boss is interested in what his boss is interested in. He might be interested in this in the beginning but at the end of the day his bosses aren't going to appreciate it because it is going to look like his team outputs less.
More Ad Hominem attacks. Are you capable of rational discourse?
"so one scientist and its case solved for you"
First of all I listed three and you just listed two although you gave a quote of something one of them said about one expedition he went on because he had integrity which only lends credibility to his other work. Then you launched into an Ad Hominem attack on the other. You do know it is a logical fallacy to discount a message by attacking the messenger rather than the content? I note you didn't have any material to attack Jane Goodall with so you skipped her.
But no, my argument isn't for the existence of Bigfoot but for the application of the same standards used to scrutinize other suspect but not yet confirmed species. You are the only one even concerned with solving the case or as you see it, who has already solved the case.
Hunter said he shot one but didn't bring back a sample. Okay, dismissed. It doesn't provide data either way.
"And playing you all for fools as you lap it up."
In other words you have nothing of substance to contribute to a discussion. Just Ad Hominem attacks.
At the end of the day I have to admit that I'm no primate expert. I highly doubt you are either. If only there were a highly credible primate expert who could end our discussion by clearly being so much better qualified that neither of us is qualified enough to argue against her opinion.
Like this chick http://en.wikipedia.org/wiki/Jane_Goodall. As an expert on primate and especially chimps she was extremely well versed in primate communications and meaningful sounds. Very much unlike the Native Americans who could so effectively replicate the calls of Sasquatch that they sold her. But what does she know. I mean I'm sure if I'd walked up to her and said I saw a primate that went oh oh ah ah she'd have been totally sold.
How about this guy? http://en.wikipedia.org/wiki/Jeffrey_Meldrum. This nutter isn't even a primate expert! Just some nut with credentials in zoology, vertebrae locomotion, and anatomical sciences. He is about the only one ever to get an article related Sasquatch through peer review. Turns out, if skeptics label something as "extraordinary" peers tend to reject articles related to it without consideration of substance. Those same peers will point to a lack of a journal published peer reviewed paper to discredit anyone who claims they have evidence to disclose though.
But why so few even if they are credible? Well considering it tantamount to career suicide to claim evidence of bigfoot, yeti, etc it isn't surprising. But more than that it turns out there really aren't all that many esteemed primatologists in the first place. There are so few that it isn't especially meaningful if one or two believe something and the others don't because there aren't enough others for one or two to not represent a substantial sample.
Why? A reclusive and intelligent species with a low population in a massive mostly unpopulated forest. I don't see any particular reason there should be. There is DNA and video. Hell we didn't have video of a giant squid until recently.
Possibly but at this point skepticism around Bigfoot is so high that DNA taken from a sample that matches no known species does not qualify as credible evidence. Such DNA has been found before, is consistent with what one would expect Bigfoots DNA to look like and the results are labeled inconclusive and dismissed as contamination by skeptics.
At this point nothing short of catching a live specimen will suffice. And even that would likely only be enough to kick off years of testing and analysis before there could be any sort of consensus. At the end of which, there would be a camp claiming that it still wasn't Bigfoot because some detail propagated in the reports about the creature was wrong.
I am all for constructive skepticism. It is a necessary and practical tool. But when the bar starts getting raised over and over again under the tired old phrase 'extraordinary claims require extraordinary evidence' a problem gets introduced. You see nobody wants to clearly delineate exactly what qualifies as extraordinary. Is the idea of a yet unknown reclusive primate living in a vast forested area really that extraordinary? Or a small population of large sea creatures living in a loch and its underwater cave system? I propose a new counter standard. "Labeling something extraordinary requires an extraordinary lack of hypothetical probability." Perpetual motion meets that standard. Bigfoot does not.
I'm not saying you take evidence for granted and fail to scrutinize it. That correlates to the number of false reports and data and there are plenty of those in Bigfoot land. But the standards by which it is authenticated should be no different than for any other unknown species.
This was about 2 years ago. The job I mentioned the story worked out well. I've seen multiple promotions and numerous pay increases in that 2 years and am actually living comfortably again and paying plenty of taxes.
