The Electoral College system has been losing popularity in recent years (notably among Democrats, for some odd reason *grin*), but I actually think it's a good thing, and here's why: No election is ever going to be perfect. In order to declare a winner with certainty, you need a very certain tally of the votes. I think we should be able to get the counted results for an election to be *very* reliable, in terms of errors, but I don't think you can ever achieve *perfection*.
When you have extremely close elections, like in the 2000 USA election between Bush and Gore, (witness how much havoc was wreaked by "Hanging Chads" and other problems), it's almost impossible to get a nationwide total that people will agree is valid, particularly if the difference between the candidates is less than 1/10 of 1 percent. You get trapped in 'recount' limbo, and 'rules lawyer' hell (where advocates for either side try to argue why certain ballots should be counted one way or another, trying to guess the intent of a vote with a hanging chad, or trying to figure out if some votes were made by people illegally voting multiple times with the names and addresses of dead people, or the same person voting multiple times under different addresses in different precincts.
The electoral college system helps 'smooth out' our inability to get *perfect exact totals*, by making the election be a district-by-district contest, where it's usually easier to decide which candidate got more votes in an individual district or state, than it is to determine the exact national total of votes. It's sort of like analog vs. digital recording of data: theoretically, analogue would be an exact represention, perfect, but we find in reality that analog recordings suffer from imperfections which distort them; digital, on the other hand, while never a truly exact/perfect representation of the data, gives us a way to record the data in such a way that we can compensate for later distortions which are introduced during transmission or duplication, and usually get much closer to perfection than analog allows.
(I would like to note that, technically, right now, the 'districts' are entire states; I do think we should break it down into smaller districts, like congressional districts or something - I don't like winner-takes-all delegate allocations at the state level, because that's too 'low resolution').
With the electoral college, if there is a problem with voting in one state or district, you can at least narrow down the 'fight' over recounts, etc, to the state or district where there is a problem or extremely close contest and don't have to worry about any other states/districts. If we went to a popular national vote, if you have a close election, recounts and rules lawyering will have to go on in every single district in the nation. That sounds ugly, and expensive to me, and more susceptible to fraud/manipulation, because the nations attention will be spread out over every state/district, instead of just worrying if the votes in say, Florida, or Ohio, or New Mexico, are accurate, and if there was fraud in those individual areas. It allows us to focus on specific places, instead of *everywhere*.
Even when it is digital, he could probably get decent reception if he got a decent outdoor antenna for his house. Most parts of the US (there are probably exceptions, but I believe it's generally true) have at least *some* over-the-air coverage that is good enough that if you have a directional antenna mounted on your roof, (or an antenna tower), you can receive at least a PBS affiliate or two, and at least one of the major commercial networks (usually all 4).
That said, I sympathize with the ideal of just streaming your TV over the Internet instead of bothering with antennas, and as others have said, Hulu will probably have it. It's probably also worth checking NBC.com, ABC.com, CBS.com, and Fox.com. Maybe PBS.org. Maybe CNBC.com, MSNBC.com, CNN.com, etc. Youtube might stream some coverage.
I'll start by confessing I'm no network engineer, but as a user, some things I'd like to see (all of which, I think, IPv6 implements?):
Trivial encryption for any type of data / all network apps required to support encryption:
It bugs me just how many network apps, from Instant Messaging clients, VoIP, etc, which arguably should have encryption, don't. Recently, I was looking into online telephony providers, and I like the idea of using a standard-based provider which uses SIP (something like a Vonage, Gizmo, Fonosip, etc), but as far as I can tell, right now *none* of the SIP telephony providers support encryption (Gizmo5 does for Internet-only calls, but not Net-to-POTS), which is pretty mind-boggling to me; so, I'll probably just go with Skype, even though I'd prefer an open standard).
Granted, not every application *needs* encryption, and in some cases, the performance overhead could be bad for the intended traffic (things like video-games, live broadcast-streamed video or audio [things like TV shows, web-seminars, etc, which maybe the person streaming the data doesn't need encrypted because it's for a general audience and is not private], etc), but crypto should be much more pervasive, so that if I *want* to turn it on in any app, I can (maybe I want to run a secure Quake server and can live with the performance degradation). I think putting it into the protocol stack could make this possible?
I think IPv6 does this with the IPSec concept, doesn't it, where all the implementation of encryption is done in the protocol stack, so that applications don't have to individually link in crypto libraries, but instead, the app basically sets a flag to true or false whether the connection should be encrypted?
The end of price-gouging for multiple public addresses:
I really think it's *stupid* to have to pay $5 or $10 per month, or whatever, for a *number*. Numbers should be free. There's an infinite supply, so the law of supply and demand should make them free. I'm already paying for Internet service, so I shouldn't have to pay more for addresses. Of course, right now, because there is a limited supply of IP addresses, you do end up paying for them (after the first) because there *aren't* an infinite (or effectively so) number of addresses.
Having a public static IP address makes things like direct connections from one person to another for things like VoIP, file transfers, VNC/RDP, games, etc much easier. Yes, there are schemes to work around NAT nowadays, but almost all of them require the use of some third-party node which *does* have a public IP address.
I sometimes hear people raise as a would-be counter argument that NAT increases security, but not really more than a simple firewall on your cable/dsl modem or WAP would do. The problem with NAT is that it is a bit more difficult, if you have multiple users behind the NAT, to all receive inbound traffic on the same port (which might happen for certain applications; e.g if you are hosting a LAN-party or you just have multiple gamers living in the same house).
Secure DNS:
This, I don't think, actually requires IPv6 (but might be made easier with IPSec?); I think it can, and will eventually, be done with IPv4, but it's still an issue with the current Internet. I'd like to always use a *trusted* DNS server no matter what network I'm roaming on. That is, always use my ISP's DNS server, or my own DNS server, instead of the DNS server of whatever WiFi network I'm currently on. I could try that without secure DNS, but there's not much guarantee that a man-in-the-middle isn't intercepting the DNS requests en-route to my 'trusted' DNS server, so I can't really trust the replies.
