That's the whole point of white space transmitters, though. As unlicensed transmitters, they are required to not interfere with licensed transmitters. The purpose of the regulations are to prevent those who are paying for exclusive rights to license spectrum (broadcasters) from being interfered with by freeloading unlicensed use of the spectrum.
As far as one unlicensed transmitter interfering with another unlicensed transmitter, the FCC doesn't care. CFR part 15 says that such unlicensed transmitters must be capable of working correctly in spite of any outside interference. They solve this through techniques like spread spectrum signaling, frequency hopping, etc. As it is right now, lots of Wi-Fi hot spots interfere with each other. Does that mean that Wi-Fi doesn't work? No, it works pretty well. Why? Because the technology was specifically designed to be robust against interference. Thus, the FCC doesn't need to care if there are multiple unlicensed devices on the same frequency. More to the point, the density of the population will likely require that they allow multiple white space services to share the same frequencies unless they're opening up a lot more spectrum than I think they are.
Being able to protect white space users from one another is the absolute farthest thing from anyone's minds, including the FCC and the hardware manufacturers. It is far easier to be robust against interference than to try to prevent it.
Wow. They're out of their frigging minds, then. I hope they're starting to build their server cluster now. If you think a site like CNN.com is high bandwidth, you ain't seen nothing yet.
Exactly. I made similar comments myself. Bear in mind that the database could potentially contain all the applications for future towers/transmitters as well, which would allow devices to black out those frequencies on the proposed start date (which is the absolute earliest the station could transmit even if everything got approved in a timely fashion). Even one download per month is likely way more than is actually needed. From submission of an application to approval can take up to a year. As long as the frequencies in every application are added to the database fairly quickly, even one download every six months is probably enough. One update per month is gravy.
That said, I think you underestimate the size of the required database a bit. It's not just TV and radio stations; there are at least 22,000 stations of those by themselves, but you also have cell frequency allocations, FCC-licensed fixed stations for police and fire dispatch, LPFM stations, licensed repeater stations for various things, licensed experimental stations, etc. I'm guessing probably more on the order of 40-50,000 stations. Plus hundreds more reserved frequencies like the ham bands, various government bands, military bands, etc., many of which vary by region. And that's for the U.S. You'd probably eventually need for this to be a worldwide database. So two or three hundred thousand stations, each needing to include GPS coordinates plus a base frequency and bandwidth with reasonable precision, plus enough information to describe the coverage footprint (which isn't just a range in miles if you do it right). I'd expect it to be several hundred megabytes when all is said and done. I mean, sure, they could be sloppy and do it in a few megs, but you'd get a lot less open spectrum that way.
The problem is not that we have too many laws, but rather that the laws are not neat. In some cases, you can say that a particular task is governed by a single set of laws, but this is rare. For the most part, a task is governed by multiple subparts of thousands of different laws, and it's utterly unmanageable trying to figure out if you are violating one or a dozen of them because of the poor organization, the lack of cross-referencing, and the general failure of the code to conform to any reasonable standards of structure.
The federal code is basically like a giant app that has been patched and tweaked to the point that it scarcely resembles the original code, has little real functioning organizational structure beyond (at best) the largest functional units (titles), contains multiple pieces of code that appear to do the same thing but generate different results, and program code that jumps haphazardly through a spaghetti-programming chain of GOTO statements into arbitrary sections of the code based on outside input from the judicial system.
What we need is to make the following changes:
All new bills/laws passed must have a sunset provision that shall be no more than five years out, at which time the law must be renewed or it goes away automatically. This will greatly reduce the clutter of temporary hacks and band-aids that otherwise will continue to degrade code quality. Once a law is renewed twice, the sunset period may be extended to 20 years Once renewed an additional two times, the sunset period may be extended to 50 years. No law may ever stand for more than 50 years without being reviewed.
All new bills/laws must have a mandatory 90 day waiting period at which time the law much be passed a second time. This will further reduce the sort of knee jerk reactionary laws that tend to be the most poorly thought out.
All existing laws should be assigned a 50 year sunset starting today. This means that the laws must be revisited within 50 years or they go away. The purpose is not to make the regulations go away, but rather to force all laws to be periodically reexamined for applicability, for people abusing the laws for nefarious purposes, for people finding loopholes in the laws, etc.
Any bill that modifies existing laws must deal with a single subject per bill. All riders must be banned, period.
To limit the number of reviews to a manageable number, if a bill modifies an existing law, review of that bill may be combined with up to two other bills that modify the same law in a substantially related way. Once such a bill moves up to the 50 year approval window, its approval may be combined into the approval process for the enclosing law (but does not extend the 50 year window for that existing law).
A law that is not constantly being reviewed for effectiveness is a bad law, period.
Don't know where you get the notion that these devices are reporting their location. Quite the opposite except insofar as accessing the central database through some mechanisms could potentially reveal that there is a device of that type at that IP number, which to some very limited degree, gives a crude approximation of a location.
The proposed method of operation is as follows:
TV and radio stations, as part of their licensing process, provide the location of their towers. This information is in an FCC database along with a list of frequencies that they use.
Devices that want to take advantage of white space periodically download a copy of that FCC database. Then, they look at their copy and see which frequencies are used by radio and TV stations and avoid those frequencies. This could occur through a direct download, through your OS's normal software update mechanism, through a weekly torrent download, whatever. It really doesn't matter.
Nobody, including the FCC, cares whether these white space devices interfere with each other. They're inherently designed to be frequency hopping, spread spectrum devices precisely so that this isn't an issue. Thus, the devices have no need to register their location in the FCC's database, and indeed, doing so would be the exact opposite of what is intended, which is to preserve the frequency ranges used by licensed broadcasters.
I'm assuming that what you are afraid of is that people will build devices that query the database on the FCC's servers directly. I can't imagine that this would be the case, for several reasons:
The FCC's antenna database almost certainly has neither the server capacity nor the bandwidth to handle that kind of traffic.
