Backgrounder - still needs some work before it can become a proper movie treatment
Following on from the lowering of T1 into the foundary cauldren at the end of "Terminator 2 : Judgement Day", the steel was eventually used to make 1000s of paper clips. These paper clips each inherited a small part of the T1 intelligence, however, because of the heat, the software resorted to the evil, malicious intent of the T1 originally shown in "The Terminator".
Individually, these paper clips were harmless. However, when kept in a box, their collective intelligence could combine, such that they were able to execute Skynet's dastardly plans.
Out of a black-ops / skunk-works-type covert, federal lab arose Detypinator, who set out to detype the Skynet, and restore the paper clips to their original, benign uses - holding lumps of paper together, and annoying the hell out of MS Word users.
I'm surprised that you see supposed "Linux" people saying, "wouldn't it be great if Microsoft open sourced Windows", or "hopefully they'll become more open source minded". It seems to me that people making these statements seem to have the view that Microsoft is fundmentally "good", and just currently misguided, and "we", the FOSS community, should "give them a chance."
Well, stop the presses. Microsoft aren't fundamentally good. They're a convicted monopolist, they use dirty tricks, they steal intellectual property from other companies they pretend they're about to partner with, to list a few of their sins.
Microsoft becoming a good FOSS citizen is never going to happen. Nobody should believe otherwise.
The whole reason that $50/h Linux admins (and therefore Linux itself) makes sense is that it doesn't require as many hours to admin.
The other thing you're overlooking is the consequences of "you get what you pay for". A $12/hour Windows admin just isn't going to be able to provide the same quality of work as a $50/hour Linux admin (otherwise, why wouldn't they charge more than $12/hour ? If they're good, they should be able to at least charge something like $30/hour ?), which again will increase the number of hours that you'll have to pay the $12/hour Windows admin. The quality of the functionaly equivalent jobs won't be the same with such as disparity between the per hourly rates.
Comparing the platforms based purely on a per hour admin rate, irrespective of the actual time and effort involved, is a way too simplistic comparison to be useful.
Notice there is nothing in that definition that indicates the origin of those ideas ? Microsoft are an innovative company, because they take ideas and use them. They aren't an inventive company, because they very often don't come up with any new ideas themselves.
IBM and Oracle are innovative companies too.
As for being inventive, I'm not sure about Oracle, however, IBM are, based on the fundamental intention of patents (registering new inventions), and based on the number of patents they are granted (more than 3000 in 2004), IBM are one of the most inventive, if not the most inventive organisation in the world.
My point isn't that its so inter-twined with IP that its impossible...
And that is the point I'm disagreeing with. Not impossible to move it to user space, however IP does rely on ICMP to the point where placing it in user space would likely increase unreliability.
ICMP is quite a light weight and simple protocol. The IPv4 ICMP source code file in Linux 2.6 is only 1138 lines, going by what "wc -l icmp.c" says. Moving it to user space would increase that count, which naturally increases the possiblity of bugs.
Moving ICMP to user space wouldn't mitigate the problems that this article is about anyway. More thorough validation of ICMP packets, and possibly a few fixes to the base protocol itself would be best.
Hosts using PMTU Discovery MUST detect decreases in Path MTU as fast
as possible. Hosts MAY detect increases in Path MTU, but because
doing so requires sending datagrams larger than the current estimated
PMTU, and because the likelihood is that the PMTU will not have
increased, this MUST be done at infrequent intervals. An attempt to
detect an increase (by sending a datagram larger than the current
estimate) MUST NOT be done less than 5 minutes after a Datagram Too
Big message has been received for the given destination, or less than
1 minute after a previous, successful attempted increase. We
recommend setting these timers at twice their minimum values (10
minutes and 2 minutes, respectively).
Run ipsentinel, and configure it to "own" all possible IP addresses i.e. 0/0. It'll break ARP which means nothing will work, unless static ARP entries are configured on the hosts.