Email origin forging:
It's entirely too easy to forge the "From:" address on emails on the current Internet. Yes, you can use signing/encryption software to get around this (PGP/GPG, or the SSL certificat
"One-shot electronic money transfer are the future, I wish my credit card made them easy to do for everything. Maybe they should go talk to those cell-phone money guys in India."
I have a card with a bank that has a one-time credit card number system I can use for one-shot transactions, or recurring paymemnts. I go to the website for the bank, and after logging in, I can request a number be created, I can specify how many months until it expires (2 month minimum), and I can specify a spending limit. It generates a number instantly, which I can use for payment. Once the number has been used once (unless I specified that it the recurring payment option), it cannot be used again. With the recurring payment numbers, I believe once a merchant has used it, it becomes 'tied' to that merchant, so that no other merchants can bill it. With the recurring payment, I can specify the per-payment limit, and it can only be billed at most once per month (I think).
It does add some inconvenience to transactions, and it occurs to me that if anyone ever were able to get into my online account, they could generate themselves some numbers and go shopping, but I think that it's more secure than the alternative of using one number for transactions with every merchant. I don't know for sure, but I think that if there were a problem, like that SOE billing system problem that caused a credit card to get locked, if you used one of those generated credit card numbers when you setup payment with the affected merchant, it might not affect transactions on the same account with other numbers and other merchants.
I'm curious. I know some people don't have good credit, so they can't get 'credit cards', or they are underage. What are the requirements for the 'pre-paid' credit cards I see advertised from time to time? I would assume, being pre-paid, that you don't need to have good credit? Not sure about the age requirements though - can a minor get a pre-paid card?
I'm not sure why more people don't use pre-paid credit cards for online transactions if they cannot get a 'traditional' credit card?
A couple things about T-mobile puzzle me a little bit. ..
1) T-mobile has this pretty cool looking option called "Unlimited Hotspot Calling" which you can add to some plans, which allows you, if you have a phone which supports the tech, to make calls over WiFi instead of the cell network, and while you are on the wifi, you get unlimited minutes. Here's the really cool trick, which Skype and other VoIP apps can't match - the calls can be transparently handed back and forth between WiFi and the cell network. You call someone while driving your car, and get within range of a WiFi network that you are able to connect to (either because you've registered the necessary WEP/WPA keys, or because it is open), and the call will switch to the WiFi network and you'll stop using minutes. Make a call while on the WiFi, and then leave the range of the WiFi network, and the call will switch to the cell phone network.
Guess what? G1 does not (at least currently) support the Unlimited Hotspot Calling feature. I don't think there is any technical reason why it couldn't - it's got WiFi, and is fully programmable, so it seems like T-Mobile's engineers could implement it (and perhaps they will; to be fair, Android 1.0 was just released pretty recently, and it's probably not trivial to create the code necessary to implement the unlimited hotspot calling; additionally, that technology might be from a third-party (like a Motorola or Cisco or something) which T-mobile just bought equipment from to implement the feature, so they might have to wait for the third-party to decide if they want to bother with creating the necessary code for Android.
Still, I really hope they do bring that to the G1, as it seems like pretty cool tech (if it works; never actually tried it myself, but I'd like to).
2) T-mobile is, compared to Sprint, ATT, and Verizon, the 'small fish in the pond'. Seems to me that in an environment where you are, at best, the number 4 player in the market, you would be extremely price-competitive with the larger players, to try to increase your market share, but T-mobile doesn't seem to offer particularly better deals than the larger players. I suppose that might be because the company probably can't afford to run at significant losses just to build market share (they do, after all, have to remain financially solvent), but it seems like that "Unlimited Hotspot Calling" feature is one which could allow them to give a lot of potential customers a great value proposition without actually costing T-mobile a lot of money.
VoIP calls are, I believe, ridiculously cheap (most of the Pay-as-you-go VoIP services seem to run at about 2 cents a minute for calls in the USA, and I assume that those companies are making a profit at 2 cents a minute, which means the wholesale cost of VoIP must be pretty ridiculously low). I would think T-mobile would be *all over* trying to market the hotspot calling feature as a way to bring in lots of new customers, but, not really. They seem to barely market the feature at all, and currently, they have a very small selection of phones that even support the feature. It might be they are having a hard time getting phone manufacturers to release models which support the tech, and so maybe without a good selection of phones, they don't feel it's wise to heavily market it, I'm not sure.
The one other somewhat puzzling (though not completely) thing about their Unlimited Hotspot Calling feature is that you must have a plan which costs at least $40/mo before you can add the UHSC, which means the minimum cost is $50/mo. For people who talk a lot, that's not a bad deal, but for people like me who talk a relatively small amount (though I'd like to talk more, but I can't afford expensive plans with lots of minutes), we get basically no benefit the way they currently have things structured.
T-mobile has a basic plan for $30/mo which comes with, I think 200 Anytime minutes, and unlimited weekends I think. I would
I didn't realize there is a version of Skype for Android. I've checked the Skype website to see if there were any announcements of such a beast, even tried a general Google for Skype for Android, and haven't seen it. Where'd you find Skype for Android? The closest thing I found was something called iSkoot, which doesn't appear to actually use Wifi?
Wasn't Bioware swallowed up by EA? Although, to be fair, I think EA might have been smart enough to buy the company, and leave it autonomous enough to not ruin it, but we'll see.
Yeah, this compulsory list sounds great in theory, but think how much time, per customer, it would add to checkins. You can already easily be stuck in a line for 20 minutes or longer at most airports, just to check in and check one bag *without* having to go through the contents item-by-item with the airline personel. If you had to go through everyone's stuff, it would take 20+ minutes per customer, meaning that if you have 10 people in line ahead of you, you're waiting for 200+ minutes.
