If the money went to the university, I'd agree with you. This does not, and cuts into the publisher's profit margins by attempt to charge the publishers for the bandwidth the textbooks would class services would use.
The universities and colleges will get very upset.
They won't ignore it, because it sets a horrid precedent. A few letters to the school about the students' landlords deliberately interfering with access to the school's subject matter would be only the start of the legal and social issues. Even if it's not settled by lawsuit, it's the sort of quality of service issue that comes up at contract renewals and approvals for off-campus housing accreditation needed to allow direct rent payment from grants or student loans.
> Right now the network is more or less open, except that we block access (by court order) to certain sites at the request of various copyright holder
Look into the history of DMCA "takedown orders". It can get quite odd: take a look at https://torrentfreak.com/fox-d... for how mentioning the takedown orders can lead to a takedown order.
> Specifically, they are interested in targeting commercial providers of services directed at college students, such as textbook rental firms, online booksellers, and so on. With approximately 35,000 residents,
All of these provide relatively static content, not streaming content. Low bandwidth will simply not affect their business models, especially since the big bandwidth users are the streaming services, gaming, and Bittorrent in most of your resident's homes, If you've the expertise and equipment to even consider this kind of throttling, you should have some network monitors already in place to verify this claim, or you should be able to justify getting a loan of a network traffic monitor for just such analysis. And given the overwhelming bandwidth use of such high bandwidth applications, the relatively low bandwidth needs of textbooks and other critical student specific services won't even notice the loss of quality service unless you effectively block them.
If you block them entirely, the book publishers can call the FCC and contact the schools and their clients, and they _will_ get upset with you. They can also contact your upstream ISP, who will be very unhappy with you muscling into their bandwidth throttling "turf".
The _behavior_ of redirecting failed DNS lookups to an advertising server is unsurprising. Roughly 10 years ago, Verisign did much the same thing to to the master servers for *.com', and broke the concept of getting a "no such record" result for everyone in the world using ".com" addresses.
Many, many people were _extremely_ upset when this unannounced change occurred. It broke tools worldwide that were used to verify DNS configuraitons, and it routed email that was misspelled or had faild DNS to Verizon's advertising DNS IP addresses. I was never sure if Verisign bothered to do anything with all the DNS connection requests, FTP requests, SSH requests, or everyehing else redirected to their sites, but it left Verisign in charge of a tremendous amount of data and potential network manipulation.
People, and software, have become more accustomed to such DNS abuse. But it's still problematic if you don't realize it's going on.
Or they just hide the errors, present them as someone else's fault, or it's "not on the tasklist" and thus never gets addressed. I've certainly seen all of these, with both open source and closed source. But closed source is more prone to pretending the problems do not exist, especially when the major security groups have agreed not to publish flaws that there is not yet a patch for.
They're radioactive. The radioactivity, particularly neutron radiation, makes the surrounding material also slightly radioactive. So it's awkward, dangerous, and self defeating to generate even more low level radioactive material trying to handle it.
Concrete or steel alone, we have some data. Radioactive sludge, I'm afraid not. We're still prone to events like this (http://www.atonkstail.com/2014/05/clay-litter-bad-for-kitties-good-for.html) , where the organic cat litter was used for the sludge rather than clay based cat litter. The results were were unfortunate, but unsurprising from the kind of optimistic design that tests one canister with one kin dof sludge for 10 years and extrapolates to 5000 canisters for "hundreds of years" years filled with kilotons of material.
> The above-ground casks can easily last that long.
Really, no. Between simple exposure and the fascinating chemical interactions with the low level radio-active material of numerous kinds, there's no evidence that the inexpensive containers will last that long. It's much like a start-up companies sales chart: a few early bits of information are extrapolated into a hopelessly optimistic long term graph that is unlikely to be relevent even for the next six months, much less the next 50 years.
There's a reasonable Scientific American article about this at http://www.scientificamerican..... They're apparently only rated for 100 years, and I consider that _extremely_ optimistic.
> Nothing is NSA-proof, therefore we should just scrap TLS and transmit everything in plaintext, right?
I'm afraid that this is a common approach. I've seen numerous system designers refuse to enable encryption on their internal websites or unternal softwre on the basis that "if they're inside our networks, we'e got much bigger problems". It's one of the mantras of bad engineers, much like "there's no point to documentation, just read the code".
>> they're probably untrainable in either social or technical standards >> they're probably untrainable in either social or technical standards
> What an impolite word to use. I'd never hire you.
Thank you for demonstrating that you don't actually know the meaning of the word "polite". The problem is not the employers providing, and expecting, minimum standards of courtesy somehow interfering with the self-expression of the crude. Ignoring those standards of basic courtesy has gone with racism, sexual abuse, and poisonous commentary about colleagues that ruin a workplace far more than the crudities themselves. If they can't even control their language, how can one expect them to follow basic software or hardware safety practices? If they reject the rules as "irrational", how can one know that they actually check their code for error messages or always connect the big green wire to ground?
