Well, I know a lot of people will say this proves that God doesn't exist. However, if the scientists were able to successfully create the sensation and sight of eating a jelly donut, without the actual donut being present, would you then conclude that donuts don't exist?
Or, more generally, the successful creation of an effect neither proves nor disproves the existance of other possible causes for the effect.
For a while, I thought my fan might have been broken because my laptop was getting very hot. Then I realized that, a few months ago I had messed with the power setting and turned off that technology to make sure I was getting maximum performance out of something. I forgot to turn it back on, and this resulted in the machine running flat-out all the time and getting very hot. Something jogged my memory, I went back to the power settings, and it works fine now. Even DVD playback doesn't force it to run flat-out, so if you have this technology you should definitely use it.
Of course it's only easy to feel the heat with a notebook. If you have a desktop you could be wasting power and not even know it unless you check the settings.
Well, it was 3 billion light years away. That means it was 3 billion and 6 years ago. This has to some kind of record, even for Slashdot. Come on guys, get with it.
This reminds me of Netscape. It seems like their web browser reached its optimal functionality with 3.x. 4.x was bloated, and then Netscape was rescued by the AOL buyout.
It seems like the MS development process produced its optimal output with XP. Now it's past peak. All they can do within the context of their current development process is add new features that aren't necessarily what people want. At the same time, they can't solve persistant problems, the biggest being security.
MS can't be rescued by a buyout--they're too big. It looks like MS needs a whole new process. I think it would be really cool if they got more friendly towards Open Source. I'm not talking about the whole OS, just some parts. Perhaps they could even do a *NIX core like Apple did, and create a "Wine Killer" that would run Windows Apps in a *NIX-like environment, flawlessly. Also, it'd be nice to have well-known things like libjpeg, zlib, etc, installed as part of the OS DLLs, instead of having to staticly link that stuff, or fight DLL wars.
Late in high school, some geek friends asked if I wanted to attend a course in PL/1 with them. I thought about it for a day or so, and came to the conclusion that by the time I got to the end of college and was out in the working world, PL/1 might be history. Then, when I got out into the working world I could learn whatever specific language I had to learn. I didn't take the course.
I'm not saying I call it right all the time. I was really skeptical about Java when it first came out, and it seems to have had a lot more staying power than I ever thought it would. OTOH, not only have I not had to code in PL/1, I've never even been at a company that was maintaining it.
We'll scan a person with one of these things and tell what they're actually thinking.'"
Jim, Jim, what's Jack thinking?
Umm...
Well, come-on, Jim. What is it?
Umm... he's thinking that we're a bunch of lamers because we're scanning him with the BB-1600, and everybody who's anybody has a MBB-8, which is what he's got.
Ah, come on. They both work. The MBB-8 just comes in more colors.
When you have an international standard, everything you need to do is clear and straightforward
We know that because it's written in ISO Standard Document 42332/J, Chapter 6, subsection 3, paragraph 2.
Easy setups aside, so what if it's clear and straightforward? If the standard says "bend over" that's very clear and straightforward. If the standard says "everybody has to do things the same way", that's very clear and straightforward.
So, Google, you want it easy. Well, sorry Google, that's Sloth. Sloth is one of the Seven Deadly Sins, isn't it? So, quit doing evil. If you try to jam a one-size-fits-all solution down our throats, if you try to run roughshod over sovereignty via international organizations and deal-making, just to satisfy your Sloth, that's evil. So stop it.
If a politician had complained about RST, how many Slashdotters would have blasted the guy for being an idiot and not knowing how firewalls work? How many *decades* have firewalls been sending RST packets, with nobody complaining? Note, I'm not defending them shutting down BT. I'm just saying that if they are using a RST packet, that's been an accepted practice for quite some time. It's the application of that RST packet that's an issue here.
It handled the load just fine when I updated. Now, I know at least one person that I'm reluctant to tell about this, because he's already adicted to GE. We've talked about the possibility of them putting a flight sim in there before--it seemed like a natural idea, since we had both done really lame "flights" just with the momentum of the existing controls.
