Uhh, no the patent office doesn't give out patents for realizing that X is profitable. It gives patents for developing technology X.
Under your argument Apple should get a patent for a windowing OS despite X/Xerox parc doing it first as they were the first ones to realize it might be profitable on PCs. Hell redhat should be able to get a software patent on Linux because they realized it was profitable to sell.
The patent system simply isn't supposed to work that way. It is only supposed to give patents that aren't obvious to those skilled in the art. Meaning those video geeks working at home are the people who count not the public at large.
If TiVO wants to get a buisness method patent on selling this technology that is another matter but realizing it is profitable simply doesn't cut it for a patent on the technology.
Ohh, of course I'm not arguing that these other patents should be upheld but not TiVO's. If we are going to go down the route of upholding these obvious patents then we should do the same with TiVO as well.
However, I think it is inconsistant to cheer for TiVO winning their case but jeer at NTP or Amazon for their obvious patents. If we are going to bitch about people getting patents on obvious things we can't do it only when they are companies we don't like.
Fine then give them a patent on making commodity hardware circa year such and such do this. If the non-obvious part of their design was just coding well enough to make it work with current hardware then they can have that patent but it would be totally worthless.
This is like arguing that the first person to do multi-tasking on a PC deserves a patent for realizing that multi-tasking wasn't only something that supercomputers can do. Admitedly at that time it probably wasn't obvious that a PC had enough power to do this but that is hardly justification for saying that multi-tasking on a PC isn't obvious for the purposes of patenting.
I think one of the reasons they didn't include such functionality is pressure from broadcasters/advertisers. The ability to record and watch simultaneously almost inevitably entails the ability to skip commercials.
The fact that it wasn't yet on the market hardly shows that it wasn't obvious. It may not have been obvious that this feature would make a profit/be popular but the patent office doesn't grant patents on realizing that something would make money. The patent office grants patents based on the technology/idea being non-obvious (though buisness method patents do make this a bit fuzzy but this isn't relevant here).
The technology to watch and record at the same time is really obvious. Even the idea that this would be nice is pretty obvious. What do you want to bet that if we went back to usenet archives before TiVO we could find someone wishing he could time delay with his VCR or suggesting that this be done with a HD? TiVO just realized that this would be a sucesful product.
On the one hand I'm happy to see a company that created a truly new product be rewarded.
On the other hand I'm not sure that what TiVO did wasn't obvious. Using a HD rather than a videotape was surely obvious and automating the process of recording shows is not only obvious but far from original. All TiVO did is put both of these things together in a convient package and I find it hard to formulate any principled rule that would call TiVO's developments non-obvious but NPT's blackberry patent obvious.
The difference seems only to be that geeks like TiVO and TiVO hasn't being suing individuals who decide to set up their computers to tape shows for them. But if they do get their patent enforced that is exactly what they *could* do in the future.
In the end I tend to think the TiVO patent should have been rejected as obvious. However, I think the only reason TiVO didn't make money is the monopoly cable companies and satellite companies have on their markets. When data service becomes the commidity everyone buys and there is free competition amount content providers on an open protocal companies like TiVO won't be shut out by monopolists who can make it difficult for TiVO to penetrate their markets.
It is far from clear for me. It is obvious that the EURid rules were really horrible things but if they were no more specific than reprinted on these blogs these companies may be perfectly within the rules.
After all if I am some domain registrar the rules don't seem to prevent me from creating 500 new companies, each of which has me as it's only customer, and use them to buy up names. After all each of these companies is genuienly intending to become a registrar, just of only a few names and is trivially giving all it's customers access to these domains since it only has one customer.
Maybe the formal rules were more specific but as it is it is far from clear these companies were violating any rules.
By carrier I meant cell phone carrier. Unlike the other responder I wasn't imagining a situation where you totally avoided phone companies all together. In fact such a situation hardly seems possible as someone is going to be carrying your packets and they will know you are making a call whether you are doing so using an internet phone or a regular phone (you don't really think the telcos aren't tracking the amount of Skype traffic on their networks do you?)
I just meant a situation where sprint can't anally rape me for minutes I use over my wi-fi connection and I fear they will figure out a way to do this up until the point I can have a wi-fi conversation totally without their knowledge. I don't care if my DSL company knows about it, at least unless they start trying to charge me more for voice packets than other packets of similar QOS.
This was a very interesting article but I'm afraid I disagree with the fundamental assertion, that software patents are (in principle not as applied) no different than any other type of patent.
The first way that writing software differs from discovering a new kind of lithography or building a widget is that the underlying building blocks of software are completely understood. While occasionally we have theories that tell us what will happen if we try using such and such materials in lithography frequently in the physical world discoveries arise when we discover materials that behave in ways our theory couldn't easily predict. However, given executable (not even machine code) for a new piece of software and a processor reference manual what that software will do is completely predictable. Sure it isn't the case that every new physical discovery wasn't predictable like this but the fact that some are means the cost-benefit analysis for software patents is going to be less compelling than it is for physical patents.
The second huge difference is level of application. Software patents are things that are of immediate and direct use in real world applications. Even genuienly novel software patents are processes that almost certainly would have be developed without the incentive of software patents. For physical patents often types of fundamental materials research wouldn't be worth doing if you didn't know you could profit from the fruits of your labours if someone else comes along and figures out how to make something useful with it.
Thirdly the sheer ease of creating software makes software patents a worse deal. Creating a monopoly on some high-tech industrial process causes a few big companies to pay royalties and the cost of dealing with patent issues (excluding royalties paid) is fairly small. On the other hand every nerd with a computer can write software so the amount of innovation restricted by a software patent has the potential to be very large and the proliferation of many many companies and investors makes just traking the existance of patents in your field a very expensive endeavor.
In fact the author of the piece himself makes a very strong case that software patents are different when he argues that they are pretty much irrelevant to innovation in software even though they can be critical to innovation in other fields. Merely noting that we could scrap software patents without harming innovation is a *very* compelling argument to treat software patents differently than other patents. After all the only reason to have patents is to encourage innovation.
Ultimately it seems that the only things in software that are truly novel enough and reusable enough to warrant patent protection are interesting new algorithms. Perhaps something like shell sort might warrant patent protection but this kind of research seems to be occuring mostly at universities and has little need to be encouraged by patents. Besides once you start letting this sort of thing get patented there is no line you can draw to prevent math in general from being patented and as a mathematician this might make me rich there was a very good reason the creators of patent systems didn't want the sieve of erasthonthes (sp) to be patentable.
I really liked how the people interviewed in the article kept saying something like "consumers might expect calling over wi-fi would be free". As if they were somehow being unreasonable or uninformed.
