Who the heck is going to have experience with the product, given that it isn't even beta yet?
Is there any benefit on the licensing side? I understood there was no reduced cost for any licenses for virtual machines. MS requires you have a legal copy of the OS for each virtual machine.
Yes, latency and throughput are related, but The point about TCP windows is likely bogus for this application. Most modern TCP implementations include the window scaling option, which will allow scaling to quite high data transfer rates. At these low data rates (a few megabits, or even lower for games) the windows are unlikely to cramp your style (by limiting bandwidth) One usually wants bigger windows for high volume transfers (say > 10 Mbytes/second) that you would see on a LAN.
built for debian... In the name of distribution neutrality... it uses an open package, portable format "OpenPKG", which installs a parallel build environment (it's own gcc, binutils, etc...) beside the linux one, packages for OpenPKG are RPM 4 based. sounds quite painful to install.
Because "dangerous" can mean a lot of different things.
To me, dangerous is leaving this sort of activity to the private sector, with no limitations on what they can look at, or who they can share the information with, or why they are looking, or any requirement for a decent audit trail. For me, dangerous is knowing that people who really want to find the dirt will do it anyways, but if the act of looking up information is itself tracked, then there is a chance of tracking the trackers. If we just put our heads in the sand and say that the goverment must not do this, then the private sector will supply the demand, with far fewer strings attached.
Second, you seem to think that the moment a pattern is detected, the SWAT teams will surround the house. I don't think any sane implementationof this sort of data mining would work that way. The "hit" would end up on somebody's desk, then they start looking for other corroborations, maybe they'll spy on the person for a while. Sound terrible? Well, what about how arabs are practically strip searched at the border now, and how visiting Canadians, living in Canada for many years, but of arab origin, are getting deported to lovely vacation spots like Syria.
Case 1.
Treatment of Arabs.
I prefer targeted observation over blacklisting and harassment of entire groups.
It's impossible to do, without simply annihilating the Constitutionally-guaranteed rights of so many people that you end up giving up the very thing you're fighting to protect.
You say that if someone is listening, then that in itself chills speech. If no-one is listening, what point is there in speaking in the first place? The point of free speech is freedom from harassment having made "dangerous" declarations. George Washington did not rally the states' population without letting any pro-Empire folk know what he was onto. If powerful people want to take someone out, they can hire spies, gather dirt, and then generate evidence for prosecution for a non-crime, just look at the Monica fiasco. It is not what you know: "Clinton indulged in a petty peccadillo, then lied about it to Congress." That is normally a criminal offense, but common sense prevailed, and the consensus seems to have been: "Lying to Congress about your sex life, which is none of their business, is no big deal." So Clinton survived just fine.
Second, how do you differentiate between the one George Washington, and the dozens of Timothy Macveighs, or Bin Ladens, The problem is not what those people say, it is what they do. If George Washington were arrested (after all, he moved around in plain sight), many people would rush to his aide, and the Brits would not have been able to keep him, regardless of what they knew. When MacVeigh was exposed (too late!), the consensus was such that the government could punish the individual without undue objections being raised by the population. The difference between an authoritirian state and a democracy lies not in what the state knows about it's population, but what it is allowed to do about it. The same goes for government spooks. It is not what they know that is important, but what they can do with the knowledge. As soon as they make a move to accuse someone, the audit trail has to become available, along with any information gathered.
When I refer to oversight, I do not mean a small number of insiders. I mean full FOIA access. I mean the right for accused to see who was gathering information on them, once they are accused of something. You presented 1984 as an argument that oversight is in-effective. In 1984, there was no free press, a single political party, and no right to free speech. That doesn't remind me of the Western world today.
A stable multi-party political system is critical, a free press is critical, a rational appreciation of modern technology is critical. It is be
hello? They're politicians! They already are stalked that way, and watched by hoards of reporters looking for the slightest gaff. For them, it is just justice for everyone to be tracked the way public figures are.
Question is: why do we think it is important to have the nation aware of blue dresses and activities between consenting adults in private, but not important to be able to look for patterns in the behaviour of large numbers of people to find the 1 in a million dangerous types?
