The problem is that global warming is not a yes or no question. All sides agree on the basic facts:
There is medium term warming since 1700s.
Human activity is having an effect.
Natural causes and cycles (solar cycle, 1200 year cycle) are having an effect.
The disagreement is on the weight to assign to natural and human causes. Worse, activists on both sides try to pretend that it is *all* human activity or *all* natural causes.
In my opinion, whether the cause is primarily human, or natural, it is pretty much a done deal. We can expect to reach temperatures at least as high as the medieval period in the next few hundred years, more depending on the extent of human influence. Instead of bickering, we should be making long term migration plans. Places like Netherlands and Florida might not be good long term real estate investments. Places like Siberia and Northern Canada might pay off. Looking at what is known about local climates in the medieval temperature maximum would be a good start.
How can such migrations be handled equitably? If the area is currently barren, a homestead policy might be effective. But there will be unforeseen shifts in climate as well. Deserts may bloom. Farmland may become desert.
What security is worth
on
Hack IIS6 Contest
·
· Score: 5, Insightful
If the bounty is an Xbox, that means that IIS6 security is robust enough to protect assets worth up to about $200.
I'm probably naive, not living there to see the problems, but I am impressed with the welfare law. As a conservative, I prefer private charity, but the Mutual Obligation feature of AU state charity sounds like it captures the true spirit of constructive charity (as opposed to tying dole payments to the number of illegitimate children like we did here in the states until recently). When providing private charity to an individual, that is exactly the expectation I have. It also reflects the successful stories of private charity passed down from the Great Depression.
Protection rackets have territories. You pay whoever currently controls your territory. If a competing salesman comes by, you let your current "protector" know, and they duke it out. You keep paying the winner.
The description was admittedly brief, but read it again. There are multiple smart cards, any one of which can sign something. You could have one for each developer. Or one for each key developer.
Basically, you need 1 of the n2 smart cards and k1 of the n1 developers to sign.
Share secret between m1 developers and m2 smart cards (like iButton/Java Ring). The signing key is shared between the smart cards, and one developer key. k1 of the m1 developers can reconstruct the developer key. The signing key is reconstructed with k = 2 from the developer key and any smart card key. It exists momentarily only inside the secure smart card memory, and signing takes place inside the smart card.
So to sign, requires cooperation of k1 developers, and 1 smart card.
Several months ago, we made the mistake of testing a new webmail server using AOL, but forgetting to actually add the DNS record first:-). The negative result is *still* cached at AOL. Bummer for users trying to use the webmail.
On the plus side, I've used AOL to find out what the IP of names *used* to be while researching problems. Kind of handy that way.
The device reads 128K byte pages - as opposed to the typical 512 byte sectors common with magnetic disks and 2048 byte sectors common with CD-ROM.
For WORM applications, this is not that big a deal. However, for R/W applications, some serious file system and virtual memory redesign is needed.
Not to worry - these holo drives wear out quickly with repeated rewriting just like CD-RW, so they are not providing paging space anytime soon. But it is fun to think about.
Publishing something does not protect your work from being exploited by predatory patents thanks to the USPTO not caring about prior art. True, in theory you could challenge the bogus patent in court. But who has the money to do that, when you just wanted to give it away? There is nothing more maddening than getting a cease and desist letter telling you to stop using something you invented.
UTC is locked to TAI except with leap seconds added at certain well-defined intervals. ... The designers of POSIX, BSD, and/or Mach were apparently unable to comprehend this.
So what you are saying is that the "GMT" label is incorrect, and should be
"UTC"? A fair point, but lighten up. The Posix library is designed to
convert a monotonic time value to common localtime formats - not to do
astronomy. But this is from Java - with homebrew extensions for leapseconds.
That is not the only labelling problem with the Java Timezone code, in
addition to hardwiring the base timezone label to "GMT" (even though that
is incorrect when leapsecond corrected), Sun made some very bad
assumptions in their Calendar and DateFormat classes:
Standard time is always precisely equal to base offset: never true
with leap seconds, so label is always "?DT".
There are never more than 2 local variations, so EWT never gets displayed.
I hand corrected the EST label, but forgot about the GMT thing.
There is no point in doing the atomic time thing for most computer clocks,
but with nearly half a minute of leap seconds, and NTP accurate to within a
few hundred milliseconds, I wish library writers would account for them - at
least optionally.
Perhaps you can answer a question for me about NTP. When the seconds clock
for unix is synchronized to NTP, does that make the default GMT (no leap
seconds) conversion display the "correct" time? Or would a UTC (leap second
corrected) convertsion display the "correct" time? In other words, does
NTP skip/add seconds to keep default unix time conversions displaying the right
value?
