The gameboy costs 16% of the PSP. IIRC games are half the price of PSP games too, especially if you pick them up used at EB or something. The PSP is also significantly larger than either the GBA or GBASP (not to mention the micro).
I've seen lots of solutions posted, the simplest probably being triangulation with a directional antenna.
Another solution is to combine a GPS unit (Or just a map of your office since you know where you are in it) with the detailed signal strength that apps like netstumbler can produce. As you walk around the office you're plotting signal strength points on a map. It would shortly become quite clear. Given enough points you don't even need to do any math or draw any lines. With very few points you can still work it out.
Ah, not so bad then. However you still need the tracker and web server, and you don't have the option of seeding a torrent trackerless. So you need at least 3 processes, which is admittedly better than one per torrent plus 2 extra.
I don't think that solves his problem though, since he would still need once instance per torrent file, unless I'm mistaken.
The answer is an application like Azureus. It handles everything itself. It serves up the torrent files with a built in HTTP server, it acts as the tracker for all torrents, and it acts as the client to seed each torrent. It allows very powerful rules about when to seed which file, how fast/how much, and so on.
So the net result is you can handle the entire hosting procedure with just one application/instance.
Where did you get the impression that UWB was shot down? The FCC approved UWB for public use. Yes, they limited it to a fraction of what we would have liked (A fraction of what still wouldn't have caused interference), but they DID approve it.
Current limitations on UWB allow WLAN (802.11g) type ranges, though the levels people were hoping for would have allowed immense ranges. A company called Pulse~Link is doing the UWB WLAN research (Cringley told me about them in an email conversation), and I assume that other companies are working along those lines too.
In fact, the FCC's limitations ONLY apply to wireless UWB. There are absolutely no limitations on what kinds of things you can do with UWB over wired solutions. It turns out that Ultra Wideband works great over wires. Pulse~Link is pushing it for cable TV providers. Apparently they can push more than 1gbit through a UWB stream over a cable TV line without affecting the existing television and cable internet services going over the cable lines. The UWB signals even survive the point at which the cable lines are aggregated into fiber optic cables. This could very well be the future of cable internet 5 or 10 years down the line.
I suggest you check out the Pulse~Link site (http://www.pulselink.net/) if you are still convinced that UWB got shot down. I don't have any connection to Pulse~Link, the only reason I'm mentioning them exclusively is that they're the only company with which I have read up on, and I trust Cringley.
It's not getting screwed because you're getting the same price as before. Or lower, since most registrars give you a discount for longer timeframes.
Regardless of if you register 10 times for 1 year or 1 time for 10 years, what does it matter? You may be worse off price-wise registerring 10 times.
Are you inferring that everything other than.net and.com are "third class" domain names? That'd be a mistake. For one thing, I merely suggested moving away from.net, not.com, so moving from.net to.com may be all you need to do. On the other hand the country code TLDs are generally considered to be just as good, at least in those countries. And of course there is always.org. I for one wouldn't mind a nice.ca address.
Everything supports it? Funny, most manufacturers don't seem to ship floppy drives with their machines anymore, barley any laptops have them (Laptops outsold desktops in the US recently), and many people leave them out of their home machines.
Of most of the people I know that have them, they only put in floppy drives because they had them lying around. A few who build new machines from scratch didn't get floppy drives and didn't pull them out of their old machines.
As far as I can tell, floppies are nearly dead, so it's a little bit dangerous to try to use them. USB keys are much safer these days since everybody has USB ports. Many more people have USB ports than floppy drives.
No, that's what NoMoreNicksLeft means. What I mean is, take fwdout, except instead of people donating their phone lines they donate money (or internet connectivity at the node location). That money is used to pay for internet connectivity and a business line with multiple outgoing lines in the target city.
In which case, as I mentioned, you are trading one limitation for another. Instead of being limited to in-network calls, you are limited to skype-to-skype calls. If you have limitations it isn't unlimited;)
The trick is still the connection between the phone and the PC. My cell phone has a data cable, but it's limited to treating the phone as a modem, or transfering files to/from. I don't know how one would make a voice connection over the low-speed cellphone USB cable that comes with the phones.
Perhaps one could take this skype cable this company makes and roll one's own software.
That's the thing though, you get donations to pay for a business line and you can make more than one connection to the POTS (Plain Old Telephone System). You only need one such node per city.
ipsec can be secure, sure, I guess it just depends on ease of configuration. I never looked much into ipsec, but it doesn't seem terribly easy to set up, compared to iptables. Especially compared to web frontends for iptables.
But then again, didn't look much into it, so for all I know Microsoft has some super easy to use tool that they don't include in XP, or that they hide.