You'd have a hard time connecting to a wireless network if you have no idea what frequencies it might be using. Wi-Fi scans a dozen channels and it takes a few seconds to find all the base stations. Imagine scanning thousands of possible channels. This means that you really need to narrow down the possible frequencies ahead of time, which requires knowing where you are and what frequencies are available before you try to connect to the network.
The list of TV and radio stations in the U.S. doesn't change quickly. Because stations are required to submit a full proposal to the FCC prior to starting tower construction, the data could be several months out of date and you still would have minimal risk of stepping on a live station's frequencies. Thus, an offline cache of the database would be just as effective without the privacy risk you're worrying about, and would be much more useful in terms of establishing an initial connection as well.
In short, I think it's safe to say that any such devices will use a cached copy of the database and will probably update their copy fairly infrequently. Even one update per month would likely be more than sufficient to guarantee compliance.
With the use of sensors, you have the "ABC problem". You have three collinear stations, A, B, and C. Station A is a licensed FCC broadcasting station. Station B is a receiver for that station. Station C is a frequency-hopping device that looks for an empty channel. Because station C is too far away, it cannot "hear" station A, but it is still close enough to station B to cause interference. Now granted, this is less likely when you're talking about multiple orders of magnitude difference in transmission power, but it is still possible, particularly when the transmitter might be inside a concrete structure with semi-directional leaks. This is a technologically unsolvable problem as long as you are depending on station C being able to somehow sense station A.
With geolocation, since all broadcast TV and radio stations are required by law to register with the FCC, including tower location, HAAT, a detailed map of estimated signal strengths based on topographical features, etc., you can come up with a much better idea of what frequencies are safe.
Yup. Truth is an absolute defense against a libel suit. She's pretty safe.
But just to pick nits... most geeks don't wear pants? So they walk around with their dangly bits hanging out? Maybe her coworkers are just perverts.... Did she ever think about that possibility? I don't know anybody who goes around wearing "nice pants" if that's what she means... at least not outside of marketing people. Wearing jeans puts you into the "normal" category. And although many geeks do spend every day in T-shirts, it isn't really a requirement unless you work at a Linux startup. Perhaps what she meant was that geeks tend not to be willing to go much out of their way as far as dressing up for a date (usually because they were still checking something into a CVS/subversion/git repo three minutes before they walked out the door).
Really, this is just a personality thing. A normal person looks at a list of tasks and prioritizes them, says "These won't get done," and does what he/she can. A geek says, "How can I get all of these things done in spite of myself." This tends to lead to much confusion when trying to understand the geek personality, but once you understand the root cause, it's much easier to find workarounds.
I'm a little puzzled by the choice of movies, too. Yes, I've seen Tron, but that's about as likely to come up in a conversation about movies as PA-RISC is to come up in a conversation about CPUs. Get with the times. It's all about Battlestar Galactica, Stargate, Eureka, the re-envisioned Star Trek franchise, Star Wars (in spite of Episode I), maybe Lost, maybe a few others.
And books like 1984, Fahrenheit 451, Dune, Enemy Mine, etc. That's how you get geek cred. I mean sure, most of us haven't read all of those books, but we know enough about them to carry on a conversation. Bonus points if you've read many of them.
First, Many large companies gets so much mail that the business comes with their own truck and picks it up from the USPS branch. Under those circumstances, as far as the USPS is concerned, the package or envelope is delivered the day it gets sorted and put into the pile to go to that company. However, this usually occurs after the company makes their pickup run for the day. In effect this means, the USPS almost invariably claims the package or envelope has been delivered a day before it really gets to the company. Whether this adds a day or not depends on what time of day the package or envelope got to the USPS office initially (whether it was before or after the company's pickup). So that's potentially one extra day.
Next, with the air mail, it probably flew out at the end of the day to NYC and arrived the next morning for processing. Truck drivers don't generally drive overnight when they can help it. That's two extra days.
Next, the truck has to break that trip up into two days worth of driving, assuming it's more than an 11 hour drive or that the truck driver didn't start in San Francisco. That's three extra days.
Next, because it is going by truck, it is probably indirect through Los Angeles, so it has to be sorted. That's four extra days.
Add one day lost in the back of and/or run over by a truck somewhere, and there you go.:-D
No need to get into things like acceleration. That would be inaccurate anyway. A better car analogy is capacity versus maximum speed. Consider a 50 passenger bus versus a customized tour bus. Both vehicles will get you there and back in an hour, but one carries 50 passengers and one carries half a dozen. The 50 passenger bus has high bandwidth, with one hour latency. The tour bus has medium bandwidth, with one hour latency. A Ferrari has low latency (it gets you there and back in 15 minutes, assuming you don't mind the tickets), but also doesn't have great bandwidth. A space shuttle has low latency (for a very limited set of destinations) and is moderately high bandwidth (at least for non-human cargo, if you use the cargo bay).
Fair point. However, the redirects are just the tip of the iceberg when you're talking about high latency connections. With modern web content, you have images that cause DNS requests followed by a request for the URL, scripts that cause DNS requests followed by a URL load request, and lots of other things going on that all cause suffering under high latency conditions. The web (and to some degree, the Internet as a whole) simply wasn't designed to degrade gracefully when round-trip packet latency goes much beyond about 100-200 milliseconds. Above that, throughput goes to hell in a handbasket rather quickly. At satellite latencies, it starts to be a problem. At multi-second mobile phone latencies, it is nearly unusable.
Honestly, I've concluded that the original, walled garden approach to mobile telecom was right in a perverse way. Sure, the walling was wrong, but the notion of (by default) using a proxy to make all of your web requests on a mobile phone is actually a great idea as long as it is not mandatory. In principal, a properly designed (non-transparent) proxy setup can cut the latency of web browsing almost in half by performing the DNS lookup on the proxy server instead of on the client, thus incurring the round-trip latency only once per request. That's a huge usability win when you're dealing with RTL measured in seconds. It also cuts the outgoing packet count roughly in half (most non-POST HTTP requests fit into 1-2 packets), which also improves latency for everybody else.