Hey, thanks for the tip. I turned off desktop effects (I'm using Ubuntu, and I believe that the Desktop Effects feature is Compiz?), and after turning that off, it did seem to improve the playback of flash quite a bit. (It also seemed to fix full-screen playback; previously, whenever I tried to switch a stream to fullscreen view, it would immediately snap back to windowed view, and I couldn't figure out why it was doing that). I didn't realize there was a conflict with Compiz. Maybe it's not Adobe's problem after all.
I've got my laptop setup to dual-boot Windows and Linux. I've noticed that, generally speaking, the Windows version of Flash can play higher resolution videos at full screen with better framerates, than the Linux version of Flash. For example, I pretty frequently go to hulu.com to watch episodes of TV shows that I've missed. Many of the videos have a 480 resolution option (which, really, is just standard definition, but 'normal' resolution on hulu is like 360 or something like that). Under Windows, the 480 resolution videos will *generally* speaking flame back pretty smoothly with very few dropped frames, but under Linux, the Flash player is like watching a slide show.
Also, outside of video playback, I've noticed that other functionality of Flash is often just much slower under Linux, like full-screen animations (as an example, Marvel.com has a digital 'comic book' viewer implemented in Flash. When you turn a page, it does an animation which looks like the page physically turning, sort of. Under Windows, it takes like a second, but under Linux it takes like 2-3 seconds.
I suspect that the Windows version of Flash takes advantage of various video acceleration features of the nVidia drivers (probably via DirectX), but that the Linux version does not use such acceleration features. I am using a recent, accelerated video driver from nVidia, so I believe that the acceleration features are *available* in Linux, but simply that the Flash for Linux does not use them.
One thing I've wondered about the idea of wireless internet access for the masses is, what happens when you have 2000+ people in a single city block (think of places like NYC, LA, Chicago, etc) trying to use a wireless Internet service? Does the system get completely bogged down with lag?
I assume that the planet is so cratered because of no/minimal atmosphere?
I can't tell, exactly, from the photograph, but are there any mountain ranges on Mercury? The image makes it look, except for the craters, pretty uniform in the surface elevation?
Can anyone give us a sense of scale for those craters? I'm guessing that, to be as well defined as they appear from space, they must be something like 1/4 mile or 1/2 mile deep?
You know, I'm just thinking - your comment about them turning off the hydro plants at night made me think of something - if someone can ever come up with high-temperature superconductors, it occurs to me that Norway and the other European countries might be able to have a nice way to make some additional import revenue by selling their excess hydro power to the US, Canada, and Central/South America, and maybe parts of Asia(?). I believe, because of timezone differences, it's still afternoon and evening in the Americas when it's night and early morning in Europe, no?
Then, as the US is shutting off the lights, so to speak, at night, we could be selling excess generation capacity to Europe, the Middle East, and Africa.
Oh well, yet another good reason why economical high-temp Super Conductors would be one the most important discoveries of the century (in whatever century it happens), if it ever happens.
I get up in the morning, my cars all charged up. Great. So, now I'm ready to drive to work, which might be a 40 or 50 mile commute. My car might have a maximum electric range of 60 or 70 miles, say. So, I get to work, but my car batteries are almost dead. I'd sure like to be able to plugin at work, so that 8 hours later, when I'm done working, I can drive home. But that means charging my car during the day as well as at night.
If my car is a hybrid, I can at least drive home on gasoline, but to get the most benefit from electric vehicles, you want people doing most of their daily driving using the electric instead of the gas, which means you pretty much have to accept that people will be charging during the day, as well as at night.
As for differential pricing of electricity day vs night, I think they already do that some places in the US. I have some friends who live in Arizona, and I believe that they have a differential pricing scheme out there (because air conditioners use up so much electricity during the day, I think), so people wait till night to do things like laundry, running the dishwasher, etc).
I would be concerned about the privacy implications of using Yahoo Mail, Hotmail, or GMail for your student and faculty email. Now, granted, a lot of college students will be using one of those three for their personal email accounts *anyhow*, but for faculty in particular, and even some students, there could be some real downsides to using a third-party email provider.
For example, I don't know what Uni you're from, but a lot of Universities have faculty and students who are involved in research which might be of a nature where it might not be good to have them sending emails through a third-party. For example, professors and/or students working on Defense dept, Energy department, or CIA/NSA research (although, it might be that in such a situation, they would be using a more secure email system run by the government agency they are collaborating with, instead of the University email, anyhow, so maybe that's not such a concern).
Still, in general, I don't like the privacy implications of using Yahoo, Microsoft, or GMail for university email systems.
You might ask the representatives what guarantees of privacy they are willing to make to the University and it's students, faculty, and staff. I think I would hold them to a higher standard than what the normal Yahoo, MSN, or Gmail privacy statements offer.
Umm, in the case where a single word is being obfuscated, there is no way around the technique you mentioned. Heck, you can often do what you are talking about without a computer. Just read the text, and you can often guess what a missing word is by context.
Assume, for a moment, that 2 or 3 lines of text have been 'blacked-out', and it suddenly gets extremely challenging for either human or computer to guess the missing text. Or, let the redacted text be something like a telephone number, password, social security number, etc, and again, just knowing the number of characters missing doesn't help you, because of the huge possibility-space in that case.
If you think a black, or mono-colored box, is ugly, then as someone else said, randomize the pixels, don't just re-arrange them according to a fixed algorithm (unless that algorithm uses randomness to determine which two pixels to swap per-pass, that might work out ok, since the randomness would make it nearly impossible to figure out what order to swap back the pixels in).
I've, from time to time, mused about the possibility of trying to create a game with a 'programming' mini-game. This might not obviously be programming to most users - maybe it would use some sort of icon-based programming (which it sounds like LittleBigPlanet sort of has with this parts system). Maybe this could be a system to let users create their own spells in a magic game, or used as a 'hacking' mini-game in a sci-fi game (something like Bioshock or Mass Effect, but replace the simplistic GUI puzzle 'hacking' mini-game with a slightly more robust mini-game which actually encourages people to learn real programming techniques), or maybe the ability to give a ship or other piece of equipment new abilities in a sci-fi game.