If this stuff leaks into the environment, what will happen?
One thing that has always irked me about Sucralose proponents, is that after
assuring me that (almost?!) all of the stuff goes right through your system,
they only have blank stares when I ask what effect all that Sucralose getting
flushed down millions of toilets is going to have. Anyone have a pointer to
even a premininary study on the environmental half life of the stuff, and maybe
add some to a pond to see if anything happens?
At first I thought that OSX might be doing leap seconds. But zoneinfo
seems to think that DST switch occurs at 2am localtime, and there are
more than 7 leap seconds.
right/US/Eastern 4/03/2005 7:00:21 GMT = April 3, 2005 1:59:59 AM EST right/US/Eastern 4/03/2005 7:00:22 GMT = April 3, 2005 3:00:00 AM EDT
Excellent questions. In addition to John Q. Public, I also like to see
questions about the impact on Paoulo Primitive (who lives a primitive
lifestyle whether by choice or circumstance). I hate technologies that "burn
our bridges", and make the "live off the land" disaster fallback option more
difficult, or even impossible. If the YellowStone Supervolcano blows in your
lifetime, you'll be sorry for all the times we screwed the Native Americans,
Eskimos and RainForest dwellers.
Oh, concerning YellowStone, Native Americans in the Arizona area experienced
several inches of Volcanic ash in early AD. This dramatically increased the
fertility of the soil - and their crops once they discovered that leaving the
ash layer in place and planting seed underneath was the best practice. The
ash layer kept the ground warm and compensated for the cooling from the
"nuclear" winter - sort of a super mulch. It might be worth making some contingincy plans for a
YellowStone eruption, including compiling the experiences of survivors of
similar events in the past.
...and you can delete any electronic copies that may have been made as well.
A quick review for those not familar with "trusted" computing. The hardware uses digital signatures to enforce running an approved BIOS only, which in turn enforces running an approved OS, which in turn will only run approved applications. Documents are encrypted, and the approved applications can phone home to determine whether you are allowed to read a document. If the document is on a delete list, it is immediately erased. Microsoft Media Player already implements this system - except for the hardware enforcement. Microsoft Office is next. Evil Media companies, and Microsoft, want to make the hardware enforcement required by law on all computing devices.
In the not too distant future, having obtained a copy of an incriminating document, you could keep it stored on a banned Linux system running on illegal hacked hardware, and given Microsoft's expertise with security, probably crack the encryption in a reasonable amount of time due to some stupid design flaw (e.g. random seed for session key is derived from Document time stamp). However, the resulting evidence would not be admissable in court. So stock up on tin foil hats.
In our small 5 person company, we are very cash strapped. We wanted to try RHEL in a production setting to make sure all our software works with it, etc. Enter CentOS. We tried it out for a $50 donation (about what we would have paid for a RedHat CD in the old days). Having installed some guinea pigs, and solved the problems (e.g. Sun/IBM Java 1 doesn't run, so memory must be upgraded to handle bloated Java 2), we can now recommend RHEL with confidence.
A big problem we have with RedHat is confusing contracts and pricing.
since they will have to pay for development twice: Once to create a new technology, and again to pay patent royalties to patent leeches that come along and patent it afterward.
Don't tell me about prior art - it doesn't stop the USPTO from granting patents, and it doesn't help overturn them unless you spend more than the royalties on lawyers.
I'm not sure that the actual research is changing. It's just that instead of it being public, it is being kept as military "trade secrets" (i.e. classified) so they don't have to deal with the patent leeches.
will say, "look there! Those open source loonies *do* care about IP when their own IP is getting stolen." (Never mind that this is about Copyright, not stupid Patents.)
Oh wait, maybe we *should* care about RIAA copyrights, even if they are a$$#01es.
I did. My parents did. My wife's parents did. My brothers and sisters did. My aunts and uncles did. My cousin got raped, so that doesn't count. The kids in my college Christian club that were virgins before coming to Christ stayed that way. The "born again virgins" stayed that way also.
Not only can people be drug-free, but they can live their lives both sexually pure and sexually fulfilled. Our current culture doesn't help, but for those with the backbone to swim against the tide, it can be done. The spiritual danger is in becoming proud and self-righteous, thinking you're better than others because you avoided some particularly self destructive sins.
The only safe sex is married sex.