Simple. Same reason latency for packets varies, but in reverse. There's a huge volume of mail that moves from San Francisco to NYC, so they have air mail routes that optimize this. Because the distance from San Francisco to, for example, San Diego is relatively short and relatively low volume, they take it by truck, so it takes longer to get there, and probably stops in LA on the way (which is probably a latency disaster due to the amount of mail they process---going through LA can be like a corrupt BGP packet causing backbone traffic to be routed through your little home DSL router). With network traffic, it's the reverse; longer hauls are more likely to go to satellite, which spikes the latency way up. Shorter distances are more likely to be by cable, which has a lower latency. The point is that the haul mechanism determines the latency, and different links have different haul mechanisms depending on distance and expected load.
Just wanted to be clear about the sort of sites I'm talking about. I don't hate sites like YouTube or Flash games that use Flash because the alternative isn't really practical. I'm talking about sites like some car companies that think that their entire website should be nothing but one big pile of Flash. Such designs invariably translate into much longer page load times, clumsy interfaces with lots of unnecessary flashiness that gets in the way of usability, etc. If you're using Flash for what can and should reasonably be done with basic HTML elements, your site is a big, steaming pile of fail. More to the point, if I can't even navigate your site without loading Flash, you're doing it wrong.
BTW, it's <blockquote>...</blockquote>. You've been using BBCode too long.
first, even if you have click2flash, flash block, or whatever, you "can" run flash. you're just choosing to block it for now. ios (for now) has absolutely no capability for playing flash.
The point was that if you've taken the time to install that, you've made a conscious decision to avoid Flash most of the time, and many of those folks won't be too happy about having to load Flash just to get basic navigation on your site. It's an annoyance that tends to drive away such visitors. The reason I have it installed is that nearly every single browser crash I experienced had Flash somewhere in the backtrace. Now, with click2flash, my browser doesn't crash two or three times per day (and more to the point, hasn't crashed in months). If I get to a site that requires Flash, usually, I just close the window. I can't imagine I'm alone in this.
second, i had to install flash. it didn't come as part of the browser. it's just when my browser detected a 'flash file' it suggest i download the flash plugin. which i did (and yes, i subsequently downloaded flash block)
*shrugs* It came preinstalled in Mac OS X. I assumed it came preinstalled with most Windows OEMs as well. Either way, it's not something people think about downloading. It's just a couple of clicks in a plug-in dialog, which means that their market penetration is due largely to sites like YouTube that used to require it. People don't wake up one morning and think, "I really ought to have Flash installed on my computer." Give it a couple of years with HTML5-capable browsers and we'll see how many non-Farmville users still have Flash installed. My guess is that those numbers will drop significantly as fewer people hit pages that need it.
third, given most 'apps' on those devices make calls to port 80 making any number of micro calls to the website, there's a skew to how much "traffic" those devices account for (like, i dunno, poorly designed wordpress apps that constantly update from wordpress.com then subsequently throw away most of the data they retrieved)
Unless you are using the WebKit viewer bits on iOS, though, the browser string will not be that of WebKit on iOS, and as such, those poorly designed apps won't typically even show up as being iOS-based devices in any browser stats. If you are seeing bogus network accesses, it's likely that you're going to see them on the desktop, too. With the exception of calling JavaScript functions, UIWebView doesn't really provide much opportunity to inflate statistics. There's no cache control, no injection of arbitrary JavaScript, and normally people would not construct a web view just to capture data, tear it down, and then display part of the data in a local-origin web view. That's what the various NSURL and CFURL bits are for, which again, don't send a user agent string unless you explicitly set one, AFAIK.
fourth, when you "assume" it makes you an ass (i forget how it goes but it sounds about right).
I never thought I would say it, but I can now quite easily envision a day very soon when all my new media (games, movies, music, TV shows, books, etc.) will belong to studios, software companies, publishers, etc.--with me just renting it. There will be no such thing as buying a used book, or a used videogame. I will never be able to resell any media that I "buy." If the studio decides to have a moritorium [wikipedia.org] on a movie (like Disney so often does), they will just be able to flick a switch at any time and turn my copy of that movie off. Publishers will be able to edit all my books retroactively. When a director decides he doesn't like the ending of his movie, he can change it and force that change on everyone who owns it. If a studio goes bankrupt and takes down their servers, all my movies from them will turn to digital dust. If a judge issues a court order, all copies of a piece of media will evaporate with a single command from a media server somewhere. And when my internet goes down, so does every piece of media I own.
And in a few decades, we'll be able to say, "We've always been at war with [insert country or group]." What, you didn't think 1984 had been averted, did you? It just was an idea whose time had not yet come.
I somehow misread the post as being Adobe's direct download stats. Download.com's stats are pretty uninteresting since most people who do download it do so in response to a browser dialog box and would presumably thus get it straight from Adobe. In that case, we don't know much more than we did before. *sigh*
Thanks for the numbers. There are about two billion computer users out there on the Internet. (Source: Internet World Stats) So barely one in a hundred computer users have downloaded Flash 10 deliberately. Although one percent is a huge percentage of computer users, it's still almost two orders of magnitude smaller than the installed base....
Clicking on ads is not necessarily a good indictor of an ad's impact. Indeed, most clicks on ads are accidental. The more important question is whether the ad produces mindshare. Fair point about zooming. On the other hand, users can ignore ads on a desktop by not looking at them, so I'm not sure how different it really is....
And the statistic is highly misleading anyway. Saying that 97% of computers can run Flash doesn't tell the whole story.
First, a lot of us use tools like click2flash that report themselves AS Flash, but are NOT Flash. Why do we do this? Because we got fed up with all the idiotic Flash-based adds that make buzzing sounds at random in background windows and make us jump straight out of our chairs. These people have Flash and put up with it when necessary, but generally avoid it. Those folks are difficult to distinguish from actual Flash "users", yet they suffer a degraded experience on Flash-heavy sites, and are less likely to come back.
Second, people have Flash largely because it came preinstalled. I don't know of anyone who has actually gone out of their way to install Flash. This means that those statistics could change on a dime.
