...they also say that if you have an IPv6-reachable service up by 6Bone Day, to write in and let them know.
Examples of services that can be configured to be IPv6-reachable would be websites using Apache or Roxen, the games of Empire or PennMUSH - dunno if Freeciv is IPv6-aware or not, the Yum RPM updater - not sure about Apt, BIND, and probably a fair few others I can't think of right now. (They don't specify if games would be acceptable or not - but anyone who hosts a celebration on 6-6-6 is probably not hopelessly straight-laced.)
If you want to set something up, I'd encourage it. All you need do is set up a tunnel with one of the standard free brokerage services, enable IPv6 on your box, and have at it. I'd particularly encourage it if it means there's a snowball's chance in hell (!) of IPv6 being seen as used.
...that it's "unlikely" and that there's "not much" history to back it up. They don't offer an alternative theory, or an explanation for the "football" in the name if the game did evolve independently. I totally approve of skeptical thinking, but am hesitent when it sounds like it's getting close to cynical thinking, and hold to the principle that it's better to have an idea that's wrong but can be modified than to have no ideas at all.
Of course, that last part may explain why I never did well on the statistics coursework.
...is if they could do a "development" respin with all of the development RPMs as well as just the updates. If you think doing a "yum update" for just the vanilla updates is bad, that's nothing compared to the development stuff. True, the development stuff is aimed more for developers and testers than the general public, but (a) the number of testers/developers bound to be limited by how practical it is to become one, and (b) it's generally a better idea to have as close to a "clean" and well-determined configuration as possible, when setting up such a box, which would seem to be precluded if you are replacing virtually every file on the system.
Now, I understand that testers aren't their primary audience, that building the ISO images & setting up the installers is not always a trivial task, and that the development files change frequently enough that large updates are almost inevitable without an automated ISO build script, so I am not saying that they "should" build such images - merely that it would be very handy for some of us if they did.
Football, when it started, was not restricted to the use of feet. Indeed, banning the use of hands only started after an enterprising young lad from the city of Rugby picked the ball up and ran with it to the goal. (Even today, this variant is called "Rugby Football", even though feet are rarely used outside of a scrum.)
So although the name appears to be logical, history would imply that it is not quite as simple. The origins of the game are sufficiently ancient and obscure - although it almost certainly started in the British Isles - that there is no absolute guarantee that the original name was even in English and therefore may be merely phonetically derived from an earlier name. That happens a lot. In that case, the modern name would have no meaning whatsoever.
Sprint suffered a cascading router failure in and around their transatlantic gateway some time back - Augustish 1995, I think - which caused an almost total outage between the UK and the US for a period of about two weeks. That's the only prolonged failure of any significance that I can remember.
For temporary slow-downs, certainly major events cause problems, and most of those are indeed caused by streaming. More specifically, unicast streaming. If streaming was predominatly multicast, there would be no meaningful load imposed, no matter how many people had broadband.
It could be a neutral mod - +0 - Car Analogy. In fact, you could have a whole range of descriptive mods. These would be useful when the score is appropriate (so you don't want to give it a +1 or a -1) but the classification just doesn't fit. These could include a +0 on-topic, +0 mostly harmless, etc.
...would simply be to put Acrobat Reader's signature into Microsoft's antivirus software. Another would be to block javascript connections from any application called "acroread" - Acrobat's reader uses javascript for DRM, so firewalling it would make it impossible to open protected Acrobat files. That could even be added as a "security feature" - they've broken apps that way before on that very excuse. PDF could automatically be associated with Notepad every time you reboot. There are lots of ways Microsoft could make life VERY hard on Acrobat users, without breaking other applications, or at least not too many of them.
This is not to say that they will do any of these things. I've neither seen nor heard any evidence that Microsoft is going to "knife the baby" the way they have done with other companies they didn't see eye-to-eye with. My knowledge is limited to the fact that they have "knifed the baby" with other companies and have therefore demonstrated a capacity to think in such terms. I think it reasonable to give Microsoft the benefit of the doubt - to an extent. That extent being that we should keep our eyes and minds open to the possibility of malign intent. However, that's only any good if we know what "malign intent" looks like.
My original posting - and any follow-ups of mine - may well be incorrect on specifics. I'm not perfect (yet, anyways). If it gets people to think about what actually would demonstrate "good intent" or "malign intent", then it has served its purpose, even if I'm wrong on every single method. (Now, if I did believe I was wrong on every single point, I would have written other points - I might invite "vigorous debate" at times, but I try hard not to troll or flamebait. I'm also practicing on admitting when I am wrong, but when things are this speculative, it can be hard to tell.)
The problem is that Microsoft can drag things out a while. They lost in the DR-DOS court case, but that was started in the days of Windows 3.1 and finished in the days of Windows 2003. The costs to Microsoft were insignificant, but were probably rather more substantial to the plaintiffs. The EU anti-trust case was "extended" by a few years by Microsoft foot-dragging and will likely take years, if not decades, to go through all of Microsoft's inevitable appeals. By the time the case is over, the EU's threat to restrict/block Vista will be of no consequence - Microsoft will be two, if not three, versions down the road and just about ready to ditch Vista anyway, and the interest earned on the money not paid in fines now will likely more than cover the fines that are imposed. I can't see how Microsoft can possibly lose in any meaningful way.
Microsoft also lost their case regarding integrating things into Windows 95. The appeals court overturned it on the technicality that Microsoft had renamed the product Windows 98.
Now, they have lost cases in the past and been forced to back things out relatively quickly. Their piracy of disk compression tools in DOS 6.0 forced them to release 6.2 with those tools absent. Assorted tools in Windows 5.0 that carried the copyright of others were also amazingly missing from 6.2. Then there was the Java fiasco, where they "embraced and extended" rather beyond their license to do so. There are probably more recent examples of cases where they've lost and had to remove stuff as a result. So, yes, it can happen and has happened. The question then becomes one of whether Microsoft has learned not to do that, or whether they've merely learned not to be quite so obvious about it. My hope is the former, my fear is the latter.
