Is this an assumption you're making, or have the developers actually stated "expect 100-300% improvement in speed once we remove the debugging code"?
If this is just an assumption, it would take a lot of debug code to slow it down to where people complain about the speed.
Not to mention that if turning off debugging code really made a huge difference, don't you think many people would turn it off and recompile to see how well it ran without it? Have you seen any reports of this? I haven't, and this tells me that it's not debugging code that's the problem.
The last version out the door was very flaky, hideously slow, and not particularly complete, IMHO.
Uh, how is this "good, VERY good"?
And how can it STILL be slow? Didn't they learn anything from the garbage known as 4.6? I mean, I'm willing to cut them slack since it isn't finished, but that tells me that they didn't improve a lot of the algorithms that made 4.6 so brain damaged (e.g., table processing).
IE may not be perfect, but I hope you're not implying that the buggy, slow piece of crap known as Netscape is better. (Did I mention how slow and buggy it is?)
The simplest explanation is usually the correct one.
The brain is an immensely complicated instrument, but it seems to me that there is plenty enough complexity in the chemical-based neuron connections to eventually produce a good theory for how it works.
Why is this not sufficient? Why do many physicists feel that there "must" be a quantum component to it? Given that there is no quantum component to DNA (that we know of), which is the foundation of everything, I find it hard to believe that there would be one in the brain.
You are not a machine full of facts, you are a man full of opinions, so maybe you should restate that as...
Actually, I state it that way intentionally in order to challenge someone to name a client app that was better. I mean, what you say is true, but rather obvious, don't you think?
In fact, you may have noticed I just changed my sig to reflect this very question. A CVS client might or might not be better; I'm not sure. But my intention is to stimulate thought as to why Open Source has failed so miserably in the area of client apps, while it has been quite successful for server apps. Particularly here on Slashdot where Open Source dogma needs some shaking up.
In fact, some of the development apps are actually starting to catch up (read: not suck), which is not surprising since that would be an area that people would naturally want to improve upon.
So, just for you, I will restate the question as "normal user apps" rather than "client apps", which is what I really meant.
you forget that the Linux Zillion+one clients are also as good if not better...
But that's just it. Where is this mythical "better" one? But telnet is such a simple, puny program that it's really beside the point. Where is a major, "killer app" for normal users that is clearly better than Windows?
I could argue that telnet sucks in windows but I'm sure you could explain to me why it is the best.
I didn't say that every Microsoft application was the best; I said that there is no client application under Linux that is better than the equivalent one that runs under Windows. If you don't like the standard one, there are zillions of other ones.
Well, I guess I agree that twenty years might be a little excessive. Still, I think this crime is comparable to a lot of other white-collar crimes. Maybe five years.
I guess I'm just fed up with punks like this who feel they have to ruin everything for everyone. When I say "send a strong message", I mean not only the punishment, but these cases should be high-profile. "Interfere with the internet infrastructure == go to jail.".
I mean, what if we had roving bands of teenagers who went around shooting out tires on trucks on the freeway (let's say at low speed)? And it caused massive traffic jams because of all the stuck trucks. While they aren't technically killing anyone, they are costing a lot of wasted time and money, and should/would be delt with harshly. I see these fools in the same light.
Oh no, poor little Yahoo/CNN/Amazon/Buy.com might "lose" (in this context lose=not gain) a few million (like they'd miss it anyways).
Once again, corporations are not living entities. They are owned by real people. Have a 401K? Then you are possibly a shareholder in these companies, or one of the companies that do business with these companies.
Are you so blind from jealousy and envy of other's success that you can't see that real, average people are affected by events such as these?
Why should I take time out of my day for these guys? Besides, didn't your mommy teach you that its not nice to be a tattle-tale?
Ah, the David Cash philosophy. Watch your friend Jeremy Strohmeyer rape and kill a little girl, but it's none of your business.
Yeah, the FBI was a bunch of thugs for tracking down that guy.
Largely to get more money than anyone else infers that you have some very large advantage over others with similar levels of work.
In my experience, that is simply not true -- on balance. Does it happen? Of course; there will always be bad people in the world. But yes, on balance, those that work the hardest get the biggest rewards. I think where you get off track is in the definition of "hardest". Ditch diggers work very hard, but that doesn't mean they deserve to be millionaires. On the other hand, the president of a large multi-national corporation probably looks to a lot of people like he has a cushy job. However, what he has is the ability to manage a monster organization like that, and not many people can do it. That's an incredibly difficult job.
However the people who are in other countries are in fact largely there because of policies that their government's took in the past which essentially made their countries less avaible for advancement.
