Among other things, Game Theory gives you the optimum strategy for playing Scissors, Paper, Stone.
Scissors cuts Paper. Paper covers Stone. Stone breaks Scissors.
It turns out that the optimum strategy is to choose each with 1/3 probability. It also turns out that if either player plays optimally, it doesn't matter what the other player does.
Another cutie is the Prisoner's Dilemma. It is to both prisoners' advantage if both keep mum, but it's to each prisoner's advantage if he rats and his partner keeps mum.
The first kind is called a zero-sum game. Each player's gain is the other player's loss. It also applies very well to questions of where and when to attack.
Non-zero-sum games get "interesting". There is probably a good case to be made for altruism since it doesn't do much good to be the best of your species if your species goes extinct.
Seems like there have been some reports that Microsoft Windows is more stable under VMWare than on the bare metal. It does make sense that the virtualized system as presented by VMWare is more regular and less messy than the bare metal.
Yes, but. The bugs taken individually won't show up. It's when they get together (and breed or something) that you get troubles.
all they can do is point out the obvious is far from useless if it wasn't obvious to the reader. O.T. Sounds like you're a couple of years ahead of the curve, so please be kind to those of us trying to play catch up.
Think of it in terms of electrical power generators generating power for manufacturing et cetera. In terms of such as distribution priorities and who get what when, it does make sense to think of them as making chips for Linux.
Ok, but then the slashdot article was some sensation laden journalism. Ok, so Slashdot doesn't pick on just Microsoft.;-) The protocol allows conforming implementations that allow leakage of information. That the protocol also allows conforming implementations that do not leak the information does not remove the flaw from the protocol.
I don't know of any places that go back on their numbering. Funny, most that I've ever seen do go back on their numbering. (Switch to a new volume and restart numbering with 1) How likely is it that something like FOCUS has only been published 7 times previously. The Copyright is 1996-2003, so they can't have just started up.
So it took 7.5 years for Microsoft to be hit with Slapper. WHY things don't work, even if the root causes are very old, are suff that matters (present tense) instead of just stuff that mattered (past tense).
What MS missed was the underground trend in the Unix world to tighten up the technolgoy that was already in place -- better file permissions, services that don't require root, "secure by default", etc. Hmmmm, I wonder how much of that is due to the OpenBSD crowd. Considering its origins, I think Unix would tend to have security appropriate for a department, about the level of the locks on office doors and desks. For security you would use physical isolation and/or something actually designed for security. All things considered, Unix security does pretty well. It does seem optimized to give about the most effective security for the lease amount of effort.
Now there is a solid subsection of the 'Linux community' that actually represents the majority of vocal Linux users, who are bitter 'Anything-But-Microsoft' naysayers. [emphasis added] I'm sorry you think we're fouling things up, but methinks you're maybe misidentifying the 'We need ONE UNIFIED DESKTOP to DEFEAT MICROSOFT' people. I'm sure that Microsoft would love to have just ONE UNIFIED DESKTOP to target instead of such as KDE and Gnome which seem to do a fairly good job of running each other's applications. I'm sure a lot of it is the tendency for newbies to want to offer "helpful" suggestions without taking the trouble to understand the situation, but I'm enough of a conspiracy theorist to think some of it is an attempt by Microsoft to try to get the target to stand still so it can be shot at. There's an interesting comparison in the handling of the latest round of exploits between Open Source and Microsoft. Open Source maybe could have handled it better and smoother, but the handling would have had to be incredibly worse for the exploits to have actually accomplished much of anything. If the patch for slapper hadn't been out already, Microsoft would have had an incredibly hard time of it.
minor changes must have minor effects Try that on the edge of a cliff. Thrashing was discovered on a time-sharing system when adding one more user caused system response time for everybody to become incredibly slow. Super-cooled solution. One speck of dust and it freezes.
If you get nailed on an audit because of a mistake in their software, they pay the bill Like they'll cover penalties for your interpretation of what their words really meant. You might get a better guarantee from a Xerox machine.
Speaking as someone with a strong dislike of things like sound cards and screen saver, no. Stick a multimedia entertainment center on a production server. A decent OS will keep them from messing with each other too much. The main reason we're currently using Linux on servers is an early test where a fancy-schmancy screensaver did not impact the responsiveness of a telnet session.