Anyone know to what extent this idea has been tried in the past by any other games? The only thing that comes to my mind is a game I saw a few years ago (can't remember what it was called now), where the player was in some sort of base on Mars or one of the moons of Jupiter or something, and the player created these autonomous vehicles by combining parts (chasis, engine, wheels, breaks, batteries, and various 'logic components') using a wiring system (which is sort of like programming). Then the vehicles would be pitted against each other in a sort of arena. Sometimes you would be racing an obstacle course, other times the vehicles were fighting each other (you could get weapons which you could wire up to the vehicle).
I imagine that, for the game to gain any popularity, this should be a fairly optional part of the game, since most users might get a little overwhelmed by it, if it were complex enough to be fairly powerful.
There's still the economic argument - if a lot of people start using electric cars, we will likely go through at least a period of time where electricity costs increase due to the increased demand. It might be that over time, new power plants will come online, which might reduce the cost, but I'm very afraid of what electric vehicles will do to electricity prices in the US, if they ever become popular.
No, this isn't another "Global Warming doesn't exist" troll. This is about energy. Electricity, specifically. It has to be generated somehow. Sure, hydro, wind, and solar are great sounding ideas. To some extent, they already work, and no doubt can be made better if investment is done in R&D, and deployment, for those technologies. But, imagine something on the order of 100 million cars which need to be charged, possibly twice or thrice, every day. That is a massive amount of electricity which is not currently being generated that needs to be generated.
Where will it come from? Right now, probably *not* from clean energy sources. People don't want to build more nuke plants; I believe hydro electric's generating capacity is already pretty much tapped out and there's not a lot of sites to build new damns; Wind and Solar are both spikey, and currently just aren't available in nearly the quantities necessary anyhow.
That leaves coal, oil, natural gas, and maybe something like propane (but, I think most propane comes from oil, doesn't it?)
Even allowing for generation from carbon-based fuels, it would require new power plants to be built and brought online, which if you believe in global warming, you don't want to encourage, because once a power company has invested in building the plant, the government is probably not going to shut it down (at least in the USA), so that means that you now have many additional carbon-based power plants and will continue to have them for 50-100 years.
So, unless people want to build a lot of fission nuclear power plants (which the public doesn't seem too enthusiastic about; McCain suggested it and that whole conversation died in about 30 seconds), electric vehicles simply are not a viable solution to global warming. Maybe, someday, if we can get cheap, clean, safe fusion nuclear power, electric vehicles might become a great option, but not right now. If electric *did* become super cheap in massive quantities, electric vehicles still might not be the best option - with cheap electricity, something like hydrogen or some other type of synthesized fuel might be a better option. Hydrogen has containment/safety issues, but still, chemical fuels that you can carry in a tank, and nearly instantly re-fill at a fueling station are likely to always be a better solution than electric, unless someone can find a way to instantly recharge an electric vehicle.
Man, I would love to just see how all the Apple customers react if iTMS got shut down, and they began to lose access to their purchased music (I would presume the music would continue to work on devices for which the songs had already been authorized, so it wouldn't be catastrophic, but still). I suppose if worse comes to worse, they could at least burn their music to CD's, since iTunes does allow that (will that work if the Apple servers were shut down?). Once ripped to CD, they could re-rip to MP3, but the process would be painful if you have many CD's worth of music to burn and re-rip, and would result in at least a small amount of sound quality degradation (though, in their credit, the iTunes aac's are pretty high quality to begin with, so I think if you encode to a high bit-rate MP3, sound would still be pretty good).
Meanwhile, people who bought unencrypted MP3s (or Oggs if you can find em anywhere) from places like Amazon, Walmart, and Rhapsody, don't have to worry about losing access to their collection (as long as they remember to backup regularly).
I've always wondered, why does it take so ridiculously long to reboot a computer? The basic problem seems to be that you want to send every program on the computer a signal to quit (sighup or sigterm or something like that), and those processes then each begin doing things like writing data to disk, closing file descriptors, running destructors to 'properly free' allocated memory, etc.
Now, here's the thing - yes, programs need 'cleanup' for a reboot, but I think simply writing data to disk and closing files would be much faster than completely shutting down programs, and doing all the *unnecessary* memory cleanup that implies. Yes, filesystems need to be flushed/unmounted, but I think that probably 80 percent of the stuff that happens during a typical computer shutdown or reboot is just unnecessary, particularly doing things like running destructors on complex data structures (like the 800 megs of model, texture, animation, particle effect, sound, and state [e.g. character inventory, stats, etc] data that a game might have loaded in memory, for example). Seems like all you need to do for a shutdown or reboot is send a signal to programs to write data and close files, then when that is finished, umount filesystems, then send the appropriate instructions or interrupts to the CPU to trigger the shutdown/reboot. Who cares if programs don't free() their memory or properly terminate if you are rebooting or shutting down?
"Sure, their email system is not capable of handling mass emails from the 300,000,000 people in the U.S. that are pissed off about the proposed bailout. .."
I also wonder if you don't sometimes end up with certain motivated individuals and PACS setting up systems to automatically send the same email 20 times/sec to representatives. Note, I will distinguish between a 'form-mailer' page where actual constituents go to the page, fill in their name, and click submit to send a form-email with their name on it, as that, at least nominally, is really a 'new' letter from a unique individual.
How does this thing deal with plug-in/add-on based systems like Firefox or Eclipse, where new capabilities get added to the executable through dlls (or java classes, I guess, in the case of Eclipse? - Although, with regards to Java, I wonder if this system would work at all, since I think the kernel never exactly 'sees' Java programs or classes as executables, but only the JRE, which already has all the system calls built into it?)