FWIW, the F-16 is actually quite forgiving. You just have to gain altitude. I managed to get it up to 50000 ft. I think they cripped the thing, because it was struggling to do 300 at 50,000 ft. I want to see the SR-71 and/or some fantasy craft that can fly into space added at some point. Oh... and yes, an autopilot would be nice. Once you get away from level flight, it's quite difficult to zero out the shimmy, and so far all my flights have ended by spinning into the ground. I managed to take the F16 from SFO to Lake Tahoe, and plant it into the Nevada desert. If I had not tried to turn around, it probably would have circumnavigated in level flight. Oh crap, but as I type this the screen is wobbling. That's just wrong. I need to step away.
Still not sure if I should tell my GE-addict friend about this.
An awful lot of the early internet documents were written by bureaucrats. Ever read an RFC? There's plenty of "design by committee" stuff in them, and obscure features that usually end up being headaches. Oh, and the ARPANET was a creation of the DoD bureaucracy.
IIRC, Larry Wall, creator of Perl, which was really the only server-side option for a few years, was a GSer too.
Of course, if youre definition of bureaucrat is something other than "works for the government, creates and/or follows their arcane procedures", then my bad.
I could hardly consider development environment something without support of unit-testing built-in
OK. Enough already. How much effort does it really take to add an extra target to your makefile, run the unit test and diff the expected output against the actual output? Every "unit testing framework" I've ever looked at has been thousands of lines of cruft just so you can do something like "assert(output < 2);".
I've got an open mind though. Convince me. Please don't cite Cxxtest as an example. That's exactly the kind of junk I'm talking about. Pulling in Perl or Python and munging a header into a test app, just for simple assertions? Pulleeze.
The one thing he's known for, the Linux kernel, isn't something I particularly like (BSD--more liberal license, Windows--better desktop, Linux? I only use it because of work); but I tend to agree with him on a lot of things. That he would downplay the controversy, and point out that it only illustrates bias doesn't surprise me. He seems to have a gift for cooling things down, for steering clear of immature games and sticking to a clear analysis of the situation.
Good grief. I'm whole-heartedly with Guitard on this. Having recently been turned on to some of the less sweet chocolates (the kind that list their cocoa percentage), the last thing I would want to see is someone using this cheap grease and trying to pass it off as good chocolate.
An application shouldn't consume excessive resources.
If an application attempts to consume excessive resources, the OS should not allow that. Appropriate responses might include failing to provide the resource, or terminating the application. A BSOD terminates everything--plainly not the correct response.
Now, considering the.MOV+Toshiba+Vista situation... something in kernel space is plainly wrong. It might be MS code or Toshiba code. We don't know. Something in application space might be wrong too; but it makes sense to fix bugs in code that runs in the kernel first. Then, if the application "hangs" or triggers a more manageable exception (The little popup window that doesn't crash the whole OS, whatever they call that on Vista) then we know that the application had problems too. If that doesn't happen, then the application was innocent all along.
Still b0rked, since it's not likely that a user would intentionally input such a value.
Wow! That's even worse than ANDing with 0xFF, since it now corrupts octets other than the one that was enterred incorrectly. Which configuration tool developers were so obsessed with memory footprint and cycles that they chose not to validate user input?
One guy at HP was a world class crackpot. For every good idea he had, he was flooding his managers with 100 ideas that ran the gamout from Rube Goldberg, conspiracies, implementation prohibitive schemes to down-right illegal-by-the-laws-of-physics-and-animal-husbandr y. Needless to say, he was ignored most of the time.
I can identify with that somewhat. Over the years I've learned how not to be that guy. You go into "creative mode" where anything is possible. Then, you need to step out of it. Two good ways to do that: 1. Develop your own inner "reality checker". 2. Somewhat more embarassing, but more effective, is to bounce your nutty ideas off TRUSTED colleagues rather than taking them straight to the boss. The TRUSTED part is important because if your colleague is a jerk he might tell you the idea is bad, then give it to the boss himself. Telling small groups of colleagues can solve this too. You are not likely to have your ideas uncredited by anybody unless you are so unlikeable that everybody wants to frustrate you out of office. If everybody wants you out of office, you need to start washing more than once a month, or cut down on the 15 minute screaming sessions, or stop accusing everybody of stealing the sugar packets from your desk, or whatever it is you're doing. Well, I could go on, but I think you get the idea...