Microsoft has enough money to buy Adobe out in a hostile bid, pay off any regulators, build the CEV, and dance on the grave of Acrobat on Mars, should they so choose.
Microsoft has more than enough lawyers to bankrupt Adobe through litigation - a tactic they've not used that much, but isn't unknown to them, or to the cut-throat world of corporate business in general.
Microsoft can certainly splurge the business world with FUD and scare tactics - they've killed competitors off that way.
They've stolen plenty of technology before - it wouldn't be hard for them to put Acrobat-reading code into Office without permission, then drag things out in the court for a few decades until it becomes irrelevant.
They could also modify Windows to make running Acrobat impossible (it worked in killing DR-DOS) or impractical (their method of disembowling Netscape).
Are any of these likely? We'll almost certainly have an answer by the time Vista comes out, and quite possibly by the time XP or 2003 have another service pack. We do know for certain that they've never liked being spurned and their competitors have a strange habit of dying terrible deaths.
Are Adobe aware of this? Oh, almost certain. They're almost guaranteed to be just as aware that allies of Microsoft have an equally strange reduction in their life expectancy. In the most recent case, Microsoft worked with anti-virus companies then bought one out and produced their own. Oddly, it seems to recognize some of the competing anti-virus products as hostile and destroy them. Curious, that.
My guess is that Adobe takes the line that if they're going to die, they might as well die with their boots on, and there's a slim chance Microsoft won't dare crush them with all of the other anti-trust cases going on, but that if they go with Microsoft, they don't stand a chance. That's based on what has visibly occurred, however, and Adobe's actual (but not necessarily stated) reasons may be very different.
...that Open Source encryption software is exempt from access controls, provided the US Government is notified of where the source is, and that there are no constraints on binary-only encrypted software. So, if you are using a standard Open Source library for encryption that has been properly registered by its authors, I wouldn't see that there was a problem. If you don't already do so, it might be a good idea to state what encryption software is used - it would make it harder for anyone to reasonably claim it might be proprietary (as proprietary algorithms aren't exempt). A lawyer would be able to give a clearer answer, but bear in mind that 100% of all court cases involve lawyers losing.
Alternative Business Model
on
Death By DMCA
·
· Score: 2, Interesting
Also known as the British Broadcasting Corporation (BBC for short). It is arguably a subscription channel - if you have a television, you pay your subscription, in the form of the license fee - and it is 100% advert-free. When it's bad, it's pretty bad (but still better than any of the US cable channels at their worst). When it's good, it's damn good.
How to translate this into the US system? Well, let's start with PBS, as that is the only public broadcasting service the US has. To be on-par, PBS would need to eliminate the sponsorship system, eliminate the adverts, get better backing from Government, broaden the content, improve the quality of productions to modern cinematic standards (documentaries shouldn't have the 1930s Pathe Newsreel feel to them), and carry out independent work (history should not be read from a textbook, and news should not be read from AP bulletins).
Next come the existing openly subscriber channels and pay-per-view. These should, really, be reaping the full cost of everything (plus profit) from the material they sell. If they don't, then the material is either grossly undervalued or grossly inferior. People are generally happy to pay for things that are worth the cost to them, so either the pricing is incorrect or the material is. Or both.
Finally, the "free" advert-laden channels. In the end, adverts cost the producers of the advertised material money. This money will end up being added to the cost of goods. Since the cost of material doesn't depend on who is paying for it, this will work out to be comparable to any of the subscriber channel costs. Only, you're paying for it whether you watch those channels or not! It's a tax on goods, going through the corporations rather than through Government, but it's still a tax. Since it is a tax, why not have it collected by the people collecting taxes anyway? It won't change how much you end up paying for your cost of living, but it will add about 15-20 minutes of material to every show, increasing the value of watching them.
(If you're going to pay $X extra because of an invisible sales tax created by advertising, it makes no difference to you in your overall costs if - instead - those same goods were $X cheaper and you had an $X flat tax to cover broadcasting in that area. $X - $X = 0.)
Actually, that's not totally correct. Those in adverts get paid royalties for every time the advert is shown. This costs the advertisers more, which they'll defray by making you pay for it by more expensive goods. This will not be exactly the same as the increase in production costs by making shows 15 minutes longer. In some cases, the cost of the adverts will be more. In other cases, the cost of the shows will be more. You would need to quantify this, to prove conclusively that the BBC model of the license fee would work in these cases. My suspicion is that you'll find that the license fee is indeed the superior model, but in either case, the difference has to be insignificant as none of the other costs for those channels is going to vary.
(You asked for an alternative model. You didn't ask for one Americans would stomach. I know perfectly well that even if every household in America saved hundreds or even thousands of dollars a year from a license model, and even if it meant program quality skyrocketted far beyond the wildest imaginings of anyone alive today, you'd be risking an armed uprising before Americans would consider a new overt tax from Government, no matter how covertly they were being taxed by corporations already.)
...for a contest that could potentially draw an entry from virtually every scientist, academic, researcher, or otherwise pro-understanding individual in the United States. The potential exists for a few tens of millions of entries, especially given the current disillusionment.
eg: NASA is currently cutting back or eliminating many science missions in order to pay for the next-gen shuttle, which assumes Congress won't cut NASA's budget over the next ten years -and- there are no cost overruns anywhere, according to the New Scientist.
eg: The US has spent a miniscule fraction of what it pledged and committed in the fight against bird flu, according to the World Bank. Whether an epidemic ever occurs is irrelevant in this. What matters is that even hard-nosed financial institutions are getting concerned. When the economists think Government is spending too little, it's time to be worried.
(I'm not singling out the US because it's particularly bad amongst nations - it actually does better than most - but because that's what the contest is about. Had this been an international contest, I'm sure I could find alarming attitudes in every civilized nation on the planet. It wasn't that long ago that the South African Health Minister promoted garlic as a cure for AIDS. Although I suppose there might be a lot of vampires in South African politics.)