A surprisingly rational statement. However, it's the unequal distribution of capitalism that keeps their economies down. In other words, the lack of the corporations that you loath.
I really don't think that using the internet will ever supplant the traditional means of shopping at all.
Why does it have to be all-or-nothing with you? Even Jeff Bezos says that he doesn't think e-commerce will supplant bricks-and-mortor. But that doesn't mean it won't be huge, particularly for business-to-business. B2B will probably be larger than the consumer space, because that's where linking supply-chains really makes sense.
Incidently it is real hipocracy to think that corporations are evil and must be destroyed one minute and the next are the perfect angels of the universe the next. Which one is it? Make up your mind right here and now before you people do even more contradiction and say the Windows is the best and that the moon is composed of cheddar cheese.
Again, why does everything have to be all-or-nothing with you? Corporations are not living entities; they are owned by real people with real lives and real families. Are there evil people in the world that have abused workers or consumers? Of course. But so what? That's why we have laws. What does that have to do with the legal construction known as a corporation?
And by the way, Windows is the best. Of course, the rub is in the definition of "best". Most consumers define "best" as the platform that supports the most applications, which is where work gets done. And the client end-user applications under Windows are far superior to anything else, particularly Linux. Not one client application under Linux is superior to the ones in Windows. Not one.
Gnerally there are groups that I would think have a better chance to "fend for themselves" so to speak. I think we all could agree that Microsoft is not entitled to such protection because they most likely could easily hire their own private army of assassins to do so form of quasi-legal garbage and just might get away with it.
Uh, and exactly who is supposed to decide who gets protection under the law? Perhaps everyone who has over a certain amount of money should be just thrown in jail, since we know they couldn't have actually earned it. They must have stolen it by exploiting "average" citizens.
In fact, you've convinced me. By your standards, I think you're too rich to deserve protection under the law. I mean, it's pretty darn easy for you with your expensive computer, etc, when people are starving around the world.
Nope can't say that I use the internet on a daily basis to satisify my hunger for stuff. I have only bought on the internet 2 times for a total of 3 items and that was only because I couldn't very easily get what I wanted at a store (debian CDs).
So all that matters is what's important to you, I see. Yeah, that's a rational outlook.
I can't even now see that a large portion of money is actually being transfered online versus traditional methods I would love some hard data to back up your claims.
I already stated that it's "not a huge deal right now", but the time to nip it in the bud is when it's not a huge deal.
First of all, corporations are owned and run by citizens. And what exactly does "average" mean? Anyone not like you is automatically a non-citizen and not deserving of protection under the law?
Second, even if the attacks are against corporations not affiliated with you personally, others just might want to use the services they offer. Some of us even like the services they offer. Not to mention that attacks against them cause problems for sites in the general subnet vicinity (which might be some non-profit socialist site that you like).
Lastly, the FBI is involved because this is a very serious matter. It was an attack on the economic infrastructure. Maybe it's not a huge deal right now, but the net is becoming more and more important to the economy (particularly business-to-business services), and it's time to nip these idiots in the bud, and throw them in jail for twenty years to send a very strong message.
Stallman will think of a reason to take credit. Mozilla == Free Software == Related to source of JWZ's money == Stallman's nightclub. "Any Nightclub that has been built on the fruits of free software MUST BE called GNU/Nightclub!"
P.S. For the humor impaired: Yes, I know his fortune was built on Netscape's proprietary software... IT'S A JOKE.
A proven axiom that Intel mysteriously ignored with the Itanium is that backward compatibility with the existing software base is paramount for the success of a microprocessor. If the market did not have a choice, there is a possibility that, despite the Itanium's lack of backward compatibility, it still might succeed, but, with the availability of the Seldgehammer (and to a lesser degree the Athlon), Itanium's failure is almost guaranteed.
Huh? As far as I remember, Itanium has always had backward compatibility; it was supposed to have a Crusoe-like translator to change x86 instructions to VLIW instructions. Did this get tossed, or is Tom smoking something?
Before the buyout/merger, would Slashdot have posted anti-Slashdot articles at all? I don't particularly remember any, although Slashdot wasn't as high-profile back then, so external critical articles may not have existed.
Personally, I don't feel that Slashdot has any particular responsibility to put themselves on a whipping post to "prove" their independence.
Oracle on the other hand is as much as 14x slower than MySQL, and has serious performance issues with things like connecting to the database. You absolutely must use database connection caching to get acceptable performance. Not to mention the obscene pricing...
I agree on connections, and really agree on the price, but I have to say the 14x thing is pretty exaggerated. As usual, it depends on what you're doing, and I wouldn't be surprised if you can find a slew of operations that Oracle is 14x faster than MySQL (large databases come to mind).