I think you're a bit misguided here. Why would people be wanting to offload TCP onto special TOE NICs if TCP performance wasn't important (not that I agree with doing that myself) ? TCP's processing overhead is higher than IP's, avoiding context switches for TCP, by putting it in the kernel, is a worth while benefit of putting it in kernel space rather than userspace.
ICMP also would have to be in the kernel even if TCP and UDP weren't. ICMP is used by IP itself, for things such as fragmentation time-out messages, protocol unreachables and PMTU discovery.
People didn't just blindly decide to place these things in the kernel space, there are good and sound reasons why they were put there, and I'm sure they've also been re-evaluated quite often since, with the same conclusion being arrived at evey time. Kernel space is best for performance for things that are very commonly used, and for ensuring that the facility is available to all applications that run on the OS, as well as the mechanisms within the kernel that rely on them (e.g., UDP relying on ICMP).
Because things that need it are in the kernel
on
Examining ICMP Flaws
·
· Score: 2, Informative
Putting things in the kernel rather than user space isn't purely a performance decision, which seems to be the logic behind your comment. If that was the case, there'd be a whole lot of things that should be moved to user space, such as the dumb terminal or low speed serial port handling.
A lot of things should be put in user space if they can. When those things are put in the kernel, it is usually because the kernel is a better location than userspace. Sometimes it is for performance, sometimes for other reasons. It is a trade off of course.
UDP relies on ICMP to generate unreachable port messages. Coming up with a setup where the in kernel UDP code calls a userspace ICMP process to generate those messages is more complicated than having ICMP as part of the kernel.
While the dictionaries that I have checked do not list "noone" as a word, the formation is consistent with putting "any", "some", "every", and "no" in front of "thing", "where", "body, and "one". Of the sixteen possibilities, only "noone" is not a word.
That's probably why people do it. The reason why I think it is probably the exception is that it breaks the repeated vowel rule, where the vowel in question is pronounced "long" rather than "short".
English is a bit of a "mutt" of a language, with many words from other languages, and many constructions which don't necessarily make sense. Specifically regarding these joinings of words with "body or one", if you think about it, there actually isn't an obvious reason why it was done in the first place. How is "nobody" any better than "no body", other than saving a space ? Of course we're conditioned to use "nobody" so "no body" looks odd now.
I don't normally have a problem with other peoples' misspellings, however, I aways trip up on "noone" to the point where I'm also starting to get annoyed when I see it. What's worse is that it seems to be becoming more and more popular.
The problem with "noone" is that you get mentally half way into thinking "noon" and then realise that it's about to be "noon-e", a nonsense word because of the "e" on the end. Then you realise that the person is really trying to mean "no one".
If people insist on joining the two words together, at least a hyphen would be nice, as in "no-one". That also looks odd, and that should be a sign that the hypen should really be a space.
I'd suggest you reconsider who you are calling a retard, because us "Software RAID array retards" will never, ever have those sorts of problems, as long as an ATA interface is available for our RAID'ed drives on a box that can run Linux.
Hotmail people will have to check their spam folder so regularly for for things that aren't actually spam that Sender-ID will just annoy them so much that they'll abandon Hotmail. What's the point of spam filters when you have to check each and every email yourself anyway to make sure ?
For this push to adopt Sender-ID to succeed, Hotmail would probably have to be the source of more than 50% of the emails sent on the Internet. I'm pretty sure that is the case.
Admittedly I haven't read a lot of "The Art of War", however, his fundamental principle was "It is best to win without fighting". IOW, only engage in war when you have no other choice.
You don't seem to be distinguishing between cards that have relatively minor optimisations to increase performance, such as the Intel and Broadcom ones you've mentioned, and the card that is the topic of this article, which runs the whole TCP/IP stack (ie. managing the connections including sequence numbers, round trip times etc. etc. etc. independently of the OS).