Third, it assumes that all people use the web equally. For some sites, iOS-based devices (iPhone, iPad, iPod Touch) make up as much as 11% of their traffic by volume. When it comes to ad revenue, the ratings don't matter. The share matters. It doesn't matter if they make up only 1% of the total number of Internet-equipped devices. What matters is their percentage of the traffic.
Fourth, it ignores the assumption that people buying iPads and iPhones are more likely to have disposable income than people buying a random Windows PC. Thus, for many advertisers, one iPhone user is equivalent to several netbook users. Once you understand that, suddenly even a 1% share becomes much more significant, and a 10% share becomes a showstopper.
The times I've bought books online, they've not come in 5 different boxes each with their own popcorn wrap, but rather 5 books in a single box.
Books are easy. They're not particularly fragile, and people mostly buy them several at a time, even dozens at a time. That makes them very much not normal when it comes to online shopping. Most online shopping is for high ticket items like electronics. In that space, most people don't buy a whole bunch of things together unless they buy a bundle, and electronics need a lot more protection during shipping.
Besides, what's the cost of maintaining a furnished showroom for books, with air conditioning, etc., instead of a single densely packed warehouse?
This isn't about costs. It's about pollution. Most of a business's costs have nothing to do with pollution. The cost of renting space, the cost of having more employees, etc. have minimal impact on pollution. (Yes, the employees use gas to get to work, but that's a drop in the bucket.)
Again, books are not the norm. There's a great deal more product differentiation in books than in most other product lines. When you can't get a particular brand of soap, you buy a different one. You probably don't drive to another store to get that brand, and you certainly don't drive very far. One book is not equivalent to another in the same way. This requires significant floor space for products that won't get purchased for years at a time. This greatly inflates the square footage needed for bookstores, and thus their environmental impact.
Books are also just about the only area where stores encourage customers to make special requests. This means that many bookstores get many of their goods in fairly small quantities. There is minimal difference between shipping two or three copies of a single book to a bookstore and shipping two or three different books from a warehouse directly to the customer.
So yes, books are an exception where online shopping actually might be better for the environment. I'd expect it to depend on the size of the store, their special order rate, and whether they sell used books. (The efficiency of selling used books locally would likely tip the environmental balance in favor of the local store when compared with online sales.) Again, though, books are an exception precisely because of their nature and the inherent difficulty in stocking sufficient breadth of products at a local store. Although you might see that same pattern in other highly specialized areas (e.g. professional video gear), it isn't typical of the online-versus-store debate. Most of those sorts of specialty shops have long since gone out of business outside of major metropolitan areas because there just isn't enough of a market for them to be viable. I'm almost surprised bookstores still exist for the same reason. Give it ten years, and this exception may be gone, too.:-)
Yeah.. because bulk packaged stuff is filled to the brim on a trailer. Oh wait. it isn't. Pallets consume some of the space (5% or so?) and then they don't fill to the top of the trailer, although why exactly I don't know. But about 10% of the headroom on the trailer isn't used. So.. ~15% of that volume is wasted. Less volume wasted in the boxes, more volume wasted outside of them.
If a carrier is shipping individual packages across the country, I believe they usually put them on pallets for the long haul runs and fasten them down with cellophane or similar. Besides, even if the carrier does not use pallets, they're still not going to fill the truck all the way up. There are limits to how high you can stack things before the weight of the stack damages the packages at the bottom. In short, you have all the same problems when things are packed individually, plus all that wasted air space of the outer box that increases its volume anywhere from 20% (decent boxing) up to over a thousand percent (you'd be amazed) of the original volume.
Also.. the cost of heating and cooling the stores isn't accounted for by you. But most products purchased by internet do not require temperature control (And good thing because couriers don't do temp controlled trucks).
The store is likely to be there, for the most part, whether you choose to buy the product there or not. As such, that's basically a sunk cost and shouldn't really contribute to this discussion. The only way it would matter would be if there were so much mail ordering of products that the stores went out of business. And even if the stores went out of business, there are enough people who need to get products quickly that a store's closing would drive farther to other stores that are open, which means more fuel for a lot of people. That's likely to far outweigh any savings in energy due to not having to heat or cool the building. And if the last store in an area closed, people would start ordering products with next-day delivery, which usually requires shipping by air, which is extremely inefficient. I really can't imagine that you'd break even....
Also, nobody schedules deliveries to be just in time anymore. So yeah, a week late delivery wouldn't screw a store's inventory over.. oh wait.. Not to mention that I don't know why you think merely being out of a product is no big deal. Shit, that means the customer drives to your store, looks for a product, can't find it, and then drives to another store to buy it. Not only do you not get a sale, but the customer incurs more travel costs (with the concurrent pollution emitted).
Yes, but the impact is the opposite of what you think it will be. You're assuming that this means that if a delivery is delayed because of a truck not being full, a store runs out of something. In practice, if done correctly, because the entire shipment schedule is approximately known ahead of time, the delay of that non-full truck can be factored into the schedule, and the product can simply be put on an earlier truck. Thus, if a product's arrival is delayed because a truck isn't full, it is delayed because that product is not needed until after that truck is full. Otherwise, i would not have been on that truck, but rather, a previous truck. Alternatively, it might have taken a different route entirely.
The whole point of just-in-time deliveries is that you can not only reduce the products in warehouses, but also accurately predict your shipping needs to maximize efficiency. This level of planning simply cannot be matched by single one-off package delivery because a product cannot be shipped before you order it. As such, just-in-time deliveries actually widen the efficiency gap between local stores and mail order.
And lastly.. distribution trucks offloading stuff at a store, could but generally don't, fill all the space that they drop off at the st
Fine, then only deliver packages on days when the person is getting junk mail anyway. That's at least three or four days a week....:-)
But seriously, yeah, online shopping is much less efficient, if only because instead of transporting the goods crammed together with no extra packaging material, you're transporting them individually wrapped in extra protective packaging. It's the volume, not the weight, that makes individual shipping inefficient. It seems pretty obvious that internet shopping would be less efficient.
