If you run djbdns, you don't need IXFR. You can rsync or scp the data file. IPV6 is a missing feature and I hope, now that the software is in the public domain, that someone fixes it.
As of December, the software is in the public domain. No licensing issues. If you would like to take on a distribution that has been patched, please feel free.
DJBDNS actually IS superior. It is lighter weight and more secure. And it has addressed this issue as best you can for more than 5 years.
porsche is the most profitable car company in the world and they do not do mass market sales. they have some lower-end models, but total volume shipments of 30,000 cars/year does not make you mass market.
mass market is not necessarily the real money - obviously, a bigger market is better, but you can also make a lof money in a niche market.
i realize that the current market is setup (with phone subsidies) so the carriers have incentive to keep you from using phones with another carrier. but is there any technological barrier? is there any reason that cdma cell phones couldn't be paired with SIM cards?
no, there is no technological barrier. further, some people believe that the importance of supporting next-generation provisioning and wi-fi/3g roaming will lead more of the carriers to support gsm on their networks. you could easily support the gsm provisioning/billing/roaming features on top of a cdma transport. in fact, i believe that some cdma phones with gsm/tdma chipsets built-in for global roaming have been announced/discussed.
it will be interesting to see how long verizon can maintain this technological provincialism. based on their dominance in the marketplace, i would imagine they'll be able to maintain for some time.
I agree - ESR is a lunatic gun nut, and not a particularly good hacker. Eric sure didn't need the GPL for CML2. And does anybody else remember a few years ago when he got into a fight with someone over a project they were working on, and threatened to kill him?
He sure has an eye for the spotlight, though - almost as good as John "Mad Dog" "Never Done Much But Show Up For the Victory Parade" Hall.
actually, Diebold was updating software on the voting machines well after the certification and very close to the time of election in CA in the primaries for the 2004 election.
That situation and scenario has since been corrected - Diebold can/will no longer update any equipment once written certification is received from the sec. of state for each state.
As a former sceptic (my company has a financial interest in insuring the accuracy of those machines) I can tell you that 1. DRE is the future and 2. the machines are as good as anything we have now.
The reason for 1. is simple - HAVA and the ADA. The Help America Vote Act sets out $4bn for US states to modernize their voting systems, but requires that they be fully ADA compliant to do so. You can't be fully ADA compliant without an electronic voting system (think about the legally blind, the wholly blind, those who don't speak english, those who can't physically access a voting machine, etc.) Diebold's systems allow for multiple languages, aural voting for the blind, and for portable systems to be brought securely to those who cannot access the polling center (physically disabled).
Further, DRE increases accuracy - the most common voting errors are 1. inscrutable ballots 2. undervoting and 3. showing up at the wrong precinct. DRE can make ballots much more legible (color-coded, large type), can reduce undervoting (the machine makes you confirm that you want to skip a vote) and present the correct ballot for your precinct even if you show up at the wrong one.
As for issue 2., security - please note that Diebold's system, even if you don't agree with it, is no less secure than the ones we have now. Do you know what happens to your ballot after you cast it on paper - ballot theft and stuffing is all too commonplace now, and is harder with the electronic systems. Do you have any idea if the machine properly records your vote when you use an electromechanical voting machine- you don't, you just trust it because it is familiar.
You should note that some of the biggest sceptics (Avi Rubin @ Johns Hopkins) have changed their opinions considerably since actually participating in a vote. In 20 years, this will be a funny story in the history books.
well, they are right, there is no such thing as a 911-only phone. if you want a phone number, you need to pay for service.
however, any phone in the united states is required to permit 911 dialing at all times under all circumstances. if you truly only want the phone to work to dial 911, disconnect servce and leave the phones plugged in. in an emergency, pick it up and dial 911 - the call must go through or the phone company is in a lot of trouble.
you won't have a number, but you'll have what you want.
Essentially, the issue with the license (you can't sue Yahoo!) is that if you end up in a dispute with Yahoo! over a patent issue or unrelated issue, you could lose your rights to this software. That violates the spirit and principles of free software, and because it makes Yahoo! into a 900-pound gorilla, is probably not Reasonable and Non-Discriminatory.
