> in total about 10 thousand euros of lost sales for Cisco/Linksys because of that one crap router they saddled me with for Christmas 2008
So their filter against non-profitable clients has worked as expected.
Each time a human at Linksys touches a customer, the company incurs at least 5 euro in costs. Since Linksys relies on retail volume and not consultation for their consumer sales, it's to their financial advantage to never hear from customers once the sale has been made, and especially to their advantage not to have to respond to unending lists of complaints or questions from detail-oriented customers. That same 10,000 euro of kit sold to 200 customers who do not generally know enough to complain is much more profitable to Linksys than if it were sold to you since you have both the aptitude and time to complain, but not effectively. (If you had complained effectively, you would have received a successful resolution from Linksys and both parties would have benefitted directly.)
Instead, they've successfully outsourced through you, and with no compensation to you, a few hundred euro of support costs to their competitors, and avoided losing their very thin margin on 10,000 euro of sales. And since you only deal in 10,000 euro of kit a year spread out over many sites and much time (and thus many purchase orders and incidents requiring human intervention), you're no big future loss either since selling one 10,000 euro pizza box to one customer is about 10 minutes of work for anyone in corporate sales, plus they would get to sell a support contract to go with it.
The attack relies on the attacker being able to guess the victim router's internal IP address, and to associate a host name of their choice with that internal address. Most routers will use their manufacturers' default addresses which are easy to guess. Since DNS rebinding relies on chance, forcing the attacker to make more incorrect guesses lowers the success rate of the attack. Therefore, attackers are unlikely to attempt to guess all of 10/8 or 192.168/16 etc. (tens of thousands of possibilities) when the vast majority of router addresses are at their defaults of 10.(0|8).(0|1).1 or 192.168.(0|1|123).1 etc. (around a dozen possibilities).
The population you rage for has close to zero members.
People who "rush into things" conduct a minimum of research. Apple's support forums do not show up in the first five pages of Bing or Google searches for "[apple] iphone 4".
People already aware of issues with the phone through Consumer Reports or another source would not need their information validated by Apple, especially since Apple offers no such validation.
If the caller persists despite your objections, you can always go to the police about harassment and obtain a police report number, and inform the caller of such. If the same caller persists, it quickly becomes criminal harassment. This mechanism pre-dates and operates outside of do not call lists.
Your post demonstrates another weakness of GIMP: the few knowledgeable and vocal members who publicly treat potential newcomers with distain, but yet wonder why they don't flock to GIMP and its abusive zealots en masse.
> As for tables, I make them in Excel then link them into Word. That is (to me) a heck of a lot easier than typing extra syntactic markup to get tables.
Word/Excel/Illustrator > Print to PDF > embed in LaTeX. Much easier to use a real DTP program to set and tune tables and figures, or when the entire imaginable area of a page needs to be accessed.
I'll take another look at TeX when it gains the ability to sanely line wrap underlined text such as URIs, and when it starts to respect j/k rules designed into typefaces.
>You don't have to stand on the shoulders of all giants simultaneously
Yes. That is the broad argument that all the apparently successful programmers here who've not heard of Knuth have tried to make. "some of the literature in the subject" is not the same as "Knuth's literature on the subject".
> Why not?
To restate a point made repeatedly elsewhere in this discussion: Hearing of a person's work, and understanding the work are two very different things.
At least some of the Knuth advocates in this discussion have argued that having "heard of" some "giant" in a field confers some qualification to work in that field. My contention is that merely hearing of Pauli's work does not qualify me as a theoretical chemist.
The danger in substituting actual knowledge with a reference to knowledge is that the knower starts to believe and operate in their false authority, while not having a strong incentive to become knowledgeable in those names they cite.
>But you can do that much more effectively when you know the old paradigm and can explain what's wrong with it.
That depends. To meaningfully "know" a paradigm means to have internalized the assumptions, structures, and "best practices" of that paradigm; that is, to be bound in thinking to the paradigm's current rules which tend not to recognise problems outside the comfortable parts of the problem domain. This is especially true of professionals and researchers who have heavily invested in physical or intellectual capital based on those assumptions. Contemporary examples include media distribution (assumes producers, exclusively a push model, and consumers), acute care (assumes the most effective solution is the one which resolves the most symptoms without necessarily addressing the root cause), and computing (assumes that vN model is reality, that the vN model is the best model, and that the vN model should be imposed on new computing substrates).
N00bs who are not wedded to such assumptions may have an easier time conceiving problems or elsewise thinking outside the existing paradigm.