Having an input parse the octet as an int and ANDing with 0xFF is not the same as actually having 345 as an "octet" or using a different base. In other words, 89!=345. I would submit that if you have an API that takes 345 and ANDs it with 0xFF without informing you of an error, that API is broken. It'd be far better to have it throw an error, since somebody almost certainly fat-fingered the IP. A 32-bit unsigned representing the full IP doesn't translate meaningfully into a decimal dotted quad with any value greater than 255 either.
However, if each octet was in octal notation, then yes indeed 024.076.0345.0200==20.62.229.128.
Is there actually in IP stack configuration or networking tool that lets you input 345 decimal, and automaticly ANDs it with 0xFF? If there is, please tell me so I can either patch it if it's OSS, or avoid it like the plague if it isn't.
"The man" doesn't learn. Not in the traditional sense. The man can only be usurped. Then there's a new man, and the cycle repeats. This usurping doesn't necessarily have to be violent. Every shift in power in Congress or the Presidency is a mild form of usurping. Even if it's *somewhat* violent, it doesn't have to totally destroy society (e.g., the US civil rights movement). Of course, examples of violent revolution are, unfortunately, all too common, and the "new boss" is often worse than the old boss; but the point is, the boss or "the man" itself doesn't learn, at least I can't think of any examples.
Slashdot Mirror
Well, I know a lot of people will say this proves that God doesn't exist. However, if the scientists were able to successfully create the sensation and sight of eating a jelly donut, without the actual donut being present, would you then conclude that donuts don't exist?
Or, more generally, the successful creation of an effect neither proves nor disproves the existance of other possible causes for the effect.
The generation that invented those toys stood in front of a blackboard. It remains to be seen what the kids with the toys will invent.
For a while, I thought my fan might have been broken because my laptop was getting very hot. Then I realized that, a few months ago I had messed with the power setting and turned off that technology to make sure I was getting maximum performance out of something. I forgot to turn it back on, and this resulted in the machine running flat-out all the time and getting very hot. Something jogged my memory, I went back to the power settings, and it works fine now. Even DVD playback doesn't force it to run flat-out, so if you have this technology you should definitely use it.
Of course it's only easy to feel the heat with a notebook. If you have a desktop you could be wasting power and not even know it unless you check the settings.
Well, it was 3 billion light years away. That means it was 3 billion and 6 years ago. This has to some kind of record, even for Slashdot. Come on guys, get with it.
This reminds me of Netscape. It seems like their web browser reached its optimal functionality with 3.x. 4.x was bloated, and then Netscape was rescued by the AOL buyout.
It seems like the MS development process produced its optimal output with XP. Now it's past peak. All they can do within the context of their current development process is add new features that aren't necessarily what people want. At the same time, they can't solve persistant problems, the biggest being security.
MS can't be rescued by a buyout--they're too big. It looks like MS needs a whole new process. I think it would be really cool if they got more friendly towards Open Source. I'm not talking about the whole OS, just some parts. Perhaps they could even do a *NIX core like Apple did, and create a "Wine Killer" that would run Windows Apps in a *NIX-like environment, flawlessly. Also, it'd be nice to have well-known things like libjpeg, zlib, etc, installed as part of the OS DLLs, instead of having to staticly link that stuff, or fight DLL wars.
Late in high school, some geek friends asked if I wanted to attend a course in PL/1 with them. I thought about it for a day or so, and came to the conclusion that by the time I got to the end of college and was out in the working world, PL/1 might be history. Then, when I got out into the working world I could learn whatever specific language I had to learn. I didn't take the course.