For the impoverished who aren't homeless and/or mentally ill a website should be accessible enough. They can use the local library. I see the homeless at the library so they must have provisions for it but every public library I've been to has wanted a couple pieces of evidence that I live inside their tax district. Maybe they are getting on the roster at a homeless shelter or something and using that address. Another example of people trying to 'prevent abuse' throwing up red tape hurdles for those in need to cross and forcing them to learn to work the system.
You don't have to starve to be hungry. You can get a day with many generous people at McDonalds and have 5 without and still be fat. You might not be dying but you will be hungry an awful lot.
I doubt they will ever be addressed under our system. There is too much tug of war to avoid paying for assistance programs. As a consequence those who don't want to pay for them make sure there are enough roadblocks that take the form of checks against abuse or limitations to who qualifies that many people can't get assistance or get too little or can't apply it where they need it. The benefits are also very low.
My S/O and I experienced a term where I was unemployed. At the worst of it we were in a long term stay studio. My S/O is on disability and essentially that entire income went to paying just enough to keep a phone from being turned off and paying to stay in the studio. At that time work wasn't obtainable, not even part time retail or fast food work let alone things in my field. That left us with no money for food. In order to get benefits I essentially had to lie and say we ate and prepared meals separately and were separate households. My S/O and I then each had benefits although the S/O got almost nothing in comparison and that still left us scrounging at the end of the month and yes we would still have a few days without food.
Even with that I was only allowed benefits for three months without finding employment. It didn't matter if I looked and had documentation to prove it. The final month I finally got a job and got the first paycheck just in time for the Food stamp cutoff. We got very lucky, there was no way I could get the job on public transportation and they were raising the cost of our room which we obviously couldn't have done. Luckily we found another rat apartment that required no deposit and was two blocks from the new place of employment, got it with rent just low enough to cover UHaul if we counted the miles very carefully, and got the job which paid me just about the time we'd run out of food.
Was this some great effort that was the result of the pressure to get a job from the benefits? Nope. I was already actively looking both for decent and p/t lousy jobs during that entire period so there was nothing more to do. It was sheer dumb luck. Actually I happened to get two offers in the same week.
I spend 2.5-3x our combined monthly benefit after perjury on a weekly basis for food to cook at home. What I spend is too much but the benefit is way way too low. And there isn't really assistance for housing (it exists but the vouchers are in extremely limited supply and there is a 6+ month waiting list to be rejected). There are no more cash benefits so those are out. Medical benefits, you can forget it unless you have a child. And cutting off people's benefits if they can't find a job is inexcusable. In fact benefits should continue for a couple months after you find full time work that causes you to exceed the benefit threshold. If you get or have a job that pays less than the benefit threshold you should get full benefits, it should be all or none. You shouldn't punish people for finding employment that is just keeping them in a hole of poverty. We also need utility assistance, not one or two bills in a program that is underfunded and only applies in the worst of winter or some such either but actual full utility assistance.
Beer and Hammers being made of atoms does equate to beer being made of hammers.
That was sort of the point. His arguments amounted to pointing out some random correlation or common trait and then jumping to conclusion that isn't even suggested by the correlation or common trait. The article at least makes an argument that follows the correlation they've found and the argument being mocked was merely pointing out there is additional support beyond that correlation.
At the end of the day the only thing we KNOW are that there are patterns and we've labeled things and built models around them. A pattern is nothing but a group of correlations. Perform an experiment and the result is a correlation to the experiment. Repeat it again, and it remains a correlation. In other words, at some point ALL evidence is nothing but correlation.
You seem to be talking about convicting beyond a reasonable doubt which is a much higher standard than I look for before I'm to suggest something might be a cause. I'm pretty sure the threshold for suggesting something that logically follows the apparent evidence may be case is no higher than that there be some apparent evidence.
You also missed the point. If lead exposure was reduced and reduced violence alone occurred you would merely have correlation as you suggest. But the argument I was defending was not making that case. He was (correctly) arguing that there is already known link between violent tendencies and lead exposure.