Re:My undocumented research...
on
Contrabandwidth
·
· Score: 1
One thing communes in America have in common: you can leave if you don't like it. There might be financial penalties (since you gave them property when you became an official member), but you can leave and start over. Cults in America often have social barriers to leaving. Members are made to be totally dependent emotionally and socially on the cult, and suffer great emotional distress when they leave. Cuba has physical and legal barriers to leaving - giving it some of the oppressive nature of a cult. I'll be more inclined to call Cuba a success when people aren't forced to stay there. (I know, sometimes the US seems to be part of the physical and legal force keeping people there.)
Also, is Castro ever going to retire and appoint a successor? Will there be a formal system for choosing successors? Hereditary? Appointment by previous leader? Will they all be lifelong office holders? Are there any provisions for removing a leader gone truly haywire (like the "Apostle")? Short of "regime change" by the US?
My undocumented research...
on
Contrabandwidth
·
· Score: 5, Interesting
...when I was fresh out of college might prove interesting. I was very interested in communes, and collectives, and visited several, and researched more. I found many long term and successful communes, and read about many failed ones". This was a personal interest, and I didn't think to document any of this, so take my conclusion with a grain of salt.
Conclusion: communes work well when they are small enough. Rule of thumb based on extended visits: communes where every member has personal contact with the current leader, and where the leader is replaceable, work well. The further removed the least prominent members are from the leaders - the more dicatorial and cult-like the commune becomes. Cuba is way past my "works well" limit. Counter-example: small cults with very charismatic leaders (e.g. Jim Jones, Heaven's gate) (although they also fail the "replaceable" test).
One Christian group in Vermont that I spent 2 weeks with had a leader who declared himself (while I was there!) an "Apostle" whose authority cannot be questioned of removed. Fortunately, the group was under oversight by an international board with members from every subscribing commune. The "Apostle" was removed.
Interestingly, while every group I visited was a "commune" in the sense that all property was legally owned by the corporate entity, a huge part of what made them tick was "ownership" of a different sort. The head chemist at the Vermont group (in charge of making soap and perfumes) was so excited about his products and workmanship that he "owned" his unique position in the commune. One of the things the "Apostle" talked about doing was switching around jobs so that everyone would learn "humility". Kind of reminded me of the part in the Communist Manefesto were everone is supposed to be able to do anyone elses job.
This is where I began to see that there is a big difference between "commune" and "Communism". Classic Communism as a political philosophy is bunk. However you handle production and distribution of goods, every person is unique, and brings unique gifts and talents to an organisation. A political philosophy that tries to make everyone interchangeable is just another means of oppression. Capitalism becomes oppressive in the same way when it tries to make every employee an interchangeable cog in the machine.
RedHat 9, LTSP 3.0, Gnome 2.2.
Keep in mind, that the X servers do not run on the server (or rather, only the console X server does). LTSP is superb! It can even run simple arcade games like breakout on the thin clients (but not 3D games like Tux racer). Here are the biggest memory users from top from the system at this moment with 3 active users (including me on an LTSP thin client):
I want my system to be robust against buggy programs with:
Infinite loops
fast memory leaks
Infinite recursion
I/O hogging
all of the above
Not because of security, but because I've personally made all those mistakes at one time or another, and hate having the system grind to a halt - preventing me from diagnosing the problem.
I like limiting things, especially memory, to prevent this.
If you only run, not develop, applications, your priorities might be different. The defaults for RedHat 9, at any rate, seem to be geared toward the user, not the developer. This is fine - it just means that I need to tune the defaults.
My point was that neither system is broken (wasn't that your point?). You just need to tune the defaults if they don't match your usage.
My machine is shared between 4 LTSP workstations and has only 385M ram. There is no 3D graphics card, and no UT (it *does* have Railroad Tycoon - works fine on LTSP). I don't want a buggy app taking down the whole system. For a single user game machine, ulimited physical memory could make sense. It is no worse than running Windows should you encounter a program with a fast memory leak.
It's not BSD that has the problem (with default settings), it is linux. It is not reading from/dev/zero that is the problem, it is the unlimited memory allocated by grep while reading a logically infinite file. I can be argued that this is a bug in grep (and I would agree), but memory leaks should not crash your system.
The default for memory is unlimited, which does indeed create a DOS "attack" for "grep bomb" and other inadvertant application bugs.
This is a case of bad defaults, not a kernel problem. I recommend a max physical memory of no more than 1/4 physical memory, preferrably less.
AIX also has cruddy defaults, but ulimit -m limits physical RAM, not virtual RAM. That way, a single process with run-away memory use will just start swapping like crazy and let the rest of the system keep running. Of course, even then a dozen or so such processes will still bring the system to a crawl. I would like to see physical RAM limited by user id.
Slashdot Mirror
- There is medium term warming since 1700s.
- Human activity is having an effect.
- Natural causes and cycles (solar cycle, 1200 year cycle) are having an effect.