There is something cool, though, about taking a $40 router and turning into a surprisingly powerful router/firewall:)
I'm not sure what they call the ones that work they way you describe, but the idea is you call a local number, get another dial tone, then call a long distance number.
On the other hand, imagine a free (opensource?) service that tracked all such numbers in all locations, and also when they were in use (Each "node" could report to the master server when it was in use). The only trick about this is getting the numbers to the people. As in, person A wants to make a call, and person B's node is ready, but how does person A get person B's number.
On a cellphone this is somewhat easy, since they are data enabled devices. A phone (or better, smartphone) could query the master server for a number. However doing it from home might be harder. You could always get a number through a computer, and then dial the node.
Another possible solution is to have a small number of "master" nodes in each calling area. This node would serve only to stay connected long enough to forward your call to an available node. It would only have to be connected for as long as it took to redirect the call and hang up.
On the other hand, this isn't terribly easier than making the call from the computer in the first place, and each call uses up two nodes. All we really need is one node, to get is into the POTS. So here is what I propose.
Opensource projects often survive on donations. Well, do the same thing, and use the donations to support one node in each major city. Start with a few of the largest cities (or calling areas, often multiple area codes are local calls.) In each (or one for starters) city, build a node. Nothing complicated, somebody gets a business line capable of making multiple outgoing calls. The cost of the node is the business phone line and an internet connection to support it.
I don't know how much bandwidth SIP/IAX uses, but I'm going to guess 24kbit per connection, since that's way more than enough for POTS quality voice. A residential DSL/cable line with 640kbit of usable bandwidth could support 26 simultaneous calls. Probably costs $40 per month canadian. I have no idea how much a business line costs.
So the trick to getting this to work is simply donations. Monetary donations can go towards the business line, and internet bandwidth can be donated directly or covered by cash donations. The more donations and popular your service, the larger and more cost effective the broadband line. A cogent line for $1000/mth could support 3500 calls, or a large number of users.
I have no idea if this could work. Just a crazy idea. Who knows. I think it's feasible. Just have to start small and grow.
Buying a $20 computer isn't exactly a recipe for reliability... And your solution is still more expensive than a WRT54G (or similar device) running OpenWRT. And for a $20 PC, I would almost think that the WRT54G would outperform the PC.
Before anybody complains that skype-to-skype calls are free, keep in mind that that isn't truely unlimited; that's the exact same restriction that the mobile phone companies put on their same-network unlimited plans, in that the person you are calling must be on skype as well.
There are two types of unlimited. Unlimited minutes to any local number, and unlimited minutes to ANY long distance or international number. Skype-to-skype isn't to anybody, only people with skype.
Don't get me wrong, this whole plan is genious, and it allows people to get skype's SkypeOut rates for their cellphones, and if the computer is hooked up to the POTS itself then free local.
He's quite a bit too late. Companies already offer truely unlimited (local calls) cell phone plans. Fido does this in Canada for $40/mth canadian ($30 or $35 US per month). Plus the usual hidden fees.
Of course, it's not free long distance, but neither is skype; skypeout might be cheap, but it isn't free.
They seem to be referring to software to put on existing servers. It would be hard to build a decent OpenBSD machine for under $100 US.
Of course if they DID want additional hardware, the absolute cheapest general-purpose linux box is the Linksys WRT54G. At least, it becomes a general purpose box as soon as you throw OpenWRT on it. Just set up the iptables rules however you like. You may want to disable the wireless functionality.
I've seen the WRT54G selling for as little as $50 CDN, which is probably about $40 US. It doesn't get much cheaper than that for a linux box.
Indeed, I first signed up for a passport when Windows XP first came out and I started using then "Windows Messenger". Even then, YEARS ago, I was able to use a yahoo address for my passport. So as far as I know, Passport has always supported arbitrary email addresses, or at least it has as long as anybody has cared about it.
Meh, point still stands. They are not in the same market.
0 05B8G1/qid=1121461204/sr=8-4/ref=pd_bbs_4/102-8200 465-4309714?v=glance&s=videogames&n=507846
Gameboy on Amazon (used):
http://www.amazon.com/exec/obidos/tg/detail/-/B00
The gameboy costs 16% of the PSP. IIRC games are half the price of PSP games too, especially if you pick them up used at EB or something. The PSP is also significantly larger than either the GBA or GBASP (not to mention the micro).
GBA: $50 US on ebay if you're lucky
PSP: $299 new, still fraking expensive on eBay
The GBA isn't going anywhere. Anybody who considers the PSP and GBA direct competitors are idiots.