None of that makes the post I responded to anything but a host of rhetoric tossed at the gp because with no evidence (correlation or otherwise) is skeptical and in an emotional I resort to rhetoric sort of way not a productive let me provide constructive criticism way.
Except you've made a leap transforming beers to hammers. Your premise do not actually support your argument (which is obviously intentional) and you suggest it is similar his argument and then proceed to beat that strawman down. You follow up with a false dichotomy suggesting that either your argument is valid or his cannot be valid.
The problem is that his argument is supported by his premise where yours is not.
is leap is that lead is proven to cause people to become violent, therefore it is reasonable that the documented decline in known sources of lead poisoning could be related to a reduction in violence. This logically follows and his premise is supported.
Bullets are known to cause death. Therefore it is reasonable to suggest that the bullets I'm firing into the crowd might be responsible for the dead people in the crowd.
Cannabis is known to get you high. Therefore it is reasonable to suggest that the marijuana found in the stoned teenagers posession might have been what he used to get high.
Now lets try yours:
Bullets cause death. Knives cause death. Therefore bullets are made of knives. Cats have claws. Dogs have claws. Therefore dogs are made of cats.
poor people didn't have any stuff 40-50 years ago and don't have stuff now. Poverty isn't the people with low incomes living in trailer parks and low income housing, They are the people pushing shopping carts on the streets with children in tow. Although neither of these groups has as much stuff as you imply and both spend time hungry with no food.
"The only thing a CPU can execute is machine code. That's what a JIT VM does, it turns the bytecode into machine code"
How is it that you think interpreted code gets executed without turning into the machine code that is the only thing the CPU can understand? Here is a hint. It doesn't. You can compile Perl, using the interpreter, into an intermediate platform independent bytecode that can be saved and executed (aka turned into machine code) on other platforms.
"My last experience with Perl was replacing a network monitoring system written in Perl with Java. The Perl version would fall over when presented with > 2 million events a day. The new Java code handled > 100 million with lower latency. How's that for an anecdote?"
I'll raise you anecdotes. My current experience with Perl is a network monitoring system written in Perl. The Perl version is currently handling well over a billion events a day with low latency. I can't speak as to its upper limit.
What does a JIT VM have to do with anything? But yes, actually there is, or something that gives the same benefits. It's called an interpreter. Internally the Perl 5 interpreter compiles to an intermediate machine independent bytecode which it then executes. You can keep that bytecode and run later just like you do with java. Outside of a few specially crafted research problems designed to show performance increases with JIT there is no reason to desire it anyway.
There are plenty of major applications written in both Perl and Java. I can't think of the last time I noticed the performance of a Perl app. Java on the other hand is a constant nightmare. Of course there is OO.org but lately I've been playing with Minecraft and Minecraft servers. OMG what a slow memory hog that thing is! There is literally no reason for it to be either. Write the same thing in C and the memory footprint would be megs and not gigs and it would run smoothly on old low end pc's and not just high end ones. I can't think of any instance where I was impressed with the speed of a java app only cases where they didn't get in the way. Another example, look at all things android. What was Google thinking basing it on Java when they have limited cpu and memory resources?
3 times slower is hardly "on par." I tend to define real world as actual large complex java projects in the real world. So things like OO or Minecraft or any of the other large complex projects put out for public consumption. I can think of java programs where speed didn't get in the way but I can't really think of anything in java that is fast and efficient.
You: "Unlike static compilers JIT compiler can gather runtime information and use it when deciding which branch to prefer etc"
Me in the post you replied to:
"there is nothing stopping you from building an application specific tuned bytecode with dynamic optimization in C that will perform as fast or faster than the generic one used by java"
There are no advantages to a JIT. If you actually have one of the rare use cases where it provides a significant benefit you can build a task specific solution and a parser in you statically compiled language of choice to add a dynamic capability. At the end of the day the JVM is written in a statically compiled language. It is literally impossible for it have any sort of capability that can't be implemented in a statically compiled language.
Yup, despite it being broken as written and his complete lack of commenting.