I just don't know how this project can possibly reach its true potential with such limited backing. Most who could enter a truly biting cartoon won't be bothered, because there won't be any perceived value. If getting into the final rounds constituted a publication in a peer-reviewed forum, then perhaps there would be more interest. Money from pro-science organizations towards prizes would have been good, too.
For those on Slashdot with no artistic talents - enter anyway. Most scientists can't draw worth a damn, so it'll be purely down to the ideas in the cartoon anyway. Besides, there are valid reasons for believing the readers here have a broader understanding of the state of science and the attitudes around it - those focused totally on their subject won't have time to read up on anything much outside of their speciality and so won't be able to so easily draw on attitudes and perceptions that are universal.
I believe "correctly encrypted" would be more correctly written as "will not be extractable by brute-forcing the key of an otherwise uncompromised encryption algorithm, within the next century, assuming classical computation and the continuation of Moore's Law, by any means affordable to a person or organization in the income bracket of greatest concern" - only, the latter is a lot more wordy and doesn't really add that much.
Yes, you do need all that. A compromised encryption algorithm allows a person to reduce the effective key length and thereby deduce the actual encryption key used comparitively quickly. You've got to assume classical computers - with one exception - as quantum computers should be able to go through all the possible keys exponentially faster. You've also got to assume a price bracket, because an infinite amount of money would allow you to do just about anything.
The one exception is for One Time Pads. These decrypt into EVERY possible string of equal length to the encrypted string, so it is totally impossible to tell which solution is the right one. One Time Pads are unbreakable without the key. Totally. It isn't even a question of nobody having a means of breaking them, it can't be done. The problem is that the key has to be as long as the data you are trying to encrypt. It trades one problem for another of equal size. So although they are theoretically perfect, they are also practically useless in most cases. Oh, and the "one time" bit is for real - the key, to be secure, can only be used once. You have to make a new key each time, with each key being genuinely random - pseudo-random isn't random enough.
No, you can validate code reasonably well within bounds. For example, you can certainly prove that a module has no residual impact and that inputs within a given range will produce outputs that are valid (maybe not correct, but at least valid). That does not require a full formal proof.
I'd also argue that although formal proofs are horribly complex and a nightmare for any significant code, they are not as impossible as is often claimed. Code doesn't have to be proven linearly. If you can prove a function correct (assuming that all calls function correctly, and the initial conditions are valid), then you can prove any program that is sufficiently modular correct from the bottom up. However, that is neither here nor there, as most large, complex programs would take a year or so and a few billion dollars to prove, and that's simply not economic. Not economic != not possible.
Software is (almost) indistinguishable from pure mathematics (what is doable in one is doable in the other, and what is not doable in one is not doable in the other). The only real distinction is the nature of the I/O. However, other than that, I would agree that software is pretty unique.
I'll offer an alternative. How about "software companies that maintain patents, trade secrets, or any other intellectual property rights over interfaces, algorithms, trivial solutions or data formats are liable for the consequences of holding those intellectual property rights, where it can be reasonably shown that the holding of said intellectual property rights denied the consumer access to information revealing flaws in the product or access to solutions to those flows, when said flaws caused significant damage when the product was used for the purpose for which it was sold, within those limits that a reasonable expert would expect the product to have been tested against."
(Phew, that's long, but if you want something bulletproof, then sometimes you need a LOT of #ifdefs in there.)
Open Source is, almost by definition, continuous development and therefore continuous beta. There is no "finished product". (I suspect that this is part of why Linus dropped the whole development vs. stable cycles concept, as the difference is purely a line in the sand. It has nothing substantial behind it.)
It is arguable that this is the case for all software. So long as corporations are honest about the flaws, so long as they do not make unsubstantiated claims as to the quality of their product, that's fair.
This is why you really want to state clearly the assumptions made on entry to every function (the preconditions) and the consequences that logically follow from applying what you think the function does to the data passed in, given those preconditions (the postconditions).
You then want the automated tool to map out the paths through the code, substituting every variable for an equivalent expression consisting of the inputs that would go into producing that variable, for each path. (This works for loops, so long as you can decompose the condition for the loop.)
Finally, the tool would need to determine one or more ranges of inputs for which at least one variable CANNOT meet the preconditions or postconditions set, in some path, but which are not expressly prohibited from being input at that point.
The tool would then be able to show you the path that is buggy (but not where - the results may be valid but incorrect at ANY point prior to the invalid data showing up). The programmer could then decide if the input should be allowed, or use the information about the path to actually debug the code.
This is not a foolproof method:
It only catches situations where the tool can force an impossible result.
The less decomposable the data is the less accurate it would be. (eg: If you have a random-length loop, but the loop variable is a byte in size, the tool would be forced to assume the loop can range from 0 to 255, even if the random function was much more limited.)
It's only as accurate as your preconditions and postconditions. (If you miss out a condition, no automated tool could ever test for it. If the condition is inaccurate, you'll get false positives and/or false negatives.)
Most programs have, in effect, an infinite number of possible paths, once totally unrolled. You can't test them all in the general case. Well, not unless you're very bored and immortal. This means this method cannot prove a program bug-free, only that it contains a bug within the number of paths checked. (You CAN test any number of paths that have no resultant impact on the state of the program, because once you've proved one cycle correct for each such path, you've proved every possible combination of every possible complexity for all such paths.)
Decomposition of unrolled paths is the ONLY way to test for unexpected side-effects, self-modifying code and other inconsistant systems. Because it is a limited approach, however, such code should really be avoided where possible.
Decomposition without unrolling (a common form of static testing) is useless in any case where there is any kind of non-fatal error that slowly accumulates, as everything is generally going to be valid on a single pass through. Static testing without decomposition (an even more common form of static testing) is unlikely to catch most non-fatal bugs.
Slow poisoning through an accumulation of non-fatal bugs that - in and of themselves - don't stray too far from what's expected are probably the most common.