Slashdot Mirror
The Electoral College system has been losing popularity in recent years (notably among Democrats, for some odd reason *grin*), but I actually think it's a good thing, and here's why: No election is ever going to be perfect. In order to declare a winner with certainty, you need a very certain tally of the votes. I think we should be able to get the counted results for an election to be *very* reliable, in terms of errors, but I don't think you can ever achieve *perfection*.
When you have extremely close elections, like in the 2000 USA election between Bush and Gore, (witness how much havoc was wreaked by "Hanging Chads" and other problems), it's almost impossible to get a nationwide total that people will agree is valid, particularly if the difference between the candidates is less than 1/10 of 1 percent. You get trapped in 'recount' limbo, and 'rules lawyer' hell (where advocates for either side try to argue why certain ballots should be counted one way or another, trying to guess the intent of a vote with a hanging chad, or trying to figure out if some votes were made by people illegally voting multiple times with the names and addresses of dead people, or the same person voting multiple times under different addresses in different precincts.
The electoral college system helps 'smooth out' our inability to get *perfect exact totals*, by making the election be a district-by-district contest, where it's usually easier to decide which candidate got more votes in an individual district or state, than it is to determine the exact national total of votes. It's sort of like analog vs. digital recording of data: theoretically, analogue would be an exact represention, perfect, but we find in reality that analog recordings suffer from imperfections which distort them; digital, on the other hand, while never a truly exact/perfect representation of the data, gives us a way to record the data in such a way that we can compensate for later distortions which are introduced during transmission or duplication, and usually get much closer to perfection than analog allows.
(I would like to note that, technically, right now, the 'districts' are entire states; I do think we should break it down into smaller districts, like congressional districts or something - I don't like winner-takes-all delegate allocations at the state level, because that's too 'low resolution').
With the electoral college, if there is a problem with voting in one state or district, you can at least narrow down the 'fight' over recounts, etc, to the state or district where there is a problem or extremely close contest and don't have to worry about any other states/districts. If we went to a popular national vote, if you have a close election, recounts and rules lawyering will have to go on in every single district in the nation. That sounds ugly, and expensive to me, and more susceptible to fraud/manipulation, because the nations attention will be spread out over every state/district, instead of just worrying if the votes in say, Florida, or Ohio, or New Mexico, are accurate, and if there was fraud in those individual areas. It allows us to focus on specific places, instead of *everywhere*.
Even when it is digital, he could probably get decent reception if he got a decent outdoor antenna for his house. Most parts of the US (there are probably exceptions, but I believe it's generally true) have at least *some* over-the-air coverage that is good enough that if you have a directional antenna mounted on your roof, (or an antenna tower), you can receive at least a PBS affiliate or two, and at least one of the major commercial networks (usually all 4).
That said, I sympathize with the ideal of just streaming your TV over the Internet instead of bothering with antennas, and as others have said, Hulu will probably have it. It's probably also worth checking NBC.com, ABC.com, CBS.com, and Fox.com. Maybe PBS.org. Maybe CNBC.com, MSNBC.com, CNN.com, etc. Youtube might stream some coverage.
I'll start by confessing I'm no network engineer, but as a user, some things I'd like to see (all of which, I think, IPv6 implements?):
Trivial encryption for any type of data / all network apps required to support encryption:
It bugs me just how many network apps, from Instant Messaging clients, VoIP, etc, which arguably should have encryption, don't. Recently, I was looking into online telephony providers, and I like the idea of using a standard-based provider which uses SIP (something like a Vonage, Gizmo, Fonosip, etc), but as far as I can tell, right now *none* of the SIP telephony providers support encryption (Gizmo5 does for Internet-only calls, but not Net-to-POTS), which is pretty mind-boggling to me; so, I'll probably just go with Skype, even though I'd prefer an open standard).
Granted, not every application *needs* encryption, and in some cases, the performance overhead could be bad for the intended traffic (things like video-games, live broadcast-streamed video or audio [things like TV shows, web-seminars, etc, which maybe the person streaming the data doesn't need encrypted because it's for a general audience and is not private], etc), but crypto should be much more pervasive, so that if I *want* to turn it on in any app, I can (maybe I want to run a secure Quake server and can live with the performance degradation). I think putting it into the protocol stack could make this possible?
I think IPv6 does this with the IPSec concept, doesn't it, where all the implementation of encryption is done in the protocol stack, so that applications don't have to individually link in crypto libraries, but instead, the app basically sets a flag to true or false whether the connection should be encrypted?
The end of price-gouging for multiple public addresses:
I really think it's *stupid* to have to pay $5 or $10 per month, or whatever, for a *number*. Numbers should be free. There's an infinite supply, so the law of supply and demand should make them free. I'm already paying for Internet service, so I shouldn't have to pay more for addresses. Of course, right now, because there is a limited supply of IP addresses, you do end up paying for them (after the first) because there *aren't* an infinite (or effectively so) number of addresses.
Having a public static IP address makes things like direct connections from one person to another for things like VoIP, file transfers, VNC/RDP, games, etc much easier. Yes, there are schemes to work around NAT nowadays, but almost all of them require the use of some third-party node which *does* have a public IP address.
I sometimes hear people raise as a would-be counter argument that NAT increases security, but not really more than a simple firewall on your cable/dsl modem or WAP would do. The problem with NAT is that it is a bit more difficult, if you have multiple users behind the NAT, to all receive inbound traffic on the same port (which might happen for certain applications; e.g if you are hosting a LAN-party or you just have multiple gamers living in the same house).
Secure DNS:
This, I don't think, actually requires IPv6 (but might be made easier with IPSec?); I think it can, and will eventually, be done with IPv4, but it's still an issue with the current Internet. I'd like to always use a *trusted* DNS server no matter what network I'm roaming on. That is, always use my ISP's DNS server, or my own DNS server, instead of the DNS server of whatever WiFi network I'm currently on. I could try that without secure DNS, but there's not much guarantee that a man-in-the-middle isn't intercepting the DNS requests en-route to my 'trusted' DNS server, so I can't really trust the replies.