The fact is that Oracle is pretty much the standard for very large web sites. I mean, I wouldn't use MySQL for a "real" web site where I cared about the data (like an e-commerce application), since MySQL doesn't support commit/rollback.
Even if we accept all your premises (such as a purely mechanistic universe), it doesn't imply that time would repeat in exactly the same way. It could be that the properties of the next bang depend on the end conditions of the previous crunch, rather than all properties being "reset" as you imply.
Uh... all the learning in the world is not going to change the fact that all the client applications under Linux are far inferior to the applications under Windows.
I don't know; maybe I'm weird but I just don't feel this emotional connection to Slashdot. It's a very vibrant community, I enjoy the discussions, but it's just a web site. To be honest with you, I think it's comments like yours that give people like Jon Katz an inflated sense of their own importance. Sites like this do have an effect, but it's an aggregated effect of all the various discussion sites. Slashdot is just one tiny piece of that effect.
And if Slashdot should die, another site will rise and take its place.
Hmmm; you are correct. I forgot they only GPL'd a very old version. I wonder if they stick with MySQL out of some sort of momentum, or if they would consider using Oracle?
Oracle is faster than you might think. You do have to architect it correctly, however, and that primarily means keeping connections open. I agree that starting the gears turning is a slow process, but once you've got them spinning, it can really move data.
Learn about styles. They are the reason you're having so much frustration with Word. I've seen this a lot when someone comes from Wordperfect (I did too).
Once you understand how Word works, you will find it far, far better than Wordperfect. It's extremely powerful.
You said that it would be "trival" to prove that GC is slower than hand management. If its trivial, it should be easy for someone of your broad knowledge of computer science to provide it. If you're going to make that kind of statement, the burden of proof is on you.
You're not paying attention. I gave you the proof. Someone with your awesome abilities should be able to blow it down in one whiff.
CMU floating point bench marks
LOL! A prime example of "lies, damn lies, and benchmarks". A trivial floating point benchmark proves absolutely nothing. Of course that type of problem is going lend itself well to compiler optimization. I could write similar things in APL, and it would probably be very fast, but that would say nothing about APL's viability for general problem domains.
You would be a bit more credible if you gave me a non-trivial program written in both languages. Say, a C++ compiler and optimizer. Or how about a Photoshop-like graphics manipulation tool. Or heck, how about a version of Quake in Lisp?
What is your theory on why people don't use Lisp for any non-trivial project? Is my apparent level of ignorance so rampant in the industry?
You must be in academia.
Quote from Harlequin: "By supporting a variety of manual policies and tuning, Harlequin delivers significantly better performance than can be achieved with other vendors' existing 'one size fits all' allocators."
Apparently they suffer under the same delusions that I do.
Slashdot Mirror
Is this an assumption you're making, or have the developers actually stated "expect 100-300% improvement in speed once we remove the debugging code"?
If this is just an assumption, it would take a lot of debug code to slow it down to where people complain about the speed.
Not to mention that if turning off debugging code really made a huge difference, don't you think many people would turn it off and recompile to see how well it ran without it? Have you seen any reports of this? I haven't, and this tells me that it's not debugging code that's the problem.
--
The last version out the door was very flaky, hideously slow, and not particularly complete, IMHO.
Uh, how is this "good, VERY good"?
And how can it STILL be slow? Didn't they learn anything from the garbage known as 4.6? I mean, I'm willing to cut them slack since it isn't finished, but that tells me that they didn't improve a lot of the algorithms that made 4.6 so brain damaged (e.g., table processing).
--
IE may not be perfect, but I hope you're not implying that the buggy, slow piece of crap known as Netscape is better. (Did I mention how slow and buggy it is?)
--
The simplest explanation is usually the correct one.
The brain is an immensely complicated instrument, but it seems to me that there is plenty enough complexity in the chemical-based neuron connections to eventually produce a good theory for how it works.
Why is this not sufficient? Why do many physicists feel that there "must" be a quantum component to it? Given that there is no quantum component to DNA (that we know of), which is the foundation of everything, I find it hard to believe that there would be one in the brain.
--
Actually, I state it that way intentionally in order to challenge someone to name a client app that was better. I mean, what you say is true, but rather obvious, don't you think?
In fact, you may have noticed I just changed my sig to reflect this very question. A CVS client might or might not be better; I'm not sure. But my intention is to stimulate thought as to why Open Source has failed so miserably in the area of client apps, while it has been quite successful for server apps. Particularly here on Slashdot where Open Source dogma needs some shaking up.