If you write software, and you want it to be a standard, you should really have to make the standard totally free. Otherwise, we could end up with another Rambus situation.
it's kind of sad, but these researchers went in to the lab for this. i mean, Uni of Queensland - Wallaby Campus is probably just another school looking to improve its reputation through quality academic research, and that's great, but this project is more fluff than stuff.
it's kind of like that eddie izzard bit about the english reaction to the US moon race in the 60's - Uni of Queensland is committed to putting a man on the surface of the earth and returning him safely to, uh, well, the earth!
or maybe the creators did think about it, but because you're given the fact that you can just jam the frequency and bring the whole thing down, it DOESN'T MAKE ANY DIFFERENCE!
this just in...wireless networks are open to a range of attack vectors generally closed to wired networks...competitive interefence leads to signal degradation and loss of service...film at 11
seriously, and i haven't even read the article yet, what could possibly be the news here. i'm imagining that, what, certain tiny packet sequences have a disproportionately large disruptive impact on the protocol by causing extended resets and delays? how is that any different from the recent tcp packet spoofing attacks except in free space?
it would still be easier to get a big antenna and a transceiver and just blanket the spectrum.
And you won't. You won't find the reinsurance coverage for it and you're balance sheet can't support it.
You're insane. You can't make money in patent insurance. The defense costs are too high, and the losses to large if you lose. That's why no one writes it.
Swiss Re has been known to write the occasional policy. Of course, it's with a $5m retention and indemnity only, and you're generally paying about 40% rate on-line.
Uhh, "northern European reinsurance" companies like Swiss Re and Munich Re pretty much wrote the book on property insurance.
I think you've got it backwards - they arrange their reinsurance programs to mitigate the total exposure, freeing the primary insurer from having to get everything right. If the primary was so confident in their ability to manage the exposures, they wouldn't need as much reinsurance, would they?
What about the fact that the coverage they are offering is already available from most of the largest E&O providers in the market (AIG, ACE, Hiscox, Chubb, Zurich, CNA, etc.) 1. cheaper 2. with higher limits 3. with more experienced claims handling staff 4. not tied to specific best practices and techniques (without which coverage does not attach) and 5. as part of a larger E&O program that will provide more coverage for the more likely problems - bugs.
Seriously, this is really just a lot of fluff at this point. $100k in defense costs and $1m in limits is nothing for large companies that buy $50m - $100m liability towers, and a small company looking for coverage can buy $1m in limits (including software copyright) for a lot less than $30k.
And for the record, Bruce, reinsurance is not a great answer. Just look at the problems the London markets are having getting Swiss Re to pay their claims. For a small, poorly capitalized company like OSRM, more than 3 or 4 losses (which their adverse selection will guarantee) will leave them cash flow negative, unable to continue functioning while they wait for the reinsurers to cut a check.
Yes, but not nearly as many. It's not 'adapting' so much as 'not learning' - the immune system is much more immature and you can keep it from learning that the foreign organs are present.
very old people need fewer drugs as well, as their immune systems are simply weaker.
No, it cannot. Open sourcing software limits your liability, but you are always liable for products that you sell. Disclaimers can help, but they are not the solution.
And since I, too, sell this kind of insurance, I can tell you that it's getting to be a pretty good thing to have if you develop software.
Ok, interesting point, but there is no opportunity to review the source code in proprietary software systems, whereas open source software at least holds out the possibility of peer review. A good example would be the infamous backdoor that Thompson put in the earliest incarnations of the Unix login program. And since we are well aware of the history of backchannel government intervention in proprietary software systems, dating at least as far back as DES, there is no reason to believe that the walls to intervention are any higher; at best, they are just more opaque.
It's been said time and again, but it's an old adage that governments would be well-served to heed is: Security through obscurity is no security at all.
x86 is extremely bloated when compared to MIPS processors. What they need to do is make a RISC or MIPS based processor with x86 backwards compatability, and slowly phase over.
uhh, with all due respect, if we have learned anything in the past thirty years from the success of windows, unix, the as/400 and finally x86, it's that architectures are the hardest thing in the world to change due to the massive installed base, and that it's usually better to extend what you have.
just look at os/2, the MCA bus, and now itanium. why would i migrate to a new ISA and lose all the software that I already have when I can just grow my current one?
and x86 isn't that bloated, and cisc isn't that bad. just look at p4 vs. athlon - the tremendous clock speeds realized by the p4's use of an extended pipeline (which is a risc-like optimization) have a tremendous downside - you lose a lot of time resetting the cache if you miss a branch. so for interative programs, as opposed to massive number crunching (and that can be addressed cheaper using MPP and clustering), risc is something of a dog.
finally, you can't say that the desktop is not important to itanium when the line between servers, workstations, and desktops gets blurrier all the time, and the largest growing segment of the market is the low-to-mid-size server.
high-end servers may carry a premium price and have a higher margin, but like lenin said, quantity has a quality all its own.
Slashdot Mirror
If you run djbdns, you don't need IXFR. You can rsync or scp the data file. IPV6 is a missing feature and I hope, now that the software is in the public domain, that someone fixes it.
http://cr.yp.to/distributors.html
As of December, the software is in the public domain. No licensing issues. If you would like to take on a distribution that has been patched, please feel free.