...I would have considered that a gross overestimate for dramatic purposes on the part of the park expert. Within the past year, ice worms were discovered in the Cascades in Washington State, 13 new species of scorpion were found in California, and they've now found that 2.5 km cave system in Israel with who knows how many species that are totally new to science.
99% of the microbes in Yellowstone Park being undiscovered (or, at least, unidentified) sounds a whole lot more reasonable to me today, and I am completely convinced that enough remains unidentified on Earth that alien origins for any organism would be impossible to prove by any process that eliminates the known. There is simply too much unknown.
The claims of no DNA aren't convincing, either. DNA isn't hard to extract, but not while it is still in a puddle on the ground. Of course, not everything there need be alive - plenty of chemicals cause colours. A few unusual compounds could make for pretty colours but not produce DNA.
(Also, survivable high-speed atmospheric impacts would require any bacteria to be in a meteorite large enough to reach the surface AND large enough to hold all the bacteria being found AND deep enough inside for the fragment to be blasted off a planet(oid) without killing everything in it. I'd be much more ok with this claim if the find was in a sizable, FRESH crater.)
In general, if a product is faulty then lemon laws apply - no matter what aspect of that product is faulty. If a drug turns out to kill those that take it, then the pharmacutical company is liable and the complainants need not establish which specific molecule is at fault, whether the compound responsible was produced in-house or purchased on eBay, etc. If the product is not fit for the purpose for which it was bought and sold, that is complete and sufficient.
How would this work in a software product? I would argue that any liability regulation as applied to software needs to make it very clear that liability is limited to the purpose for which the product is sold. (In other words, there would be a document which can be easily accessed by the consumer which states what the software is known to run on, what the software is known to do, and what the software is known NOT to do.)
Let's say there's a bug in the Linux kernel that prevents it running on processors made of swiss cheese (such as the Itanium). That bug is declared as part of the product. Part of the purpose for which it is now being sold is to NOT run it on processors made of swiss cheese. It would be absurd to hold a company or person liable for selling you a product that does what it says it does.
This means that developers would need to clearly document what they know FOR CERTAIN works, and what they know FOR CERTAIN does not. (IANAL, so how do I know that this is even vaguely plausible? Because people do stupid things with otherwise functional products and yet civilization is still essentially intact.) Clear, quality documentation will not kill Open Source. It stands an excellent chance of improving it, because others will have a clearer idea of what isn't working (yet) and why.
Now, what about all those people who sue for no obvious reason, just because they see a chance of getting some quick cash? Well, the documentation should prevent such people from actually getting said cash, because it is clearly stated what purpose(s) the product is usable for. Absolutely no use outside of those limits would count.
However, legal cases aren't cheap, so you'd probably want something extra in there. I'd suggest something along the lines of "developers are not liable for the consequences of abuse of the product" (just to make things clear) and "whereupon it is shown that the case is frivolous, malicious or criminally stupid, the plaintiff is liable for all legal costs by the defendant, plus damages to their reputation". This should limit the number of cases and might even help fund Open Source developers where court cases result from FUD or attempted robbery by the suit addicts.
(It might even force companies to tone down the anti-Open Source FUD - each case won by the developers would damage the credibility of FUD perpetrators. It would become too expensive to keep believing them.)
It is perfectly true that the "unwashed masses" do not get a "Service Level Agreement" (SLA). However, they DO get a rating for their connection, and (provided the network is neutral) SHOULD be guaranteed "best effort" for packet delivery. What is being described in the article does not sound like "best effort", and the inability to reach the claimed "rated speed" (presumably even at off-peak times) suggests that the actual rating for the line is much lower than that advertised.
Now, there are certain exceptions. In general, you can't drive a dense network at much beyond 1/3 the rated speed - thin-wire ethernet was bad for that - so you can expect similar sorts of problems on a shared line such as cable. The entire design of cable - a single line with taps off it - is exactly what thick-wire and thin-wire ethernet were like.
However, the article mentions DSL. DSL is not a shared line, it is essentially a dedicated line. The service only becomes shared at the teleco's CO (as that's where the DSL modems are, on the other side). At that point, everyone gets plugged into one or more routers. Now, when you change the speed of the modem, they simply program the DSL modem on their end to take a slower connection. They do not (at least, if they are network neutral) mess with the routers to change the priority of your network traffic.
Interestingly, when I worked for a company that got SDSL installed (no service agreement), the engineer ramped up the listed speed beyond what we'd paid for, but the actual speed we ended up with was what we'd bought . This doesn't conflict with what I've just said - we were on the edge of the service area and the speed we were supposed to get simply didn't operate. At all. Apparently, if the copper is poor, not all frequencies are guaranteed to work, and it's not an upper limit - lower speeds can be affected too.
Anyway, to the poster of the original story, I'd strongly suggest getting an INDEPENDENT person that you can trust to check the phone wiring from the DSL modem as far out as practical. At the very least, check the wiring in the house. It is possible that poor wiring, a rusty connector or a loose connection somewhere is killing the speed. If that is the case, then fixing the problem would be very cheap and easy, and would save a LOT of money - you'd have more bandwidth without shelling out the extra cash.
If the wiring is good, then the fault lies with the ISP, and I'd suggest calling a consumer advocacy group for advice on what to do - if, indeed, you can do anything. If only a handful of people care enough to actually do anything, you probably can't - although there are usually multiple DSL providers in an area, and some are better than others.
If a LOT of people are VERY frustrated AND willing to spend hard cash to get this fixed once and for all, you might want to investigate the pros and cons of setting up a DSL cooperative. The teleco can't deny you equal access to the CO (that's law), but industrial-strength network equipment (DSL modems, high-end routers, T3 or T4 line) - that isn't cheap. And, yes, you probably would need to go to a T3 or T4 in order to make the whole thing fast enough to pay for itself. This is NOT a recommended option, without some serious funding behind it. However, if the funding is there, it is the one path you can take that (a) guarantees you the results you want, (b) guarantees the ISP has consequences it WILL notice, and (c) guarantees you the undivided attention of every disenchanted geek and abusive ISP on the planet - at least, for a week or two.