As far as I'm concerned real wi-fi phones which don't even let your carrier know how many wi-fi minutes you are using can't come soon enough. I hate the high prices, ridiculous options and general blood sucking (prices for ringtones) and can't wait till they are the ones begging the technology companies to include support for their off wi-fi network you use when you leave the city or have at least started offering wi-fi type service in cities.
Ultimately of course the upshot of all of this is that we will be paying more for DSL/landline phones as well as for remote cell phone service. In both the landline phone market and the cell phone market massive fixed costs are amortized over a huge number of phone calls. The fixed line phone calls then in effect subsidize DSL service (the phone companies make money on it but wouldn't if they had to do all the maintence/set up the phone lines just for DSL). Similarly all the cellphone calls made in big cities subsidize building cell phone towers in more rural locations. As the distinction between different sorts of data transmission inevitably disappears the price per unit of reasonably low latency Kb must equalize. I mean it really is absurd that it is cheaper to use your phone line for DSL and utilize Skype than it is to call on a real phone. This will force the price of DSL up as it becomes less subsidized by phone calls and the existance of Wi-Fi phones will remove the ability of the cell companies to subsidize the less used more rural towers (unless of course they are just doing things in a really inefficent fashion compared to google/earthlink in SF)
At least this is what happens if the DSL prices aren't constrained by local laws, in which case we will just see more tricks trying to offer tiered access charging for cell phone use (instead of by Mb) or other stupid money generating tricks.
I certainly don't have aspergers (tho some people seem to act as if all math people do) but I couldn't tell you either of those things.
In fact I suspect most people couldn't pick out the features in virtue of which they new someone was bored much better than someone with aspergers. I suspect they would do better because they can immediatly tell whether someone is bored or not but for both cases it would seem to be something you need to learn explicitly.
It seems that much of our processing of faces and bodily expressions occurs entierly unconciously. People *look bored* to us and if we step back we can break this down into vacant eyed, glancing away, fidgety etc.. but phenomenologically they are presented as being bored to us not as having certain facial features from which we can then infer boredom (usually).
I think they have even worked out the parts of the brain responsible for this sort of processing but I could just be high.
I hate professors like this who think they know better than you how you learn best. Hell even if you are wrong it's your fucking education.
Besides, even if this helps students learn that class better it makes sure all the students we graduate for college expect to be babied and have their study habits dictated to them. What happens when they get out in the real world and they don't have people telling them to come to class and to put away their computers and etc..
Most of all though this pisses me off so much because it is very similar to professors I had in college who required class attendance or decided that it would be better to present the material in this non-standard way (which was just a bit better) so you couldn't learn out of a book.
NEWS FLASH different people learn differently. I can't learn jack squat from lecture in technical courses, i.e., math. When I have to attend lecture I just waste my time and make sure I hate the class. Yet because I went to a school where *most* professors didn't care what you did as long as you got your work done I went to virtually no classes through all of college (made it entire terms without seeing a professor) and I graduated caltech and am now a math grad student at berkeley.
When I teach I make a point of only imposing discipline as necessery to stop students from interfering with other student's learning. If a student wants to stare at the ceiling or even read a novel that's fine with me. I will advise them to pay attention but it is their buisness and they are either good enough or learn differently that they will do well on their exams or they will learn the lesson and now understand how to handle their own study habits.
While it is certainly an interesting statistic I'm troubled by the author's statement which says something to the effect of "if anything this information should be used by parents to make an educated deciscion whether or not their child should play." He might claim neutrality all he wants but implicitly this suggests that it is reasonable that a parent try to keep their child away from swear words. A position I happen to disagree with.
I mean imagine if someone said, "I gathered data on the frequency of black people at school X. But I'm not suggesting black people should be kept out. If anything parents should use this information to determine if they want their kids to be in this enviornment or be home schooled." This would clearly be a way of saying, "I think it is reasonable to try and keep your kids away from black people." Similarly this comment here says it is reasonable to keep your kids away from swears.
I was being an idiot. He doesn't actually seem to be setting up a public wiki to host these kind of calculations but creating a spreadsheet sharing system for inside a company. Still usefull but much more boring.
Hopefully someone will come along and do what I suggested anyway.
I'm not entierly sure how many buisnesses will be willing to trust such a system. One needs to be carefull that your competitor hasn't inserted some special logic to screw things up when *your* data is entered into the system. Unlike wikipedia where everyone can see any devious attempts to change things code or calculations can seem fine and work fine on most data but contain subtle bugs.
However, for plenty of individuals and personal buisnesses this seems like a great idea. All the time there are common calculational tasks we do that are programed over and over again. Super simple examples are dates being turned into days of the week or dollar amounts in various years being turned into 1970 (or some other amount) dollars. Some of these are tought because they require the right data (rate of inflation for each year) others are tough because they just involve lots of special cases. In any case having a wiki where people can use other people's work and build off of it seems like a great idea.
In particular it seems like a code repository where you can just run the code right there. This is the biggest problem with code repositories now, often downloading the already written code and figuring out how the API works just takes too much time and since there is no guarantee that someone will download a whole code repository people tend to reinvent the wheel many many times.
It would be great if ultimately this could support more than just a spreadsheet interface. A general wiki for *online* DHTML calculation tools, especially if it also had functionality to let you download the code or acess it as a web serice, would just be awesome. Maybe this already exists somewhere and I don't know about it but the idea seems great.
Where ohh where will the magical internet take us next?
I think most of us dislike HDCP and hope it fails. Unfortunatly most consumers don't care enough to vote with their feet, or are too afraid everyone else is going to defect so they buy HDCP enabled systems too. Situation seems hopeless doesn't it?
Not so fast. What if we could get the major video card manufacturers to make sure everyone thinks this new technology works on their cards when it really doesn't. When HDCP content comes down the line these people will have a horrible experience when they buy the content and it doesn't work and are going to be loath to shell out more money for another 'HDCP enabled' video card. Then the public might dismiss HDCP video as a buggy and annoying product.
Of course it sounds like a horrible PR move for any video card company to make, even counting the extra sales it gets when people upgrade, but we (at least those of us who didn't buy one of the crippled cards) can be happy about their screw-up.
As a side note does anyone know what the anti-trust law is with regard to this sort of content protection scheme. In particular in situations like this which have clauses excluding open source video players, and given that there are certainly companies that do package open source video players as part of their buisness model (google now too I think) why isn't the consortium that defines these content protection standards in violation of the anti-trust laws.