Email origin forging:
It's entirely too easy to forge the "From:" address on emails on the current Internet. Yes, you can use signing/encryption software to get around this (PGP/GPG, or the SSL certificat
"One-shot electronic money transfer are the future, I wish my credit card made them easy to do for everything. Maybe they should go talk to those cell-phone money guys in India."
I have a card with a bank that has a one-time credit card number system I can use for one-shot transactions, or recurring paymemnts. I go to the website for the bank, and after logging in, I can request a number be created, I can specify how many months until it expires (2 month minimum), and I can specify a spending limit. It generates a number instantly, which I can use for payment. Once the number has been used once (unless I specified that it the recurring payment option), it cannot be used again. With the recurring payment numbers, I believe once a merchant has used it, it becomes 'tied' to that merchant, so that no other merchants can bill it. With the recurring payment, I can specify the per-payment limit, and it can only be billed at most once per month (I think).
It does add some inconvenience to transactions, and it occurs to me that if anyone ever were able to get into my online account, they could generate themselves some numbers and go shopping, but I think that it's more secure than the alternative of using one number for transactions with every merchant. I don't know for sure, but I think that if there were a problem, like that SOE billing system problem that caused a credit card to get locked, if you used one of those generated credit card numbers when you setup payment with the affected merchant, it might not affect transactions on the same account with other numbers and other merchants.
I'm curious. I know some people don't have good credit, so they can't get 'credit cards', or they are underage. What are the requirements for the 'pre-paid' credit cards I see advertised from time to time? I would assume, being pre-paid, that you don't need to have good credit? Not sure about the age requirements though - can a minor get a pre-paid card?
I'm not sure why more people don't use pre-paid credit cards for online transactions if they cannot get a 'traditional' credit card?
A couple things about T-mobile puzzle me a little bit. . .
1) T-mobile has this pretty cool looking option called "Unlimited Hotspot Calling" which you can add to some plans, which allows you, if you have a phone which supports the tech, to make calls over WiFi instead of the cell network, and while you are on the wifi, you get unlimited minutes. Here's the really cool trick, which Skype and other VoIP apps can't match - the calls can be transparently handed back and forth between WiFi and the cell network. You call someone while driving your car, and get within range of a WiFi network that you are able to connect to (either because you've registered the necessary WEP/WPA keys, or because it is open), and the call will switch to the WiFi network and you'll stop using minutes. Make a call while on the WiFi, and then leave the range of the WiFi network, and the call will switch to the cell phone network.
Guess what? G1 does not (at least currently) support the Unlimited Hotspot Calling feature. I don't think there is any technical reason why it couldn't - it's got WiFi, and is fully programmable, so it seems like T-Mobile's engineers could implement it (and perhaps they will; to be fair, Android 1.0 was just released pretty recently, and it's probably not trivial to create the code necessary to implement the unlimited hotspot calling; additionally, that technology might be from a third-party (like a Motorola or Cisco or something) which T-mobile just bought equipment from to implement the feature, so they might have to wait for the third-party to decide if they want to bother with creating the necessary code for Android.
Still, I really hope they do bring that to the G1, as it seems like pretty cool tech (if it works; never actually tried it myself, but I'd like to).
2) T-mobile is, compared to Sprint, ATT, and Verizon, the 'small fish in the pond'. Seems to me that in an environment where you are, at best, the number 4 player in the market, you would be extremely price-competitive with the larger players, to try to increase your market share, but T-mobile doesn't seem to offer particularly better deals than the larger players. I suppose that might be because the company probably can't afford to run at significant losses just to build market share (they do, after all, have to remain financially solvent), but it seems like that "Unlimited Hotspot Calling" feature is one which could allow them to give a lot of potential customers a great value proposition without actually costing T-mobile a lot of money.
VoIP calls are, I believe, ridiculously cheap (most of the Pay-as-you-go VoIP services seem to run at about 2 cents a minute for calls in the USA, and I assume that those companies are making a profit at 2 cents a minute, which means the wholesale cost of VoIP must be pretty ridiculously low). I would think T-mobile would be *all over* trying to market the hotspot calling feature as a way to bring in lots of new customers, but, not really. They seem to barely market the feature at all, and currently, they have a very small selection of phones that even support the feature. It might be they are having a hard time getting phone manufacturers to release models which support the tech, and so maybe without a good selection of phones, they don't feel it's wise to heavily market it, I'm not sure.
The one other somewhat puzzling (though not completely) thing about their Unlimited Hotspot Calling feature is that you must have a plan which costs at least $40/mo before you can add the UHSC, which means the minimum cost is $50/mo. For people who talk a lot, that's not a bad deal, but for people like me who talk a relatively small amount (though I'd like to talk more, but I can't afford expensive plans with lots of minutes), we get basically no benefit the way they currently have things structured.
T-mobile has a basic plan for $30/mo which comes with, I think 200 Anytime minutes, and unlimited weekends I think. I would
I didn't realize there is a version of Skype for Android. I've checked the Skype website to see if there were any announcements of such a beast, even tried a general Google for Skype for Android, and haven't seen it. Where'd you find Skype for Android? The closest thing I found was something called iSkoot, which doesn't appear to actually use Wifi?
Wasn't Bioware swallowed up by EA? Although, to be fair, I think EA might have been smart enough to buy the company, and leave it autonomous enough to not ruin it, but we'll see.
Yeah, this compulsory list sounds great in theory, but think how much time, per customer, it would add to checkins. You can already easily be stuck in a line for 20 minutes or longer at most airports, just to check in and check one bag *without* having to go through the contents item-by-item with the airline personel. If you had to go through everyone's stuff, it would take 20+ minutes per customer, meaning that if you have 10 people in line ahead of you, you're waiting for 200+ minutes.