The disagreement is on the weight to assign to natural and human causes. Worse, activists on both sides try to pretend that it is *all* human activity or *all* natural causes.In my opinion, whether the cause is primarily human, or natural, it is pretty much a done deal. We can expect to reach temperatures at least as high as the medieval period in the next few hundred years, more depending on the extent of human influence. Instead of bickering, we should be making long term migration plans. Places like Netherlands and Florida might not be good long term real estate investments. Places like Siberia and Northern Canada might pay off. Looking at what is known about local climates in the medieval temperature maximum would be a good start.
How can such migrations be handled equitably? If the area is currently barren, a homestead policy might be effective. But there will be unforeseen shifts in climate as well. Deserts may bloom. Farmland may become desert.
If the bounty is an Xbox, that means that IIS6 security is robust enough to protect assets worth up to about $200.
I'm probably naive, not living there to see the problems, but I am impressed with the welfare law. As a conservative, I prefer private charity, but the Mutual Obligation feature of AU state charity sounds like it captures the true spirit of constructive charity (as opposed to tying dole payments to the number of illegitimate children like we did here in the states until recently). When providing private charity to an individual, that is exactly the expectation I have. It also reflects the successful stories of private charity passed down from the Great Depression.
Protection rackets have territories. You pay whoever currently controls your territory. If a competing salesman comes by, you let your current "protector" know, and they duke it out. You keep paying the winner.
Basically, you need 1 of the n2 smart cards and k1 of the n1 developers to sign.
So to sign, requires cooperation of k1 developers, and 1 smart card.
On the plus side, I've used AOL to find out what the IP of names *used* to be while researching problems. Kind of handy that way.
For WORM applications, this is not that big a deal. However, for R/W applications, some serious file system and virtual memory redesign is needed.
Not to worry - these holo drives wear out quickly with repeated rewriting just like CD-RW, so they are not providing paging space anytime soon. But it is fun to think about.
Publishing something does not protect your work from being exploited by predatory patents thanks to the USPTO not caring about prior art. True, in theory you could challenge the bogus patent in court. But who has the money to do that, when you just wanted to give it away? There is nothing more maddening than getting a cease and desist letter telling you to stop using something you invented.
...
The designers of POSIX, BSD, and/or Mach were apparently unable to comprehend this.
So what you are saying is that the "GMT" label is incorrect, and should be "UTC"? A fair point, but lighten up. The Posix library is designed to convert a monotonic time value to common localtime formats - not to do astronomy. But this is from Java - with homebrew extensions for leapseconds. That is not the only labelling problem with the Java Timezone code, in addition to hardwiring the base timezone label to "GMT" (even though that is incorrect when leapsecond corrected), Sun made some very bad assumptions in their Calendar and DateFormat classes:
- Standard time is always precisely equal to base offset: never true
with leap seconds, so label is always "?DT".
- There are never more than 2 local variations, so EWT never gets displayed.
I hand corrected the EST label, but forgot about the GMT thing.There is no point in doing the atomic time thing for most computer clocks, but with nearly half a minute of leap seconds, and NTP accurate to within a few hundred milliseconds, I wish library writers would account for them - at least optionally.
Perhaps you can answer a question for me about NTP. When the seconds clock for unix is synchronized to NTP, does that make the default GMT (no leap seconds) conversion display the "correct" time? Or would a UTC (leap second corrected) convertsion display the "correct" time? In other words, does NTP skip/add seconds to keep default unix time conversions displaying the right value?
One thing that has always irked me about Sucralose proponents, is that after assuring me that (almost?!) all of the stuff goes right through your system, they only have blank stares when I ask what effect all that Sucralose getting flushed down millions of toilets is going to have. Anyone have a pointer to even a premininary study on the environmental half life of the stuff, and maybe add some to a pond to see if anything happens?
Oh, concerning YellowStone, Native Americans in the Arizona area experienced several inches of Volcanic ash in early AD. This dramatically increased the fertility of the soil - and their crops once they discovered that leaving the ash layer in place and planting seed underneath was the best practice. The ash layer kept the ground warm and compensated for the cooling from the "nuclear" winter - sort of a super mulch. It might be worth making some contingincy plans for a YellowStone eruption, including compiling the experiences of survivors of similar events in the past.
A quick review for those not familar with "trusted" computing. The hardware uses digital signatures to enforce running an approved BIOS only, which in turn enforces running an approved OS, which in turn will only run approved applications. Documents are encrypted, and the approved applications can phone home to determine whether you are allowed to read a document. If the document is on a delete list, it is immediately erased. Microsoft Media Player already implements this system - except for the hardware enforcement. Microsoft Office is next. Evil Media companies, and Microsoft, want to make the hardware enforcement required by law on all computing devices.