This topic seems to just make people to lazy to think. It's super easy to say censorship bad, google is bowing to chinese censorship hence google bad. However, If anyone bothered to take 5 minutes to think about what actions would be most likely to improve the rights of the chinese people they would realize that this argument is just stupid. Now perhaps there is some other argument critisizing google for doing this but these all get into complex social and political questions that are disagreed on by reasonable knowledgeable people. At worst one can critisize google for making a bad judgement not for being nypocritical or evil (trying to do good and fucking up isn't evil).
The position of Yahoo and MS is a bit more doubtfull in this. There is a big difference between being the 1st big company to do this and the 3rd. Besides evidence has come to light suggesting they are helping repress the chinese people more than they have to.
But to accuse google of hypocripsy in this case makes about as much sense as accusing a billionaire who pays some terrorist a ransom to get a family member/employee/whoever of hypocripsy because he said he didn't give money to terrorist. The billionaire's statement that he doesn't give money to terrorists should obviously be understood as saying he doesn't approve of the terrorists actions and will not aid them in killing/bombing by donating money. Of course there is a genuine question about whether paying them off in this case to *STOP* them from killing someone is worth the extra money/moral boost it gives the terrorists but this is a tough and unclear point.
Google is similarly in a hostage situation and it would be just as wrong to accuse google of selling out. What is being held hostage though is not google's profits (though those do hang in the balance too) but the chinese people's access to information and the internet. IF GOOGLE DOESN'T CENSHOR GOOGLE.CN THE CHINESE GOVERNMENT CENSORS THE RESULTS ANYWAY. It just isn't the case that if google took some principled stand that the chinese people would have any more access to the internet. In fact google seems to have done a remarkably poor job of their censoring (the misspelling issue is well documented) so they probably give the chinese more access to info than they did if they had refused. Quite simply the choice is between a censored version of google being availible to the chinese and a slow, buggy, frequently inaccessable censored version of google being availible to the chinese. Also I suspect by doing self-censorship rather than having chinese filters do it the censorship system will move much slower but I'm not sure.
Yes, had google been the first company to do this there would have been concerns. Perhaps china might have relaxed its censorship policy if western companies hadn't been willing to go along (although how do we know this isn't what happened? By making a comprimise these companies might have gotten china to loosen their policies...though I wouldn't count on it) or the chinese would have been more outraged if the censorship resulted in that much inconveince. However google was not the first company. MS and Yahoo were already there and they provide acceptable search results. Any damage was already done when these companies made their deciscion, google could either refuse to go along and just let censored MS and Yahoo cites take all the traffic or jump in themselves.
SO LONG AS GOOGLE IS LESS REPRESIVE THAN MS AND YAHOO THE CHINESE USERS HAVE MORE FREEDOMS WITH GOOGLE.CN THAN THEY HAD BEFORE.
Given the records of MS and Yahoo I suspect this is probably the case. Even if all the western companies had refused it is unclear whether this would have been good or bad. After all if china developed its own native search company the government would have far more power over that entity and fewer ties with the west would mean less long term pressure to change.
It is particularly rich to see congress critisize google and these other companies whe
This is why programs are released as BETAs!! That way people who want to take risks, live on the edge and know enough that they can repair their systems can try the software out and report any problems. If you are installing a beta it's your responsibility to go do a search and see if there are known bugs that will screw with your computer. Sure alot of us get lazy and don't do this, I know I often don't, but if it them screws up your system in a reasonable way (e.g. they aren't knowingly distributing software that wipes every users HD but just have a bug in their code) you only have yourself to blame.
I certainly have almost no sympathy for any buisness who was installing beta software (for something like this..that wasn't really needed) on anything but a test platform.
Unless I'm just missing an important part of the story, like MS suggested this beta product was safe to run on your buisness computers, this seems to be just the standard stuff one should expect from a BETA. Sometimes they get things wrong.
It is far from clear that this would make the application more robust and not less robust instead. All the magical features you get from managed code stem from the ability of the code to do automatic type inference and other types of validity checking. If the objects you are manipulating pass through an unmanaged section you can't automatically infer anything about them (except perhaps that they occupy x number of bytes and I don't know if you can even do this). Essentially you need to treat the objects from your C code as untrusted input and this is going to require a great deal of code and hence an increased likelihood of bugs offsetting the benefits of managed code.
Worse just the fact of having the interface between the two languages can make garbage collection and other issues a nightmare. Who has the responsibility to dispose of objects? Can the managed code garbage collect objects it inherits from C++? If not all the sudden you need to do garbage collecting for some objects in your nice garbage collected language and when these start getting mixed with other objects things can get messy...there is a lot of chance for error with the wrong objects not getting GCed or incorrectly getting GCed.
Also as the other poster pointed out whether of not managed code is very helpfull depends a lot on what kind of stability you need. If what is important is that the application just doesn't crash managed code is wonderfull. However, if you need to guarantee that it gets the right answer/doesn't screw up the overall computation (as say one might in some type of simulation) it is less beneficial. This is especially true when it is taking in data from unmanaged code. Essentially this just means you catch an error saying 'sorry object in wrong format' or 'unexpected value' and if there is no way to recover from this and continue executing you are screwed.
Besides if you are writing something like a scientific application where it is very important that the computational core doesn't crash but the GUI/data input type stuff is relatively unimportant if it can be restarted you might as well just make the boundaries between C++ and managed code into boundaries between processes. Since you have to do something like an API anyway to interface your languages you might as well just do it as seperate processes so you don't need to worry about crashing at all.
I tried to say that I agreed with much the EFF does at the start of my post. I meant to include in this the lawsuit against AT&T and similar measures to reduce government power/access. I expect the EFF is doing similarly good things in lobbying for changes to the buisness records exception to the search warrant requirement (what causes the increased government access with this product). Though I think they could do better in this area by working with google.
I guess I should have been more clear that I meant only to critisize the way they address the narrow issue of privacy as it involves giving companies or otherwise allowing other private individuals access to your personal information. While the things I mentioned in the first paragraph can quite rightly be characterized as issues of privacy (hell so can abortion) I tend to classify those actions more as civil liberties. Since the major concern in these areas is often the ability of the government to control you through *legal* use of this information or other harrasment I think it is reasonable to distingush this issue from their policy about our free interaction with various companies.
Or to put it really simply and leave my confusing use of terminology behind what I mean is this. I do think the EFF is doing a great job when it comes to lobbying the government for better protections against government intrusion. However, it is their positions on volountary data exchange/record keeping by the private sector that I view as misguided. In fact for the reasons I outlined above I think their attitude on these sort of issues is interfering with their ability to restrict government snooping. Of course people need to be warned that their data is less protected with google desktop but it makes no more sense for the EFF to tell me I shouldn't use google desktop than it would for them to tell me I shouldn't publish my diary online since the government can now read it. The EFF could better (though they are doing alot now as you point out) by cooperating with google and getting them to warn users about the additional risk they are taking and then working together to increase the protection from subpeona.
