State law is strict: if Californians travel to a state with a 5 percent tax and shop there, the law requires them to cough up the 3.25 percent difference when they return. Online purchases are taxed as well.
But compliance is spotty at best. California's Board of Equalization estimates the state lost $1.34 billion in 2003 because residents aren't paying use taxes--and attributes $208 million of that to online purchases.
This reminds me of the RIAA's definition of "lost revenue". The state didn't lose anything... with a law as badly thought out as this, any money they did collect should be treated as a windfall. When you create a law where the only possibility of any compliance at all is people's innate honesty, just be glad that so many people are basically honest and bank what you can.
MacHeads are obviously more likely to pay for all of their software (as well as their media files) instead of pirating it, right?
So that's why the report included Microsoft Office as a cost on the Mac side and not on the PC side. Microsoft's saying "It's OK, go ahead and boost a copy of Office from work... so long as it's on Windows".
You don't need to invoke quantum mechanics to explain why a bunch of software written over a period of several billion years using a random walk algorithm and just about every bad software development model (cargo cult programming, spaghetti inheritance, copy-and-paste,...) doesn't always follow the optimal course of action.
I blame Penrose. Not because I have a logical reason for it, but because blaming Penrose makes my R-complex feel good. Yummy.
It's nothing about storing data in a temporary directory, those flags when set simultaneously actually keep a file(s) in RAM like tmpfs or just partial parts of the file unless there isn't enough RAM to keep there.
Oh, you mean it's a workaround for Windows appalling virtual memory subsystem. *sigh*
One example is the lack of the FILE_ATTRIBUTE_TEMPORARY and FILE_FLAG_DELETE_ON_CLOSE attributes not existing on older Windows versions
Gee, I have written applications for 30 years without those APIs and I haven't found "wrap open() and close() and remove everything in your temporary directory when you start up" driving up the cost to create an application "insanely". ESPECIALLY when they've already written that code.
If the support for the old version take too much time it's reasonable to stop supporting.
I've seen people say things like this, but I've never seen a breakdown of the missing features or APIs in Windows 2000 that MAKE support cost so much. I've seen plenty of people ask, without any answers ever being given, for any number of applications. This has even less justification than the excuses people give for not supporting Linux or Mac OS, or for not writing portable code in the first place... and THOSE get short shrift on Slashdot.
There is zero way to verify that the code you just compiled is the same code that is showing up on the screen.
Yes, I read "On Trusting Trust" too.
My main point - having the source code should not make people feel as rosy as it seem to make them feel.
You were claiming that it provides effectively NO benefits at all, which is what I was objecting to. You don't sound like a stupid fellow, so I'm pretty sure you know better.
When Microsoft deployed ActiveX installation and launch over HTTP and email with Active Desktop in 1997 they made Windows inherently insecure in a way that nobody had ever imagined anyone would be stupid enought to do. In fact it used to be a joke, the "Good Times" virus... a virus so effective it would run without you even opening the email message it was contained in. EVERYONE knew it was a joke, because EVERYONE knew nobody would be so stupid as to deliberately allow untrusted content to automatically run.
Nobody but Microsoft was that stupid, anyway.
Jesus Christ, man, the fundamental desing of Internet Explorer is so f-ing bad that over 10 years later I am STILL aghast that ANYONE would defend it, or any OS that depends on it. What the HELL are you smoking? DO you honestly not understand just how amazingly stupid this is? Honestly? By the bowels of Christ, consider that you might be mistaken.
No, and I'm not going to worry about Thompson hiding a backdoor in login.c, either.
You wrote, and I quote, "There is no way to verify that, so this is not reassuring in the slightest, unless you don't know how software works."... yes, this reassuring, because you can still verify that you're running the code you compiled. You can modify the open source code so it behaves the way you want, so it doesn't automatically download the code. There's many many differences in the level of appropriate trust between "closed source undocumented code that resists disabling" and "open source peer-reviewed code modified to support the policies you want".