In fact, some of the development apps are actually starting to catch up (read: not suck), which is not surprising since that would be an area that people would naturally want to improve upon.
So, just for you, I will restate the question as "normal user apps" rather than "client apps", which is what I really meant.
But that's just it. Where is this mythical "better" one? But telnet is such a simple, puny program that it's really beside the point. Where is a major, "killer app" for normal users that is clearly better than Windows?
--
I didn't say that every Microsoft application was the best; I said that there is no client application under Linux that is better than the equivalent one that runs under Windows. If you don't like the standard one, there are zillions of other ones.
--
Well, I guess I agree that twenty years might be a little excessive. Still, I think this crime is comparable to a lot of other white-collar crimes. Maybe five years.
I guess I'm just fed up with punks like this who feel they have to ruin everything for everyone. When I say "send a strong message", I mean not only the punishment, but these cases should be high-profile. "Interfere with the internet infrastructure == go to jail.".
I mean, what if we had roving bands of teenagers who went around shooting out tires on trucks on the freeway (let's say at low speed)? And it caused massive traffic jams because of all the stuck trucks. While they aren't technically killing anyone, they are costing a lot of wasted time and money, and should/would be delt with harshly. I see these fools in the same light.
--
Once again, corporations are not living entities. They are owned by real people. Have a 401K? Then you are possibly a shareholder in these companies, or one of the companies that do business with these companies.
Are you so blind from jealousy and envy of other's success that you can't see that real, average people are affected by events such as these?
Ah, the David Cash philosophy. Watch your friend Jeremy Strohmeyer rape and kill a little girl, but it's none of your business.
Yeah, the FBI was a bunch of thugs for tracking down that guy.
--
In my experience, that is simply not true -- on balance. Does it happen? Of course; there will always be bad people in the world. But yes, on balance, those that work the hardest get the biggest rewards. I think where you get off track is in the definition of "hardest". Ditch diggers work very hard, but that doesn't mean they deserve to be millionaires. On the other hand, the president of a large multi-national corporation probably looks to a lot of people like he has a cushy job. However, what he has is the ability to manage a monster organization like that, and not many people can do it. That's an incredibly difficult job.
A surprisingly rational statement. However, it's the unequal distribution of capitalism that keeps their economies down. In other words, the lack of the corporations that you loath.
Why does it have to be all-or-nothing with you? Even Jeff Bezos says that he doesn't think e-commerce will supplant bricks-and-mortor. But that doesn't mean it won't be huge, particularly for business-to-business. B2B will probably be larger than the consumer space, because that's where linking supply-chains really makes sense.
Again, why does everything have to be all-or-nothing with you? Corporations are not living entities; they are owned by real people with real lives and real families. Are there evil people in the world that have abused workers or consumers? Of course. But so what? That's why we have laws. What does that have to do with the legal construction known as a corporation?
And by the way, Windows is the best. Of course, the rub is in the definition of "best". Most consumers define "best" as the platform that supports the most applications, which is where work gets done. And the client end-user applications under Windows are far superior to anything else, particularly Linux. Not one client application under Linux is superior to the ones in Windows. Not one.
--
Uh, and exactly who is supposed to decide who gets protection under the law? Perhaps everyone who has over a certain amount of money should be just thrown in jail, since we know they couldn't have actually earned it. They must have stolen it by exploiting "average" citizens.
In fact, you've convinced me. By your standards, I think you're too rich to deserve protection under the law. I mean, it's pretty darn easy for you with your expensive computer, etc, when people are starving around the world.
So all that matters is what's important to you, I see. Yeah, that's a rational outlook.
I already stated that it's "not a huge deal right now", but the time to nip it in the bud is when it's not a huge deal.
I think I've been probably been taken by a troll.
--
Pardon my flame, but what an idiotic question.
First of all, corporations are owned and run by citizens. And what exactly does "average" mean? Anyone not like you is automatically a non-citizen and not deserving of protection under the law?
Second, even if the attacks are against corporations not affiliated with you personally, others just might want to use the services they offer. Some of us even like the services they offer. Not to mention that attacks against them cause problems for sites in the general subnet vicinity (which might be some non-profit socialist site that you like).
Lastly, the FBI is involved because this is a very serious matter. It was an attack on the economic infrastructure. Maybe it's not a huge deal right now, but the net is becoming more and more important to the economy (particularly business-to-business services), and it's time to nip these idiots in the bud, and throw them in jail for twenty years to send a very strong message.
--
I don't know; I thought it was pretty funny.