This article is about a separate card running the full TCP/IP stack, and it is this card that Phil Karn's criticisms apply to. I don't think can turn that into a criticism of all on card optimisations, such as TCP Segment Offload, which is what I understand is one of the the optimisations that the Intel and Broadcom cards implement. I think you're expanding his opinion wider than it probably is.
As an alternative example of opinions on NIC hardware optimisations, Dave Miller, the main networking subsystem maintainer for Linux, does not like TCP Offload Engines (which is what this article card is) at all, yet he's quite happy to come up with new and better ways to use TCP Segment Offload. Here's an example of his opinion on TOE. Here's his recent post regarding new TSO code he has written and how he is happy with its performance.
It's called "TCP Onloading". Here's an article about it (paid subscription required for the whole article - the abstract is free). Notice that it's coming from Intel research labs.
I could only read the abstract. I really can't see how to interpret it as saying that the whole TCP/IP stack was going to be moved on to the chipset.
Intel currently supports TCP segmentation offload (pushing the packet segmentation task from the TCP stack onto the hardware), and is moving to push the entire TCP stack to a dedicated processor + NIC combo.
I'm curious, where have Intel said they're going to push the whole TCP/IP stack onto a dedicated processor ? I've seen a bit of speculation they're going to do it, but not a statement from Intel themselves. Does one exist ?
You couldn't be more wrong. Since the 90s, the boundary of what the NIC should do and what the OS should do has been repeatedly re-examined, and industry leaders in networking have successfully deployed products that big-iron servers rely on.
I don't necessarily think that just because some customers have bought them, means its a good, well thought out and useful idea. A lot of people buy Britney Spears music, does that make it good or just popular ?
I suggest having a read of any threads on the Linux netdev mailing list, with "TOE" in the title. There has been quite a lot of discussion about the pros and cons, and it seems that the consensis of the main Linux kernel networking hackers is that TOE isn't a good idea.
If by tought, you mean that they are aware of it, actually, yes.
No, I mean detailed, explicit sexual knowledge that 13 year olds have. If you wouldn't be happy about a seven year old having that level of knowledge about sex, then I think you shouldn't be happy about a seven year old having the same level of detailed and explicity knowledge of violence.
The major issue here is that the child in question is around half the age the violence is judged to be suitable for. I'd have no problem with a parent saying, for example, that their 11 year old is mature enough to see a 13+ rated movie. 11 is close enough to 13, and the 11 year old may be more mature than their age if they have older siblings. That is their parents decision, and they know the child better than I do. However, when it is such a significant difference between the child's age and the suitability rating, I then have to question the parent's ability to make a proper and sound judgement on the maturity of their child. At a certain point, society has to be watching out for the child, because the parent isn't. I personally think in this case, the parent in question isn't, which is why I think it is child abuse.
You could argue that it is none of my business about how a child is brought up, however I then think you'd also be saying that it would be none of my business if the parent was making their child participate in child pornography. Society has a responsiblity to look out for children who can't defend themselves, sometimes even from their parents. Exposing a young child to a level of violence that is way beyond their years is a form of mental abuse, and is no different to physical abuse, sexual or otherwise. Society shouldn't accept it or permit it.
I don't know if it is against the law in the US to show a seven year old a 13+ movie, as it is here in Australia. If it is, I'm tempted to contact the relivent authorities.
I don't have a problem with him wanting to be able to pay bills etc. There is a level of practicality that has to be accepted with any open source project.
What I don't understand is why it was Microsoft. The only reason that I could find acceptable was that nobody else wanted him. I'd find that a bit hard to believe.
Maybe they made a good offer, and he naively believes he can change them from the inside.
P.S., good luck with the filesystem. Sorry to hear about the divorce, hopefully it works out somewhat amicably, as much as these things can.
Slashdot Mirror
Backgrounder - still needs some work before it can become a proper movie treatment
Following on from the lowering of T1 into the foundary cauldren at the end of "Terminator 2 : Judgement Day", the steel was eventually used to make 1000s of paper clips. These paper clips each inherited a small part of the T1 intelligence, however, because of the heat, the software resorted to the evil, malicious intent of the T1 originally shown in "The Terminator".