Also, any successful new paradigm (has great explanatory value) must at least re-explore problems considered solved by the old paradigm if the old paradigm is to be displaced. Have a good conversation with anyone who has practiced within a single industry for a couple of generations (or read Thomas Kuhn's "The Structure of Scientific Revolutions" for a theoretical treatment of the same knowledge) if you would like a better understanding of this point. Or, closer to the ground, recall that almost every new technology (the result of a minor or major paradigm shift) starts out with substantial performance gaps when compared with the leading edge of the old as it works through and deals with old problems: solid state drives and latency; MP3 players and size; LCD panels and pixel switching time; digital cameras and resolution; etc. Expert readings and knowledge of servos from the old paradigms would have been non-useful, if not hindering, in thinking about how to solve the new problems. In some cases, "what's wrong with" an old paradigm may become the strength of the new paradigm (as with anti-bacterial moulds).
"The Statute of Anne, short title Copyright Act 1709 8 Anne c.19; long title An Act for the Encouragement of Learning, by vesting the Copies of Printed Books in the Authors or purchasers of such Copies, during the Times therein mentioned, was the first copyright statute in the Kingdom of Great Britain (thus the United Kingdom, see Copyright law of the United Kingdom). It was enacted in 1709 and entered into force on 10 April 1710. It is generally considered to be the first fully-fledged copyright statute. It is named for Queen Anne, during whose reign it was enacted. The Statute of Anne is now seen as the origin of copyright law.[1]"
To take the logical extension, mcvos' argument appears not to be that new knowledge can build on previous knowledge, but that new knowledge must build on all previous references to knowledge. This is dangerous since references to knowledge are not generally themselves knowledge, and because requiring new thinking to be limited to the assumptions set out in previous theories would systematically prevent the emergence of new ideas (not in or derived from the previous theory).
Thinking inside the box and working to define the nature of the box to exquisite levels of precision is a great way to get tenure and such, but rarely do such activities create new boxes or provide reasons to obsolete old ones.
Unlike you, I do not fear being wrong. "Wrong" is one of only two outcomes in a system of oppositional thinking, but it is the default and most probable outcome of most meaningful research and theory. In fact, most theories from which we develop applications have been inadequate and subsequently subsumed or supplanted by more accurate and useful theories with different dependencies. Look at the histories of medicine, mechanics, cosmology, chemistry, theism, justice, sociology, and any other field of endeavor for long lists of obsolete but at the time definitive and functional theories that rightly evoke "WTF?" today.
Being wrong and acknowledging that fact is the only way we arrive at new knowledge. Please join those of us working for the new science of tomorrow, while acknowledging the accomplishments of the past.
Knuth, and several hundred million other people of his time, invented and discovered all sorts of wonderful things. But to claim that no other discoveries can be made without using his inventions and discoveries implies that his work represents the final and correct summation about those particular aspects of the universe on which he worked.
Every human advance in history has been shown by further advances to be obsolete. Newton's theories were considered almost reality for 300 years, and continue to help us to generate reasonably good approximations of the physical world. But then relativity provided a much more accurate and powerful theory about the physical world without reference to Newton's kinematics, yet we do not consider that strange. We continue to use both for different purposes. But then again, we no longer practice blood-letting, nor burn witches, around which elaborate theories and self-consistent sciences were developed.
I take no issue with individuals worshipping particular texts; I simply prefer to live and explore the real world in addition to the models (however elegant) recorded in the cannon.
Surprise 1: Not every programmer makes breakthroughs, nor are they supposed to.
Surprise 2: It's not only possible, but routine to use, rediscover and surpass the work of previous scientists without knowing their individual names. AI and economics have and continue to both built glorious structures on top of the work of dozens of nameless enlightenment-era and ancient human naturalists. Almost every network administrator has used or built on the work of Dave Farber without knowing of his existence.
Hearing of Darwin and reading Darwin are almost completely independent events. For example, a basic reading of Darwin would assert that inheritance is exclusively unidirectional from parent to offspring within a line of descent within one species. Darwin's basic theory that beneficial mutations accrue based on environment precludes the discovery of plasmids and non-genetic inheritance in plants.
I note also that your argument claims to pertain to persons responsible for breakthroughs in biology, and yet you didn't object to the absence of Lovelace in my list.
Must anyone who works with life have heard of read LaPlace, Darwin, Watson and Crick, and Latour?
Must anyone who works with computers have heard of read the thousands of scripts of ancient and modern natural and religious philosophers setting out the foundations for logic?
How many engineers have read Newton or Schumpeter or would even be able to read their works?