I'm not saying that there's no possibiilty whatsoever there could be a subtle backdoor in the code even after review and rebuilding, I'm just pointing out that it's silly to take an absolute position that "there's no way to verify [it]" or that it's "not reassuring in the slightest".:p
OTOH, in the realm of just annoying, is that a device emulator we use frequently takes about 90 seconds to load and can't just be left running -- you have to restart it for each recompile. It's like the testing cycle is make as many changes as possible, compile, go get a beverage or take a pee, come back, it should be just about ready to run.
You poor things. My first job we got two test runs a day, and if you made a typo on your coding pad you had to wait in line for the one working keypunch so you could correct the cards without waiting for another run to the service bureau they had punching production cards for us in the name of "efficiency".
Kids today don't even know what "desk checking" is for.
I guess I'm just stupid but can't you just make your URL shorter in the first place?
That was also my first thought, well, my second thought. My first thought was "what does 'rev' mean?". Reverse? Revised? Whts wrng with usng full wrds?
They don't have the advertising clause in there any more, but the acknowledgment clause is still intact. Don't feel bad, there's been lots of people who've confused BSDL with public domain and ripped off the BSDL and replaced it with something like the GPL and tripped over that one.
But releasing code that's known to be covered under a valid patent would undermine the spirit of open source no matter what license you use.
No doubt. But the only reason for them NOT using the MIT/BSD style license would be if they wanted to add more restrictions than just the patent. So I suspect the patents aren't the only issue... they don't just want to release with an OS license plus a patent, they want something else as well.
Slashdot Mirror
State law is strict: if Californians travel to a state with a 5 percent tax and shop there, the law requires them to cough up the 3.25 percent difference when they return. Online purchases are taxed as well.
But compliance is spotty at best. California's Board of Equalization estimates the state lost $1.34 billion in 2003 because residents aren't paying use taxes--and attributes $208 million of that to online purchases.
This reminds me of the RIAA's definition of "lost revenue". The state didn't lose anything... with a law as badly thought out as this, any money they did collect should be treated as a windfall. When you create a law where the only possibility of any compliance at all is people's innate honesty, just be glad that so many people are basically honest and bank what you can.
Is that Bill Gates? So that's why he retired!
MacHeads are obviously more likely to pay for all of their software (as well as their media files) instead of pirating it, right?
So that's why the report included Microsoft Office as a cost on the Mac side and not on the PC side. Microsoft's saying "It's OK, go ahead and boost a copy of Office from work... so long as it's on Windows".
You don't need to invoke quantum mechanics to explain why a bunch of software written over a period of several billion years using a random walk algorithm and just about every bad software development model (cargo cult programming, spaghetti inheritance, copy-and-paste, ...) doesn't always follow the optimal course of action.
I blame Penrose. Not because I have a logical reason for it, but because blaming Penrose makes my R-complex feel good. Yummy.
They aren't just trying to stick it to the heavy users...they are trying to stick it to everyone.
That should have been your subject line. :)
Try faking that with gummy bears.
Any "Dr Who" fan knows you need to use Jelly Babies.
It's nothing about storing data in a temporary directory, those flags when set simultaneously actually keep a file(s) in RAM like tmpfs or just partial parts of the file unless there isn't enough RAM to keep there.
Oh, you mean it's a workaround for Windows appalling virtual memory subsystem. *sigh*
One example is the lack of the FILE_ATTRIBUTE_TEMPORARY and FILE_FLAG_DELETE_ON_CLOSE attributes not existing on older Windows versions
Gee, I have written applications for 30 years without those APIs and I haven't found "wrap open() and close() and remove everything in your temporary directory when you start up" driving up the cost to create an application "insanely". ESPECIALLY when they've already written that code.
Try again.
If the support for the old version take too much time it's reasonable to stop supporting.
I've seen people say things like this, but I've never seen a breakdown of the missing features or APIs in Windows 2000 that MAKE support cost so much. I've seen plenty of people ask, without any answers ever being given, for any number of applications. This has even less justification than the excuses people give for not supporting Linux or Mac OS, or for not writing portable code in the first place... and THOSE get short shrift on Slashdot.