Individually, these paper clips were harmless. However, when kept in a box, their collective intelligence could combine, such that they were able to execute Skynet's dastardly plans.
Out of a black-ops / skunk-works-type covert, federal lab arose Detypinator, who set out to detype the Skynet, and restore the paper clips to their original, benign uses - holding lumps of paper together, and annoying the hell out of MS Word users.
going to MS is just naive.
I'm surprised that you see supposed "Linux" people saying, "wouldn't it be great if Microsoft open sourced Windows", or "hopefully they'll become more open source minded". It seems to me that people making these statements seem to have the view that Microsoft is fundmentally "good", and just currently misguided, and "we", the FOSS community, should "give them a chance."
Well, stop the presses. Microsoft aren't fundamentally good. They're a convicted monopolist, they use dirty tricks, they steal intellectual property from other companies they pretend they're about to partner with, to list a few of their sins.
Microsoft becoming a good FOSS citizen is never going to happen. Nobody should believe otherwise.
The whole reason that $50/h Linux admins (and therefore Linux itself) makes sense is that it doesn't require as many hours to admin.
The other thing you're overlooking is the consequences of "you get what you pay for". A $12/hour Windows admin just isn't going to be able to provide the same quality of work as a $50/hour Linux admin (otherwise, why wouldn't they charge more than $12/hour ? If they're good, they should be able to at least charge something like $30/hour ?), which again will increase the number of hours that you'll have to pay the $12/hour Windows admin. The quality of the functionaly equivalent jobs won't be the same with such as disparity between the per hourly rates.
Comparing the platforms based purely on a per hour admin rate, irrespective of the actual time and effort involved, is a way too simplistic comparison to be useful.
From the Cambridge Advanced Learner's Dictionary :
innovative - using new methods or ideas
Notice there is nothing in that definition that indicates the origin of those ideas ? Microsoft are an innovative company, because they take ideas and use them. They aren't an inventive company, because they very often don't come up with any new ideas themselves.
IBM and Oracle are innovative companies too.
As for being inventive, I'm not sure about Oracle, however, IBM are, based on the fundamental intention of patents (registering new inventions), and based on the number of patents they are granted (more than 3000 in 2004), IBM are one of the most inventive, if not the most inventive organisation in the world.
And that is the point I'm disagreeing with. Not impossible to move it to user space, however IP does rely on ICMP to the point where placing it in user space would likely increase unreliability.
ICMP is quite a light weight and simple protocol. The IPv4 ICMP source code file in Linux 2.6 is only 1138 lines, going by what "wc -l icmp.c" says. Moving it to user space would increase that count, which naturally increases the possiblity of bugs.
Moving ICMP to user space wouldn't mitigate the problems that this article is about anyway. More thorough validation of ICMP packets, and possibly a few fixes to the base protocol itself would be best.
Hosts using PMTU Discovery MUST detect decreases in Path MTU as fast as possible. Hosts MAY detect increases in Path MTU, but because doing so requires sending datagrams larger than the current estimated PMTU, and because the likelihood is that the PMTU will not have increased, this MUST be done at infrequent intervals. An attempt to detect an increase (by sending a datagram larger than the current estimate) MUST NOT be done less than 5 minutes after a Datagram Too Big message has been received for the given destination, or less than 1 minute after a previous, successful attempted increase. We recommend setting these timers at twice their minimum values (10 minutes and 2 minutes, respectively).
Run ipsentinel, and configure it to "own" all possible IP addresses i.e. 0/0. It'll break ARP which means nothing will work, unless static ARP entries are configured on the hosts.
If it is John Nagle behind "animats", then he knows very well what he is talking about.
I think you're a bit misguided here. Why would people be wanting to offload TCP onto special TOE NICs if TCP performance wasn't important (not that I agree with doing that myself) ? TCP's processing overhead is higher than IP's, avoiding context switches for TCP, by putting it in the kernel, is a worth while benefit of putting it in kernel space rather than userspace.