That was about as ugly as Perl gets. The ugliest Perl gets is routines involving complex data structures. A hash full of hashes filled with arrays which contain and element that is a hash of hashes for example. Perl objects aren't great (it isn't really and OO language) but in most cases they can replace these type of ugly constructs with something that is not only easier to read but easier for the programmer to follow himself.
Other than that, there are a handful of variables that Perl uses or populates on a regular basis. These wouldn't be obvious to a non-perl programmer.
Slashdot Mirror
I'm sure you'd know better than my anecdotal experience.
In my world Novell Netware ran this space until post NT 4 and NT 4 wasn't released until 1996. By the time we were seeing and recommending Linux/Samba combinations as replacements Linux had already established itself running web boxes. But then I wasn't working in the enterprise back then. Most of the companies we did work for had under 50 seats.
Very interesting stuff. I didn't know about Samba's history on commercial UNIX. But that's all ancient history anyway.
Over the years I have to say I've found Samba to be my tool of choice in many roles and it is an absolutely indispensable part of my toolkit. I haven't had a chance to put Samba 4 through the paces yet but some of the new capabilities suggest that I might be able to eliminate windows entirely from the piece of network I'm working in now. So thank you Jeremy for your role in all of that!
Jeremy I have no doubt you know when Samba was started but started and widely adopted are two different things.
I certainly saw no sign of serious samba deployment until after serious Linux deployment and in most organization that was after the Linux boxes were first deployed (at least deployments that stuck) in roles that were previously owned by UNIX such as web/ftp/dns and for a while email but exchange ended up winning email. It wasn't until after a company already had Linux working and proven as a reliable source of a cheap dust collector that just works before they would risk more directly desktop interfacing tasks to it like file/print and auth.
Yup outlook/exchange. I've never understood what is so great a technical challenge in developing open solutions that replace these well but when something pops up it seems to focus on only one piece of this or it provides calendar and email but as separate pieces within the suite and not tying and integrating all the pieces together the way exchange does.
I'm sorry but you simply don't speak for what is the case in corporate america. The only place you see windows or samba needing to replace it is on desktops, file and print servers and often authentication ties (so as to support single sign-on) and exchange. In most of those cases it is only used because it easy and popular. And thanks to Samba 4 you could drop windows from all of that but exchange and the desktops. At the end of the day the main reason you use windows/samba for print/auth is that you already have it for your desktop+outlook combo.
This all represents a very small part of the corporate server world. You aren't going to find windows boxes serving webpages or anywhere in the data center. Most applications (especially the internal apps that drive a workplace) are going to be running on Linux and served up through a web interface.
Perhaps it is just me. When you say 'corporate america' my mind goes to at least Fortune 1000. I suppose technically you could include all the small businesses that don't really have an infrastructure beyond desktop/file/print/exchange but are legally structured as a s or c corp.
The boss is interested in what his boss is interested in. He might be interested in this in the beginning but at the end of the day his bosses aren't going to appreciate it because it is going to look like his team outputs less.
More Ad Hominem attacks. Are you capable of rational discourse?
"so one scientist and its case solved for you"
First of all I listed three and you just listed two although you gave a quote of something one of them said about one expedition he went on because he had integrity which only lends credibility to his other work. Then you launched into an Ad Hominem attack on the other. You do know it is a logical fallacy to discount a message by attacking the messenger rather than the content? I note you didn't have any material to attack Jane Goodall with so you skipped her.
But no, my argument isn't for the existence of Bigfoot but for the application of the same standards used to scrutinize other suspect but not yet confirmed species. You are the only one even concerned with solving the case or as you see it, who has already solved the case.
"DODGY videos done by very DODGY people."
http://en.wikipedia.org/wiki/Ad_hominem
Hunter said he shot one but didn't bring back a sample. Okay, dismissed. It doesn't provide data either way.
"And playing you all for fools as you lap it up."
In other words you have nothing of substance to contribute to a discussion. Just Ad Hominem attacks.
At the end of the day I have to admit that I'm no primate expert. I highly doubt you are either. If only there were a highly credible primate expert who could end our discussion by clearly being so much better qualified that neither of us is qualified enough to argue against her opinion.