So don't anyone stop giving money to the EFF because of this argument. If I wasn't a poor grad student I would still be giving money. However, I think they could do even better.
Ultimately my difference with the EFF is about whether it is worth fighting for privacy in the sense of other people not being able to find out about your life. When we lived in small communities our privacy in this sense was minimal because neighbors could easily share and sift through the many clues about our lives we drop all the time. For awhile after most of the population moved to cities the number of people and thus the shear amount of data overwhelmed our capacity to process the information giving us a certain sort of anonymity (not true anonymity but just a high probability that the people who saw you every week on the other side of town wouldn't communicate this to anyone who could put 2 and 2 together and infer you were having an affair). It is inevitable that the increase in processing power will return us to something much more like the small town situation...but unless we are carefull with the added danger of identity theft.
If you really believe that this sort of privacy, i.e., people not being able to figure out what you were up to/what you liked to read etc, is valuable and possible to save the EFF's position makes perfect sense. This sort of privacy is a collective good (if everyone else's lives are out there on the web to see someone can fill in the gaps and figure out your life) and their urge for people not to use google makes sense. On the other hand if you believe as I do that it is a lost cause the EFFs attempts to cling to this type of privacy are hindering their ability to protect the important civil rights which have been implicitly aided by this kind of privacy but may recede unless we do something. Sure it's true that if
It doesn't matter if google will give their all. All that matters is that they are going to do more than the next guy. If people's stuff is going to end up on servers anyway we want to pick the company who will do the most *even if it isn't that much* to go first.
The idea that you can defend your privacy on your own just doesn't work. If everyone you correspond with puts stuff on servers you lose privacy anyway. Even if my suggestion is a slim chance it is the only chance we have.
Also hiring a lobbyist, no matter what position they support isn't the sort of thing that gets you made into road kill. You might not ultimately win but congressmen aren't going to get vindictive because you took them out to dinner. I think the best hope in this area is lobbying lawmakers not the courts (though we should pursue both avenues).
Sorry, I made a mistake. Someone who had posted before had made the remarks I had meant to be critisizing and I stupidly just skimmed your post and thought it was the same.
Well I guess the egg is on my face. I'd mod myself down if I could (hmm..I wonder if that would be a good feature)
I'm getting more and more annoyed at the privacy strategies of organizations like the EFF. I'm generally a big fan of them, I think I've even donated money (as a grad student that's saying alot), but like most online privacy activists they have a very short shighted view of privacy.
The attitude of many of these privacy activisits is somehow that we can draw a line in the sand and refuse to give out any more information about ourselves to anyone. Not letting google have our information is just another example but traffic cameras, search histories, purchase histories are others. I'm not going to argue about whether this would be worthwhile if we could do it here but it simply isn't achievable. At least for the near future people are going to continue to give more and more of their information to companies. Giving advice like 'don't use google desktop' and stuff like this just marginalizes the EFF in terms of privacy. If they took a more pragmatic approach they could do a great deal more good.
In particular in this case the EFF should recognize that SOONER OR LATER PEOPLE ARE GOING TO PUT PERSONAL INFORMATION ON SERVERS THEY DON'T OWN. The conveince benefits are just too great for people not to want their stuff accesable everywhere and the cost of running your own servers is just too great. Rather than telling people not to use any of these products and convincing a few people with privacy paranoia they should be concentrating on improving the protections that information will have.
Personally if I wanted any company to be the first one to do this it would be google. They are the most likely to mount a serious legal defense against any subpeonas or other legal challenges. The EFF should be working with google to beef up the legal and technical defenses not fighting a losing battle to keep everyone's information on their own computers. For a first step how about bargaining with google and withdrawing their recommendation that no one use google desktop (though of course still warn people about the possible risk) in return for a promise/money from google to lobby for tougher protections for such data. Someone is going to do this eventually and I would rather have the law and precedent shaped by google than MS (who might not even tell about subpeonas). Basically at some point the distinction between personal info and buisness records needs to be fixed for internet stuff.
As an aside I think the greater goal of restricting the personal info that is availible on you just isn't compatible with personal freedoms. Every one of us leaks tons of information in a thousand personal interactions a day. Many of these interactions happen in public and in full view of strangers/aquintances. Unless we impose draconian laws banning people from using wearable computers that aid name recognition and record snippets of what they see for later use or abridge people's freedom of speech to post video blog entries about what happened to them today eventually powerfull search technology will make tons of information availible online. The EFF should be figuring out ways to handle this loss of privacy gracefully (so poor people don't lose more than the rich etc..) and minimize the harmfull impact not trying to put the digital genie back in the bottle.
You are missing his point (as the post right below you suggests). The first claim he makes is that there is a transverse red-shift from gravity in addition to the normal one predicted by GR.
In other words not only is there are redshit if we fire a lazer up into space from the earth (i.e. light leaving a gravity well) but even if we just shine a laser from one point on the earth's surface to another there should be a small redshift as well. His argument is that one would expect to see such a reshift in a accelerating frame because the light is traveling farther than it would at constant velocity.
Personally I'm skeptical of this argument at the moment because whether or not one would see a redshift is going to depend on the effect of that acceleration on the clocks. As the rocket speeds up the time dilation from SR increases as well, perhaps the right amount to compensate for the increased difference. At the very least the thought experiment doesn't produce a clear result (and it is always possible that multiple solutions are compatible with it).
As an aside the question of whether there is a global constant progression of time or it differs from location to location is just a matter of naming. The scientific community has decided to call the effects from acceleration/velocity changes in the passage of time because such a description seems to be more productive and simpler. However, one could describe the same phenomena by saying time progresses at the same rate everywhere but all physical processes slow down/speed up. Or to say it another way the Lorentzian theory of an ether with shrinking rulers and faster clocks is experimentally equivalent to SR and the same thing should be possible to do with GR (so long as there are no closed curves in time e.g. time travel)
This statement is just ignorant and wrong. I don't want to flame the original poster, he probably just didn't think about it too much and made a mistake, but it is even worse to mod this up to 5.
It is a well known fact that a huge percentage of processor time is used by a small percentage of the code. It is very important for that code to be fast because it is inside a loop being executed millions of times and perhaps this requires a language like C/C++ however other parts of the code which execute once for every million times the performance critical code runs can often be orders of magnitude slower without affecting overal performance.