Windows is no more or no less vulnerable than Linux when you have console access as has been pointed out repeatedly.
Windows is vulnerable when you have console access. Linux is vulnerable when you have console access. All vulnerabilities are created equal. Windows is just as vulnerable as Linux. (or CP/M or DOS)
Actually Linux is effectively less vulnerable since people tend to question why it was rebooted. A freshly rebooted Windows system is considered "normal".
Ultimately you will live or die based on those users. Anything else, like network-based or not, is minor in comparison. It's like the centralized versus decentralized argument. Good centralized beats bad decentralized and good decentralized beats bad centralized. It's very hard to centralize everything on the network, but it's also very hard to communicate when you're isolated. Carving up a mainframe into a gaggle of pc's is a losing proposition. IMNSHO *that* is what "killed the mainframes". Trying to make a mainframe out of a gaggle of pc's doesn't really work either. Maybe that's why IBM is selling bigger and better mainframes these days.
Now you know the real reason for the interest in long uptimes. Boot into single-user and do whatever. Reboot system into normal. No traces. If the system stays up, it's much much harder not to leave traces.
Microsoft actually happens to adapt to things very quickly at times Like... security? Well to give credit, they did manage to get something up on microsoft.com in about 24 hours as opposed to 3 days for Code Red.
20 years ago, IBM was as much of a monopolist as Microsoft is today. Yes, but. Seems like IBM has always had a much stronger ethical sense than Microsoft. I don't know that IBM lost their anti-trust action, but I'm reasonably sure that it was sufficiently invasive and painful that they do not want to repeat the experience. Seems like FUD started not so much with IBM as with keypunch supervisors promoted to IT executives who were terrified of doing anything to provoke IBM's ire and receive support that while technically meeting requirements would be not quite at the level required.
Slashdot Mirror
The trick is to just not open unexpected attachments especially from people you know.
I think my favorite of von Neumann was "Theory of Self-Reporducing Automata".
Among other things, Game Theory gives you the optimum strategy for playing Scissors, Paper, Stone.
Scissors cuts Paper.
Paper covers Stone.
Stone breaks Scissors.
It turns out that the optimum strategy is to choose each with 1/3 probability.
It also turns out that if either player plays optimally, it doesn't matter what the other player does.
Another cutie is the Prisoner's Dilemma.
It is to both prisoners' advantage if both keep mum, but it's to each prisoner's advantage if he rats and his partner keeps mum.
The first kind is called a zero-sum game. Each player's gain is the other player's loss. It also applies very well to questions of where and when to attack.
Non-zero-sum games get "interesting". There is probably a good case to be made for altruism since it doesn't do much good to be the best of your species if your species goes extinct.
Seems like there have been some reports that Microsoft Windows is more stable under VMWare than on the bare metal. It does make sense that the virtualized system as presented by VMWare is more regular and less messy than the bare metal.
Yes, but. The bugs taken individually won't show up. It's when they get together (and breed or something) that you get troubles.
I don't think even Microsoft is that stupid.
(But I could be wrong;-)
all they can do is point out the obvious
is far from useless if it wasn't obvious to the reader.
O.T. Sounds like you're a couple of years ahead of the curve, so please be kind to those of us trying to play catch up.
Think of it in terms of electrical power generators generating power for manufacturing et cetera. In terms of such as distribution priorities and who get what when, it does make sense to think of them as making chips for Linux.
Ok, but then the slashdot article was some sensation laden journalism. ;-)
Ok, so Slashdot doesn't pick on just Microsoft.
The protocol allows conforming implementations that allow leakage of information. That the protocol also allows conforming implementations that do not leak the information does not remove the flaw from the protocol.
No, too many cooks fork the broth.
This begets competition, back-porting, and maybe even eventual convergence.
I don't know of any places that go back on their numbering.
Funny, most that I've ever seen do go back on their numbering.
(Switch to a new volume and restart numbering with 1)
How likely is it that something like FOCUS has only been published 7 times previously. The Copyright is 1996-2003, so they can't have just started up.
So it took 7.5 years for Microsoft to be hit with Slapper.
WHY things don't work, even if the root causes are very old, are suff that matters (present tense) instead of just stuff that mattered (past tense).