Like this chick http://en.wikipedia.org/wiki/Jane_Goodall. As an expert on primate and especially chimps she was extremely well versed in primate communications and meaningful sounds. Very much unlike the Native Americans who could so effectively replicate the calls of Sasquatch that they sold her. But what does she know. I mean I'm sure if I'd walked up to her and said I saw a primate that went oh oh ah ah she'd have been totally sold.
Or maybe this nut http://en.wikipedia.org/wiki/Grover_Krantz. I mean sure dude is an expert in evolutionary anthropology and primatology but what does that even mean?
How about this guy? http://en.wikipedia.org/wiki/Jeffrey_Meldrum. This nutter isn't even a primate expert! Just some nut with credentials in zoology, vertebrae locomotion, and anatomical sciences. He is about the only one ever to get an article related Sasquatch through peer review. Turns out, if skeptics label something as "extraordinary" peers tend to reject articles related to it without consideration of substance. Those same peers will point to a lack of a journal published peer reviewed paper to discredit anyone who claims they have evidence to disclose though.
But why so few even if they are credible? Well considering it tantamount to career suicide to claim evidence of bigfoot, yeti, etc it isn't surprising. But more than that it turns out there really aren't all that many esteemed primatologists in the first place. There are so few that it isn't especially meaningful if one or two believe something and the others don't because there aren't enough others for one or two to not represent a substantial sample.
Why? A reclusive and intelligent species with a low population in a massive mostly unpopulated forest. I don't see any particular reason there should be. There is DNA and video. Hell we didn't have video of a giant squid until recently.
Bah, any fly DNA discovered with this technique would likely be labeled contamination so we probably won't find new flies this way.
Possibly but at this point skepticism around Bigfoot is so high that DNA taken from a sample that matches no known species does not qualify as credible evidence. Such DNA has been found before, is consistent with what one would expect Bigfoots DNA to look like and the results are labeled inconclusive and dismissed as contamination by skeptics.
At this point nothing short of catching a live specimen will suffice. And even that would likely only be enough to kick off years of testing and analysis before there could be any sort of consensus. At the end of which, there would be a camp claiming that it still wasn't Bigfoot because some detail propagated in the reports about the creature was wrong.
I am all for constructive skepticism. It is a necessary and practical tool. But when the bar starts getting raised over and over again under the tired old phrase 'extraordinary claims require extraordinary evidence' a problem gets introduced. You see nobody wants to clearly delineate exactly what qualifies as extraordinary. Is the idea of a yet unknown reclusive primate living in a vast forested area really that extraordinary? Or a small population of large sea creatures living in a loch and its underwater cave system? I propose a new counter standard. "Labeling something extraordinary requires an extraordinary lack of hypothetical probability." Perpetual motion meets that standard. Bigfoot does not.
I'm not saying you take evidence for granted and fail to scrutinize it. That correlates to the number of false reports and data and there are plenty of those in Bigfoot land. But the standards by which it is authenticated should be no different than for any other unknown species.
That would be a waste of perfectly good beer! *punches you in the upper arm* Party foul!
Wait... what kind of beer are we talking about?
This was about 2 years ago. The job I mentioned the story worked out well. I've seen multiple promotions and numerous pay increases in that 2 years and am actually living comfortably again and paying plenty of taxes.
For the impoverished who aren't homeless and/or mentally ill a website should be accessible enough. They can use the local library. I see the homeless at the library so they must have provisions for it but every public library I've been to has wanted a couple pieces of evidence that I live inside their tax district. Maybe they are getting on the roster at a homeless shelter or something and using that address. Another example of people trying to 'prevent abuse' throwing up red tape hurdles for those in need to cross and forcing them to learn to work the system.
You don't have to starve to be hungry. You can get a day with many generous people at McDonalds and have 5 without and still be fat. You might not be dying but you will be hungry an awful lot.
What does the census have to do with anything? You know the true poor in this country aren't even counted in the census.