Firefox developers know that their product will be used on Windows; they promote such usage. Through their collection of crash reporter data over several years and versions, they have reason to know about the common ways in which Windows fails when used in conjunction with Firefox. Not defending against such classes of failures (with code, documentation, education, or other means) when Mozilla knows and want their software to be used in conjunction with known and common failure modes is at least irresponsible on the personal part of each contributor, if not outright unprofessional of the developers and the organization.
The cleaning agent companies figured this out long ago, and responsibly advertise "do not mix X with Y" since they know the combination is harmful. Going back to 3.6 (if not much earlier), Mozilla knows that mixing Firefox, Skype, and Windows causes badness.
Even in the unlikely event that Skype and Microsoft are both completely uncooperative, a responsible thing to do would be to advertise "do not mix Firefox, Skype, and Windows under such and such circumstances" since Mozilla can anticipate the failure of their own product, and the negative impression of Firefox that such failure should leave upon the user.
For users of the components that crash most (Skype and UserCallWinProcCheckWow) in ways that crash them, the chance of a crash far exceed 1% since they are likely to want to use those components with Firefox repeatedly.
The installed base for the latest release is not the entire 250 million. 2.5 to 25 million would be a more reasonable estimate based on previous very deliberate and focused download Firefox campaigns which resulted in low millions of downloads in short periods.
210,000 crashes out of a population of 21-210 million runtime instances over the last two weeks, with basically a long tail distribution is conservatively a 0.1-1% chance of a crash. The long tail of the kinds of crashes indicates that whatever the source of the issue, Firefox has very poor generic exception handling capabilities such that most *classes* of possible badness have not been considered in the code design or implementation, let alone ways to gracefully recover from such badness.
The data show that Skype, UserCallWinProcCheckWow, and Firefox's own plugin handler are the top three crash instigators accounting for 10% of all crashes. This points to Firefox testers or developers not having tested well with a well-known use case (Skype), somehow using the Windows primitive incorrectly despite it being a well known and tested quantity in other applications, and a reasonable lack of understanding of how Firefox's new plugin handler may fail. Not surprisingly, Skype and UserCallWinProcCheckWow top the list of crash instigators in 3.6: http://crash-stats.mozilla.com/topcrasher/byversion/Firefox/3.6 That Firefox developers have known about those two issues for some time (having created several bugs around them) but have made no demonstrated progress in fixing them indicates a lack of willingness or ability (or both) to adequately understand and address the underlying issues.
Together, the lack of successful efforts to fix or defend against the top few specific issues at head of the distribution, along with the lack of successful efforts to address the exceedingly long tail, indicate that Firefox development is not currently being conducted with a reflexive or sustainable process.
Far less than 0.1% of Firefox users may have encountered issues. Far less than 0.1% of Internet users are active in the free software movement. There's no logical basis to trivialize the needs or views of a user population based only on their size as a fraction of the whole unless the goal is to achieve or enforce some boring homogeneity.
"He also cited the potential for refinements to TCP, DNS, and SSL/TLS to make the web a much faster place"
The core Internet protocol and infrastructure was and remains a conduit of innovation/because/ it is agnostic to HTTP and all other protocols. Optimizing for one small subset of its protocols and for a single kind of contemporary usage would discourage all kinds of innovation using protocols we've not conceived yet, and would be the single largest setback the modern Internet has seen.
Cxbx and derivatives appear to be making progress, but are confounded by needing to re-implement and test around 400 poorly documented kernel APIs and Direct* calls. Many calls are sufficiently different from those found in their desktop Windows analogues to prevent borrowing from Wine or passing through directly to the host Windows OS. Given that the scope of the task is approximately as large as implementing Wine but with more difficult speed and multimedia requirements, I'd expect to see a generally playable Xbox emulator in another six years.
You ever stop and think about WHY healthcare is public?
That's only a series of mirrors. Primary health care in Canada is public insofar as doctors (essentially private contractors) get paid by the government. The current payment structure rewards practitioners on each visit, so there's a perverse incentive for practitioners to minimize the time spent on each patient, and to maximize throughput. My family physician expertly sees 30 patients an hour (mind like a steel trap) during peak periods (business tower yuppies and recent immigrants) and takes half the week off to not exceed quotas. Various colleagues of mine have had to visit their GPs or the medical clinic over 30 times a year because the payment schedule and performance measures for practitioners incentivizes treating every tangent of the condition before being allowed to address the root cause.
When you allow people with more money to have better healthcare you are putting their lives at a greater value than those less fortunate.
Yes. In Canada, this is most clearly evident when comparing the availability of routine dental and eye care (essentially entirely privatized and not covered under mandatory provincial health insurance) available to insured and uninsured individuals.