I'm not saying I call it right all the time. I was really skeptical about Java when it first came out, and it seems to have had a lot more staying power than I ever thought it would. OTOH, not only have I not had to code in PL/1, I've never even been at a company that was maintaining it.
We'll scan a person with one of these things and tell what they're actually thinking.'"
Jim, Jim, what's Jack thinking?
Umm...
Well, come-on, Jim. What is it?
Umm... he's thinking that we're a bunch of lamers because we're scanning him with the BB-1600, and everybody who's anybody has a MBB-8, which is what he's got.
Ah, come on. They both work. The MBB-8 just comes in more colors.
Yeah. Mac fan boys. Piss me off.
When you have an international standard, everything you need to do is clear and straightforward
We know that because it's written in ISO Standard Document 42332/J, Chapter 6, subsection 3, paragraph 2.
Easy setups aside, so what if it's clear and straightforward? If the standard says "bend over" that's very clear and straightforward. If the standard says "everybody has to do things the same way", that's very clear and straightforward.
So, Google, you want it easy. Well, sorry Google, that's Sloth. Sloth is one of the Seven Deadly Sins, isn't it? So, quit doing evil. If you try to jam a one-size-fits-all solution down our throats, if you try to run roughshod over sovereignty via international organizations and deal-making, just to satisfy your Sloth, that's evil. So stop it.
what free alternatives exist that can replace this much-needed service?"
Easy. Get up and walk off some of those calories.
If a politician had complained about RST, how many Slashdotters would have blasted the guy for being an idiot and not knowing how firewalls work? How many *decades* have firewalls been sending RST packets, with nobody complaining? Note, I'm not defending them shutting down BT. I'm just saying that if they are using a RST packet, that's been an accepted practice for quite some time. It's the application of that RST packet that's an issue here.
It handled the load just fine when I updated. Now, I know at least one person that I'm reluctant to tell about this, because he's already adicted to GE. We've talked about the possibility of them putting a flight sim in there before--it seemed like a natural idea, since we had both done really lame "flights" just with the momentum of the existing controls.
FWIW, the F-16 is actually quite forgiving. You just have to gain altitude. I managed to get it up to 50000 ft. I think they cripped the thing, because it was struggling to do 300 at 50,000 ft. I want to see the SR-71 and/or some fantasy craft that can fly into space added at some point. Oh... and yes, an autopilot would be nice. Once you get away from level flight, it's quite difficult to zero out the shimmy, and so far all my flights have ended by spinning into the ground. I managed to take the F16 from SFO to Lake Tahoe, and plant it into the Nevada desert. If I had not tried to turn around, it probably would have circumnavigated in level flight. Oh crap, but as I type this the screen is wobbling. That's just wrong. I need to step away.
Still not sure if I should tell my GE-addict friend about this.
...hot alien babes?
...a cup holder?
...a flying cup?
...tea?
...an application for a Darwin award?
all the rest were bureaucrats.
An awful lot of the early internet documents were written by bureaucrats. Ever read an RFC? There's plenty of "design by committee" stuff in them, and obscure features that usually end up being headaches. Oh, and the ARPANET was a creation of the DoD bureaucracy.
IIRC, Larry Wall, creator of Perl, which was really the only server-side option for a few years, was a GSer too.
Of course, if youre definition of bureaucrat is something other than "works for the government, creates and/or follows their arcane procedures", then my bad.
I could hardly consider development environment something without support of unit-testing built-in
OK. Enough already. How much effort does it really take to add an extra target to your makefile, run the unit test and diff the expected output against the actual output? Every "unit testing framework" I've ever looked at has been thousands of lines of cruft just so you can do something like "assert(output < 2);".
I've got an open mind though. Convince me. Please don't cite Cxxtest as an example. That's exactly the kind of junk I'm talking about. Pulling in Perl or Python and munging a header into a test app, just for simple assertions? Pulleeze.
Good point about gnome/kde. Nobody's perfect. As for v2/v3, it seems like a classic "dammned if you do, damned if you don't" situation.
