Dust. Very fine dust mixed with air in the right proportion is explosive. Flour mills are particularly susceptible. If the grain is stored wet, it will almost certainly release methane which is explosive in the right proportions.
>>wrongly assumes that information has anything to do with nature.
If I'm not mistaken, information has almost everything to do with nature. You increase order in your corner of the world at the expense of more than compensating disorder everywhere else. When you have enough order in your corner, you call it life.
Single-user makes some convenient assumptions about the one-to-one relation between user (singular) and computer (singular). The unix multi-user paradigm translates much better into a single-user (me), multiple-machine scenario.
Another bad decision from the past. "One objective of the Windows NT security model is to ensure that the programs a user runs don't have greater access to object than the user does." Sounds good at first, but it implies that if I can change my password, I can change anyone else's. More importantly, it means that I cannot programmatically protect access to data. If the user can access information through a program, he/she can access the underlying data directly.
Human nature? Right.
In part, the sunk-cost falacy. That something is worth what you paid for it.
To demonstrate the force behind it, after buying a new car, carefully read the ads and other promotional literature for for competing models, and just try to keep any sense of objectivity.
That's funny.
On my NT, click on Start, Help, NT Commands, and there is ftp. Click on Start, Programs, Accessories, and there is telnet. Of course being Microsoft, they are inferior to the unix commands.
>Unlike many other vendors, IBM really seems to 'get' Linux.
I think you are right. I think IBM is looking at least 5-10 years into the future. The real reason for mainframes is the small core of logic that cannot be distributed to multiple machines and still function properly.
One advantage of multiple "independent" sites on a single S/390, if company A and company B must syncronize something without significantly disrupting operations, it should be possible on the mainframe.
>Although it's not for the better.
Anything to win a benchmark by a few percentage points. Nevermind the effects on other running systems or the stability of the system. Reminds me of the "goto considered harmful". It's not that goto's are bad, it's that the resulting mess of spaghetti logic, where things are connected that don't quite fit exactly, leads to systems that are not only broken, but unrepairable.
In-line code modifications can be useful, but should be attempted only be people who understand machine language intimately and also understand the relevant system conventions intimately.
>Didn't Godel prove that no logical system can be both complete and internally consistent?
It's any logical system that includes arithmetic. Some simple logical systems like first order predicate calculus can be both complete and consistent. (I's been a while, so flame me if I've got it wrong;)
Embrace and extend and...
Looks like a lot of Microsoft Astroturfers have gotten moderator points. Yours is one of the few good comments. For something to really scare Microsoft, imagine OpenBSD on the desktop. Not so much the final result, but all the lovely intermediate steps along the way.
The media feeds on the media. Linux and Open Source are newsworthy so they do something on Linux and Open Source. They do what they can with the resources they have, and the result is "would somebody get a clue? Please!" You might be able to get a good take on the situation from Harvard or Yale, from a standpoint of History or Sociology -- but don't hold your breath.
With the current state of affairs, the/. commentary may be the only extant useful source of information.
Obviously, your economics courses were not covered by a lemon law.
Free and open markets require informed buyers and informed sellers. If something is sold as a working computer, it should be a working computer, and Joe Consumer should not be the one to unknowingly absorb the risk.
Good question. Worst choice would be that Joe Consumer is responsible. Best choice would be Intel who should supply non-defective replacements. Seems like the law does no more than any reputable firm would do anyway just to keep from enraging its customers.
The legislation should (unless they muck it up badly) protect you much the same, but probably not quite as well, as CompUSA's return policies.
The legislation would prevent CompUSA from saying "Doesn't work? Too bad, sucker."
One critical question is who is in the better position to assume the risk that it is a lemon.
A dealer, even a small dealer without a service department has a responsibility to the buying public, unless it is sold explicitly as is.
A used monitor at a flea market, unless the seller has a lot of them, pretty much has to be at the risk of the buyer. The buyer and the seller cannot afford teams of lawyers for each purchase, so the government, legal system, whatever needs to set the rules and expectations that apply to a sale. Laws protecting Joe Consumer buying a computer, television, stereo, whatever seems like a good idea.
Seems like there was a PCP (Primary Control Program) (Single Tasking) prior to MFT. Don't know if that counts in your definition of a real OS. Seems like virtual memory was introduced in 370/135. Kinda fun when the dinky 135 will run jobs too big to fit on the 165.
Not that simple. The supply is NOT stable. Demand goes up and it become profitable for the suppliers to switch from low volume high markup to high volume low markup.
I'm still convinced that the Mindcraft benchmarks were rigged, in the choice of hardware if nothing else. Complaining doesn't do a lot of good, though, so best to fix things and get on with it.
Judging from his comment on 2.4 versus 2.2, I think Linux is progressing from hacker-class to enterprise-class. Maybe I'm reading too much between the lines, but it seems like Linux has gone from struggling to arrived, with the gratious comments regarding Windows placating someone who is no longer a contender.
Yeah, getting it do do anything worthwhile is a real "brainer". Wanting to use it is a no-brainer in the sense of requiring no brain (as opposed to not requiring a brain);)
Somehow I think that Unix, both commercial and *BSD will be around for a long, long time.
The support for Linux by the commercial Unix companies is just raising the bar. Eventually they will sell you the binary, supported and/or give you the source, unsupported.
Symbiosis is mutual parasitism.Linux makes an excellent testbed for the advancement of Unix.
Slashdot Mirror
Dust. Very fine dust mixed with air in the right proportion is explosive. Flour mills are particularly susceptible. If the grain is stored wet, it will almost certainly release methane which is explosive in the right proportions.
>>wrongly assumes that information has anything to do with nature.