In a system with mandatory access controls, the absolutely worst thing possible is a user who can do anything. Superusers are actually a Bad Idea, from a security standpoint. What you want are specialized users who can do what they need (without interference) but cannot escape that box.
Generally, if I use a root account, it's because I'm doing a mix of admin and other stuff. I could equally well use a user who is specifically configured for what I want. Root's powers go well beyond what I'm likely to use at any given time.
The primary reason people don't like role-based computing is that it's bloody hard to get it configured correctly. A big source of complaints is for servers like Apache, which do a lot of very different tasks and can (potentially) access many different types of service.
But hard is not the same as impossible. It does require effort, though, and a lot of planning in terms of what you are doing. Role-based computing is not designed for sit-down-and-hack operators, it's designed for people who architect their systems and know the interrelationships involved.
This is not to criticise the sit-down-and-hack folk - for a start, I'm usually one of them, and for another, the architects may be great admins but they're generally poor coders outside of the "mission critical" applications (life support, for example). Sit-down-and-hack types get things done, they get things done fast, and they get things done now. Software from such coders is frequently buggy, which is why peer-review is essential. Ultimately, though, hackers (in this correct use of the term) don't need all the powers of root - though usually far more than Windows allots to general users.
The problem with having a set number of specially-designed caricature user-types (which is the Windows model) is that users invariably end up way too restricted or way too powerful. REAL role-based computing is as fine-grained as you like, with the ideal restrictions being solely that you can't do what you wouldn't want to do anyway.
This is not to say hardened Linux, Trusted Irix, or any other system out there, is vastly better. I believe that there's a few thousand times as much room for improvement as there have been improvements since the first time-sharing OS' were developed. However, if we fail to criticise what is blatantly incorrect design, nobody will ever design anything better. If the inferior design is considered "good enough", there will be no motivation to design anything better - and no incentive for users to switch to it.
It is absolutely vital for the health of the industry that critics wrench every last defect that they can find out of a system and hold them to the light. In Open Source, this is part of the normal software lifecycle and is usually done on mailing lists, bugtrackers and pubs across the globe. For closed source, especially with the limitations on discovery placed by assorted US laws, we rely on tech journalists to do this work.
Sure, the journalist in question could have done a better job. They could probably have found twenty times the faults, and compressed the article to half the size in the process. But instead of telling me why they're not really "problems", maybe you should be telling journalists to explain their conclusions better and to dig a little deeper into the subject.
Hell, Windows 2000 reputedly came out with 65,536 known, documented bugs. Where were the investigative journalists, the go-getters, hunting through every last scrap of available information, questioning/bribing Microsoft employees for every last drop of data? If journalists don't have any problems doing this for celebrity movie stars who have done no harm and have no real capacity to - ever, then why not do this in an industry where a crashed computer could cost billions in some cases, or a breeched server could compromise tens of millions of bank accounts or credit cards? In both cases, we're talking nine or ten figure sums. Telephon
There is absolutely no information you can enter that someone else can't.I'm sure most kids (who want to) have ALL their parents' credit card numbers written down somewhere - and many have probably been asked to use them for online ordering.
Surely you saw the bit on saltwater crustaceans.:) The real reason that DNF hasn't been released is that they've not evolved hands to operate the keyboard. The code was finished five and a half million years ago.
Slashdot Mirror
Examples of services that can be configured to be IPv6-reachable would be websites using Apache or Roxen, the games of Empire or PennMUSH - dunno if Freeciv is IPv6-aware or not, the Yum RPM updater - not sure about Apt, BIND, and probably a fair few others I can't think of right now. (They don't specify if games would be acceptable or not - but anyone who hosts a celebration on 6-6-6 is probably not hopelessly straight-laced.)
If you want to set something up, I'd encourage it. All you need do is set up a tunnel with one of the standard free brokerage services, enable IPv6 on your box, and have at it. I'd particularly encourage it if it means there's a snowball's chance in hell (!) of IPv6 being seen as used.
Of course, that last part may explain why I never did well on the statistics coursework.
Now, I understand that testers aren't their primary audience, that building the ISO images & setting up the installers is not always a trivial task, and that the development files change frequently enough that large updates are almost inevitable without an automated ISO build script, so I am not saying that they "should" build such images - merely that it would be very handy for some of us if they did.
So although the name appears to be logical, history would imply that it is not quite as simple. The origins of the game are sufficiently ancient and obscure - although it almost certainly started in the British Isles - that there is no absolute guarantee that the original name was even in English and therefore may be merely phonetically derived from an earlier name. That happens a lot. In that case, the modern name would have no meaning whatsoever.
For temporary slow-downs, certainly major events cause problems, and most of those are indeed caused by streaming. More specifically, unicast streaming. If streaming was predominatly multicast, there would be no meaningful load imposed, no matter how many people had broadband.
It could be a neutral mod - +0 - Car Analogy. In fact, you could have a whole range of descriptive mods. These would be useful when the score is appropriate (so you don't want to give it a +1 or a -1) but the classification just doesn't fit. These could include a +0 on-topic, +0 mostly harmless, etc.
...it's so HIPP!
This is not to say that they will do any of these things. I've neither seen nor heard any evidence that Microsoft is going to "knife the baby" the way they have done with other companies they didn't see eye-to-eye with. My knowledge is limited to the fact that they have "knifed the baby" with other companies and have therefore demonstrated a capacity to think in such terms. I think it reasonable to give Microsoft the benefit of the doubt - to an extent. That extent being that we should keep our eyes and minds open to the possibility of malign intent. However, that's only any good if we know what "malign intent" looks like.
My original posting - and any follow-ups of mine - may well be incorrect on specifics. I'm not perfect (yet, anyways). If it gets people to think about what actually would demonstrate "good intent" or "malign intent", then it has served its purpose, even if I'm wrong on every single method. (Now, if I did believe I was wrong on every single point, I would have written other points - I might invite "vigorous debate" at times, but I try hard not to troll or flamebait. I'm also practicing on admitting when I am wrong, but when things are this speculative, it can be hard to tell.)