I doubt they will ever be addressed under our system. There is too much tug of war to avoid paying for assistance programs. As a consequence those who don't want to pay for them make sure there are enough roadblocks that take the form of checks against abuse or limitations to who qualifies that many people can't get assistance or get too little or can't apply it where they need it. The benefits are also very low.
My S/O and I experienced a term where I was unemployed. At the worst of it we were in a long term stay studio. My S/O is on disability and essentially that entire income went to paying just enough to keep a phone from being turned off and paying to stay in the studio. At that time work wasn't obtainable, not even part time retail or fast food work let alone things in my field. That left us with no money for food. In order to get benefits I essentially had to lie and say we ate and prepared meals separately and were separate households. My S/O and I then each had benefits although the S/O got almost nothing in comparison and that still left us scrounging at the end of the month and yes we would still have a few days without food.
Even with that I was only allowed benefits for three months without finding employment. It didn't matter if I looked and had documentation to prove it. The final month I finally got a job and got the first paycheck just in time for the Food stamp cutoff. We got very lucky, there was no way I could get the job on public transportation and they were raising the cost of our room which we obviously couldn't have done. Luckily we found another rat apartment that required no deposit and was two blocks from the new place of employment, got it with rent just low enough to cover UHaul if we counted the miles very carefully, and got the job which paid me just about the time we'd run out of food.
Was this some great effort that was the result of the pressure to get a job from the benefits? Nope. I was already actively looking both for decent and p/t lousy jobs during that entire period so there was nothing more to do. It was sheer dumb luck. Actually I happened to get two offers in the same week.
I spend 2.5-3x our combined monthly benefit after perjury on a weekly basis for food to cook at home. What I spend is too much but the benefit is way way too low. And there isn't really assistance for housing (it exists but the vouchers are in extremely limited supply and there is a 6+ month waiting list to be rejected). There are no more cash benefits so those are out. Medical benefits, you can forget it unless you have a child. And cutting off people's benefits if they can't find a job is inexcusable. In fact benefits should continue for a couple months after you find full time work that causes you to exceed the benefit threshold. If you get or have a job that pays less than the benefit threshold you should get full benefits, it should be all or none. You shouldn't punish people for finding employment that is just keeping them in a hole of poverty. We also need utility assistance, not one or two bills in a program that is underfunded and only applies in the worst of winter or some such either but actual full utility assistance.
Beer and Hammers being made of atoms does equate to beer being made of hammers.
That was sort of the point. His arguments amounted to pointing out some random correlation or common trait and then jumping to conclusion that isn't even suggested by the correlation or common trait. The article at least makes an argument that follows the correlation they've found and the argument being mocked was merely pointing out there is additional support beyond that correlation.
At the end of the day the only thing we KNOW are that there are patterns and we've labeled things and built models around them. A pattern is nothing but a group of correlations. Perform an experiment and the result is a correlation to the experiment. Repeat it again, and it remains a correlation. In other words, at some point ALL evidence is nothing but correlation.
You seem to be talking about convicting beyond a reasonable doubt which is a much higher standard than I look for before I'm to suggest something might be a cause. I'm pretty sure the threshold for suggesting something that logically follows the apparent evidence may be case is no higher than that there be some apparent evidence.
You also missed the point. If lead exposure was reduced and reduced violence alone occurred you would merely have correlation as you suggest. But the argument I was defending was not making that case. He was (correctly) arguing that there is already known link between violent tendencies and lead exposure.
None of that makes the post I responded to anything but a host of rhetoric tossed at the gp because with no evidence (correlation or otherwise) is skeptical and in an emotional I resort to rhetoric sort of way not a productive let me provide constructive criticism way.
Except you've made a leap transforming beers to hammers. Your premise do not actually support your argument (which is obviously intentional) and you suggest it is similar his argument and then proceed to beat that strawman down. You follow up with a false dichotomy suggesting that either your argument is valid or his cannot be valid.
The problem is that his argument is supported by his premise where yours is not.
is leap is that lead is proven to cause people to become violent, therefore it is reasonable that the documented decline in known sources of lead poisoning could be related to a reduction in violence. This logically follows and his premise is supported.