Also, instead of the product traveling from the manufacturer across the country to a store and then a few miles to your house, the product travels across country to an online store warehouse, then across the country again to your house. In the best case, the travel is comparable; in the worst case, the travel is doubled. Of course, much of this problem could be eliminated if manufacturers provided low-cost fulfillment services for their own products. Then an online store site could be merely a storefront, and could contact the manufacturer and have them ship the product directly to the customer. I've actually seen this done with some products (particularly products from Audio Technica), but this practice doesn't seem to be the norm. Most web stores use regional fulfillment warehouses, which may be closer to the retailers in terms of travel distance than if the web store had just one fulfillment warehouse, but it is still far less efficient than a full distribution network. Also, even if the manufacturer provided the fulfillment, it would still be less efficient unless that manufacturer shipped a whole pallet full of the product out at once---a delay that most online shoppers would not want to wait for.
What some folks don't realize about big box stores is that their stores (or trailers sitting behind their stores) can be used in part as temporary warehouse space for transporting goods to other stores. Imagine four stores, A, B, and C, in a line. You need to get a product from point A to point B and some of them to point C. You also have products that need to go from somewhere else (non-linear) to point C. So you run one truck from that second location to point B and drop off products. You then run a different truck from point A to point B, which drops off part of its load and picks up the additional goods before carrying everything to point D. Similarly, if there is extra space in the trucks, the central warehouses can preemptively ship extra products out that the store will eventually use. Those products sit in the truck until they are needed, and if another nearby store needs them, those products can be diverted locally instead of having to potentially run an extra truck from a central warehouse. In effect, this allows for much more distributed warehousing than is feasible with only a few warehouses scattered throughout the country.
Adding to the efficiency is the fact that most deliveries, fresh produce notwithstanding, are not time-dependent, unlike your package. If they get there a week later than expected, it's usually not a big deal; in the worst case, the store might be out of a product, but usually they keep ahead of things so that there is some wiggle room. What matters is minimizing the amount of distance that a product travels and filling every truckload as full as possible. The result is that delivery is highly efficient.
With individual package delivery, although some effort is made to do the same sorts of optimizations, the timeliness of the delivery limits the extent to which this is possible. Similarly, the inability to pre-distribute goods using otherwise wasted truck or plane space makes individual package delivery inherently less efficient.
Slashdot Mirror
That's the whole point of white space transmitters, though. As unlicensed transmitters, they are required to not interfere with licensed transmitters. The purpose of the regulations are to prevent those who are paying for exclusive rights to license spectrum (broadcasters) from being interfered with by freeloading unlicensed use of the spectrum.
As far as one unlicensed transmitter interfering with another unlicensed transmitter, the FCC doesn't care. CFR part 15 says that such unlicensed transmitters must be capable of working correctly in spite of any outside interference. They solve this through techniques like spread spectrum signaling, frequency hopping, etc. As it is right now, lots of Wi-Fi hot spots interfere with each other. Does that mean that Wi-Fi doesn't work? No, it works pretty well. Why? Because the technology was specifically designed to be robust against interference. Thus, the FCC doesn't need to care if there are multiple unlicensed devices on the same frequency. More to the point, the density of the population will likely require that they allow multiple white space services to share the same frequencies unless they're opening up a lot more spectrum than I think they are.
Being able to protect white space users from one another is the absolute farthest thing from anyone's minds, including the FCC and the hardware manufacturers. It is far easier to be robust against interference than to try to prevent it.
Wow. They're out of their frigging minds, then. I hope they're starting to build their server cluster now. If you think a site like CNN.com is high bandwidth, you ain't seen nothing yet.
Exactly. I made similar comments myself. Bear in mind that the database could potentially contain all the applications for future towers/transmitters as well, which would allow devices to black out those frequencies on the proposed start date (which is the absolute earliest the station could transmit even if everything got approved in a timely fashion). Even one download per month is likely way more than is actually needed. From submission of an application to approval can take up to a year. As long as the frequencies in every application are added to the database fairly quickly, even one download every six months is probably enough. One update per month is gravy.
That said, I think you underestimate the size of the required database a bit. It's not just TV and radio stations; there are at least 22,000 stations of those by themselves, but you also have cell frequency allocations, FCC-licensed fixed stations for police and fire dispatch, LPFM stations, licensed repeater stations for various things, licensed experimental stations, etc. I'm guessing probably more on the order of 40-50,000 stations. Plus hundreds more reserved frequencies like the ham bands, various government bands, military bands, etc., many of which vary by region. And that's for the U.S. You'd probably eventually need for this to be a worldwide database. So two or three hundred thousand stations, each needing to include GPS coordinates plus a base frequency and bandwidth with reasonable precision, plus enough information to describe the coverage footprint (which isn't just a range in miles if you do it right). I'd expect it to be several hundred megabytes when all is said and done. I mean, sure, they could be sloppy and do it in a few megs, but you'd get a lot less open spectrum that way.
The problem is not that we have too many laws, but rather that the laws are not neat. In some cases, you can say that a particular task is governed by a single set of laws, but this is rare. For the most part, a task is governed by multiple subparts of thousands of different laws, and it's utterly unmanageable trying to figure out if you are violating one or a dozen of them because of the poor organization, the lack of cross-referencing, and the general failure of the code to conform to any reasonable standards of structure.
The federal code is basically like a giant app that has been patched and tweaked to the point that it scarcely resembles the original code, has little real functioning organizational structure beyond (at best) the largest functional units (titles), contains multiple pieces of code that appear to do the same thing but generate different results, and program code that jumps haphazardly through a spaghetti-programming chain of GOTO statements into arbitrary sections of the code based on outside input from the judicial system.
What we need is to make the following changes:
A law that is not constantly being reviewed for effectiveness is a bad law, period.
Don't know where you get the notion that these devices are reporting their location. Quite the opposite except insofar as accessing the central database through some mechanisms could potentially reveal that there is a device of that type at that IP number, which to some very limited degree, gives a crude approximation of a location.