DJBDNS actually IS superior. It is lighter weight and more secure. And it has addressed this issue as best you can for more than 5 years.
porsche is the most profitable car company in the world and they do not do mass market sales. they have some lower-end models, but total volume shipments of 30,000 cars/year does not make you mass market.
mass market is not necessarily the real money - obviously, a bigger market is better, but you can also make a lof money in a niche market.
i realize that the current market is setup (with phone subsidies) so the carriers have incentive to keep you from using phones with another carrier.
but is there any technological barrier? is there any reason that cdma cell phones couldn't be paired with SIM cards?
no, there is no technological barrier. further, some people believe that the importance of supporting next-generation provisioning and wi-fi/3g roaming will lead more of the carriers to support gsm on their networks. you could easily support the gsm provisioning/billing/roaming features on top of a cdma transport. in fact, i believe that some cdma phones with gsm/tdma chipsets built-in for global roaming have been announced/discussed.
it will be interesting to see how long verizon can maintain this technological provincialism. based on their dominance in the marketplace, i would imagine they'll be able to maintain for some time.
I agree - ESR is a lunatic gun nut, and not a particularly good hacker. Eric sure didn't need the GPL for CML2. And does anybody else remember a few years ago when he got into a fight with someone over a project they were working on, and threatened to kill him?
He sure has an eye for the spotlight, though - almost as good as John "Mad Dog" "Never Done Much But Show Up For the Victory Parade" Hall.
it's constitutionally forbidden. control of elections is ceded to county commissioners, under the appropriate secretary of state for that state.
actually, Diebold was updating software on the voting machines well after the certification and very close to the time of election in CA in the primaries for the 2004 election.
That situation and scenario has since been corrected - Diebold can/will no longer update any equipment once written certification is received from the sec. of state for each state.
As a former sceptic (my company has a financial interest in insuring the accuracy of those machines) I can tell you that 1. DRE is the future and 2. the machines are as good as anything we have now.
The reason for 1. is simple - HAVA and the ADA. The Help America Vote Act sets out $4bn for US states to modernize their voting systems, but requires that they be fully ADA compliant to do so. You can't be fully ADA compliant without an electronic voting system (think about the legally blind, the wholly blind, those who don't speak english, those who can't physically access a voting machine, etc.) Diebold's systems allow for multiple languages, aural voting for the blind, and for portable systems to be brought securely to those who cannot access the polling center (physically disabled).
Further, DRE increases accuracy - the most common voting errors are 1. inscrutable ballots 2. undervoting and 3. showing up at the wrong precinct. DRE can make ballots much more legible (color-coded, large type), can reduce undervoting (the machine makes you confirm that you want to skip a vote) and present the correct ballot for your precinct even if you show up at the wrong one.
As for issue 2., security - please note that Diebold's system, even if you don't agree with it, is no less secure than the ones we have now. Do you know what happens to your ballot after you cast it on paper - ballot theft and stuffing is all too commonplace now, and is harder with the electronic systems. Do you have any idea if the machine properly records your vote when you use an electromechanical voting machine- you don't, you just trust it because it is familiar.
You should note that some of the biggest sceptics (Avi Rubin @ Johns Hopkins) have changed their opinions considerably since actually participating in a vote. In 20 years, this will be a funny story in the history books.
i look forward to our new gtk+ win32 overlords?
That's both a thought for the day, a way of life, and the SSID on the AP in my apt.
Strange that even though it's soooo unfashionable, I still found about 8 leeches hanging off it last week.
well, they are right, there is no such thing as a 911-only phone. if you want a phone number, you need to pay for service.
however, any phone in the united states is required to permit 911 dialing at all times under all circumstances. if you truly only want the phone to work to dial 911, disconnect servce and leave the phones plugged in. in an emergency, pick it up and dial 911 - the call must go through or the phone company is in a lot of trouble.
you won't have a number, but you'll have what you want.
There is a serious objection to "you may not sue Yahoo!" language.
Apache article on patent issues in free software
Essentially, the issue with the license (you can't sue Yahoo!) is that if you end up in a dispute with Yahoo! over a patent issue or unrelated issue, you could lose your rights to this software. That violates the spirit and principles of free software, and because it makes Yahoo! into a 900-pound gorilla, is probably not Reasonable and Non-Discriminatory.
If you write software, and you want it to be a standard, you should really have to make the standard totally free. Otherwise, we could end up with another Rambus situation.