I don't care where you live but that's not right. The first case we had of someone being able to pay more for better healthcare was last year I think. I'm not sure what loophole they used to legally do this. There was a social uproar about it.
Health clubs are great for that. Essentially, some practitioners (GPs, specialists, nurses, surgeons, etc.) get together to own a private club which also provides instant on-site medical care (not illegal). Membership in said club costs several kilobux per year (not illegal). The practitioners end up seeing only a handful of patients per day compared to their clinic or professional corporation counterparts (not illegal since there's no minimum quota on the number of patients seen), and everyone involved gets paid from dividends on the club several times more than the wages other practitioners receive. (Incidentally, this is not much different than the model used by most non-acute freestanding GPs and specialists, health clinics, etc. since every practitioner gets to arbitrarily determine who gets to be a patient, when they receive care, and to discontinue care whenever they view the patient to be uncooperative.)
Within the acute care (hospital) system, because the operational model remains one adapted from World War I field medicine, there are very few mechanisms to keep individuals accountable (or even to be aware of what they're doing) on a system level, so total costs are forced to increase by 5-10 per cent per year without an understanding of where the money goes. The unions claim that keeping records which would show that their members are performing to their job descriptions would compromise patient care, and the public somehow agrees that it's a good idea to keep putting more money into a system which produces patient care of declining quality as measured by outcomes and quality of life.
Most Americans and even Canadians do not realize that we already have an unsustainable two-tiered health care system, wherein the system has driven so many practitioners out of the country that most Canadians must wait on-site for 6-24 hours for urgent care, and sometimes years for non-emergency surgery.
> That's both the benefit and the drawbacks of closed source - nobody will know the problem exists but nobody will be around to fix it either.
How does the attribute of open source enable users of Firefox or Apache httpd to find out about problems that exist but are filed away in the not publicly accessible security sections of their respective bugzillas and listservs?
With respect to the image handling, the failures on the part of PHP fell outside of the exception handling system presented to the user, and outside its internal facilities. Its attempts to parse some images walled the machines to the extent that local and remote ttys became unresponsive.
A well implemented system should include enough safeguards to prevent the system from failing so spectacularly on easily anticipated use cases, regardless of the inputs that code monkeys, such as allegedly myself, can mash up and throw at it.
In this case, PHP has compromised itself by adding some unsafe logic before or after passing valid data to a known working library. That fault would present itself as much to bad code monkeys and esteemed PHP professionals such as yourself.
Speaking of which, could you please direct me to the PHP professionals code of conduct or the like? I'd imagine that it's the same boilerplate that most others use, but I've not been able to locate one through my simple code monkey Google ful.
When recently helping a colleague with a CMS migration of a few million rows, I had to stop tracking the number of times and ways in which using string functions on not particularly interesting user-submitted data would cause PHP to segfault. It made no difference whether we ran the migration scripts from the command line or an Apache instance, nor whether the host was OS X, Windows, or the various Linux distros we had on hand. And then we got into the image parsing functions, which choke on files saved with particular non-spectacular options in Photoshop (and others) when the underlying GD libraries handle with no issue. Thankfully, we were only using PHP to get the data into the new system, and not as the foundation of the new system itself.
We can only observe and measure things that have happened, not things that will happen. Science enables us to organize those observations to draw predictions about things that could happen in the future, and the likelihood of their occurrence given a set of conditions which we may or may not be able to manipulate.
Half of a a lump of element with a half life will decay into something else in a certain time, but there's no way to determine which half decays.
Through repeated observation, we know that adding a small amount of alcohol to the bloodstream of several chordates has yielded poor motor coordination in many cases, and that the effect is causal. Without doing the experiment, all we can say is that alcohol has a high, but not 100% probability, that any member of a species of chordate, when alcohol is added to their bloodstream, of experiencing diminished motor control, and we couldn't even say with certainty by how much the control is diminished. Through repeated sampling, you could arrive at an average with high confidence, which would be useful to predict some kind of average impairment for a particular quanta of alcohol, but even then there would be exceptions.
It is the assertion of direct and unvarying relationships between cause and effect which hasn't been science since we got past the mechanical model of the universe a bit after Newton's time. That type of thinking might not even be a sophisticated way of handling faith. Even the Vatican acknowledges exceptions to such assertions.
Slashdot Mirror
> in total about 10 thousand euros of lost sales for Cisco/Linksys because of that one crap router they saddled me with for Christmas 2008
So their filter against non-profitable clients has worked as expected.