What MS missed was the underground trend in the Unix world to tighten up the technolgoy that was already in place -- better file permissions, services that don't require root, "secure by default", etc.
Hmmmm, I wonder how much of that is due to the OpenBSD crowd.
Considering its origins, I think Unix would tend to have security appropriate for a department, about the level of the locks on office doors and desks. For security you would use physical isolation and/or something actually designed for security. All things considered, Unix security does pretty well. It does seem optimized to give about the most effective security for the lease amount of effort.
Now there is a solid subsection of the 'Linux community' that actually represents the majority of vocal Linux users, who are bitter 'Anything-But-Microsoft' naysayers. [emphasis added]
I'm sorry you think we're fouling things up, but methinks you're maybe misidentifying the 'We need ONE UNIFIED DESKTOP to DEFEAT MICROSOFT' people. I'm sure that Microsoft would love to have just ONE UNIFIED DESKTOP to target instead of such as KDE and Gnome which seem to do a fairly good job of running each other's applications. I'm sure a lot of it is the tendency for newbies to want to offer "helpful" suggestions without taking the trouble to understand the situation, but I'm enough of a conspiracy theorist to think some of it is an attempt by Microsoft to try to get the target to stand still so it can be shot at. There's an interesting comparison in the handling of the latest round of exploits between Open Source and Microsoft. Open Source maybe could have handled it better and smoother, but the handling would have had to be incredibly worse for the exploits to have actually accomplished much of anything. If the patch for slapper hadn't been out already, Microsoft would have had an incredibly hard time of it.
Much like chess is not a spectator sport.
Until the media can be realistic portraying chess, it doesn't stand a chance with programming.
minor changes must have minor effects
Try that on the edge of a cliff.
Thrashing was discovered on a time-sharing system when adding one more user caused system response time for everybody to become incredibly slow.
Super-cooled solution. One speck of dust and it freezes.
If you get nailed on an audit because of a mistake in their software, they pay the bill
Like they'll cover penalties for your interpretation of what their words really meant. You might get a better guarantee from a Xerox machine.
Speaking as someone with a strong dislike of things like sound cards and screen saver, no. Stick a multimedia entertainment center on a production server. A decent OS will keep them from messing with each other too much. The main reason we're currently using Linux on servers is an early test where a fancy-schmancy screensaver did not impact the responsiveness of a telnet session.
Windows is no more or no less vulnerable than Linux when you have console access as has been pointed out repeatedly.
Windows is vulnerable when you have console access.
Linux is vulnerable when you have console access.
All vulnerabilities are created equal.
Windows is just as vulnerable as Linux. (or CP/M or DOS)
Actually Linux is effectively less vulnerable since people tend to question why it was rebooted. A freshly rebooted Windows system is considered "normal".
And there goes your uptime.
Ever have someone lose a password?
Ultimately you will live or die based on those users. Anything else, like network-based or not, is minor in comparison. It's like the centralized versus decentralized argument. Good centralized beats bad decentralized and good decentralized beats bad centralized. It's very hard to centralize everything on the network, but it's also very hard to communicate when you're isolated. Carving up a mainframe into a gaggle of pc's is a losing proposition. IMNSHO *that* is what "killed the mainframes". Trying to make a mainframe out of a gaggle of pc's doesn't really work either. Maybe that's why IBM is selling bigger and better mainframes these days.
Now you know the real reason for the interest in long uptimes.
Boot into single-user and do whatever. Reboot system into normal. No traces.
If the system stays up, it's much much harder not to leave traces.
Microsoft actually happens to adapt to things very quickly at times ... security?
Like
Well to give credit, they did manage to get something up on microsoft.com in about 24 hours as opposed to 3 days for Code Red.
20 years ago, IBM was as much of a monopolist as Microsoft is today.
Yes, but.
Seems like IBM has always had a much stronger ethical sense than Microsoft. I don't know that IBM lost their anti-trust action, but I'm reasonably sure that it was sufficiently invasive and painful that they do not want to repeat the experience. Seems like FUD started not so much with IBM as with keypunch supervisors promoted to IT executives who were terrified of doing anything to provoke IBM's ire and receive support that while technically meeting requirements would be not quite at the level required.
Without backing down in the slightest, I completely agree with you.