Just as an example consider something like the GIMPS project or one of the RC whatever challenges. It is vitally important that the tight inner loops be very fast but you could write the graphic output and network components in a shell script that calls command line packet/output programs written in java and it wouldn't slow the overall app much.
Slashdot Mirror
Uhh, no the patent office doesn't give out patents for realizing that X is profitable. It gives patents for developing technology X.
Under your argument Apple should get a patent for a windowing OS despite X/Xerox parc doing it first as they were the first ones to realize it might be profitable on PCs. Hell redhat should be able to get a software patent on Linux because they realized it was profitable to sell.
The patent system simply isn't supposed to work that way. It is only supposed to give patents that aren't obvious to those skilled in the art. Meaning those video geeks working at home are the people who count not the public at large.
If TiVO wants to get a buisness method patent on selling this technology that is another matter but realizing it is profitable simply doesn't cut it for a patent on the technology.
Ohh, of course I'm not arguing that these other patents should be upheld but not TiVO's. If we are going to go down the route of upholding these obvious patents then we should do the same with TiVO as well.
However, I think it is inconsistant to cheer for TiVO winning their case but jeer at NTP or Amazon for their obvious patents. If we are going to bitch about people getting patents on obvious things we can't do it only when they are companies we don't like.
Fine then give them a patent on making commodity hardware circa year such and such do this. If the non-obvious part of their design was just coding well enough to make it work with current hardware then they can have that patent but it would be totally worthless.
This is like arguing that the first person to do multi-tasking on a PC deserves a patent for realizing that multi-tasking wasn't only something that supercomputers can do. Admitedly at that time it probably wasn't obvious that a PC had enough power to do this but that is hardly justification for saying that multi-tasking on a PC isn't obvious for the purposes of patenting.
I think one of the reasons they didn't include such functionality is pressure from broadcasters/advertisers. The ability to record and watch simultaneously almost inevitably entails the ability to skip commercials.
The fact that it wasn't yet on the market hardly shows that it wasn't obvious. It may not have been obvious that this feature would make a profit/be popular but the patent office doesn't grant patents on realizing that something would make money. The patent office grants patents based on the technology/idea being non-obvious (though buisness method patents do make this a bit fuzzy but this isn't relevant here).
The technology to watch and record at the same time is really obvious. Even the idea that this would be nice is pretty obvious. What do you want to bet that if we went back to usenet archives before TiVO we could find someone wishing he could time delay with his VCR or suggesting that this be done with a HD? TiVO just realized that this would be a sucesful product.
On the one hand I'm happy to see a company that created a truly new product be rewarded.
On the other hand I'm not sure that what TiVO did wasn't obvious. Using a HD rather than a videotape was surely obvious and automating the process of recording shows is not only obvious but far from original. All TiVO did is put both of these things together in a convient package and I find it hard to formulate any principled rule that would call TiVO's developments non-obvious but NPT's blackberry patent obvious.
The difference seems only to be that geeks like TiVO and TiVO hasn't being suing individuals who decide to set up their computers to tape shows for them. But if they do get their patent enforced that is exactly what they *could* do in the future.
In the end I tend to think the TiVO patent should have been rejected as obvious. However, I think the only reason TiVO didn't make money is the monopoly cable companies and satellite companies have on their markets. When data service becomes the commidity everyone buys and there is free competition amount content providers on an open protocal companies like TiVO won't be shut out by monopolists who can make it difficult for TiVO to penetrate their markets.
Legally I mean.
It is far from clear for me. It is obvious that the EURid rules were really horrible things but if they were no more specific than reprinted on these blogs these companies may be perfectly within the rules.
After all if I am some domain registrar the rules don't seem to prevent me from creating 500 new companies, each of which has me as it's only customer, and use them to buy up names. After all each of these companies is genuienly intending to become a registrar, just of only a few names and is trivially giving all it's customers access to these domains since it only has one customer.
Maybe the formal rules were more specific but as it is it is far from clear these companies were violating any rules.
By carrier I meant cell phone carrier. Unlike the other responder I wasn't imagining a situation where you totally avoided phone companies all together. In fact such a situation hardly seems possible as someone is going to be carrying your packets and they will know you are making a call whether you are doing so using an internet phone or a regular phone (you don't really think the telcos aren't tracking the amount of Skype traffic on their networks do you?)
I just meant a situation where sprint can't anally rape me for minutes I use over my wi-fi connection and I fear they will figure out a way to do this up until the point I can have a wi-fi conversation totally without their knowledge. I don't care if my DSL company knows about it, at least unless they start trying to charge me more for voice packets than other packets of similar QOS.
This was a very interesting article but I'm afraid I disagree with the fundamental assertion, that software patents are (in principle not as applied) no different than any other type of patent.
The first way that writing software differs from discovering a new kind of lithography or building a widget is that the underlying building blocks of software are completely understood. While occasionally we have theories that tell us what will happen if we try using such and such materials in lithography frequently in the physical world discoveries arise when we discover materials that behave in ways our theory couldn't easily predict. However, given executable (not even machine code) for a new piece of software and a processor reference manual what that software will do is completely predictable. Sure it isn't the case that every new physical discovery wasn't predictable like this but the fact that some are means the cost-benefit analysis for software patents is going to be less compelling than it is for physical patents.
The second huge difference is level of application. Software patents are things that are of immediate and direct use in real world applications. Even genuienly novel software patents are processes that almost certainly would have be developed without the incentive of software patents. For physical patents often types of fundamental materials research wouldn't be worth doing if you didn't know you could profit from the fruits of your labours if someone else comes along and figures out how to make something useful with it.
Thirdly the sheer ease of creating software makes software patents a worse deal. Creating a monopoly on some high-tech industrial process causes a few big companies to pay royalties and the cost of dealing with patent issues (excluding royalties paid) is fairly small. On the other hand every nerd with a computer can write software so the amount of innovation restricted by a software patent has the potential to be very large and the proliferation of many many companies and investors makes just traking the existance of patents in your field a very expensive endeavor.
In fact the author of the piece himself makes a very strong case that software patents are different when he argues that they are pretty much irrelevant to innovation in software even though they can be critical to innovation in other fields. Merely noting that we could scrap software patents without harming innovation is a *very* compelling argument to treat software patents differently than other patents. After all the only reason to have patents is to encourage innovation.
Ultimately it seems that the only things in software that are truly novel enough and reusable enough to warrant patent protection are interesting new algorithms. Perhaps something like shell sort might warrant patent protection but this kind of research seems to be occuring mostly at universities and has little need to be encouraged by patents. Besides once you start letting this sort of thing get patented there is no line you can draw to prevent math in general from being patented and as a mathematician this might make me rich there was a very good reason the creators of patent systems didn't want the sieve of erasthonthes (sp) to be patentable.