The proposed method of operation is as follows:
Nobody, including the FCC, cares whether these white space devices interfere with each other. They're inherently designed to be frequency hopping, spread spectrum devices precisely so that this isn't an issue. Thus, the devices have no need to register their location in the FCC's database, and indeed, doing so would be the exact opposite of what is intended, which is to preserve the frequency ranges used by licensed broadcasters.
I'm assuming that what you are afraid of is that people will build devices that query the database on the FCC's servers directly. I can't imagine that this would be the case, for several reasons:
In short, I think it's safe to say that any such devices will use a cached copy of the database and will probably update their copy fairly infrequently. Even one update per month would likely be more than sufficient to guarantee compliance.
No, it's actually very good news.
With the use of sensors, you have the "ABC problem". You have three collinear stations, A, B, and C. Station A is a licensed FCC broadcasting station. Station B is a receiver for that station. Station C is a frequency-hopping device that looks for an empty channel. Because station C is too far away, it cannot "hear" station A, but it is still close enough to station B to cause interference. Now granted, this is less likely when you're talking about multiple orders of magnitude difference in transmission power, but it is still possible, particularly when the transmitter might be inside a concrete structure with semi-directional leaks. This is a technologically unsolvable problem as long as you are depending on station C being able to somehow sense station A.
With geolocation, since all broadcast TV and radio stations are required by law to register with the FCC, including tower location, HAAT, a detailed map of estimated signal strengths based on topographical features, etc., you can come up with a much better idea of what frequencies are safe.
Yup. Truth is an absolute defense against a libel suit. She's pretty safe.
But just to pick nits... most geeks don't wear pants? So they walk around with their dangly bits hanging out? Maybe her coworkers are just perverts.... Did she ever think about that possibility? I don't know anybody who goes around wearing "nice pants" if that's what she means... at least not outside of marketing people. Wearing jeans puts you into the "normal" category. And although many geeks do spend every day in T-shirts, it isn't really a requirement unless you work at a Linux startup. Perhaps what she meant was that geeks tend not to be willing to go much out of their way as far as dressing up for a date (usually because they were still checking something into a CVS/subversion/git repo three minutes before they walked out the door).
Really, this is just a personality thing. A normal person looks at a list of tasks and prioritizes them, says "These won't get done," and does what he/she can. A geek says, "How can I get all of these things done in spite of myself." This tends to lead to much confusion when trying to understand the geek personality, but once you understand the root cause, it's much easier to find workarounds.
I'm a little puzzled by the choice of movies, too. Yes, I've seen Tron, but that's about as likely to come up in a conversation about movies as PA-RISC is to come up in a conversation about CPUs. Get with the times. It's all about Battlestar Galactica, Stargate, Eureka, the re-envisioned Star Trek franchise, Star Wars (in spite of Episode I), maybe Lost, maybe a few others.
And books like 1984, Fahrenheit 451, Dune, Enemy Mine, etc. That's how you get geek cred. I mean sure, most of us haven't read all of those books, but we know enough about them to carry on a conversation. Bonus points if you've read many of them.
Sounds like they're incompetent, then. :-)
I said more likely to, not likely to.
It's not too hard to believe.
First, Many large companies gets so much mail that the business comes with their own truck and picks it up from the USPS branch. Under those circumstances, as far as the USPS is concerned, the package or envelope is delivered the day it gets sorted and put into the pile to go to that company. However, this usually occurs after the company makes their pickup run for the day. In effect this means, the USPS almost invariably claims the package or envelope has been delivered a day before it really gets to the company. Whether this adds a day or not depends on what time of day the package or envelope got to the USPS office initially (whether it was before or after the company's pickup). So that's potentially one extra day.
Next, with the air mail, it probably flew out at the end of the day to NYC and arrived the next morning for processing. Truck drivers don't generally drive overnight when they can help it. That's two extra days.
Next, the truck has to break that trip up into two days worth of driving, assuming it's more than an 11 hour drive or that the truck driver didn't start in San Francisco. That's three extra days.
Next, because it is going by truck, it is probably indirect through Los Angeles, so it has to be sorted. That's four extra days.
Add one day lost in the back of and/or run over by a truck somewhere, and there you go. :-D
Longer hops (hops that cross oceans, more specifically) are more likely to be satellite hops, at which point latency goes south rather rapidly.
No need to get into things like acceleration. That would be inaccurate anyway. A better car analogy is capacity versus maximum speed. Consider a 50 passenger bus versus a customized tour bus. Both vehicles will get you there and back in an hour, but one carries 50 passengers and one carries half a dozen. The 50 passenger bus has high bandwidth, with one hour latency. The tour bus has medium bandwidth, with one hour latency. A Ferrari has low latency (it gets you there and back in 15 minutes, assuming you don't mind the tickets), but also doesn't have great bandwidth. A space shuttle has low latency (for a very limited set of destinations) and is moderately high bandwidth (at least for non-human cargo, if you use the cargo bay).
Fair point. However, the redirects are just the tip of the iceberg when you're talking about high latency connections. With modern web content, you have images that cause DNS requests followed by a request for the URL, scripts that cause DNS requests followed by a URL load request, and lots of other things going on that all cause suffering under high latency conditions. The web (and to some degree, the Internet as a whole) simply wasn't designed to degrade gracefully when round-trip packet latency goes much beyond about 100-200 milliseconds. Above that, throughput goes to hell in a handbasket rather quickly. At satellite latencies, it starts to be a problem. At multi-second mobile phone latencies, it is nearly unusable.
Honestly, I've concluded that the original, walled garden approach to mobile telecom was right in a perverse way. Sure, the walling was wrong, but the notion of (by default) using a proxy to make all of your web requests on a mobile phone is actually a great idea as long as it is not mandatory. In principal, a properly designed (non-transparent) proxy setup can cut the latency of web browsing almost in half by performing the DNS lookup on the proxy server instead of on the client, thus incurring the round-trip latency only once per request. That's a huge usability win when you're dealing with RTL measured in seconds. It also cuts the outgoing packet count roughly in half (most non-POST HTTP requests fit into 1-2 packets), which also improves latency for everybody else.