ICMP also would have to be in the kernel even if TCP and UDP weren't. ICMP is used by IP itself, for things such as fragmentation time-out messages, protocol unreachables and PMTU discovery.
People didn't just blindly decide to place these things in the kernel space, there are good and sound reasons why they were put there, and I'm sure they've also been re-evaluated quite often since, with the same conclusion being arrived at evey time. Kernel space is best for performance for things that are very commonly used, and for ensuring that the facility is available to all applications that run on the OS, as well as the mechanisms within the kernel that rely on them (e.g., UDP relying on ICMP).
Putting things in the kernel rather than user space isn't purely a performance decision, which seems to be the logic behind your comment. If that was the case, there'd be a whole lot of things that should be moved to user space, such as the dumb terminal or low speed serial port handling.
A lot of things should be put in user space if they can. When those things are put in the kernel, it is usually because the kernel is a better location than userspace. Sometimes it is for performance, sometimes for other reasons. It is a trade off of course.
UDP relies on ICMP to generate unreachable port messages. Coming up with a setup where the in kernel UDP code calls a userspace ICMP process to generate those messages is more complicated than having ICMP as part of the kernel.
That's probably why people do it. The reason why I think it is probably the exception is that it breaks the repeated vowel rule, where the vowel in question is pronounced "long" rather than "short".
English is a bit of a "mutt" of a language, with many words from other languages, and many constructions which don't necessarily make sense. Specifically regarding these joinings of words with "body or one", if you think about it, there actually isn't an obvious reason why it was done in the first place. How is "nobody" any better than "no body", other than saving a space ? Of course we're conditioned to use "nobody" so "no body" looks odd now.
I don't normally have a problem with other peoples' misspellings, however, I aways trip up on "noone" to the point where I'm also starting to get annoyed when I see it. What's worse is that it seems to be becoming more and more popular.
The problem with "noone" is that you get mentally half way into thinking "noon" and then realise that it's about to be "noon-e", a nonsense word because of the "e" on the end. Then you realise that the person is really trying to mean "no one".
If people insist on joining the two words together, at least a hyphen would be nice, as in "no-one". That also looks odd, and that should be a sign that the hypen should really be a space.
You get a swimming turtle if you are IPv6 connected.
you can read the papers here. I recommend the first few pages of "NewArch Final Technical Report" to find out answers to your questions
NewArch Project: Future-Generation Internet Architecture
Things that NATs break
Fix all those, and you'll have invented IPv6 ...
when they're hardware RAID controller failed, and they weren't able to get an exact replacement.
The Adaptec 2110S RAID card in Fark's database server is dying.
I'd suggest you reconsider who you are calling a retard, because us "Software RAID array retards" will never, ever have those sorts of problems, as long as an ATA interface is available for our RAID'ed drives on a box that can run Linux.
you might be interested in another advantage of your set up that you may not be aware of.
Looks like I'm getting a bit too lazy with the preview button.
Hotmail people will have to check their spam folder so regularly for for things that aren't actually spam that Sender-ID will just annoy them so much that they'll abandon Hotmail. What's the point of spam filters when you have to check each and every email yourself anyway to make sure ?
For this push to adopt Sender-ID to succeed, Hotmail would probably have to be the source of more than 50% of the emails sent on the Internet. I'm pretty sure that is the case.
Admittedly I haven't read a lot of "The Art of War", however, his fundamental principle was "It is best to win without fighting". IOW, only engage in war when you have no other choice.
You don't seem to be distinguishing between cards that have relatively minor optimisations to increase performance, such as the Intel and Broadcom ones you've mentioned, and the card that is the topic of this article, which runs the whole TCP/IP stack (ie. managing the connections including sequence numbers, round trip times etc. etc. etc. independently of the OS).