If we don't engage in far more aggressive information gathering, then we are stuck with mindless random checks based on race i.e. "If you're arab, you might just be a terrrorist!" sticking entire groups in the "suspect" category. That is not right. It will be very hard to do properly, but it is just plain dumb not to try.
Privacy should not about hoarding or hiding information. It needs to be about managing it ethically and responsibly. Sure they track us. But in there is the audit trail of who did the tracking, and why.
The privacy of individuals is a lost cause. There is video along the highways, in stores, automated tellers, at work, and in web cams in a lot of homes. Facial recognition is coming along. within 10 years, the "personal information" companies will be exchanging will be your complete daily routine. "Data mining" will mean being able to lookup where every person was every moment of the day. Law enforcement will have access. It would be too stupid for them not to have access.
It is not a question of whether you think this future is good or bad. It will be too cheap for it not to happen.
Information wants to be free. In the future, your life will be free too.
A transparent multi-national conglomerate (ie. WorldComm, Enron) is necessary for people who own it (the stockholders) to control it. How else do we evaluate how our employees are doing ?
This isn't the same species, but how about swimming for 18 days at a time, and diving to 500m. That doesn't sound terribly idle. (http://www.penguins.cl/king-penguins.htm)
For this species, they migrate from Tierra Del Fuego to Brazil. That's got to be a decent swim. http://animaldiversity.ummz.umich.edu/accou nts/sph eniscus/s._magellanicus$narrative.html
Or how about "The satellite data have already revealed that some penguins swim up to 300 miles from their nesting area to find food while their mates sit in the nests on their eggs. Such a foraging journey can take nearly three weeks, leaving the penguins' newborn chicks at risk of starving before the parent returns. This is particularly worrisome since the number of Magellanic penguin chicks surviving to adulthood has declined in recent years." from ttp://www.artsci.washington.edu/new sletter/Autumn9 9/Boersma.htm
I'd rather have them swim from dawn to dusk as they would in the wild, than sit there and gawk at the tourists. They need the exercise. This is probably the best thing that could happen to them. Mind you, I bet they're costing the zoo a fortune in herring.
never used it. it's been around for many years, used to be a SCO app (so it's politcally a propos;-) and they're still in business. Seems about the complexity you are looking for.
wondrous google is. http://www.google.com/linux?hl=en&lr=&ie=ISO- 8859- 1&q=laboratory+instrument+software&btnG=Google+Sea rch gives... seismologists' tools on linus is: http://www.seismo.unr.edu/ftp/pub/ichinose/LINUX/s eismolinux.html
A cool article about an linux-based instrument for use in U-2 spy planes. http://www.linuxdevices.com/articles/AT62 74728044. html
A linux journal article about linux in a lab. http://www.linuxjournal.com/article.php?sid= 2596
which, in the resources link, includes a link to the "Scientific Applications on Linux" page... http://SAL.KachinaTech.com/index.shtml
Spammers will just team up with hackers with a few thousand machines that are "owned" to do the encryption, and then have those machines send out the mail via a few open relays, so they can keep their stable of machines "safe."
the grid: It isn't just for finding aliens any more.
What you're describing is a rural condo. You mentioned the group is used to working together, but it is probably pressing your luck to assume everyone will get along well forever. You probably want individual "star" topology media going to a central point which should be a locked up place with only the equipment (telco termination + connections to the house) in it. The shack or bunker should be powered independently from any of the units (Joe sells, and the new guy is a luddite who wants no bandwidth, and finds all that equipment gobbles up power. The last thing you want is to have A cutting of B because A doesn't want to pay for bandwidth, but B is behind him, or "A."
I work in a large organization. The people with the biggest investment in MS are the techies themselves, and they (being basically self-taught) are loathe to consider the unknown. There is a pocket of technical people who use linux, but the vast majority on the business side of things are fully endroned. Unless the whole technical support organization is on side, it will not happen. So some sort of gradual introduction and familiarization is needed.
In this guy's case, I assume he is IT, so I figured it might be OK.