Each time a human at Linksys touches a customer, the company incurs at least 5 euro in costs. Since Linksys relies on retail volume and not consultation for their consumer sales, it's to their financial advantage to never hear from customers once the sale has been made, and especially to their advantage not to have to respond to unending lists of complaints or questions from detail-oriented customers. That same 10,000 euro of kit sold to 200 customers who do not generally know enough to complain is much more profitable to Linksys than if it were sold to you since you have both the aptitude and time to complain, but not effectively. (If you had complained effectively, you would have received a successful resolution from Linksys and both parties would have benefitted directly.)
Instead, they've successfully outsourced through you, and with no compensation to you, a few hundred euro of support costs to their competitors, and avoided losing their very thin margin on 10,000 euro of sales. And since you only deal in 10,000 euro of kit a year spread out over many sites and much time (and thus many purchase orders and incidents requiring human intervention), you're no big future loss either since selling one 10,000 euro pizza box to one customer is about 10 minutes of work for anyone in corporate sales, plus they would get to sell a support contract to go with it.
The attack relies on the attacker being able to guess the victim router's internal IP address, and to associate a host name of their choice with that internal address. Most routers will use their manufacturers' default addresses which are easy to guess. Since DNS rebinding relies on chance, forcing the attacker to make more incorrect guesses lowers the success rate of the attack. Therefore, attackers are unlikely to attempt to guess all of 10/8 or 192.168/16 etc. (tens of thousands of possibilities) when the vast majority of router addresses are at their defaults of 10.(0|8).(0|1).1 or 192.168.(0|1|123).1 etc. (around a dozen possibilities).
The population you rage for has close to zero members.
People who "rush into things" conduct a minimum of research. Apple's support forums do not show up in the first five pages of Bing or Google searches for "[apple] iphone 4".
People already aware of issues with the phone through Consumer Reports or another source would not need their information validated by Apple, especially since Apple offers no such validation.
If the caller persists despite your objections, you can always go to the police about harassment and obtain a police report number, and inform the caller of such. If the same caller persists, it quickly becomes criminal harassment. This mechanism pre-dates and operates outside of do not call lists.
Your post demonstrates another weakness of GIMP: the few knowledgeable and vocal members who publicly treat potential newcomers with distain, but yet wonder why they don't flock to GIMP and its abusive zealots en masse.
> As for tables, I make them in Excel then link them into Word. That is (to me) a heck of a lot easier than typing extra syntactic markup to get tables.
Word/Excel/Illustrator > Print to PDF > embed in LaTeX. Much easier to use a real DTP program to set and tune tables and figures, or when the entire imaginable area of a page needs to be accessed.
I'll take another look at TeX when it gains the ability to sanely line wrap underlined text such as URIs, and when it starts to respect j/k rules designed into typefaces.
If the bandwidth provider's top priority is maximizing cash input from whatever source, they would be the wrong bandwidth provider for this project.
>You don't have to stand on the shoulders of all giants simultaneously
Yes. That is the broad argument that all the apparently successful programmers here who've not heard of Knuth have tried to make. "some of the literature in the subject" is not the same as "Knuth's literature on the subject".
> Why not?
To restate a point made repeatedly elsewhere in this discussion: Hearing of a person's work, and understanding the work are two very different things.
At least some of the Knuth advocates in this discussion have argued that having "heard of" some "giant" in a field confers some qualification to work in that field. My contention is that merely hearing of Pauli's work does not qualify me as a theoretical chemist.
The danger in substituting actual knowledge with a reference to knowledge is that the knower starts to believe and operate in their false authority, while not having a strong incentive to become knowledgeable in those names they cite.
>But you can do that much more effectively when you know the old paradigm and can explain what's wrong with it.
That depends. To meaningfully "know" a paradigm means to have internalized the assumptions, structures, and "best practices" of that paradigm; that is, to be bound in thinking to the paradigm's current rules which tend not to recognise problems outside the comfortable parts of the problem domain. This is especially true of professionals and researchers who have heavily invested in physical or intellectual capital based on those assumptions. Contemporary examples include media distribution (assumes producers, exclusively a push model, and consumers), acute care (assumes the most effective solution is the one which resolves the most symptoms without necessarily addressing the root cause), and computing (assumes that vN model is reality, that the vN model is the best model, and that the vN model should be imposed on new computing substrates).
N00bs who are not wedded to such assumptions may have an easier time conceiving problems or elsewise thinking outside the existing paradigm.