I'm afraid you have it backwards. If an employer, or employer, is too petty or stupid to use language politely, they're probably untrainable in either social or technical standards and all their work will require constant review and repair before publication.
I'm afraid I'm not "moving the goalposts". I was reasonably specific: "please name 2 environments that actually benefited from using Infiniband or Firewire." You've cited several that _use_ Infiniband, which is a good step towards that goal. But it's not clear that they actually benefited. As Wikipedia correctly points out: "InfiniBand has no standard programming API. Be prepared to develop and write TCP and UDP.". The result has been demonstrably unstable and incompatible with other vendors' hardware or software, and the lockin and instability due to subtle discrepancies in interpretations of the "OpenFabrics Allicance" from even different departments of the same vendor was nightmarish. The adapters were, in my experience, bug filled and fragile. The time wasted with downtime and replacements, especially due to firmware bugs which often _could not be resolved_ on the original model sold to my partners, cost more performance than was gained by their high bandwidth and low latency.
It's conceivable it's gotten better in the last 3 years, but it's typical of HP servers. They have important features, but flaws are ignored while the project fails: I've seen an IT director fired because he decided the data center would be "all HP", and the new hardware sat in boxes for 6 months while he tried to negotiate a return because it had the wrong specs. The new director had the unopened boxes shipped back, and the stacking fee paid within 48 hours, and the hardware replaced with equivalent hardware from Dell with the new specs delivered within another 48 hours, at a net cost matching the original HP servers.
I'm afraid that "HPC clusters" is a technology, not an environment. And "cloud providers" is an amazingly vague term, and filled with many alternative technolgies and architectures.
Please do name real companies or projects that used or benefited from Infiniband.
> doesn't really have to deal with true capitalism
Is that like finding a "true faith" or a "true soulmate" or a "true democracy"? It's a goal, and a laudable one, but not a complete reality at any level.
> they don't have to worry about them complaining about things like working conditions
Or being forced to use a style guide. Or using PHP instead of Perl or Ruby. Or Jboss instead of Tomcat, or NFS instead of CIFS, emacs instead of vi, Or actually working on the project on top of the task list instead of their personal "startup company" side project.
The outsourced engineers I work with are right now are very good, but they are also fairly senior. They've worked their way up from 'call center' jobs, and I applaud their efforts. They've not tended to delve into the underlying issues, but that's a managerial pressure problem, and I'm helping with that.
They're being rewarded for closing tasks on the tasklist or closing tickets, not solving a ticket so well it never recurs and never shows up on the tasklist. They are also _cheap_, in the sense of avoiding new architectures, new paradigms, or re-inventing the wheel because they think they can do a better job themself. The desire to re-invent from scratch in order to solve one small problem is common to my American colleagues, and I've enjoyed the opportunity to help negotiate and rethink the approaches.
> Rock Solid meaning slick bomb-proof drivers as well as a machine that didn't crap out
I'm afraid HP desktops took on the "planned obsolescence" model around the time they bought Compaq. They've done the same for personal color printers, which are _much_ more expensive if you try to make them robust. Instead they make their money on the ink, and they make their deesktop and laptop money on the high turnover.
Unfortunately, similar attitudes seem to have infested their servers, which are no longer the reliable standard they used to represent. They do occasionally have leading edge features, but please name 2 environments that actually benefited from using Infiniband or Firewire.
A company that has that kind of resources has already created their own patent suite, which is cheaper _for a large company_.
The most effective way to eliminate patent trolls would be to discard software patents as a clear hindrance to creativity, which they are in their current state. But I don't see that as feasible in the current political marketplace.
My point was not about their poverty: My point was that if and as they progress economically, they're also going to want more energy. I wasn't suggesting that solar wind was viable, only that it's probably the only to gather enough deuterium and tritium to satisfy the demands of fusion fuels if that technology is ever even _viable_. And if one is bothering to invest in large solar mirrors or gathering arrays, why not use them for direct solar power collection, which is workable with technology today.
"Very Safe" for a metal with such pyrokinetic behavior is not necessarily how I'd describe it. But some casual research does show it be more promising than uranium for longer term fuel sources: I'll keep this in mind: thank you for the pointer.
We've figured out the basics of _that_ fusion source. But it does not scale down well. _All_ of the test fusion reactors use deuterium or tritium, not plain hydrogen.
Slashdot Mirror
If the money went to the university, I'd agree with you. This does not, and cuts into the publisher's profit margins by attempt to charge the publishers for the bandwidth the textbooks would class services would use.