This article is about a separate card running the full TCP/IP stack, and it is this card that Phil Karn's criticisms apply to. I don't think can turn that into a criticism of all on card optimisations, such as TCP Segment Offload, which is what I understand is one of the the optimisations that the Intel and Broadcom cards implement. I think you're expanding his opinion wider than it probably is.
As an alternative example of opinions on NIC hardware optimisations, Dave Miller, the main networking subsystem maintainer for Linux, does not like TCP Offload Engines (which is what this article card is) at all, yet he's quite happy to come up with new and better ways to use TCP Segment Offload. Here's an example of his opinion on TOE. Here's his recent post regarding new TSO code he has written and how he is happy with its performance.
It's called "TCP Onloading". Here's an article about it (paid subscription required for the whole article - the abstract is free). Notice that it's coming from Intel research labs.
I could only read the abstract. I really can't see how to interpret it as saying that the whole TCP/IP stack was going to be moved on to the chipset.
Intel currently supports TCP segmentation offload (pushing the packet segmentation task from the TCP stack onto the hardware), and is moving to push the entire TCP stack to a dedicated processor + NIC combo.
I'm curious, where have Intel said they're going to push the whole TCP/IP stack onto a dedicated processor ? I've seen a bit of speculation they're going to do it, but not a statement from Intel themselves. Does one exist ?
You couldn't be more wrong. Since the 90s, the boundary of what the NIC should do and what the OS should do has been repeatedly re-examined, and industry leaders in networking have successfully deployed products that big-iron servers rely on.I don't necessarily think that just because some customers have bought them, means its a good, well thought out and useful idea. A lot of people buy Britney Spears music, does that make it good or just popular ?
I suggest having a read of any threads on the Linux netdev mailing list, with "TOE" in the title. There has been quite a lot of discussion about the pros and cons, and it seems that the consensis of the main Linux kernel networking hackers is that TOE isn't a good idea.
Here is a quick TOE netdev search link for you.
and have a read of why the interrupt problem isn't a problem anymore, at least on Linux. Note the date too - October 2001.
LWN.net
NAPI has been implemented in the kernel.org kernels for a number of years now.
If by tought, you mean that they are aware of it, actually, yes.
No, I mean detailed, explicit sexual knowledge that 13 year olds have. If you wouldn't be happy about a seven year old having that level of knowledge about sex, then I think you shouldn't be happy about a seven year old having the same level of detailed and explicity knowledge of violence.
The major issue here is that the child in question is around half the age the violence is judged to be suitable for. I'd have no problem with a parent saying, for example, that their 11 year old is mature enough to see a 13+ rated movie. 11 is close enough to 13, and the 11 year old may be more mature than their age if they have older siblings. That is their parents decision, and they know the child better than I do. However, when it is such a significant difference between the child's age and the suitability rating, I then have to question the parent's ability to make a proper and sound judgement on the maturity of their child. At a certain point, society has to be watching out for the child, because the parent isn't. I personally think in this case, the parent in question isn't, which is why I think it is child abuse.
You could argue that it is none of my business about how a child is brought up, however I then think you'd also be saying that it would be none of my business if the parent was making their child participate in child pornography. Society has a responsiblity to look out for children who can't defend themselves, sometimes even from their parents. Exposing a young child to a level of violence that is way beyond their years is a form of mental abuse, and is no different to physical abuse, sexual or otherwise. Society shouldn't accept it or permit it.
I don't know if it is against the law in the US to show a seven year old a 13+ movie, as it is here in Australia. If it is, I'm tempted to contact the relivent authorities.
I don't have a problem with him wanting to be able to pay bills etc. There is a level of practicality that has to be accepted with any open source project.
What I don't understand is why it was Microsoft. The only reason that I could find acceptable was that nobody else wanted him. I'd find that a bit hard to believe.
Maybe they made a good offer, and he naively believes he can change them from the inside.
P.S., good luck with the filesystem. Sorry to hear about the divorce, hopefully it works out somewhat amicably, as much as these things can.