Microsoft also lost their case regarding integrating things into Windows 95. The appeals court overturned it on the technicality that Microsoft had renamed the product Windows 98.
Now, they have lost cases in the past and been forced to back things out relatively quickly. Their piracy of disk compression tools in DOS 6.0 forced them to release 6.2 with those tools absent. Assorted tools in Windows 5.0 that carried the copyright of others were also amazingly missing from 6.2. Then there was the Java fiasco, where they "embraced and extended" rather beyond their license to do so. There are probably more recent examples of cases where they've lost and had to remove stuff as a result. So, yes, it can happen and has happened. The question then becomes one of whether Microsoft has learned not to do that, or whether they've merely learned not to be quite so obvious about it. My hope is the former, my fear is the latter.
Are any of these likely? We'll almost certainly have an answer by the time Vista comes out, and quite possibly by the time XP or 2003 have another service pack. We do know for certain that they've never liked being spurned and their competitors have a strange habit of dying terrible deaths.
Are Adobe aware of this? Oh, almost certain. They're almost guaranteed to be just as aware that allies of Microsoft have an equally strange reduction in their life expectancy. In the most recent case, Microsoft worked with anti-virus companies then bought one out and produced their own. Oddly, it seems to recognize some of the competing anti-virus products as hostile and destroy them. Curious, that.
My guess is that Adobe takes the line that if they're going to die, they might as well die with their boots on, and there's a slim chance Microsoft won't dare crush them with all of the other anti-trust cases going on, but that if they go with Microsoft, they don't stand a chance. That's based on what has visibly occurred, however, and Adobe's actual (but not necessarily stated) reasons may be very different.
...that Open Source encryption software is exempt from access controls, provided the US Government is notified of where the source is, and that there are no constraints on binary-only encrypted software. So, if you are using a standard Open Source library for encryption that has been properly registered by its authors, I wouldn't see that there was a problem. If you don't already do so, it might be a good idea to state what encryption software is used - it would make it harder for anyone to reasonably claim it might be proprietary (as proprietary algorithms aren't exempt). A lawyer would be able to give a clearer answer, but bear in mind that 100% of all court cases involve lawyers losing.
How to translate this into the US system? Well, let's start with PBS, as that is the only public broadcasting service the US has. To be on-par, PBS would need to eliminate the sponsorship system, eliminate the adverts, get better backing from Government, broaden the content, improve the quality of productions to modern cinematic standards (documentaries shouldn't have the 1930s Pathe Newsreel feel to them), and carry out independent work (history should not be read from a textbook, and news should not be read from AP bulletins).
Next come the existing openly subscriber channels and pay-per-view. These should, really, be reaping the full cost of everything (plus profit) from the material they sell. If they don't, then the material is either grossly undervalued or grossly inferior. People are generally happy to pay for things that are worth the cost to them, so either the pricing is incorrect or the material is. Or both.
Finally, the "free" advert-laden channels. In the end, adverts cost the producers of the advertised material money. This money will end up being added to the cost of goods. Since the cost of material doesn't depend on who is paying for it, this will work out to be comparable to any of the subscriber channel costs. Only, you're paying for it whether you watch those channels or not! It's a tax on goods, going through the corporations rather than through Government, but it's still a tax. Since it is a tax, why not have it collected by the people collecting taxes anyway? It won't change how much you end up paying for your cost of living, but it will add about 15-20 minutes of material to every show, increasing the value of watching them.
(If you're going to pay $X extra because of an invisible sales tax created by advertising, it makes no difference to you in your overall costs if - instead - those same goods were $X cheaper and you had an $X flat tax to cover broadcasting in that area. $X - $X = 0.)
Actually, that's not totally correct. Those in adverts get paid royalties for every time the advert is shown. This costs the advertisers more, which they'll defray by making you pay for it by more expensive goods. This will not be exactly the same as the increase in production costs by making shows 15 minutes longer. In some cases, the cost of the adverts will be more. In other cases, the cost of the shows will be more. You would need to quantify this, to prove conclusively that the BBC model of the license fee would work in these cases. My suspicion is that you'll find that the license fee is indeed the superior model, but in either case, the difference has to be insignificant as none of the other costs for those channels is going to vary.
(You asked for an alternative model. You didn't ask for one Americans would stomach. I know perfectly well that even if every household in America saved hundreds or even thousands of dollars a year from a license model, and even if it meant program quality skyrocketted far beyond the wildest imaginings of anyone alive today, you'd be risking an armed uprising before Americans would consider a new overt tax from Government, no matter how covertly they were being taxed by corporations already.)
eg: NASA is currently cutting back or eliminating many science missions in order to pay for the next-gen shuttle, which assumes Congress won't cut NASA's budget over the next ten years -and- there are no cost overruns anywhere, according to the New Scientist.
eg: The US has spent a miniscule fraction of what it pledged and committed in the fight against bird flu, according to the World Bank. Whether an epidemic ever occurs is irrelevant in this. What matters is that even hard-nosed financial institutions are getting concerned. When the economists think Government is spending too little, it's time to be worried.
(I'm not singling out the US because it's particularly bad amongst nations - it actually does better than most - but because that's what the contest is about. Had this been an international contest, I'm sure I could find alarming attitudes in every civilized nation on the planet. It wasn't that long ago that the South African Health Minister promoted garlic as a cure for AIDS. Although I suppose there might be a lot of vampires in South African politics.)
I just don't know how this project can possibly reach its true potential with such limited backing. Most who could enter a truly biting cartoon won't be bothered, because there won't be any perceived value. If getting into the final rounds constituted a publication in a peer-reviewed forum, then perhaps there would be more interest. Money from pro-science organizations towards prizes would have been good, too.