Bullets are known to cause death. Therefore it is reasonable to suggest that the bullets I'm firing into the crowd might be responsible for the dead people in the crowd.
Cannabis is known to get you high. Therefore it is reasonable to suggest that the marijuana found in the stoned teenagers posession might have been what he used to get high.
Now lets try yours:
Bullets cause death. Knives cause death. Therefore bullets are made of knives.
Cats have claws. Dogs have claws. Therefore dogs are made of cats.
poor people didn't have any stuff 40-50 years ago and don't have stuff now. Poverty isn't the people with low incomes living in trailer parks and low income housing, They are the people pushing shopping carts on the streets with children in tow. Although neither of these groups has as much stuff as you imply and both spend time hungry with no food.
hashes are not arrays. Unless you consider the alternative term 'associative array' to be more meaningful than the internal data representation.
"The only thing a CPU can execute is machine code. That's what a JIT VM does, it turns the bytecode into machine code"
How is it that you think interpreted code gets executed without turning into the machine code that is the only thing the CPU can understand? Here is a hint. It doesn't. You can compile Perl, using the interpreter, into an intermediate platform independent bytecode that can be saved and executed (aka turned into machine code) on other platforms.
"My last experience with Perl was replacing a network monitoring system written in Perl with Java. The Perl version would fall over when presented with > 2 million events a day. The new Java code handled > 100 million with lower latency. How's that for an anecdote?"
I'll raise you anecdotes. My current experience with Perl is a network monitoring system written in Perl. The Perl version is currently handling well over a billion events a day with low latency. I can't speak as to its upper limit.
What does a JIT VM have to do with anything? But yes, actually there is, or something that gives the same benefits. It's called an interpreter. Internally the Perl 5 interpreter compiles to an intermediate machine independent bytecode which it then executes. You can keep that bytecode and run later just like you do with java. Outside of a few specially crafted research problems designed to show performance increases with JIT there is no reason to desire it anyway.
There are plenty of major applications written in both Perl and Java. I can't think of the last time I noticed the performance of a Perl app. Java on the other hand is a constant nightmare. Of course there is OO.org but lately I've been playing with Minecraft and Minecraft servers. OMG what a slow memory hog that thing is! There is literally no reason for it to be either. Write the same thing in C and the memory footprint would be megs and not gigs and it would run smoothly on old low end pc's and not just high end ones. I can't think of any instance where I was impressed with the speed of a java app only cases where they didn't get in the way. Another example, look at all things android. What was Google thinking basing it on Java when they have limited cpu and memory resources?
3 times slower is hardly "on par." I tend to define real world as actual large complex java projects in the real world. So things like OO or Minecraft or any of the other large complex projects put out for public consumption. I can think of java programs where speed didn't get in the way but I can't really think of anything in java that is fast and efficient.
You:
"Unlike static compilers JIT compiler can gather runtime information and use it when deciding which branch to prefer etc"
Me in the post you replied to:
"there is nothing stopping you from building an application specific tuned bytecode with dynamic optimization in C that will perform as fast or faster than the generic one used by java"
There are no advantages to a JIT. If you actually have one of the rare use cases where it provides a significant benefit you can build a task specific solution and a parser in you statically compiled language of choice to add a dynamic capability. At the end of the day the JVM is written in a statically compiled language. It is literally impossible for it have any sort of capability that can't be implemented in a statically compiled language.
I'd have to disagree. I mean sure it is spaced out nicely but it isn't especially readable. It all just sort of blurs together.
Yup, despite it being broken as written and his complete lack of commenting.
That was about as ugly as Perl gets. The ugliest Perl gets is routines involving complex data structures. A hash full of hashes filled with arrays which contain and element that is a hash of hashes for example. Perl objects aren't great (it isn't really and OO language) but in most cases they can replace these type of ugly constructs with something that is not only easier to read but easier for the programmer to follow himself.
Other than that, there are a handful of variables that Perl uses or populates on a regular basis. These wouldn't be obvious to a non-perl programmer.