Sure you can try a piecemeal approach. But it is probably more effective to go the whole 9 yards. for twenty people, switch the clients over to linux as well. Otherwise, you're going to be spending plenty on MS-Office & Virus software per year. There will be an initial pain of transition, but after the hump you will be way better off.
Many, many organizations have IT groups who consist of people who started out changing printer ribbons and "worked their way up." They learned whatever they needed in their environment, painfully. That is they learned all the quirks of MS stuff slowly over time.
Reality is, that many organizations know only MS. and the people with the biggest "investment" in MS are those who spent the last ten years learning the foibles. The IT staff are the biggest obstacle to widespread adoption of Linux & OSS. They will estimate huge training costs because they need to get the same level of expertise that they themselves have acquired with MS.
It is painful to watch, but true. I don't know what the solution is. It is chicken and egg.
Various proprietary UNIXes (ie Cray, NEC) had proper batch systems ten years ago. The basic hooks are a good thing, and could be used for many modern uses:
standardized user accounting, groups, and accounts. grouping of processes by job-id (so that you cannot just daemonize your running job, to change your process group, and avoid being killed at job end.) The ability to asssign different scheduling priorities and limits to jobs (not users or accounts.) The limits should include the normal cpu time and memory, but also cpu count, and things like residency time.
Once you have hooks for the above, you will have decent support from various workload managers: NQS, Sun Grid Engine, Platform, PBS, IBM WLM, etc...
Slashdot Mirror
Who the heck is going to have experience with
the product, given that it isn't even beta yet?
Is there any benefit on the licensing side?
I understood there was no reduced cost for any licenses for virtual machines. MS requires you have a legal copy of the OS for each virtual machine.
(It's a voice over in the backgound)
sorry, had to...
Yes, latency and throughput are related, but
The point about TCP windows is likely bogus for this application. Most modern TCP implementations include the window scaling option, which will allow scaling to quite high data transfer rates. At these low data rates (a few megabits, or even lower for games) the windows are unlikely to cramp your style (by limiting bandwidth) One usually wants bigger windows for high volume transfers (say > 10 Mbytes/second) that you would see on a LAN.
and be done with it.
built for debian...
In the name of distribution neutrality...
it uses an open package, portable format "OpenPKG", which installs a parallel build environment (it's own gcc, binutils, etc...)
beside the linux one, packages for OpenPKG are RPM 4 based.
sounds quite painful to install.
Because "dangerous" can mean a lot of different things.
To me, dangerous is leaving this sort of activity to the private sector, with no limitations on what they can look at, or who they can share the information with, or why they are looking, or any requirement for a decent audit trail. For me, dangerous is knowing that people who really want to find the dirt will do it anyways, but if the act of looking up information is itself tracked, then there is a chance of tracking the trackers. If we just put our heads in the sand and say that the goverment must not do this, then the private sector will supply the demand, with far fewer strings attached.
Second, you seem to think that the moment a pattern is detected, the SWAT teams will surround the house. I don't think any sane implementationof this sort of data mining would work that way. The "hit" would end up on somebody's desk, then they start looking for other corroborations, maybe they'll spy on the person for a while. Sound terrible? Well, what about how arabs are practically strip searched at the border now, and how visiting Canadians, living in Canada for many years, but of arab origin, are getting deported to lovely vacation spots like Syria. Case 1. Treatment of Arabs. I prefer targeted observation over blacklisting and harassment of entire groups.
It's impossible to do, without simply annihilating the Constitutionally-guaranteed rights of so many people that you end up giving up the very thing you're fighting to protect.
You say that if someone is listening, then that in itself chills speech. If no-one is listening, what point is there in speaking in the first place? The point of free speech is freedom from harassment having made "dangerous" declarations. George Washington did not rally the states' population without letting any pro-Empire folk know what he was onto. If powerful people want to take someone out, they can hire spies, gather dirt, and then generate evidence for prosecution for a non-crime, just look at the Monica fiasco. It is not what you know: "Clinton indulged in a petty peccadillo, then lied about it to Congress." That is normally a criminal offense, but common sense prevailed, and the consensus seems to have been: "Lying to Congress about your sex life, which is none of their business, is no big deal." So Clinton survived just fine.