For those on Slashdot with no artistic talents - enter anyway. Most scientists can't draw worth a damn, so it'll be purely down to the ideas in the cartoon anyway. Besides, there are valid reasons for believing the readers here have a broader understanding of the state of science and the attitudes around it - those focused totally on their subject won't have time to read up on anything much outside of their speciality and so won't be able to so easily draw on attitudes and perceptions that are universal.
In Soviet Russia, original orders are worked in you!
Yes, you do need all that. A compromised encryption algorithm allows a person to reduce the effective key length and thereby deduce the actual encryption key used comparitively quickly. You've got to assume classical computers - with one exception - as quantum computers should be able to go through all the possible keys exponentially faster. You've also got to assume a price bracket, because an infinite amount of money would allow you to do just about anything.
The one exception is for One Time Pads. These decrypt into EVERY possible string of equal length to the encrypted string, so it is totally impossible to tell which solution is the right one. One Time Pads are unbreakable without the key. Totally. It isn't even a question of nobody having a means of breaking them, it can't be done. The problem is that the key has to be as long as the data you are trying to encrypt. It trades one problem for another of equal size. So although they are theoretically perfect, they are also practically useless in most cases. Oh, and the "one time" bit is for real - the key, to be secure, can only be used once. You have to make a new key each time, with each key being genuinely random - pseudo-random isn't random enough.
I'd also argue that although formal proofs are horribly complex and a nightmare for any significant code, they are not as impossible as is often claimed. Code doesn't have to be proven linearly. If you can prove a function correct (assuming that all calls function correctly, and the initial conditions are valid), then you can prove any program that is sufficiently modular correct from the bottom up. However, that is neither here nor there, as most large, complex programs would take a year or so and a few billion dollars to prove, and that's simply not economic. Not economic != not possible.
Software is (almost) indistinguishable from pure mathematics (what is doable in one is doable in the other, and what is not doable in one is not doable in the other). The only real distinction is the nature of the I/O. However, other than that, I would agree that software is pretty unique.
I'll offer an alternative. How about "software companies that maintain patents, trade secrets, or any other intellectual property rights over interfaces, algorithms, trivial solutions or data formats are liable for the consequences of holding those intellectual property rights, where it can be reasonably shown that the holding of said intellectual property rights denied the consumer access to information revealing flaws in the product or access to solutions to those flows, when said flaws caused significant damage when the product was used for the purpose for which it was sold, within those limits that a reasonable expert would expect the product to have been tested against."
(Phew, that's long, but if you want something bulletproof, then sometimes you need a LOT of #ifdefs in there.)
It is arguable that this is the case for all software. So long as corporations are honest about the flaws, so long as they do not make unsubstantiated claims as to the quality of their product, that's fair.
You then want the automated tool to map out the paths through the code, substituting every variable for an equivalent expression consisting of the inputs that would go into producing that variable, for each path. (This works for loops, so long as you can decompose the condition for the loop.)
Finally, the tool would need to determine one or more ranges of inputs for which at least one variable CANNOT meet the preconditions or postconditions set, in some path, but which are not expressly prohibited from being input at that point.
The tool would then be able to show you the path that is buggy (but not where - the results may be valid but incorrect at ANY point prior to the invalid data showing up). The programmer could then decide if the input should be allowed, or use the information about the path to actually debug the code.
This is not a foolproof method:
Decomposition of unrolled paths is the ONLY way to test for unexpected side-effects, self-modifying code and other inconsistant systems. Because it is a limited approach, however, such code should really be avoided where possible.
Decomposition without unrolling (a common form of static testing) is useless in any case where there is any kind of non-fatal error that slowly accumulates, as everything is generally going to be valid on a single pass through. Static testing without decomposition (an even more common form of static testing) is unlikely to catch most non-fatal bugs.
Slow poisoning through an accumulation of non-fatal bugs that - in and of themselves - don't stray too far from what's expected are probably the most common.
99% of the microbes in Yellowstone Park being undiscovered (or, at least, unidentified) sounds a whole lot more reasonable to me today, and I am completely convinced that enough remains unidentified on Earth that alien origins for any organism would be impossible to prove by any process that eliminates the known. There is simply too much unknown.
The claims of no DNA aren't convincing, either. DNA isn't hard to extract, but not while it is still in a puddle on the ground. Of course, not everything there need be alive - plenty of chemicals cause colours. A few unusual compounds could make for pretty colours but not produce DNA.
(Also, survivable high-speed atmospheric impacts would require any bacteria to be in a meteorite large enough to reach the surface AND large enough to hold all the bacteria being found AND deep enough inside for the fragment to be blasted off a planet(oid) without killing everything in it. I'd be much more ok with this claim if the find was in a sizable, FRESH crater.)
How would this work in a software product? I would argue that any liability regulation as applied to software needs to make it very clear that liability is limited to the purpose for which the product is sold. (In other words, there would be a document which can be easily accessed by the consumer which states what the software is known to run on, what the software is known to do, and what the software is known NOT to do.)
Let's say there's a bug in the Linux kernel that prevents it running on processors made of swiss cheese (such as the Itanium). That bug is declared as part of the product. Part of the purpose for which it is now being sold is to NOT run it on processors made of swiss cheese. It would be absurd to hold a company or person liable for selling you a product that does what it says it does.
This means that developers would need to clearly document what they know FOR CERTAIN works, and what they know FOR CERTAIN does not. (IANAL, so how do I know that this is even vaguely plausible? Because people do stupid things with otherwise functional products and yet civilization is still essentially intact.) Clear, quality documentation will not kill Open Source. It stands an excellent chance of improving it, because others will have a clearer idea of what isn't working (yet) and why.
Now, what about all those people who sue for no obvious reason, just because they see a chance of getting some quick cash? Well, the documentation should prevent such people from actually getting said cash, because it is clearly stated what purpose(s) the product is usable for. Absolutely no use outside of those limits would count.
However, legal cases aren't cheap, so you'd probably want something extra in there. I'd suggest something along the lines of "developers are not liable for the consequences of abuse of the product" (just to make things clear) and "whereupon it is shown that the case is frivolous, malicious or criminally stupid, the plaintiff is liable for all legal costs by the defendant, plus damages to their reputation". This should limit the number of cases and might even help fund Open Source developers where court cases result from FUD or attempted robbery by the suit addicts.