The universities and colleges will get very upset.
They won't ignore it, because it sets a horrid precedent. A few letters to the school about the students' landlords deliberately interfering with access to the school's subject matter would be only the start of the legal and social issues. Even if it's not settled by lawsuit, it's the sort of quality of service issue that comes up at contract renewals and approvals for off-campus housing accreditation needed to allow direct rent payment from grants or student loans.
> Right now the network is more or less open, except that we block access (by court order) to certain sites at the request of various copyright holder
Look into the history of DMCA "takedown orders". It can get quite odd: take a look at https://torrentfreak.com/fox-d... for how mentioning the takedown orders can lead to a takedown order.
> Specifically, they are interested in targeting commercial providers of services directed at college students, such as textbook rental firms, online booksellers, and so on. With approximately 35,000 residents,
All of these provide relatively static content, not streaming content. Low bandwidth will simply not affect their business models, especially since the big bandwidth users are the streaming services, gaming, and Bittorrent in most of your resident's homes, If you've the expertise and equipment to even consider this kind of throttling, you should have some network monitors already in place to verify this claim, or you should be able to justify getting a loan of a network traffic monitor for just such analysis. And given the overwhelming bandwidth use of such high bandwidth applications, the relatively low bandwidth needs of textbooks and other critical student specific services won't even notice the loss of quality service unless you effectively block them.
If you block them entirely, the book publishers can call the FCC and contact the schools and their clients, and they _will_ get upset with you. They can also contact your upstream ISP, who will be very unhappy with you muscling into their bandwidth throttling "turf".
The _behavior_ of redirecting failed DNS lookups to an advertising server is unsurprising. Roughly 10 years ago, Verisign did much the same thing to to the master servers for *.com', and broke the concept of getting a "no such record" result for everyone in the world using ".com" addresses.
http://slashdot.org/story/03/0...
Many, many people were _extremely_ upset when this unannounced change occurred. It broke tools worldwide that were used to verify DNS configuraitons, and it routed email that was misspelled or had faild DNS to Verizon's advertising DNS IP addresses. I was never sure if Verisign bothered to do anything with all the DNS connection requests, FTP requests, SSH requests, or everyehing else redirected to their sites, but it left Verisign in charge of a tremendous amount of data and potential network manipulation.
People, and software, have become more accustomed to such DNS abuse. But it's still problematic if you don't realize it's going on.
Or they just hide the errors, present them as someone else's fault, or it's "not on the tasklist" and thus never gets addressed. I've certainly seen all of these, with both open source and closed source. But closed source is more prone to pretending the problems do not exist, especially when the major security groups have agreed not to publish flaws that there is not yet a patch for.
They're radioactive. The radioactivity, particularly neutron radiation, makes the surrounding material also slightly radioactive. So it's awkward, dangerous, and self defeating to generate even more low level radioactive material trying to handle it.
Concrete or steel alone, we have some data. Radioactive sludge, I'm afraid not. We're still prone to events like this (http://www.atonkstail.com/2014/05/clay-litter-bad-for-kitties-good-for.html) , where the organic cat litter was used for the sludge rather than clay based cat litter. The results were were unfortunate, but unsurprising from the kind of optimistic design that tests one canister with one kin dof sludge for 10 years and extrapolates to 5000 canisters for "hundreds of years" years filled with kilotons of material.
> The above-ground casks can easily last that long.
Really, no. Between simple exposure and the fascinating chemical interactions with the low level radio-active material of numerous kinds, there's no evidence that the inexpensive containers will last that long. It's much like a start-up companies sales chart: a few early bits of information are extrapolated into a hopelessly optimistic long term graph that is unlikely to be relevent even for the next six months, much less the next 50 years.
There's a reasonable Scientific American article about this at http://www.scientificamerican..... They're apparently only rated for 100 years, and I consider that _extremely_ optimistic.
> Nothing is NSA-proof, therefore we should just scrap TLS and transmit everything in plaintext, right?
I'm afraid that this is a common approach. I've seen numerous system designers refuse to enable encryption on their internal websites or unternal softwre on the basis that "if they're inside our networks, we'e got much bigger problems". It's one of the mantras of bad engineers, much like "there's no point to documentation, just read the code".
>> they're probably untrainable in either social or technical standards
>> they're probably untrainable in either social or technical standards
> What an impolite word to use. I'd never hire you.
Thank you for demonstrating that you don't actually know the meaning of the word "polite". The problem is not the employers providing, and expecting, minimum standards of courtesy somehow interfering with the self-expression of the crude. Ignoring those standards of basic courtesy has gone with racism, sexual abuse, and poisonous commentary about colleagues that ruin a workplace far more than the crudities themselves. If they can't even control their language, how can one expect them to follow basic software or hardware safety practices? If they reject the rules as "irrational", how can one know that they actually check their code for error messages or always connect the big green wire to ground?