Also, any successful new paradigm (has great explanatory value) must at least re-explore problems considered solved by the old paradigm if the old paradigm is to be displaced. Have a good conversation with anyone who has practiced within a single industry for a couple of generations (or read Thomas Kuhn's "The Structure of Scientific Revolutions" for a theoretical treatment of the same knowledge) if you would like a better understanding of this point. Or, closer to the ground, recall that almost every new technology (the result of a minor or major paradigm shift) starts out with substantial performance gaps when compared with the leading edge of the old as it works through and deals with old problems: solid state drives and latency; MP3 players and size; LCD panels and pixel switching time; digital cameras and resolution; etc. Expert readings and knowledge of servos from the old paradigms would have been non-useful, if not hindering, in thinking about how to solve the new problems. In some cases, "what's wrong with" an old paradigm may become the strength of the new paradigm (as with anti-bacterial moulds).
http://en.wikipedia.org/wiki/Statute_of_Anne
"The Statute of Anne, short title Copyright Act 1709 8 Anne c.19; long title An Act for the Encouragement of Learning, by vesting the Copies of Printed Books in the Authors or purchasers of such Copies, during the Times therein mentioned, was the first copyright statute in the Kingdom of Great Britain (thus the United Kingdom, see Copyright law of the United Kingdom). It was enacted in 1709 and entered into force on 10 April 1710. It is generally considered to be the first fully-fledged copyright statute. It is named for Queen Anne, during whose reign it was enacted.
The Statute of Anne is now seen as the origin of copyright law.[1]"
To take the logical extension, mcvos' argument appears not to be that new knowledge can build on previous knowledge, but that new knowledge must build on all previous references to knowledge. This is dangerous since references to knowledge are not generally themselves knowledge, and because requiring new thinking to be limited to the assumptions set out in previous theories would systematically prevent the emergence of new ideas (not in or derived from the previous theory).
Thinking inside the box and working to define the nature of the box to exquisite levels of precision is a great way to get tenure and such, but rarely do such activities create new boxes or provide reasons to obsolete old ones.
Unlike you, I do not fear being wrong. "Wrong" is one of only two outcomes in a system of oppositional thinking, but it is the default and most probable outcome of most meaningful research and theory. In fact, most theories from which we develop applications have been inadequate and subsequently subsumed or supplanted by more accurate and useful theories with different dependencies. Look at the histories of medicine, mechanics, cosmology, chemistry, theism, justice, sociology, and any other field of endeavor for long lists of obsolete but at the time definitive and functional theories that rightly evoke "WTF?" today.
Being wrong and acknowledging that fact is the only way we arrive at new knowledge. Please join those of us working for the new science of tomorrow, while acknowledging the accomplishments of the past.
Knuth, and several hundred million other people of his time, invented and discovered all sorts of wonderful things. But to claim that no other discoveries can be made without using his inventions and discoveries implies that his work represents the final and correct summation about those particular aspects of the universe on which he worked.
Every human advance in history has been shown by further advances to be obsolete. Newton's theories were considered almost reality for 300 years, and continue to help us to generate reasonably good approximations of the physical world. But then relativity provided a much more accurate and powerful theory about the physical world without reference to Newton's kinematics, yet we do not consider that strange. We continue to use both for different purposes. But then again, we no longer practice blood-letting, nor burn witches, around which elaborate theories and self-consistent sciences were developed.
I take no issue with individuals worshipping particular texts; I simply prefer to live and explore the real world in addition to the models (however elegant) recorded in the cannon.
Why are you afraid of being wrong?
Surprise 1: Not every programmer makes breakthroughs, nor are they supposed to.
Surprise 2: It's not only possible, but routine to use, rediscover and surpass the work of previous scientists without knowing their individual names. AI and economics have and continue to both built glorious structures on top of the work of dozens of nameless enlightenment-era and ancient human naturalists. Almost every network administrator has used or built on the work of Dave Farber without knowing of his existence.
Hearing of Darwin and reading Darwin are almost completely independent events. For example, a basic reading of Darwin would assert that inheritance is exclusively unidirectional from parent to offspring within a line of descent within one species. Darwin's basic theory that beneficial mutations accrue based on environment precludes the discovery of plasmids and non-genetic inheritance in plants.
I note also that your argument claims to pertain to persons responsible for breakthroughs in biology, and yet you didn't object to the absence of Lovelace in my list.
Must anyone who works with life have heard of read LaPlace, Darwin, Watson and Crick, and Latour?
Must anyone who works with computers have heard of read the thousands of scripts of ancient and modern natural and religious philosophers setting out the foundations for logic?
How many engineers have read Newton or Schumpeter or would even be able to read their works?