I really liked how the people interviewed in the article kept saying something like "consumers might expect calling over wi-fi would be free". As if they were somehow being unreasonable or uninformed.
As far as I'm concerned real wi-fi phones which don't even let your carrier know how many wi-fi minutes you are using can't come soon enough. I hate the high prices, ridiculous options and general blood sucking (prices for ringtones) and can't wait till they are the ones begging the technology companies to include support for their off wi-fi network you use when you leave the city or have at least started offering wi-fi type service in cities.
Ultimately of course the upshot of all of this is that we will be paying more for DSL/landline phones as well as for remote cell phone service. In both the landline phone market and the cell phone market massive fixed costs are amortized over a huge number of phone calls. The fixed line phone calls then in effect subsidize DSL service (the phone companies make money on it but wouldn't if they had to do all the maintence/set up the phone lines just for DSL). Similarly all the cellphone calls made in big cities subsidize building cell phone towers in more rural locations. As the distinction between different sorts of data transmission inevitably disappears the price per unit of reasonably low latency Kb must equalize. I mean it really is absurd that it is cheaper to use your phone line for DSL and utilize Skype than it is to call on a real phone. This will force the price of DSL up as it becomes less subsidized by phone calls and the existance of Wi-Fi phones will remove the ability of the cell companies to subsidize the less used more rural towers (unless of course they are just doing things in a really inefficent fashion compared to google/earthlink in SF)
At least this is what happens if the DSL prices aren't constrained by local laws, in which case we will just see more tricks trying to offer tiered access charging for cell phone use (instead of by Mb) or other stupid money generating tricks.
I certainly don't have aspergers (tho some people seem to act as if all math people do) but I couldn't tell you either of those things.
In fact I suspect most people couldn't pick out the features in virtue of which they new someone was bored much better than someone with aspergers. I suspect they would do better because they can immediatly tell whether someone is bored or not but for both cases it would seem to be something you need to learn explicitly.
It seems that much of our processing of faces and bodily expressions occurs entierly unconciously. People *look bored* to us and if we step back we can break this down into vacant eyed, glancing away, fidgety etc.. but phenomenologically they are presented as being bored to us not as having certain facial features from which we can then infer boredom (usually).
I think they have even worked out the parts of the brain responsible for this sort of processing but I could just be high.
I hate professors like this who think they know better than you how you learn best. Hell even if you are wrong it's your fucking education.
Besides, even if this helps students learn that class better it makes sure all the students we graduate for college expect to be babied and have their study habits dictated to them. What happens when they get out in the real world and they don't have people telling them to come to class and to put away their computers and etc..
Most of all though this pisses me off so much because it is very similar to professors I had in college who required class attendance or decided that it would be better to present the material in this non-standard way (which was just a bit better) so you couldn't learn out of a book.
NEWS FLASH different people learn differently. I can't learn jack squat from lecture in technical courses, i.e., math. When I have to attend lecture I just waste my time and make sure I hate the class. Yet because I went to a school where *most* professors didn't care what you did as long as you got your work done I went to virtually no classes through all of college (made it entire terms without seeing a professor) and I graduated caltech and am now a math grad student at berkeley.
When I teach I make a point of only imposing discipline as necessery to stop students from interfering with other student's learning. If a student wants to stare at the ceiling or even read a novel that's fine with me. I will advise them to pay attention but it is their buisness and they are either good enough or learn differently that they will do well on their exams or they will learn the lesson and now understand how to handle their own study habits.
By doing what? Releasing a software package which does exactly what it says it does?
Might as well say the people who wrote FTP overstepped the mark as it doesn't stop people from sending sensitive data outside the company.
While it is certainly an interesting statistic I'm troubled by the author's statement which says something to the effect of "if anything this information should be used by parents to make an educated deciscion whether or not their child should play." He might claim neutrality all he wants but implicitly this suggests that it is reasonable that a parent try to keep their child away from swear words. A position I happen to disagree with.
I mean imagine if someone said, "I gathered data on the frequency of black people at school X. But I'm not suggesting black people should be kept out. If anything parents should use this information to determine if they want their kids to be in this enviornment or be home schooled." This would clearly be a way of saying, "I think it is reasonable to try and keep your kids away from black people." Similarly this comment here says it is reasonable to keep your kids away from swears.
I was being an idiot. He doesn't actually seem to be setting up a public wiki to host these kind of calculations but creating a spreadsheet sharing system for inside a company. Still usefull but much more boring.
Hopefully someone will come along and do what I suggested anyway.
I'm not entierly sure how many buisnesses will be willing to trust such a system. One needs to be carefull that your competitor hasn't inserted some special logic to screw things up when *your* data is entered into the system. Unlike wikipedia where everyone can see any devious attempts to change things code or calculations can seem fine and work fine on most data but contain subtle bugs.
However, for plenty of individuals and personal buisnesses this seems like a great idea. All the time there are common calculational tasks we do that are programed over and over again. Super simple examples are dates being turned into days of the week or dollar amounts in various years being turned into 1970 (or some other amount) dollars. Some of these are tought because they require the right data (rate of inflation for each year) others are tough because they just involve lots of special cases. In any case having a wiki where people can use other people's work and build off of it seems like a great idea.
In particular it seems like a code repository where you can just run the code right there. This is the biggest problem with code repositories now, often downloading the already written code and figuring out how the API works just takes too much time and since there is no guarantee that someone will download a whole code repository people tend to reinvent the wheel many many times.
It would be great if ultimately this could support more than just a spreadsheet interface. A general wiki for *online* DHTML calculation tools, especially if it also had functionality to let you download the code or acess it as a web serice, would just be awesome. Maybe this already exists somewhere and I don't know about it but the idea seems great.
Where ohh where will the magical internet take us next?
I think most of us dislike HDCP and hope it fails. Unfortunatly most consumers don't care enough to vote with their feet, or are too afraid everyone else is going to defect so they buy HDCP enabled systems too. Situation seems hopeless doesn't it?
Not so fast. What if we could get the major video card manufacturers to make sure everyone thinks this new technology works on their cards when it really doesn't. When HDCP content comes down the line these people will have a horrible experience when they buy the content and it doesn't work and are going to be loath to shell out more money for another 'HDCP enabled' video card. Then the public might dismiss HDCP video as a buggy and annoying product.
Of course it sounds like a horrible PR move for any video card company to make, even counting the extra sales it gets when people upgrade, but we (at least those of us who didn't buy one of the crippled cards) can be happy about their screw-up.