I'm afraid you have it backwards. If an employer, or employer, is too petty or stupid to use language politely, they're probably untrainable in either social or technical standards and all their work will require constant review and repair before publication.
Like the f2c toolkit, for converting Fortran to C?
I don't think you could write he parser in R, or in Julia.
I'm afraid I'm not "moving the goalposts". I was reasonably specific: "please name 2 environments that actually benefited from using Infiniband or Firewire." You've cited several that _use_ Infiniband, which is a good step towards that goal. But it's not clear that they actually benefited. As Wikipedia correctly points out: "InfiniBand has no standard programming API. Be prepared to develop and write TCP and UDP.". The result has been demonstrably unstable and incompatible with other vendors' hardware or software, and the lockin and instability due to subtle discrepancies in interpretations of the "OpenFabrics Allicance" from even different departments of the same vendor was nightmarish. The adapters were, in my experience, bug filled and fragile. The time wasted with downtime and replacements, especially due to firmware bugs which often _could not be resolved_ on the original model sold to my partners, cost more performance than was gained by their high bandwidth and low latency.
It's conceivable it's gotten better in the last 3 years, but it's typical of HP servers. They have important features, but flaws are ignored while the project fails: I've seen an IT director fired because he decided the data center would be "all HP", and the new hardware sat in boxes for 6 months while he tried to negotiate a return because it had the wrong specs. The new director had the unopened boxes shipped back, and the stacking fee paid within 48 hours, and the hardware replaced with equivalent hardware from Dell with the new specs delivered within another 48 hours, at a net cost matching the original HP servers.
I'd assume it inconvenienced _you_ far more than me, especially if subscribers were alarmed.
I'm afraid that "HPC clusters" is a technology, not an environment. And "cloud providers" is an amazingly vague term, and filled with many alternative technolgies and architectures.
Please do name real companies or projects that used or benefited from Infiniband.
> doesn't really have to deal with true capitalism
Is that like finding a "true faith" or a "true soulmate" or a "true democracy"? It's a goal, and a laudable one, but not a complete reality at any level.
> they don't have to worry about them complaining about things like working conditions
Or being forced to use a style guide. Or using PHP instead of Perl or Ruby. Or Jboss instead of Tomcat, or NFS instead of CIFS, emacs instead of vi, Or actually working on the project on top of the task list instead of their personal "startup company" side project.
The outsourced engineers I work with are right now are very good, but they are also fairly senior. They've worked their way up from 'call center' jobs, and I applaud their efforts. They've not tended to delve into the underlying issues, but that's a managerial pressure problem, and I'm helping with that.
They're being rewarded for closing tasks on the tasklist or closing tickets, not solving a ticket so well it never recurs and never shows up on the tasklist. They are also _cheap_, in the sense of avoiding new architectures, new paradigms, or re-inventing the wheel because they think they can do a better job themself. The desire to re-invent from scratch in order to solve one small problem is common to my American colleagues, and I've enjoyed the opportunity to help negotiate and rethink the approaches.
> Rock Solid meaning slick bomb-proof drivers as well as a machine that didn't crap out
I'm afraid HP desktops took on the "planned obsolescence" model around the time they bought Compaq. They've done the same for personal color printers, which are _much_ more expensive if you try to make them robust. Instead they make their money on the ink, and they make their deesktop and laptop money on the high turnover.
Unfortunately, similar attitudes seem to have infested their servers, which are no longer the reliable standard they used to represent. They do occasionally have leading edge features, but please name 2 environments that actually benefited from using Infiniband or Firewire.
Just in case other people notice, the SSL certificate for the Slashdot login page expired today.
A company that has that kind of resources has already created their own patent suite, which is cheaper _for a large company_.
The most effective way to eliminate patent trolls would be to discard software patents as a clear hindrance to creativity, which they are in their current state. But I don't see that as feasible in the current political marketplace.
My point was not about their poverty: My point was that if and as they progress economically, they're also going to want more energy. I wasn't suggesting that solar wind was viable, only that it's probably the only to gather enough deuterium and tritium to satisfy the demands of fusion fuels if that technology is ever even _viable_. And if one is bothering to invest in large solar mirrors or gathering arrays, why not use them for direct solar power collection, which is workable with technology today.
"Very Safe" for a metal with such pyrokinetic behavior is not necessarily how I'd describe it. But some casual research does show it be more promising than uranium for longer term fuel sources: I'll keep this in mind: thank you for the pointer.
We've figured out the basics of _that_ fusion source. But it does not scale down well. _All_ of the test fusion reactors use deuterium or tritium, not plain hydrogen.