Firefox developers know that their product will be used on Windows; they promote such usage. Through their collection of crash reporter data over several years and versions, they have reason to know about the common ways in which Windows fails when used in conjunction with Firefox. Not defending against such classes of failures (with code, documentation, education, or other means) when Mozilla knows and want their software to be used in conjunction with known and common failure modes is at least irresponsible on the personal part of each contributor, if not outright unprofessional of the developers and the organization.
The cleaning agent companies figured this out long ago, and responsibly advertise "do not mix X with Y" since they know the combination is harmful. Going back to 3.6 (if not much earlier), Mozilla knows that mixing Firefox, Skype, and Windows causes badness.
Even in the unlikely event that Skype and Microsoft are both completely uncooperative, a responsible thing to do would be to advertise "do not mix Firefox, Skype, and Windows under such and such circumstances" since Mozilla can anticipate the failure of their own product, and the negative impression of Firefox that such failure should leave upon the user.
For users of the components that crash most (Skype and UserCallWinProcCheckWow) in ways that crash them, the chance of a crash far exceed 1% since they are likely to want to use those components with Firefox repeatedly.
The installed base for the latest release is not the entire 250 million. 2.5 to 25 million would be a more reasonable estimate based on previous very deliberate and focused download Firefox campaigns which resulted in low millions of downloads in short periods.
210,000 crashes out of a population of 21-210 million runtime instances over the last two weeks, with basically a long tail distribution is conservatively a 0.1-1% chance of a crash. The long tail of the kinds of crashes indicates that whatever the source of the issue, Firefox has very poor generic exception handling capabilities such that most *classes* of possible badness have not been considered in the code design or implementation, let alone ways to gracefully recover from such badness.
The data show that Skype, UserCallWinProcCheckWow, and Firefox's own plugin handler are the top three crash instigators accounting for 10% of all crashes. This points to Firefox testers or developers not having tested well with a well-known use case (Skype), somehow using the Windows primitive incorrectly despite it being a well known and tested quantity in other applications, and a reasonable lack of understanding of how Firefox's new plugin handler may fail. Not surprisingly, Skype and UserCallWinProcCheckWow top the list of crash instigators in 3.6:
http://crash-stats.mozilla.com/topcrasher/byversion/Firefox/3.6
That Firefox developers have known about those two issues for some time (having created several bugs around them) but have made no demonstrated progress in fixing them indicates a lack of willingness or ability (or both) to adequately understand and address the underlying issues.
Together, the lack of successful efforts to fix or defend against the top few specific issues at head of the distribution, along with the lack of successful efforts to address the exceedingly long tail, indicate that Firefox development is not currently being conducted with a reflexive or sustainable process.
Far less than 0.1% of Firefox users may have encountered issues. Far less than 0.1% of Internet users are active in the free software movement. There's no logical basis to trivialize the needs or views of a user population based only on their size as a fraction of the whole unless the goal is to achieve or enforce some boring homogeneity.
/disappointed.
"He also cited the potential for refinements to TCP, DNS, and SSL/TLS to make the web a much faster place"
The core Internet protocol and infrastructure was and remains a conduit of innovation /because/ it is agnostic to HTTP and all other protocols. Optimizing for one small subset of its protocols and for a single kind of contemporary usage would discourage all kinds of innovation using protocols we've not conceived yet, and would be the single largest setback the modern Internet has seen.
A Shatner Speech... from... the... Throne would be... genuinely... epic... but only... that one time.
Cxbx and derivatives appear to be making progress, but are confounded by needing to re-implement and test around 400 poorly documented kernel APIs and Direct* calls. Many calls are sufficiently different from those found in their desktop Windows analogues to prevent borrowing from Wine or passing through directly to the host Windows OS. Given that the scope of the task is approximately as large as implementing Wine but with more difficult speed and multimedia requirements, I'd expect to see a generally playable Xbox emulator in another six years.
... to boldly go ...
You ever stop and think about WHY healthcare is public?
That's only a series of mirrors. Primary health care in Canada is public insofar as doctors (essentially private contractors) get paid by the government. The current payment structure rewards practitioners on each visit, so there's a perverse incentive for practitioners to minimize the time spent on each patient, and to maximize throughput. My family physician expertly sees 30 patients an hour (mind like a steel trap) during peak periods (business tower yuppies and recent immigrants) and takes half the week off to not exceed quotas. Various colleagues of mine have had to visit their GPs or the medical clinic over 30 times a year because the payment schedule and performance measures for practitioners incentivizes treating every tangent of the condition before being allowed to address the root cause.