--tkr
actually, something more just occurred to me...
it's kind of sad, but these researchers went in to the lab for this. i mean, Uni of Queensland - Wallaby Campus is probably just another school looking to improve its reputation through quality academic research, and that's great, but this project is more fluff than stuff.
it's kind of like that eddie izzard bit about the english reaction to the US moon race in the 60's - Uni of Queensland is committed to putting a man on the surface of the earth and returning him safely to, uh, well, the earth!
or maybe the creators did think about it, but because you're given the fact that you can just jam the frequency and bring the whole thing down, it DOESN'T MAKE ANY DIFFERENCE!
this just in...wireless networks are open to a range of attack vectors generally closed to wired networks...competitive interefence leads to signal degradation and loss of service...film at 11
seriously, and i haven't even read the article yet, what could possibly be the news here. i'm imagining that, what, certain tiny packet sequences have a disproportionately large disruptive impact on the protocol by causing extended resets and delays? how is that any different from the recent tcp packet spoofing attacks except in free space?
it would still be easier to get a big antenna and a transceiver and just blanket the spectrum.
move along, nothing to see here.
i hear, but GenRe still seems to have figured out how to make it work...
thank god someone around here gets it.
And you won't. You won't find the reinsurance coverage for it and you're balance sheet can't support it.
You're insane. You can't make money in patent insurance. The defense costs are too high, and the losses to large if you lose. That's why no one writes it.
Swiss Re has been known to write the occasional policy. Of course, it's with a $5m retention and indemnity only, and you're generally paying about 40% rate on-line.
Uhh, "northern European reinsurance" companies like Swiss Re and Munich Re pretty much wrote the book on property insurance.
I think you've got it backwards - they arrange their reinsurance programs to mitigate the total exposure, freeing the primary insurer from having to get everything right. If the primary was so confident in their ability to manage the exposures, they wouldn't need as much reinsurance, would they?
What about the fact that the coverage they are offering is already available from most of the largest E&O providers in the market (AIG, ACE, Hiscox, Chubb, Zurich, CNA, etc.) 1. cheaper 2. with higher limits 3. with more experienced claims handling staff 4. not tied to specific best practices and techniques (without which coverage does not attach) and 5. as part of a larger E&O program that will provide more coverage for the more likely problems - bugs.
Seriously, this is really just a lot of fluff at this point. $100k in defense costs and $1m in limits is nothing for large companies that buy $50m - $100m liability towers, and a small company looking for coverage can buy $1m in limits (including software copyright) for a lot less than $30k.
And for the record, Bruce, reinsurance is not a great answer. Just look at the problems the London markets are having getting Swiss Re to pay their claims. For a small, poorly capitalized company like OSRM, more than 3 or 4 losses (which their adverse selection will guarantee) will leave them cash flow negative, unable to continue functioning while they wait for the reinsurers to cut a check.
Yes, but not nearly as many. It's not 'adapting' so much as 'not learning' - the immune system is much more immature and you can keep it from learning that the foreign organs are present.
very old people need fewer drugs as well, as their immune systems are simply weaker.
No, it cannot. Open sourcing software limits your liability, but you are always liable for products that you sell. Disclaimers can help, but they are not the solution.
And since I, too, sell this kind of insurance, I can tell you that it's getting to be a pretty good thing to have if you develop software.
Ok, interesting point, but there is no opportunity to review the source code in proprietary software systems, whereas open source software at least holds out the possibility of peer review. A good example would be the infamous backdoor that Thompson put in the earliest incarnations of the Unix login program. And since we are well aware of the history of backchannel government intervention in proprietary software systems, dating at least as far back as DES, there is no reason to believe that the walls to intervention are any higher; at best, they are just more opaque.
It's been said time and again, but it's an old adage that governments would be well-served to heed is: Security through obscurity is no security at all.
Thanks for listening,
x86 is extremely bloated when compared to MIPS processors. What they need to do is make a RISC or MIPS based processor with x86 backwards compatability, and slowly phase over.
oh, like itanium? good idea.
uhh, with all due respect, if we have learned anything in the past thirty years from the success of windows, unix, the as/400 and finally x86, it's that architectures are the hardest thing in the world to change due to the massive installed base, and that it's usually better to extend what you have.
just look at os/2, the MCA bus, and now itanium. why would i migrate to a new ISA and lose all the software that I already have when I can just grow my current one?
and x86 isn't that bloated, and cisc isn't that bad. just look at p4 vs. athlon - the tremendous clock speeds realized by the p4's use of an extended pipeline (which is a risc-like optimization) have a tremendous downside - you lose a lot of time resetting the cache if you miss a branch. so for interative programs, as opposed to massive number crunching (and that can be addressed cheaper using MPP and clustering), risc is something of a dog.
finally, you can't say that the desktop is not important to itanium when the line between servers, workstations, and desktops gets blurrier all the time, and the largest growing segment of the market is the low-to-mid-size server.
high-end servers may carry a premium price and have a higher margin, but like lenin said, quantity has a quality all its own.
this is not good news for intel.