As a side note does anyone know what the anti-trust law is with regard to this sort of content protection scheme. In particular in situations like this which have clauses excluding open source video players, and given that there are certainly companies that do package open source video players as part of their buisness model (google now too I think) why isn't the consortium that defines these content protection standards in violation of the anti-trust laws.
The position of Yahoo and MS is a bit more doubtfull in this. There is a big difference between being the 1st big company to do this and the 3rd. Besides evidence has come to light suggesting they are helping repress the chinese people more than they have to.
But to accuse google of hypocripsy in this case makes about as much sense as accusing a billionaire who pays some terrorist a ransom to get a family member/employee/whoever of hypocripsy because he said he didn't give money to terrorist. The billionaire's statement that he doesn't give money to terrorists should obviously be understood as saying he doesn't approve of the terrorists actions and will not aid them in killing/bombing by donating money. Of course there is a genuine question about whether paying them off in this case to *STOP* them from killing someone is worth the extra money/moral boost it gives the terrorists but this is a tough and unclear point.
Google is similarly in a hostage situation and it would be just as wrong to accuse google of selling out. What is being held hostage though is not google's profits (though those do hang in the balance too) but the chinese people's access to information and the internet. IF GOOGLE DOESN'T CENSHOR GOOGLE.CN THE CHINESE GOVERNMENT CENSORS THE RESULTS ANYWAY. It just isn't the case that if google took some principled stand that the chinese people would have any more access to the internet. In fact google seems to have done a remarkably poor job of their censoring (the misspelling issue is well documented) so they probably give the chinese more access to info than they did if they had refused. Quite simply the choice is between a censored version of google being availible to the chinese and a slow, buggy, frequently inaccessable censored version of google being availible to the chinese. Also I suspect by doing self-censorship rather than having chinese filters do it the censorship system will move much slower but I'm not sure.
Yes, had google been the first company to do this there would have been concerns. Perhaps china might have relaxed its censorship policy if western companies hadn't been willing to go along (although how do we know this isn't what happened? By making a comprimise these companies might have gotten china to loosen their policies...though I wouldn't count on it) or the chinese would have been more outraged if the censorship resulted in that much inconveince. However google was not the first company. MS and Yahoo were already there and they provide acceptable search results. Any damage was already done when these companies made their deciscion, google could either refuse to go along and just let censored MS and Yahoo cites take all the traffic or jump in themselves.
SO LONG AS GOOGLE IS LESS REPRESIVE THAN MS AND YAHOO THE CHINESE USERS HAVE MORE FREEDOMS WITH GOOGLE.CN THAN THEY HAD BEFORE.
Given the records of MS and Yahoo I suspect this is probably the case. Even if all the western companies had refused it is unclear whether this would have been good or bad. After all if china developed its own native search company the government would have far more power over that entity and fewer ties with the west would mean less long term pressure to change.
It is particularly rich to see congress critisize google and these other companies whe
This is why programs are released as BETAs!! That way people who want to take risks, live on the edge and know enough that they can repair their systems can try the software out and report any problems. If you are installing a beta it's your responsibility to go do a search and see if there are known bugs that will screw with your computer. Sure alot of us get lazy and don't do this, I know I often don't, but if it them screws up your system in a reasonable way (e.g. they aren't knowingly distributing software that wipes every users HD but just have a bug in their code) you only have yourself to blame.
I certainly have almost no sympathy for any buisness who was installing beta software (for something like this..that wasn't really needed) on anything but a test platform.
Unless I'm just missing an important part of the story, like MS suggested this beta product was safe to run on your buisness computers, this seems to be just the standard stuff one should expect from a BETA. Sometimes they get things wrong.
It is far from clear that this would make the application more robust and not less robust instead. All the magical features you get from managed code stem from the ability of the code to do automatic type inference and other types of validity checking. If the objects you are manipulating pass through an unmanaged section you can't automatically infer anything about them (except perhaps that they occupy x number of bytes and I don't know if you can even do this). Essentially you need to treat the objects from your C code as untrusted input and this is going to require a great deal of code and hence an increased likelihood of bugs offsetting the benefits of managed code.
Worse just the fact of having the interface between the two languages can make garbage collection and other issues a nightmare. Who has the responsibility to dispose of objects? Can the managed code garbage collect objects it inherits from C++? If not all the sudden you need to do garbage collecting for some objects in your nice garbage collected language and when these start getting mixed with other objects things can get messy...there is a lot of chance for error with the wrong objects not getting GCed or incorrectly getting GCed.
Also as the other poster pointed out whether of not managed code is very helpfull depends a lot on what kind of stability you need. If what is important is that the application just doesn't crash managed code is wonderfull. However, if you need to guarantee that it gets the right answer/doesn't screw up the overall computation (as say one might in some type of simulation) it is less beneficial. This is especially true when it is taking in data from unmanaged code. Essentially this just means you catch an error saying 'sorry object in wrong format' or 'unexpected value' and if there is no way to recover from this and continue executing you are screwed.
Besides if you are writing something like a scientific application where it is very important that the computational core doesn't crash but the GUI/data input type stuff is relatively unimportant if it can be restarted you might as well just make the boundaries between C++ and managed code into boundaries between processes. Since you have to do something like an API anyway to interface your languages you might as well just do it as seperate processes so you don't need to worry about crashing at all.
I tried to say that I agreed with much the EFF does at the start of my post. I meant to include in this the lawsuit against AT&T and similar measures to reduce government power/access. I expect the EFF is doing similarly good things in lobbying for changes to the buisness records exception to the search warrant requirement (what causes the increased government access with this product). Though I think they could do better in this area by working with google.
I guess I should have been more clear that I meant only to critisize the way they address the narrow issue of privacy as it involves giving companies or otherwise allowing other private individuals access to your personal information. While the things I mentioned in the first paragraph can quite rightly be characterized as issues of privacy (hell so can abortion) I tend to classify those actions more as civil liberties. Since the major concern in these areas is often the ability of the government to control you through *legal* use of this information or other harrasment I think it is reasonable to distingush this issue from their policy about our free interaction with various companies.
Or to put it really simply and leave my confusing use of terminology behind what I mean is this. I do think the EFF is doing a great job when it comes to lobbying the government for better protections against government intrusion. However, it is their positions on volountary data exchange/record keeping by the private sector that I view as misguided. In fact for the reasons I outlined above I think their attitude on these sort of issues is interfering with their ability to restrict government snooping. Of course people need to be warned that their data is less protected with google desktop but it makes no more sense for the EFF to tell me I shouldn't use google desktop than it would for them to tell me I shouldn't publish my diary online since the government can now read it. The EFF could better (though they are doing alot now as you point out) by cooperating with google and getting them to warn users about the additional risk they are taking and then working together to increase the protection from subpeona.