-1 Impractical
Hey, thanks for the tip. I turned off desktop effects (I'm using Ubuntu, and I believe that the Desktop Effects feature is Compiz?), and after turning that off, it did seem to improve the playback of flash quite a bit. (It also seemed to fix full-screen playback; previously, whenever I tried to switch a stream to fullscreen view, it would immediately snap back to windowed view, and I couldn't figure out why it was doing that). I didn't realize there was a conflict with Compiz. Maybe it's not Adobe's problem after all.
I've got my laptop setup to dual-boot Windows and Linux. I've noticed that, generally speaking, the Windows version of Flash can play higher resolution videos at full screen with better framerates, than the Linux version of Flash. For example, I pretty frequently go to hulu.com to watch episodes of TV shows that I've missed. Many of the videos have a 480 resolution option (which, really, is just standard definition, but 'normal' resolution on hulu is like 360 or something like that). Under Windows, the 480 resolution videos will *generally* speaking flame back pretty smoothly with very few dropped frames, but under Linux, the Flash player is like watching a slide show.
Also, outside of video playback, I've noticed that other functionality of Flash is often just much slower under Linux, like full-screen animations (as an example, Marvel.com has a digital 'comic book' viewer implemented in Flash. When you turn a page, it does an animation which looks like the page physically turning, sort of. Under Windows, it takes like a second, but under Linux it takes like 2-3 seconds.
I suspect that the Windows version of Flash takes advantage of various video acceleration features of the nVidia drivers (probably via DirectX), but that the Linux version does not use such acceleration features. I am using a recent, accelerated video driver from nVidia, so I believe that the acceleration features are *available* in Linux, but simply that the Flash for Linux does not use them.
One thing I've wondered about the idea of wireless internet access for the masses is, what happens when you have 2000+ people in a single city block (think of places like NYC, LA, Chicago, etc) trying to use a wireless Internet service? Does the system get completely bogged down with lag?
I assume that the planet is so cratered because of no/minimal atmosphere?
I can't tell, exactly, from the photograph, but are there any mountain ranges on Mercury? The image makes it look, except for the craters, pretty uniform in the surface elevation?
Can anyone give us a sense of scale for those craters? I'm guessing that, to be as well defined as they appear from space, they must be something like 1/4 mile or 1/2 mile deep?
You know, I'm just thinking - your comment about them turning off the hydro plants at night made me think of something - if someone can ever come up with high-temperature superconductors, it occurs to me that Norway and the other European countries might be able to have a nice way to make some additional import revenue by selling their excess hydro power to the US, Canada, and Central/South America, and maybe parts of Asia(?). I believe, because of timezone differences, it's still afternoon and evening in the Americas when it's night and early morning in Europe, no?
Then, as the US is shutting off the lights, so to speak, at night, we could be selling excess generation capacity to Europe, the Middle East, and Africa.
Oh well, yet another good reason why economical high-temp Super Conductors would be one the most important discoveries of the century (in whatever century it happens), if it ever happens.
Charging cars at night sounds great but. . .
I get up in the morning, my cars all charged up. Great. So, now I'm ready to drive to work, which might be a 40 or 50 mile commute. My car might have a maximum electric range of 60 or 70 miles, say. So, I get to work, but my car batteries are almost dead. I'd sure like to be able to plugin at work, so that 8 hours later, when I'm done working, I can drive home. But that means charging my car during the day as well as at night.
If my car is a hybrid, I can at least drive home on gasoline, but to get the most benefit from electric vehicles, you want people doing most of their daily driving using the electric instead of the gas, which means you pretty much have to accept that people will be charging during the day, as well as at night.
As for differential pricing of electricity day vs night, I think they already do that some places in the US. I have some friends who live in Arizona, and I believe that they have a differential pricing scheme out there (because air conditioners use up so much electricity during the day, I think), so people wait till night to do things like laundry, running the dishwasher, etc).
I would be concerned about the privacy implications of using Yahoo Mail, Hotmail, or GMail for your student and faculty email. Now, granted, a lot of college students will be using one of those three for their personal email accounts *anyhow*, but for faculty in particular, and even some students, there could be some real downsides to using a third-party email provider.
For example, I don't know what Uni you're from, but a lot of Universities have faculty and students who are involved in research which might be of a nature where it might not be good to have them sending emails through a third-party. For example, professors and/or students working on Defense dept, Energy department, or CIA/NSA research (although, it might be that in such a situation, they would be using a more secure email system run by the government agency they are collaborating with, instead of the University email, anyhow, so maybe that's not such a concern).
Still, in general, I don't like the privacy implications of using Yahoo, Microsoft, or GMail for university email systems.
You might ask the representatives what guarantees of privacy they are willing to make to the University and it's students, faculty, and staff. I think I would hold them to a higher standard than what the normal Yahoo, MSN, or Gmail privacy statements offer.
Umm, in the case where a single word is being obfuscated, there is no way around the technique you mentioned. Heck, you can often do what you are talking about without a computer. Just read the text, and you can often guess what a missing word is by context.
Assume, for a moment, that 2 or 3 lines of text have been 'blacked-out', and it suddenly gets extremely challenging for either human or computer to guess the missing text. Or, let the redacted text be something like a telephone number, password, social security number, etc, and again, just knowing the number of characters missing doesn't help you, because of the huge possibility-space in that case.
If you think a black, or mono-colored box, is ugly, then as someone else said, randomize the pixels, don't just re-arrange them according to a fixed algorithm (unless that algorithm uses randomness to determine which two pixels to swap per-pass, that might work out ok, since the randomness would make it nearly impossible to figure out what order to swap back the pixels in).
I've, from time to time, mused about the possibility of trying to create a game with a 'programming' mini-game. This might not obviously be programming to most users - maybe it would use some sort of icon-based programming (which it sounds like LittleBigPlanet sort of has with this parts system). Maybe this could be a system to let users create their own spells in a magic game, or used as a 'hacking' mini-game in a sci-fi game (something like Bioshock or Mass Effect, but replace the simplistic GUI puzzle 'hacking' mini-game with a slightly more robust mini-game which actually encourages people to learn real programming techniques), or maybe the ability to give a ship or other piece of equipment new abilities in a sci-fi game.