(It might even force companies to tone down the anti-Open Source FUD - each case won by the developers would damage the credibility of FUD perpetrators. It would become too expensive to keep believing them.)
Now, there are certain exceptions. In general, you can't drive a dense network at much beyond 1/3 the rated speed - thin-wire ethernet was bad for that - so you can expect similar sorts of problems on a shared line such as cable. The entire design of cable - a single line with taps off it - is exactly what thick-wire and thin-wire ethernet were like.
However, the article mentions DSL. DSL is not a shared line, it is essentially a dedicated line. The service only becomes shared at the teleco's CO (as that's where the DSL modems are, on the other side). At that point, everyone gets plugged into one or more routers. Now, when you change the speed of the modem, they simply program the DSL modem on their end to take a slower connection. They do not (at least, if they are network neutral) mess with the routers to change the priority of your network traffic.
Interestingly, when I worked for a company that got SDSL installed (no service agreement), the engineer ramped up the listed speed beyond what we'd paid for, but the actual speed we ended up with was what we'd bought . This doesn't conflict with what I've just said - we were on the edge of the service area and the speed we were supposed to get simply didn't operate. At all. Apparently, if the copper is poor, not all frequencies are guaranteed to work, and it's not an upper limit - lower speeds can be affected too.
Anyway, to the poster of the original story, I'd strongly suggest getting an INDEPENDENT person that you can trust to check the phone wiring from the DSL modem as far out as practical. At the very least, check the wiring in the house. It is possible that poor wiring, a rusty connector or a loose connection somewhere is killing the speed. If that is the case, then fixing the problem would be very cheap and easy, and would save a LOT of money - you'd have more bandwidth without shelling out the extra cash.
If the wiring is good, then the fault lies with the ISP, and I'd suggest calling a consumer advocacy group for advice on what to do - if, indeed, you can do anything. If only a handful of people care enough to actually do anything, you probably can't - although there are usually multiple DSL providers in an area, and some are better than others.
If a LOT of people are VERY frustrated AND willing to spend hard cash to get this fixed once and for all, you might want to investigate the pros and cons of setting up a DSL cooperative. The teleco can't deny you equal access to the CO (that's law), but industrial-strength network equipment (DSL modems, high-end routers, T3 or T4 line) - that isn't cheap. And, yes, you probably would need to go to a T3 or T4 in order to make the whole thing fast enough to pay for itself. This is NOT a recommended option, without some serious funding behind it. However, if the funding is there, it is the one path you can take that (a) guarantees you the results you want, (b) guarantees the ISP has consequences it WILL notice, and (c) guarantees you the undivided attention of every disenchanted geek and abusive ISP on the planet - at least, for a week or two.
Generally, if I use a root account, it's because I'm doing a mix of admin and other stuff. I could equally well use a user who is specifically configured for what I want. Root's powers go well beyond what I'm likely to use at any given time.
The primary reason people don't like role-based computing is that it's bloody hard to get it configured correctly. A big source of complaints is for servers like Apache, which do a lot of very different tasks and can (potentially) access many different types of service.
But hard is not the same as impossible. It does require effort, though, and a lot of planning in terms of what you are doing. Role-based computing is not designed for sit-down-and-hack operators, it's designed for people who architect their systems and know the interrelationships involved.
This is not to criticise the sit-down-and-hack folk - for a start, I'm usually one of them, and for another, the architects may be great admins but they're generally poor coders outside of the "mission critical" applications (life support, for example). Sit-down-and-hack types get things done, they get things done fast, and they get things done now. Software from such coders is frequently buggy, which is why peer-review is essential. Ultimately, though, hackers (in this correct use of the term) don't need all the powers of root - though usually far more than Windows allots to general users.
The problem with having a set number of specially-designed caricature user-types (which is the Windows model) is that users invariably end up way too restricted or way too powerful. REAL role-based computing is as fine-grained as you like, with the ideal restrictions being solely that you can't do what you wouldn't want to do anyway.
This is not to say hardened Linux, Trusted Irix, or any other system out there, is vastly better. I believe that there's a few thousand times as much room for improvement as there have been improvements since the first time-sharing OS' were developed. However, if we fail to criticise what is blatantly incorrect design, nobody will ever design anything better. If the inferior design is considered "good enough", there will be no motivation to design anything better - and no incentive for users to switch to it.
It is absolutely vital for the health of the industry that critics wrench every last defect that they can find out of a system and hold them to the light. In Open Source, this is part of the normal software lifecycle and is usually done on mailing lists, bugtrackers and pubs across the globe. For closed source, especially with the limitations on discovery placed by assorted US laws, we rely on tech journalists to do this work.
Sure, the journalist in question could have done a better job. They could probably have found twenty times the faults, and compressed the article to half the size in the process. But instead of telling me why they're not really "problems", maybe you should be telling journalists to explain their conclusions better and to dig a little deeper into the subject.
Hell, Windows 2000 reputedly came out with 65,536 known, documented bugs. Where were the investigative journalists, the go-getters, hunting through every last scrap of available information, questioning/bribing Microsoft employees for every last drop of data? If journalists don't have any problems doing this for celebrity movie stars who have done no harm and have no real capacity to - ever, then why not do this in an industry where a crashed computer could cost billions in some cases, or a breeched server could compromise tens of millions of bank accounts or credit cards? In both cases, we're talking nine or ten figure sums. Telephon
There is absolutely no information you can enter that someone else can't.I'm sure most kids (who want to) have ALL their parents' credit card numbers written down somewhere - and many have probably been asked to use them for online ordering.
Surely you saw the bit on saltwater crustaceans. :) The real reason that DNF hasn't been released is that they've not evolved hands to operate the keyboard. The code was finished five and a half million years ago.
So THAT is where Eric Griffiths, Rod Davis and Colin Hanton vanished to, after leaving the band!