Second, how do you differentiate between the one George Washington, and the dozens of Timothy Macveighs, or Bin Ladens, The problem is not what those people say, it is what they do. If George Washington were arrested (after all, he moved around in plain sight), many people would rush to his aide, and the Brits would not have been able to keep him, regardless of what they knew. When MacVeigh was exposed (too late!), the consensus was such that the government could punish the individual without undue objections being raised by the population. The difference between an authoritirian state and a democracy lies not in what the state knows about it's population, but what it is allowed to do about it. The same goes for government spooks. It is not what they know that is important, but what they can do with the knowledge. As soon as they make a move to accuse someone, the audit trail has to become available, along with any information gathered.
When I refer to oversight, I do not mean a small number of insiders. I mean full FOIA access. I mean the right for accused to see who was gathering information on them, once they are accused of something. You presented 1984 as an argument that oversight is in-effective. In 1984, there was no free press, a single political party, and no right to free speech. That doesn't remind me of the Western world today.
A stable multi-party political system is critical, a free press is critical, a rational appreciation of modern technology is critical. It is be
hello? They're politicians! They already are stalked that way, and watched by hoards of reporters looking for the slightest gaff. For them, it is just justice for everyone to be tracked the way public figures are.
Question is: why do we think it is important to have the nation aware of blue dresses and activities between consenting adults in private, but not important to be able to look for patterns in the behaviour of large numbers of people to find the 1 in a million dangerous types?
If we don't engage in far more aggressive information gathering, then we are stuck with mindless random checks based on race i.e. "If you're arab, you might just be a terrrorist!" sticking entire groups in the "suspect" category. That is not right. It will be very hard to do properly, but it is just plain dumb not to try.
Privacy should not about hoarding or hiding information. It needs to be about managing it ethically and responsibly. Sure they track us. But in there is the audit trail of who did the tracking, and why.
The privacy of individuals is a lost cause.
There is video along the highways, in stores, automated tellers, at work, and in web cams in a lot of homes. Facial recognition is coming along. within 10 years, the "personal information" companies will be exchanging will be your complete daily routine. "Data mining" will mean being able to lookup where every person was every moment of the day.
Law enforcement will have access. It would be too stupid for them not to have access.
It is not a question of whether you think this future is good or bad. It will be too cheap for it not to happen.
Information wants to be free. In the future, your life will be free too.
The question is how are we going to deal with it.
A transparent multi-national conglomerate (ie. WorldComm, Enron) is necessary for people who own it (the stockholders) to control it. How else do we evaluate how our employees are doing ?
no need to single out governments...
This isn't the same species, but how about swimming for 18 days at a time, and diving to 500m. That doesn't sound terribly idle. (http://www.penguins.cl/king-penguins.htm)
For this species, they migrate from Tierra Del Fuego to Brazil. That's got to be a decent swim.
http://animaldiversity.ummz.umich.edu/acco
Or how about "The satellite data have already revealed that some penguins swim up to 300 miles from their nesting area to find food while their mates sit in the nests on their eggs. Such a foraging journey can take nearly three weeks, leaving the penguins' newborn chicks at risk of starving before the parent returns. This is particularly worrisome since the number of Magellanic penguin chicks surviving to adulthood has declined in recent years."
from
ttp://www.artsci.washington.edu/ne
What's the big deal?