Simple. Same reason latency for packets varies, but in reverse. There's a huge volume of mail that moves from San Francisco to NYC, so they have air mail routes that optimize this. Because the distance from San Francisco to, for example, San Diego is relatively short and relatively low volume, they take it by truck, so it takes longer to get there, and probably stops in LA on the way (which is probably a latency disaster due to the amount of mail they process---going through LA can be like a corrupt BGP packet causing backbone traffic to be routed through your little home DSL router). With network traffic, it's the reverse; longer hauls are more likely to go to satellite, which spikes the latency way up. Shorter distances are more likely to be by cable, which has a lower latency. The point is that the haul mechanism determines the latency, and different links have different haul mechanisms depending on distance and expected load.
Just wanted to be clear about the sort of sites I'm talking about. I don't hate sites like YouTube or Flash games that use Flash because the alternative isn't really practical. I'm talking about sites like some car companies that think that their entire website should be nothing but one big pile of Flash. Such designs invariably translate into much longer page load times, clumsy interfaces with lots of unnecessary flashiness that gets in the way of usability, etc. If you're using Flash for what can and should reasonably be done with basic HTML elements, your site is a big, steaming pile of fail. More to the point, if I can't even navigate your site without loading Flash, you're doing it wrong.
BTW, it's <blockquote>...</blockquote>. You've been using BBCode too long.
Kind of what I was thinking, but not quite. I was suddenly hearing the song, "Attack of the Killer Tomatoes" in my head....
The point was that if you've taken the time to install that, you've made a conscious decision to avoid Flash most of the time, and many of those folks won't be too happy about having to load Flash just to get basic navigation on your site. It's an annoyance that tends to drive away such visitors. The reason I have it installed is that nearly every single browser crash I experienced had Flash somewhere in the backtrace. Now, with click2flash, my browser doesn't crash two or three times per day (and more to the point, hasn't crashed in months). If I get to a site that requires Flash, usually, I just close the window. I can't imagine I'm alone in this.
*shrugs* It came preinstalled in Mac OS X. I assumed it came preinstalled with most Windows OEMs as well. Either way, it's not something people think about downloading. It's just a couple of clicks in a plug-in dialog, which means that their market penetration is due largely to sites like YouTube that used to require it. People don't wake up one morning and think, "I really ought to have Flash installed on my computer." Give it a couple of years with HTML5-capable browsers and we'll see how many non-Farmville users still have Flash installed. My guess is that those numbers will drop significantly as fewer people hit pages that need it.
Unless you are using the WebKit viewer bits on iOS, though, the browser string will not be that of WebKit on iOS, and as such, those poorly designed apps won't typically even show up as being iOS-based devices in any browser stats. If you are seeing bogus network accesses, it's likely that you're going to see them on the desktop, too. With the exception of calling JavaScript functions, UIWebView doesn't really provide much opportunity to inflate statistics. There's no cache control, no injection of arbitrary JavaScript, and normally people would not construct a web view just to capture data, tear it down, and then display part of the data in a local-origin web view. That's what the various NSURL and CFURL bits are for, which again, don't send a user agent string unless you explicitly set one, AFAIK.
Pot, kettle, black.
And in a few decades, we'll be able to say, "We've always been at war with [insert country or group]." What, you didn't think 1984 had been averted, did you? It just was an idea whose time had not yet come.
I somehow misread the post as being Adobe's direct download stats. Download.com's stats are pretty uninteresting since most people who do download it do so in response to a browser dialog box and would presumably thus get it straight from Adobe. In that case, we don't know much more than we did before. *sigh*
Thanks for the numbers. There are about two billion computer users out there on the Internet. (Source: Internet World Stats) So barely one in a hundred computer users have downloaded Flash 10 deliberately. Although one percent is a huge percentage of computer users, it's still almost two orders of magnitude smaller than the installed base....
Clicking on ads is not necessarily a good indictor of an ad's impact. Indeed, most clicks on ads are accidental. The more important question is whether the ad produces mindshare. Fair point about zooming. On the other hand, users can ignore ads on a desktop by not looking at them, so I'm not sure how different it really is....
And the statistic is highly misleading anyway. Saying that 97% of computers can run Flash doesn't tell the whole story.
First, a lot of us use tools like click2flash that report themselves AS Flash, but are NOT Flash. Why do we do this? Because we got fed up with all the idiotic Flash-based adds that make buzzing sounds at random in background windows and make us jump straight out of our chairs. These people have Flash and put up with it when necessary, but generally avoid it. Those folks are difficult to distinguish from actual Flash "users", yet they suffer a degraded experience on Flash-heavy sites, and are less likely to come back.
Second, people have Flash largely because it came preinstalled. I don't know of anyone who has actually gone out of their way to install Flash. This means that those statistics could change on a dime.
Third, it assumes that all people use the web equally. For some sites, iOS-based devices (iPhone, iPad, iPod Touch) make up as much as 11% of their traffic by volume. When it comes to ad revenue, the ratings don't matter. The share matters. It doesn't matter if they make up only 1% of the total number of Internet-equipped devices. What matters is their percentage of the traffic.
Fourth, it ignores the assumption that people buying iPads and iPhones are more likely to have disposable income than people buying a random Windows PC. Thus, for many advertisers, one iPhone user is equivalent to several netbook users. Once you understand that, suddenly even a 1% share becomes much more significant, and a 10% share becomes a showstopper.
Books are easy. They're not particularly fragile, and people mostly buy them several at a time, even dozens at a time. That makes them very much not normal when it comes to online shopping. Most online shopping is for high ticket items like electronics. In that space, most people don't buy a whole bunch of things together unless they buy a bundle, and electronics need a lot more protection during shipping.
This isn't about costs. It's about pollution. Most of a business's costs have nothing to do with pollution. The cost of renting space, the cost of having more employees, etc. have minimal impact on pollution. (Yes, the employees use gas to get to work, but that's a drop in the bucket.)