I've seen lots of solutions posted, the simplest probably being triangulation with a directional antenna.
Another solution is to combine a GPS unit (Or just a map of your office since you know where you are in it) with the detailed signal strength that apps like netstumbler can produce. As you walk around the office you're plotting signal strength points on a map. It would shortly become quite clear. Given enough points you don't even need to do any math or draw any lines. With very few points you can still work it out.
Ah, not so bad then. However you still need the tracker and web server, and you don't have the option of seeding a torrent trackerless. So you need at least 3 processes, which is admittedly better than one per torrent plus 2 extra.
I don't think that solves his problem though, since he would still need once instance per torrent file, unless I'm mistaken.
The answer is an application like Azureus. It handles everything itself. It serves up the torrent files with a built in HTTP server, it acts as the tracker for all torrents, and it acts as the client to seed each torrent. It allows very powerful rules about when to seed which file, how fast/how much, and so on.
So the net result is you can handle the entire hosting procedure with just one application/instance.
More info at http://azureus.sourceforge.net/
Where did you get the impression that UWB was shot down? The FCC approved UWB for public use. Yes, they limited it to a fraction of what we would have liked (A fraction of what still wouldn't have caused interference), but they DID approve it.
Current limitations on UWB allow WLAN (802.11g) type ranges, though the levels people were hoping for would have allowed immense ranges. A company called Pulse~Link is doing the UWB WLAN research (Cringley told me about them in an email conversation), and I assume that other companies are working along those lines too.
In fact, the FCC's limitations ONLY apply to wireless UWB. There are absolutely no limitations on what kinds of things you can do with UWB over wired solutions. It turns out that Ultra Wideband works great over wires. Pulse~Link is pushing it for cable TV providers. Apparently they can push more than 1gbit through a UWB stream over a cable TV line without affecting the existing television and cable internet services going over the cable lines. The UWB signals even survive the point at which the cable lines are aggregated into fiber optic cables. This could very well be the future of cable internet 5 or 10 years down the line.
I suggest you check out the Pulse~Link site (http://www.pulselink.net/) if you are still convinced that UWB got shot down. I don't have any connection to Pulse~Link, the only reason I'm mentioning them exclusively is that they're the only company with which I have read up on, and I trust Cringley.
It's not getting screwed because you're getting the same price as before. Or lower, since most registrars give you a discount for longer timeframes.
.net and .com are "third class" domain names? That'd be a mistake. For one thing, I merely suggested moving away from .net, not .com, so moving from .net to .com may be all you need to do. On the other hand the country code TLDs are generally considered to be just as good, at least in those countries. And of course there is always .org. I for one wouldn't mind a nice .ca address.
Regardless of if you register 10 times for 1 year or 1 time for 10 years, what does it matter? You may be worse off price-wise registerring 10 times.
Are you inferring that everything other than
Not entirely screwed; VeriSlime must give 6 months notice, during which time you can make long many-year registrations.
If in several years time the cost of renewing the domains is too high, you can move to alternate TLDs, because you have enough time to do so.
No, they outfitted the canadarm with supersensative 3D mapping devices.
Real smart, when charged with a crime flee to countries the US has extradition treaties with. Let me know how that goes when you try it.
Everything supports it? Funny, most manufacturers don't seem to ship floppy drives with their machines anymore, barley any laptops have them (Laptops outsold desktops in the US recently), and many people leave them out of their home machines.
Of most of the people I know that have them, they only put in floppy drives because they had them lying around. A few who build new machines from scratch didn't get floppy drives and didn't pull them out of their old machines.
As far as I can tell, floppies are nearly dead, so it's a little bit dangerous to try to use them. USB keys are much safer these days since everybody has USB ports. Many more people have USB ports than floppy drives.
No, that's what NoMoreNicksLeft means. What I mean is, take fwdout, except instead of people donating their phone lines they donate money (or internet connectivity at the node location). That money is used to pay for internet connectivity and a business line with multiple outgoing lines in the target city.
See, my idea is to centralize it (per city).
In which case, as I mentioned, you are trading one limitation for another. Instead of being limited to in-network calls, you are limited to skype-to-skype calls. If you have limitations it isn't unlimited ;)
The trick is still the connection between the phone and the PC. My cell phone has a data cable, but it's limited to treating the phone as a modem, or transfering files to/from. I don't know how one would make a voice connection over the low-speed cellphone USB cable that comes with the phones.
Perhaps one could take this skype cable this company makes and roll one's own software.
That's the thing though, you get donations to pay for a business line and you can make more than one connection to the POTS (Plain Old Telephone System). You only need one such node per city.