In the not too distant future, having obtained a copy of an incriminating document, you could keep it stored on a banned Linux system running on illegal hacked hardware, and given Microsoft's expertise with security, probably crack the encryption in a reasonable amount of time due to some stupid design flaw (e.g. random seed for session key is derived from Document time stamp). However, the resulting evidence would not be admissable in court. So stock up on tin foil hats.
A big problem we have with RedHat is confusing contracts and pricing.
Don't tell me about prior art - it doesn't stop the USPTO from granting patents, and it doesn't help overturn them unless you spend more than the royalties on lawyers.
I'm not sure that the actual research is changing. It's just that instead of it being public, it is being kept as military "trade secrets" (i.e. classified) so they don't have to deal with the patent leeches.
Oh wait, maybe we *should* care about RIAA copyrights, even if they are a$$#01es.
Not only can people be drug-free, but they can live their lives both sexually pure and sexually fulfilled. Our current culture doesn't help, but for those with the backbone to swim against the tide, it can be done. The spiritual danger is in becoming proud and self-righteous, thinking you're better than others because you avoided some particularly self destructive sins.
The only safe sex is married sex.
Also, is Castro ever going to retire and appoint a successor? Will there be a formal system for choosing successors? Hereditary? Appointment by previous leader? Will they all be lifelong office holders? Are there any provisions for removing a leader gone truly haywire (like the "Apostle")? Short of "regime change" by the US?
Conclusion: communes work well when they are small enough. Rule of thumb based on extended visits: communes where every member has personal contact with the current leader, and where the leader is replaceable, work well. The further removed the least prominent members are from the leaders - the more dicatorial and cult-like the commune becomes. Cuba is way past my "works well" limit. Counter-example: small cults with very charismatic leaders (e.g. Jim Jones, Heaven's gate) (although they also fail the "replaceable" test).
One Christian group in Vermont that I spent 2 weeks with had a leader who declared himself (while I was there!) an "Apostle" whose authority cannot be questioned of removed. Fortunately, the group was under oversight by an international board with members from every subscribing commune. The "Apostle" was removed.
Interestingly, while every group I visited was a "commune" in the sense that all property was legally owned by the corporate entity, a huge part of what made them tick was "ownership" of a different sort. The head chemist at the Vermont group (in charge of making soap and perfumes) was so excited about his products and workmanship that he "owned" his unique position in the commune. One of the things the "Apostle" talked about doing was switching around jobs so that everyone would learn "humility". Kind of reminded me of the part in the Communist Manefesto were everone is supposed to be able to do anyone elses job.
This is where I began to see that there is a big difference between "commune" and "Communism". Classic Communism as a political philosophy is bunk. However you handle production and distribution of goods, every person is unique, and brings unique gifts and talents to an organisation. A political philosophy that tries to make everyone interchangeable is just another means of oppression. Capitalism becomes oppressive in the same way when it tries to make every employee an interchangeable cog in the machine.
- Infinite loops
- fast memory leaks
- Infinite recursion
- I/O hogging
- all of the above
Not because of security, but because I've personally made all those mistakes at one time or another, and hate having the system grind to a halt - preventing me from diagnosing the problem. I like limiting things, especially memory, to prevent this.If you only run, not develop, applications, your priorities might be different. The defaults for RedHat 9, at any rate, seem to be geared toward the user, not the developer. This is fine - it just means that I need to tune the defaults.
My point was that neither system is broken (wasn't that your point?). You just need to tune the defaults if they don't match your usage.
My machine is shared between 4 LTSP workstations and has only 385M ram. There is no 3D graphics card, and no UT (it *does* have Railroad Tycoon - works fine on LTSP). I don't want a buggy app taking down the whole system. For a single user game machine, ulimited physical memory could make sense. It is no worse than running Windows should you encounter a program with a fast memory leak.
It's not BSD that has the problem (with default settings), it is linux. It is not reading from /dev/zero that is the problem, it is the unlimited memory allocated by grep while reading a logically infinite file. I can be argued that this is a bug in grep (and I would agree), but memory leaks should not crash your system.
This is a case of bad defaults, not a kernel problem. I recommend a max physical memory of no more than 1/4 physical memory, preferrably less.
AIX also has cruddy defaults, but ulimit -m limits physical RAM, not virtual RAM. That way, a single process with run-away memory use will just start swapping like crazy and let the rest of the system keep running. Of course, even then a dozen or so such processes will still bring the system to a crawl. I would like to see physical RAM limited by user id.