I'd rather have them swim from dawn to dusk
as they would in the wild, than sit there and gawk at the tourists. They need the exercise. This is probably the best thing that could happen to them. Mind you, I bet they're costing the zoo a fortune in herring.
never used it. it's been around for many years, used to be a SCO app (so it's politcally a propos
http://www.linuxcanada.com/pos.html
you can download evaluation rpms for free.
wondrous google is.
http://www.google.com/linux?hl=en&lr=&ie=ISO
gives...
seismologists' tools on linus is: http://www.seismo.unr.edu/ftp/pub/ichinose/LINUX/
A cool article about an linux-based instrument for use in U-2 spy planes.
http://www.linuxdevices.com/articles/AT6
A linux journal article about linux in a lab.
http://www.linuxjournal.com/article.php?sid
which, in the resources link, includes a link to
the "Scientific Applications on Linux" page...
http://SAL.KachinaTech.com/index.shtml
Spammers
will just team up with hackers with a few thousand machines that are "owned" to do the encryption, and then have those machines send out the mail via a few open relays, so they can keep their stable of machines "safe."
the grid: It isn't just for finding aliens any more.
What you're describing is a rural condo.
You mentioned the group is used to working together, but it is probably pressing your luck to assume everyone will get along well forever. You probably want individual "star" topology media going to a central point which should be a locked up place with only the equipment (telco termination + connections to the house) in it. The shack or bunker should be powered independently from any of the units (Joe sells, and the new guy is a luddite who wants no bandwidth, and finds all that equipment gobbles up power. The last thing you want is to have A cutting of B because A doesn't want to pay for bandwidth, but B is behind him, or "A."
extract from a sendmail.mc :-)
INPUT_MAIL_FILTER(`spamassassin', `S=local:/var/run/spamassassin.sock, F=, T=C:15m;S:4m;R:4m;E:10m')
INPUT_MAIL_FILTER(`mimedefang', `S=unix:/var/spool/MIMEDefang/mimedefang.sock, F=T, T=S:60s;R:60s;E:5m')
define(`confPRIVACY_FLAGS', `authwarnings,novrfy,noexpn,restrictqrun')dnl
I work in a large organization. The people with the biggest investment in MS are the techies themselves, and they (being basically self-taught) are loathe to consider the unknown. There is a pocket of technical people who use linux, but the vast majority on the business side of things are fully endroned. Unless the whole technical support organization is on side, it will not happen. So some sort of gradual introduction and familiarization is needed.
In this guy's case, I assume he is IT, so I figured it might be OK.
sounds like the answer to your question is no.
Sure you can try a piecemeal approach.
But it is probably more effective to go the whole
9 yards. for twenty people, switch the clients over
to linux as well. Otherwise, you're going to be spending plenty on MS-Office & Virus software per year. There will be an initial pain of transition, but after the hump you will be way better off.
only if you went to the bathroom during commercials, but that's another story.
Many, many organizations have IT groups who consist of people who started out changing printer ribbons and "worked their way up." They learned whatever they needed in their environment, painfully. That is they learned all the quirks of MS stuff slowly over time.
Reality is, that many organizations know only MS. and the people with the biggest "investment" in MS are those who spent the last ten years learning the foibles. The IT staff are the biggest obstacle to widespread adoption of Linux & OSS. They will estimate huge training costs because they need to get the same level of expertise that they themselves have acquired with MS.
It is painful to watch, but true. I don't know what the solution is. It is chicken and egg.
Data recovery. It's been a month since the last
backup,... The dog chewed on the hard drive and my paper's due tommorrow....
People probably need to get the data off the drive to copy it onto the new one.
Various proprietary UNIXes (ie Cray, NEC) had proper batch systems ten years ago. The basic hooks are a good thing, and could be used for many modern uses:
standardized user accounting, groups, and accounts. grouping of processes by job-id (so that you cannot just daemonize your running job, to change your process group, and avoid being killed at job end.) The ability to asssign different scheduling priorities and limits to jobs (not users or accounts.) The limits should include the normal cpu time and memory, but also cpu count, and things like residency time.
Once you have hooks for the above, you will have decent support from various workload managers: NQS, Sun Grid Engine, Platform, PBS, IBM WLM, etc...
The outlook for this thread is grim.
Hopefully in will fade to invisio soon.
They've just standardized HDTV for cable.
HDTV's still do not have digital outputs, and
VCR's & DVD's do not have completely digital output yet.
Another few years, it should settled out.
So if the air is denser, then it should have more thermal mass, and should provide better cooling.
Anybody tried OC'ing in one of those chambers?