So don't anyone stop giving money to the EFF because of this argument. If I wasn't a poor grad student I would still be giving money. However, I think they could do even better.
Ultimately my difference with the EFF is about whether it is worth fighting for privacy in the sense of other people not being able to find out about your life. When we lived in small communities our privacy in this sense was minimal because neighbors could easily share and sift through the many clues about our lives we drop all the time. For awhile after most of the population moved to cities the number of people and thus the shear amount of data overwhelmed our capacity to process the information giving us a certain sort of anonymity (not true anonymity but just a high probability that the people who saw you every week on the other side of town wouldn't communicate this to anyone who could put 2 and 2 together and infer you were having an affair). It is inevitable that the increase in processing power will return us to something much more like the small town situation...but unless we are carefull with the added danger of identity theft.
If you really believe that this sort of privacy, i.e., people not being able to figure out what you were up to/what you liked to read etc, is valuable and possible to save the EFF's position makes perfect sense. This sort of privacy is a collective good (if everyone else's lives are out there on the web to see someone can fill in the gaps and figure out your life) and their urge for people not to use google makes sense. On the other hand if you believe as I do that it is a lost cause the EFFs attempts to cling to this type of privacy are hindering their ability to protect the important civil rights which have been implicitly aided by this kind of privacy but may recede unless we do something. Sure it's true that if
It doesn't matter if google will give their all. All that matters is that they are going to do more than the next guy. If people's stuff is going to end up on servers anyway we want to pick the company who will do the most *even if it isn't that much* to go first.
The idea that you can defend your privacy on your own just doesn't work. If everyone you correspond with puts stuff on servers you lose privacy anyway. Even if my suggestion is a slim chance it is the only chance we have.
Also hiring a lobbyist, no matter what position they support isn't the sort of thing that gets you made into road kill. You might not ultimately win but congressmen aren't going to get vindictive because you took them out to dinner. I think the best hope in this area is lobbying lawmakers not the courts (though we should pursue both avenues).
Sorry, I made a mistake. Someone who had posted before had made the remarks I had meant to be critisizing and I stupidly just skimmed your post and thought it was the same.
Well I guess the egg is on my face. I'd mod myself down if I could (hmm..I wonder if that would be a good feature)
I'm getting more and more annoyed at the privacy strategies of organizations like the EFF. I'm generally a big fan of them, I think I've even donated money (as a grad student that's saying alot), but like most online privacy activists they have a very short shighted view of privacy.
The attitude of many of these privacy activisits is somehow that we can draw a line in the sand and refuse to give out any more information about ourselves to anyone. Not letting google have our information is just another example but traffic cameras, search histories, purchase histories are others. I'm not going to argue about whether this would be worthwhile if we could do it here but it simply isn't achievable. At least for the near future people are going to continue to give more and more of their information to companies. Giving advice like 'don't use google desktop' and stuff like this just marginalizes the EFF in terms of privacy. If they took a more pragmatic approach they could do a great deal more good.
In particular in this case the EFF should recognize that SOONER OR LATER PEOPLE ARE GOING TO PUT PERSONAL INFORMATION ON SERVERS THEY DON'T OWN. The conveince benefits are just too great for people not to want their stuff accesable everywhere and the cost of running your own servers is just too great. Rather than telling people not to use any of these products and convincing a few people with privacy paranoia they should be concentrating on improving the protections that information will have.
Personally if I wanted any company to be the first one to do this it would be google. They are the most likely to mount a serious legal defense against any subpeonas or other legal challenges. The EFF should be working with google to beef up the legal and technical defenses not fighting a losing battle to keep everyone's information on their own computers. For a first step how about bargaining with google and withdrawing their recommendation that no one use google desktop (though of course still warn people about the possible risk) in return for a promise/money from google to lobby for tougher protections for such data. Someone is going to do this eventually and I would rather have the law and precedent shaped by google than MS (who might not even tell about subpeonas). Basically at some point the distinction between personal info and buisness records needs to be fixed for internet stuff.
As an aside I think the greater goal of restricting the personal info that is availible on you just isn't compatible with personal freedoms. Every one of us leaks tons of information in a thousand personal interactions a day. Many of these interactions happen in public and in full view of strangers/aquintances. Unless we impose draconian laws banning people from using wearable computers that aid name recognition and record snippets of what they see for later use or abridge people's freedom of speech to post video blog entries about what happened to them today eventually powerfull search technology will make tons of information availible online. The EFF should be figuring out ways to handle this loss of privacy gracefully (so poor people don't lose more than the rich etc..) and minimize the harmfull impact not trying to put the digital genie back in the bottle.
You are missing his point (as the post right below you suggests). The first claim he makes is that there is a transverse red-shift from gravity in addition to the normal one predicted by GR.
In other words not only is there are redshit if we fire a lazer up into space from the earth (i.e. light leaving a gravity well) but even if we just shine a laser from one point on the earth's surface to another there should be a small redshift as well. His argument is that one would expect to see such a reshift in a accelerating frame because the light is traveling farther than it would at constant velocity.
Personally I'm skeptical of this argument at the moment because whether or not one would see a redshift is going to depend on the effect of that acceleration on the clocks. As the rocket speeds up the time dilation from SR increases as well, perhaps the right amount to compensate for the increased difference. At the very least the thought experiment doesn't produce a clear result (and it is always possible that multiple solutions are compatible with it).
As an aside the question of whether there is a global constant progression of time or it differs from location to location is just a matter of naming. The scientific community has decided to call the effects from acceleration/velocity changes in the passage of time because such a description seems to be more productive and simpler. However, one could describe the same phenomena by saying time progresses at the same rate everywhere but all physical processes slow down/speed up. Or to say it another way the Lorentzian theory of an ether with shrinking rulers and faster clocks is experimentally equivalent to SR and the same thing should be possible to do with GR (so long as there are no closed curves in time e.g. time travel)
This statement is just ignorant and wrong. I don't want to flame the original poster, he probably just didn't think about it too much and made a mistake, but it is even worse to mod this up to 5.
It is a well known fact that a huge percentage of processor time is used by a small percentage of the code. It is very important for that code to be fast because it is inside a loop being executed millions of times and perhaps this requires a language like C/C++ however other parts of the code which execute once for every million times the performance critical code runs can often be orders of magnitude slower without affecting overal performance.
Just as an example consider something like the GIMPS project or one of the RC whatever challenges. It is vitally important that the tight inner loops be very fast but you could write the graphic output and network components in a shell script that calls command line packet/output programs written in java and it wouldn't slow the overall app much.