Stallman will think of a reason to take credit. Mozilla == Free Software == Related to source of JWZ's money == Stallman's nightclub. "Any Nightclub that has been built on the fruits of free software MUST BE called GNU/Nightclub!"
P.S. For the humor impaired: Yes, I know his fortune was built on Netscape's proprietary software... IT'S A JOKE.
--
Tom says...
Huh? As far as I remember, Itanium has always had backward compatibility; it was supposed to have a Crusoe-like translator to change x86 instructions to VLIW instructions. Did this get tossed, or is Tom smoking something?
--
IIRC, Rob and Jeff have contracts that force them to work for a couple of years. I doubt you'd be seeing them go anywhere.
They're not bound into slavery ("force"?). It just would probably cost them a bunch of stock options if they left.
--
Before the buyout/merger, would Slashdot have posted anti-Slashdot articles at all? I don't particularly remember any, although Slashdot wasn't as high-profile back then, so external critical articles may not have existed.
Personally, I don't feel that Slashdot has any particular responsibility to put themselves on a whipping post to "prove" their independence.
--
Oracle on the other hand is as much as 14x slower than MySQL, and has serious performance issues with things like connecting to the database. You absolutely must use database connection caching to get acceptable performance. Not to mention the obscene pricing...
I agree on connections, and really agree on the price, but I have to say the 14x thing is pretty exaggerated. As usual, it depends on what you're doing, and I wouldn't be surprised if you can find a slew of operations that Oracle is 14x faster than MySQL (large databases come to mind).
The fact is that Oracle is pretty much the standard for very large web sites. I mean, I wouldn't use MySQL for a "real" web site where I cared about the data (like an e-commerce application), since MySQL doesn't support commit/rollback.
--
Even if we accept all your premises (such as a purely mechanistic universe), it doesn't imply that time would repeat in exactly the same way. It could be that the properties of the next bang depend on the end conditions of the previous crunch, rather than all properties being "reset" as you imply.
--
Uh... all the learning in the world is not going to change the fact that all the client applications under Linux are far inferior to the applications under Windows.
People use applications, not operating systems.
--
There's no transcript, so what's the essence of the argument why I wouldn't want WYSIWYG while I'm working with a document?
--
I don't know; maybe I'm weird but I just don't feel this emotional connection to Slashdot. It's a very vibrant community, I enjoy the discussions, but it's just a web site. To be honest with you, I think it's comments like yours that give people like Jon Katz an inflated sense of their own importance. Sites like this do have an effect, but it's an aggregated effect of all the various discussion sites. Slashdot is just one tiny piece of that effect.
And if Slashdot should die, another site will rise and take its place.
--
Interestingly, MySQL is not open source.
Hmmm; you are correct. I forgot they only GPL'd a very old version. I wonder if they stick with MySQL out of some sort of momentum, or if they would consider using Oracle?
--
Oracle is faster than you might think. You do have to architect it correctly, however, and that primarily means keeping connections open. I agree that starting the gears turning is a slow process, but once you've got them spinning, it can really move data.
--
Learn about styles. They are the reason you're having so much frustration with Word. I've seen this a lot when someone comes from Wordperfect (I did too).
Once you understand how Word works, you will find it far, far better than Wordperfect. It's extremely powerful.
--
As users we have to learn to see beyond the fancy user interfaces, and and focus more on the usability of the information we are producing.
Uh, so you're saying it's not useful to see on the screen exactly what will be printed? It's just useless flash?
--
You said that it would be "trival" to prove that GC is slower than hand management. If its trivial, it should be easy for someone of your broad knowledge of computer science to provide it. If you're going to make that kind of statement, the burden of proof is on you.
You're not paying attention. I gave you the proof. Someone with your awesome abilities should be able to blow it down in one whiff.
CMU floating point bench marks
LOL! A prime example of "lies, damn lies, and benchmarks". A trivial floating point benchmark proves absolutely nothing. Of course that type of problem is going lend itself well to compiler optimization. I could write similar things in APL, and it would probably be very fast, but that would say nothing about APL's viability for general problem domains.
You would be a bit more credible if you gave me a non-trivial program written in both languages. Say, a C++ compiler and optimizer. Or how about a Photoshop-like graphics manipulation tool. Or heck, how about a version of Quake in Lisp?
What is your theory on why people don't use Lisp for any non-trivial project? Is my apparent level of ignorance so rampant in the industry?
You must be in academia.
Quote from Harlequin: "By supporting a variety of manual policies and tuning, Harlequin delivers significantly better performance than can be achieved with other vendors' existing 'one size fits all' allocators."
Apparently they suffer under the same delusions that I do.
--