The one thing he's known for, the Linux kernel, isn't something I particularly like (BSD--more liberal license, Windows--better desktop, Linux? I only use it because of work); but I tend to agree with him on a lot of things. That he would downplay the controversy, and point out that it only illustrates bias doesn't surprise me. He seems to have a gift for cooling things down, for steering clear of immature games and sticking to a clear analysis of the situation.
And henceforth, cheap fatty oil is "freedom".
Good grief. I'm whole-heartedly with Guitard on this. Having recently been turned on to some of the less sweet chocolates (the kind that list their cocoa percentage), the last thing I would want to see is someone using this cheap grease and trying to pass it off as good chocolate.
An application shouldn't consume excessive resources.
If an application attempts to consume excessive resources, the OS should not allow that. Appropriate responses might include failing to provide the resource, or terminating the application. A BSOD terminates everything--plainly not the correct response.
Now, considering the .MOV+Toshiba+Vista situation... something in kernel space is plainly wrong. It might be MS code or Toshiba code. We don't know. Something in application space might be wrong too; but it makes sense to fix bugs in code that runs in the kernel first. Then, if the application "hangs" or triggers a more manageable exception (The little popup window that doesn't crash the whole OS, whatever they call that on Vista) then we know that the application had problems too. If that doesn't happen, then the application was innocent all along.
It sounds like ignorance really is bliss. Thanks both of you for filling me in on this reference though.
Well then: Full PC for sale. Includes OEM XP CD-ROM. PC and CD-ROM shipped separately. You agree not to hold me liable if the PC is lost in shipping.
Still b0rked, since it's not likely that a user would intentionally input such a value.
Wow! That's even worse than ANDing with 0xFF, since it now corrupts octets other than the one that was enterred incorrectly. Which configuration tool developers were so obsessed with memory footprint and cycles that they chose not to validate user input?
I can identify with that somewhat. Over the years I've learned how not to be that guy. You go into "creative mode" where anything is possible. Then, you need to step out of it. Two good ways to do that: 1. Develop your own inner "reality checker". 2. Somewhat more embarassing, but more effective, is to bounce your nutty ideas off TRUSTED colleagues rather than taking them straight to the boss. The TRUSTED part is important because if your colleague is a jerk he might tell you the idea is bad, then give it to the boss himself. Telling small groups of colleagues can solve this too. You are not likely to have your ideas uncredited by anybody unless you are so unlikeable that everybody wants to frustrate you out of office. If everybody wants you out of office, you need to start washing more than once a month, or cut down on the 15 minute screaming sessions, or stop accusing everybody of stealing the sugar packets from your desk, or whatever it is you're doing. Well, I could go on, but I think you get the idea...
Having an input parse the octet as an int and ANDing with 0xFF is not the same as actually having 345 as an "octet" or using a different base. In other words, 89!=345. I would submit that if you have an API that takes 345 and ANDs it with 0xFF without informing you of an error, that API is broken. It'd be far better to have it throw an error, since somebody almost certainly fat-fingered the IP. A 32-bit unsigned representing the full IP doesn't translate meaningfully into a decimal dotted quad with any value greater than 255 either.
However, if each octet was in octal notation, then yes indeed 024.076.0345.0200==20.62.229.128.
Is there actually in IP stack configuration or networking tool that lets you input 345 decimal, and automaticly ANDs it with 0xFF? If there is, please tell me so I can either patch it if it's OSS, or avoid it like the plague if it isn't.
I don't see what I'm missing here.
Wow! You really are a super-hacker. I could never even get a stack to accept that, let alone have those packets route.
"The man" doesn't learn. Not in the traditional sense. The man can only be usurped. Then there's a new man, and the cycle repeats. This usurping doesn't necessarily have to be violent. Every shift in power in Congress or the Presidency is a mild form of usurping. Even if it's *somewhat* violent, it doesn't have to totally destroy society (e.g., the US civil rights movement). Of course, examples of violent revolution are, unfortunately, all too common, and the "new boss" is often worse than the old boss; but the point is, the boss or "the man" itself doesn't learn, at least I can't think of any examples.