There is zero way to verify that the code you just compiled is the same code that is showing up on the screen.
Yes, I read "On Trusting Trust" too.
My main point - having the source code should not make people feel as rosy as it seem to make them feel.
You were claiming that it provides effectively NO benefits at all, which is what I was objecting to. You don't sound like a stupid fellow, so I'm pretty sure you know better.
Security has to be designed in.
When Microsoft deployed ActiveX installation and launch over HTTP and email with Active Desktop in 1997 they made Windows inherently insecure in a way that nobody had ever imagined anyone would be stupid enought to do. In fact it used to be a joke, the "Good Times" virus... a virus so effective it would run without you even opening the email message it was contained in. EVERYONE knew it was a joke, because EVERYONE knew nobody would be so stupid as to deliberately allow untrusted content to automatically run.
Nobody but Microsoft was that stupid, anyway.
Jesus Christ, man, the fundamental desing of Internet Explorer is so f-ing bad that over 10 years later I am STILL aghast that ANYONE would defend it, or any OS that depends on it. What the HELL are you smoking? DO you honestly not understand just how amazingly stupid this is? Honestly? By the bowels of Christ, consider that you might be mistaken.
No, and I'm not going to worry about Thompson hiding a backdoor in login.c, either.
You wrote, and I quote, "There is no way to verify that, so this is not reassuring in the slightest, unless you don't know how software works."... yes, this reassuring, because you can still verify that you're running the code you compiled. You can modify the open source code so it behaves the way you want, so it doesn't automatically download the code. There's many many differences in the level of appropriate trust between "closed source undocumented code that resists disabling" and "open source peer-reviewed code modified to support the policies you want".
I'm not saying that there's no possibiilty whatsoever there could be a subtle backdoor in the code even after review and rebuilding, I'm just pointing out that it's silly to take an absolute position that "there's no way to verify [it]" or that it's "not reassuring in the slightest". :p
Korean Internet users now have to submit their resident registration codes, the Korean equivalent of social security numbers, and names, before posting files or commenting on Web sites with more than 100,000 daily visitors
Have they become what they hate?
OTOH, in the realm of just annoying, is that a device emulator we use frequently takes about 90 seconds to load and can't just be left running -- you have to restart it for each recompile. It's like the testing cycle is make as many changes as possible, compile, go get a beverage or take a pee, come back, it should be just about ready to run.
You poor things. My first job we got two test runs a day, and if you made a typo on your coding pad you had to wait in line for the one working keypunch so you could correct the cards without waiting for another run to the service bureau they had punching production cards for us in the name of "efficiency".
Kids today don't even know what "desk checking" is for.
OK, "mod parent up funny". :)
Quit diluting the meaning of DRM to simply mean everything stupidly any-business that the companies are doing to reduce their own revenue.
What he said.
Build your own updater, or wait for someone to do that, to replace Google's version. There's only one copy of Google Updater running on your computer.
Gosh whillikers! You don't sound family-friendly at all!
Family friendly means you don't want someone else's values rammed down your throat.
Mein Kampf? The Protocols of the Elders of Zion? That's OK then?
I guess I'm just stupid but can't you just make your URL shorter in the first place?
That was also my first thought, well, my second thought. My first thought was "what does 'rev' mean?". Reverse? Revised? Whts wrng with usng full wrds?
Yah, this is the obvious one. Of course, that's not wacky enough for Wired.
I'd love to see a symphony orchestra play this.
In New York? The animals are hiding. They're edible. Robots aren't.
They don't have the advertising clause in there any more, but the acknowledgment clause is still intact. Don't feel bad, there's been lots of people who've confused BSDL with public domain and ripped off the BSDL and replaced it with something like the GPL and tripped over that one.
But releasing code that's known to be covered under a valid patent would undermine the spirit of open source no matter what license you use.
No doubt. But the only reason for them NOT using the MIT/BSD style license would be if they wanted to add more restrictions than just the patent. So I suspect the patents aren't the only issue... they don't just want to release with an OS license plus a patent, they want something else as well.