I admit SkypeOut's quality isn't great, but the cost pretty nice. Still more expensive than local calls, but really cheap for long distance.
Can you provide more information about these termination providers?
ipsec can be secure, sure, I guess it just depends on ease of configuration. I never looked much into ipsec, but it doesn't seem terribly easy to set up, compared to iptables. Especially compared to web frontends for iptables.
:)
But then again, didn't look much into it, so for all I know Microsoft has some super easy to use tool that they don't include in XP, or that they hide.
There is something cool, though, about taking a $40 router and turning into a surprisingly powerful router/firewall
You just described a VOIP provider ;)
I'm not sure what they call the ones that work they way you describe, but the idea is you call a local number, get another dial tone, then call a long distance number.
On the other hand, imagine a free (opensource?) service that tracked all such numbers in all locations, and also when they were in use (Each "node" could report to the master server when it was in use). The only trick about this is getting the numbers to the people. As in, person A wants to make a call, and person B's node is ready, but how does person A get person B's number.
On a cellphone this is somewhat easy, since they are data enabled devices. A phone (or better, smartphone) could query the master server for a number. However doing it from home might be harder. You could always get a number through a computer, and then dial the node.
Another possible solution is to have a small number of "master" nodes in each calling area. This node would serve only to stay connected long enough to forward your call to an available node. It would only have to be connected for as long as it took to redirect the call and hang up.
On the other hand, this isn't terribly easier than making the call from the computer in the first place, and each call uses up two nodes. All we really need is one node, to get is into the POTS. So here is what I propose.
Opensource projects often survive on donations. Well, do the same thing, and use the donations to support one node in each major city. Start with a few of the largest cities (or calling areas, often multiple area codes are local calls.) In each (or one for starters) city, build a node. Nothing complicated, somebody gets a business line capable of making multiple outgoing calls. The cost of the node is the business phone line and an internet connection to support it.
I don't know how much bandwidth SIP/IAX uses, but I'm going to guess 24kbit per connection, since that's way more than enough for POTS quality voice. A residential DSL/cable line with 640kbit of usable bandwidth could support 26 simultaneous calls. Probably costs $40 per month canadian. I have no idea how much a business line costs.
So the trick to getting this to work is simply donations. Monetary donations can go towards the business line, and internet bandwidth can be donated directly or covered by cash donations. The more donations and popular your service, the larger and more cost effective the broadband line. A cogent line for $1000/mth could support 3500 calls, or a large number of users.
I have no idea if this could work. Just a crazy idea. Who knows. I think it's feasible. Just have to start small and grow.
Sorry for the blatant plug, but you should be looking into an autonomic solution:
http://nitix.com/
WebHostingTalk hasn't been owned by EV1 for a long time. It isn't even hosted there anymore, it's hosted at RackSpace now.
Buying a $20 computer isn't exactly a recipe for reliability... And your solution is still more expensive than a WRT54G (or similar device) running OpenWRT. And for a $20 PC, I would almost think that the WRT54G would outperform the PC.
Before anybody complains that skype-to-skype calls are free, keep in mind that that isn't truely unlimited; that's the exact same restriction that the mobile phone companies put on their same-network unlimited plans, in that the person you are calling must be on skype as well.
There are two types of unlimited. Unlimited minutes to any local number, and unlimited minutes to ANY long distance or international number. Skype-to-skype isn't to anybody, only people with skype.
Don't get me wrong, this whole plan is genious, and it allows people to get skype's SkypeOut rates for their cellphones, and if the computer is hooked up to the POTS itself then free local.
He's quite a bit too late. Companies already offer truely unlimited (local calls) cell phone plans. Fido does this in Canada for $40/mth canadian ($30 or $35 US per month). Plus the usual hidden fees.
Of course, it's not free long distance, but neither is skype; skypeout might be cheap, but it isn't free.
They seem to be referring to software to put on existing servers. It would be hard to build a decent OpenBSD machine for under $100 US.
Of course if they DID want additional hardware, the absolute cheapest general-purpose linux box is the Linksys WRT54G. At least, it becomes a general purpose box as soon as you throw OpenWRT on it. Just set up the iptables rules however you like. You may want to disable the wireless functionality.
I've seen the WRT54G selling for as little as $50 CDN, which is probably about $40 US. It doesn't get much cheaper than that for a linux box.
Still, I think he meant more software-wise.
Indeed, I first signed up for a passport when Windows XP first came out and I started using then "Windows Messenger". Even then, YEARS ago, I was able to use a yahoo address for my passport. So as far as I know, Passport has always supported arbitrary email addresses, or at least it has as long as anybody has cared about it.