When you allow people with more money to have better healthcare you are putting their lives at a greater value than those less fortunate.
Yes. In Canada, this is most clearly evident when comparing the availability of routine dental and eye care (essentially entirely privatized and not covered under mandatory provincial health insurance) available to insured and uninsured individuals.
I don't care where you live but that's not right. The first case we had of someone being able to pay more for better healthcare was last year I think. I'm not sure what loophole they used to legally do this. There was a social uproar about it.
Health clubs are great for that. Essentially, some practitioners (GPs, specialists, nurses, surgeons, etc.) get together to own a private club which also provides instant on-site medical care (not illegal). Membership in said club costs several kilobux per year (not illegal). The practitioners end up seeing only a handful of patients per day compared to their clinic or professional corporation counterparts (not illegal since there's no minimum quota on the number of patients seen), and everyone involved gets paid from dividends on the club several times more than the wages other practitioners receive. (Incidentally, this is not much different than the model used by most non-acute freestanding GPs and specialists, health clinics, etc. since every practitioner gets to arbitrarily determine who gets to be a patient, when they receive care, and to discontinue care whenever they view the patient to be uncooperative.)
Within the acute care (hospital) system, because the operational model remains one adapted from World War I field medicine, there are very few mechanisms to keep individuals accountable (or even to be aware of what they're doing) on a system level, so total costs are forced to increase by 5-10 per cent per year without an understanding of where the money goes. The unions claim that keeping records which would show that their members are performing to their job descriptions would compromise patient care, and the public somehow agrees that it's a good idea to keep putting more money into a system which produces patient care of declining quality as measured by outcomes and quality of life.
Most Americans and even Canadians do not realize that we already have an unsustainable two-tiered health care system, wherein the system has driven so many practitioners out of the country that most Canadians must wait on-site for 6-24 hours for urgent care, and sometimes years for non-emergency surgery.
> That's both the benefit and the drawbacks of closed source - nobody will know the problem exists but nobody will be around to fix it either.
How does the attribute of open source enable users of Firefox or Apache httpd to find out about problems that exist but are filed away in the not publicly accessible security sections of their respective bugzillas and listservs?
With respect to the image handling, the failures on the part of PHP fell outside of the exception handling system presented to the user, and outside its internal facilities. Its attempts to parse some images walled the machines to the extent that local and remote ttys became unresponsive.
A well implemented system should include enough safeguards to prevent the system from failing so spectacularly on easily anticipated use cases, regardless of the inputs that code monkeys, such as allegedly myself, can mash up and throw at it.
In this case, PHP has compromised itself by adding some unsafe logic before or after passing valid data to a known working library. That fault would present itself as much to bad code monkeys and esteemed PHP professionals such as yourself.
Speaking of which, could you please direct me to the PHP professionals code of conduct or the like? I'd imagine that it's the same boilerplate that most others use, but I've not been able to locate one through my simple code monkey Google ful.
When recently helping a colleague with a CMS migration of a few million rows, I had to stop tracking the number of times and ways in which using string functions on not particularly interesting user-submitted data would cause PHP to segfault. It made no difference whether we ran the migration scripts from the command line or an Apache instance, nor whether the host was OS X, Windows, or the various Linux distros we had on hand. And then we got into the image parsing functions, which choke on files saved with particular non-spectacular options in Photoshop (and others) when the underlying GD libraries handle with no issue. Thankfully, we were only using PHP to get the data into the new system, and not as the foundation of the new system itself.
"Would probably" is the essence of science.
We can only observe and measure things that have happened, not things that will happen. Science enables us to organize those observations to draw predictions about things that could happen in the future, and the likelihood of their occurrence given a set of conditions which we may or may not be able to manipulate.
Half of a a lump of element with a half life will decay into something else in a certain time, but there's no way to determine which half decays.
Through repeated observation, we know that adding a small amount of alcohol to the bloodstream of several chordates has yielded poor motor coordination in many cases, and that the effect is causal. Without doing the experiment, all we can say is that alcohol has a high, but not 100% probability, that any member of a species of chordate, when alcohol is added to their bloodstream, of experiencing diminished motor control, and we couldn't even say with certainty by how much the control is diminished. Through repeated sampling, you could arrive at an average with high confidence, which would be useful to predict some kind of average impairment for a particular quanta of alcohol, but even then there would be exceptions.
It is the assertion of direct and unvarying relationships between cause and effect which hasn't been science since we got past the mechanical model of the universe a bit after Newton's time. That type of thinking might not even be a sophisticated way of handling faith. Even the Vatican acknowledges exceptions to such assertions.