Again, books are not the norm. There's a great deal more product differentiation in books than in most other product lines. When you can't get a particular brand of soap, you buy a different one. You probably don't drive to another store to get that brand, and you certainly don't drive very far. One book is not equivalent to another in the same way. This requires significant floor space for products that won't get purchased for years at a time. This greatly inflates the square footage needed for bookstores, and thus their environmental impact.
Books are also just about the only area where stores encourage customers to make special requests. This means that many bookstores get many of their goods in fairly small quantities. There is minimal difference between shipping two or three copies of a single book to a bookstore and shipping two or three different books from a warehouse directly to the customer.
So yes, books are an exception where online shopping actually might be better for the environment. I'd expect it to depend on the size of the store, their special order rate, and whether they sell used books. (The efficiency of selling used books locally would likely tip the environmental balance in favor of the local store when compared with online sales.) Again, though, books are an exception precisely because of their nature and the inherent difficulty in stocking sufficient breadth of products at a local store. Although you might see that same pattern in other highly specialized areas (e.g. professional video gear), it isn't typical of the online-versus-store debate. Most of those sorts of specialty shops have long since gone out of business outside of major metropolitan areas because there just isn't enough of a market for them to be viable. I'm almost surprised bookstores still exist for the same reason. Give it ten years, and this exception may be gone, too. :-)
Where to begin....
If a carrier is shipping individual packages across the country, I believe they usually put them on pallets for the long haul runs and fasten them down with cellophane or similar. Besides, even if the carrier does not use pallets, they're still not going to fill the truck all the way up. There are limits to how high you can stack things before the weight of the stack damages the packages at the bottom. In short, you have all the same problems when things are packed individually, plus all that wasted air space of the outer box that increases its volume anywhere from 20% (decent boxing) up to over a thousand percent (you'd be amazed) of the original volume.
The store is likely to be there, for the most part, whether you choose to buy the product there or not. As such, that's basically a sunk cost and shouldn't really contribute to this discussion. The only way it would matter would be if there were so much mail ordering of products that the stores went out of business. And even if the stores went out of business, there are enough people who need to get products quickly that a store's closing would drive farther to other stores that are open, which means more fuel for a lot of people. That's likely to far outweigh any savings in energy due to not having to heat or cool the building. And if the last store in an area closed, people would start ordering products with next-day delivery, which usually requires shipping by air, which is extremely inefficient. I really can't imagine that you'd break even....
Yes, but the impact is the opposite of what you think it will be. You're assuming that this means that if a delivery is delayed because of a truck not being full, a store runs out of something. In practice, if done correctly, because the entire shipment schedule is approximately known ahead of time, the delay of that non-full truck can be factored into the schedule, and the product can simply be put on an earlier truck. Thus, if a product's arrival is delayed because a truck isn't full, it is delayed because that product is not needed until after that truck is full. Otherwise, i would not have been on that truck, but rather, a previous truck. Alternatively, it might have taken a different route entirely.
The whole point of just-in-time deliveries is that you can not only reduce the products in warehouses, but also accurately predict your shipping needs to maximize efficiency. This level of planning simply cannot be matched by single one-off package delivery because a product cannot be shipped before you order it. As such, just-in-time deliveries actually widen the efficiency gap between local stores and mail order.
Fine, then only deliver packages on days when the person is getting junk mail anyway. That's at least three or four days a week.... :-)
But seriously, yeah, online shopping is much less efficient, if only because instead of transporting the goods crammed together with no extra packaging material, you're transporting them individually wrapped in extra protective packaging. It's the volume, not the weight, that makes individual shipping inefficient. It seems pretty obvious that internet shopping would be less efficient.
Also, instead of the product traveling from the manufacturer across the country to a store and then a few miles to your house, the product travels across country to an online store warehouse, then across the country again to your house. In the best case, the travel is comparable; in the worst case, the travel is doubled. Of course, much of this problem could be eliminated if manufacturers provided low-cost fulfillment services for their own products. Then an online store site could be merely a storefront, and could contact the manufacturer and have them ship the product directly to the customer. I've actually seen this done with some products (particularly products from Audio Technica), but this practice doesn't seem to be the norm. Most web stores use regional fulfillment warehouses, which may be closer to the retailers in terms of travel distance than if the web store had just one fulfillment warehouse, but it is still far less efficient than a full distribution network. Also, even if the manufacturer provided the fulfillment, it would still be less efficient unless that manufacturer shipped a whole pallet full of the product out at once---a delay that most online shoppers would not want to wait for.
What some folks don't realize about big box stores is that their stores (or trailers sitting behind their stores) can be used in part as temporary warehouse space for transporting goods to other stores. Imagine four stores, A, B, and C, in a line. You need to get a product from point A to point B and some of them to point C. You also have products that need to go from somewhere else (non-linear) to point C. So you run one truck from that second location to point B and drop off products. You then run a different truck from point A to point B, which drops off part of its load and picks up the additional goods before carrying everything to point D. Similarly, if there is extra space in the trucks, the central warehouses can preemptively ship extra products out that the store will eventually use. Those products sit in the truck until they are needed, and if another nearby store needs them, those products can be diverted locally instead of having to potentially run an extra truck from a central warehouse. In effect, this allows for much more distributed warehousing than is feasible with only a few warehouses scattered throughout the country.
Adding to the efficiency is the fact that most deliveries, fresh produce notwithstanding, are not time-dependent, unlike your package. If they get there a week later than expected, it's usually not a big deal; in the worst case, the store might be out of a product, but usually they keep ahead of things so that there is some wiggle room. What matters is minimizing the amount of distance that a product travels and filling every truckload as full as possible. The result is that delivery is highly efficient.
With individual package delivery, although some effort is made to do the same sorts of optimizations, the timeliness of the delivery limits the extent to which this is possible. Similarly, the inability to pre-distribute goods using otherwise wasted truck or plane space makes individual package delivery inherently less efficient.