Anyone know to what extent this idea has been tried in the past by any other games? The only thing that comes to my mind is a game I saw a few years ago (can't remember what it was called now), where the player was in some sort of base on Mars or one of the moons of Jupiter or something, and the player created these autonomous vehicles by combining parts (chasis, engine, wheels, breaks, batteries, and various 'logic components') using a wiring system (which is sort of like programming). Then the vehicles would be pitted against each other in a sort of arena. Sometimes you would be racing an obstacle course, other times the vehicles were fighting each other (you could get weapons which you could wire up to the vehicle).
I imagine that, for the game to gain any popularity, this should be a fairly optional part of the game, since most users might get a little overwhelmed by it, if it were complex enough to be fairly powerful.
Well. . . those seem like reasonable answers.
There's still the economic argument - if a lot of people start using electric cars, we will likely go through at least a period of time where electricity costs increase due to the increased demand. It might be that over time, new power plants will come online, which might reduce the cost, but I'm very afraid of what electric vehicles will do to electricity prices in the US, if they ever become popular.
No, this isn't another "Global Warming doesn't exist" troll. This is about energy. Electricity, specifically. It has to be generated somehow. Sure, hydro, wind, and solar are great sounding ideas. To some extent, they already work, and no doubt can be made better if investment is done in R&D, and deployment, for those technologies. But, imagine something on the order of 100 million cars which need to be charged, possibly twice or thrice, every day. That is a massive amount of electricity which is not currently being generated that needs to be generated.
Where will it come from? Right now, probably *not* from clean energy sources. People don't want to build more nuke plants; I believe hydro electric's generating capacity is already pretty much tapped out and there's not a lot of sites to build new damns; Wind and Solar are both spikey, and currently just aren't available in nearly the quantities necessary anyhow.
That leaves coal, oil, natural gas, and maybe something like propane (but, I think most propane comes from oil, doesn't it?)
Even allowing for generation from carbon-based fuels, it would require new power plants to be built and brought online, which if you believe in global warming, you don't want to encourage, because once a power company has invested in building the plant, the government is probably not going to shut it down (at least in the USA), so that means that you now have many additional carbon-based power plants and will continue to have them for 50-100 years.
So, unless people want to build a lot of fission nuclear power plants (which the public doesn't seem too enthusiastic about; McCain suggested it and that whole conversation died in about 30 seconds), electric vehicles simply are not a viable solution to global warming. Maybe, someday, if we can get cheap, clean, safe fusion nuclear power, electric vehicles might become a great option, but not right now. If electric *did* become super cheap in massive quantities, electric vehicles still might not be the best option - with cheap electricity, something like hydrogen or some other type of synthesized fuel might be a better option. Hydrogen has containment/safety issues, but still, chemical fuels that you can carry in a tank, and nearly instantly re-fill at a fueling station are likely to always be a better solution than electric, unless someone can find a way to instantly recharge an electric vehicle.
This story sounds quite remarkably like the beginning of the movie "Brazil".
Man, I would love to just see how all the Apple customers react if iTMS got shut down, and they began to lose access to their purchased music (I would presume the music would continue to work on devices for which the songs had already been authorized, so it wouldn't be catastrophic, but still). I suppose if worse comes to worse, they could at least burn their music to CD's, since iTunes does allow that (will that work if the Apple servers were shut down?). Once ripped to CD, they could re-rip to MP3, but the process would be painful if you have many CD's worth of music to burn and re-rip, and would result in at least a small amount of sound quality degradation (though, in their credit, the iTunes aac's are pretty high quality to begin with, so I think if you encode to a high bit-rate MP3, sound would still be pretty good).
Meanwhile, people who bought unencrypted MP3s (or Oggs if you can find em anywhere) from places like Amazon, Walmart, and Rhapsody, don't have to worry about losing access to their collection (as long as they remember to backup regularly).
I've always wondered, why does it take so ridiculously long to reboot a computer? The basic problem seems to be that you want to send every program on the computer a signal to quit (sighup or sigterm or something like that), and those processes then each begin doing things like writing data to disk, closing file descriptors, running destructors to 'properly free' allocated memory, etc.
Now, here's the thing - yes, programs need 'cleanup' for a reboot, but I think simply writing data to disk and closing files would be much faster than completely shutting down programs, and doing all the *unnecessary* memory cleanup that implies. Yes, filesystems need to be flushed/unmounted, but I think that probably 80 percent of the stuff that happens during a typical computer shutdown or reboot is just unnecessary, particularly doing things like running destructors on complex data structures (like the 800 megs of model, texture, animation, particle effect, sound, and state [e.g. character inventory, stats, etc] data that a game might have loaded in memory, for example). Seems like all you need to do for a shutdown or reboot is send a signal to programs to write data and close files, then when that is finished, umount filesystems, then send the appropriate instructions or interrupts to the CPU to trigger the shutdown/reboot. Who cares if programs don't free() their memory or properly terminate if you are rebooting or shutting down?
"Sure, their email system is not capable of handling mass emails from the 300,000,000 people in the U.S. that are pissed off about the proposed bailout. . ."
I also wonder if you don't sometimes end up with certain motivated individuals and PACS setting up systems to automatically send the same email 20 times/sec to representatives. Note, I will distinguish between a 'form-mailer' page where actual constituents go to the page, fill in their name, and click submit to send a form-email with their name on it, as that, at least nominally, is really a 'new' letter from a unique individual.
How does this thing deal with plug-in/add-on based systems like Firefox or Eclipse, where new capabilities get added to the executable through dlls (or java classes, I guess, in the case of Eclipse? - Although, with regards to Java, I wonder if this system would work at all, since I think the kernel never exactly 'sees' Java programs or classes as executables, but only the JRE, which already has all the system calls built into it?)