If I'm not mistaken, information has almost everything to do with nature. You increase order in your corner of the world at the expense of more than compensating disorder everywhere else. When you have enough order in your corner, you call it life.
But how do I install a group with the same name as a user?
Single-user makes some convenient assumptions about the one-to-one relation between user (singular) and computer (singular). The unix multi-user paradigm translates much better into a single-user (me), multiple-machine scenario.
Another bad decision from the past. "One objective of the Windows NT security model is to ensure that the programs a user runs don't have greater access to object than the user does." Sounds good at first, but it implies that if I can change my password, I can change anyone else's. More importantly, it means that I cannot programmatically protect access to data. If the user can access information through a program, he/she can access the underlying data directly.
Human nature? Right.
In part, the sunk-cost falacy. That something is worth what you paid for it.
To demonstrate the force behind it, after buying a new car, carefully read the ads and other promotional literature for for competing models, and just try to keep any sense of objectivity.
Control-Alt-F2 and the crashed machine is running just fine. On Linux.
That's funny.
On my NT, click on Start, Help, NT Commands, and there is ftp. Click on Start, Programs, Accessories, and there is telnet. Of course being Microsoft, they are inferior to the unix commands.
>Unlike many other vendors, IBM really seems to 'get' Linux.
I think you are right. I think IBM is looking at least 5-10 years into the future. The real reason for mainframes is the small core of logic that cannot be distributed to multiple machines and still function properly.
One advantage of multiple "independent" sites on a single S/390, if company A and company B must syncronize something without significantly disrupting operations, it should be possible on the mainframe.
>Although it's not for the better.
Anything to win a benchmark by a few percentage points. Nevermind the effects on other running systems or the stability of the system. Reminds me of the "goto considered harmful". It's not that goto's are bad, it's that the resulting mess of spaghetti logic, where things are connected that don't quite fit exactly, leads to systems that are not only broken, but unrepairable.
In-line code modifications can be useful, but should be attempted only be people who understand machine language intimately and also understand the relevant system conventions intimately.
>Didn't Godel prove that no logical system can be both complete and internally consistent? ;)
It's any logical system that includes arithmetic. Some simple logical systems like first order predicate calculus can be both complete and consistent. (I's been a while, so flame me if I've got it wrong
Embrace and extend and ...
Looks like a lot of Microsoft Astroturfers have gotten moderator points. Yours is one of the few good comments. For something to really scare Microsoft, imagine OpenBSD on the desktop. Not so much the final result, but all the lovely intermediate steps along the way.
The media feeds on the media. Linux and Open Source are newsworthy so they do something on Linux and Open Source. They do what they can with the resources they have, and the result is "would somebody get a clue? Please!" You might be able to get a good take on the situation from Harvard or Yale, from a standpoint of History or Sociology -- but don't hold your breath. /. commentary may be the only extant useful source of information.
With the current state of affairs, the
Obviously, your economics courses were not covered by a lemon law.
Free and open markets require informed buyers and informed sellers. If something is sold as a working computer, it should be a working computer, and Joe Consumer should not be the one to unknowingly absorb the risk.
Good question. Worst choice would be that Joe Consumer is responsible. Best choice would be Intel who should supply non-defective replacements. Seems like the law does no more than any reputable firm would do anyway just to keep from enraging its customers.
The other side of the pond. The side with honest computer sellers ;)
The legislation should (unless they muck it up badly) protect you much the same, but probably not quite as well, as CompUSA's return policies.
The legislation would prevent CompUSA from saying "Doesn't work? Too bad, sucker."
One critical question is who is in the better position to assume the risk that it is a lemon.
A dealer, even a small dealer without a service department has a responsibility to the buying public, unless it is sold explicitly as is.
A used monitor at a flea market, unless the seller has a lot of them, pretty much has to be at the risk of the buyer. The buyer and the seller cannot afford teams of lawyers for each purchase, so the government, legal system, whatever needs to set the rules and expectations that apply to a sale. Laws protecting Joe Consumer buying a computer, television, stereo, whatever seems like a good idea.
Seems like there was a PCP (Primary Control Program) (Single Tasking) prior to MFT. Don't know if that counts in your definition of a real OS. Seems like virtual memory was introduced in 370/135. Kinda fun when the dinky 135 will run jobs too big to fit on the 165.
OpenBSD 2.7 does NOT feel like the 70's. Or do you mean Unix on the PDP-11's?
Internet appliance. ...
Then you refrigerator dumps core and your washing machine doesn't and
Not that simple. The supply is NOT stable. Demand goes up and it become profitable for the suppliers to switch from low volume high markup to high volume low markup.
I'm still convinced that the Mindcraft benchmarks were rigged, in the choice of hardware if nothing else. Complaining doesn't do a lot of good, though, so best to fix things and get on with it.
Judging from his comment on 2.4 versus 2.2, I think Linux is progressing from hacker-class to enterprise-class. Maybe I'm reading too much between the lines, but it seems like Linux has gone from struggling to arrived, with the gratious comments regarding Windows placating someone who is no longer a contender.
Yeah, getting it do do anything worthwhile is a real "brainer". Wanting to use it is a no-brainer in the sense of requiring no brain (as opposed to not requiring a brain) ;)
>Microsoft is in business to make money.
How does Microsoft make money from Internet Explorer?
Somehow I think that Unix, both commercial and *BSD will be around for a long, long time.
The support for Linux by the commercial Unix companies is just raising the bar. Eventually they will sell you the binary, supported and/or give you the source, unsupported.
Symbiosis is mutual parasitism.Linux makes an excellent testbed for the advancement of Unix.